l-OA United States
Environmental Protection
Agency
PB2006-102459
EPA/600/R-05/092
Relationships Between Questionnaire
Responses and Children's Pesticide
Exposure Measurements
RESEARCH AND DEVELOPMENT
-------
-------
EPA/600/R-05/092
August 2005
Relationships Between Questionnaire Responses and
Children's Pesticide Exposure Measurements
by
Carol B. Thompson1, Lynda S. Harrison1
Richard A. Fenske, Ph.D., MPH,2
Gary L. Robertson (Retired)3, Stephen C. Hern (Retired)3
1 Anteon Corporation
Las Vegas, NV 89119
2 Department of Environmental and Occupational Health Sciences
University of Washington
Seattle, WA 98195
3National Exposure Research Laboratory
Las Vegas NV 89193
Work Assignment Manager: Kim R. Rogers
National Exposure Research Laboratory
Las Vegas, NV 89193-3478
GS A Contract No. GS-09K-99-BHD-0001
Task No. 9T1Z004TMA
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Executive Summary
Children are widely acknowledged to be more vulnerable than adults to many environmental
health hazards, including pesticides, because they are more exposed, and because they may
have elevated susceptibility. The effect of relatively low levels of pesticide exposure in
children is an area of great scientific uncertainty and has, therefore, become the focus of
substantial research and regulatory activities. The work of the Pesticides in Young Children
Border States Program to identify the major exposure risk factors for intervention requires
studying children across the exposure measurement distribution, especially those with higher
exposure measurements. Questions that could be used for exposure classification as a
prescreening tool would help produce either an enriched population (i.e., a larger percentage
of individuals with higher exposure levels) or would eliminate individuals with lower
exposure levels from further review. Considerable savings in time and money may then be
realized in selecting the desired population for a study.
The objective of this project is to identify questions that indicate a higher likelihood of
predicting a child's level of exposure to pesticides as input to future study designs. This
report reviews the state of the science in relating questionnaire responses to environmental
and biological measurements, primarily for children, based on results and data from previous
exposure studies. A two-part approach was used for this evaluation:
• A literature review of previous exposure studies to summarize the existence of such
quantitative and qualitative relationships, and
• An analysis of Phase II of the Pesticide Exposure and Health Effects on Children
Initiative (Yuma Study), which contained questionnaires and measurements.
The literature review of previous exposure studies identified 20 publications that met the
criteria set for this evaluation. These publications were reviewed in detail to determine the
relationships that were considered and statistically analyzed in each study. Relationship, as
used here, is defined as a systematic correspondence between the values of two variables,
that is, questionnaire responses and analytical measurements. Detailed information about
each relationship was compiled to allow for more in-depth evaluations by other researchers
as their interests dictated. The questions were grouped into categories for evaluation, and
questions showing overall significance across the publications were identified.
From the 20 relevant publications, 603 statistically significant and non-significant
relationships across 117 questions in 14 question categories were identified. Eighty-six
percent of the relationships were in the categories of residential pesticide use, household
characteristics, household occupation, residential proximity to agricultural fields, subject's
personal characteristics, and family hygienic practices. These six categories represent
questions whose relationships with exposure measurements have both a strong theoretical
basis and the quality of being reasonably evaluated through the study designs. Sixty-six
percent of the relationships considered metabolites in urine measurements, primarily with
DAP-based metabolites, and 31 percent of the relationships considered dust measurements.
Three risk factors related to the take-home or para-occupational exposure pathways were
11
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
analyzed as separate question categories: household occupation, family hygiene practices,
and work exposure/practices.
The relationships for each question and chemical/metabolite combination were reviewed to
determine the question's effectiveness for differentiating exposure levels. Generally the
questions showing the most effectiveness were:
• residential pesticide use (inside and outside)
• occupation of household members
• child's characteristics (age, ethnicity, family income)
• family hygiene practices.
Several other questions, which were tested less extensively in the publications, also showed
some effectiveness:
• pets
• household location: urban vs non-urban
• dietary behaviors (organic food)
• exposure levels of household members
• health status (diseases)
• smoking behaviors
• proximity to agricultural fields (for house dust only).
The Yuma Study was conducted from October 1999 through February 2000 for 152
households of permanent residents of the area with children in kindergarten or first grade.
The children's urine samples were measured for the six most common dialkylphosphates
(DAPs) associated with OP pesticides. Dust samples collected from each household and
from classrooms were measured for specific organophosphorous (OP), organochlorine,
pyrethroid, and carbamate pesticides. One set of statistical analyses considered the urine,
dust, and questionnaire data to identify any associations between questionnaire data and
pesticide exposure levels. Traditional statistical techniques were performed to test
predefined hypotheses on the principal participant children and their siblings between 2 and
11 years of age. Of the questions analyzed, recent indoor pesticide use, household members
working in agriculture, and distance from home to agricultural fields have statistically
significant relationships with the ethylated DAP sum and with individual ethyl and methyl
DAPs; however, the direction of some relationships was opposite of what might be expected.
A second set of statistical analyses was performed only on principal participant children from
the eight schools and two grades in the initial study design. A data mining approach,
Classification and Regression Trees (CART), was used to identify potential predictors
without specifying a priori hypotheses between the biomarker measurements and the
questions and dust measurements. Six scenarios with increasing levels of measurement
burden from questions, household dust, and school dust were analyzed for the ethylated and
methylated DAPs sums.
in
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Although there are differences in the subpopulations considered and the statistical analyses
performed, a summary of the questions selected under either approach for both DAP sums
provides a view of the effective differentiators from the Yuma Study. The best use of these
results is as "indicators" of predictors that are more useful in differentiating the exposure
levels.
Table ES-1 Comparison of Questions Selected from the Yuma Study for the Sum of Methylated and
Ethylated DAPs Based on Two Analysis Approaches
Sum of Methylated DAPs
Sum of Ethylated DAPs
Recent use of pesticides inside home
Child's characteristics (height, weight)
Child's characteristics (weight, ethnicity)
Other adult in household working in agriculture
Proximity to agricultural fields, spraying conditions
Proximity to agricultural fields, spraying conditions
Child's time spent away from home
Where in house child spends time
Child's school
Father's occupation
Diet - local fruits/vegetables
Questions that seem to have stronger relationships with the exposure levels across both the
literature review and the Yuma Study include the following:
• occupation of adults living in household
• residential pesticide use
• residential proximity to spraying and agricultural fields
• characteristics of the subject that may indicate potential exposure activities
• family hygiene practices that may mitigate the take-home pathway exposure
• where the child spends time (in home, away from home)
• diet with respect to locally-grown fresh fruits and vegetables.
This report reviews the state of the science in relating questionnaire responses to
environmental and biological measurements primarily for children. Future studies that use
biological monitoring and questionnaires should draw upon this and other recent research to
refine study protocols with the following recommendations.
Based on the literature review, 41 questions were identified as effective in differentiating
exposure levels of at least one chemical/metabolite in urine and dust measurements. These
questions are offered as a resource of recommended questions with specific chemicals or
metabolites for future study designs. Note that the questions were evaluated here as a
screening tool to create an enriched population of participants with higher exposure levels.
Thus, their future use is better suited to similar purposes.
iv
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table ES-2 Questions Considered Effective Differentiators of Children's Pesticide Exposure Levels
(Extracted from Table 2.11)
Medium
Q Category
Q Description
Urine
Dust
Residential Pesticide
Use
X
Were the "bedrooms" in the house treated with
pesticides?
X
Were the "closets" in the house treated with pesticides?
X
Was the "dining room" in the house treated with
pesticides?
X
Was the "living room" in the house treated with
pesticides?
X
Was an "other room" in the house treated with pesticides?
X
X
Was the outside of the house treated with pesticides?
X
Was the garden treated with pesticides?
X
Was the lawn or yard treated with pesticides?
X
Level of household pesticide use
X
Number of times personally applied pesticides inside the
house
X
Number of times personally applied pesticides outside
the house
X
Was the inside or outside of the house treated with
pesticides by a family member?
X
Did you personally mix pesticide inside the house?
Household
characteristics
X
Is the property used as a farm?
X
Number of persons living in household
X
Do you have pets in the house?
X
Do you have pets inside or outside the house?
X
Does household have a garden or vegetable garden?
Household occupation
X
Number of agricultural workers in household
X
Applicator vs farm worker
X
Applicator vs non-applicator
X
X
Applicator and farm worker vs reference
X
Applicator vs reference
X
Fieldworker vs pesticide handler
X
Did head of household spray fields?
X
Was a household member recently involved in fieldwork?
v
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Medium
Q Category
Q Description
Urine
Dust
X
Are household members involved in tree thinning?
X
Number of household members with high pesticide
contact jobs
Residential proximity to
agricultural fields
X
X
Proximity of home to pesticide-treated farmland/orchard
Residential location
X
Urban vs non-urban
Subject's personal
characteristics
X
Age
X
Ethnicity
X
Income
Child's behaviors
X
Hand wipe concentration per unit area
Dietary behaviors
X
Was diet conventional or organic?
Family hygiene
practices
.X
Are work clothes worn inside the house?
X
Number of weeks since last house was last vacuumed
Related exposure levels
X
Number of adult household members with high metabolite
levels
Health
X
Have you ever had bowel disease?
X
Have you ever had intestinal disease?
X
Have you ever had ulcers?
Analyses of the association between questionnaire data and pesticide metabolites in
children's urine are conducted on the assumption that the urinary metabolite measurements
provide an accurate estimate of children's exposure. Metabolites under study are processed
and excreted relatively quickly in humans (1-3 days), which is in contrast to the general
nature, in terms of the time frame of a particular activity or behavior, of most questions asked
of parents or children. It is therefore worthwhile to consider the variability in measurements
in urinary pesticide metabolites. Recent studies suggest that if complete 24- or 48-hour urine
samples are collected rather than spot urine samples, it may be possible to better identify
major risk factors for exposure.
vi
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Validity of questionnaire data is an essential consideration in epidemiologic studies, and
future studies of children's pesticide exposure should be preceded by validation studies.
Such studies might include validating the basis for classifying each applicator's exposure
through biological monitoring, or evaluating the correlations between self-reported
behavioral data from potential participants and the urinary metabolite data.
Studies of children's pesticide exposure should work to improve the quality of data related to
behavior. At present, researchers rely primarily on parental reports of behavior for young
children. Yet the validity of parental reports has not been scrutinized in a systematic fashion.
There is clearly a need for more objective measures of children's activities and behaviors in
conjunction with systematic biological monitoring to ensure identification of key predictors
of children's exposure to pesticides.
vii
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Disclaimer
Notice: This document is a preliminary draft. It has not been formally reviewed or released
by the U.S. Environmental Protection Agency and should not at this stage be construed to
represent Agency policy. This manuscript is being circulated for comments on its technical
merit and potential for policy implications. Do not cite or quote. Mention of trade names or
commercial products does not constitute endorsement or recommendation by EPA for use.
viii
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Abstract
Children are deemed to be more vulnerable than adults to many environmental health
hazards, including pesticides, and the effect of relatively low levels of pesticide exposure in
children has become the focus of substantial research and regulatory activities. To identify
the major exposure risk factors for intervention, the Pesticides in Young Children Border
States Program requires studying children across the exposure measurement distribution,
especially those with higher exposure measurements. Questions that could be used for
exposure classification as a pre-screening tool may prove to be a cost-effective way to select
the desired population for a study. A two-part approach was implemented to identify
questions that indicate a higher likelihood of predicting a child's level of exposure to
pesticides as input to future study designs:
• A literature review of previous exposure studies to evaluate questions used, and
• An analysis of a recent children's pesticide exposure study.
From the 20 relevant exposure study publications, 603 relationships, statistically significant
and not, across 117 questions in 14 question categories were identified. The relationships for
each question and chemical/metabolite combination were reviewed to determine the
question's effectiveness for differentiating exposure levels. Generally the questions showing
the most effectiveness were: residential pesticide use (inside and outside), occupation of
household members, child's characteristics (age, ethnicity, income, and family hygienic
practices. Several other questions, which were used less extensively in the studies, also
showed some effectiveness.
Data from a recent study of children's exposure to pesticides conducted in Yuma, Arizona,
was analyzed from two perspectives: traditional statistical analyses on predefined hypotheses
of potential risk factors, and a data mining approach to explore the relationships existing in
the data. Both analyses evaluated the relationships between the dialkylphosphate
(organophosphate pesticide) metabolite levels in the children's urine samples, the pesticide
levels in the household and school dust samples, and the questionnaire responses.
Questions that seem to have stronger relationships with the exposure levels across both the
literature review and the analysis of the exposure study include the following: occupation of
adults living in household, residential pesticide use, residential proximity to spraying and
agricultural fields, characteristics of the subject that may indicate potential exposure
activities, family hygienic practices that may mitigate the take home pathway exposure,
where the child spends time (in home, away from home), and diet with respect to locally-
grown fresh fruits and vegetables.
This report reviews the state of the science in relating questionnaire responses to
environmental and biological measurements, primarily for children. Future studies that use
biological monitoring and questionnaires should draw upon this and other recent research to
refine study protocols with the recommendations noted.
IX
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Acknowledgements
No large project like this report is accomplished by the co-authors alone, and we would like
to gratefully acknowledge those who helped in this process. Andy Tsang, Anteon
Corporation, helped with some of the statistical analyses and facilitated the transition of the
relationship database entries for the tables in Appendices B, C, and D. Dr. Peter Stephan,
Anteon Corporation, was responsible for the technical preparation of the document, which
included valuable comments for making the information-intensive tables more readable.
Guadalupe Chapa, Fellow of Associated Schools of Public Health, provided comments and
clarifications. Thank you.
None of this work would be possible without the continued commitment of the researchers
focused on children's exposure to pesticides, both through the publications we reviewed and
their willingness to answer questions related to their work. There are many other researchers
in this area whose work, though not directly relevant to this report, is part of the set of
building blocks upon which this area of research depends. We hope that those engaged in
this field find the results of this work useful in their future research efforts. Finally, we
gratefully acknowledge all of the families who have participated in these studies; without
their cooperation we would not be able to progress in our understanding of these important
issues.
x
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Contents
Executive Summary
Disclaimer
Abstract
Acknowledgements
1.0 INTRODUCTION 1-1
1.1 Introduction 1-1
1.2 Children's Exposure to Pesticides 1-1
1.3 Populations of Concern 1-2
1.4 Exposure Pathways and Patterns 1-2
1.5 Questionnaire Data 1-3
1.6 Exposure Measurements 1-4
1.7 Border States Program: Pesticides in Young Children 1-5
1.8 Motivation and Goal of the Project 1-5
1.9 General Approach and Report Contents 1-6
2.0 SUMMARY AND RECOMMENDATIONS 2-1
2.1 Introduction 2-1
2.2 Methods 2-1
2.3 Results 2-3
2.3.1 Literature Review 2-3
2.3.2 Children's Pesticide Exposure Study (Yuma Study) 2-6
2.4 Summary of Results from Two Approaches 2-11
2.5 Recommendations 2-14
2.5.1 Effective Differentiators of Exposure Level 2-14
2.5.2 Urinary Metabolite Monitoring 2-18
2.5.3 Questionnaire Validation 2-19
2.5.4 Objective Measures of Children's Behaviors 2-19
3.0 METHODOLOGY 3-1
3.1 General Description of Approach 3-1
3.2 Literature Review Methods 3-1
3.2.1 Sources 3-1
3.2.2 A Database of Relationships 3-3
3.3 Children's Pesticide Exposure Study 3-4
3.3.1 B ackground 3-4
3.3.2 Stage 1 - Data Preparation 3-5
3.3.3 Stage 2 - Review of Basic Relationships 3-10
3.3.4 Stage 3 - Classification Approach 3-10
xi August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
4.0 RESULTS AND DISCUSSION 4-1
4.1 Introduction 4-1
4.2 Literature Review 4-1
4.2.1 Publications Reviewed for Relationships 4-2
4.2.2 Description of Relationship Information 4-11
4.2.2.1 Content 4-11
4.2.2.2 Organization 4-12
4.2.2.3 Assumptions and Caveats 4-16
4.2.3 Description of Relationships Presented 4-17
4.2.4 Presentation of Source Relationships 4-19
4.2.4.1 Category 1: Residential Pesticide Use 4-19
4.2.4.2 Category 2: Household Characteristics 4-22
4.2.4.3 Category 3: Residential Sources (Environmental Measures) . 4-24
4.2.4.4 Category 4: Household Occupation 4-26
4.2.4.5 Category 5: Residential Proximity to Agricultural Fields 4-29
4.2.4.6 Category 6: Residential Location 4-31
4.2.4.7 Summary of Results from Source Relationships 4-33
4.2.5 Presentation of Behavior Relationships 4-36
4.2.5.1 Category 7: Subject's Personal Characteristics 4-36
4.2.5.2 Category 8: Child's Behaviors 4-38
4.2.5.3 Category 9: Dietary Behaviors 4-39
4.2.5.4 Category 10: Family Hygiene Practices 4-41
4.2.5.5 Category 11: Smoking-Related Activities 4-43
4.2.5.6 Category 12: Work Exposure/Practices 4-44
4.2.5.7 Summary of Results from Behavior Relationships 4-45
4.2.6 Presentation of Other Relationships 4-46
4.2.6.1 Category 13: Related Exposure Levels 4-47
4.2.6.2 Category 14: Health 4-48
4.2.6.3 Summary of Results from Other Relationships 4-49
4.2.7 Summary of Results from Literature Review 4-50
4.3 Children's Pesticide Exposure Study (Yuma Study) 4-52
4.3.1 Relationships Explored in the Yuma Study Report 4-55
4.3.1.1 Relationships Between Questions and DAP Metabolites 4-55
4.3.1.2 Relationships Between Dust Measurements and DAP
Metabolites 4-59
4.3.1.3 Summary of Results 4-65
4.3.2 Results from the Data Mining Approach 4-67
4.3.2.1 Subpopulation Selected for Analysis 4-67
4.3.2.2 Preliminary Analyses 4-68
4.3.2.3 Analysis for Underlying Structure 4-68
4.3.2.4 Classification Analyses 4-71
4.3.2.5 Comparison of Questionnaire Responses for High and Low
Ends of Measurements 4-73
4.3.2.6 Summary of Results 4-74
4.4 Effective Predictors of Pesticide Exposure Levels 4-76
5.0 REFERENCES 5-1
xii August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendix A References from the Literature Review
Appendix B Overview Tables for Relationships from the Literature Review
Appendix C Detail Tables for Relationships from the Literature Review
Appendix D Comment Tables for Relationships from the Literature Review
Appendix E Questions Tracked in the Literature Review
Appendix F Definition of Chemical Measurement Variables Used in the Analysis of the
Yuma Study
Appendix G Data Mining Methodology and Results for the Yuma Study
Xlll
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Tables
Table 2.1 Distribution of Questions and Relationships Across 14 Question
Categories 2-3
Table 2.2 Distribution of Questions and Relationships Across Medium
Measured 2-4
Table 2.3 Questions and Dust Measurements from the Yuma Study Having
Strong Relationships with the DAPs, Adjusted for Creatinine
(CDC 2002) 2-7
Table 2.4 Questions and Measurements Included in the Yuma Study CART
Analysis Scenarios for Each DAP Sum 2-8
Table 2.5 Questions and Dust Measurements from the Yuma Study Having
Strong Relationships with the Sum of Methylated DAPs Based on
the Data Mining Approach 2-9
Table 2.6 Questions and Dust Measurements from the Yuma Study Having
Strong Relationships with the Sum of Ethylated DAPs Based on
the Data Mining Approach (Excluding CHLDTM3) 2-10
Table 2.7 Comparison of Selected Predictors from Yuma Study Report
(CDC 2002) and Data Mining Approach for Methylated Sum of
DAPs 2-11
Table 2.8 Comparison of Selected Predictors from Yuma Study Report
(CDC 2002) and Data Mining Approach for Ethylated Sum of
DAPs 2-12
Table 2.9 Summary of Predictors Selected as Useful in Differentiating
Children's Pesticide Exposure Levels Across Two Approaches 2-13
Table 2.10 Description of Code Names and Groups Assigned to the
Chemicals and Metabolites, Sorted by Code 2-14
Table 2.11 Questions Considered Effective Differentiators of Children's
Pesticide Exposure Levels Based on a Literature Review of
Previous Exposure Studies 2-15
Table 3.1 Questionnaire Variables from the Yuma Study Used in Data
Mining Analyses, Sorted by the Questionnaire Order 3-6
Table 3.2 Analytical Measurement Variables from the Yuma Study Used in
Data Mining Analyses 3-9
Table 4.2.1 Brief Descriptions of the Studies Included in the 20 Relevant
Publications 4-3
Table 4.2.2 Information Extracted from Relevant Publications for Each
Relationship 4-11
Table 4.2.3 Distribution of Relationships Across Risk Factor Groups and
Question Categories of Questions Used to Organize Sections
4.2.4, 4.2.5, 4.2.6, Appendix B, Appendix C, and Appendix D 4-13
Table 4.2.4 Distribution of Relationships across Question Categories and
Mediums Measured 4-15
Table 4.2.5 Cross-Reference for Relationship Tables by Question Category
Group 4-18
xiv
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.6.a Codes and Descriptions for Chemicals/Metabolites with
Significant Relationships for Questions in the Residential
Pesticide Use Category 4-19
Table 4.2.6.b Distribution of Significant Medium/Question Relationships for
Residential Pesticide Use Questions, by Medium 4-20
Table 4.2.6.C Residential Pesticide Use Questions and Chemicals/Metabolites
with Overall Significant Relationships 4-21
Table 4.2.7.a Codes and Descriptions for Chemicals/Metabolites with
Significant Relationships for Questions in the Household
Characteristics Category 4-22
Table 4.2.7.b Distribution of Significant Medium/Question Relationships with
Household Characteristics Questions, by Medium 4-23
Table 4.2.7.C Household Characteristics Questions and Chemicals/Metabolites
with Overall Significant Relationships 4-24
Table 4.2.8.a Codes and Descriptions for Chemicals/Metabolites with
Significant Relationships for Questions in the Residential Sources
Category 4-25
Table 4.2.8.b Distribution of Significant Medium/Question Relationships with
Residential Sources Questions, by Medium 4-25
Table 4.2.8.C Residential Sources Questions and Chemicals/Metabolites with
Overall Significant Relationships 4-26
Table 4.2.9.a Codes and Descriptions for Chemicals/Metabolites with
Significant Relationships for Questions in the Household
Occupation Category 4-27
Table 4.2.9.b Distribution of Significant Medium/Question Relationships with
Household Occupation Questions, by Medium 4-28
Table 4.2.9.C Household Occupation Questions and Chemicals/Metabolites with
Overall Significant Relationships 4-29
Table 4.2.10.a Codes and Descriptions for Chemicals/Metabolites with
Significant Relationships for Questions in the Residential
Proximity to Agricultural Fields Category 4-30
Table 4.2.10.b Distribution of Significant Medium/Question Relationships with
Residential Proximity to Agricultural Fields Questions, by
Medium 4-30
Table 4.2.10.c Residential Proximity to Agricultural Fields Questions and
Chemicals/Metabolites with Overall Significant Relationships 4-31
Table 4.2.11 .a Codes and Descriptions for Chemicals/Metabolites with
Significant Relationships for Questions in the Residential
Location Category 4-32
Table 4.2.11 .b Distribution of Significant Medium/Question Relationships with
Residential Location Questions, by Medium 4-32
Table 4.2.11 .c Residential Location Questions and Chemicals/Metabolites with
Overall Significant Relationships 4-33
Table 4.2.12 Questions from Source Categories Considered Overall
Statistically Significant, by Medium 4-34
xv
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.13.a Codes and Descriptions for Metabolites with Significant
Relationships for Questions in the Subject's Personal
Characteristics Category 4-36
Table 4.2.13.b Distribution of Significant Medium/Question Relationships with
Subject's Personal Characteristics Questions, by Medium 4-37
Table 4.2.13.c Subject's Personal Characteristics Questions and Metabolites with
Overall Significant Relationships 4-37
Table 4.2.14.a Codes and Descriptions for Metabolites with Significant
Relationships for Questions in the Child's Behaviors Category 4-38
Table 4.2.14.b Distribution of Significant Medium/Question Relationships with
Child's Behaviors Questions, by Medium 4-39
Table 4.2.14.c Child's Behaviors Questions and Metabolites with Overall
Significant Relationships 4-39
Table 4.2.15.a Codes and Descriptions for Metabolites with Significant
Relationships for Questions in the Dietary Behaviors Category 4-40
Table 4.2.15.b Distribution of Significant Medium/Question Relationships with
Dietary Behaviors Questions, by Medium 4-40
Table 4.2.15.c Dietary Behaviors Questions and Metabolites with Overall
Significant Relationships 4-40
Table 4.2.16.a Codes and Descriptions for Chemicals/Metabolites with
Significant Relationships for Questions in the Family Hygiene
Practices Category 4-41
Table 4.2.16.b Distribution of Significant Medium/Question Relationships with
Family Hygiene Practices Questions, by Medium 4-42
Table 4.2.16.c Family Hygiene Practices Questions and Chemicals/Metabolites
with Overall Significant Relationships 4-42
Table 4.2.17.a Codes and Descriptions for Metabolites with Significant
Relationships for Questions in the Smoking-Related Activities
Category 4-43
Table 4.2.17.b Distribution of Significant Medium/Question Relationships with
Smoking-Related Activities Questions, by Medium 4-43
Table 4.2.17.c Smoking-Related Activities Questions and Metabolites with
Overall Significant Relationships 4-44
Table 4.2.18.a Codes and Descriptions for Chemicals/Metabolites with
Significant Relationships for Questions in the Work
Exposure/Practices Category 4-44
Table 4.2.18.b Distribution of Significant Relationships with Work
Exposure/Practices Questions, by Medium 4-45
Table 4.2.18.c Work Exposure/Practices Questions and Chemicals/Metabolites
with Overall Significant Relationships 4-45
Table 4.2.19 Questions from Behavior Question Categories Considered Overall
Statistically Significant, by Medium 4-46
Table 4.2.20.a Codes and Descriptions for Metabolites with Significant
Relationships for Questions in the Related Exposure Levels
Category 4-47
xvi
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.20.b Distribution of Significant Relationships with Related Exposure
Levels Questions, by Medium 4-47
Table 4.2.20.C Related Exposure Levels Questions and Metabolites with Overall
Significant Relationships 4-48
Table 4.2.21 .a Codes and Descriptions for Metabolites with Significant
Relationships for Questions in the Health Category 4-48
Table 4.2.21 .b Distribution of Significant Relationships with Health Questions,
by Medium 4-49
Table 4.2.21 .c Health Questions and Metabolites with Overall Significant
Relationships 4-49
Table 4.2.22 Questions from Other Question Categories Considered Overall
Statistically Significant, by Medium 4-50
Table 4.3.1 Number of Yuma Study Principal Participants, by School and
Grade Level 4-53
Table 4.3.2 Number of Principal Participants Where Yuma Study Dust
Samples Were Collected, by School and Grade Level 4-54
Table 4.3.3 Pesticides Measured in Yuma Study Household and School Dust
Samples 4-54
Table 4.3.4 Results of Regression Models with DMOP and DEOP,
Unadjusted and Adjusted for Creatinine, for 152 Households 4-56
Table 4.3.5 Results of Regression Models with Individual DAP Metabolites,
Unadjusted and Adjusted for Creatinine, for 152 Households 4-57
Table 4.3.6 Results Comparing Distance from Home to Agricultural Fields
with Six DAP Metabolites, Unadjusted and Adjusted for
Creatinine, for Principal Participants 4-59
Table 4.3.7 Results of Regression Models with DMOP and DEOP,
Unadjusted and Adjusted for Creatinine, and the Ten Pesticides
Most Detected in Household Dust Samples for 152 Households 4-60
Table 4.3.8 Results of Regression Models with Individual DAP Metabolites,
Unadjusted and Adjusted for Creatinine, and the Ten Pesticides
Most Detected in Household Dust Samples for 152 Households 4-61
Table 4.3.9 Results of Regression Models with DMOP and DEOP,
Unadjusted and Adjusted for Creatinine, and the Seven Pesticides
Most Detected in Household and School Dust Samples, for
Principal Participants 4-62
Table 4.3.10 Results of Regression Models with Individual DAP Metabolites,
Unadjusted and Adjusted for Creatinine, and the Seven Pesticides
Most Detected in Household and School Dust Samples, for
Principal Participants 4-64
Table 4.3.11 Questions and DAP Metabolites with Significant Relationships in
the Yuma Study Based on Tables 4.3.4, 4.3.5, and 4.3.6 4-66
Table 4.3.12 Number of Yuma Study Core Principal Participants, by School
and Grade Level 4-67
Table 4.3.13 First Ten Principal Components from Two Scenarios Using Yuma
Study Data 4-70
xvii
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.3.14 Cross-Reference for CART Analyses Performed on Yuma Study
Data 4-71
Table 4.3.15 Categories of Selected Predictors from CART Analyses of DAP
Sums for Yuma Study Participant Children 4-72
Table 4.3.16 Results from Non-statistical Comparison of Questionnaire
Responses Between High and Low Ends of Measurement Sum
Distributions 4-74
Table 4.3.17 Comparison of Selected Predictors from Yuma Study Report and
Data Mining Approach for Sum of Ethylated DAPs 4-75
Table 4.3.18 Comparison of Selected Predictors from Yuma Study Report and
Data Mining Approach for Sum of Methylated DAPs 4-75
Table 4.4.1 Summary of Predictor Categories Selected as Useful in
Differentiating Children's Pesticide Exposure Levels Across Two
Approaches 4-76
xviii
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Figures
Figure 1.1 Activities, Pathways, and Routes Related to a Child's Exposure to
Pesticides 1-3
xix
August 2005
-------
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
1.0 INTRODUCTION
1.1 Introduction
The Pesticides in Young Children Border States Program (USEPA 2002) includes a series of
studies designed to develop and implement an approach for examining the cumulative risks
and potential health effects in children from repeated exposure to pesticides via multiple
sources and pathways. The work of this program includes identifying the major exposure
risk factors for intervention that require studying the full range of the exposure measurement
distribution for children, especially for children with high exposure measurements. Various
screening processes have been studied to help identify populations of interest. This report
presents an evaluation of questions and environmental measures used in previous exposure
studies as potential indicators of pesticide exposure (e.g., metabolite levels in children's
urine).
1.2 Children's Exposure to Pesticides
Pesticides represent a wide range of chemicals that are used in agricultural production, vector
control, and food preservation, as well as in residential environments for aesthetic and pest
control purposes. Many pesticides currently registered in the United States have documented
health effects, including acute toxicity and carcinogenicity. The effect of relatively low
levels of pesticide exposure in children is an area of great scientific uncertainty, and has,
therefore, become the focus of substantial research and regulatory activities. It is widely
acknowledged that children are more vulnerable to many environmental health hazards than
adults, including pesticides, because they are more exposed and because they may have
elevated susceptibility (Needham and Sexton 2000).
The 1993 National Academy of Sciences report, Pesticides in the Diets of Infants and
Children, highlighted this concern, and pointed out the complexities involved in the
evaluation of aggregate exposures and cumulative risks (NRC 1993). The Food Quality
Protection Act (FQPA) of 1996 called upon the U.S. Environmental Protection Agency (U.S.
EPA) to implement risk assessment procedures that would be protective of children in their
dietary exposures to pesticides, and that would factor in exposures from other sources, such
as residential pesticide use (http://www.epa.gov/opppspsl/fqpa/).
The last decade has seen a substantial increase in studies aimed at characterizing children's
exposure to pesticides. The U.S. EPA Science to Achieve Results (STAR) grant program
(http://www.epa.gov/ncer/grants/), established in 1996, has provided an ongoing source of
funds that are distributed to investigators throughout the country, based on peer review of
proposed projects for scientific excellence. Several of the STAR grant programs have
addressed children's exposure to pesticides (http://www.epa.gov/ncer/grants/ and USEPA
1997). A substantial research program on children's pesticide exposure (USEPA 2000b,
USEPA 2003) has also been developed within the Agency, and in conjunction with other
federal agencies, such as the Centers for Disease Control and Prevention (CDC). The work
produced by these studies affords the first opportunity to evaluate systematically the
effectiveness of particular exposure assessment methods.
1-1
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
1.3 Populations of Concern
Recent studies (Aprea 2000, Lu 2001, Adgate 2001, Curl 2003, Heudorf 2004, Morgan 2004)
have tended to focus on relatively young children - either pre-school or elementary school
age. It is believed that such children may be at greater risk, both in terms of exposure and
susceptibility. Several sub-groups of children have been studied because they were believed
to be at particularly high risk. For example, numerous studies (Koch 2002, Shalat 2003,
Fenske 2002, Royster 2002) have focused on children in agricultural communities due to the
high use rates of pesticides in and around these communities. Children of minority or
disadvantaged groups have also been examined (Mills and Zahm 2001, Grossman 2001,
Krinsley 1998, McCauley 2001, Quandt 2004) as part of a broader environmental justice
initiative within the federal government. Finally, children whose parents are exposed to
pesticides in the workplace have received special attention (Loewenherz 1997, Lu 2000,
Azaroff 1999, Curl 2002), as it is well documented that workplace contaminants can
contribute to the exposure of children. Participant recruitment strategies for these studies
have ranged from convenience to probability-based sampling approaches.
1.4 Exposure Pathways and Patterns
Children's exposure to pesticides typically involves multiple pathways and multiple routes as
shown in Figure 1.1, which is adapted from Cohen Hubal (2000a) and Cohen Hubal (2000b).
1-2
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Inhalation
Activity
Patterns
Source
•indoor
residential
•outdoor
residential
•industrial
•agricultural
Exposure
Media
Outdoor
•air
•water
•soil
•plants
•surfaces
Indoor
water
house dust
surfaces
clothes
Dermal
Contact
Contact
Activities
Respiratory
Tract
Inhalation
Exposure
Rate
Nondietary Activities
Ingestion
Skin
Surface
Dermal
Exposure
Rate
Dermal Uptake
R.T. Uptake
Absorbed
Mouthing
Activities
-M Diet
Dietary Ingestion
G.I.
Tract
Ingestion
Exposure
Rate
G.I. Uptake
Figure 1.1 Activities, Pathways, and Routes Related to a Child's Exposure to Pesticides
(adapted from Cohen Hubal 2000a and Cohen Hubal 2000b)
Although most of the pathways and routes for children are similar to those for adults, the
types of, and amount of time spent at, activities will differ from adults, and between children
of different ages. As a result, the assessment of such exposures is among the most
challenging tasks faced by exposure assessment scientists. In addition to exposure from diet
and drinking water, children naturally explore their environment. Contact with surfaces,
frequent mouthing activities, and even the consumption of soil or dust can contribute to
exposures, and the temporal and spatial patterns of these exposures can be unpredictable
(Black 2005).
1.5 Questionnaire Data
Questionnaires have long been used as a primary source of exposure data in epidemiologic
studies. Most epidemiologic investigations are initiated after exposure has begun, and
structured interviews or questionnaires are used to reconstruct historical exposure patterns.
Such an approach raises questions regarding recall accuracy and possible bias (Teitelbaum
2002).
In contrast, contemporaneous studies of children's pesticide exposure often include some
1-3
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
type of exposure measurement. In such study designs, questionnaires are generally used to
identify important risk factors or exposure pathways. The exposure measurements are
usually considered objective measures of exposure and are treated statistically as outcome
variables. Questionnaire data are more easily collected than environmental measurements
and are, therefore, more cost-effective. In large population studies questionnaires may be the
only means available to ascertain exposure information. Yet questionnaires carry with them
some important limitations. Questions are often posed in very general terms (e.g.,
"ever/never") and so fail to specify the temporal pattern of the exposure. Questions may also
lack specificity in regard to behavior (e.g., does a child practice hand-to-mouth behavior), so
that the frequency of the behavior remains unknown. While many researchers have
attempted to address these limitations, there is also a limit to the number of questions that can
be posed and the level of detail that can be requested before study participants become
reluctant or unable to continue with the study.
Two large-scale research endeavors have made great strides in the development and use of
questionnaires in exposure assessment and epidemiology. First, the National Human
Exposure Assessment Survey (NHEXAS) (Sexton 1995b) was developed by the U.S. EPA
with the specific goal of improving the quality of exposure data. Studies sponsored under
this program have included probability-based subject recruitment, carefully tested
questionnaires and diaries, and accompanying environmental measurements for a variety of
environmental contaminants including pesticides. Second, the Agricultural Health Study
(http://www.aghealth.org). led by the National Cancer Institute (NCI), has incorporated a
prospective study design, using questionnaires at the outset of the study, and periodically
throughout the life of the study. NCI has also worked collaboratively with U.S. EPA to
examine the validity of questionnaire responses through the collection of exposure
measurements (Dosemeci 2002). These studies are likely to add significant new knowledge
to the field of exposure assessment.
1.6 Exposure Measurements
Measurements used in studies of children's exposure to pesticides have focused primarily on
samples collected in the children's microenvironments (e.g., homes, daycare centers, special
play areas). Sampling media have included soil, house dust, and wipes of surfaces. The
hands of children have also been washed or wiped to provide a relative indicator of dermal
exposure. An important feature of many of these studies has been the quantitation of
children's behavior (e.g., frequency of hand-to-mouth contact), although most data on
behavior have been collected in studies that have not included exposure measurements
(Cohen Hubal 2000b). Finally, biological exposure methods have been used to evaluate
pesticide exposure (Barr 1999, Barr and Needham 2002). Urine sampling has been used most
frequently in studies of children's exposure, as it does not involve invasive sampling. Saliva
monitoring of pesticides has proven feasible in animal models (Lu 2003). Current studies
include the collection and analysis of saliva samples from young children in an effort to
measure pesticides directly rather than as metabolic byproducts
(www.sph.emory.edu/eoh/facultv/Lu.htmn.
Environmental and biological samples are considered objective measures of exposure, and
1-4
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
they are sometimes viewed as representing a kind of "gold standard" when compared with
questionnaire data. However, these measurements can have substantial analytic variability.
For example, house dust is a complex matrix that may vary from location to location within a
residence. Extraction procedures may produce a range of values for the same measured
sample. Metabolites in urine may require derivatization before analysis, a step that can
introduce variability in the analytical results.
Furthermore, exposure measurements can vary over time. This is most pronounced in the
case of spot urine samples collected to measure exposure to pesticides that are rapidly
excreted (1-3 days). Significant variability can be observed day to day for the same
individual as well as for a group of individuals whose exposures are presumably the same.
Thus, it is important to consider both the quality and the potential for variability in exposure
measurements when assessing the utility of questionnaire data in predicting exposure levels.
1.7 Border States Program: Pesticides in Young Children
Research for the Pesticides in Young Children Border States Program is being, and has been,
conducted in the U.S.-Mexico Border States of Arizona, California, New Mexico, and Texas
as part of the Environmental Health Workgroup on the U.S.-Mexico Border program
(http://www.epa.gov/orsearth) which was developed with the passage of the North American
Free Trade Agreement (NAFTA). A three-phase approach was undertaken to address the
project objectives. Phase I was a planning phase. It included a review of existing
environmental pesticide exposure and health data, and the identification/review of techniques
for measuring pesticides and pesticide biomarkers in environmental and biological media
Phase II evaluated the extent and distribution of pesticide exposure in children living in the
border region with the intent of identifying those children with the highest levels of exposure.
Phase II also included methods development and evaluation studies to fill data gaps needed
for the design of Phase III. The initial Phase II analyses suggest that the existing Phase II
studies have not identified a definitive population for the Phase III activities.
The planned Phase Ilia would include a more complete monitoring of children classified in
Phase II as "high end exposures." Follow-up on these children would include detailed
measurements of their environmental exposure and biological monitoring for levels of
metabolites. From the Phase Ilia effort, a study would be designed to evaluate the
relationships between pesticide exposures and selected health outcomes and to define
specific hypotheses to be tested. An epidemiological study (Phase Illb) may then be
performed to examine the specific hypotheses about the impact of pesticide exposure on
health status/outcome of children. In order for the Pesticides in Young Children Border
States Program to move towards the Phase III goals, better exposure classification tools are
needed to identify a subset of children likely to have higher exposure levels so that a Phase
Ilia study can be performed in a cost effective manner.
1.8 Motivation and Goal of the Project
The current goal of many exposure assessment studies is to collect information using
questionnaires and diaries, environmental measurements and biomarkers, to develop an
1-5
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
understanding of the levels of chemicals to which people are exposed and to understand how
these exposures might occur. This approach is useful when the objective of the study is to
determine the population distribution of exposure to a chemical. A different approach is
needed; however, when the objective is to identify those individuals with higher exposure
levels for intervention and additional study. In many exposure assessment studies, the
chemical/metabolite levels from most of the participants are either not detectable or below
the level of concern for the chemical of interest. Thus, a large proportion of the available
funding for a study may be used to collect data that has little value in identifying and
subsequently protecting the most highly exposed individuals. Since the collection and
analysis of environmental and biological samples is an expensive portion of the exposure
evaluation process, considerable savings in time and money may be realized if questionnaires
could be used in an effective manner to predict those individuals with higher exposure levels.
Because of the potential reduced costs and other benefits, the Pesticides in Young Children
Border States Program would like to employ questionnaires in the Phase Ilia study for
exposure classification to aid in selecting the desired population for study. This selection
would produce either an enriched population (i.e., a larger percentage of individuals with
higher exposure levels) or would eliminate from further analysis individuals with lower
exposure levels. These interests require having questions that can predict the likelihood that
an individual has been exposed to pesticides.
1.9 General Approach and Report Contents
The work presented in this report offers researchers another tool for selecting the questions to
be included in a study's design. In some studies, experts using a Delphi consensus process
(http://www.scu.edu.au/schools/gcm/ar/arp/delphi.htmn to consider hypothetical and
observed relationships described in the research literature that are pertinent to the study's
interests. This report reviews the state of the science in terms of how well questionnaire
responses, from previous exposure studies, are statistically related to environmental and
biological measurements for children. The approach for evaluating these relationships was
twofold, and includes:
• A literature review of previous exposure studies to summarize the existence of such
quantitative and qualitative relationships, and
• An analysis of a recent children's pesticide exposure study in Yuma, Arizona, which
included questionnaires and measurements.
Section 2 of this report provides a summary of this project's results and recommendations for
additional work. It describes both the questions and categories of questions that were found
to be the most useful in differentiating children's pesticide exposure levels to pesticides in
the context of future study designs. Section 3 describes the methodology used for the
literature review and the statistical analysis of the Yuma study data. Section 4 includes
details of the results from the literature review and from the Yuma Study data. Section 5 lists
the references cited in this report. Appendix A lists the publications included in the literature
review. Appendices B, C, and D present the overview, detail, and comments tables
describing the relationships between questionnaire responses and pesticide exposure
1-6
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
measurements extracted from the literature review, as summarized in Section 4. Appendix E
describes the questions for which relationships from the literature review were tracked.
Appendix F lists the chemicals used in the Yuma Study analyses, and describes the molar
weighting process used to create the combinations of the chemicals/metabolites. Appendix G
describes the methodology and provides detailed results of the data mining approach for the
Yuma Study as summarized in Section 4.
1-7
August 2005
-------
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
2.0 SUMMARY AND RECOMMENDATIONS
2.1 Introduction
Children are widely acknowledged to be more vulnerable than adults to many environmental
health hazards, including pesticides, because they are more exposed, and because they may
have elevated susceptibility (Needham and Sexton 2000). The effect of relatively low levels
of pesticide exposure in children is an area of great scientific uncertainty and has, therefore,
become the focus of substantial research and regulatory activities. The work of the
Pesticides in Young Children Border States Program (USEPA 2002) includes identifying the
major exposure risk factors for intervention, which requires studying children across the
exposure measurement distribution, especially those with higher exposure measurements.
The goal of many exposure assessment studies is to collect information using questionnaires
and time-activity diaries, environmental measurements and biomarkers, to develop an
understanding of the levels of chemicals to which people are exposed and to understand how
the exposures might occur. In the typical exposure assessment study, the measurements from
most of the participants are either not detectable or below the level of concern for the
chemical of interest; thus, a large proportion of available funding for a study may be spent
collecting data that has little value in identifying, and subsequently protecting, the most
highly exposed individuals. Considerable savings in time and money may be realized if
questionnaires could be used for exposure classification that would aid in selecting the
desired population for a study. Such a screening tool would help produce either an enriched
population (i.e., a larger percentage of individuals with higher exposure levels) or would
eliminate individuals with lower exposure levels from further review.
The objective of this project is to identify questions that indicate a higher likelihood of
predicting a child's level of exposure to pesticides as input to future study designs. The work
presented in this report reviews how well questionnaire responses, based on previous
exposure studies, are statistically related to environmental and biological measurements for
children. The approach for evaluating these relationships was twofold, and includes:
• A literature review of previous exposure studies to summarize the existence of such
quantitative and qualitative relationships, and
• An analysis of a recent children's pesticide exposure study in Yuma, Arizona, which
contained questionnaires and measurements.
This section describes both the questions and categories of questions that were found to be
the most useful in differentiating children's pesticide exposure levels to pesticides in the
context of future study designs, discusses issues in developing effective screening tools, and
includes recommendations for additional work.
2.2 Methods
Questions for a new exposure study design are usually selected based on theoretically-
defined or hypothesis-driven relationships, the results of relationships tested in previous
2-1
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
exposure studies, and the interests of the new study. Relationship, as used in this report, is
defined as a systematic correspondence between the values of two variables, that is,
questionnaire responses and analytical measurements. This correspondence may or may not
be statistically significant. Some questions or categories of questions become the typical
selections for exposure studies because of the results from previous studies. This report
reviews the state of the science in relating questionnaire responses to environmental and
biological measurements, primarily for children, based on results and data from previous
exposure studies. A two-part approach evaluated questions from the studies to identify those
showing strong, that is, statistically significant, relationships with pesticide exposure levels.
One part of the approach was based on a literature review of previous exposure studies. A
search through several resources identified over 100 citations that might meet the criteria set
for this evaluation. The abstracts and full publications, where necessary, were reviewed with
respect to the following criteria:
• Was pesticide exposure studied?
• Were relationships between questions and measurements from monitoring, and
preferably urine, samples described? and
• Were children included as part of the population studied?
The 20 publications that were selected as being relevant for evaluating the usefulness of
questions were reviewed in detail to determine the relationships that were considered and
statistically analyzed in each study. A simple MS Excel database was created to track the
relationships between questions, and environmental and biological measurements, as noted in
the publications, whether or not the statistical tests of the relationships were statistically
significant. Detailed information about each relationship was compiled to allow for more in-
depth evaluations by researchers as their interests dictated. The questions were grouped into
categories for evaluation and questions showing overall significance across the publications
were highlighted. It should be noted that the 20 publications represent 14 distinct studies.
Details on the publication selection process and the extraction of the relationship information
are described in section 4.2.
The second part of the approach reviewed a recent study of children's exposure to pesticides
from Phase II of the Pesticide Exposure and Health Effects on Children Initiative (section
1.6). Some analyses had already been performed to meet the study's objectives and were
summarized in a report (CDC 2002). Subsequent analyses of the data using a data mining
approach were then performed for this project to identify relationships that exist in the data
rather than ones that are predetermined by the study's hypotheses. Preliminary bivariate
analyses and principal component analyses were performed to fine-tune the analysis
approach and to help understand some of the underlying structures within the data. The main
type of analysis performed was Classification and Regression Trees (CART). This technique
was used to investigate several scenarios of questions and dust measurements as potential
predictors for the sum of the ethylated dialkylphosphates (DAPs) and the sum of the
methylated DAPs (Appendix F). The questions and dust measurements that were selected as
predictors from the CART analysis in more than 50% of the scenarios were identified as
useful in differentiating exposure levels.
2-2
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
2.3 Results
2.3.1 Literature Review
Relationships are evaluated as the means for testing a study's hypotheses, and results of these
evaluations are included in publications based on the study. From the 20 relevant
publications, 603 relationships across 117 questions in 14 question categories were
identified. All statistically significant and non-significant relationships were noted except for
a large number of non-significant relationships alluded to in Sexton (2003) (section 4.2.2.2).
Appendix E lists the questions included under each of the 14 categories.
Table 2.1 Distribution of Questions and Relationships Across 14 Question Categories
Group
Category
#
questions3
#
relationships'3,0
%
relationships
#
publications'1
Source
Residential pesticide use
32
100
16.6
11
Source
Household characteristics
17
73
12.1
7
Source
Residential sources
(environmental
measurements)
3
13
2.2
4
Source
Household occupation
16
115
19.1
11
Source
Residential proximity to
agricultural fields
2
72
11.9
10
Source
Residential location
5
14
2.3
4
Behavior
Subject's personal
characteristics
6
78
13.0
9
Behavior
Child's behaviors
6
20
3.3
4
Behavior
Dietary behaviors
4
16
2.7
3
Behavior
Family hygiene practices
11
81
13.4
7
Behavior
Smoking-related activities
3
4
0.7
1
Behavior
Work exposure practices
4
4
0.7
2
Other
Related exposure levels
2
5
0.8
1
Other
Health
6
8
1.3
1
Total
117
603
100
a See Appendix E for list of questions tracked in the literature review.
b # relationships obtained by totaling numbers from tables in Appendix B for each category.
c See section 4.2.2.2 regarding relationships from Sexton (2003).
d Number of publications from the relevant list that were sources for the relationships.
Eighty-six percent of the relationships were in the categories of residential pesticide use,
household characteristics, household occupation, residential proximity to agricultural fields,
subject's personal characteristics, and family hygiene practices (Table 2.1). These six
2-3
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
categories represent questions whose relationships with exposure measurements have both a
strong theoretical basis, and could be reasonably evaluated through the study designs.
Table 2.2 Distribution of Questions and Relationships Across Medium Measured
Medium
# questions3,13
# relationships0,11
% relationships
# publications6
Urine - DAP
47
267
44.3
12
Urine - non-DAP
68
129
21.4
5
Dust
42
187
31.0
8
Indoor Air
2
3
0.5
1
Outdoor air
1
2
0.3
1
Personal air
2
4
0.7
1
Soil
2
8
1.3
1
Solid food
2
3
0.5
1
Total
Id
CD
603
100
a See Appendix E for list of questions tracked in the literature review.
b Some questions were related to measurements in more than one medium.
c See section 4.2.2.2 regarding relationships from Sexton (2003).
d # relationships obtained by totaling numbers from tables in Appendix B for each category.
e Number of publications from the relevant list that were sources for the relationships.
f Some questions are used with more than one medium, thus, the total differs from the total in Table 2.1.
Sixty-six percent of the relationships considered metabolites in urine measurements (Table
2.2). The majority of these relationships were with DAP-based metabolites. Dust
measurements have the next largest number of relationships with 31%. Both urine and dust
have been shown in other studies, mostly with adults, to be more useful indicators of
exposure level, and easier to collect from participants than other media. Few studies
extended measurement collection to other media. Studies with measurements of other
environmental media in conjunction with children's urine are not plentiful and were not
evaluated.
The questions identified in the publications were reviewed for their ability to differentiate
children's pesticide exposure levels based on whether the majority of the question's
relationships were found to be statistically significant (p < 0.05) or marginally significant
(0.05 < p < 0.10). This criterion summarizes the results of the relationships at a very high
level, and does not take into account any differences in populations sampled or in
interview/measurement situations for the studies described in the publications. Thus, the
relationships for a question or category of interest should be reviewed in more detail with
information in Section 4 and in Appendices B, C and D to determine their applicability to a
particular situation.
Summary statistics for the exposure measurements, when available, were extracted from the
publications for each relationship to provide researchers with additional information to help
understand the relationships analyzed. When judging the appropriateness of a question for a
2-4
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
future study, the researcher should also consider the difference between statistical and
practical significance. The p-value associated with each relationship analyzed measures the
strength of the relationship from the statistical perspective. Measures of central tendency,
e.g., means, medians and coefficients from a multiple or logistic regression, give insights into
the strength of the relationship from a practical level. The difference in the medians of two
groups may be statistically significant, but both median values may be lower than
measurement levels of interest for mitigating potential exposure. Thus, the magnitude of the
difference between the two groups may not be useful for practical considerations.
Dust and urine measurements were found in 97% of the relationships. Measurements for the
other media were found in only two of the 20 publications: Sexton (2003) and Simcox
(1995). The relationships for each question and chemical/metabolite combination were
reviewed to determine the question's effectiveness in differentiating the exposure levels. Not
all question/chemical combinations were evaluated in the studies to the same extent. The
number of relationships in which a question is evaluated, especially when the question is
used with more than one study population, gives additional credence to the question as a
potential differentiator. Generally the questions showing the most effectiveness are:
• residential pesticide use (inside and outside)
• occupation of household members
• child's characteristics (age, ethnicity, income)
• family hygiene practices
• household dust.
Several other questions also show some effectiveness:
• pets
• household location (urban vs non-urban)
• dietary behaviors (organic food)
• exposure levels of household members
• health status (diseases)
• smoking behaviors
• proximity to agricultural fields (for house dust only).
The number of relationships evaluated for the second group of questions was small,
indicating that their effectiveness has not been tested as extensively as for the questions in
the first group.
For urine measurements, questions showing usefulness as indicators of a child's pesticide
exposure level cover the areas of residential pesticide use both indoors and outdoors,
household occupation, subject's personal characteristics, family hygiene practices, and
smoking behavior. Each of these indicators seems plausible, in that such relationships have
been seen in previous investigations of environmental exposures (e.g., lead exposure in
children). Some smoking activities were identified as potential differentiators (section
4.2.5.5); however, considerations regarding the study population in which they occurred and
the very limited transferability of any pesticides through second-hand smoke makes this
2-5
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
question less effective for purposes of this project. For dust measurements, the questions
showing usefulness as indicators of a child's pesticide exposure level cover the areas of
household occupation, residential proximity to spraying, and family hygienic behavior. Each
of these indicators also seems plausible in terms of pesticides being present in the child's
environment. These questions represent potential exposure from the take-home pathway and
from agricultural pesticide spraying. A list of the specific questions and chemical/metabolite
combinations found to be generally effective as differentiators of the exposure levels are
presented in section 2.5.
The set of question categories used in this report provides one perspective for organizing the
relationships. Three risk or exposure factors related to the take-home or para-occupational
exposure pathway were analyzed as separate categories in this report: household occupation,
family hygiene practices, and work exposure/practices. Household occupation was
considered a source that would result in measurable differences in children's pesticide
exposures, because it may represent a surrogate for the actual exposure levels of household
members employed in agriculture. Children may be exposed to agricultural chemicals
through this pathway and their exposure levels are dependent on the occupational status,
work, handling, and hygiene practices of agricultural workers in their households.
Two other risk factors examined in this report also contribute to the para-occupational
exposure pathway. Family hygiene practices and work exposure/practices were considered
behavioral practices that could modify pesticide exposure to agricultural workers and their
family members. There were fewer relationships in these two categories because the studies
under review were primarily environmental exposures studies conducted in agricultural
communities with a focus on children. If these studies had been strictly occupational
exposure assessment studies, more questions related to the work and family hygiene practices
might have been included in these studies.
2.3.2 Children's Pesticide Exposure Study (Yuma Study)
A study of children's exposure to pesticides was conducted in Yuma, Arizona for 152
households. In cooperation with eight local schools in the study area, families who were
permanent residents of the area with children in kindergarten or first grade were self-selected
to participate in a study conducted from October 1999 through February 2000. A urine
sample was collected from each of these children (principal participants) and from any
sibling in the household between the ages of 2 and 11 years. The urine samples were
measured for the six most common dialkylphosphate (DAP) metabolites associated with OP
pesticides. A dust sample was collected from each household, and from classrooms, with
principal participants. These samples were measured for specific organophosphorous (OP),
organochlorine, pyrethroid, and carbamate pesticides. A questionnaire regarding
characteristics and practices of the family and the principal participant child was
administered to each household. The study included 152 children as principal participants
and 127 siblings. A total of 244 urine samples were available for analysis. Dust samples
were available from 152 households and from 25 kindergarten and first-grade classrooms in
six of the participating schools.
2-6
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
In the Yuma Study report (CDC 2002), the urine, dust, and questionnaire data were analyzed
in order to describe levels of pesticide exposure and to identify any associations between
questionnaire data and pesticide exposure levels. Traditional statistical techniques were
performed to test predefined hypotheses. Urine measurements from all children in a
household, household and school dust measurements, and responses for a selected subset of
questions were included in the statistical analyses.
A second set of statistical analyses was performed on the Yuma study data specifically for
this project. The analyses included only principal participants from the eight schools and two
grades in the initial study design. A data mining approach was used to identify potential
predictors without specifying a priori hypotheses between the biomarker measurements and
the questions and dust measurements, that is, the approach was used to explore the
relationships that exist in the data (Hand 1999).
The statistical analyses in CDC (2002) focused on six DAPs (DEP, DETP, DEDTP, DMP,
DMTP, DMDTP) and the molar-weighted sums of the ethylated and methylated DAPs
(Appendix F).
Table 2.3 Questions and Dust Measurements from the Yuma Study Having Strong Relationships with
the DAPs, Adjusted for Creatinine (CDC 2002)
DEP, DETP,
or DEDTP'1
DEAP"
DMP, DMTP,
or DMDTP''
DMAP'
Questions1'0
Used pesticide inside home in last month
X
X
X
Distance from home to agricultural field
X
Father working in agriculture
X
Other adult in house working in agriculture
X
X
X
Father, mother or other adult working in
agriculture
X
Household and/or School Dust1
carbaryl
X
X
chlorpyrifos9
xg
xg
X
X
cis-permethrin
X
X
X
X
cy-permethrin
X
X
X
X
diazinon9
xg
X
X
gamma-chlordane
X
proxopur
X
X
X
trans-permethrin
X
X
X
X
a DEP = diethylphosphate, DETP = diethylthiophosphate, DEDTP = diethyldithiophosphate
DMP = dimethylphosphate, DMTP = dimethylthiophosphate, DMDTP = dimethyldithiophosphate.
2-7
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
b DEAP is a summary variable made from summing molar weights of DEP, DETP and DEDTP, and is noted as
DEOP in CDC (2002). (Concentrations < Limit of detection (LOD) were replaced with LOD/2.)
c DMAP is a summary variable made from summing molar weights of DMP, DMTP and DMDTP, and is noted as
DMOP in CDC (2002). (Concentrations < LOD were replaced with LOD/2.)
d See Tables 4.3.4, 4.3.5, and 4.3.6 for relationships with specific metabolites or sums.
e Full description of questions can be found in Table 3.1.
f See Tables 4.3.7, 4.3.8, 4.3.9, and 4.3.10 for relationships with specific metabolites or sums
g These are the only OPs for which relationships with ethylated DAPs are expected. All other significant
relationships may be indicative of heavy pesticide use, although they do not correspond to the metabolite
found.
Of the questions analyzed, recent indoor pesticide use, household members working in
agriculture, and distance from home to agricultural fields have statistically significant
relationships with the ethylated DAP sum, and individual ethyl and methyl DAPs (Table 2.3).
The directions for some of the relationships, however, are the opposite of what might be
expected based on current knowledge, that is, an exposure activity is not related to a higher
measurement level. Dust measurements of chlorpyrifos and the permethrins are strongly
related with both the ethylated and methylated DAP sums. Some of the significant
relationships between house/school dust measurements and the DAP measurements are
unexpected. These may be indicators of heavy pesticide use, although they do not
correspond to the metabolite found. The regression coefficients for these statistical analyses
are very small and may indicate that the relationships are not necessarily practically
significant. The report authors note:
The regression models in which the slopes were small but were statistically
significant may suggest either that a) true associations existed, but the
numbers of significance were less than the numbers measured in the
statistical programs or b) the associations were meaningless and based
solely [on] the probability of finding statistical significance if enough tests
were run. (CDC 2002)
Another set of analyses was conducted for this project using the data mining technique
Classification and Regression Trees (CART). Six scenarios (Table 2.4) of potential
predictors containing questions, and house and school dust measurements were evaluated for
the sums of ethylated and methylated DAPs (Appendix F). The scenarios covered three
levels of increasing measurement burden (questions, household dust, and school dust) with
two sets of questions for each level.
Table 2.4 Questions and Measurements Included in the Yuma Study CART Analysis Scenarios for
Each DAP Sum
Scenario
Full Set of
Questions3
Limited Set of
Questions'3
House Dust
Measurements
School Dust
Measurements
1
X
2
X
3
X
X
4
X
X
5
X
X
X
2-8
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Scenario
Full Set of
Questions3
Limited Set of
Questions'3
House Dust
Measurements
School Dust
Measurements
6
X
X
X
a Full set includes limited set of questions.
b Questions from full set considered more likely to be predictors of children's pesticide exposure level.
Only the limited set of questions was included in all six scenarios. Thus, the remaining
questions and dust measurements did not have as many opportunities to be selected as
predictors in the CART analyses. The criterion used to determine whether a question or dust
measurement had a strong relationship with one of the DAP sums was that it was selected in
the CART analyses a majority of times (> 50%) based on the number of scenarios in which
the predictor was included. Thus, for a house dust measurement to denote a strong
relationship with the biomarker measurement, the dust measurement would have to be
selected in at least three of the four CART analyses. This type of criterion identifies
predictors that are strong, because they are more universal across the scenarios.
The CART analyses were conducted using responses from the principal participant children
in the initial study design, that is, the children who were in one of the eight schools and were
in kindergarten or first grade. All six scenarios were conducted with 130 principal
participants. An explanation of the CART technique and details of the CART analyses can
be found in Appendix G.
Table 2.5 Questions and Dust Measurements from the Yuma Study Having Strong Relationships with
the Sum of Methylated DAPs Based on the Data Mining Approach
Predictor
Description
LTD
Q'1
%
Scenarios
Predictor
Selected
Number of
Scenarios
with
Predictor
Included
Questions
HEIGHT
Child's height (inches)
X
100
6
NCATWRKD
Father's occupation — categories
67
3
SCHOOL
Child's school
X
83
6
WEIGHT
Child's weight (lbs)
X
67
6
WHEEL
Distance between home and field - rotary
wheel
67
3
WHERTIME
Room where child spends most awake time
X
100
6
WHNCHEMO
Last time field treated with pesticides?
100
3
House Dust Measurement Sumsb
WCHLPYRF
Weighted chlorpyrifos
100
4
WDIAZNON
Weighted diazinon
75
4
WPERMSUM0
Weighted sum of cis-permethrin and trans-
permethrin
100
4
2-9
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
School Dust Measurement Sums"
None
a X — Question was in the limited subset and thus included in all six scenarios.
b Description of measurement sums can be found in Appendix F.
c Although these relationships do not correspond to the metabolite found, they may be indicative of heavy
pesticide use or may be a surrogate for some other exposure event.
Questions related to agricultural fields and child size, and household dust measurements were
selected as having strong relationships for the sum of methylated DAPs; however, school
dust measurements were not selected (Table 2.5). Initial analyses for the sum of ethylated
DAPs included CHLDTM3 (Child spends time at school) as a strong predictor, however, it
was difficult to understand the responses in the context of the population analyzed, that is,
children in kindergarten and first grade. Analyses were then performed excluding
CHLDTM3. Subsequently, CHLDTM3 was considered a possible indicator of additional
time spent at school, which might reflect additional exposure from the home environment for
the "NO" respondents because they were not spending more time at school (Table G.3.6).
Questions relating to diet, residential pesticide use, time spent at home, and agricultural
fields, and measurements from both the house and school were selected for the sum of the
ethylated DAPs excluding CHLDTM3 (Table 2.6).
Table 2.6 Questions and Dust Measurements from the Yuma Study Having Strong Relationships with
the Sum of Ethylated DAPs Based on the Data Mining Approach (Excluding CHLDTM3)
Predictor
Description
LTD
Q'1
%
Scenarios
with
Predictor
Selected
Number of
Scenarios
with
Predictor
Included
Questions
ETHNIC
Child's ethnic and racial background
67
3
HOURAWAY
Number hours/wk child not at home
X
100
6
NRMSPRYD
Number of rooms sprayed last month
100
3
VEGGIES
How often child eats local fresh fruit/veg?
67
3
WEIGHT
Child's weight (lbs)
X
100
6
WHEEL
Distance between home and field - rotary
wheel
100
3
WHNCHEMO
Last time field treated with pesticides?
100
3
House Dust Measurement Sums'1
WCHLPYRF0
Weighted chlorpyrifos
75
4
WDUSTBAL
Weighted sum of dust analytes except
OP pesticides
100
4
WDUSTSUM
Weighted sum of all dust analytes
100
4
WPERMSUM
Weighted sum of cis-permethrin and
trans-permethrin
100
4
School Dust Measurement Sums'1
2-10
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor
Description
LTD
Q'1
%
Scenarios
with
Predictor
Selected
Number of
Scenarios
with
Predictor
Included
SWCHLPYR0
Weighted chlorpyrifos
100
2
SWOPBAL
Weighted sum of OP pesticides except
chlorpyrifos, diazinon, permethrins, and
o-phenylphenol
100
2
SWOPSUM
Weighted sum of OP pesticides
100
2
a X — Question was in the limited subset and thus included in all six scenarios.
b Description of measurement sums can be found in Appendix F.
c Although these relationships do not correspond to the metabolite found, they may be indicative of heavy
pesticide use or may be a surrogate for some other exposure event.
2.4 Summary of Results from Two Approaches
The Yuma Study report (CDC 2002) looked at each question or measurement individually
and included siblings as well as principal participants using a general linear estimating model
with repeated measures for 152 households. The potential risk or exposure factors selected
for analysis were the subset of the full set of questions that were available for siblings as well
as principal participants, that is, the child's physical characteristics and household
characteristics or practices. The data mining approach used all the questions and
measurements simultaneously in CART analyses for only 130 principal participants in
kindergarten and first grade. Given these and other differences, it may be useful, with
caution, to look at a summary of the predictors selected under both approaches to evaluate
the universal strength of the predictors. It should be noted that some of the significant
relationships between house/school dust measurements and the DAP measurements may be
indicators of heavy pesticide use, although they do not correspond to the metabolite found.
Table 2.7 Comparison of Selected Predictors from Yuma Study Report" (CDC 2002) and Data Mining
Approachbfor Methylated Sum of DAPsc
Yuma Study Report
Data Mining Approach
No Questions
Child's characteristics (height, weight)
Proximity to agricultural fields, spraying conditions
Father's occupation
Where in house child spends timed
Child's schoold
Household dust®: diazinon, chlorpyrifos, permethrins,
carbaryl
Household dust®: diazinon, chlorpyrifos, permethrins
School dust®: diazinon, permethrins,
School dust: none
a Based on Tables 4.3.4, 4.3.7, and 4.3.9 and the molar-weighted sum of methylated DAPs (adjusted for
creatinine).
b Based on Table G.3.5 and log (molar-weighted sum of methylated DAPs-adjusted for creatinine).
c See definition in Appendix F.
2-11
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
d Questions were not analyzed in CDC (2002) because responses were not available for siblings.
e Although these relationships do not correspond to the metabolite found, they may be indicative of heavy
pesticide use or may be a surrogate for some other exposure event.
Table 2.8 Comparison of Selected Predictors from Yuma Study Report3 (CDC 2002) and Data Mining
Approachb for Ethylated Sum of DAPsc
Yuma Study Report
Data Mining Approach
Recent use of pesticides inside home
Recent use of pesticides inside home
Child's characteristics (weight, ethnicity)
Other adult in household working in agriculture
Proximity to agricultural fields, spraying conditions
Child's time spent away from homed
Diet - local fruits/vegetablesd
Household dust®: OPs, permethrins, non-OPs
Household dust®: OPs, permethrins, non-OPs
School dust®: permethrins
School dust®: OPs
a Based on Tables 4.3.4, 4.3.7, and 4.3.9 and the molar-weighted sum of ethylated DAPs (adjusted for
creatinine).
b Based on Table G.3.4 without CHLDTM3 as a potential predictor and log (molar-weighted sum of ethylated
DAPs-adjusted for creatinine).
c See definition in Appendix F.
d Questions were not analyzed in CDC (2002) because responses were not available for siblings.
e Although these relationships do not correspond to the metabolite found, they may be indicative of heavy
pesticide use or may be a surrogate for some other exposure event.
The analyses in the Yuma Study report (CDC 2002) consider questions and measurements
that would apply as risk factors to the siblings as well as the principal participants, and for
which there were available responses. These factors may affect explanations of the
variability of the pesticide metabolite levels across siblings within a household. The data
mining approach focuses the analyses on a group of children with less diverse characteristics
in terms of school and grade level and includes all questions regarding the principal
participants. For the sum of methylated DAPs, no pesticides in household dust were similar
across both approaches and no questions were found significant in the Yuma Study report
(Table 2.7). For the sum of ethylated DAPs, recent use of pesticides inside the home, and
OPs, non-OPs, and permethrins in the household dust stand out as differentiators of
children's pesticide exposure level across both approaches (Table 2.8). The difference in
analysis techniques and the difference in participants included in the analyses may help
explain the differences in the predictors selected across the two approaches. Also, some
questions regarding sibling activities were not analyzed in CDC (2002) because that
information was not collected as part of the study design. The best use of these results is as
"indicators" of predictors that are more useful in differentiating the exposure levels.
Two approaches were taken in this project to identify questions that were useful in
differentiating children's pesticide exposure levels as a screening tool for selecting
participants of interest in future exposure studies. One approach reviewed relationships with
2-12
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
questions described in the literature from previous exposure studies. The second approach
reviewed relationships with questions based on a study of children's pesticide exposure in
Yuma, Arizona.
Table 2.9 Summary of Predictors Selected as Useful in Differentiating Children's Pesticide Exposure
Levels Across Two Approaches
Literature Review3
Yuma Studyb
Residential pesticide use
Residential pesticide use
Petsc
Occupation of household members
Occupation of household members
Household location: urban vs non-urbanc
Child's personal characteristics
Child's personal characteristics
Dietary behaviors (organic food)0
Dietary behaviors (local fruits/vegetables)
Family hygiene practices
Exposure levels of household members0
Health status (diseases)0
(Proximity to agricultural fields)d
Proximity to agricultural fields, spraying conditions
Where child spent time at home/not, or within home
a Based on the "c" tables: Tables 4.2.6.C - 4.2.21.C.
b Based on Tables G.3.5 and G.3.7.
c Only a small number of relationships evaluated these questions.
d Proximity to agricultural fields for the literature review was related to dust measurements only.
The types of questions that seem to be strong differentiators of children's pesticide exposure
levels based on both approaches are:
• occupation of adults living in household
• residential pesticide use
• residential proximity to spraying and agricultural fields
• characteristics of the subject that may indicate potential exposure activities
• family hygiene practices that may mitigate the take-home pathway exposure
• where the child spends time (in home, away from home)
• diet with respect to locally-grown fresh fruits and vegetables (Table 2.9).
It seems clear from this review that children's proximity to pesticide use can increase the
likelihood of their exposures, whether the source is residential pesticide use, agricultural
pesticide use near the residence, or pesticide exposure in the workplace that results in
residential contamination. It is also evident from one study that replacement of
conventionally produced fresh fruits and vegetables (i.e., pesticides used in production) with
organic produce can result in substantial decreases in urinary pesticide metabolite levels.
2-13
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Future studies that use biological monitoring and questionnaires should draw upon recent
research to refine study protocols. Several suggestions are provided in the following section
of recommendations.
2.5 Recommendations
2.5.1 Effective Differentiators of Exposure Level
Based on an evaluation of the relationships found in the literature review, forty-two questions
were identified as effective in differentiating exposure levels of at least one
chemical/metabolite in Table 2.11. These questions are offered as a resource of
recommended questions with specific chemicals or metabolites for future study designs.
Note that the questions were evaluated here as a screening tool to create an enriched
population of participants with higher exposure levels. Thus their future use is better suited
to similar purposes.
The chemicals and metabolites found in the publications were assigned to seven groups, for
presentation purposes, based on medium and type of chemical metabolite measured (Table
2.10).
Table 2.10 Description of Code Names and Groups Assigned to the Chemicals and Metabolites, Sorted
by Code
Chemicals/Metabolites
Medium
Grouping
Code
Description
urine
1-Non-DAP
1NAP
1-Naphthol
urine
1-Non-DAP
4NITR
4-Nitrophenol
other3
6-Chemical
ATZ
Atrazine
urine
1-Non-DAP
ATZM
Atrazine mercapturate
other
6-Chemical
AZM
Azinphosmethyl
other
6-Chemical
AZMPH
Azinphosmethyl+Phosmet
other
6-Chemical
CHLR
Chlorpyrifos
urine
3-DAP Sum
DAP1
DMP+DMTP+DMDTP+DEP+DETP+DEDTP
urine
4-DAP Detect
DAP2
DEP, DETP, DEDTP, DMP, DMTP
(at least one detectable measurement)
urine
5-DAP High
DAP3
DEP, DETP, DEDTP, DMP, DMTP
(at least one high measurement)13
urine
2-DAP
DEDTP
Diethyldithiophosphate (DEDTP)
urine
2-DAP
DEP
Diethylphosphate (DEP)
urine
2-DAP
DETP
Diethylthiophosphate (DETP)
urine
2-DAP
DMDTP
Dimethyldithiophosphate (DMDTP)
urine
2-DAP
DMP
Dimethylphosphate (DMP)
urine
2-DAP
DMTP
Dimethylthiophosphate (DMTP)
other
6-Chemical
EPAR
Ethyl parathion
2-14
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Chemicals/Metabolites
Medium
Grouping
Code
Description
urine
3-DAP Sum
ETHL1
DEP+DETP
urine
3-DAP Sum
ETHL2
DEP+DETP+DEDTP
urine
4-DAP Detect
ETHL3
DEP, DETP, DEDTP
(at least one detectable measurement)
other
6-Chemical
MAL
Malathion
urine
1-Non-DAP
MDA
Malathion dicarboxylic acid
urine
3-DAP Sum
MTHL1
DMTP+DMDTP
urine
3-DAP Sum
MTHL2
DMP+DMTP+DMDTP
urine
4-DAP Detect
MTHL3
DMTP (detectable measurement)
urine
4-DAP Detect
MTHL4
DMP, DMTP
(at least one detectable measurement)
urine
5-DAP High
MTHL5
DMP, DMTP
(at least one high measurement)13
urine
7-M eta bo lite NA
NA
Specific metabolite was not provided
other
6-Chemical
OPSUM
OP sumc
other
6-Chemical
PHSM
Phosmet
urine
1-Non-DAP
TCPY
3,5,6-T richloro-2-pyridinol
a Medium is noted as urine or other (any other medium sampled).
b See definition of high measurement in Azaroff (1999)
c OP sum = azinphosmethyl, chlorpyrifos, malathion, and phosmet.
Forty-eight questions across 12 question categories were considered effective differentiators
of the exposure measurement levels of at least one chemical/metabolite evaluated in the
relevant publications (Table 2.11). Their effectiveness was determined by whether a
majority (> 50%) of the relationships for a given chemical/metabolite were statistically or
marginally significant.
Table 2.11 Questions Considered Effective Differentiators of Children's Pesticide Exposure Levels
Based on a Literature Review of Previous Exposure Studies
Medium
Q Category
Q#a
Q Description11
Chemicals/
Metabolites c
Dust
Residential pesticide use
Q119
Outside Treatedd
CHLR
Household characteristics
Q202
Property Used As a Farmd
CHLR
Q213
Size of Household
AZM
Residential sources
(environmental
measures)
Q303
Outdoor Soil
EPAR
Household occupation
Q401
Agricultural Workers in
Household
AZM
2-15
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Medium
Q Category
Q#a
Q Description11
Chemicals/
Metabolites c
Q404
Applicator vs Farmworker
AZMPH, EPAR
Q405
Applicator vs Non-applicator
CHLR, EPAR
Q407
Applicator and Farm worker
vs Reference
AZM, AZMPH, CHLR,
EPAR, PHSM
Q412
Fieldworker vs Pesticide
Handler
AZM
Q415
Tree Thinning
OPSUM
Q416
Number in Household with
High Pesticide Contact
OPSUM
Residential proximity to
agricultural fields
Q501
Proximity of Home to
Pesticide-Treated
Farmland/Orchard
AZMPH, EPAR
Residential location
Q605
Vehicle vs House
AZM
Family hygiene practices
Q1006
Work Clothes Worn Indoors
AZM, OPSUM
Q1009
Number of Weeks Since
Last Vacuuming
OPSUM
Indoor Air
Household characteristics
Q202
Property Used As a Farmd
CHLR
Personal Air
Residential pesticide use
Q102
Inside Treated
CHLR
Q124
Level of Pesticide Used
ATZ
Soil
Household occupation
Q409
Farmer and Farm Worker
vs Reference
AZM
Solid Food
Residential pesticide use
Q119
Outside Treatedd
CHLR
Urine
Residential pesticide use
Q104
Inside Treated - Bedroom
TCPY
Q106
Inside Treated - Closets
TCPY
Q108
Inside Treated - Dining
Room
TCPY
Q111
Inside Treated — Living
Room
TCPY
Q117
Inside Treated — Other
Room
TCPY
Q119
Outside Treatedd
MDA, TCPY
Q120
Garden Treated
TCPY, ETHYL1,
METHYL2
Q121
Lawn/Yard Treatedd
TCPY
Q124
Level of Pesticide Used
MDA, TCPY
2-16
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Medium
Q Category
Q#a
Q Description11
Chemicals/
Metabolites c
Q125
Frequency Personal
Application Inside
TCPY
Q126
Frequency Personal
Application Outside
TCPY
Q127
Inside/Outside Treated by
Family Member
ETHYL3, METHYL3,
METHYL4, DAP2,
DAP3
Q130
Personally Mixed Pesticide
Inside
TCPY
Household characteristics
Q208
Pets in House
METHYL2
Q209
Pets Inside/Outside Housed
MDA
Q211
Existence of Garden or
Vegetable Gardend
ETHYL1, MDA
Residential sources
(environmental
measures)
Q301
Household Dust
METHYL2, NA
Household occupation
Q402
Household Member
Spraying Fields
DAP2, DAP3, ETHYL3,
METHYL3, METHYL4,
METHYL5
Q403
Recent Fieldwork
DAP2, DAP3,
METHYL4, METHYL5
Q406
Applicator vs Reference
DMTP
Q407
Applicator and Farm Worker
vs Reference
DMTP, METHYL1
Residential proximity to
agricultural fields
Q501
Proximity of Home to
Pesticide-Treated
Farmland/Orchard
DMTP
Residential location
Q601
Urban vs Non-urban
TCPY
Subject's personal
characteristics
Q702
Age
DAP1, METHYL2
Q703
Ethnicity
1NAP, MDA
Q705
Income
1NAP, MDA, TCPY,
DMTP, DAP1
Child's behaviors
Q806
Loading from Hand Wipe
DAP1
Dietary behaviors
Q904
Organic Diet
METHYL2
Smoking-related activities
Q1101
Current Smoker®
TCPY
Q1102
Subject Smoked®
TCPY
Related exposure levels
Q1302
High Levels in Adult
Household Members
DAP2, DAP3,
METHYL4
Health
Q1403
Bowel Disease
TCPY
Q1405
Intestinal Disease
TCPY
Q1406
Ulcers
TCPY
a For some of the significant relationships, the effect of the exposure factor was not in the direction expected.
2-17
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
See Appendix C for details on specific relationships.
b See Appendix C for specific question phrasings included under each question description.
c Chemicals or metabolites for which > 50% of the relationships with the question were statistically or marginally
significant. See Table 2.10 for chemical/metabolite description.
d See section 4.2.2 regarding relationships from Sexton (2003).
e Included only in Krinsley (1998) (section 4.2.5.5).
2.5.2 Urinary Metabolite Monitoring
A substantial proportion of the analysis in this report has focused on the association between
questionnaire data and pesticide metabolites in children's urine. These analyses have been
conducted on the assumption that the urinary metabolite measurements provide an accurate
estimate of children's exposure; that is, if statistical associations were not observed, it was
concluded that the questionnaire information was probably not a useful indicator of
children's pesticide exposure. Yet we know that the metabolites under study are processed
and excreted relatively quickly in humans (1-3 days) and, therefore, represent recent
exposures. In contrast, most of the questions asked of parents or children were of a general
nature in terms of the time frame of a particular activity or behavior. It is, therefore,
worthwhile to consider the variability in measurements in urinary pesticide metabolites.
Nearly all of the studies examined in this report have used spot urine samples as the outcome
that is compared to questionnaire data. A number of these studies have collected at least two
spot samples from children, but only one collected complete urine samples over a fixed time
period (Curl 2003). Several recent exposure studies have observed that intra-individual
variability in pesticide metabolite concentrations in urine can be high (Macintosh 1999,
Adgate 2001, Koch 2002). In these studies, an attempt was made to address this issue by
collecting samples on a repeated basis: Macintosh (1999) collected up to six samples from
each of up to 80 adult participants in the Maryland NHEXAS study, but the samples were
approximately eight weeks apart; Koch (2002) collected samples from pre-school children on
a bi-weekly basis for approximately one year. In both of these studies, the urine samples
were essentially independent from one another in relation to exposure sources, although in
the Koch study the 4-6 week agricultural spray season was identified as a time of elevated
exposure. Adgate (2001) introduced more of a TEAM (Total Exposure Assessment and
Monitoring) study design, that is, multiple samples over time, by collecting three morning
voids from children in the course of one week. Such repeated measures would have a better
chance of separating high and low exposed children if, for instance, a pesticide application
had occurred at the residence at the beginning of the week. However, none of these study
designs addresses directly the high day-to-day variability that seems to be the norm for
pesticide metabolite excretion in children, even when creatinine adjustments are performed.
In contrast Curl (2003) collected a full 24-hour urine sample to compare conventional and
organic dietary behavior, and was able to demonstrate a large difference in exposure between
these two groups. Krieger (2001) also collected 24-hour urine samples from children after
the use of total aerosol release devices (foggers) in residences and was able to discern clear
patterns in child exposure levels over time.
Two occupational exposure studies may serve as useful models for the design of future
studies of children's pesticide exposure that involve urinary metabolite monitoring.
Arbuckle (2002) examined the relationship between self-reported behaviors during pesticide
2-18
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
applications and urinary excretion of two herbicides—2,4-dichlorophenoxyacetic acid (2,4-D)
and 4-chloro-2-methylphenoxyacetic acid (MCPA). Urine samples were complete 24 hour
voids from the beginning of application through the following day. With this sampling
scheme the questionnaire's prediction of exposure had a sensitivity of 57% and a specificity
of 86% for 2,4-D; for MCPA the sensitivity and specificity were 92% and 67%, respectively.
A multivariate analysis was able to identify several variables as predictive of urinary
metabolite concentrations. Harris (2002) studied commercial pesticide applicator exposure
collecting two consecutive 24-hour urine samples from each participant. Investigators then
modeled weekly exposure and dose based on knowledge of the amount of pesticide used by
each applicator. This analysis was able to identify two major exposure factors: type of
nozzle used and use of gloves during application. These studies suggest that if complete 24
or 48 hour urine samples are collected, it may be possible to identify major risk factors for
exposure.
2.5.3 Questionnaire Validation
Few of the studies analyzed in this report have used validated questionnaires as a part of their
examination of children's pesticide exposure. Questionnaire validation includes a test for
accuracy (i.e., determine if the answer reported on the questionnaire by the study participant
is correct), usually by comparison of the study instrument results with a "gold standard" for
some subset of the study population. For example, answers to a question regarding a child's
absence from school could be checked against school attendance records. Validation may
also include tests for reliability (i.e., determine if the study participant provides the same
answer to the question when tested on several occasions.) Validity of questionnaire data is
an essential consideration in epidemiologic studies and future studies of children's pesticide
exposure should be preceded by validation studies. A good example of this approach is
available from the ongoing Agricultural Health Study conducted by the National Cancer
Institute in collaboration with other federal agencies in the United States (Alavanja 1994).
Dosemeci (2002) developed a quantitative metric for applicator exposure based on an
analysis of the existing scientific literature. This metric provides quantitative adjustment
factors for certain behaviors (e.g., use of gloves) reported in questionnaires and provides the
basis for classifying each applicator's exposure for epidemiologic analysis. A critical
component of the development of this model has been its validation through biological
monitoring. The U.S. Environmental Protection Agency has conducted a study of pesticide
applicators, collecting urinary metabolite data and comparing these to the questionnaire data
collected by the National Cancer Institute (Thomas 2004). This work has demonstrated good
correlations between self-reported behavioral data from applicators and the urinary
metabolite data. It would behoove those involved in the study of children's pesticide
exposure to consider this approach in the development of epidemiologic investigations.
2.5.4 Objective Measures of Children's Behaviors
Studies of children's pesticide exposure should work to improve the quality of data related to
behavior. At present, researchers rely primarily on parental reports of behavior for young
children. Yet the validity of parental reports has not been scrutinized in a systematic fashion.
Duplicate diet sampling over 24 hours (Macintosh 1999, Fenske 2002) is a good example of
2-19
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
an objective measure of pesticide exposure for the dietary pathway. In a recent study,
researchers have found that NHEXAS-style parental diaries of children's time-location
(macro-activities) were not accurate when compared with global positioning system (GPS)
measurements over a 24 hour period (Elgethun 2003, Elgethun 2004, Elgethun 2005).
Similarly, parental reports of children's contact with objects and mouthing behavior (micro-
activities) are not necessarily accurate when compared to videotaping (Reed 1999, Black
2005). There is clearly a need for more objective measures of children's activities and
behaviors in conjunction with systematic biological monitoring to ensure identification of
key predictors of children's exposure to pesticides.
2-20
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
3.0 METHODOLOGY
3.1 General Description of Approach
Reconciling the sources and outcomes of exposure is a complex process because of the
multitude of potential sources, interactions between sources and other factors, and timing
issues between an actual exposure and evidence of the exposure. When preparing for an
exposure study, investigators are likely to take into consideration both the hypothetical and
observed relationships described in the research literature for their study design. Examples
of hypothetical or theoretical relationships are found in the general environmental health
paradigm (Sexton 1995a) as models for source, concentration, exposure, and dose. Examples
of observed relationships are those identified in data analyses from an exposure study as in
Clayton (1999) which described the result of examining question-measurement relationships
for the NHEXAS Region 5 Study.
As a reference for the design efforts of the Pesticides in Young Children Border States
Program, and for other future exposure studies, this project compiled the observed
relationships between questions and measures of children's exposure to pesticides. The
assimilation and review of these relationships was performed using two approaches:
• A literature review of previous exposure studies to summarize the existence of such
quantitative and qualitative relationships, and
• An analysis of a recent children's pesticide exposure study in Yuma, Arizona, which
included questionnaires and measurements.
Relationship, as used in this report, is defined as a systematic correspondence between the
values of two variables from an exposure study, that is, questionnaire responses and
analytical measurements. This correspondence may or may not be statistically significant.
Similarities and differences in the results from the two approaches are discussed in section 4
(Results and Discussion).
3.2 Literature Review Methods
3.2.1 Sources
The literature review began with a search of several online citation indexes available at the
University of Nevada, Las Vegas library using keywords pertinent to this project's objective.
The following indexes were searched: MEDLINE (PubMed), Medline (FirstSearch),
Infotrieve, NTIS (National Technical Information Service), Wiley Interscience Journals,
Environmental Sciences And Pollution Management, and Toxline. The keywords survey,
questionnaire, children, pesticide, measurement, and biomonitoring were used in
combination to search the indexes. When an index limited the number of keywords used,
multiple searches were performed using subsets of the keyword list.
3-1
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Abstracts of over 100 citations from the search results were evaluated to determine if they fit
the project's focus. To be considered for the next level of review, a publication was required
to:
• address the pesticide exposure of children,
• have collected monitoring samples, preferably urine, and
• indicate the use of a survey or questionnaire in the study.
Two types of publications were excluded from further review: those describing studies of
infants or pre-natal situations and those that did not include an evaluation of relationships
between questions and monitoring measurements. The former type of publication was not
included because the exposure scenarios for children at such very young ages are somewhat
different from those for children of toddler age and up. Including such publications in this
review might add another layer of variability in evaluating the literature results.
Other sources of publications or pertinent results were also considered. They included:
• references cited in the relevant articles from the first round of searches,
• Masters' theses that were the basis for some of the relevant articles,
• Status Report on Biological Monitoring Research Relevant to Aggregate Exposure
Assessment under the Food Quality Protection Act (Fenske 1998), and
• Report on the Phase II NAFTA studies (USEPA 2002).
A more in-depth review of the publications considered potentially pertinent was performed
with an adjusted set of criteria. To be included in the next round, a publication was required
to:
• study pesticide exposure,
• describe relationships between questions and measurements from monitoring
samples, and
• include children as part of the population studied.
These criteria expanded the base of articles with studies of children and adults for potential
take-home exposure, while narrowing the list of pertinent articles to those that evaluated
relationships. No limitation was placed on the pesticides considered; however, most of the
relevant articles described organophosphorous (OP) pesticides because they are generally the
most toxic of the pesticides. Although the primary interest was in biomarker data,
relationships for any medium were noted.
Based on the second level of review, the publications were sorted into two groups: relevant
and not applicable. The relevant publications were the basis for identifying the relationships
to be reviewed (Table A.l). The rest of the publications were considered not applicable to
this project's objective (Table A.2).
3-2
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
3.2.2 A Database of Relationships
A simple database was created using MS Excel to track the relationships noted in the set of
relevant publications and to facilitate presentation of the relationships as tables for this
report. For each relationship in the database, the following information was recorded:
• citation abbreviation,
• question,
• medium of the measurement,
• chemical or metabolite measured,
• type of statistical analysis performed to evaluate the relationship,
• statistical significance of the analysis,
• p-value for the statistical analysis (as available),
• study sub-population included in the analysis,
• groups compared in the analysis (depending on the type of analysis performed),
• descriptive statistics or parameters produced by the analysis (as available),
• descriptors for the chemical measurements, such as log transformed, adjusted for
creatinine, etc., and
• comments.
An attempt was made to extract the maximum amount of information from each publication
for the database. Any of the very few instances of interpretation or assumptions are noted in
Appendix D.
To evaluate the track record of potentially effective questions for future studies, this report
includes relationships that are both statistically significant and non-significant. Several
publications listed or alluded to questions that were asked in the study interviews, but their
relationships with a measurement were not addressed in the publication. Phone or email
contact was made with the respective principal author to determine the status of the missing
relationship descriptions while recognizing the boundaries of unpublished research. In most
cases, it was determined that the relationships were excluded by the authors because the
relationships were not significant, were never analyzed, were analyzed and reported in
another article, or were to be reported in future publications. Information about these
unaddressed relationships was not included in the database.
Lastly, to facilitate the presentation of the literature review, several levels of organization
were added to the database. The questions were first grouped into 14 categories, such as
residential pesticide use, dietary habits, and household occupation. These categories were
then grouped into three super categories of risk factors: source, behavior, and other. An
abbreviated version of the question was assigned to each relationship to allow questions with
similar intent, but slightly different phrasing, to be presented together.
The relationships extracted from the relevant publications are presented in Appendices B, C,
and D. An evaluation of the questions' usefulness in differentiating levels of pesticide
exposure in children is presented in section 4.2 (Results and Discussion) and section 2
(Summary and Recommendations).
3-3
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
3.3 Children's Pesticide Exposure Study
A recent study of children's exposure to pesticides from Phase II of the Pesticide Exposure
and Health Effects on Children Initiative (Section 1.6) was also considered. A report (CDC
2002) describing the study and its evaluation of predefined hypotheses was reviewed in a
manner similar to the literature review. Data from the study were also made available for
exploratory analysis to determine if relationships, other than those predefined by the study's
objective, might surface.
3.3.1 Background
The U. S. Centers for Disease Control and Prevention (CDC), specifically the Health Studies
Branch, and the Toxicology Branch of the National Center for Environmental Health, in
collaboration with the U. S. Environmental Protection Agency (U.S. EPA), and the Arizona
Department of Health Services, conducted a study of pesticide exposure in children living in
Yuma County, Arizona. The Children's Pesticide Exposure Study is one of the studies
funded by the Environmental Health Working Group in the Border 2012 Program (USEPA
2004b), through the Pesticide Exposure and Health Effects on Children Initiative, to assess
the association of health outcomes in children with chronic exposure to pesticides. The study
collected objective measures of pesticide exposure in the children to help determine the need
for mitigation and prevention strategies for children and families living near the border. Its
objective was to determine the impact of living, or attending school, near pesticide-treated
fields on children's exposure to organophosphorous (OP) pesticides. Subsequently, this study
will be referred to as the Yuma Study.
In cooperation with eight local schools in the study area, families who were permanent
residents of the area with children in kindergarten or first grade were self-selected to
participate in a study conducted from October 1999 through February 2000. Promatores, that
is, lay health-care workers from a local non-government agency in Yuma, recruited a
convenience sample of participants by sending informational flyers home with children in
kindergarten and first grade, by approaching parents at Women, Infant and Children (WIC)
clinics, and by referrals from other participants. The data collection was performed during a
time period when large quantities of OP pesticides were expected to be applied to crops. The
promatores performed the data collection including the administration of a questionnaire
regarding characteristics and practices of the family and principal participant child. A urine
sample was collected from each of these children (principal participants) and from any
sibling in the household between the ages of 2 and 11 years. The urine samples were
measured for the six most common dialkylphosphate (DAP) metabolites associated with OP
pesticides. A dust sample was also collected from each household, and from classrooms,
with principal participants. These samples were measured for specific OP, organochlorine,
pyrethroid, and carbamate pesticides.
The study included 152 children as principal participants and 127 siblings. A total of 244
urine samples were available for analysis. Dust samples were available from 152 households
and from 25 kindergarten and first-grade classrooms in six of the participating schools. The
3-4
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
study analyzed the urine, dust, and questionnaire data to describe levels of pesticide
exposure, and to identify any associations between questionnaire data and pesticide exposure
levels.
The study report (CDC 2002) describes the statistical analysis approach that evaluated the
predefined hypotheses between children's pesticide exposure and risk factors like distance
from agricultural fields. Most of the analyses in this study were performed on urine
measurements for the principal participant and the siblings in each household, included
measures of intra-household correlation, and compared the measurements of principal
participants with their siblings. Geometric means and 95% confidence intervals were
calculated for variables that were not normally distributed. When considering the
relationships between risk factors and measures of pesticide metabolites, regression models
on log-transformed concentrations, controlling for intra-house correlation, were used. These
relationships were evaluated for metabolite concentrations adjusted, and not adjusted, for
creatinine. Regression models and Spearman correlations evaluated associations between the
concentrations of the urinary metabolites, adjusted and unadjusted for creatinine, and
household or school dust.
As a supplement to the initial findings in the Yuma Study report (CDC 2002), the study's
data were also evaluated using a data mining approach. Data mining describes an analysis
approach that searches through data for relationships that may or may not be defined a priori.
This process is exploratory in nature in comparison to a confirmatory analysis that is
interested in determining whether a proposed relationship adequately explains the observed
set of data (Hand 1999). The data mining approach used in this project focused on
identifying relationships that would be useful in classifying children by their OP pesticide
exposure level, with a specific interest in being able to identify children with high or low
exposure levels. The first stage of this approach prepared the data for analysis, the second
stage reviewed basic relationships in the data, and the third stage performed classification
type analyses. The data manipulation and analysis steps were carried out with SPSS versions
11.5 and 12.0 (SPSS, Inc., Chicago, IL), and S-Plus version 6 (Insightful, Inc., Seattle WA).
3.3.2 Stage 1 - Data Preparation
The Yuma Study data were reviewed to determine the types of analyses to be performed.
Adjustments were made to the data only to facilitate analyses and not to change the intent of
any responses. These adjustments included changes in data formats, the addition of code
values to describe certain situations, the creation of additional variables based on the original
data, and the identification of subgroups within the study to be used for the analyses. Steps
were taken to assure the quality of any changes made to the data and for any additional
variables created. A general description of the adjustments made to the Yuma Study data is
described in Appendix G.
After making these adjustments, the questions from the study were reviewed to determine
which would be used in the analyses. Questions that expanded on an "other" response, or
that were open-ended questions, such as "type of pesticide used in the field," were excluded
from analysis. In Table 3.1, the "Type" column denotes whether a variable was originally
3-5
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
used as a question in the Yuma Study, or whether a variable was created for the additional
analyses in this report. (A variable is defined as the set of participant responses for a
specific question that are assigned codes for analysis.) The "Brief Description" is used to
identify the questions in subsequent tables. The "Extended Description" includes the full
statement of the question. "Principal child" and "participant" are used interchangeably in
Table 3.1 to refer to the principal participant child.
Table 3.1 Questionnaire Variables from the Yuma Study Used in Data Mining Analyses, Sorted by the
Questionnaire Order
Type3
Name
Brief Description
Extended Description
Original
AGE
Age of principal child
Age of principal child calculated from date
of birth
Original
SEX
Child's gender
Gender of principal child
Original
HEIGHT
Child's height (inches)
Measurement of principal child's height
without shoes (inches)
Original
WEIGHT
Child's weight (lbs)
Measurement of principal child's weight
without shoes or other heavy articles (lbs)
Original
SCHOOL
Child's school
School where principal child attends
Original
GRADE
Child's grade
What grade is the principal child in?
Original
ETHNIC
Child's ethnic and racial background
Child's ethnic and racial background
Original
LIVEYEAR
Number of years child lives at this address
Number of years child lives at this address
Original
LIVEAREA
Children/respondent lives in area part-time
Children/respondent live in area < 10
months/year
Original
PEOPLIVE
Number of people in household including
participant
Number of people in household including
participant
Original
YOUNGSIB
Number of children in household < 11
years old
Number of additional children in household
> 2 years and < 11 years old?
Original
CHEMINHS
Pesticides used inside home last month?
Were chemicals to control insects used
inside the house during the last month?
Original
WHOCHEMI
Who applied pesticides inside the house?
Who applied chemicals inside the house?
Original
LIVINGRM
Living room treated with pesticides?
Was living room treated with pesticides?
Original
FAMILYRM
Family room treated with pesticides?
Was family room treated with pesticides?
Original
DININGRM
Dining room treated with pesticides?
Was dining room treated with pesticides?
Original
KITCHEN
Kitchen treated with pesticides?
Was kitchen treated with pesticides?
Original
BATHROOM
Bathroom treated with pesticides?
Was bathroom treated with pesticides?
Original
BEDROOM
Bedroom treated with pesticides?
Was bedroom treated with pesticides?
Original
CHILDBED
Child's bedroom treated with pesticides?
Was child's bedroom treated with
pesticides?
Original
BASEMENT
Basement treated with pesticides?
Was basement treated with pesticides?
Additional
NRMSPRYD
Number of rooms sprayed last month
Number of rooms in house sprayed with
pesticides in past month
3-6
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Type3
Name
Brief Description
Extended Description
Original
OTHERRM
Other rooms treated with pesticides?
Were other rooms in the house treated
with pesticide?
Original
OFTCHEMI
How often is home treated for pests?
How often is participant's home treated for
pests?
Original
CHEMOUTH
Pesticides used outside home last month?
Were chemicals to control insects used on
the exterior or foundation of the house
during the last month?
Original
WHOCHEMO
Who applied pesticides outside house?
Who applied chemicals outside house?
Original
FARFIELD
Distance between home and agricultural
field
How far is participant's home from a field
where crops are grown?
Original
CLOSEAPP
Distance between home and nearest
application of pesticides
In past month, how close to participant's
home was the nearest application of
agricultural or gardening chemicals?
Original
GPS
Distance between home and field using
GPS
How far is participant's home from a field
where crops are grown?
Original
WHEEL
Distance between home and field - rotary
wheel
Distance from home to field measured
with rotary wheel - categories
Original
HOWCHEMO
How pesticides were applied to fields
How were agricultural chemicals applied to
field close to participant's home?
Original
WHNCHEMO
Last time field treated with pesticides?
When was the last time the field was
sprayed or treated with pesticides?
Original
VEGGIES
How often child eats local fresh fruit/veg?
During the year, how often does principal
child eat locally grown fresh fruits or
vegetables?
Original
WASHVEGI
How often wash local fresh fruit/veg before
eating?
How often are the locally grown fresh fruits
and vegetables washed before they are
eaten?
Original
HOURAWAY
Number hours/wk child not at home
During school year, about how many
hours per week does principal child spend
away from home?
Additional
CHLDTM1
Child spends time in another home?
Principal child routinely spends time away
from home - in another home
Additional
CHLDTM2
Child spends time at day care center?
Principal child routinely spends time away
from home - at day care center
Additional
CHLDTM3
Child spends time at school?
Principal child routinely spends time away
from home - at school
Additional
CHLDTM4
Child spends time at sport event?
Principal child routinely spends time away
from home - at sport event
Additional
CHLDTM5
Child spends time playing in field?
Principal child routinely spends time away
from home - playing in field
Additional
CHLDTM6
Child spends time playing in irrigation
water?
Principal child routinely spends time away
from home - playing in irrigation water
Additional
CHLDTM7
Child spends time playing outside?
Principal child routinely spends time away
from home - playing outside
Original
WHERTIME
Room where child spends most awake
time
Room where principal child spends most
of their awake time
Original
SPRAYFLD
Child in yard when fields sprayed or
dusted?
Does principal child play outside in the
yard when the fields are sprayed or
dusted?
Original
WATERSR1
Drinking water source - public/commercial
Source of drinking water in participant's
home is public/commercial
3-7
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Type3
Name
Brief Description
Extended Description
Original
WATERSR2
Drinking water source - private well
Source of drinking water in participant's
home is private well
Original
WATERSR3
Drinking water source - cistern
Source of drinking water in participant's
home is cistern
Original
DADWORK
Is the father currently employed?
Is the father currently employed?
Original
NCATWRKD
Father's occupation - categories
Father's occupation - categories
Additional
DADPEST
Are pesticides used where father works?
Are pesticides used where father works? -
categories
Additional
DADCON2
Father's occupation - location and
pesticide use
Does father work indoors or outdoors and
with or without pesticides?
Original
MOMWORK
Mother now employed (not as housewife)?
Is the mother currently employed?
Original
NCATWRKM
Mother's occupation - categories
Mother's occupation - categories
Additional
MOMPEST
Are pesticides used where mother works?
Are pesticides used where mother works?
- categories
Additional
MOMCON2
Mother's occupation - location and
pesticide use
Does mother work indoors or outdoors and
with or without pesticides?
Original
ADLTPEST
Non-parent in home works where
pesticides used?
Is there another person living in the house
(other than parent) who works in a place
where pesticides are used?
Original
ADTPSWK
Non-parent in home works where
pesticides used?
Any adult in household works where
pesticides used?
Original
NUMADLTS
Number of additional adults in home
Number of non-parent adults in home
working with pesticides
Original
CHILDFLD
Child worked in fields last month?
Has principal child been to the work
field(s) during past month?
Original
WHENFILD
Last time child was in work field
When was the last time principal child was
in the work field?
Additional
WHERMD1
Family med care at private medical clinic
Where principal child's family receives
medical care - private medical clinic
Additional
WHERMD2
Family med care at health dept clinic
Where principal child's family receives
medical care - local health department
clinic
Additional
WHERMD3
Family med care at other med clinic
Where principal child's family receives
medical care - other medical clinic
Additional
WHERMD4
Family med care in Mexico
Where principal child's family receives
medical care - Mexico
Additional
WHERMD5
No access to medical care
Where principal child's family receives
medical care - no access
Additional
WHERMD6
Family med care at other place
Where principal child's family receives
medical care - at other facility
Additional
WHERMD7
Family med care - do not know
Where principal child's family receives
medical care - do not know
Original
POISON
Anyone treated for pesticide poison?
Has anyone in the household been treated
for pesticide poisoning in past year?
Original
HOWCHILD
Child's health in general
Description of principal child's health in
general
3-8
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Type3
Name
Brief Description
Extended Description
Original
LICE
Child treated for head lice past six
months?
Has principal child been treated for head
lice in past six months?
Original
INSURED
Is child covered by medical insurance?
Is principal child covered by medical
insurance?
a Original variables existed in the data set provided from the Yuma Study. Additional variables were created based on the
original variables.
Code values were reassigned for non-responses, conditional questions, and to create an
underlying continuum of potential exposure. See Appendix G for information on coding
schemes and additional variables created.
The chemical/metabolite measurements for the urine, house dust, and school dust samples
that were used in the data mining analyses were analyzed as molar-weighted sums (Table
3.2). Appendix F includes a list of all the chemicals and metabolites measured in the Yuma
Study, the specific chemicals included in each sum, and an example of how the sums are
calculated. For the urinary metabolite sums, the log of the sum was used as the dependent
variable.
Table 3.2 Analytical Measurement Variables from the Yuma Study Used in Data Mining Analyses
Name
Description3
Urine from Child
LWETHSUM
Log of weighted sum of DEP, DETP, and DEDTP (adjusted for creatinine)b
LWMETHSM
Log of weighted sum of DMP, DMTP, and DMDTP (adjusted for creatinine)c
Household Dust
WCHDNSUM
Weighted sum of alpha-chlordane and gamma-chlordane
WCHLPYRF
Weighted chlorpyrifos
WCYPRMET
Weighted cy-permethrin
WDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and 4,4'DDT
WDIAZNON
Weighted diazinon
WDUSTBAL
Weighted sum of dust analytes except chlorpyrifos, diazinon, permethrins, and
o-phenylphenol
WDUSTSUM
Weighted sum of all dust analytes
WOPBAL
Weighted sum of OP pesticides except chlorpyrifos and diazinon
WOPHNYLP
Weighted o-phenylphenol
WOPSUM
Weighted sum of OP pesticides
WPERMSUM
Weighted sum of cis-permethrin and trans-permethrin
3-9
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Name
Description3
School Dust
SWCHDNSM
Weighted sum of alpha-chlordane and gamma-chlordane
SWCHLPYR
Weighted chlorpyrifos
SWCYPRME
Weighted cy-permethrin
SWDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and 4,4'DDT
SWDIAZNO
Weighted diazinon
SWDSTBAL
Weighted sum of dust analytes except chlorpyrifos, diazinon, permethrins, and
o-phenylphenol
SWDUSTSM
Weighted sum of all dust analytes
SWOPBAL
Weighted sum of OP pesticides except chlorpyrifos and diazinon
SWOPHNYL
Weighted o-phenylphenol
SWOPSUM
Weighted sum of OP pesticides
SWPERMSM
Weighted sum of cis-permethrin and trans-permethrin
a See Appendix F for definition of weighted sums.
bDEP- diethylphosphate, DETP - diethylthiophosphate, DEDTP - diethyldithiophosphate
c DMP - dimethylphosphate, DMTP - dimethylthiophosphate, DMDTP - dimethyldithiophosphate.
3.3.3 Stage 2 - Review of Basic Relationships
In Stage 2, relationships between questionnaire and analytical measurement data were
reviewed. This stage was exploratory rather than inferential and helped determine the sets of
variables to be analyzed in Stage 3. As the basic analyses were performed and seeming
inconsistencies in the relationships appeared, the data were reviewed. Stage 2 included
evaluating simple indicators of high exposure levels, identifying relationships between
questionnaire variables, and refining the group of participants to be included in the Stage 3
analyses.
3.3.4 Stage 3 - Classification Approach
The Stage 3 analyses were performed using the data mining technique Classification and
Regression Trees (CART). This method divides the study population into subsets of
participants where the between-subset variability of the dependent variable (e.g.,
LWETHSUM from Table 3.2) is maximized and the within-subset variability is minimized.
The predictors or independent variables (questions and dust measurements in this study) can
be nominal, ordinal, or continuous in nature and are the basis for defining the subsets. CART
attempts to identify a model of predictors and their interactions that optimally classify
subjects by the dependent variable, in this case, the child's measured exposure level for a
specific metabolite. The output from CART is a classification map, or tree, that describes the
subsets of the study population in terms of the dependent variable values and provides
characteristics of the subsets in terms of the predictors and the predictors' values. This is
3-10
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
analogous to a regression equation without a functional form. The predictors selected in the
CART analyses can help differentiate children's pesticide exposure levels as measured by the
molar-weighted DAP sums.
Details of the methodology for the data mining approach can be found in Appendix G.
3-11
August 2005
-------
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
4.0 RESULTS AND DISCUSSION
4.1 Introduction
The objective of this project is to evaluate questions that are potential indicators of pesticide
metabolite levels in children's urine and to identify the more useful questions as input for
future study design. The evaluation reviewed relationships between questions and exposure
measurements under two approaches:
• A literature review of previous exposure studies to summarize the existence of such
quantitative and qualitative relationships, and
• An analysis of a recent children's pesticide exposure study in Yuma, Arizona, which
included questionnaires and measurements.
As work proceeded, it became obvious that environmental measurements were also useful in
completing the link between questions and metabolite levels. The results from each approach
are described and discussed in this section. A summary of the results is also presented in
section 2.
4.2 Literature Review
Most of the pertinent studies on children's exposure to pesticides began in the 1990's, a
much shorter period of study than for adults. The publications reviewed were those
published through early 2003. Multiple searches of the literature resulted in over 100
citations (section 3.2). The abstracts and publications were reviewed against the first set of
criteria, which required a pertinent publication to:
• address the pesticide exposure of children,
• have collected monitoring samples, preferably urine, and
• indicate the use of a survey or questionnaire in the study.
Of the publications meeting the first set of criteria, 64 were reviewed against a second set of
criteria, which required a pertinent publication to:
• study pesticide exposure,
• describe relationships between questions and measurements from monitoring
samples, and
• include children as part of the population studied.
These criteria categorized 20 of the 64 publications as "relevant" (Table A.l) and the
remaining 44 publications as "not applicable" (Table A.2). Twenty publications of the
relevant publications were peer-reviewed journal articles. Four of the twenty relevant
publications were subsequently categorized as not applicable. Bradman (1997), Mills (2001),
O'Rourke (2000), and Thompson (2003) included study designs that were pertinent to this
review; however, either the particular publication did not include the required relationship
information, or the study was a pilot that contained only a small number of subjects. The
4-1
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Bradman (1997) and Mills (2001) publications had sample sizes less than 10. The O'Rourke
(2000) and Thompson (2003) publications included information on pesticide metabolites in
children's urine, but did not attempt to draw associations between this information and
questionnaire data. Four Masters' theses were subsequently added to the relevant
publications list. Two of the theses produced journal articles that are among the relevant
publications, and were included because they provided details not available in the articles:
the Carrel (1996) thesis was published as Loewenherz (1997), and was expanded upon in Lu
(2000); the Koch (1999) thesis was published as Koch (2002). The other two Masters' theses
were included because their findings have not been published in the peer-reviewed literature:
the Grossman (2001) thesis was based on the same field study reported on by Curl (2002)
and Thompson (2003), but the analysis conducted by Grossman was not included in either of
these publications. The Krinsley (1998) thesis work has not been published elsewhere.
Future references to the relevant publications list will denote the 20 publications from which
relationship information was extracted. These 20 publications covered aspects of 14
different exposure studies. Appendix A includes references for all publications reviewed.
Results from the literature review are presented as information on each "relationship"
between a question and exposure measurement described in the relevant publications.
Relationship, as used in this report, is defined as a systematic correspondence between the
values of two variables from an exposure study, that is, questionnaire responses and
analytical measurements. This correspondence may or may not be statistically significant.
As the review of the publications progressed, the scope of relationships considered for this
report was expanded to include relationships with environmental media as well as biomarkers
to enhance the information base relating to potential exposure pathways. When reviewing
individual relationships to ensure comparability, the reader should be cognizant of study and
analysis particulars. Unless stated otherwise in the Results and Discussion section,
significant and statistically significant will be interchangeable.
In the process of extracting pertinent relationship information, each publication was reviewed
several times. Since the publications were not consistent in the manner or level of
information provided for the relationships, a structure was developed to capture the variety of
information available. The objective of this review process was to ensure that the
relationship information was extracted correctly, as the authors intended it to be interpreted,
and to make few if any assumptions. In a few instances, authors were contacted to clarify
information presented in the publications.
4.2.1 Publications Reviewed for Relationships
The studies on which the 20 relevant publications are based were assigned abbreviated
citation references to be used in this and other sections of the report. The studies as used in
the publications are briefly described with information about the study's location, population,
media, pesticides measured, and types of questions asked (Table 4.2.1). Some studies
generated more than one publication. Related publications and a study number are included
for those instances.
4-2
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.1 Brief Descriptions of the Studies Included in the 20 Relevant Publications
Citation
Reference3
Location
Population
Media
Studied
Pesticides Studied13
Type of Questions Asked
Related
Work
Ad gate
2001
STUDY 1c
Minnesota: Urban
(Minneapolis/St.
Paul) and non-
urban areas (Rice
and Goodhue
counties)
102 children 3-13 years old —
Preferences were for
households with more
frequent pesticide use, more
than 1 eligible child, use of a
private well in non-urban
areas, children having
greater potential for recent
exposure to target pesticides.
Urine
Metabolites: 1-naphthol,
atrazine mercapturate,
malathion dicarboxylic
acid, 3,5,6-trichloro-2-
pyridinol.
Characteristics of the participating child
and housing, usual frequency of activities
over a period of a month or year, detailed
(daily) time and location information of
activities for the child, and information on
less than daily activities during the
monitoring period.
Sexton 2003
Aprea 2000
STUDY 2
Tuscany, Italy
195 children 6-7 years old —
Children were enrolled in
elementary schools in Siena,
Italy, which does not have
major industries. Population
is employed mostly at banks,
hospitals, universities, or as
shopkeepers, and
professionals.
Urine
Metabolites: DMP,
DMTP, DMDTP, DEP,
DETP, DEDTP
Lifestyle and dietary habits — sex, date of
birth, weight, height, school, class, father
and mother's occupations, illness and
hospitalization of child, existence of
garden or vegetable garden, existence of
ornamental plants in house, purchase of
cut flowers for the house, domestic
animals in house, use of pesticides inside
or outside house, food and drink ingested
day before urine sample, and ate lunch at
school. Diet, parent's occupation, height,
weight and height/weight ratio were used
for qualitative classification of population.
4-3
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Citation
Reference3
Location
Population
Media
Studied
Pesticides Studied13
Type of Questions Asked
Related
Work
Azaroff
1999
STUDY 3
Rural El Salvador
103 farmer households and
family members at least 8
years old — Families were
recruited from five
agricultural communities who
lived there during planting
season. Household members
who lived in the home during
planting season and were
able to answer questions
were included.
Urine
Metabolites: DMP,
DMTP, DMDTP, DEP,
DETP, DEDTP
Application of pesticides to crops, age
and sex of household members, laundry
practices, field work, and pesticide use for
the house.
Carrel
1996d
STUDY 4
Washington
(Douglas and
Chelan counties)
88 children no more than 6
years old — Two family types
were selected: pesticide
applicator in family living
near sprayed orchard, and
family with no pesticide
applicator living further from
orchard. One child selected
per family.
Urine
Metabolites: DMP,
DMTP, DMDTP
Occupational and residential pesticide
use, cleaning activities, laundry practices,
protective equipment use, proximity to
spray sites, and child activity.
Published as
Loewenherz
1997;
expanded in
Lu 2000;
analysis of
diethyl
metabolites in
Fenske 2002
4-4
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Citation
Reference3
Location
Population
Media
Studied
Pesticides Studied13
Type of Questions Asked
Related
Work
Curl 2003
STUDY 5
Seattle,
Washington
43 children 2-5 years old -
Children were recruited
based on whether their juice,
fresh fruit, and fresh
vegetable consumption was
either nearly all organic or
nearly all conventional.
Urine
Metabolites: DMP,
DMTP, DMDTP, DEP,
DETP
Age and weight of child, parental age,
and occupation, annual family income,
home ownership, length of time at the
current residence, housekeeping
practices, residential pesticide use (in the
home, on the home structure, in the
garden, on the lawn, and on pets), time
since last pesticide application and who
applied the pesticide, child behaviors:
thumb-sucking, hand washing, hand-to-
mouth activity, and amount of time spent
outside of home. Food diary for child with
type and amount of food and beverage
and whether each item was organic or
not. Food diary was used for
classification of child's diet.
Curl 2002
STUDY 6
Washington
218 farm-worker households
in 24 agricultural
communities — One farm-
worker actively involved in
field work or pesticide
application and one 2-6 year
old child were sampled from
each household.
Urine, dust
Pesticides:
azinphosmethyl,
malathion, methyl
parathion, phosmet,
chlorpyrifos, diazinon
Metabolites: DMP,
DMTP, DMDTP, DEP,
DETP
Types of agricultural job tasks,
occupational pesticide exposure,
perceived health effects of pesticide
exposure, occupation and personal
protective practices, and demographics.
Grossman
2001
Fenske
2002
STUDY 4
Central
Washington
75 homes and 109 children
up-to-6-years old — Three
family types were selected:
pesticide applicator in family
living near sprayed orchard,
pesticide applicator in family
living further from orchard,
and no pesticide applicator in
family living further from
orchard.
Urine,
dust,
dermal
wipe
Pesticides: chlorpyrifos,
ethyl parathion.
Metabolites: 3,5,6-
trichloro-2-pyridinol, 4-
nitrophenol.
Occupational and residential pesticide
use, hygienic and housekeeping
practices, child behavior and activity, and
proximity of home to pesticide-treated
fruit orchard.
Carrel 1996,
Lu 2000,
Loewenherz
1997
4-5
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Citation
Reference3
Location
Population
Media
Studied
Pesticides Studied13
Type of Questions Asked
Related
Work
Grossman
2001d
STUDY 6
Lower Yakima
Valley,
Washington
148 households with children
2-6 years old — Hispanic
farm-worker households
were selected from For
Healthy Kids, a community
intervention study of take-
home pesticide exposures.
Dust
Pesticides:
azinphosmethyl
Sociodemographic characteristics and
acculturation, agricultural tasks,
knowledge about pesticides and related
health effects, perceived exposure to
pesticides, workplace facilities, and work
and home practices related to pesticide
exposure.
Curl 2002
Koch 2002
STUDY 7
Central
Washington
44 children 2-5 years old —
Households were recruited
from a Women, Infants and
Children (WIC) clinic in a fruit
tree production region. One
child per family was selected.
Urine
Metabolites: DMP,
DMTP, DMDTP, DEP,
DETP
Characteristics of the study child, parental
occupations, household pesticide use,
and children's activities
Koch 1999
Koch 1999d
STUDY 7
Central
Washington
44 children 2-5 years old —
Households were recruited
from a Women, Infants and
Children (WIC) clinic in a fruit
tree production region. One
child per family was selected.
Urine
Metabolites: DMP,
DMTP, DMDTP, DEP,
DETP
Characteristics of the study child, parental
occupations, household pesticide use,
and children's activities
Koch 2002
Krinsley
1998d,e
STUDY 8
Arizona, including
US-Mexico border
179 households that were
full-time Arizona residents
and were a subset of the
Arizona NHEXAS study.
The focus was on high risk
subgroups of minorities,
children, and US-Mexico
border residents.
Urine
Pesticides: chlorpyrifos
Metabolites:
3,5,6-trichloro-2-pyridinol
Health status, occupation, pesticide use
characteristics, home characteristics,
demographic information, behavior, time-
activity, and daily diet diaries.
4-6
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Citation
Reference3
Location
Population
Media
Studied
Pesticides Studied13
Type of Questions Asked
Related
Work
Loewenherz
1997
STUDY 4
Washington
(Douglas and
Chelan counties)
88 children no more than 6
years old — Two family types
were selected: pesticide
applicator in family living
near sprayed orchard, and
family with no pesticide
applicator living further from
orchard. One child selected
per family.
Urine
Metabolites: DMP,
DMTP, DMDTP
Occupational and residential pesticide
use, cleaning activities, laundry practices,
protective equipment use, proximity to
spray sites, and child activity.
Carrel 1996,
Fenske 2002,
Lu 2000
Lu 2001
STUDY 9
Seattle,
Washington
110 children 2-5 years old —
Families recruited from clinic
and outpatient waiting rooms
in two communities — an
urban, densely-populated
one with lower to middle
income families, and a
suburban one with middle to
upper income families. One
focus child was selected from
each family.
Urine
Metabolites: DMP,
DMTP, DMDTP, DEP,
DETP, DEDTP
Characteristics of the child, parental
occupation and family income level, home
ownership status, length of time at current
residence, housekeeping practices,
residential pesticide use regarding pets,
lawn or vegetable/flower garden,
professional application of pesticides in
last 6 months, which pesticide products
were applied, and child's activities and
behaviors.
Lu 2000
STUDY 4
Central
Washington
109 children from 9 months
to 6 years old from 76
households — Three family
types were selected:
pesticide applicator in family
living near pesticide-treated
orchard, farm-worker in
family living near pesticide-
treated orchard, and no
pesticide applicator in family
living > .25 mi from pesticide-
treated orchard.
Urine,
dust,
dermal
wipe
Pesticides:
azinphosmethyl,
phosmet. Metabolites:
DMP, DMTP, DMDTP.
Occupational and residential pesticide
use, hygienic and housekeeping
practices, child behavior and activity, and
proximity of home to pesticide-treated
fruit orchard.
Carrel 1996,
Fenske 2002,
Loewenherz
1997
4-7
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Citation
Reference3
Location
Population
Media
Studied
Pesticides Studied13
Type of Questions Asked
Related
Work
McCauley
2003
STUDY 10
Hood River,
Oregon
24 fruit-tree orchard
agricultural families with at
least one adult member
working in an orchard full
time, and with at least one 0-
7 year old child; four control
families.
Dust
Pesticides:
azinphosmethyl,
chlorpyrifos, malathion,
phosmet, diazinon,
parathion
Demographics, agricultural work practices
of all adult family members residing in the
home, self-reported protective practices
at work and upon coming home,
residential pesticide use, a household
pesticide inventory, land use and
proximal crop information, child play
locations, and precautions taken by family
during pesticide spraying events.
McCauley
2001a
STUDY 11
Oregon
(Washington and
Hood River
counties)
96 families with preschool
children — Families were
recruited from children
enrolled in Migrant Head
Start centers.
Dust
Pesticides:
azinphosmethyl
Demographics, agricultural work practices
of all adult family members residing in the
home, self-reported protective practices
at work and upon coming home,
residential pesticide use, a household
pesticide inventory, and land use and
proximal crop information.
Royster
2002
STUDY 12
Imperial County,
California
20 children 12-18 months old
- Children were recruited
during well-child visits at
clinics, when due for their
first MMR (measles, mumps,
rubella) vaccination, without
certain health issues.
Urine
Metabolites: DMP,
DMTP, DMDTP, DEP,
DETP, DEDTP
Family's occupational pesticide exposure,
the child's and family's health histories,
pesticide usage, proximity to agricultural
fields, location of residence, source of
drinking water, history of smoking within
household, and demographic
characteristics.
4-8
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Citation
Reference3
Location
Population
Media
Studied
Pesticides Studied13
Type of Questions Asked
Related
Work
Sexton
2003
STUDY 1
Minnesota: Urban
(Minneapolis/St.
Paul) and non-
urban (Rice and
Goodhue
counties) areas
102 children 3-13 year-olds —
Preferences were for
households with more
frequent pesticide use, more
than 1 eligible child, use of a
private well in non-urban
areas, children having
greater potential for recent
exposure to target pesticides.
Urine,
dust, hand
rinse, soil
Pesticides: chlorpyrifos,
diazinon, malathion,
atrazine. Metabolites:
malathion dicarboxylic
acid, 3,5,6-trichloro-2-
pyridinol
Occupant characteristics, household
characteristics, household pesticide use
and occupant activities. Characteristics of
the participating child and housing, usual
frequency of activities over a period of a
month or year, detailed (daily) time and
location information and activities for the
child, and information on less than daily
activities during the monitoring period.
Ad gate 2001
Shalat 2003
STUDY 13
Rio Bravo, Texas
52 children 7-53 months old -
- 29 households were
selected from an agricultural
community on the US-Mexico
border.
Urine,
dust, hand
rinse, soil
Pesticides:
azinphosmethyl,
chlorpyrifos, demoton O,
demoton S, diazinon,
ethion, fenithrothion, ethyl
parathion, methyl
parathion. Metabolites:
DMP, DMTP, DMDTP,
DEP, DETP, DEDTP
Medical information, occupational
information, time/activity information,
children's hand-to-mouth activities, diet,
residential pesticide use, and pets or
animals in the household.
Simcox
1995
STUDY 14
Central
Washington
59 households with at least
one child 1-6 years old —
Households included
reference families, and
agricultural families where at
least one family member
living in the home was
employed as an orchardist,
field worker, and/or pesticide
applicator.
Dust, soil
Pesticides:
azinophosmethyl,
chlorpyrifos, ethyl
parathion, phosmet
Occupational pesticide use, residential
and agricultural pesticide use in past 6
months, proximity of home to orchards,
protective practices, and family hygiene
practices.
a See Table A-1 (Appendix A) for citations.
b DEP- diethylphosphate, DETP - diethylthiophosphate, DEDTP - diethyldithiophosphate
DMP - dimethylphosphate, DMTP - dimethylthiophosphate, DMDTP - dimethyldithiophosphate.
cSome studies generated multiple publications and are identified with the same study number.
4-9
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
d Masters theses related to publications in the initial relevant list.
e Data used for the Krinsley thesis are available at EPA's Human Exposure Database System (HEDS) web site at: http://vwvw.epa.gov/heds/index.htm under the NHEXAS
Arizona Study.
4-10
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
4.2.2 Description of Relationship Information
4.2.2.1 Content
Detailed information was extracted for each relationship to provide a useful reference
tool for diverse research needs. A simple database was created using MS Excel to capture
the relationship information presented in Sections 4.2.4, 4.2.5, and 4.2.6, and in
Appendices B, C, and D. The types of information included are descriptive, general
analysis, and statistical analysis (Table 4.2.2). The data fields under the analysis types of
information refer to the results of a statistical analysis or to the groups compared in the
statistical analysis.
Table 4.2.2 Information Extracted from Relevant Publications for Each Relationship
Type of Information
Data Fields3
Descriptive
Citation reference
Question asked
Sample medium
Chemical measured
Type of measurement, e.g., concentration or loading
Log transformation indicator
Subpopulation included in the analysis
Type of statistical analysis performed
Groups compared in the analysis, if relevant
Significance indicator for analysis
Comments about the chemical measurement
Analysis: General
p-value for the statistical analysis
p-value for model or predictor indicator
Comments about the analysis
Analysis: Statistics
Units of chemical measurement
Geometric mean
Geometric standard deviation
Median
Mean
Standard deviation
Percent detectable measurements
Number of subjects
Odds ratio (from logistic regression)
Confidence interval (95% level for either Odds Ratio or Beta)
4-11
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Type of Information
Data Fields3
Beta (coefficient from regression analysis)
R2 — square of multiple correlation coefficient from regression analysis
a Analysis refers to statistical analysis unless otherwise noted.
The available statistical parameters differ across publications and relationships.
Relationships were included in the database based on the type of relationship measured
and the types of information available. Two types of relationships were not included in
the database: relationships for which no information on a statistical analysis was provided
even though statistical parameters at a group level (e.g., means by age group) were
provided, and relationships where the analysis did not fit the project's objective (e.g., a
relationship between two questions). As much of the pertinent information as possible
was extracted for each relationship from the publications for inclusion in the database.
The descriptive fields in the database (Table 4.2.2), and the study information (Table
4.2.1) set the context for each relationship that was evaluated because a study's design or
the study subgroups compared in a statistical analysis can affect the significance of the
relationship between the question and the measurement. The general and statistical
analysis fields in the database allow for additional evaluation of a question's usefulness
in understanding exposure-related activities and their potential impact. For example,
knowing that a question showed a statistically significant relationship to a particular type
of measurement in two separate relationships provides one type of information. Also
knowing that in one of the relationships the median for group A was greater than the
median for group B, and in the other relationship the median for group B was greater than
the median for group A provides a different type of information. The inconsistent
relationship between the medians of the two groups is a cue for the reader to consider the
possibility of confounding factors (e.g., related to design) in the analyses or to recognize
that the usefulness of the question in predicting exposure level may not yet be adequately
proven. The database of information, as included in Appendices, B, C, and D, allows the
reader to review the relationships from these perspectives.
4.2.2.2 Organization
Each relationship was assigned a unique ID number that will be shown in the tables
describing the relationships. This ID number was used primarily for tracking and for
preparing the information in table format. Because the breadth of information extracted
for the relationships could not easily be presented in a single table format, the
relationship information was grouped into three types of tables, an overview, details, and
comments, which are included in Appendices B, C, and D, respectively. The ID number
can be used to match information between Appendices C and D.
Another level of organization for the relationships was introduced at the question level.
The publications differ in how, and to what extent, the questions are described. Some
provide the full question, and some provide an abbreviated or generalized description of
4-12
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
the question. To analyze their usefulness, question descriptions that at least implied the
same question are grouped together, and a question number, e.g., Q102, has been
assigned to each question phrasing for ease of reference in other tables. Thus, for each
relationship, an abbreviated question phrasing was assigned. For example, the
abbreviated question phrasing or description "inside treated" includes the following
questions:
• pesticide use inside
• pesticide used inside in past 6 months
• Was there indoor pesticide application in past 6 months?
• In the past 6 months were any chemicals for the control of fleas, roaches, ants or
other insects used inside this house/apartment?
Judgments were made regarding the level of abbreviated question descriptions to use.
Since, in a few instances, the specificity of the question may affect the comparison of
relationships, the description of the question as it is presented in the publication is
included in Appendix D with supplemental information about the relationship.
Reviewing the question descriptions allows the reader to evaluate the summary results
and any influence from the question groupings.
A third level of organization arranges the question phrasings (the abbreviated question
descriptions) into 14 question categories, and three risk factor groups for presentation and
discussion purposes (Table 4.2.3). These groupings provide the organization of
information and relationships in sections 4.2.4, 4.2.5, and 4.2.6, and in Appendices B, C,
and D.
Table 4.2.3 Distribution of Relationships Across Risk Factor Groups and Question Categories" of
Questions Used to Organize Sections 4.2.4,4.2.5,4.2.6, Appendix B, Appendix C, and
Appendix D
Risk Factor
Group
Question Category
Relationships
Description
#
Description
N
%
Source
1
Residential pesticide useb
100
17
Source
2
Household characteristics'3
73
12
Source
3
Residential sources (environmental measures)
13
2
Source
4
Household occupation
115
19
Source
5
Residential proximity to agricultural fields
72
12
Source
6
Residential location
14
2
Behavior
7
Subject's personal characteristics
78
13
Behavior
8
Child's behaviors
20
3
4-13
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Risk Factor
Group
Question Category
Relationships
Description
#
Description
N
%
Behavior
9
Dietary behaviors
16
3
Behavior
10
Family hygiene practices
81
13
Behavior
11
Smoking-related activities
4
1
Behavior
12
Work exposure/practices
4
1
Other
13
Related exposure levels
5
1
Other
14
Health
8
1
Total
603
100
a Based on counts in Appendix B tables.
bSee note in the following paragraph regarding relationships from Sexton (2003).
Sexton (2003) evaluated many relationships between questions in the residential pesticide
use and household characteristics categories, and measurements of atrazine, diazinon,
malathion, and chlorpyrifos in personal air, indoor air, outdoor air, solid food, beverages,
dust, soil, and urine under several statistical analysis scenarios. The majority of the
relationships analyzed were not statistically significant. Since these relationships
represented a large number of relationships for which no additional statistical information
is provided, they were not included in the relationship database. The reader should be
cognizant of this exclusion because it affects the percentages of statistically significant
versus non-significant results in subsequent summary tables. However, if included, the
large number of non-significant analyses would give more weight to the results from this
publication than perhaps reasonable. Each table affected by this exclusion will be
footnoted for the reader's awareness. All statistically significant relationships, and any
non-significant relationships specifically noted in the publications' tables or text, were
included in the database.
Six question categories had the largest number of relationships: residential pesticide use,
household characteristics, household occupation, residential proximity to agricultural
fields, subject's personal characteristics, and family hygiene practices. These categories
are likely considered the most appropriate questions for the type of exposures studied
because of expected or proven relationships. The fact that these categories have more
relationships than the other categories may provide more credence overall to predictors
selected from these categories. That judgment will specifically depend on the number of
relationships available for a given combination of question and chemical or metabolite.
Potential predictors from the other eight categories may also be useful, but have not been
tested enough to make adequate judgments.
The number of relationships for the urine and dust media overwhelms the number from
any other medium, with twice as many relationships for urine as for dust (Table 4.2.4).
4-14
August 2005
-------
I/O
o
o
(N
C/3
3
ao
Number of Relationships, by Medium
All
Media
100
73
CO
115
72
-3"
O
00
00
CD
00
"sr
-3"
LO
00
603
Solid
Food
CO
CO
SO
Soil
00
00
CO
Personal
Air
0.7
Outdoor
Air
CM
CM
CO
O
Indoor
Air
CM
-
CO
so
Dust
00
00
00
59
42
-
00
co
187
O
CO
Urine
00
54
LO
00
"3-
30
CO
o
00
00
CD
33
-
LO
00
396
65.7
Question Category
Description
Residential pesticide useb
Household characteristics'3
Residential sources (environmental measurements)
Household occupation
Residential proximity to agricultural fields
Residential location
Subject's personal characteristics
Child's behaviors
Dietary behaviors
Family hygiene practices
Smoking-related activities
Work exposure/practices
Related exposure levels
Health
Total
Percent of Total
--
CM
CO
LO
CD
00
a>
o
CM
CO
CO
o
o
CM
X
0
CO
E
o
4—
CO
Q_
_TO
0
TO
O)
0
CO
CN1
_0
_q
CO
_0
_Q
TO
-i—>
CD
X
"O
0 0
Q- c
Q- £
< E
CO
TO
O ^
° TO
C Q-
O 0
"D -C
0
CO 0
TO 0
CD CO
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
4.2.2.3 Assumptions and Caveats
Great care was taken and quality assurance was applied when extracting the relationship
information from the publications. Each publication was reviewed several times to identify
all hypothesized relationships noted. The ID# for each relationship was noted on a copy of
the publication for cross-checking and the information extracted was reviewed by more than
one person. The intent was to extract information without assumptions or interpretation.
There are a few situations where judgments or assumptions were made to provide as much,
and as consistent, information as possible (Appendix D). Relationships mentioned in a
publication's text, but not specifically included in tables, were also extracted.
In several publications, questions were identified as part of the study's interview process;
however, no results from analyzing relationships were mentioned. To determine the reason
for the absence, and to glean any additional information for the database, an author of the
publication was contacted. In most instances, the relationships in question were excluded
from the publication because they were not significant or were to be included in future
publications. In four instances, these contacts led to the inclusion of four Masters' theses as
relevant citations: Carrel (Loewenherz) (1996), Grossman (2001), Koch (1999), and Krinsley
(1998). Complete copies of the Grossman and Krinsley theses were reviewed. Parts of the
Koch and Carrel theses were made available for this report in response to specific questions
in the related publications.
Significance levels for the statistical analyses are reported in various ways, even within a
publication. For example, p-values may be specified as a value (p = 0.042) or as an interval
(p < 0.05). Sometimes the significance level is noted only as an indicator, that is, significant,
not significant, marginally significant, or as a trend. Since knowing the p-value rather than a
general indicator of significance allows the reader to make decisions based on their research
objectives, the p-values were added to the database as a separate field. In cases where the
significance indicator rather than the p-value is given, the publication was reviewed to
identify the p-value used as the critical value for statistical significance. In all of the
publications, the p-value for identifying significant relationships was 0.05. When the
marginally significant or trend indicator was noted, the critical values are either 0.10 or 0.20.
When no p-value is noted for a relationship, one was added based on the publication's
significance indicator. For example, if the publication's critical value for significance is p =
0.05, p > 0.05 was noted in the database for a relationship identified as not significant. If the
publication's critical value for being marginally significant was p = 0.10, p > 0.10 is noted in
the database for a relationship identified as not marginally significant.
The extent of inferences that can be made from the relationships presented must be taken into
account. Most of the studies conducted were convenience samples. Analyses from such
studies are descriptive of the particular group sampled, and may or may not generalize to
similar populations. The value of these relationships, however, is that they identify potential
trends that may exist in the populations. When evaluating the effectiveness of a question for
a particular research situation, it is also important to review and understand the similarities
and differences in the samples taken, the type of measurements taken and analyzed, and the
4-16
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
subgroups compared across the relationships available for the question. The reader may need
to review the pertinent publications for some of these details.
4.2.3 Description of Relationships Presented
The amount and variety of information available across the relationships made it difficult to
create a presentation format that was easy to read without being burdensome in other
respects. The selected formats are more compact, but require more introduction to, and
scrutiny by, the reader. This section briefly describes the content and organization of the
tables in the upcoming sections, and gives an introduction to the related tables in Appendices
B, C, and D.
The question categories (Table 4.2.3) provide the framework for the presentation and
discussion of the results from the literature review. Questions for each of the three risk
factors are discussed in separate sections. Source relationships are presented in section 4.2.4,
behavior relationships are presented in section 4.2.5, and other relationships are included in
section 4.2.6. The questions included under each risk factor and question category can be
found in Appendix E. For each question category, three types of summary tables are
presented as part of the evaluation and discussion. These tables describe the effectiveness of
the questions in differentiating exposure levels by describing the extent of statistically
significant relationships for each question and metabolite/chemical combination.
The first table type, "a," for each question category, e.g., Table 4.2.6.a, lists the coded names
and descriptions for the chemicals/metabolites with significant relationships in the question
category. Thus, of all the chemicals/metabolites measured in relationships with questions
from this category (Tables B.3.1.1.a-g), only the 11 listed in Table 4.2.6.a had significant
relationships. The second table type, "b," e.g., Table 4.2.6.b, shows the number and
percentage of significant relationships by medium for the category. Thus, in Table 4.2.6.b,
there are three significant relationships for ATZ (atrazine) in personal air and residential
pesticide use questions. Four relationships for personal air were found for this category, and
75% of the relationships are statistically significant (Table B.3.1.1.f). The third table type,
"c," e.g., Table 4.2.6.C, lists each question/medium/chemical-metabolite combination for
which a majority (>50%) of the relationships are either significant or marginally significant.
Thus, in Table 4.2.6.c, the question Q119-outside treated, overall has statistically significant
relationships for MDA and TCPY in urine (Table B.3.1.1 .a) and for chlorpyrifos (CHLR) in
dust (Table B.3.1.1.c) and solid food (Table B.3.1.1.g). This majority criterion will be
described as "overall" in subsequent tables. Note that questions with spotty levels of
significant relationships are not included in the type "c" table, although they are included in
the type "a", and type "b" tables, and in the appendices.
It is important for the reader to review any results of interest with the details of the
relationships, since information in the following tables is summarized across different studies
and thus, across different populations, questionnaire instruments, and analytical measurement
techniques. A starting point for this level of review is Appendix C, where the reader can
examine both the defining situation and the results for individual relationships associated
with a specific combination of questions, chemicals measured, and significance level.
4-17
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendices B, C, and D include specific information about the relationships through
overview, detail, and comment tables, respectively. The tables in each appendix are
organized by question category, and then question within the category. A complete list of
questions by category is included in Appendix E. The overview tables in Appendix B are a
high level summary of the relationships found in the literature review, by question, medium,
and chemical/metabolite measured. They provide a general indication of each question's
effectiveness in identifying the exposure level to a specific chemical or metabolite. The
detail tables in Appendix C present the specific statistical analysis and descriptive
information for each relationship counted in that question category's overview table.
Additional information about each relationship with respect to the subpopulation analyzed,
the chemical measurement, and the statistical analysis are included in the comment tables in
Appendix D. Instructions for reading these tables are included in each appendix.
Information in both summary and detailed forms can be found for each question category as
shown in Table 4.2.5.
Table 4.2.5 Cross-Reference for Relationship Tables by Question Category Group
Category
Section #
Table #a
Overview
Table #
Detailed
Table #
Comment
Table #
Group
#
Description
Results
Results
Appendix
B
Appendix
C
Appendix
D
Source
1
Residential pesticide use
4.2.4.1
4.2.6.x
B.3.1.1
C.3.1.1
D.3.1.1
Source
2
Household characteristics
4.2.4.2
4.2.7.x
B.3.1.2
C.3.1.2
D.3.1.2
Source
3
Residential sources
(environmental
measures)
4.2.4.3
4.2.8.x
B.3.1.3
C.3.1.3
D.3.1.3
Source
4
Household occupation
4.2.4.4
4.2.9.x
B.3.1.4
C.3.1.4
D.3.1.4
Source
5
Residential proximity to
agricultural fields
4.2.4.5
4.2.10.x
B.3.1.5
C.3.1.5
D.3.1.5
Source
6
Residential location
4.2.4.6
4.2.11.x
B.3.1.6
C.3.1.6
D.3.1.6
Behavior
7
Subject's personal
characteristics
4.2.5.1
4.2.13.x
B.3.2.1
C.3.2.1
D.3.2.1
Behavior
8
Child's behaviors
4.2.5.2
4.2.14.x
B.3.2.2
C.3.2.2
D.3.2.2
Behavior
9
Dietary behaviors
4.2.5.3
4.2.15.x
B.3.2.3
C.3.2.3
D.3.2.3
Behavior
10
Family hygiene practices
4.2.5.4
4.2.16.x
B.3.2.4
C.3.2.4
D.3.2.4
Behavior
11
Smoking-related activities
4.2.5.5
4.2.17.x
B.3.2.5
C.3.2.5
D.3.2.5
Behavior
12
Work exposure/practices
4.2.5.6
4.2.18.x
B.3.2.6
C.3.2.6
D.3.2.6
Other
13
Related exposure levels
4.2.6.1
4.2.20.x
B.3.3.1
C.3.3.1
D.3.3.1
Other
14
Health
4.2.6.2
4.2.21.x
B.3.3.2
C.3.3.2
D.3.3.2
a x in this column refers to the three table types, a, b, and c, described above.
The results for question categories under each of the three risk factors are presented in the
following sections: source (section 4.2.4), behavior (section 4.2.5), and other (section 4.2.6).
4-18
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Within each of these major sections, there is a subsection for each question category.
Finally, a parallel construction of three summary table types (a, b, and c) is included for each
question category.
4.2.4 Presentation of Source Relationships
Questions related to pesticide sources produced 387, or 64%, of the relationships found in
this review (Table 4.2.3). The term "source" is used broadly here to include purposeful
application of pesticides in the residential environment, measurements of pesticide levels in
the residential environment, and incidental or accidental introduction of pesticides into the
residential environment. The questions included in each of these categories can be found in
Appendix E.
4.2.4.1 Category 1: Residential Pesticide Use
This category of questions (Appendix E) focuses on the purposeful application of pesticides
in or around the residence, including indoor treatments for pests, outdoor treatments for
insects, weeds, and other garden pests, and commercial applications of residential property.
The chemicals/metabolites measured in the study samples having the most medium/question
relationships in this category include: azinphosmethyl+phosmet, DAPs, and TCPY (Tables
B.3.1.1.a-g); however, azinphosmethyl+phosmet did not have any significant relationships
with questions from this category (Table B.3.1.1.c).
Table 4.2.6.a Codes and Descriptions for Chemicals/Metabolites with Significant Relationships for
Questions in the Residential Pesticide Use Category
Code(s)
Medium3
Description13
ATZ
other
Atrazine
CHLR
other
Chlorpyrifos
DAP2
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one detectable measurement)
DAP3
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one high measurement)0
ETHL1, ETHYL1
urine
DEP+DETP
ETHL3, ETHYL3
urine
DEP, DETP, DEDTP
(at least one detectable measurement)
MDA
urine
Malathion dicarboxylic acid
MTHL2, METHYL2
urine
DMP+DMTP+DMDTP
MTHL3, METHYL3
urine
DMTP (detectable measurement)
4-19
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Code(s)
Medium3
Description13
MTHL4, METHYL4
urine
DMP, DMTP
(at least one detectable measurement)
TCPY
urine
3,5,6-T richloro-2-pyridinol
a Medium is noted as urine or other (any other medium sampled).
b DEP = diethylphosphate, DETP = diethylthiophosphate, DEDTP = diethyldithiophosphate
DMP = dimethylphosphate, DMTP = dimethylthiophosphate, DMDTP = dimethyldithiophosphate.
c See definition of high measurement in Azaroff (1999).
The residential pesticide use category includes 100 or 17% of the relationships extracted
from the relevant publications (Table 4.2.3). This category of questions has the second
highest occurrence of relationships. It is considered an important exposure source because
children may be exposed during the application procedures, or may contact pesticide residues
in the residential environment soon after application when residue levels can be relatively
high.
Table 4.2.6.b Distribution of Significant Medium/Question Relationships for Residential Pesticide Use
Questions, by Medium
Medium/Question Relationships3 "
Significant0
Medium
Sampled
Chemicals/Metabolites Measuredd
N
%e
Total
N
Urine
MDA, TCPY, ETHYL1, ETHYL3, METHYL2, METHYL3, METHYL4,
DAP2, DAP3
30
38
81
Dust
CHLR
1
13
8
Indoor air
0
0
2
Outdoor air
0
0
2
Personal air
ATZ, CHLR
3
75
4
Solid food
CHLR
1
33
3
Total
35
35
100
a See the paragraph immediately following Table 4.2.3, above, regarding relationships from Sexton (2003).
b Based on counts in Tables B.3.1.1.a through B.3.1.1 .g.
c Significant (p < 0.05) and marginally significant (p < 0.10).
d See descriptions in Table 4.2.6.a.
e Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Relationships between questions and urine sample concentrations account for 81% of
the relationships in this category (Tables B.3.1.1.a-g). This is much higher than the
overall percentage of relationships with urine measurements (65.6%).
4-20
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
• The percent of relationships with dust measurements, 8%, is much lower than the
overall percentage of 31% (Table B.3.1.1 .c). This is not unexpected, since only a few
of the studies for this category included dust sampling.
• TCPY and DAPs are the predominant metabolites measured in this category's urine-
based relationships, 37% and 31%, respectively (Tables B.3.1.1.a-b).
• For TCPY, about 50% of the relationships are significant; for the DAPs, about 23%
are significant (Table B.3.1.1.a). This difference may be due to the higher specificity
of the TCPY metabolite; i.e., it is specific for chlorpyrifos or chlorpyrifos methyl,
whereas the DAPs can be the result of multiple OP pesticides.
• Thirty-eight percent of the relationships with urine metabolites are significant or
marginally significant (Table 4.2.10.b).
• Only one of the relationships with dust chemicals is significant (Table 4.2. lO.b).
Relationships with other environmental measurements were found; however, they were
found only in Sexton (2003) and the number of relationships was small (Tables B.3.1.1.d-g).
The relationships were in the direction expected, that is, the activity was associated with
higher exposure measurements.
See Table 4.2.5 for tables with related information for the questions in this category.
Table 4.2.6.C Residential Pesticide Use Questions and Chemicals/Metabolites with Overall3 Significant
Relationships'"
Q#
Description
Medium
Chemicals/Metabolites Analyzed0
Q102
Inside Treated
Personal Air
CHLR
Q104
Inside Treated - Bedroom
Urine
TCPY
Q106
Inside Treated - Closets
Urine
TCPY
Q108
Inside Treated - Dining Room
Urine
TCPY
Q111
Inside Treated - Living Room
Urine
TCPY
Q117
Inside Treated - Other Room
Urine
TCPY
Q119
Outside Treated
Urine
MDA, TCPY
Dust
CHLR
Solid Food
CHLR
Q120
Garden Treated
Urine
TCPY, ETHYL1, METHYL2
Q121
Lawn/yard Treated
Urine
TCPY
Q124
Level of Pesticide Use
Urine
MDA, TCPY
Personal Air
ATZ
Q125
Frequency Personal Application Inside
Urine
TCPY
Q126
Frequency Personal Application Outside
Urine
TCPY
Q127
Inside/Outside Treated by Family Member
Urine
ETHYL3, METHYL3, METHYL4, DAP2,
DAP3
Q130
Personally Mixed Pesticide Inside
Urine
TCPY
a Overall indicates that > 50% of the question/medium/chemical relationships are significant.
4-21
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
b See the paragraph immediately following Table 4.2.3, above, regarding relationships from Sexton (2003).
c See descriptions in Table 4.2.6.a.
In most instances, and where information regarding the direction of the relationships is
provided, the significant or marginally significant medium/question relationships are in
agreement with the expectation that the exposure or activity is associated with a higher
measurement level (Table C.3.1.1). Many of the significant relationships in Sexton (2003)
(e.g., ID#s 562 and 567) show an effect opposite of what is expected. The publication
speculated that the unexpected direction occurs either due to chance given the large number
of relationships tested and/or in instances having a large number of non-detects. Note that
questions regarding room-specific treatment are only available from Krinsley (1998) where
the majority of respondents are adults. Overall the questions selected from this category
(Table 4.2.6.c) appear to be useful predictors of exposure level for the chemicals and
metabolites noted.
4.2.4.2 Category 2: Household Characteristics
Questions in this category (Appendix E) focus on unusual circumstances related to the
household characteristics that might be associated with pesticide exposure. In particular, if
property was used as a farm, there was a presumption that pesticide use might be greater or
different than for other residences. Also, the movement of pets in and out of the house might
lead to the track-in of pesticides that would not occur otherwise.
The chemicals/metabolites measured in the study samples having the most medium/question
relationships in this category include: azinphosmethyl, DAPs, a sum of selected OP
pesticides, and TCPY (Tables B.3.1.2.a-d).
Table 4.2.7.a Codes and Descriptions for Chemicals/Metabolites with Significant Relationships for
Questions in the Household Characteristics Category
Code(s)
Medium3
Description13
AZM
other
Azinphosmethyl
CHLR
other
Chlorpyrifos
ETHL1, ETHYL1
urine
DEP+DETP
MDA
urine
Malathion dicarboxylic acid
MTHL2, METHYL2
urine
DMP+DMTP+DMDTP
OPSUM
other
OP Sum
a Medium is noted as urine or other (any other medium sampled).
b DEP = diethylphosphate, DETP = diethylthiophosphate;
DMP = dimethylphosphate, DMTP = dimethylthiophosphate, DMDTP = dimethyldithiophosphate
OP Sum = azinphosmethyl, chlorpyrifos, malathion, and phosmet.
The household characteristics category of questions includes 73 or 12% of the relationships
extracted from the relevant publications (Table 4.2.3). This category of questions falls into
4-22
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
the mid-range occurrence level of relationships. It is considered an important exposure
source because young children spend a majority of their time in this environment — their
residence. Further the pesticide does not degrade as quickly indoors as it does outdoors,
because there is less sunshine and air circulation indoors.
Table 4.2.7.b Distribution of Significant Medium/Question Relationships with Household
Characteristics Questions, by Medium
Medium/Question Relationships '3
Significant0
Total
Medium Sampled
Chemicals/Metabolites Measuredd
N
%e
N
Urine
ETHYL1.MDA, METHYL2
4
7
54
Dust
AZM, CHLR, OPSUM
3
17
18
Indoor air
CHLR
1
100
1
Total
8
11
73
a See the paragraph immediately following Table 4.2.3, above, regarding relationships from Sexton (2003).
b Based on counts in Tables B.3.1.2.a through B.3.1.2.d.
c Significant (p < 0.05) and marginally significant (p < 0.10).
d See descriptions in Table 4.2.7.a.
e Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Relationships between questions and urine sample concentrations account for 74% of
the relationships in this category (Tables B.3.1.2.a-b). This is slightly higher than the
overall percentage of relationships with urine measurements (65.6%).
• The percent of relationships with dust measurements, 25%, is slightly lower than the
overall percentage of 31% (Table B.3.1.2.c).
• TCPY and DAPs are the predominant metabolites measured in this category's urine-
based relationships (Tables B.3.1.2.a-b). Fifty percent of these relationships are with
ethylated DAPs, and 31% are with methylated DAPs. This difference may be due to
the higher specificity of the TCPY metabolite; i.e., it is specific for chlorpyrifos or
chlorpyrifos methyl, whereas the DAPs can be the result of multiple OP pesticides.
• Seven percent of all the relationships with urine metabolites are significant (Table
4.2.7.b).
• Seventeen percent of the relationships with dust chemicals, and the one relationship
with the indoor air chemical are significant (Table 4.2.7.b).
See Table 4.2.5 for tables with related information for the questions in this category.
4-23
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.7.c Household Characteristics Questions and Chemicals/Metabolites with Overall"
Significant Relationships'"
Q#
Description
Medium
Metabolites Analyzedc
Q202
Property Used As a Farm
Dust
CHLR
Indoor Air
CHLR
Q209
Pets Inside/Outside House
Urine
MDA
Q211
Existence of Garden or
Vegetable Garden
Urine
ETHYL1, MDA
Q213
Size of Household
Dust
AZM
a Overall indicates that > 50% of the question/medium/chemical relationships are significant.
b See the paragraph immediately following Table 4.2.3, above, regarding relationships from Sexton (2003).
c See descriptions in Table 4.2.7.a.
In most instances, and where information regarding the direction of the relationships is
provided, the significant or marginally significant medium/question relationships are in
agreement with the expectation that the exposure or activity is associated with a higher
measurement level (Table C.3.1.2). Q202 includes only two relationships from Sexton
(2003), and although they are significant and marginally significant, the relationships are in
the opposite direction from what is expected. For example, property used as a farm would be
expected to have higher measurement levels because of additional uses of pesticides;
however, the levels in dust and indoor air were lower for farm property. The publication
speculated that the unexpected direction occurred either due to chance given the large
number of relationships tested and/or in instances having a large number of non-detects.
Thus, the question may not be a useful predictor of a child's exposure level. The other
questions selected from this category (Table 4.2.7.c) appear to be useful in predicting
exposure level for the chemical or metabolite noted.
4.2.4.3 Category 3: Residential Sources (Environmental Measures)
This category of questions (Appendix E) focused on relationships between measurements of
pesticides in the soil of residential environments and pesticides in house dust, as well as
between measurements of pesticides in house dust and/or soil and pesticide metabolite levels
in urine. In these cases the source of the pesticides in the environment was not known.
Pesticide contamination could have occurred for any number of reasons. While the specific
reason was not known in most cases, there was clear evidence that pesticides were present in
the residential environment.
The chemicals/metabolites measured in the study samples are azinphosmethyl, chlorpyrifos,
DAPs, ethyl parathion, and phosmet (Tables B.3.1.3.a-b).
4-24
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.8.a Codes and Descriptions for Chemicals/Metabolites with Significant Relationships for
Questions in the Residential Sources Category
Code(s)
Medium3
Description13
AZM
other
Azinphosmethyl
CHLR
other
Chlorpyrifos
EPAR
other
Ethyl parathion
MTHL2, METHYL2
urine
DMP+DMTP+DMDTP
NA
urine
Not available in publication
PHSM
urine
Phosmet
a Medium is noted as urine or other (any other medium sampled)
b DMP = dimethylphosphate, DMTP = dimethylthiophosphate, DMDTP = dimethyldithiophosphate
The residential source category of questions includes 13 or 2% of the relationships extracted
from the relevant publications (Table 4.2.3). This category of questions falls into the low-
range occurrence level of relationships. Relatively few of the studies under review collected
house dust and urine samples concurrently and only Simcox (1995) collected soil and house
dust samples concurrently. The very low pesticide concentrations found in soil in this study
led later investigators to focus on house dust and other sources rather than soil.
Table 4.2.8.b Distribution of Significant Medium/Question Relationships with Residential Sources
Questions, by Medium
Medium/Question Relationships3
Significant3
Total
Medium Sampled
Chemicals/Metabolites Measured0
N
%d
N
Urine
METHYL2, NA
4
80
5
Dust
AZM, CHLR, EPAR, PHSM
5
63
8
Total
9
69
13
a Based on counts in Tables B.3.1.3.a and B.3.1.3.b.
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.8.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Relationships in this category were found only between urine or dust sample
concentrations and other environmental measurements (Tables B.3.1.3.a-b).
• The relationships with urine sample concentrations account for 38% of the
relationships in this category (Table B.3.1.3.a). This is much lower than the overall
percentage of relationships with urine measurements (65.6%).
• The percent of relationships with dust measurements, 62%, is much higher than the
overall percentage of 31% (Table B.3.1.3 .b). The higher percent of relationships with
4-25
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
dust as compared to urine is likely due to the role of house dust as a reservoir for
pesticides in the home, whereas metabolite measurements reflect only exposure that
may have occurred in the last 1-3 days.
• Eighty percent of the relationships with urine metabolites are significant (Table
4.2.8.b).
• Sixty-three percent of the relationships with dust chemicals are significant (Table
4.2.8.b).
See Table 4.2.5 for tables with related information for the questions in this category.
Table 4.2.8.C Residential Sources Questions and Chemicals/Metabolites with Overall11 Significant
Relationships
Q#
Description
Medium
Chemicals/Metabolites Analyzed13
Q301
Household Dust
Urine
METHYL2, NA
Q303
Outdoor Soil
Dust
EPAR
a Overall indicates that > 50% of the question/medium/chemical relationships are significant.
b See descriptions in Table 4.2.8.a.
The relationships in this question category compared measurements, environmental to
environmental, or environmental to urinary, and the statistical analyses used were
correlations or regression analysis (Table C.3.1.3). When the direction of the relationships
was provided for the significant or marginally significant relationships, the measurements
generally increased together as expected. Thus, house dust and soil measurements (Table
4.2.8.c) may be considered useful in predicting exposure level.
4.2.4.4 Category 4: Household Occupation
Children who live in households where one or more of the adults has occupational exposures
to pesticides may be at risk for increased exposure. This para-occupational exposure has been
well demonstrated in studies of lead battery workers, asbestos workers, and others. A
number of studies have been conducted recently to examine the extent to which pesticides
used in the workplace are found in the home and whether this exposure pathway contributes
to the body burden of children living in those homes. A list of the questions included in this
category can be found in Appendix E.
The chemicals/metabolites measured in the study samples having the most medium/question
relationships in this category include: azinphosmethyl, chlorpyrifos, DAPs, ethyl parathion,
and phosmet (Tables B.3.1.4.a-d).
4-26
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.9.a Codes and Descriptions for Chemicals/Metabolites with Significant Relationships for
Questions in the Household Occupation Category
Code(s)
Medium3
Description13
AZM
other
Azinphosmethyl
AZMPH
other
Azinphosmethyl+Phosmet
CHLR
other
Chlorpyrifos
DAP2
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one detectable measurement)
DAP3
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one high measurement)0
DMTP
urine
Dimethylthiophosphate (DMTP)
EPAR
other
Ethyl parathion
ETHL3, ETHYL3
urine
DEP, DETP, DEDTP
(at least one detectable measurement)
MTHL1, METHYL1
urine
DMTP+DMDTP
MTHL3, METHYL3
urine
DMTP (detectable measurement)
MTHL4, METHYL4
urine
DMP, DMTP
(at least one detectable measurement)
MTHL5, METHYL5
urine
DMP, DMTP
(at least one high measurement)0
OPSUM
other
OP Sum
PHSM
other
Phosmet
a Medium is noted as urine or other (any other medium sampled).
b DEP = diethylphosphate, DETP = diethylthiophosphate, DEDTP = diethyldithiophosphate
DMP = dimethylphosphate, DMTP = dimethylthiophosphate, DMDTP = dimethyldithiophosphate
OP Sum = azinphosmethyl, chlorpyrifos, malathion, and phosmet.
c See definition of high measurement in Azaroff (1999).
The household occupation category of questions includes 115 or 19% of the relationships
extracted from the relevant publications (Table 4.2.3). This category of questions has the
highest level of relationship occurrence. The pesticides used in agricultural workplaces can
normally be identified with a high degree of specificity and the presence of these compounds
in the home environment is clear evidence of workplace-to-residence chemical transmission.
4-27
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.9.b Distribution of Significant Medium/Question Relationships with Household Occupation
Questions, by Medium
Medium/Question Relationships3
Significant13
Total
Medium Sampled
Chemicals/Metabolites Measured0
N
%d
N
Urine
DAP2, DAP3, DMTP, ETHYL3, METHYL1, METHYL3,
METHYL4, METHYL5
19
40
48
Dust
AZM, AZMPH, CHLR, EPAR, OPSUM, PHSM
28
47
59
Soil
AZM
1
13
8
Total
48
42
115
a Based on counts in Tables B.3.1.4.a through B.3.1.4.d.
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.9.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Relationships between questions and urine sample concentrations account for 42% of
the relationships in this category (Tables B.3.1.4.a-b). This is much lower than the
overall percentage of relationships with urine measurements (65.6%).
• The percent of relationships with dust measurements, 51%, is much higher than the
overall percentage of 31% (Table B.2.1,4.c). Dust is a convenient and stable medium
for assaying the presence of agricultural chemicals in the home.
• Ethylated and methylated DAPs are the predominant metabolites measured in this
category's urine-based relationships, 23% and 56%, respectively (Tables B.3.1.4.a-b).
• For ethylated DAPs, 18% of the relationships are significant (Table B.3.1.4.a-b).
• For methylated DAPs, 48% of the relationships are significant (Table B.3.1.4.a-b).
• Forty percent of the relationships with urine metabolites are significant or marginally
significant (Table 4.2.9.b).
• Azinphosmethyl, chlorpyrifos, ethyl parathion, and phosmet are the predominant
chemicals measured in this category's dust-based measurements. The percent of
significant relationships for these chemicals is 42, 43, 56, and 8, respectively (Table
B.3.1.4.C).
• Overall 47% of the relationships with dust chemicals are significant, and one
relationship (13%) with a soil chemical is significant (Table 4.2.9.b).
See Table 4.2.5 for tables with related information for the questions in this category.
4-28
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.9.C Household Occupation Questions and Chemicals/Metabolites with Overall3 Significant
Relationships
Q#
Description
Medium
Chemicals/Metabolites Analyzed13
Q401
Agricultural Workers in
Household
Dust
AZM
Q402
Household Member Spraying
Fields
Urine
DAP2, DAP3, ETHYL3, METHYL3, METHYL4,
METHYL5
Q403
Recent Fieldwork
Urine
DAP2, DAP3, METHYL4, METHYL5
Q404
Applicator vs Farmworker
Dust
AZMPH, EPAR
Q405
Applicator vs Non-applicator
Dust
CHLR, EPAR
Q406
Applicator vs Reference
Urine
DMTP
Q407
Applicator and Farmworker vs
Reference
Urine
DMTP, METHYL1
Dust
AZM, AZMPH, CHLR, EPAR, PHSM
Q409
Farmer and Farmworker vs
Reference
Soil
AZM
Q412
Fieldworker vs Pesticide
Handler
Dust
AZM
Q415
Tree Thinning
Dust
OPSUM
Q416
Number in household with
high pesticide contact
Dust
OPSUM
a Overall indicates that > 50% of the question/medium/chemical relationships were significant.
b See descriptions in Table 4.2.9.a.
Based on the available information, the significant or marginally significant
medium/question relationships seem to be in agreement with the expectation that the
exposure or activity is associated with a higher measurement level (Table C.3.1.4). In some
instances, no information regarding the direction of the relationships was provided.
Questions under the Family Hygiene Practices and Work Practices/Exposures categories are
also related to this exposure pathway. Overall, the questions selected in this category (Table
4.2.9.c) appear to be useful in predicting pesticide exposure levels.
4.2.4.5 Category 5: Residential Proximity to Agricultural Fields
Pesticide spray application remains a concern for families in agricultural communities and
may contribute to a child's exposure. This is of particular concern as new housing
developments are situated adjacent to working farms and where agricultural workers are
housed, within or on the boundaries of agricultural fields. The distance between the
residence and agricultural fields has been used as a surrogate metric for home contamination
that can result from pesticide application spraying events. The accuracy of this metric is open
to question, particularly when it is self-reported. More advanced methods of characterizing
the link between agricultural pesticide use and human exposure are a topic of current
scientific inquiry. A list of the questions included in this category can be found in Appendix
E.
4-29
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The chemicals/metabolites measured in the study samples having the most medium/question
relationships in this category include: azinphosmethyl, chlorpyrifos, ethyl parathion, and
phosmet (Tables B.3.1.5.a-b).
Table 4.2.10.a Codes and Descriptions for Chemicals/Metabolites with Significant Relationships for
Questions in the Residential Proximity to Agricultural Fields Category
Code(s)
Medium3
Description13
AZM
other
Azinphosmethyl
AZMPH
other
Azinphosmethyl+Phosmet
CHLR
other
Chlorpyrifos
DMTP
urine
Dimethylthiophosphate (DMTP)
EPAR
other
Ethyl parathion
MTHL1, METHYL1
urine
DMTP+DMDTP
a Medium is noted as urine or other (any other medium sampled).
b DMP = dimethylphosphate, DMTP = dimethylthiophosphate.
The residential proximity to agricultural fields category of questions includes 72 or 12% of
the relationships extracted from the relevant publications (Table 4.2.3). This category of
questions falls into the mid-range occurrence level of relationships. The possibility of
misclassification of exposure potential through use of a simple residential proximity metric is
relatively high. In most cases, it is not known when, or even if, the nearby fields were treated
with pesticides nor is it known what compounds may have been used. Factors, such as wind
direction and application procedures, are important variables that are not accounted for in the
use of residential proximity.
Table 4.2.10.b Distribution of Significant Medium/Question Relationships with Residential Proximity
to Agricultural Fields Questions, by Medium
Medium/Question Relationships3
Significant13
Total
Medium Sampled
Chemicals/Metabolites Measured0
N
%d
N
Urine
DMTP, METHYL1
4
13
30
Dust
AZM, AZMPH, CHLR, EPAR
16
38
42
Total
20
28
72
a Based on counts in Tables B.3.1.5.a and B.3.1.5.b.
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.10.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
4-30
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The relationships in this question category can be summarized as follows:
• Relationships between questions and urine sample concentrations account for 42% of
the relationships in this category (Table B.3.1.5.a). This is much lower than the
overall percentage of relationships with urine measurements (65.6%).
• The percent of relationships with dust measurements, 58%, is much higher than the
overall percentage of 31% (Table B.3.1.5.b). Dust is a convenient and stable medium
for attempting to track the impact of agricultural pesticide use on residential
environments.
• Methylated DAPs are the predominant metabolites measured {61%) in this category's
urine-based relationships, but only 20% of those relationships are significant (Table
B.3.1.5.a). Many of the studies that have explored this relationship have been
conducted in agricultural regions where methyl DAPs are used for insect control.
• Thirteen percent of the relationships with urine metabolites are significant (Table
4.2.10.b).
• Azinphosmethyl, chlorpyrifos, ethyl parathion, and phosmet are the predominant
chemicals measured in this category's dust-based measurements. The percent of
significant relationships for these chemicals is 31, 38, 56, and 0, respectively (Table
B.3.1.5.b).
• Overall 38% of the relationships with dust chemicals are significant (Table 4.2.10.b).
See Table 4.2.5 for tables with related information for the questions in this category.
Table 4.2.10.C Residential Proximity to Agricultural Fields Questions and Chemicals/Metabolites with
Overall11 Significant Relationships
Q#
Description
Medium
Chemicals/Metabolites Analyzed13
Q501
Proximity of Home to
Pesticide-Treated
Farmland/Orchard
Urine
DMTP
Dust
AZMPH, EPAR
a Overall indicates that > 50% of the question/medium/chemical relationships were significant.
b See descriptions in Table 4.2.10.a
In most instances, the significant or marginally significant media/question relationships are in
agreement with the expectation that the exposure or activity is associated with a higher
measurement level (Table C.3.1.5). In some instances, no information regarding the direction
of the relationships is provided. Thus, proximity of the home to pesticide-treated farmland or
orchards appears to be useful for predicting exposure for the chemicals and metabolites noted
(Table 4.2.lO.c).
4.2.4.6 Category 6: Residential Location
This category of questions (Appendix E) was developed to capture aspects of residential
location other than proximity to agricultural fields. In particular, some studies have
4-31
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
compared urban and rural residential environments on the assumption that rural
environments might provide a greater opportunity for children's exposure to pesticides.
Also, some studies have examined the relationship between pesticide concentrations in house
dust and vehicle dust, based on the theory that the vehicle may serve as a vector for pesticide
transmission into the home and as a direct source of exposure when children are transported.
The chemicals/metabolites measured in the study samples are 1-naphthol, malathion
dicarboxylic acid, TCPY, ethylated DAPs, and azinphosmethyl (Tables B.3.1.6.a-b);
however, only azinphosmethyl and TCPY had any significant relationships.
Table 4.2.11.a Codes and Descriptions for Chemicals/Metabolites with Significant Relationships for
Questions in the Residential Location Category
Code(s)
Medium3
Description
AZM
other
Azinphosmethyl
TCPY
urine
3,5,6-T richloro-2-pyridinol
a Medium is noted as urine or other (any other medium sampled)
The residential location category of questions includes 14 or 2% of the relationships
extracted from the relevant publications (Table 4.2.3). This category of questions falls into
the low-range occurrence level of relationships. Only a few studies have examined these
relationships, so data for this category is limited.
Table 4.2.11.b Distribution of Significant Medium/Question Relationships with Residential Location
Questions, by Medium
Medium/Question Relationships3
Significant13
Total
Medium Sampled
Chemicals/Metabolites Measured0
N
%d
N
Urine
TCPY
3
23
13
Dust
AZM
1
100
1
Total
4
29
14
a Based on counts in Tables B.3.1.6.a and B.3.1.6.b.
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.11.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Relationships between questions and urine sample concentrations account for 93% of
the relationships in this category (Table B.3.1.6.a). This is much higher than the
overall percentage of relationships with urine measurements (65.6%). Most studies
4-32
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
that have explored these relationships have focused on urinary metabolite
measurements.
• There is only one relationship with a dust sample concentration (Table B.3.1.6.b).
Studies of this kind have not been conducted frequently. They focus on worker
commuter vehicles as part of an exposure pathway for children, and this has only
recently been recognized as a potential contributor to exposure.
• TCPY is the predominant metabolite measured (36%) in this category's urine-based
relationships, but only 3 (60%) of its relationships are significant (Table B.3.1.6.a). It
is not surprising to find that the relationships between questions and TCPY
metabolites were not found to be significant. Chlorpyrifos, the parent compound of
the TCPY metabolite, has been until recently the most widely used OP pesticide in
the United States. Furthermore, diet is an important source of chlorpyrifos exposure,
so a simple categorization of homes as urban or rural would be unlikely to
demonstrate differential body burdens in children.
• Overall, 23% of the relationships with urine metabolites are significant (Table
4.2.1l.b)
• The one relationship with azinphosmethyl in dust is also significant (Table 4.2.11 .b).
See Table 4.2.5 for tables with related information for the questions in this category.
Table 4.2.11.C Residential Location Questions and Chemicals/Metabolites with Overall3 Significant
Relationships
Q#
Description
Medium
Chemicals/Metabolites Analyzed13
Q601
Urban vs Non-urban
Urine
TCPY
Q605
Vehicle vs House
Dust
AZM
a Overall indicates that > 50% of the question/medium/chemical relationships were significant.
b See descriptions in Table 4.2.11.a.
In most instances where information regarding the direction of the relationships is provided,
the significant or marginally significant media/question relationships were in agreement with
the expectation that the exposure or activity would be associated with a higher measurement
level (Table C. 3.1.6). Thus, although the number of relationships evaluated for questions in
this category is small, the questions selected (Table 4.2.1 l.c) appear to be useful in predicting
exposure level for the chemical and metabolite noted.
4.2.4.7 Summary of Results from Source Relationships
The six question categories under the source risk factor represent sources of exposure in the
residential environment. Thirty-five questions from these categories are considered overall
statistically significant (and effective differentiators of pesticide exposure levels) for the
chemicals/metabolites noted (Table 4.2.12). For each of the question and
chemical/metabolite combinations, the majority (> 50%) of the relationships were
statistically or marginally significant.
4-33
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.12 Questions from Source Categories Considered Overall Statistically Significant, by
Medium
Medium
Q Category
Q #a
Q Description
Chemicals/
Metabolitesb
Dust
Residential pesticide use
Q119
Outside Treated0
CHLR
Household characteristics
Q202
Property Used As a Farm0
CHLR
Residential environment
(environmental
measures)
Q213
Size of Household
AZM
Residential sources
Q303
Outdoor Soil
EPAR
Household occupation
Q401
Agricultural Workers in
Household
AZM
Q404
Applicator vs Farmworker
AZMPH, EPAR
Q405
Applicator vs Non-applicator
CHLR, EPAR
Q407
Applicator and Farm worker
vs Reference
AZM, AZMPH, CHLR,
EPAR, PHSM
Q412
Fieldworker vs Pesticide
Handler
AZM
Q415
Tree Thinning
OPSUM
Q416
Number in Household with
High Pesticide Contact
OPSUM
Residential proximity to
agricultural fields
Q501
Proximity of Home to
Pesticide-Treated
Farmland/Orchard
AZMPH, EPAR
Residential location
Q605
Vehicle vs House
AZM
Indoor Air
Household characteristics
Q202
Property Used As a Farm0
CHLR
Personal Air
Residential pesticide use
Q102
Inside Treated
CHLR
Q124
Level of Pesticide Use0
ATZ
Soil
Household occupation
Q409
Farmer and Farm worker vs
Reference
AZM
Solid Food
Residential pesticide use
Q119
Outside Treated0
CHLR
Urine
Residential pesticide use
Q104
Inside Treated - Bedroom
TCPY
Q106
Inside Treated - Closets
TCPY
Q108
Inside Treated - Dining
Room
TCPY
4-34
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Medium
Q Category
Q#a
Q Description
Chemicals/
Metabolites'1
Q111
Inside Treated — Living
Room
TCPY
Q117
Inside Treated — Other
Room
TCPY
Q119
Outside Treated0
MDA, TCPY
Q120
Garden Treated
TCPY, ETHYL1,
METHYL2
Q121
Lawn/Yard Treatedc
TCPY
Q124
Level of Pesticide Usec
MDA, TCPY
Q125
Frequency Personal
Application Inside
TCPY
Q126
Frequency Personal
Application Outside
TCPY
Q127
Inside/Outside Treated by
Family Member
ETHYL3, METHYL3,
METHYL4, DAP2,
DAP3
Q130
Personally Mixed Pesticide
Inside
TCPY
Household characteristics
Q208
Pets in House
METHYL2
Q209
Pets Inside/Outside House0
MDA
Q211
Existence of Garden or
Vegetable Garden0
ETHYL1, MDA
Residential sources
(environmental
measures)
Q301
Household Dust
METHYL2, NA
Household occupation
Q402
Household Member
Spraying Fields
DAP2, DAP3, ETHYL3,
METHYL3, METHYL4,
METHYL5
Q403
Recent Fieldwork
DAP2, DAP3,
METHYL4, METHYL5
Q406
Applicator vs Reference
DMTP
Q407
Applicator and Farm worker
vs Reference
DMTP, METHYL1
Residential proximity to
agricultural fields
Q501
Proximity of Home to
Pesticide-Treated
Farmland/Orchard
DMTP
Residential location
Q601
Urban vs Non-urban
TCPY
a For some of the significant relationships, the effect of the exposure factor was not in the direction expected.
See Appendix C for details on specific questions.
b Chemicals or metabolites for which > 50% of the relationships with the question were statistically or marginally
significant. (See "a" tables: Tables 4.2.6.a through 4.2.11.a for descriptions.)
c See Section 4.2.2 regarding relationships from Sexton (2003).
4-35
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Household occupation was a strong differentiator for pesticide levels in dust which relate to
the take-home pathway exposure. Residential pesticide use and household occupation were
strong differentiating categories for the urine metabolite levels (Table 4.2.12).
4.2.5 Presentation of Behavior Relationships
Many exposure studies include questions that focus on the behavior of household members
or children. The temporal and spatial patterns of children's activities are important variables
in exposure assessment, generally referred to as macro-activities. Additionally, activities
conducted in specific microenvironments, such as crawling, contact with objects, hand-to-
mouth behavior, and object-to-mouth behavior - generally referred to as micro-activities -
are thought to contribute significantly to dermal, oral, and respiratory exposures among
children. Behavior accounted for 203, or 34%, of observed relationships in this review
(Table 4.2.3).
4.2.5.1 Category 7: Subject's Personal Characteristics
A number of studies have collected demographic information, such as age, gender, ethnicity,
and income level, and have explored possible associations with pesticide metabolite levels in
urine. These analyses have been undertaken in an effort to determine if there are consistent
trends related to subject information that is often readily available through census or other
databases. A list of the questions included in this category can be found in Appendix E.
The chemicals/metabolites measured in the study samples having the most medium/question
relationships in this category are DAPs (Tables B.3.2.1.a-b).
Table 4.2.13.a Codes and Descriptions for Metabolites with Significant Relationships for Questions in
the Subject's Personal Characteristics Category
Code(s)
Medium3
Description13
1NAP
urine
1-Naphthol
DAP1
urine
DMP+DMTP+DMDTP+DEP+DETP+DEDTP
DMTP
urine
Dimethylthiophsophate (DMTP)
ETHL2, ETHYL2
urine
DEP+DETP+DEDTP
MDA
urine
Malathion dicarboxylic acid
MTHL2, METHYL2
urine
DMP+DMTP+DMDTP
TCPY
urine
3,5,6-T richloro-2-pyridinol
a Medium is noted as urine or other (any other medium sampled).
b DEP = diethylphosphate, DETP = diethylthiophosphate, DEDTP = diethyldithiophosphate
DMP = dimethylphosphate, DMTP = dimethylthiophosphate, DMDTP = dimethyldithiophosphate.
The subject's personal characteristics category of questions includes 78 or 13% of the
relationships extracted from the relevant publications (Table 4.2.3). This category of
4-36
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
questions falls into the mid-range occurrence level of relationships. Information regarding
age, gender, ethnicity and income is relatively easy to obtain, but, with the exception of age,
it is not clear that these characteristics would be related to pesticide exposures.
Table 4.2.13.b Distribution of Significant Medium/Question Relationships with Subject's Personal
Characteristics Questions, by Medium
Medium/Question Relationships3
Significant13
Total
Medium Sampled
Chemicals/Metabolites Measured0
N
%d
N
Urine
1NAP, DAP1, DMTP, ETHYL2, MDA, METHYL2, TCPY
22
28
78
Total
22
28
78
a Based on counts in Tables B.3.2.1.a and B.3.2.1.b.
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.13.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Only urine concentrations were found for this category (Tables B.3.2.1.a-b).
• Twenty-eight percent of the relationships are significant (Table 4.2.13.b).
• The percent of significant relationships for each metabolite is (Tables B.3.2.1.a-b):
• Ethylated DAPs - 12%,
• Methylated DAPs - 27%
• Ethylated+methylated DAPs - 40%
• 1-Naphthol - 50%)
• MD A - 67%
• TCPY-30%.
See Table 4.2.5 for tables with related information for the questions in this category.
Table 4.2.13.C Subject's Personal Characteristics Questions and Metabolites with Overall3 Significant
Relationships
Q#
Description
Medium
Metabolites Analyzed13
Q702
Age
Urine
DAP1, METHYL2
Q703
Ethnicity
Urine
1NAP, MDA
Q705
Income
Urine
1NAP, MDA, TCPY
a Overall indicates that > 50% of the question/medium/chemical relationships were significant.
b See descriptions in Table 4.2.13.a.
4-37
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Based on the significant relationships, younger children have higher levels of metabolites
than older children, and children have higher levels than adults (Table C.3.2.1). Thus, age
appears to be a useful predictor of pesticide exposure level in the metabolites noted (Table
4.2.13.c). The significant ethnic and income relationships found in Adgate (2001) were not
consistently in the same direction, and given the small number of relationships found in the
publications, they do not appear to be useful predictors of pesticide exposure levels.
4.2.5.2 Category 8: Child's Behaviors
This category includes children's behaviors, actions, and activities that may differentiate
children's pesticide exposure levels. Factors include both habits and hygiene practices of
children such as sucking thumbs and the frequency and timing of hand washing. Time
related activities such as the amount of time children spend in certain environments (e.g.
indoors, outdoors, or at school) can also contribute to measurable differences in their
pesticide exposure levels. A limited number of studies have included children's hand wipes
as an exposure metric and compared pesticide loading values with metabolite levels. A list
of the questions included in this category can be found in Appendix E.
The metabolites measured in the study samples are 4-nitrophenol, DAPs, and TCPY (Table
B.3.2.2.a); however, only the DAP metabolite has significant relationships with the questions
in this category (Table 4.2.14.a).
Table 4.2.14.a Codes and Descriptions for Metabolites with Significant Relationships for Questions in
the Child's Behaviors Category
Code(s)
Medium3
Description13
DAP1
urine
DMP+DMTP+DMDTP+DEP+DETP+DEDTP
a Medium is noted as urine or other (any other medium sampled).
b DEP = diethylphosphate, DETP = diethylthiophosphate, DEDTP = diethyldithiophosphate
DMP = dimethylphosphate, DMTP = dimethylthiophosphate, DMDTP = dimethyldithiophosphate.
The child's behaviors category of questions includes 20 or 3% of the relationships extracted
from the relevant publications (Table 4.2.3). This category of questions falls into the low-
range occurrence level of relationships. It is considered an important exposure factor, since
most investigators believe that a child's behavior will have a significant impact on pesticide
exposure (e.g., see Cohen Hubal (2000b), Black (2005), Reed (1999), Freeman (2005)).
4-38
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.14.b Distribution of Significant Medium/Question Relationships with Child's Behaviors
Questions, by Medium
Medium/Question Relationships3
Significant13
Total
Medium Sampled
Metabolites Measuredc
N
%d
N
Urine
DAP1
2
10
20
Total
2
10
20
a Based on counts in Table B.3.2.2.a
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.14.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Only urine-based relationships were found for this category (Table B.3.2.2.a).
• Ten percent of the relationships are significant or marginally significant (Table
4.2.14.b), and the significant relationships were with the DAP metabolite (Table
B.3.2.2.a).
See Table 4.2.5 for tables with related information for the questions in this category.
Table 4.2.14.C Child's Behaviors Questions and Metabolites with Overall" Significant Relationships
Q#
Description
Medium
Metabolites Analyzed
Q806
Loading from hand wipe
Urine
DAP1
a Overall indicates that > 50% of the question/medium/chemical relationships were significant.
No questions in this category have overall significant relationships with urine measurements
(Table C.3.2.2); however, the loading measurement from the hand wipe (Table 4.2.14.c) does
have a significant relationship. Overall, questions available in the category of child's
behaviors do not appear to be useful in predicting the child's pesticide exposure level.
4.2.5.3 Category 9: Dietary Behaviors
Diet is likely to be a major pathway of pesticide exposure for most children, yet few studies
have examined this issue directly. The U.S. EPA has made a substantial effort to develop
quantitative estimates of dietary pesticide through the combination of food consumption
surveys and analysis of pesticide residues in common food products. Nonetheless, there
remains substantial uncertainty in the ability to predict an individual's pesticide ingestion
based on food diaries or food frequency questionnaires. In this review, only one study was
placed in this category. Additional studies are underway, and should add to the understanding
4-39
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
of this pathway. A list of the questions included in this category can be found in Appendix
E.
The metabolites measured in the study samples are DAPs and TCPY (Table B.3.2.3.a);
however, only a DAP metabolite had significant relationships for this category.
Table 4.2.15.a Codes and Descriptions for Metabolites with Significant Relationships for Questions in
the Dietary Behaviors Category
Code(s)
Medium3
Description13
MTHL2, METHYL2
urine
DMP+DMTP+DMDTP
a Medium is noted as urine or other (any other medium sampled).
b DMP = dimethylphosphate, DMTP = dimethylthiophosphate, DMDTP = dimethyldithiophosphate.
The dietary behaviors category of questions includes 16 or 3% of the relationships extracted
from the relevant publications (Table 4.2.3). This category of questions falls into the low-
range occurrence level of relationships, which is due primarily to the lack of studies that have
focused on this pathway.
Table 4.2.15.b Distribution of Significant Medium/Question Relationships with Dietary Behaviors
Questions, by Medium
Medium/Question Relationships3
Significant13
Total
Medium Sampled
Metabolites Measuredc
N
%d
N
Urine
METHYL2
2
13
16
Total
2
13
16
a Based on counts in Table B.3.2.3.a.
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.15.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Only urine-based relationships were found for this category (Table B.3.2.3.a).
• The significant relationships are for a methylated DAP sum (Table B.3.2.3.a), and
account for 13% of the urine-based relationships (Table 4.2.15.b). A number of
studies have found that the methyl DAPs are more common than ethyl DAPs in urine.
See, for example, the most recent NHANES data (Barr 2004).
See Table 4.2.5 for tables with related information for the questions in this category.
4-40
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.15.C Dietary Behaviors Questions and Metabolites with Overall11 Significant Relationships
Q#
Description
Medium
Metabolites Analyzed13
Q904
Organic Diet
Urine
METHYL2
a Overall indicates that > 50% of the question/medium/chemical relationships were significant.
b See descriptions in Table 4.2.15.a
Based on the significant relationships, a conventional diet has higher pesticide metabolite
levels than an organic diet (Table C.3.2.3). Although the number of relationships for this
question is small, it appears to be useful in predicting methylated DAP metabolite levels.
4.2.5.4 Category 10: Family Hygiene Practices
Many investigators have placed an emphasis on good hygienic practices within the home as a
means of reducing children's exposure to pesticides. For example, it is common for public
health scientists and practitioners to recommend that agricultural workers remove their work
boots before entering the home and that work clothing be washed separately from the family
clothing. Thus, family hygiene is an important variable to investigate in studies of children's
pesticide exposure. A list of the questions included in this category can be found in
Appendix E.
The chemicals/metabolites measured in the study samples having the most medium/question
relationships in this category include: azinphosmethyl, chlorpyrifos, dimethylthiophosphate
(DMTP), and ethyl parathion (Tables B.3.2.4.a-b).
Table 4.2.16.a Codes and Descriptions for Chemicals/Metabolites with Significant Relationships for
Questions in the Family Hygiene Practices Category
Code(s)
Medium3
Description
AZM
other
Azinphosmethyl
DMTP
urine
Dimethylthiophosphate
OPSUM
other
OP Sumb
a Medium is noted as urine or other (any other medium sampled)
b OP Sum = azinphosmethyl, chlorpyrifos, malathion, and phosmet.
The family hygiene practices category of questions includes 81 or 13% of the relationships
extracted from the relevant publications. This category of questions falls into the mid-range
occurrence level of relationships (Table 4.2.3). Many of the studies under review have been
conducted in agricultural communities, so it is not surprising that questions related to family
hygiene would occur with some frequency.
4-41
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.2.16.b Distribution of Significant Medium/Question Relationships with Family Hygiene
Practices Questions, by Medium
Medium/Question Relationships3
Significant13
Total
Medium Sampled
Chemicals/Metabolites Measured0
N
%d
N
Urine
DMTP
2
6
33
Dust
AZM, OPSUM
3
6
48
Total
5
6
81
a Based on counts in Tables B.3.2.4.a and B.3.2.4.b.
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.16.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Relationships between questions and urine sample concentrations account for 41% of
the relationships in this category (Table B.3.2.4.a). This is much lower than the
overall percentage of relationships with urine measurements (65.6%).
• The percent of relationships with dust measurements, 59%, is much higher than the
overall percentage of 31% (Table B.3.2.4.b). Dust is used commonly as a metric for
home contamination by workplace chemicals, since its measurement is more stable
than urinary metabolites.
• DMTP is the predominant metabolite measured (48%) in this category's urine-based
relationships (Table B.3.2.4.a).
• Only 6% of the relationships with DMTP are significant (Table 4.2.16.b), and these
are the only urine-based relationships that are significant (Table B.3.2.4.a).
• For dust concentrations, azinphosmethyl, chlorpyrifos, and ethyl parathion were the
predominant chemicals measured, 25%, 21%, and 21%, respectively (Table
B.3.2.4.b).
• Overall, only six percent of the relationships with dust concentrations are significant
(Table 4.2.16.b).
See Table 4.2.5 for tables with related information for the questions in this category.
Table 4.2.16.C Family Hygiene Practices Questions and Chemicals/Metabolites with Overall11
Significant Relationships
Q#
Description
Medium
Chemicals/Metabolites Analyzed13
Q1006
Work Clothes Worn Indoors
Dust
AZM, OPSUM
Q1009
Number of Weeks Since Last
Vacuuming
Dust
OPSUM
a Overall indicates that > 50% of the question/medium/chemical relationships were significant.
b See descriptions in Table 4.2.16.a.
4-42
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
In instances where information regarding the direction of the relationships is provided, the
significant or marginally significant media/question relationships are in agreement with the
expectation that the exposure or activity is associated with a higher measurement level (Table
C.3.2.4). Both longer periods of wearing work clothes indoors and more weeks since last
vacuuming were associated with higher measurement levels of dust. Although there are a
small number of relationships for these questions, they appear to be useful in predicting
exposure levels in dust for the chemicals noted (Table 4.2.16.c).
4.2.5.5 Category 11: Smoking-Related Activities
Several studies have included the measurement of urinary cotinine, the primary metabolite of
nicotine, as a marker of children's exposure to smoking. Smoking at the workplace has been
associated with higher pesticide exposures since cigarettes or other smoking material may
become contaminated during work. It is not clear, however, that there is a plausible
hypothesis for an effect of adult smoking behavior on children's pesticide exposure. A list of
the questions included in this category can be found in Appendix E.
The only chemical/metabolite measured in the study samples was TCPY (Table B.3.2.5.a).
Table 4.2.17.a Codes and Descriptions for Metabolites with Significant Relationships for Questions in
the Smoking-Related Activities Category
Code(s)
Medium3
Description
TCPY
urine
3,5,6-T richloro-2-pyridinol
a Medium is noted as urine or other (any other medium sampled).
The smoking-related activities category of questions includes 4 or 1% of the relationships
extracted from the relevant publications (Table 4.2.3) and falls into the low-range occurrence
level. This finding is not surprising, since there would appear to be little relationship between
smoking and pesticide metabolite levels in children.
Table 4.2.17.b Distribution of Significant Medium/Question Relationships with Smoking-Related
Activities Questions, by Medium
Medium/Question Relationships3
Significant13
Total
Medium Sampled
Metabolites Measuredc
N
%d
N
Urine
TCPY
3
75
4
Total
3
75
4
a Based on counts from Table B.3.2.5.a.
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.17.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
4-43
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The relationships in this question category can be summarized as follows:
• Only urine-based relationships were found for this category (Table B.3.2.5.a).
• Seventy-five percent of the relationships with TCPY are significant (Table 4.2.17.b).
See Table 4.2.5 for tables with related information for the questions in this category.
Table 4.2.17.C Smoking-Related Activities Questions and Metabolites with Overall3 Significant
Relationships
Q#
Description
Medium
Metabolites Analyzed13
Q1101
Current Smoker
Urine
TCPY
Q1102
Subject Smoked
Urine
TCPY
a Overall indicates that > 50% of the question/medium/chemical relationships were significant.
b See descriptions in Table 4.2.17.a.
The two questions in Table 4.2.17.c were from Krinsley (1998) whose study population was
focused on adults, but included children greater than 10 years of age. For Q1102, the two
relationships are significant; however, the direction of the relationship differs depending on
the other questions included in the regression analysis. For Q1101, the direction of the effect
is opposite of what is expected; that is, higher measurement levels are not associated with
currently smoking. Thus, the relationship between smoking and TCPY levels in urine
appears not to be supported by a plausible hypothesis (Table C.3.2.5).
4.2.5.6 Category 12: Work Exposure/Practices
Work exposure and work practices may lead to children's pesticide exposure if pesticides are
transmitted from the workplace to the home. The studies under review were primarily
environmental exposures studies conducted in agricultural communities with a focus on
children. If these studies had been strictly occupational exposure assessment studies, more
questions related to the work and family hygiene practices might have been included in these
studies. A list of the questions included in this category can be found in Appendix E.
The chemicals/metabolites measured in the study samples are azinphospmethyl and TCPY
(Table B.3.2.6.a); however, none of the chemicals/metabolites have significant relationships
with questions in this category (Table 4.2.18.a).
Table 4.2.18.a Codes and Descriptions for Chemicals/Metabolites with Significant Relationships for
Questions in the Work Exposure/Practices Category
Code(s)
Medium
Description
None
urine, dust
No chemicals
4-44
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The work exposure/practices category of questions includes 4 or 1% of the relationships
extracted from the relevant publications (Table 4.2.3). This category of questions falls into
the low-range occurrence level of relationships.
Table 4.2.18.b Distribution of Significant Relationships with Work Exposure/Practices Questions, by
Medium
Media/Question Relationships3
Significant13
Total
Medium Sampled
Chemicals/Metabolites Measured0
N
%d
N
Urine
None
0
0
1
Dust
None
0
0
3
Total
0
0
4
a Based on counts in Table B.3.2.6.a.
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.18.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Relationships between questions and urine sample concentrations account for 25% of
the relationships in this category (Table B.3.2.6.a).
• Seventy-five percent of the relationships are dust-based (Table B.3.2.6.b).
• None of the relationships with urine metabolites or dust chemicals are significant or
marginally significant (Table 4.2.18.b).
See Table 4.2.5 for tables with related information for the questions in this category.
Table 4.2.18.C Work Exposure/Practices Questions and Chemicals/Metabolites with Overall"
Significant Relationships
Q#
Description
Medium
Chemicals/Metabolites Analyzed
N/A
Not applicable
Urine, dust
Not applicable
a Overall indicates that > 50% of the question/medium/chemical relationships were significant.
No questions (Table 4.2.18.c) in this category have significant relationships with the urine
and dust measurements (Table C.3.2.6) for the studies considered in this review.
4.2.5.7 Summary of Results from Behavior Relationships
The six question categories under the behavior risk factor focus on the behaviors of the
household members or children in both macro and micro environments. Nine questions from
4-45
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
five of the six question categories are considered overall statistically significant (and
effective differentiators of exposure level) for the chemicals/metabolites noted. For each of
the question and chemical/metabolite combinations, the majority (> 50%) of the relationships
were statistically or marginally significant.
Table 4.2.19 Questions from Behavior Question Categories Considered Overall Statistically
Significant, by Medium
Medium
Q Category
Q #a
Q Description
Chemicals/
Metabolites'1
Dust
Family hygiene practices
Q1006
Work Clothes Worn Indoors
AZM, OPSUM
Q1009
Number of Weeks Since Last
Vacuuming
OPSUM
Urine
Subject's personal
characteristics
Q702
Age
DAP1, METHYL2
Q703
Ethnicity
1NAP, MDA
Q705
Income
1NAP, MDA, TCPY,
DMTP, DAP1
Child's behaviors
Q806
Loading from Hand Wipe
DAP1
Dietary behaviors
Q904
Organic Diet
METHYL2
Smoking-related
activities
Q1101
Current Smoker0
TCPY
Q1102
Subject Smoked0
TCPY
a For some of the significant relationships, the effect of the exposure factor was not in the direction expected.
See Appendix C for details on specific questions.
b Chemicals or metabolites for which > 50% of the relationships with the question were statistically or marginally
significant. (See "a" tables: Tables 4.2.13.a through 4.2.18.a for descriptions.)
c Included only in Krinsley (1998) whose study population was focused on adults, but included children greater
than 10 years of age.
Family hygiene practices were the strong differentiators for pesticide levels in dust
measurements; the subject's personal characteristics, the child's behaviors, and dietary
behaviors were the strong differentiators for the pesticide metabolite levels in urine (Table
4.2.19).
4.2.6 Presentation of Other Relationships
Several other relationships were tested in the studies under review, but these were difficult to
categorize. Two types of relationships are discussed here: exposure levels in populations (in
particular, in adults living with children), and health outcomes. This category included 13, or
2%, of the relationships identified in this review.
4-46
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
4.2.6.1 Category 13: Related Exposure Levels
Several studies examined the relationship between pesticide metabolite levels in adults and
children. It was hypothesized in these studies that adults living in the same environment as
children, and perhaps consuming similar foods, would exhibit similar metabolite levels. A
list of the questions included in this category can be found in Appendix E.
Table 4.2.20.a Codes and Descriptions for Metabolites with Significant Relationships for Questions in
the Related Exposure Levels Category
Code(s)
Medium3
Description13
DAP2
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one detectable measurement)
DAP3
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one high measurement)0
MTHL4, METHYL4
urine
DMP, DMTP
(at least one detectable measurement)
a Medium is noted as urine or other (any other medium sampled).
b DEP = diethylphosphate, DETP = diethylthiophosphate, DEDTP = diethyldithiophosphate
DMP = dimethylphosphate, DMTP = dimethylthiophosphate, DMDTP = dimethyldithiophosphate
c See definition of high measurement in Azaroff (1999).
The related exposure levels category of questions includes 5 or 1% of the relationships
extracted from the relevant publications. This category of questions falls into the low-range
occurrence level for relationships (Table 4.2.3). An association between adult and child
pesticide metabolite levels suggests similar exposure sources for these populations, and may
help in understanding how to reduce or prevent exposures. However, few studies have
examined this relationship.
Table 4.2.20.b Distribution of Significant Relationships with Related Exposure Levels Questions, by
Medium
Media/Question Relationships3
Medium Sampled
Significant13
Total
Metabolites Measuredc
N
%d
N
Urine
DAP2, DAP3, METHYL4
4
80
5
Total
4
80
5
a Based on counts in Table B.3.3.1.a.
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.20.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Only urine-based relationships were found for this category (Table B.3.3.1.a).
4-47
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
• Eighty percent of the relationships are significant (Table 4.2.20.b).
See Table 4.2.5 for tables with related information for the questions in this category.
Table 4.2.20.C Related Exposure Levels Questions and Metabolites with Overall3 Significant
Relationships
Q#
Description
Medium
Metabolites Analyzed13
Q1302
High Levels in Adult
Household Members
Urine
DAP2, DAP3, METHYL4
a Overall indicates that > 50% of the question/medium/chemical relationships were significant.
b See descriptions in Table 4.2.20.a
In most instances where information regarding the direction of the relationships is provided,
the significant medium/question relationships are in agreement with the expectation that
more adults in the household with high measurement levels is associated with a higher
measurement level for the children (Table C.3.3.1). Although there are a small number of
relationships for this question, it appears to be a useful predictor of DAP levels in urine.
4.2.6.2 Category 14: Health
In some of the studies under review investigators included general health status
questionnaires. It is not clear whether such questions were included to collect general health
status information on the population, or to explore specific hypotheses related to pesticide
exposure. For example, it is not immediately evident why pesticide exposure would be
associated with intestinal disease or ulcers, unless one considers a possible change in diet to
be associated with such diseases. Nevertheless, possible associations between health
outcomes and pesticide exposure metrics were tested in some instances. A list of the
questions included in this category can be found in Appendix E.
The only metabolite measured in the study samples for this category is TCPY (Table
B.3.3.2.a).
Table 4.2.21.a Codes and Descriptions for Metabolites with Significant Relationships for Questions in
the Health Category
Code(s)
Medium3
Description
TCPY
urine
3,5,6-T richloro-2-pyridinol
a Medium is noted as urine or other (any other medium sampled).
The health category of questions includes 8 or 1% of the relationships extracted from the
relevant publications (Table 4.2.3). This category of questions falls into the low-range
occurrence level. The possible association of TCPY metabolites in urine with health
outcomes does not imply causality in either direction. Associations that do not involve the
4-48
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
nervous system are generally not supported by a well-established hypothesis; however, the
OP pesticides may affect the nervous system such that other organ systems can also be
affected. Analyses of these types of possible associations should be considered exploratory.
Table 4.2.21.b Distribution of Significant Relationships with Health Questions, by Medium
Media/Question Relationships3
Medium Sampled
Significant13
Total
Chemicals/Metabolites Measured0
N
%d
N
Urine
TCPY
5
63
8
Total
5
63
8
a Based on counts in Table B.3.3.2.a.
b Significant (p < 0.05) and marginally significant (p < 0.10).
c See descriptions in Table 4.2.21.a.
d Percent of significant relationships for medium, that is, (N*100)/Total N.
The relationships in this question category can be summarized as follows:
• Only urine-based relationships were found for this category (Table B.3.3.2.a).
• Sixty-three percent of the relationships with TCPY are significant (Table 4.2.21.b).
See Table 4.2.5 for tables with related information for the questions in this category.
Table 4.2.21.C Health Questions and Metabolites with Overall11 Significant Relationships
Description
Medium
Metabolites Analyzed13
Q1403
Bowel Disease
Urine
TCPY
Q1405
Intestinal Disease
Urine
TCPY
Q1406
Ulcers
Urine
TCPY
a Overall indicates that > 50% of the question/medium/chemical relationships were significant.
b See descriptions in Table 4.2.21.a
In most of the significant relationships, not having the disease is associated with a higher
TCPY level than having the disease; however several of these analyses included the question
as one of several in a regression analysis, and the direction of the relationship cannot be
determined by the regression coefficient alone (Table C.3.3.2). These relationships did not
necessarily consider the health outcome of pesticide exposure; however, it is possible that
health status may signal changes in dietary or other behaviors that may affect exposure
levels. One example is the inclusion or exclusion of additional vegetables and fruits in the
diet.
4.2.6.3 Summary of Results from Other Relationships
4-49
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The two question categories under the "other" risk factor were difficult to categorize. Four
questions from the two question categories are considered overall statistically significant
(and effective differentiators of pesticide exposure levels) for TCPY. For each of the
question and chemical/metabolite combinations, the majority (> 50%) of the relationships
were statistically or marginally significant.
Table 4.2.22 Questions from Other Question Categories Considered Overall Statistically Significant,
by Medium
Medium
Q Category
Q#a
Q Description
Chemicals/
Metabolitesb
Urine
Related exposure levels
Q1302
High Levels in Adult
Household Members
DAP2, DAP3,
METHYL4
Health
Q1403
Bowel Disease
TCPY
Q1405
Intestinal Disease
TCPY
Q1406
Ulcers
TCPY
a For some of the significant relationships, the effect of the exposure factor was not in the direction expected.
See Appendix C for details on specific questions.
b Chemicals or metabolites for which > 50% of the relationships with the question were statistically or marginally
significant. (See "a" tables: Tables 4.2.21.a and 4.2.22.a for descriptions.)
Neither of the question categories showed differentiating capability for pesticide levels in
dust measurement levels (Table 4.2.22). Both categories have some questions that
differentiate pesticide metabolite levels in urine.
4.2.7 Summary of Results from Literature Review
Tables 4.2.12, 4.2.19, and 4.2.22 list the questions that are strong differentiators for the
chemical or metabolite levels for each of the three risk factors, source, behavior, and other.
Dust and urine measurements were found in 97% of the relationships. Measurements for the
other media were found in only two of the 20 publications: Sexton (2003) and Simcox
(1995). The relationships for each question and chemical/metabolite combination were
reviewed to determine the question's effectiveness for differentiating the exposure levels.
Not all question/chemical combinations were evaluated in the studies to the same extent.
The number of relationships evaluated with a question, especially when the questions are
used with more than one study population, gives additional credence to the question as a
potential differentiator for a specific chemical or metabolite. Generally the questions
showing the most effectiveness were:
• residential pesticide use (inside and outside)
• occupation of household members
• child's characteristics (age, ethnicity, income)
• family hygiene practices
• household dust.
4-50
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Several other questions also show some effectiveness:
• pets
• household location (urban vs non-urban)
• dietary behaviors (organic food)
• exposure levels of household members
• health status (diseases)
• smoking behaviors
• proximity to agricultural fields (for house dust only).
The number of relationships evaluated in the publications for this second group of questions
is small, indicating that their effectiveness has not been tested as extensively as for the
questions in the first group.
For urine measurements, questions showing usefulness as indicators of a child's pesticide
exposure level cover the areas of residential pesticide use both indoors and outdoors,
household occupation, subject's personal characteristics, family hygiene practices, and
smoking behavior. Each of these indicators seems plausible, in that such relationships have
been seen in previous investigations of environmental exposures (e.g., lead exposure in
children). The smoking questions appeared only in Krinsley (1998), whose study population
was focused on adults, but included children greater than 10 years of age. Although second-
hand smoke is noted as a significant predictor, the age of the majority of the study population
and the very limited transferability of any pesticides from the smoke makes this question less
effective for purposes of this project. For dust measurements, the questions showing
usefulness as indicators of a child's pesticide exposure level cover the areas of household
occupation, residential proximity to spraying, and family hygiene behavior. Each of these
indicators also seems plausible in terms of pesticides being present in the child's
environment. These questions represent potential exposure from the take-home pathway and
from agricultural pesticide spraying.
The set of question categories used in this report (Table 4.2.3) provide one perspective for
organizing the relationships. Three risk or exposure factors, related to the take-home or para-
occupational exposure pathway, were analyzed as separate categories in this report:
household occupation, family hygiene practices, and work exposure/practices. Household
occupation is considered a source that may result in measurable differences in children's
pesticide exposures. Most of the questions in this category involve the occupational status of
household members. The occupations considered were pesticide applicators, farm workers,
pesticide handlers, growers, and reference groups (non-agricultural workers). The
occupation of household workers within the agricultural sector produced a substantial
number of statistically significant relationships for urine and dust and the corresponding
DAPs and OP parent compound levels. For these relationships, occupation may represent a
surrogate for the actual exposure levels of household members employed in agriculture.
These workers become reservoirs for the chemicals to which they are exposed at work. They
subsequently transfer these chemicals into their homes and to their family members. This
para-occupational exposure pathway involves the transport of contaminants from the
4-51
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
workplace to the residence on a worker's clothing or person (Curl 2002). Children may be
exposed to agricultural chemicals through the take-home or para-occupational pathway and
their exposure levels are dependent on the occupational status, work, handling, and hygiene
practices of agricultural workers in their households.
Two other risk factors examined in this report also contribute to the para-occupational
exposure pathway. Family hygiene practices and work exposure/practices are considered
behavioral practices that may modify pesticide exposure to agricultural workers and their
family members. There were fewer relationships in these two categories because of the
nature of the studies analyzed. The studies under review were primarily environmental
exposures studies conducted in agricultural communities with a focus on children. If these
studies had been strictly occupational exposure assessment studies, more questions related to
the work and family hygiene practices might have been included in these studies. The
findings which produced significant results such as laundering practices, vacuuming, and
removal of work clothing and boots are also integral components required to fully understand
the para-occupational exposure pathway.
4.3 Children's Pesticide Exposure Study (Yuma Study)
The second approach for evaluating questions useful in differentiating children's pesticide
exposure levels was based on information available from a recent exposure study with this
goal. The Children's Pesticide Exposure Study collected questionnaire responses and sample
measurements from 152 households in Yuma County, Arizona. Throughout this section, the
Children's Pesticide Exposure Study will be referred to as the Yuma Study.
In the Yuma Study, one child in each household was considered the principal participant.
Urine samples were collected from the principal participant and a dust sample was collected
from the household. An interview was conducted regarding the household's characteristics
and activities, and the principal participant's behaviors. Siblings from the household were
included in the study if they were available and in the appropriate age range (2-11 years old);
however, only urine samples and minimal demographic information were collected for the
siblings. For 77 of the households, one sibling was included in the study, and for 15 of the
households two siblings were included.
The study design initially focused the selection of households on eight schools and of
principal participants in kindergarten and first grade (Table 4.3.1). Seventeen children
outside the initial school and grade list were included as principal participants. There were
five "other" school categories including none. There were five "other" grade categories
including: second grade, third grade, Head Start, preschool, and none.
4-52
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.3.1 Number of Yuma Study Principal Participants, by School and Grade Level
Grade of Principal Child
Total
School Attended by
Principal Child
Kindergarten
First Grade
Other Grades
School # 1
6
4
1
11
School #2
16
25
0
41
School # 3
5
8
1
14
School #4
4
6
0
10
School # 5
12
7
3
22
School # 6
3
8
0
11
School #7
16
10
1
27
School # 8
4
1
0
5
Other Schools
1
0
10
11
Total
67
69
16
152
Urine samples were measured for the six dialkylphosphates: DEP, DETP, DEDTP, DMP,
DMTP, and DMDTP. Unadjusted urinary metabolite measurements were available for 150
of the 152 principal participants; urinary metabolite measurements adjusted for creatinine
were available for 148 of the 152 principal participants. Household dust samples were
available for the 152 households. Dust samples were also collected from rooms where
principal participants attended class in the six schools that gave permission (Table 4.3.2).
School dust measurements were available for a subset of the schools and grades. These
samples cover 82% of the principal participants from the eight schools in the initial Yuma
Study design.
4-53
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.3.2 Number of Principal Participants Where Yuma Study Dust Samples Were Collected, by
School and Grade Level
Grade of Principal Child
Total
School Attended by
Principal Child
Kindergarten
First Grade
School # 1
6
4
10
School #2
16
25
41
School # 3
5
8
13
School #4
4
6
10
School # 6
3
8
11
School #7
16
10
26
Total
50
61
111
Household and school dust samples were measured for pesticides in the classes
organophosphates, organochlorines, permethrins, and miscellaneous (Table 4.3.3).
Table 4.3.3 Pesticides Measured in Yuma Study Household and School Dust Samples
atrazine
4,4-' DDT
methyl parathion3
azinphos-methyl3
diazinon3
methoxychlor
bendiocarb
dichlorvos3
metolachlor
bensulide
dicofol
pendimethalin
benzamide
dieldrin
cis-permethrin
captan
disulfoton3
trans-permethrin
carbaryl
endosulfan 1
o-phenylphenol
carbofuran
endosulfan 2
phorate3
alpha-chlordane
ethyl parathion3
prometryn
gamma-chlordane
folpet
propoxur
chlorpyrifos3
fonophos3
simazine
chlorthal-dimethyl
heptachlor
terbufos3
cy-permethrin
hexachlorobenzene
trifluralin
4,4-' DDD
lindane
4,4-' DDE
malathion3
a Organophosphorous (OP) pesticides
4-54
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Relationships found in the Yuma Study are described in two segments. Section 4.3.1
describes the relationships that are relevant to the interests of this project, based on the Yuma
Study report (CDC 2002). Section 4.3.2 and Appendix G describe the relationships
identified from the data mining analysis. Both analysis paths use the same study data set, but
consider different subsets of the Yuma Study participants, and performed analyses for
different purposes. These differences should be taken into consideration when comparing
results from the two approaches.
4.3.1 Relationships Explored in the Yuma Study Report
The Yuma Study analyzed potential risk factors, based on the questionnaire responses, for
children in a household, that is, the principal participant and any siblings. The objective of
the study was to determine the effect and levels of pesticide exposure on children living or
attending schools near pesticide-treated fields (CDC 2002). A child's exposure level was
determined by the level of pesticide metabolites in the urine. The study's report uses the
terms risk factors, associations, and borderline, and those terms will be used in section 4.3.1
when describing the report's results. For purposes of this report, the terms risk (or exposure)
factors and questions, association and relationship, and borderline and marginal (regarding
statistical significance) are interchangeable in section 4.3.
4.3.1.1 Relationships Between Questions and DAP Metabolites
For the Yuma Study report (CDC 2002), the children's exposure levels were evaluated with
regression models controlled for intra-household correlation with household as the repeated
measurement. The potential risk factors selected for analysis were the subset of the full set
of questions that could be applied to, and were available for, siblings as well as principal
participants. These factors included the child's physical characteristics and household
characteristics or practices. Child-specific behaviors were not used for the analyses because
they were not collected on any siblings. The pesticide metabolite concentrations were log
transformed to better meet the normality assumptions of the analyses and generalized
estimating equations (SAS version 8.2, SAS Institute, Cary, NC) were used to measure the
associations. Discussions about whether to analyze urinary metabolites in children as
adjusted or unadjusted for creatinine can be found in the literature, e.g., O'Rourke (2000).
4-55
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.3.4 Results of Regression Models" with DMOP and DEOP, Unadjusted and Adjusted for
Creatinine, for 152 Households
Risk Factor
DMOPb unadjc
DMOPb adjd
DEOP6 unadjc
DEOP6 adjd
Slope'
Slope'
Slope'
Slope'
(p-value)g
(p-value)9
(p-value)9
(p-value)9
Sex of participant
-0.07
-0.21
0.01
-0.12
(0.68)
(0.25)
(0.94)
(0.30)
Age of participant
-0.02
-0.06
-0.01
-0.05
(0.70)
(0.23)
(0.71)
(0.21)
Size of participant
0.00
0.00
-0.03
-0.02
(0.98)
(0.94)
(0.40)
(0.49)
Use of lice shampoo in last
0.06
-0.01
-0.07
-0.13
year
(0.78)
(0.98)
(0.67)
(0.56)
Distance from home to
0.18
-0.05
0.00
-0.23
agricultural field
(0.36)
(0.82)
(0.99)
(0.12)
Use of pesticides inside home
0.03
0.27
0.23
0.45
in last month
(0.87)
(0.16)
(0.06)
(0.00)
Use of pesticides outside of
0.19
0.31
0.14
0.24
home in last month
(0.31)
(0.11)
(0.25)
(0.10)
Father working in agriculture
-0.12
-0.19
-0.03
-0.11
(0.54)
(0.33)
(0.83)
(0.49)
Mother working in agriculture
0.27
0.19
0.44
0.39
(0.54)
(0.68)
(0.06)
(0.18)
Father or mother working in
0.09
-0.22
0.04
-0.24
agriculture
(0.84)
(0.63)
(0.86)
(0.42)
Other adult in house working in
-0.19
-0.30
-0.23
-0.32
agriculture
(0.46)
(0.22)
(0.23)
(0.04)
Father, mother or other adult
-0.23
0.46
0.00
-0.20
working in agriculture
(0.56)
(0.25)
(0.99)
(0.46)
a Regression model included all participating children controlling for intra-household correlation with household
as a repeated measure. Slope and p-value provided. N varies according to available responses for a risk
factor.
b DMOP is a summary variable made from summing molar weights of DMP, DMTP, and DMDTP.
(Concentrations < Limit of detection (LOD) were replaced with LOD/2.)
c Unadjusted for creatinine (ug/l urine).
d Adjusted for creatinine (ug/g Creatinine).
e DEOP is a summary variable made from summing molar weights of DEP, DETP, and DEDTP. (Concentrations
< LOD were replaced with LOD/2.)
f The slope estimates the increase in the pesticide level per unit increase in the independent variable. Slopes
associated with statistically significant p-values (p < 0.05) are in bold italics. Slopes associated with borderline
statistically significant p-values (0.05 < p < 0.10) are in bold. For questions answered by a yes or no, a yes
response was assigned a value of 1 and a no response was assigned a value of 2.
g Statistically significant p-values (<0.05) are in bold italics. Borderline statistically significant p-values (0.05 < p
<0.10) are in bold.
No risk factors were found to be associated with DMOP, the sum of methylated DAPs (Table
4.3.4). For DEOP, the sum of ethylated DAPs, the strongest association is with the questions
for recent pesticide use in the home. Secondary associations are identified with questions
about the mother or another adult (not father) working in agriculture, which may relate to
4-56
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
the level of daily interactions between these adults and the children. The coding assigned to
the yes and no responses was 1 and 2, respectively. Thus, negative slopes indicate that the
"yes" respondents have a higher measurement level than the "no" respondents, and positive
slopes indicate that the "yes" respondents have a lower measurement level than the "no"
respondents. Some of the slopes are in the direction expected (negative slope) based on
current knowledge, while others like recent pesticide use in the home or mother working in
agriculture appear to be in the reverse direction (positive slope). Alternatively, the reverse
direction for mother working in agriculture may be a surrogate measure for time away from
home rather than the take-home pathway.
Statistical analyses were also performed on the individual DAPs, and in a few instances, risk
factors that are significant or marginally significant for an individual DAP metabolite are not
significant for the metabolite sum. One such example is recent pesticide use in the home for
methylated DAPs. There is a significant association for DMP, and no significant association
for DMTP, DMDTP and DMOP (Tables 4.3.4 and 4.3.5). This might indicate the impact of
combining associations having different directions of association or of combining strong and
weak associations.
Table 4.3.5 Results of Regression Models" with Individual DAP Metabolites, Unadjusted and Adjusted
for Creatinine, for 152 Households
Risk factor
DMP
Slopeb
(p-value)c
DMTP
Slopeb
(p-value)c
DMDTP
Slopeb
(p-value)c
DEP
Slopeb
(p-value)c
DETP
Slopeb
(p-value)c
DEDTP
Slopeb
(p-value)c
Unadjusted for Creatinine11
Sex of participant
-0.01
(0..96)
-0.15
(0.46)
0.02
(0.93)
-0.02
(0.85)
0.00
(0.97)
0.12
(0.31)
Age of participant
-0.02
(0.69)
0.05
(0.44)
0.04
(0.62)
-0.03
(0.41)
-0.00
(0.87)
0.07
(0.03)
Size of participant
-0.05
(0.27)
0.05
(0.44)
-0.03
(0.79)
-0.04
(0.26)
-0.02
(0.62)
-0.00
(0.97)
Use of lice shampoo in last
year
-0.01
(0.95)
0.26
(0.41)
-0.12
(0.78)
-0.07
(0.73)
-0.14
(0.19)
0.09
(0.70)
Distance from home to
agricultural field
0.08
(0.66)
0.10
(0.72)
0.31
(0.37)
-0.07
(0.63)
-0.00
(0.99)
0.30
(0.02)
Use of pesticides inside
home in last month
0.39
(0.02)
-0.04
(0.87)
-0.12
(0.70)
0.31
(0.03)
0.19
(0.09)
-0.22
(0.13)
Use of pesticides outside of
home in last month
0.12
(0.49)
0.17
(0.52)
0.16
(0.64)
0.18
(0.22)
0.06
(0.57)
-0.21
(0.16)
Father working in
agriculture
-0.30
(0.12)
-0.02
(0.95)
0.16
(0.65)
-0.13
(0.41)
0.07
(0.58)
0.27
(0.11)
Mother working in
agriculture
0.67
(0.08)
0.32
(0.57)
1.27
(0.10)
0.39
(0.16)
0.28
(0.22)
0.87
(0.08)
Father or mother working in
agriculture
-0.33
(0.32)
0.09
(0.88)
1.46
(0.04)
-0.08
(0.77)
0.09
(0.68)
0.53
(0.16)
Other adult in house
working in agriculture
-0.21
(0.39)
-0.64
(0.11)
0.14
(0.77)
-0.34
(0.08)
-0.16
(0.29)
0.33
(0.30)
4-57
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Risk factor
DMP
Slopeb
(p-value)c
DMTP
Slopeb
(p-value)c
DMDTP
Slopeb
(p-value)c
DEP
Slopeb
(p-value)c
DETP
Slopeb
(p-value)c
DEDTP
Slopeb
(p-value)c
Father, mother or other
adult working in agriculture
-0.19
(0.56)
-0.67
(0.25)
0.44
(0.51)
-0.06
(0.83)
0.01
(0.97)
0.35
(0.30)
Adjusted for Creatinine6
Sex of participant
-0.14
(0.43)
-0.29
(0.18)
-0.10
(0.71)
-0.15
(0.30)
-0.14
(0.16)
-0.01
(0.97)
Age of participant
-0.05
(0.35)
-0.00
(0.98)
-0.00
(0.97)
-0.07
(0.14)
-0.04
(0.20)
0.03
(0.31)
Size of participant
-0.05
(0.37)
0.06
(0.39)
-0.02
(0.86)
-0.04
(0.39)
-0.02
(0.63)
0.00
(0.93)
Use of lice shampoo in last
year
-0.07
(0.82)
0.17
(0.65)
-0.20
(0.63)
-0.13
(0.63)
-0.20
(0.19)
0.03
(0.89)
Distance from home to
agricultural field
-0.14
(0.53)
-0.18
(0.53)
0.05
(0.89)
-0.31
(0.08)
-0.22
(0.05;
0.06
(0.67)
Use of pesticides inside
home in last month
0.61
(0.00)
0.25
(0.33)
0.15
(0.63)
0.53
(0.00)
0.40
(0.00)
0.01
(0.94)
Use of pesticides outside of
home in last month
0.21
(0.32)
0.34
(0.19)
0.31
(0.33)
0.28
(0.11)
0.15
(0.18)
-0.10
(0.47)
Father working in
agriculture
-0.38
(0.09)
-0.10
(0.71)
0.08
(0.81)
0.22
(0.25)
-0.00
(0.99)
0.19
(0.23)
Mother working in
agriculture
0.63
(0.17)
0.18
(0.75)
1.14
(0.14)
0.33
(0.34)
0.23
(0.35)
0.79
(0.11)
Father or mother working in
agriculture
-0.60
(0.14)
-0.30
(0.63)
1.10
(0.11)
-0.37
(0.25)
-0.17
(0.42)
0.23
(0.57)
Other adult in house
working in agriculture
-0.30
(0.22)
-0.76
(0.03)
0.02
(0.96)
-0.44
(0.02)
-0.25
(0.02)
0.23
(0.44)
Father, mother or other
adult working in agriculture
-0.37
(0.33)
-0.99
(0.07)
0.14
(0.82)
-0.27
(0.39)
-0.17
(0.38)
0.13
(0.71)
a Regression model included all participating children controlling for intra-household correlation with household
as a repeated measure. Slope and p-value provided. N varies according to available responses for a risk
factor. (Concentrations < LOD were replaced by LOD/2.)
b The slope estimates the increase in the pesticide level per unit increase in the independent variable. Slopes
associated with statistically significant p-values (p < 0.05) are in bold italics. Slopes associated with borderline
statistically significant p-values (0.05 < p < 0.10) are in bold. For questions answered by a yes or no, a yes
response was assigned a value of 1 and a no response was assigned a value of 2.
c Statistically significant p-values (<0.05) are in bold italics. Borderline statistically significant p-values (0.05 < p
<0.10) are in bold.
d ug/l urine.
e ug/g Creatinine.
When looking at the individual DAPs, associations of the methylated DAPs occur with the
questions about recent pesticide use in the home and household members working in
agriculture (Table 4.3.5). As expected, based on the significant associations for DEOP
(Table 4.3.4), these risk factors are associated with some of the individual ethylated DAPs,
DEP and DETP (Table 4.3.5). Distance from home to agricultural fields also shows
significant associations with DEP and DETP (Table 4.3.5). The significant associations with
these two DAPs are likely the basis for the significant relationships with DEOP (Table 4.3.4).
4-58
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The positive valued slopes for the questions with yes/no responses show an association
opposite of what might be expected for the risk factor, that is, a yes response is predicted to
have a lower measurement level than a no response.
Statistical analyses were also performed on the risk factor distance from household to
agricultural field. The households were divided into two groups, those whose distance to the
agricultural fields was < 250 feet, and those whose distance was > 250 feet. The question
distance from home to the agricultural fields for principal participants shows statistical
significance (borderline) only for the DEDTP metabolite, with distances closer to the
agricultural fields having higher concentration values, as expected (Table 4.3.6).
Table 4.3.6 Results Comparing Distance from Home to Agricultural Fields with Six DAP Metabolites,
Unadjusted and Adjusted for Creatinine, for Principal Participants
Analyte
(unadjusted, ug/l urine)
(adjusted, ug/g Creatinine)
Ni
Mean and range of
urine samples in area
< 250 feet3
n2
Mean and range of
urine samples in
area > 250 feet3
t-
statistic
P-
value
DMP (unadjusted)
108b
4.06 (0.29-29.00)
42
3.33 (0.29- 14.00)
1.00
0.32
DMP (adjusted)
107°
7.21 (0.18-60.13)
41d
6.67(0.28-49.82)
0.31
0.76
DMTP (unadjusted)
108b
13.43 (0.09-200.00)
42
11.10 (0.09- 120.00)
0.48
0.63
DMTP (adjusted)
107°
21.15 (0.09-409.00)
41d
18.10 (0.30-223.33)
0.36
0.72
DMDTP (unadjusted)
108b
5.61 (0.04- 160.00)
42
4.02 ( 0.04-51.00)
0.67
0.50
DMDTP (adjusted)
107°
8.78 ( 0.02-215.25)
41d
6.71 (0.03-94.29)
0.53
0.59
DEP (unadjusted)
108b
3.24 (0.59-21.00)
42
2.92 (0.55- 11.00)
0.57
0.57
DEP (adjusted)
107°
5.37 (0.41 -40.32)
41d
5.61 (0.81 -39.15)
-0.19
0.85
DETP (unadjusted)
108b
1.32 (0.50-5.70)
42
1.73 (0.50-9.2)
-1.32
0.19
DETP (adjusted)
107°
2.04 (0.24-9.09)
41d
2.42 (0.67- 10.53)
-1.16
0.25
DEDTP (unadjusted)
108b
0.50 (0.08- 14.00)
42
0.24 (0.08- 1.10)
1.71
0.09
DEDTP (adjusted)
107°
0.64 (0.07- 11.75)
41d
0.39 (0.06- 1.42)
1.62
0.11
a Concentrations < LOD were replaced with LOD/2.
b Results from urine samples of two principal participants were not available.
c Results from urine samples of two principal participants were not available and creatinine level from urine
sample of one principal participant was not reported.
d Creatinine level from urine sample of one principal participant was not reported.
4.3.1.2 Relationships Between Dust Measurements and DAP Metabolites
The Yuma Study (CDC 2002) evaluated associations between the DAP metabolite levels and
levels of the ten pesticides most detected in the household dust samples. The associations
4-59
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
were evaluated with the ethylated (DEOP) and methylated (DMOP) DAP sums, adjusted and
unadjusted for creatinine (Table 4.3.7), and for the six individual DAPs, adjusted and
unadjusted for creatinine (Table 4.3.8).
Table 4.3.7 Results of Regression Models" with DMOP and DEOP, Unadjusted and Adjusted for
Creatinine, and the Ten Pesticides Most Detected in Household Dust Samples for 152
Households
Household dust pesticide
DMOPb unadjc
DMOPb adjd
DEOP6 unadjc
DEOP adjd
Slope'
Slope'
Slope'
Slope'
(p-value)9
(p-value)9
(p-value)9
(p-value)9
Trans-permethrin
0.00
0.00
0.00
0.00
(0.23)
(0.03)
(0.07)
(0.06)
Cis-permethrin
0.00
0.00
0.00
0.00
(0.00)
(0.00)
(0.00)
(0.03)
Chlorpyrifos
-0.00
-0.00
0.00
0.00
(0.00)
(0.03)
(0.20)
(0.01)
Diazinon
-0.00
-0.00
-0.00
0.00
(0.32)
(0.06)
(0.00)
(0.14)
Propoxur
-0.00
-0.00
-0.00
0.00
(0.00)
(0.61)
(0.00)
(0.01)
O-phenylphenol
-0.00
0.00
0.00
-0.00
(0.92)
(0.92)
(0.59)
(0.44)
Cy-permethrin
-0.00
-0.00
0.00
0.00
(0.07)
(0.10)
(0.06)
(0.00)
4,4'-DDT
-0.00
-0.00
-0.00
0.00
(0.01)
(0.49)
(0.18)
(0.54)
Gamma-chlordane
-0.00
0.00
-0.00
-0.00
(0.66)
(0.84)
(0.05)
(0.07)
Carbaryl
0.00
0.00
-0.00
-0.00
(0.00)
(0.00)
(0.09)
(0.00)
a Regression model included all participating children controlling for intra-household correlation with household
as a repeated measure. Slope and p-value provided. N varies according to available responses for a risk
factor.
b DMOP is a summary variable made from summing molar weights of DMP, DMTP, and DMDTP.
(Concentrations < LOD were replaced with LOD/2.)
c Unadjusted for creatinine (ug/l urine).
d Adjusted for creatinine (ug/g Creatinine).
e DEOP is a summary variable made from summing molar weights of DEP, DETP, and DEDTP. (Concentrations
< LOD were replaced with LOD/2.)
f The slope estimates the increase in the pesticide level per unit increase in the independent variable. Slopes
associated with statistically significant p-values (p < 0.05) are in bold italics. Slopes associated with borderline
statistically significant p-values (0.05 < p < 0.10) are in bold.
g Statistically significant p-values (<0.05) are in bold italics. Borderline statistically significant p-values (0.05 < p
<0.10) are in bold.
4-60
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.3.8 Results of Regression Models" with Individual DAP Metabolites, Unadjusted and Adjusted
for Creatinine, and the Ten Pesticides Most Detected in Household Dust Samples for 152
Households
Household dust
DMP
DMTP
DMDTP
DEP
DETP
DEDTP
pesticide
Slopeb
Slopeb
Slopeb
Slopeb
Slopeb
Slopeb
(p-value)c
(p-value)c
(p-value)c
(p-value)c
(p-value)c
(p-value)c
Unadjusted for Creatinined
Trans-permethrin
0.00
0.00
0.00
0.00
0.00
-0.00
(0.10)
(0.171)
(0.07)
(0.01)
(0.44)
(0.60)
Cis-permethrin
0.00
0.00
0.00
0.00
0.00
-0.00
(0.01)
(0.00)
(0.00)
(0.00)
(0.07)
(0.76)
Chlorpyrifos
-0.00
-0.00
0.00
0.00
0.00
-0.00
(0.22)
(0.05)
(0.24)
(0.20)
(0.10)
(0.13)
Diazinon
0.00
-0.00
-0.00
0.00
0.00
0.00
(0.33)
(0.13)
(0.12)
(0.00)
(0.00)
(0.74)
Propoxur
0.00
-0.00
-0.00
-0.00
-0.00
-0.00
(0.00)
(0.42)
(0.62)
(0.00)
(0.07)
(0.08)
O-phenylphenol
0.00
-0.00
-0.00
-0.00
-0.00
-0.00
(0.30)
(0.94)
(0.29)
(0.72)
(0.89)
(0.00)
Cy-permethrin
-0.00
-0.00
-0.00
0.00
0.00
-0.00
(0.52)
(0.31)
(0.68)
(0.08)
(0.02)
(0.04)
4,4'-DDT
-0.00
-0.00
-0.00
-0.00
-0.00
-0.00
(0.43)
(0.03)
(0.00)
(0.48)
(0.02)
(0.64)
Gamma-chlordane
0.00
0.00
0.00
-0.00
-0.00
0.00
(0.32)
(0.62)
(0.52)
(0.05)
(0.09)
(0.97)
Carbaryl
-0.00
0.00
-0.00
-0.00
-0.00
-0.00
(0.30)
(0.01)
(0.85)
(0.17)
(0.01)
(0.89)
Adjusted for Creatinine6
Trans-permethrin
0.00
0.00
0.00
0.00
0.00
-0.00
(0.17)
(0.01)
(0.00)
(0.07)
(0.00)
(0.82)
Cis-permethrin
0.00
0.00
0.00
0.00
0.00
-0.00
(0.13)
(0.00)
(0.00)
(0.05)
(0.00)
(0.98)
Chlorpyrifos
-0.00
-0.00
0.00
0.00
0.00
-0.00
(0.14)
(0.24)
(0.20)
(0.03)
(0.00)
(0.41)
Diazinon
-0.00
-0.00
-0.00
0.00
0.00
0.00
(0.89)
(0.03)
(0.02)
(0.23)
(0.03)
(0.41)
Propoxur
0.00
0.00
0.00
0.00
0.00
0.05
(0.49)
(0.02)
(0.26)
(0.16)
(0.00)
(0.03)
O-phenylphenol
-0.00
0.00
-0.00
-0.00
-0.00
-0.00
(0.24)
(0.39)
(0.48)
(0.55)
(0.59)
(0.00)
Cy-permethrin
0.00
-0.00
0.00
0.00
0.00
-0.00
(0.73)
(0.48)
(0.99)
(0.00)
(0.02)
(0.51)
4,4'-DDT
0.00
-0.00
-0.00
-0.00
-0.00
-0.00
(0.58)
(0.41)
(0.01)
(0.44)
(0.98)
(0.44)
Gamma-chlordane
0.00
0.00
0.00
-0.00
-0.00
0.00
(0.47)
(0.77)
(0.62)
(0.08)
(0.04)
(0.82)
4-61
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Household dust
DMP
DMTP
DMDTP
DEP
DETP
DEDTP
pesticide
Slopeb
Slopeb
Slopeb
Slopeb
Slopeb
Slopeb
(p-value)c
(p-value)c
(p-value)c
(p-value)c
(p-value)c
(p-value)c
Carbaryl
-0.00
0.00
0.00
0.00
-0.00
-0.00
(0.10)
(0.05)
(0.57)
(0.00)
(0.00)
(0.06)
a Regression model included all participating children controlling for intra-household correlation with household
as a repeated measure. Slope and p-value provided. N varies according to available responses for a risk
factor.
b The slope estimates the increase in the pesticide level per unit increase in the independent variable. Slopes
associated with statistically significant p-values (p < 0.05) are in bold italics. Slopes associated with borderline
statistically significant p-values (0.05 < p < 0.10) are in bold.
c Statistically significant p-values (<0.05) are in bold italics. Borderline statistically significant p-values (0.05 < p
<0.10) are in bold.
d ug/l urine.
e ug/g Creatinine.
The Yuma Study (CDC 2002) also evaluated associations between the DAP urinary
metabolite levels and the levels of the seven pesticides most detected across the household
and school dust samples. The associations were evaluated with the ethylated (DEOP) and
methylated (DMOP) DAP sums, adjusted and unadjusted for creatinine (Table 4.3.9), and for
the six individual DAPs, adjusted and unadjusted for creatinine (Table 4.3.10). The
statistical analyses were performed only for the principal participants whose
school/classroom dust was measured, that is, in only six of the eight schools from the initial
study design (Table 4.3.2).
Table 4.3.9 Results of Regression Models" with DMOP and DEOP, Unadjusted and Adjusted for
Creatinine, and the Seven Pesticides Most Detected in Household and School Dust Samples,
for Principal Participants
Household and school dust
DMOPb unadjc
DMOPb adjd
DEOP6 unadjc
DEOP6 adjd
pesticide
Slope'
Slope'
Slope'
Slope'
(p-value)9
(p-value)9
(p-value)9
(p-value)9
Trans-permethrin (nh = 80/n' =79)
0.00
0.00
0.00
0.00
(0.00)
(0.00)
(0.00)
(0.00)
Cis-permethrin (nh = 82/n' =81)
0.00
0.00
0.00
0.00
(0.00)
(0.00)
(0.00)
(0.00)
Chlorpyrifos (nh = 110/n1 =108)
-0.00
-0.0
0.00
0.00
(0.15)
(0.28)
(0.45)
(0.18)
Diazinon (nh = 110/n' =108)
-0.00
-0.00
-0.00
0.00
(0.14)
(0.01)
(0.00)
(0.11)
Propoxur (nh = 110/n' =108)
-0.00
0.00
-0.00
0.00
(0.56)
(0.32)
(0.00)
(0.20)
O-phenylphenol (nh = 110/n' =108)
-0.00
0.00
-0.00
-0.00
(0.74)
(0.76)
(0.97)
(0.96)
Cy-permethrin (nh = 79/n' =77)
-0.00
-0.00
-0.00
-0.00
(0.27)
(0.06)
(0.01)
(0.00)
a Regression model included only principal participants where school dust samples were collected from their
classrooms. Slope and p-value provided.
4-62
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
b DMOP is a summary variable made from summing molar weights of DMP, DMTP, and DMDTP.
(Concentrations < LOD were replaced with LOD/2.)
c Unadjusted for creatinine (ug/l urine).
d Adjusted for creatinine (ug/g Creatinine).
e DEOP is a summary variable made from summing molar weights of DEP, DETP, and DEDTP. (Concentrations
< LOD were replaced with LOD/2.)
f The slope estimates the increase in the pesticide level per unit increase in the independent variable. Slopes
associated with statistically significant p-values (p < 0.05) are in bold italics. Slopes associated with borderline
statistically significant p-values (0.05 < p < 0.10) are in bold.
g Statistically significant p-values (<0.05) are in bold italics. Borderline statistically significant p-values (0.05 < p
<0.10) are in bold.
h Number of measurements unadjusted for creatinine.
' Number of measurements adjusted for creatinine.
4-63
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.3.10 Results of Regression Models" with Individual DAP Metabolites, Unadjusted and
Adjusted for Creatinine, and the Seven Pesticides Most Detected in Household and
School Dust Samples, for Principal Participants
Household and school
dust pesticide
DMP
Slopeb
(p-value)c
DMTP
Slopeb
(p-value)c
DMDTP
Slopeb
(p-value)c
DEP
Slopeb
(p-value)c
DETP
Slopeb
(p-value)c
DEDTP
Slopeb
(p-value)c
Unadjusted for Creatinine11
Trans-permethrin (n = 80)
0.00
(0.01)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
Cis-permethrin (n = 82)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
0.00
(0.20)
Chlorpyrifos (n = 110)
-0.00
(0.00)
-0.00
(0.53)
0.00
(0.29)
0.00
(0.61)
0.00
(0.15)
-0.00
(0.51)
Diazinon (n = 110)
0.00
(0.84)
-0.00
(0.33)
-0.00
(0.33)
0.00
(0.00)
-0.00
(0.00)
-0.00
(0.00)
Propoxur (n=110)
-0.00
(0.00)
-0.00
(0.22)
0.00
(0.07)
-0.00
(0.00)
-0.00
(0.87)
-0.00
(0.31)
O-phenylphenol (n = 110)
-0.00
(0.41)
-0.00
(0.61)
-0.00
(0.65)
-0.00
(0.79)
-0.00
(0.63)
-0.00
(0.01)
Cy-permethrin (n = 79)
-0.00
(0.57)
0.00
(0.22)
-0.00
(0.01)
0.00
(0.00)
0.00
(0.01)
0.00
(0.01)
Adjusted for Creatinine6
Trans-permethrin (n = 79)
0.00
(0.01)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
0.00
(0.63)
Cis-permethrin (n = 81)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
0.00
(0.81)
Chlorpyrifos (n= 108)
-0.00
(0.01)
-0.00
(0.53)
0.00
(0.29)
0.00
(0.40)
0.00
(0.01)
0.00
(0.77)
Diazinon (n= 108)
-0.00
(0.21)
-0.00
(0.62)
0.00
(0.22)
0.00
(0.24)
0.00
(0.00)
-0.00
(0.00)
Propoxur (n= 108)
-0.00
(0.07)
0.00
(0.03)
0.00
(0.07)
0.00
(0.58)
0.00
(0.00)
0.00
(0.04)
O-phenylphenol (n= 108)
-0.00
(0.38)
0.00
(0.52)
0.00
(0.76)
0.00
(0.91)
-0.00
(0.40)
-0.00
(0.65)
Cy-permethrin (n = 77)
-0.00
(0.47)
0.00
(0.38)
-0.00
(0.00)
0.00
(0.00)
0.00
(0.01)
-0.00
(0.00)
a Regression model included only principal participants where school dust samples were collected from their
classrooms. Slope and p-value provided. (Concentrations < LOD were replaced by LOD/2.)
b The slope estimates the increase in the pesticide level per unit increase in the independent variable. Slopes
associated with statistically significant p-values (p < 0.05) are in bold italics. Slopes associated with borderline
statistically significant p-values (0.05 < p < 0.10) are in bold.
c Statistically significant p-values (<0.05) are in bold italics. Borderline statistically significant p-values (0.05 < p
<0.10) are in bold.
d ug/l urine.
e ug/g Creatinine.
Many associations between the DAP urinary metabolites and the most detected OP pesticides
in household and school dust were found; however, the regression coefficients are very small
4-64
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
(Tables 4.3.7 - 4.3.10). Thus, the associations may be statistically significant, but may not
necessarily be practically significant. The report authors note:
The regression models in which the slopes were small but were statistically
significant may suggest either that a) true associations existed, but the
numbers of significance were less than the numbers measured in the
statistical programs or b) the associations were meaningless and based
solely [on] the probability of finding statistical significance if enough tests
were run. (CDC 2002)
Some of the most detected pesticides in household and school dust samples are other than OP
pesticides and associations between the DAPs and pesticides exist regardless of the class of
pesticide. These relationships may indicate heavy pesticide use, although they do not
correspond to the metabolites found.
4.3.1.3 Summary of Results
The analyses from the Yuma Study report (CDC 2002) show that the most significant
associations with DAP urinary metabolites are questions about recent pesticide use in the
home, adult household members working in agriculture, and distance from home to
agricultural fields. Table 4.3.11 summarizes the significant associations between questions
and DAP metabolites based on Tables 4.3.4, 4.3.5, and 4.3.6. Not all of the significant
associations are in the directions expected. In some cases, it is either the concentration
adjusted, or unadjusted, for creatinine that is significant, but not both. O'Rourke (2000) and
Barr (2004) include discussions about differences in the use of the two measures for
statistical analysis.
Questions about recent pesticide use in the home, take-home pathway from the mother or
other adult working in agriculture, and distance from home to agricultural fields seem to be
the most useful in predicting ethylated DAP exposure measurements in urine. Questions
about the father working in agriculture seem to be somewhat useful in predicting methylated
DAP exposure measurements, which is to be expected since methylated OPs are commonly
used in agriculture. In many of the associations, however, the direction of the association is
the opposite of what is expected (Table 4.3.11). Many strong associations are shown
between the pesticides most detected in household and school dust samples and the DAP
metabolites. Most of the significant regression coefficients are in the direction expected for
the association based on current knowledge.
4-65
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.3.11 Questions and DAP Metabolites with Significant" Relationships in the Yuma Study Based on Tables 4.3.4,4.3.5, and 4.3.6
Questions
DAP Metabolites13
DEP
DETP
DEDTP
DEOPc
DMP
DMTP
DMDTP
DMOPd
ue
a®
u
a
u
a
u
a
u
a
u
a
u
a
u
a
Age of participant
X
Used pesticide inside home in last month
X
X
X
X
X
X
X
X
Distance from home to agricultural field
Y
Y
X
Mother working in agriculture
X
X
X
Father working in agriculture
Y
Father or mother working in agriculture
X
Other adult in house working in agriculture
Y
Y
Y
Y
Y
Father, mother or other adult working in
agriculture
Y
a Statistically significant or borderline significant.
b DEP = diethylphosphate, DETP = diethylthiophosphate, DEDTP = diethyldithiophosphate
DMP = dimethylphosphate, DMTP = dimethylthiophosphate, DMDTP = dimethyldithiophosphate.
0 DEOP is a summary variable made from summing molar weights of DEP, DETP, and DEDTP. (Concentrations < LOD were replaced with LOD/2.)
d DMOP is a summary variable made from summing molar weights of DMP, DMTP, and DMDTP. (Concentrations < LOD were replaced with LOD/2.)
6 u = unadjusted for creatinine,
a = adjusted for creatinine.
X = occurrence of risk factor associated with lower levels of pesticide metabolite.
Y = occurrence of risk factor associated with higher levels of pesticide metabolite.
4-66
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
4.3.2 Results from the Data Mining Approach
In the Yuma Study report (CDC 2002), hypotheses were defined a priori which set the
direction for the data analyses performed. Based on these hypotheses, the risk factors were
analyzed individually with the urinary DAP metabolites measured. The data mining
approach provides options for exploring the Yuma Study data for relationships between risk
factors and exposure levels without specifying a priori views, that is, based on relationships
that exist in the data. In some cases, the risk factors were analyzed in groups and interactions
between risk factors in the relationships were considered. In this section, the term
relationships will be used instead of the term associations used in Section 4.3.1, to be
consistent with the use of relationships in Section 4.2. Their meaning, however, is
considered interchangeable.
4.3.2.1 Subpopulation Selected for Analysis
Since questionnaire responses and school dust measurements were not collected for siblings,
the analyses reported here were performed only on data from principal participants. To
further limit the impact of factors relating to children not defined in the initial study design,
the principal participants from kindergarten and first grade and from the initial eight schools,
were selected as the core set of participants for the data mining analysis. Comparisons of
questionnaire responses between the 135 core participants (Table 4.3.12), and the other 17
participants (Table 4.3.1) showed little difference. Depending on the particular statistical
analysis, and the urinary metabolite or sum selected as the dependent variable, up to five
additional core participants were excluded because of a lack of, or suspicions about, the
urinary metabolite measurements. Subsequent use of the phrase principal child will denote
the core principal participant children described above.
Table 4.3.12 Number of Yuma Study Core Principal Participants, by School and Grade Level
Grade of Principal Child
Total
School Attended by
Principal Child
Kindergarten
First Grade
School # 1
6
4
10
School #2
16
25
41
School # 3
5
8
13
School #4
4
6
10
School # 5
12
7
19
School # 6
3
8
11
School #7
16
10
26
School # 8
4
1
5
Total
66
69
135
4-67
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
4.3.2.2 Preliminary Analyses
Two types of preliminary analyses were performed to begin understanding potential
relationships: bivariate analyses to identify simple indicators of exposure level and principal
component analysis to understand the underlying dimensions or structure in the data. These
analyses are described in Appendix G, sections G.2.2.1 and G.2.2.2. The bivariate analyses
included the principal participants with usable urinary metabolite measurements
(approximately 148 children), and were performed before any recoding of questionnaire
responses for conditional questions and non-responses (Appendix G, sections G.2.1.3 and
G.2.1.4). Because the questions were used as categorical grouping variables, the lack of
recoding did not affect the evaluation of the relationships. Questionnaire variables that
indicate some differences in levels for at least three of the six DAP metabolites (DEP, DETP,
DEDTP, DMP, DMTP, DMDTP) are:
Since house dust measurements are potential indicators of exposure, non-parametric
correlations between the dust and urine measurements were also performed. Fourteen of the
forty-three dust chemicals from Table 4.3.3 show some correlation with the urine
measurements. Chlorpyrifos, diazinon, endosulfan I, endosulfan II, pendimethrin, trifuralin,
and terbufos show a correlation with more than one of the DAP metabolites. The non-OPs in
this list may be indicative of heavy pesticide use, although they do not correspond to the
DAP metabolites.
4.3.2.3 Analysis for Underlying Structure
A principal component analysis (PCA) was performed to identify the dimensions (groups of
questions) explaining the most variability among the potential predictors. When considering
relationships of questions with urine measurements, questions in the same dimension can be
considered like surrogate questions, although each question in a dimension is not a
replacement for the information contained in the group of questions forming the dimension.
Questionnaire responses were recoded to ensure that responses affected by a conditioning
question would be analyzed appropriately (G.2.1.3). This type of recoding affects the
questions included in a PCA dimension. For example, the question cheminhs, about pesticide
treatment inside the house, is the condition question for all questions regarding specific
rooms that were treated. This conditioning of the code values for the room-treated questions
was a likely influence on all of the room questions being grouped together in one dimension.
Variable Name
Variable Description
Pesticides used inside home last month?
Pesticides used outside home last month?
Distance between home and nearest application of pesticides
How often wash local fruit/veg before eating?
Mother now employed (not as housewife)?
Is child covered by medical insurance?
cheminhs
chemouth
closeapp
washvegi
momwork
insured
4-68
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Two scenarios were run in the PCA. One scenario was based on 67 of the questions in Table
G.2.1; the other scenario was based on the same 67 questions and the 22 house and school
dust measurement sums in Table G.2.4. Note that the number of cases used in the two PCA
scenarios differ because school dust measurements were not available for all principal
participants. The complete listing of the principal components (PCs), or dimensions, can be
found in Table G.3.1 in Appendix G. The ten PCs explaining the most variability in the data
for each scenario are listed in Table 4.3.13.
The dimensions explaining the most variability across the two scenarios were:
• Pesticide sprayed inside house
• School and school dust measurements
• Child working in agricultural field
• Relationship of home to agricultural fields
• House dust measurements—OPs
• Adults in household working with pesticides.
Although these dimensions were not analyzed with respect to the urine measurements, they
are consistent with the findings in Stage 3 which were so analyzed. School dust
measurements took prominence in the dimensions extracted when they were included in the
second scenario. These dimensions are useful in understanding the relationships between
questions or questions and dust measurements; however, they do not directly represent
questions having statistically significant relationships with the urine measurement
concentrations. They do represent sets of questions that have more variability, which may
help differentiate pesticide metabolite levels.
4-69
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.3.13 First Ten Principal Components" from Two Scenarios Using Yuma Study Data
Questions-only Scenario (N=130)
Questions and House/School Dust Scenario (N=107)
PC#b
Dimension Description
% variability
explained
Dimension Description
% variability
explained
1
Pesticide sprayed in house and rooms sprayed
17.7
Pesticide sprayed in house and rooms sprayed
13.7
2
Child working in field
7.4
School and school dust measurements
8.8
3
Distance from home to agricultural field
4.9
Child working in field
5.8
4
Relationship to fields where pesticides sprayed
4.2
House dust measurements - OPs
4.3
5
Additional adults at home and working with pesticides
3.6
Grade, age, school dust sum and school dust
permethrins
3.9
6
Sources of drinking water
3.6
Sources of drinking water
3.6
7
Number and age of people in household
3.1
Additional adults at home and working with pesticides
3.1
8
Pesticide sprayed outside home
3.0
Relationship to fields where pesticides sprayed
3.0
9
Mother's occupation
2.9
Distance from home to agricultural field
2.9
10
Height and weight of principal participant
2.5
Household dust sum and household dust permethrins
2.6
Total for top/first 10 PCs
53
Total for top/first 10 PCs
52
Total for all 29 PCsc
86
Total for all 35 PCsc
89
Number of variables included in PCA
67
Number of variables included in PCA
89
a Based on Varimax-rotated component matrix and absolute loadings values greater than or equal to 0.6.
b PC# = principal component number.
c Based on PCs with eigenvalues >0.7 (Jolliffe 1986).
4-70
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
4.3.2.4 Classification Analyses
The technique Classification and Regression Trees (CART) (Breiman 1984) was selected as
the primary type of data mining analysis. Details of the technique can be found in Appendix
G, section G.2.4.1. The principal participants included in these analyses was limited to 130
children, those in kindergarten or first grade, from the initial eight study schools, and with
available and non-suspect urine measurement data. Twelve CART analyses were performed
(Table 4.3.14). Six of the analyses were performed with the log of the molar-weighted sum
of ethylated DAPs (LWETHSUM), and six were performed with the log of the molar-
weighted sum of the methylated DAPs (LWMETHSM).
The six CART analyses performed for each DAP sum evaluated the predictors selected
under, or relative effectiveness of, increasing levels of information and measurement
collection. Three levels of information were analyzed: questions, questions and household
dust measurements, and questions, household dust and school dust measurements. For each
of the three levels, two sets of questions were analyzed to compare the question predictors
selected or relative effectiveness of the questions. The smaller set (LTD) included questions
considered the more likely predictors of exposure levels; the larger set included all of the
questions from the study. The analysis results can be used to compare the effectiveness of
the information levels as screening tools to help identify participants with higher exposure
levels, based on whether predictors from the dust measurements are selected when questions
are available. CART analyses can handle independent variables with missing values; thus,
scenarios including school dust measurements did not have to be analyzed with a smaller
number of cases as for the PCA (section 4.3.2.3).
Table 4.3.14 Cross-Reference for CART Analyses Performed on Yuma Study Data
Predictors Included
Dependent
Variable3
Question
Groupb
House
Dust
School
Dust
Summary
Table
CART Details-
Figures in Appendix G
LWETHSUM
ALL
No
No
G.3.4 c
G.2.1.a, G.2.1 .b
LWETHSUM
LTD
No
No
G.3.4 c
G.2.2.a, G.2.2.b
LWETHSUM
ALL
Yes
No
G.3.4 c
G.2.3.a, G.2.3.b
LWETHSUM
LTD
Yes
No
G.3.4 c
G.2.4.a, G.2.4.b
LWETHSUM
ALL
Yes
Yes
G.3.4 c
G.2.5.a, G.2.5.b
LWETHSUM
LTD
Yes
Yes
G.3.4 c
G.2.6.a, G.2.6.b
LWMETHSM
ALL
No
No
G.3.5
G.2.7.a, G.2.7.b
LWMETHSM
LTD
No
No
G.3.5
G.2.8.a, G.2.8.b
LWMETHSM
ALL
Yes
No
G.3.5
G.2.9.a, G.2.9.b
LWMETHSM
LTD
Yes
No
G.3.5
G.2.10.a, G.2.10.b
LWMETHSM
ALL
Yes
Yes
G.3.5
G.2.11.a, G.2.11 .b
LWMETHSM
LTD
Yes
Yes
G.3.5
G.2.12.a, G.2.12.b
4-71
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
a LWETHSUM is log (molar-weighted sum of ethylated DAPs adjusted for creatinine); LWMETHSM is log
(molar-weighted sum of methylated DAPs adjusted for creatinine). See Appendix F for more details.
b ALL represents analyses with all 67 questions used (Table G.2.1). LTD represents analyses with 29 of the 67
questions considered to be more likely predictors.
c See also Tables G.3.4 and G.3.7 for comparisons of CART analyses with and without CHLDTM3.
For ease of presentation, these classifiers will be termed predictors, although these analyses
are not performed with the intent of offering traditional predictive tools as in regression
analysis. Instead CART is used as a tool to understand the factors and the interactions of the
factors that may affect the exposure measurement levels found in the Yuma Study
participants. A summary of the predictors selected by the CART analyses gives an overview
of the questions or measurements that were found useful in differentiating the levels of
pesticide exposure for children in the Yuma Study (Table 4.3.15).
Table 4.3.15 Categories of Selected Predictors" from CART Analyses of DAP Sums for Yuma Study
Participant Children
LWETHSUMbc
LWMETHSMd
Child's characteristics (weight, ethnicity)
Child's characteristics (height, weight)
Proximity to agricultural fields, spraying conditions
Proximity to agricultural fields, spraying conditions,
child outside when fields sprayed
Child's time spent away from home
Diet - local fruits/vegetables
Pesticide use inside home
Pesticide use inside home, where in house child
spends time
Father's occupation
Child's school
Household dust: OPs, permethrins, non-OPs
Household dust: non-OPs, permethrins
School dust: OPs
School dust: none
a Predictors selected for CART analyses across more than 50% of the scenarios.
b Log (molar-weighted sum of ethylated DAPs) - section F.3.2.
c Predictors based on CART analyses without CHLDTM3 (Table G.3.4)
d Log (molar-weighted sum of methylated DAPs) - section F.3.2
Several predictors are similar across the two DAP sums:
• child's characteristics
• proximity of home to agricultural fields
• pesticide use in the home
• permethrins (in house dust).
The ethylated sum levels consider the time spent at home, locally-grown fruits/vegetables in
the diet, and OPs in house and school dust. The methylated sum levels also consider the
father's occupation and time spent outside when the agricultural fields are sprayed. These
4-72
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
results seem plausible given the differences in the DAP metabolites expected from pesticide
use scenarios in residences and in agriculture.
Note that although the details of the CART analyses are presented in Appendix G, it is
important to recognize that the CART analyses were performed on a maximum of 130 cases.
This level of N, and the range of the DAP sum measurements used as the dependent or target
variables, may make the subpopulations identified in the CART analyses less precise than
needed for prediction. The best use of the CART results is as "indicators" of predictors that
are more useful in differentiating the exposure levels. The CART tree allows the user to note
the localized interactions (at each node's split) between predictors making up the higher or
lower exposure level subpopulations, especially for the first few levels of each tree. When
trying to sort through a large number of predictors, the ability to identify localized rather than
global interactions in a data set is one advantage CART analysis provides over traditional
regression analyses.
4.3.2.5 Comparison of Questionnaire Responses for High and Low Ends of
Measurements
A non-statistical approach was implemented to identify any predictors that could differentiate
between the high and low exposure levels based on the DAP urinary metabolites. In the
previous analyses, CART and CDC (2002), the questionnaire responses, dust measurements,
and urine measurements for all of the participant children were considered. Because the
range of the distribution of the urine and dust measurement values is limited, it seemed
reasonable to compare the information of participants from the extremes of the available
distribution. Thus, approximately 10% of the respondents from the low end of a specific
distribution and approximately 10% of the respondents from the high end of the distribution
were selected.
Twenty-one questions considered more likely to be predictors of a child's pesticide exposure
level were identified. The weighted sum of the responses for each participant was created
from 18 of the questions where the weight was added to the sum if the response indicated a
potential exposure to pesticides. Table G.3.5 in Appendix G shows the questions used in the
exposure weighted sum, and the amounts added to the sum based on the responses. The
values of this weighted sum and the responses to the 18 individual questions (and to school,
grade, and number of rooms treated) were compared between the high- and low-end values
of each measurement sum to determine if any patterns in the responses were evident.
4-73
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.3.16 Results from Non-statistical Comparison of Questionnaire Responses Between High and
Low Ends of Measurement Sum Distributions
Measurement Sum3
Questions'3 Differentiating Between the High and Low Measurement
Groups
WETHSUM +
WMETHSUM
EXPOSURE SUM d, FARFIELDc, WHNCHMOd'e, WHEELd, DADCON2d,
MOMCON2c
WOPSUM
SCHOOL, HOWCHEMOde, FARFIELDd, CLOSEAPPd e, WHEELd,
CHLDTM7d, WHENFILDd, CHLDFLDd
WDUSTSUM
SCHOOL, NRMSRYD0, HOWCHEMOde, OFTCHEMIc, FARFIELDd,
WHNCHMOde, WHEELd, SPRAYFLDd, DADCON2c, MOMCON2d
a See Appendix F for description of sums. Exposure sum is created using weighting scheme in Table G.3.5.
b See Table 3.5 for abbreviated description of question variables.
c Some difference (> 15%) in responses between participants at both ends of measurement distribution was
evident. Difference was in direction expected, that is, exposure to factor is associated with high-end
measurement value.
d Some difference (> 15%) in responses between participants at both ends of measurement distribution was
evident. Difference was not in direction expected based on current knowledge; that is, t exposure to factor is
associated with low-end measurement values.
e Some difference (> 15%) in responses between participants at both ends of measurement distribution was
evident. Difference is based on response (some exposure to factor) compared to non-response (Don't know,
No response).
The questions that point to some differentiation of the exposure levels are reasonable;
however, most of them show the difference to be in the direction opposite of what is
expected based on current knowledge (Table 4.3.16). As in the results of CDC (2002),
relationships with the responses are considered one question at a time. This view may hide
interactions with other risk factors or it may point to other factors that have a related effect.
4.3.2.6 Summary of Results
The Yuma Study report (CDC 2002) looked at each question or dust measurement
individually and included siblings as well as principal participants from 152 households
using a general linear estimating model with repeated measures. The data mining approach
used the questions and dust measurements simultaneously in CART analyses for only 130
principal participants in kindergarten and first grade. Given these and other differences, it
may be useful to look with caution at a summary of the predictors selected under both
approaches to evaluate the universal strength of the predictors for the ethylated DAP sum
(Table 4.3.17) and the methylated DAP sum (Table 4.3.18). Only questions that could be
applied, or were available for siblings as well as principal participants, were analyzed for the
Yuma Study report (CDC 2002).
4-74
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table 4.3.17 Comparison of Selected Predictors from Yuma Study Report" and Data Mining
Approachbfor Sum of Ethylated DAPsc
Yuma Study Report
Data Mining Approach
Recent use of pesticides inside home
Recent use of pesticides inside home
Other adult in household working in agriculture
Child's characteristics (weight, ethnicity)
Proximity to agricultural fields, spraying conditions
Child's time spent away from homed
Diet - local fruits/vegetablesd
Household dust: OPs, permethrins, non-OPs
Household dust: OPs, permethrins, non-OPs
School dust: permethrins
School dust: OPs
a Based on Tables 4.3.4, 4.3.7, and 4.3.9 and molar-weighted sum of ethylated DAPs (adjusted for creatinine).
b Based on Table G.3.7 without CHLDTM3 and log (molar-weighted sum of ethylated DAPs-adjusted for
creatinine).
c See definition in Appendix F.
d Question was not analyzed in CDC (2002).
Table 4.3.18 Comparison of Selected Predictors from Yuma Study Report" and Data Mining
Approachb for Sum of Methylated DAPsc
Yuma Study Report
Data Mining Approach
No Questions
Child's characteristics (height, weight)
Proximity to agricultural fields, spraying conditions
Father's occupation
Where in house child spends timed
Child's schoold
Household dust: diazinon, chlorpyrifos, permethrins,
carbaryl
Household dust: diazinon, chlorpyrifos, permethrins
School dust: diazinon, permethrins,
School dust: none
a Based on Tables 4.3.4, 4.3.7, and 4.3.9 and molar-weighted sum of methylated DAPs (adjusted for creatinine).
b Based on Table G.3.5 and log (molar-weighted sum of methylated DAPs-adjusted for creatinine).
c See definition in Appendix F.
d Question was not analyzed in CDC (2002).
The analyses in the Yuma Study report (CDC 2002) consider questions and measurements
that would apply as risk or exposure factors to the principal participants and the siblings.
These factors may affect explanations of the variability of the pesticide metabolite levels
across siblings within a household. The data mining approach focuses the analyses on a
potentially less variable group of children. For the ethylated sum of DAPs, recent use of
pesticides inside the home, and OPs, non-OPs, and permethrins in the household dust stand
out as differentiators of children's exposure level across both approaches. For the methylated
sum of DAPs, only permethrins and OPs in household dust were similar across both
approaches, since no questions were found significant in the Yuma Study report.
4-75
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
4.4 Effective Predictors of Pesticide Exposure Levels
Two approaches were considered in this project: a literature review across multiple exposure
studies and multiple metabolites, and a more in-depth review of one exposure study in Yuma
County, Arizona (Yuma Study). Although the literature review covers many different
studies, results about significant relationships may be more limited because of the focus of
each publication reviewed. For the Yuma Study, all questions asked were available for
analysis. Access to this level of detail for the studies in the literature review was not
available, although it is likely that the statistically significant relationships are noted in the
publications for questions asked in the study. Taking these differences into consideration, a
summary of the broad categories of predictors selected as differentiators of children's
pesticide exposure (based on urinary metabolites or environmental measurements) can be
enumerated as in Table 4.4.1.
Table 4.4.1 Summary of Predictor Categories Selected as Useful in Differentiating Children's Pesticide
Exposure Levels Across Two Approaches
Literature Review3
Yuma Studyb
Residential pesticide use (inside and
outside)
Residential pesticide use (inside)
Petsc
Occupation of household members
Occupation of household members
Household location: urban vs non-urbanc
Child's characteristics (age, ethnicity, family
income)
Child's characteristics (age, ethnicity, height, weight)
Child's behaviors (loading from hand wipe)0
Dietary behaviors (organic food)0
Dietary behaviors (local fruits/vegetables)
Family hygiene practices
Exposure levels of household members0
Health status (diseases)0
Smoking behavior
(Proximity to agricultural fields)d
Proximity to agricultural fields, spraying conditions
Where child spent time at home/not, or within home
Household dust
Household and school dust: permethrins, OPs and non-OPs
a Based on the "c" Tables 4.2.10.c - 4.2.20.C.
b Based on Tables 4.3.17 and 4.3.18.
c Small number of relationships using these questions categories.
d Proximity to agricultural fields for the literature review was related to dust measurements only.
4-76
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
It would therefore appear that residential use of pesticides, the occupation of household
members, certain demographic characteristics of the children, dietary behaviors, and
proximity to agricultural spraying are the strongest predictors of exposure. Household dust
levels are also predictive of exposures in some cases. Future studies should focus on more
accurate questionnaire information, and more complete urine sample collection to improve
the likelihood of identifying key risk or exposure factors for children's pesticide exposure.
4-77
August 2005
-------
-------
Relationships Between Questionnaire Responses and Children's Pesticide Measurements
5.0 REFERENCES
Adgate, J.L., Barr, D.B., Clayton, C.A., Eberly, L.E., Freeman, N.C.G., Lioy, P.J., Needham,
L.L., Pellizzari, E.D., Quackenboss, J.J., Roy, A., and Sexton, K. 2001. Measurement
of children's exposure to pesticides: analysis of urinary metabolite levels in a
probability-based sample. Environ Health Perspect 109(6): 583-590.
Agricultural Health Study. < http:// www.aghealth.org>
Alavanja, M.C., Sandler, D.P., McMaster, S.B., Zahm, S.H., McDonnell, C.J., Lynch, C.F.,
Pennybacker, M., Rothman, N., Dosemeci, M., Bond, A.E., and Blair, A. 1994.The
Agricultural Health Study. Environ Health Perspect 104: 362-9.
Aprea, C., Strambi, M., Novelli, M.T., Lunghini, L., and Bozzi, N. 2000. Biologic
monitoring of exposure to organophosphorus pesticides in 195 Italian children.
Environ Health Perspect 108: 521-5.
Arbuckle, T.E., Burnett, R., Cole, D., Teschke, K., Dosemeci, M., Bancej, C.N., and Zhang,
J. 2002. Predictors of herbicide exposure in farm applicators. IntArchOccup
Environ Health 75(6): 406-414.
Azaroff, L.S. 1999. Biomarkers of exposure to organophosphorous insecticides among
farmers' families in rural El Salvador: factors associated with exposure. Environ Res
80(2 Pt 1): 138-147.
Barr, D.B., Barr, J.R., Driskell, W.J., Hill, R.H., Jr., Ashley, D.L, Needham L.L., et al. 1999.
Strategies for biological monitoring of exposure for contemporary-use pesticides.
Toxicol Ind Health 15: 168-179.
Barr, D.B. and Needham, L.L. 2002. Analytical methods for biological monitoring of
exposure to pesticides: a review. J Chromatogr B Analyt TechnolBiomedLife Sci
778: 5-29.
Barr, D.B., and Garry, V.F. 2004. Pesticides and Children. Toxicol ApplPharmacol 198(2):
152-163.
Black, K., Shalat, S.L., Freeman, N.C., Jimenez, M., Donnelly, K.C., and Calvin, J.A. 2005.
Children's mouthing and food-handling behavior in an agricultural community on the
US/Mexico border. J Expo Anal Environ Epidemiol 15(3): 244-51.
Bradman, M.A., Harnly, M.E., Draper, W., Seidel, S., Teran, S., Wakeham, D., andNeutra,
R. 1997. Pesticide exposures to children from California's Central Valley: results of a
pilot study. J Expo Anal Environ Epidemiol 7(2): 217-234.
Breiman, Leo, Jerome H. Friedman, Richard A. Olshen, Charles J. Stone. 1984.
Classification and Regression Trees. Boca Raton, Florida: Chapman & Hall/CRC.
5-1
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Measurements
Carrel, C.L. 1996. Urinary Metabolite Monitoring in Children With Paraoccupational
Exposure to Dimethyl Organophosphorous Pesticides [thesis], Seattle (WA):
University of Washington. 141 pp. University of Washington Health Sciences
Library, call number WA 7 Th44574.
Centers for Disease Control and Prevention. 2002. Pesticide Exposure in Children Living in
Agricultural Areas along the United States-Mexico Border, Yuma County, Arizona
(Final Report). Available from rsabogal@cdc.gov.
Clayton, C. A., E. D. Pellizzari, R. W. Whitmore, R. L. Perritt, and J. J. Quackenboss. 1999.
National Human Exposure Assessment Survey (NHEXAS): Distributions and
Associations of Lead, Arsenic and Volatile Organic Compounds in EPA Region 5. J
Expo Anal Environ Epidemiol 9(5): 381-392.
Cohen Hubal, E.A., L.S. Sheldon, J.M. Burke, T.R. McCurdy, M.R. Berry, M.I. Rigas, V.G.
Zartarian, andN.C. Freeman. 2000. Children's exposure assessment: a review of
factors influencing children's exposure, and the data available to characterize and
assess that exposure. Environ Health Perspect 108(6): 475-86.
Cohen Hubal, Elaine A., Sheldon, Linds S., Aufall, Maria J., Burke, Janet M., and Thomas,
Kent W. 2000b. The challenge of assessing children's residential exposure to
pesticides. J Exp Anal Environ Epidemiol 10(6, Part 2): 638-549.
Curl, C.L., Fenske, R.A., Kissel, J.C„ Shirai, J.H., Moate, T.F., Griffith, W., et al. 2002.
Evaluation of take-home organophosphorus pesticide exposure among agricultural
workers and their children. Environ Health Perspect 110: 787-792.
Curl C.L., Fenske, R.A., Elgethun, K. 2003. Organophosphorus pesticide exposure of urban
and suburban pre-school children with organic and conventional diets. Environ
Health Perspect 111(3): 377-382.
Delphi face to face. 2005.
(accessed 2 August 2005).
Dosemici, M., M.C. Alvanja, A.S. Rowland, D. Mage, S.H. Zahm, N. Rothman, J.H. Lubin,
J. A. Hoppin, D.P. Sandler, and A. Blair. 2002. A quantitative approach for estimating
exposure to pesticides in the Agricultural Health Study. Ann Occup Hyg 46(2): 245-
260.
Elgethun, K., Fenske, R.A., Yost, M.G., and Palcisko, G.J. 2003. Time-location analysis for
exposure assessment studies of children using a novel global positioning system
instrument. Environ Health Perspect 111:115-22.
Elgethun, K. 2004. Global positioning system (GPS) tracking to characterize children's
exposure to pesticides [thesis], Seattle (WA): University of Washington, 299 pp.
5-2
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Measurements
Available from: University of Washington Library, WA 7 Th53624. Doctoral
Dissertation, University of Washington.
Elgethun, K., Yost, M.G., Fitzpatrick, C.T.E., Nyerges, T.L., and Fenske, R.A. 2005.
Comparison of GPS tracking to the NHEXAS diary timeline for time-location
measurement of children. J Exp Anal Environ Epidemiol (submitted).
Fenske, Richard. 1998. Status Report on Biological Monitoring Research Relevant to
Aggregate Exposure Assessment under the Food Quality Protection Act. Report to
Subcommittee on Aggregate Exposure Assessment, Health and Environmental
Sciences Institute, International Life Sciences Institute. October 12, 1998. Not
published.
Fenske, R.A., C. Lu, D. Barr, and L. Needham. 2002. Children's exposure to chlorpyrifos and
parathion in an agricultural community in central Washington State. Environ Health
Perspect 110(5): 549-553.
Food Quality Protection Act of 1996, Public Law 104-170, 1996. Available at
Freeman, N.C., Hore, P., Black, K., Jimenez, M., Sheldon, L., Tulve, N., Lioy, P.J. 2005.
Contributions of children's activities to pesticide hand loadings following residential
pesticide application. J Expo Anal Environ Epidemiol 15(1): 81-88.
Grossman, J.E. 2001. The take-home pathway for agricultural pesticides: contribution of
occupational factors to home contamination [thesis], Seattle (WA): University of
Washington, 168 pp. Available from: University of Washington Library, WA 7
Th50727.
Hand, David J. 1999. Statistics and Data Mining: Intersecting Disciplines. SIGKDD
Explorations 1(1): 16-9.
Harris, S.A., Sass-Kortsak, A.M., Corey, P.N., and Purdham, J.T. 2002. Development of
models to predict dose of pesticides in professional turf applicators. J Expo Anal
Environ Epidemiol 12: 130-144.
Heudorf, U., Angerer, J., and Drexler, H. 2004. Current internal exposure to pesticides in
children and adolescents in Germany: urinary levels of metabolites of pyrethroid and
organophosphorus insecticides. IntArch Occup Environ Health 77: 67-72. Epub 2003
Oct 9.
Hotelling, H. 1933. Analysis of a Complex of Statistical Variables into Principal
Components. Journal of Educational Psychology 24: 417-41,498-520.
Jackson, J. Edward. 1991. A User's Guide To Principal Components. New York: John
Wiley & Sons, Inc.
5-3
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Measurements
Jolliffe, I.T. 1986. Principal Component Analysis. New York, NY: Springer-Verlag.
Jolliffe, I.T. 2002. Principal Component Analysis. New York, NY: Springer-Verlag.
Koch, D. 1999. Longitudinal Biological Monitoring Study of Organophosphate Pesticide
Exposure among Children Living in an Agricultural Community [thesis], Seattle
(WA): University of Washington. 137pp. University of Washington Health Sciences
Library, call number WA 7 Th48906.
Koch, D., Lu, C., Fisker-Andersen, J., Jolley, L., and Fenske, R.A. 2002. Temporal
association of children's pesticide exposure and agricultural spraying: report of a
longitudinal biological monitoring study. Environ Health Perspect 110(8): 829-833.
Koch, D., Lu, C., Jolley, L., Fisker-Andersen, J.A., and Fenske, R.A. 2003. Temporal
association of children's pesticide exposure and agricultural spraying: report of a
longitudinal biological monitoring study. Environ Health Perspect 110: 829-833.
Krieger, R.I., Bernard, C.E., Dinoff, T.M., Ross, J.H., and Williams, R.L. 2001.
Biomonitoring of persons exposed to insecticides used in residences. Ann Occup Hyg
45(Suppl 1): S143-153.
Krinsley, Jeanne S. 1998. Exposure to chlorpyrifos and use of pesticides in Arizona [thesis],
Tucson (AZ): The University of Arizona, 204 pp. Available from: Arizona Health
Sciences Library, call number W4A 1998 K92E.
Loewenherz, C., Fenske, R.A., Simcox, N.J., Bellamy, G., and Kalman, D. 1997. Biological
monitoring of organophosphorus pesticide exposure among children of agricultural
workers in Central Washington State. Environ Health Perspect 105(12): 1344-1353.
Lu, C., Fenske, R.A., Simcox, N.J., and Kalman, D. 2000. Pesticide exposure of children in
an agricultural community: evidence of household proximity to farmland and take
home exposure pathways. Environ Res 84:290-302.
Lu, C., Knutson, D.E., Fisker-Anderson, J., and Fenske, R.A. 2001. Biological monitoring
survey of organophosphorus pesticide exposure among pre-school children in the
Seattle metropolitan area. Environ Health Perspect 109:299-303.
Lu, C., Showlund-Irish, R.M., and Fenske, R.A.. 2003. Biological monitoring of diazinon
exposure using saliva in an animal model. J Toxicol Environ Health A. 66(24):
2315-25.
Macintosh, D.L., Needham, L.L., Hammerstrom, K.A., and Ryan, P.B. 1999. A longitudinal
investigation of selected pesticide metabolites in urine. J Expo Anal Environ
Epidemiol. 9(5): 494-501.
5-4
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Measurements
McCauley, L.A., M.R. Lasarev, G. Higgins, J. Rothlein, J. Muniz, C. Ebbert, and J. Phillips.
2001. Work characteristics and pesticide exposures among migrant agricultural
families: a community-based research approach. Environ Health Perspect 109(5):
533-538
Magidson, J. 1993. The CHAID approach to segmentation modeling. In R. Bagozzi (Ed.),
Handbook of Marketing Research. Cambridge, Massachusetts: Blackwell.
Measurement Group Website, .
Mills, P.K., Zahm, S.H. 2001. Organophosphate pesticide residues in urine of farmworkers
and their children in Fresno County, California. Am JIndMed 40: 571-7.
Morgan, M.K., Sheldon, L.S., Croghan, C.W., Jones, P.A., Robertson, G.L., Chuang, J.C.,
Wilson, N.K., and Lyu, C.W. Exposures of preschool children to chlorpyrifos and its
degradation product 3,5,6-trichloro-2-pyridinol in their everyday environments. J
Expo Anal Environ Epidemiol. 2004 Sep 15; [Epub ahead of print]
National Research Council Committee on Pesticides in the Diets of Infants and Children.
1993. Pesticides in the Diets of Infants and Children. Washington, DC: National
Academy of Sciences.
Needham, L.L., Sexton, K. 2000. Assessing children's exposure to hazardous environmental
chemicals: an overview of selected research challenges and complexities. J Expo Anal
Environ Epidemiol 10: 611-629.
O'Rourke, M.K., P.S. Lizardi, S.P. Rogan, N.C. Freeman, A. Aguirre, C.G. Saint. 2000.
Pesticide exposure and creatinine variation among young children. J Expo Anal
Environ Epidemiol 10(6 Pt 2): 672-81.
Quandt, S.A., Arcury, T.A., Rao, P., Snively, B.M., Camann, D.E., Doran, A.M., Yau A.Y.,
Hoppin, J. A., and Jackson, D.S. 2004. Agricultural and residential pesticides in wipe
samples from farmworker family residences in North Carolina and Virginia. Environ
Health Perspect 112: 382-7
Reed, K.J., Jimenez, M., Freeman, N.C., and Lioy, P.J. 1999. Quantification of children's
hand and mouthing activities through a videotaping methodology. J Expo Anal
Environ Epidemiol 9(5): 513-20.
Roy, A., P. G. Georgopoulos, M. Ouyang, N. Freeman, and P.J. Lioy. 2003. Environmental,
dietary, demographic, and activity variables associated with biomarkers of exposure
for benzene and lead. J Expo Anal Environ Epidemiol 13: 417-426.
Royster, M.O., Hilborn, E.D., Barr, D., Carty, C.L., Rhoney, S., and Walsh. D. 2002. A pilot
study of global positioning system/geographical information system measurement of
5-5
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Measurements
residential proximity to agricultural fields and urinary organophosphate metabolite
concentrations in toddlers. J Expo Anal Environ Epidemiol 12(6): 433-440.
Sexton, K., Callahan, M.A., and Bryan, E.F. 1995a. Estimating Exposure and Dose to
Characterize Health Risks: The Role of Human Tissue Monitoring in Exposure
Assessment. Environmental Health Perspectives Suppl 103(3): 13-30.
Sexton, K., Callahan, M.A., and Bryan, E.F. 1995b. An Introduction to the National Human
Exposure Assessment Survey (NHEXAS) and Related Phase I Field Studies. J Expo
Anal Environ Epidemiol 5(3): 233-56.
Sexton, K., Adgate, J.L., Eberly, L.E., Clayton, C.A., Whitmore, R.W., and Pellizzari, E.D.
2003. Predicting children's short-term exposure to pesticides: results of a
questionnaire screening approach. Environ Health Perspect 110:123-128.
Shalat, S.L., Donnelly, K.C., Freeman, N.C.G., Calvin, J.A., Ramesh, J., Jimenez, M., Black,
K., Coutinho, C., Needham, L.L., Barr, D.B., and Ramirez, J. 2003. Nondietary
ingestion of pesticides by children in an agricultural community on the US/Mexico
border: preliminary results. J Expo Anal Environ Epidemiol 13: 42-50.
Simcox, N.J., Fenske, R.A., Wolz, S.A., Lee, I-C., Kalman, D.A. 1995. Pesticides in
household dust and soil: exposure pathways for children of agricultural families.
Environ Health Perspect 103: 1126-1134.
Teitelbaum, S.L. 2002. Questionnaire assessment of nonoccupational pesticide exposure in
epidemiologic studies of cancer. J Expo Anal Environ Epidemiol (12): 373-380.
Thomas, K. 2004. Assessing Exposure Classification in the Agricultural Health Study. 14th
Annual Conference of the International Society of Exposure Analysis, Philadelphia,
PA, October 17-21.
Thompson, B., Coronado, G.D., Grossman, J.E., Puschel, K., Solomon, C.C., Islas, I., Curl,
C.L., Shirai, J.H., Kissel, J.C., and Fenske, R.A. 2003. Pesticide take-home pathway
among children of agricultural workers: study design, methods, and baseline findings.
J Occup Environ Med 45: 42-53.
Two Crows Corporation. 1999. Introduction to Data Mining and Knowledge Discovery,
Third Ed. Potomac, Maryland.
U.S. Environmental Protection Agency (EPA). 1992. Guidelines for Exposure Assessment.
Washington D.C.: EPA/600Z-92/011. FR 57: 22888-22938.
U.S. Environmental Protection Agency (EPA). 1997. Children's exposure to pesticides.
STAR Report 1(1): 1-6. EPA/600/F-97/020.
5-6
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Measurements
U.S. Environmental Protection Agency (EPA). 2000a. Strategic Plan for the Analysis of the
National Human Exposure Assessment Survey (NHEXAS) Pilot Study Data.
U.S. Environmental Protection Agency (EPA), Office of Research and Development (ORD).
2000b. Strategy for Research on Environmental Risks to Children. EPA/600/R-
00/068.
U.S. Environmental Protection Agency (EPA). 2002. Pesticide Exposure and Potential
Health Effects in Young Children Along the U.S.-Mexico Border. Washington D.C.:
EPA/600/R-02/085. To be posted September 2005 at
.
U.S. Environmental Protection Agency (EPA). 2003. The Children's Exposure to Pesticides
and Other Persistent Organic Pollutants (CTEPP) Study.
U.S. Environmental Protection Agency (EPA). 2004a. Identifying Predictors of Exposure—
Analyses of the National Human Exposure Assessment Survey (NHEXAS)
Questionnaire and Measurement Data. Washington D.C.: EPA/600/R-03/061
U.S. Environmental Protection Agency (EPA). 2004b. US-Mexico Border 2012
Environmental Health Workgroup website, (accessed
15 July 2004).
Venable, W.N., and Ripley, B.D. 1994. Modern Applied Statistics with S-Plus. New York:
Springer-Verlag.
Wietlisbach, V. 1999. Statistical Approaches in the Development of Clinical Practice
Guidelines From Expert Panels: The Case of Laminectomy in Sciatica Patients.
Medical Care 37(8): 785-797.
5-7
August 2005
-------
-------
Appendix A
References from the Literature Review
-------
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendix A
References from the Literature Review
The following tables show three sets of citations from the literature review for this report. Table
A. 1 lists the citations that were relevant to the objectives of this report from which relationships
were extracted. Table A.2 lists citations that were reviewed, but which did not meet the
objectives of this report. The literature review for this report was conducted on literature
published through early 2003. Table A-3 includes some references published after that date that
may provide additional information regarding the type of relationships discussed in section 4.2
(Results)
Table A-l. Citations and Citation Abbreviations Referenced in the Relationship Results (Section 4,
Appendix B, and Appendix C)
Citation
Abbreviation
Adgate 2001
Aprea 2000
Azaroff 1999
Carrel 1996
Citation
Adgate JL, Barr DB, Clayton CA, Eberly LE, Freeman NCG, Lioy PJ,
Needham LL, Pellizzari ED, Quackenboss JJ, Roy A, Sexton K. 2001.
Measurement of children's exposure to pesticides: analysis of urinary
metabolite levels in a probability-based sample. Environ Health Perspect
109(6): 583-590.
Aprea C, Strambi M, Novelli MT, Lunghini L, Bozzi N. 2000. Biologic
monitoring of exposure to organophosphorus pesticides in 195 Italian
children. Environ Health Perspect 108(6): 521-525.
Azaroff LS. 1999. Biomarkers of exposure to organophosphorous
insecticides among farmers' families in rural El Salvador: factors associated
with exposure. Environ Res 80(2 Pt 1): 138-147.
Carrel CL. Urinary Metabolite Monitoring in Children With Paraoccupational
Exposure to Dimethyl Organophosphorous Pesticides [thesis], Seattle
(WA): University of Washington. 141 pp. U of WA Health Sciences Library,
call number WA 7 Th44574.
Curl 2003
Curl 2002
Fenske 2002
Grossman 2001
Curl CL, Fenske RA, Elgethun K. 2003. Organophosphorus pesticide
exposure of urban and suburban preschool children with organic and
conventional diets. Environ Health Perspect 111 (3): 377-382.
Curl CL, Fenske RA, Kissel JC, Shirai JH, Moate TF, Griffith W, Coronado
G, Thompson B. 2002. Evaluation of take-home organophosphorus
pesticide exposure among agricultural workers and their children. Environ
Health Perspect 110(12): A787-A792.
Fenske RA, Lu C, Barr D, Needham L. 2002. Children's exposure to
chlorpyrifos and parathion in an agricultural community in central
Washington state. Environ Health Perspect 110(5): 549-553.
Grossman JE. 2001. The take-home pathway for agricultural pesticides:
contribution of occupational factors to home contamination [thesis], Seattle
(WA): University of Washington, 168 p. U of WA Health Sciences Library,
call number W 7 Th50727.
A-l
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Citation
Abbreviation
Koch 1999
Koch 2002
Krinsley 1998
Loewenherz
1997
Lu 2000
Lu 2001
McCauley 2001
McCauley 2003
Royster 2002
Sexton 2003
Shalat 2003
Simcox 1995
Citation
Koch D. 1999. Longitudinal Biological Monitoring Study of
Organophosphate Pesticide Exposure among Children Living in an
Agricultural Community [thesis], Seattle (WA): University of Washington.
137pp. U of WA Health Sciences Library, call number WA 7 Th48906.
Koch D, Lu C, Fisker-Andersen J, Jolley L, Fenske RA. 2002. Temporal
association of children's pesticide exposure and agricultural spraying: report
of a longitudinal biological monitoring study. Environ Health Perspect
110(8): 829-833.
Krinsley, JS. Exposure to chlorpyrifos and use of pesticides in Arizona
[thesis], Tucson (AZ): The University of Arizona. 204 pp. Available from:
Arizona Health Sciences Library, call number W4A 1998 K92E.
Loewenherz C, Fenske RA, Simcox NJ, Bellamy G, Kalman D. 1997.
Biological monitoring of organophosphorus pesticide exposure among
children of agricultural workers in central Washington state. Environ Health
Perspect 105:1344-1353. Corrections and clarifications in Environ Health
Perspect 107(2): A61.
Lu C, Fenske RA, Simcox NJ, Kalman D. 2000. Pesticide exposure of
children in an agricultural community: evidence of household proximity to
farmland and take home exposure pathways. Environ Res 84(3): 290-302.
Lu C, Knutson DE, Fisker-Andersen J, Fenske RA. 2001. Biological
monitoring survey of organophosphorus pesticide exposure among pre-
school children in the Seattle metropolitan area. Environ Health Perspect
109(3): 299-303.
McCauley LA, Lasarev MR, Higgins G, Rothlein J, Muniz J, Ebbert C,
Phillips J. 2001. Work characteristics and pesticide exposures among
migrant agricultural families: a community-based research approach.
Environ Health Perspect 109(5): 533-538.
McCauley LA, Michaels S, Rothlein J, Muniz J, Lasarev M, Ebbert C. 2003.
Pesticide expsosure and self-reported home hygiene: practices in
agricultural families. AAOHN Journal 51(3): 113-119.
Royster MO, Hilborn ED, Barr D, Carty CL, Rhoney S, Walsh D. 2002. A
pilot study of global positioning system/geographical information system
measurement of residential proximity to agricultural fields and urinary
organophosphate metabolite concentrations in toddlers. J Expo Anal
Environ Epidemiol 12(6): 433-440.
Sexton K, Adgate JL, Eberly LE, Clayton CA, Whitmore RW, Pellizzari ED.
2003. Predicting children's short-term exposure to pesticides: results of a
questionnaire screening approach. Environ Health Perspect 110:123-128.
Shalat SL, Donnelly KC, Freeman NCG, Calvin JA, Ramesh S, Jimenez M,
Black K, Coutinho C, Needham LL, Barr DB, Ramirez J. 2003. Nondietary
ingestion of pesticides by children in an agricultural community on the
US/Mexico border: preliminary results. J Expo Anal Environ Epidemiol 13:
42-50.
Simcox NJ, Fenske RA, Wolz SA, Lee l-C, Kalman DA. 1995. Pesticides in
household dust and soil: exposure pathways for children of agricultural
families. Environ Health Perspect 103:1126-1134.
A-2
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table A-2. Citations Reviewed and Considered Not Applicable For This Report
Citation
Abbreviation
Adgate 2000a
Adgate 2000b
Akland 2000
AmNurse 2001
Arbuckle 2002
Balluz 2001
Bernstein 1999
Boone 2001
Bradman 1997
Buckley 2000
CDC 2001
Clayton 2003
Citation
Adgate JL, Kukowski A, Stroebel C, Shubat PJ, Morrell S, Quackenboss JJ,
Whitmore RW, Sexton K. 2000. Pesticide storage and use patterns in Minnesota
households with children. J Expo Anal Environ Epidemiol 10(2): 159-167.
Adgate JL, Clayton CA, Quackenboss JJ, Thomas KW, Whitmore RW, Pellizzari
ED, Lioy PJ, Shubat P, Stroebel C, Freeman NC, Sexton K. 2000. Measurement
of multi-pollutant and multi-pathway exposures in a probability-based sample of
children: practical strategies for effective field studies. J Expo Anal Environ
Epidemiol 10(6 Pt 2): 650-661.
Akland GG, Pellizzari ED, Hu Y, Roberds M, Rohrer CA, Leckie JO, Berry MR.
2000. Factors influencing total dietary exposures of young children. J Expo Anal
Environ Epidemiol 10(6 Pt 2): 710-722.
[Anonymous], Environmentally healthy homes and communities. Children's
special vulnerabilities. AmNurse 33(6): 26-38.
Arbuckle TE, Burnett R, Cole D, Teschke K, Dosemeci M, Bancej CN, Zhang J.
2002. Predictors of herbicide exposure in farm applicators. Int Arch Occup
Environ Health 75(6): 406-414.
Balluz L, Moll D, Diaz Martinez MG, Merida Colindres JE, Malilay J. 2001.
Environmental pesticide exposure in Honduras following hurricane Mitch. Bull
World Health Organ 79(4): 288-295.
Bernstein, IL; Bernstein, JA; Miller, M; Tierzieva, S; Bernstein, Dl; Lummus, Z;
Selgrade, MK; Doerfler, DL; Seligy, VL. 1999. Immune responses in farm
workers after exposure to Bacillus thuringiensis pesticides. Environ Health
Perspect 107(7): 575-582.
Boone JS, Tyler JW, Chambers JE. 2001. Transferable residues from dog fur
and plasma cholinesterase inhibition in dogs treated with a flea control dip
containing chlorpyrifos. Environ Health Perspect 109:1109-1114.
Bradman MA, Harnly ME, Draper W, Seidel S, Teran S, Wakeham D, Neutra R.
1997. Pesticide exposures to children from California's Central Valley: results of a
pilot study. J Expo Anal Environ Epidemiol 7(2):217-234.
Buckley B, Ettinger A, Hore P, Lioy P, Freeman N. 2000. Using observational
information in planning and implementation of field studies with children as
subjects. J Expo Anal Environ Epidemiol 10(6 Pt 2):695-702.
[CDC] Centers for Disease Control and Prevention (US). 2001 Mar. National
Report on Human Exposure to Environmental Chemicals. Report Number
PB2002106473. Atlanta (GA). Available from:
http://www.ntis.gov/search/product.asp?ABBR=PB2002106473&starDB=GRAHIS
T
Clayton CA, Pellizzari ED, Whitmore RW, Quackenboss JJ, Adgate J, Sefton K.
2003. Distributions, associations, and partial aggregate exposure of pesticides
and polynuclear aromatic hydrocarbons in the Minnesota Children's Pesticide
Exposure Study (MNCPES). J Expo Anal Environ Epidemiol 13(2): 100-111.
A-3
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Citation
Abbreviation
Cooper 2001 a
Cooper 2001b
Curwin 2002
Dosemeci 2002
Eskenazi 1999
Esteban 1996
Faustman 2000
Fenske 2000a
Fenske 2000b
Gladen 1998
Harris 2002
Hill 1989
Infante-Rivard
1999
Citation
Cooper SP, Burau K, Sweeney A, Robison T, Smith MA, Symanski E, Colt JS,
Laseter J, Zahm SH. 2001. Prenatal exposure to pesticides: a feasibility study
among migrant and seasonal farmworkers. Am J Ind Med 40(5): 578-585.
Cooper, SP; Darragh, AR; Vernon, SW; Stallones, L; MacNaughton, N; Robison,
T; Hanis, C; Zahm, SH. 2001. Ascertainment of pesticide exposures of migrant
and seasonal farmworker children: findings from focus groups. Am J Ind Med
40(5): 531-537.
Curwin B, Sanderson W, Reynolds S, Hein M, Alavanja M. 2002. Pesticide use
and practices in an Iowa farm family pesticide exposure study. J Agric Saf Health
8(4): 423-433.
Dosemeci M, Alavanja MC, Rowland AS, Mage D, Zahm SH, Rothman N, Lubin
JH, Hoppin JA, Sandler DP, Blair A. 2002. A quantitative approach for estimating
exposure to pesticides in the Agricultural Health Study. The Ann Occup Hyg
46(2): 245-260.
Eskenazi B, Bradman A, Castorina R. 1999. Exposures of children to
organophosphate pesticides and their potential adverse health effects. Environ
Health Perspect 107(Suppl 3): 409-419.
Esteban E, Rubin C, Hill R, Olson D, Pearce K. 1996. Association between
indoor residential contamination with methyl parathion and urinary para-
nitrophenol. J Expo Anal Environ Epidemiol 6(3): 375-387.
Faustman EM, Silbernagel SM, Fenske RA, BurbacherTM, Ponce RA. 2000.
Mechanisms underlying children's susceptibility to environmental toxicants.
Environ Health Perspect 108(Suppl 1): 13-21.
Fenske RA, Kissel JC, Lu C, Kalman DA, Simcox NJ, Allen EH, Keifer MC. 2000.
Biologically based pesticide dose estimates for children in an agricultural
community. Environ Health Perspect 108(6): 515-520.
Fenske RA, Lu C, Simcox NJ, Loewenherz C, Touchstone J, Moate TF, Allen EH,
Kissel JC. 2000. Strategies for assessing children's organophosphorus pesticide
exposures in agricultural communities. J Expo Anal Environ Epidemiol 10(6 Pt 2):
662-671.
Gladen BC, Sandler DP, Zahm SH, Kamel F, Rowland AS, Alavanja MCR. 1998.
Exposure opportunities of families of farmer pesticide applicators. Am J Ind Med
34:581-587.
Harris SA, Sass-Kortsak AM, Corey PN, Purdham JT. 2002. Development of
models to predict dose of pesticides in professional turf applicators. J Expo Anal
Environ Epidemiol 12: 130-144.
Hill RH Jr, To T, Holler JS, Fast DM, Smith SJ, Needham LL, Binder S. 1989.
Residues of chlorinated phenols and phenoxy acid herbicides in the urine of
Arkansas children. Arch Environ Contam Toxicol 18(4): 469-474.
Infante-Rivard C, Krajinovic M, Labuda D, Sinnett D. 1999. Risk of childhood
leukemia associated with exposure to pesticides and with gene polymorphisms.
EpidemiollO (5) :481-487.
A-4
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Citation
Abbreviation
Krieger2001
Lebowitz 1995
Lewis 1994
Lioy 2000
London 1998
Masley 2000
McCauley 2001
Mills 2001
Moschandreas
2001
NASS 2001
O'Rourke 2000
Park 2001
Perera 1999
Citation
Krieger Rl, Bernard CE, DinoffTM, Ross JH, Williams RL. 2001. Biomonitoring of
persons exposed to insecticides used in residences. Ann Occup Hyg 45(Suppl 1):
S143-153.
Lebowitz MD, O'Rourke MK, Gordon S, Moschandreas DJ, Buckley T, Nishioka
M. 1995. Population-based exposure measurements in Arizona: a phase I field
study in support of the National Human Exposure Assessment Survey. J Expo
Anal Environ Epidemiol 5(3): 297-325.
Lewis RG, Fortmann RC, Camann DE. 1994. Evaluation of methods for
monitoring the potential exposure of small children to pesticides in the residential
environment. Arch Environ Contam Toxicol 26(1): 37-46.
Lioy PJ, Edwards RD, Freeman N, Gurunathan S, Pellizzari E, Adgate JL,
Quackenboss J, Sexton K. 2000. House dust levels of selected insecticides and
a herbicide measured by the EL and LWW samplers and comparisons to hand
rinses and urine metabolites. J Expo Anal Environ Epidemiol 10(4): 327-340.
London L, Nell V, Thompson ML, Myers JF. 1998. Health status among farm
workers in the Western Cape-collateral evidence from a study of occupational
hazards. S African Med J. (Suid-Afrikaanse tydskrif geneeskunde) 88(9): 1096-
1101.
Masley, ML; Semchuk, KM; Senthilselvan, A; McDuffie, HH; Hanke, P; Dosman,
JA; Cessna, AJ; Crossley, MFO; Irvine, DG; Rosenberg, AM; Hagel, LM. 2000.
Health and environment of rural families: Results of a community canvass survey
in the Prairie Ecosystem Study (PECOS). J Agric Saf Health 6(2): 103-115.
McCauley LA, Beltran M, Phillips J, Lasarev M, Sticker D. 2001. The Oregon
migrant farmworker community: an evolving model for participatory research.
Environ Health Perspect 109(Suppl 3): 449-455.
Mills PK, Zahm SH. 2001. Organophosphate pesticide residues in urine of
farmworkers and their children in Fresno County, California. Am J Ind Med
40(5): 571-577.
Moschandreas DJ, Karuchit S, Kim Y, O'Rourke MK, Ari H, Lebowitz MD,
Robertson G, Gordon S, Moschandreas DJ. 1994. In-residence, multiple route
exposures to chlorpyrifos and diazinon estimated by indirect method models.
Atmosph Environ 35(12): 2201-2224.
[NASS] National Agricultural Statistics Service (US). 2001 Oct. Agricultural
chemical usage: 2000 restricted use summary. Report Number NASSZUP2001.
Washington, DC: Agricultural Statistics Board. Available from:
http://www.ntis.gov/search/product.asp?ABBR=NASSZUP2001&starDB=GRAHIS
T
O'Rourke MK, Lizardi PS, Rogan SP, Freeman NC, Aguirre A, Saint CG. 2000.
Pesticide exposure and creatinine variation among young children. J Expo Anal
Environ Epidemiol 10(6 Pt 2): 672-681.
Park JH, Spiegelman DL, Gold DR, Burge HA, Milton DK. 2001. Predictors of
airborne endotoxin in the home. Environ Health Perspect 109(8): 859-864.
Perera FP, Jedrychowski W, Rauh V, Whyatt RM. 1999. Molecular epidemiologic
research on the effects of environmental pollutants on the fetus. Environ Health
Perspect 107(Suppl 3): 451-460.
A-5
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Citation
Abbreviation
Quackenboss
2000
Rigas 2001
Seifert 2000
Sexton 2000
Stuetz 2001
Thompson
2003
Vela-Acosta
2002
Ward 2001
Wilson 2003
Zartarian 2000
Citation
Quackenboss JJ, Pellizzari ED, Shubat P, Whitmore RW, Adgate JL, Thomas
KW, Freeman NC, Stroebel C, Lioy PJ, Clayton AC, Sexton K. 2000. Design
strategy for assessing multi-pathway exposure for children: the Minnesota
Children's Pesticide. J Expo Anal Environ Epidemiol 10(2):145-158.
Rigas ML, Okino MS, Quackenboss JJ. 2001. Use of a pharmacokinetic model to
assess chlorpyrifos exposure and dose in children, based on urinary biomarker
measurements. Toxicol Sci 61 (2): 374-81.
Seifert B, Becker K, Hoffmann K, Krause C, Schulz C. 2000. The German
Environmental Survey 1990/1992 (GerES II): a representative population study. J
Expo Anal Environ Epidemiol 10(2): 103-114,
Sexton K, Greaves IA, Church TR, Adgate JL, Ramachandran G, Tweedie RL,
Fredrickson A, Geisser M, Sikorski M, Fischer G, Jones D, Ellringer P. 2000. A
school-based strategy to assess children's environmental exposures and related
health effects in economically disadvantaged urban neighborhoods. J Expo Anal
Environ Epidemiol 10(6 Pt 2): 682-694.
Stuetz W, Prapamontol T, Erhardt JG, Classen HG. Organochlorine pesticide
residues in human milk of a Hmong hill tribe living in Northern Thailand. Sci Total
Environ 273(1-3): 53-60.
Thompson B, Coronado GD, Grossman JE, Puschel K, Solomon CC, Islas I, Curl
CL, Shirai JH, Kissel JC, Fenske RA. 2003. Pesticide take-home pathway among
children of agricultural workers: study design, methods, and baseline findings. J
Occup Environ Med 45: 42-53.
Vela-Acosta MS, Bigelow P, Buchan R. 2002. Assessment of occupational health
and safety risks of farmworkers in Colorado. Am J Ind Medicine 42 (Suppl 2): 19-
27.
Ward MH, Prince JR, Stewart PA; Zahm, SH. 2001. Determining the probability of
pesticide exposures among migrant farmworkers: Results from a feasibility study.
Am J Ind Med 40(5): 538-553.
Wilson NK, Chuang JC, Lyu C, Menton R, Morgan MK. 2003. Aggregate
exposures of nine preschool children to persistent organic pollutants at day care
and at home. J Expo Anal Environ Epidemiol 13(3): 187-202.
Zartarian VG, Ozkaynak H, Burke JM, Zufall MJ, Rigas ML, Furtaw EJ Jr. 2000.
A modeling framework for estimating children's residential exposure and dose to
chlorpyrifos via dermal residue contact and nondietary ingestion. Environ Health
Perspect 108(6): 505-14
A-6
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table A-3. References Published After Early 2003 That May Provide Additional Relationship Information
Citation
Abbreviation
Arbuckle 2004
Barr2004
Coronado 2004
Freeman 2005
Garry 2004
Lu 2004
Citation
Lu C, Kedan G, Fisker-Andersen J, Kissell JC, Fenske RA. 2004.
Multipathway organophosphorus pesticide exposures of preschool children
living in agricultural and nonagricultural communities. Environ Res 96(3):
283-9.
BarrDB, Garry VF. 2004. Pesticides and Children. Toxicol Appl Pharmacol
198(2): 152-63.
Arbuckle TE, Cole DC, Ritter L, Ripley BD. Farm children's exposure to
herbicides: comparison of biomonitoring and questionnaire data.
Epidemiology 15(2): 187-94.
Freeman NC, Hore P, Black K, Jimenez M, Sheldon L, Tulve N, Lioy PJ.
Contributions of children's activities to pesticide hand loadings following
residential pesticide application. J Expo Anal Environ Epidemiol 15(1): 81-8.
Garry VF, Barr DB, Bravo R, Weerasekera G, Caltabiano LM, Whitehead
RD Jr, Olsson AO, Caudill SP, SchoberSE, Pirkle JL, Sampson EJ,
Jackson RJ, Needham LL. Concentration of dialkyl phosphate metabolites
of organophosphorus pesticides in the U.S. population. Environ Health
Perspectives 112(2): 186-200.
Coronado GD, Thompson B, Strong L, Griffith WC, Islas I. Agricultural task
and exposure to organophosphate pesticides among farmworkers. Environ
Health Perspectives 112(2): 142-7.
A-7
August 2005
-------
-------
Appendix B
Overview Tables for Relationships from Literature Review
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendix B
Overview Tables for Relationships from Literature Review
Contents
B.l Description B-l
B.2 Reference Information B-3
B.3 Overview Tables B-7
B.3.1 Source Relationships B-7
B.3.1.1 Category 1: Residential Pesticide Use B-7
B.3.1.2 Category 2: Household Characteristics B-16
B.3.1.3 Category 3: Residential Sources (Environmental Measures) B-20
B.3.1.4 Category 4: Household Occupation B-21
B.3.1.5 Category 5: Residential Proximity to Agricultural Fields B-25
B.3.1.6 Category 6: Residential Location B-27
B.3.2 Behavior Relationships B-28
B.3.2.1 Category 7: Subject's Personal Characteristics B-28
B.3.2.2 Category 8: Child's Behaviors B-31
B.3.2.3 Category 9: Dietary Behaviors B-32
B.3.2.4 Category 10: Family Hygiene Practices B-33
B.3.2.5 Category 11: Smoking-Related Activities B-36
B.3.2.6 Category 12: Work Exposure/Practices B-37
B.3.3 Other Relationships B-38
B.3.3.1 Category 13: Related Exposure Levels B-38
B.3.3.2 Category 14: Health B-39
B-ii August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Tables
Table B. 1.1 Example of Relationship Overview Table for Question Category: Residential Pesticide Use B-l
Table B.2.1 List of Columns and Associated Reference Tables in Overview Tables B-3
Table B.2.2 Chemical/Metabolite Reference Table B-4
Table B.2.3 Significance Indicator Reference Table B-5
Table B.2.4 Table Numbers Cross-Referenced between Results Section and Appendices A, B, and C, by Category Group B-6
Table B.3.1.1 Overview of Relationships for Questions in Category 1: Residential Pesticide Use - Grouped by Medium and Sorted by
Question and Citation B-7
Table B.3.1.1.a Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Urine Measurements, part 1 B-7
Table B.3.1.1.b Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Urine Measurements, part 2 B-ll
Table B.3.1.1.C Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Dust Measurements B-14
Table B.3.1.1.d Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Indoor Air Measurements B-14
Table B.3.1.1.e Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Outdoor Air Measurements B-14
Table B.3.1.1.f Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Personal Air Measurements B-15
Table B.3.1.1.g Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Solid Food Measurements B-15
Table B.3.1.2 Overview of Relationships for Questions in Category 2: Household Characteristics - Grouped by Medium and Sorted by
Question and Citation B-l6
Table B.3.1.2.a Overview of Relationships for Questions in Category 2: Household Characteristics with Urine Measurements, part 1 B-16
Table B.3.1.2.b Overview of Relationships for Questions in Category 2: Household Characteristics with Urine Measurements, part 2 B-17
Table B.3.1.2.C Overview of Relationships for Questions in Category 2: Household Characteristics with Dust Measurements B-l 8
Table B.3.1.2.d Overview of Relationships for Questions in Category 2: Household Characteristics with Indoor Air Measurements B-19
Table B.3.1.3 Overview of Relationships for Questions in Category 3: Residential Sources (Environmental Measures) - Grouped by
Medium and Sorted by Question and Citation B-20
Table B.3.1.3.a Overview of Relationships for Questions in Category 3: Residential Sources (Environmental Measures) with Urine
Measurements B-20
Table B.3.1.3.b Overview of Relationships for Questions in Category 3: Residential Sources (Environmental Measures) with Dust
Measurements B-20
Table B.3.1.4 Overview of Relationships for Questions in Category 4: Household Occupation - Grouped by Medium and Sorted by
Question and Citation B-21
B-iii
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table B.3.1.4.a
Table B.3.1.4.b
Table B.3.1.4.C
Table B.3.1.4.d
Table B.3.1.5
Table B.3.1.5.a
Table B.3.1.5.b
Table B.3.1.6
Table B.3.1.6.a
Table B.3.1.6.b
Table B.3.2.1
Table B.3.2.1.a
Table B.3.2.1.b
Table B.3.2.2
Table B.3.2.2.a
Table B.3.2.3
Table B.3.2.3.a
Table B.3.2.4
Table B.3.2.4.a
Table B.3.2.4.b
Overview of Relationships for Questions
Overview of Relationships for Questions
Overview of Relationships for Questions
Overview of Relationships for Questions
Overview of Relationships for Questions
and Sorted by Question and Citation
Overview of Relationships for Questions in Category 5: Residential Proximity to Agricultural Fields with Urine
Measurements
Overview of Relationships for Questions in Category 5: Residential Proximity to Agricultural Fields with Dust
Measurements
Overview of Relationships for Questions in Category 6: Residential Location - Grouped by Medium and Sorted by Question
and Citation
Overview of Relationships for Questions
Overview of Relationships for Questions
and Citation
in Category 4: Household Occupation with Urine Measurements, part 1 B-21
in Category 4: Household Occupation with Urine Measurements, part 2 B-22
in Category 4: Household Occupation with Dust Measurements B-23
in Category 4: Household Occupation with Soil Measurements B-24
in Category 5: Residential Proximity to Agricultural Fields - Grouped by Medium
B-25
B-25
B-26
B-27
in Category 6: Residential Location with Urine Measurements B-27
in Category 6: Residential Location with Dust Measurements B-27
in Category 7: Subject's Personal Characteristics - Grouped by Medium and Sorted
B-28
Overview of Relationships for Questions
Overview of Relationships for Questions
Overview of Relationships for Questions
by Question and Citation
Overview of Relationships for Questions in Category 7: Subject's Personal Characteristics with Urine Measurements,
part 1
B-28
Overview of Relationships for Questions in Category 7: Subject's Personal Characteristics with Urine Measurements,
part 2
B-29
Overview of Relationships for Questions in Category 8: Child's Behaviors - Grouped by Medium and Sorted by Question
and Citation
B-31
in Category 8: Child's Behaviors with Urine Measurements B-31
in Category 9: Dietary Behaviors - Grouped by Medium and Sorted by Question
B-3 2
Overview of Relationships for Questions in Category 9: Dietary Behaviors with Urine Measurements B-32
Overview of Relationships for Questions in Category 10: Family Hygiene Practices - Grouped by Medium and Sorted by
B-3 3
Overview of Relationships for Questions in Category 10: Family Hygiene Practices with Urine Measurements B-33
Overview of Relationships for Questions in Category 10: Family Hygiene Practices with Dust Measurements B-34
B-iv
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table B.3.2.5 Overview of Relationships for Questions in Category 11: Smoking-Related Activities - Grouped by Medium and Sorted by
Question and Citation B-36
Table B.3.2.5.a Overview of Relationships for Questions in Category 11: Smoking-Related Activities with Urine Measurements B-36
Table B.3.2.6 Overview of Relationships for Questions in Category 12: Work Exposure/Practices - Grouped by Medium and Sorted by
Question and Citation B-37
Table B.3.2.6.a Overview of Relationships for Questions in Category 12: Work Exposure/Practices with Urine Measurements B-37
Table B.3.2.6.b Overview of Relationships for Questions in Category 12: Work Exposure/Practices with Dust Measurements B-37
Table B.3.3.1 Overview of Relationships for Questions in Category 13: Related Exposure Levels - Grouped by Medium and Sorted by
Question and Citation B-38
Table B.3.3.1.a Overview of Relationships for Questions in Category 13: Related Exposure Levels with Urine Measurements B-38
Table B.3.3.2 Overview of Relationships for Questions in Category 14: Health - Grouped by Medium and Sorted by Question and
Citation B-39
Table B.3.3.2.a Overview of Relationships for Questions in Category 14: Health with Urine Measurements B-39
B-v
August 2005
-------
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendices B, C, and D provide specific information about the relationships extracted from the literature review and summarized in Results sections
4.2.4, 4.2.5, and 4.2.6. The information is presented as overview, detail, and comment tables. Each appendix includes one type of table for all the
question categories and relationships. This appendix presents the overview tables.
B.l Description
Table B.l.l is an example of an overview table and provides a high4evel view of the relationships found in the literature review for the source
category of residential pesticide use. The highest level of organization for this example is the sampling medium, that is, all relationships for urine are
grouped together, all relationships for dust are grouped together, etc. The next level of organization within a table is the chemical. The chemicals
analyzed for the medium are columns in the table. There may be more than one subtable for a particular medium depending on how many chemicals
or metabolites are represented in the relationships for the category. The chemicals for each medium are listed alphabetically except for the urinary
metabolites. These columns are arranged alphabetically within the following chemical groupings: non-DAPs (dialkylphosphates), single DAPs, DAP
sums, and level of DAPs. In example Table B.l.l, relationships for urine appear first, followed by those for dust. There are two subtables for the
urine relationships, one for the DAP metabolites, the other for molar-weighted sums of the DAP metabolites. There is only one table for the dust
relationships.
Table B.l.l Example of Relationship Overview Table for Question Category: Residential Pesticide Use
Urine, Part 1
Q#
Description
Citation
1NAP
4NITR
ATZM
MDA
TCPY
DEP
DETP
DEDTP
DMP
DMDTP
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
101
Pesticide Use
Krinsley 1998
1
101
Pesticide Use
Royster 2002
102
Inside Treated
Krinsley 1998
1
102
Inside Treated
Lu 2001
102
Inside Treated
Sexton 2003
1
1
103
Inside Treated -
Bathroom
Krinsley 1998
1
B-l
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Urine, Part 2
Q#
Description
Citation
ETHL1
ETHL2
MTHL1
MTHL2
ETHL3
MTHL3
MTHL4
DAP2
DAP3
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
101
Pesticide Use
Krinsley 1998
101
Pesticide Use
Royster 2002
2
2
102
Inside Treated
Krinsley 1998
102
Inside Treated
Lu 2001
1
1
102
Inside Treated
Sexton 2003
103
Inside Treated -
Bathroom
Krinsley 1998
Dust
Q#
Description
Citation
AZM
AZMPH
CHLR
OPSUM
S
NS
S
NS
S
NS
s
NS
101
Pesticide Use
McCauley 2001a
1
101
Pesticide Use
McCauley 2003
1
102
Inside Treated
Sexton 2003
1
The rows within a medium's table(s) are the questions assigned to the residential pesticide use category, and are identified by a question number and
the abbreviated question description. When a medium's information is presented in more than one subtable, e.g., urine in Table B.l.l, combinations
of question/citation rows are repeated in all of the medium's subtables. Thus in Table B.l.l Urine Part 1, the "inside treated-bathroom" question (Q#
103) appears in both urine subtables even though relationships are available only for the metabolites in the first subtable. The overview table shows
the number of significant or non-significant relationships for each combination of medium, chemical, question, and citation in the relationship
database. Two columns are included for each chemical or metabolite. The S column shows the number of relationships identified as significant; the
B-2
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
NS columns shows the number of relationships identified as not significant. Marginally significant relationships are included here under the S
column, but are specifically identified as MS in the detail tables in Appendix C. Note that the relationships counted in any cell may represent
different subpopulations compared or different types of analyses performed. Rows for alternating questions are shaded for ease of viewing.
B.2 Reference Information
To make the overview tables more compact, it was necessary to use abbreviations or codes in both the column name and contents. Table B.2.1
describes each column used in the overview tables. The column Reference Table identifies the number of a subsequent table containing information
about the codes used. For example, the column (S, NS) includes codes described in Table B.2.3.
Table B.2.1 List of Columns and Associated Reference Tables in Overview Tables
Column Type or
Name
Column Description
Column
Applies
toa
Reference Tableb
Q#
Number assigned to an abbreviated question (Sec
4.2.2.2)
b
NA
Description
Abbreviated question
b
NA
Citation
Citation reference
Appendix A - Table A.1 and Table 4.2.1
(Chemical columns)
Chemical, metabolite, or molar-weighted sum
a
Appendix B - Table B.2.2
(S, NS)
Significance indicator
Appendix B - Table B.2.3
a The entry "a" is a dependent variable, in this case a chemical analytical measurement. The entry "b" is an independent variable or predictor, usually a question.
b NA- Not applicable
B-3
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table B.2.2 Chemical/Metabolite Reference Table
Grouping3
Code
Medium
Description
1-Non-DAP
1NAP
urine
1-Naphthol
1-Non-DAP
4NITR
urine
4-Nitrophenol
6-Chemical
ATZ
otherb
Atrazine
1-Non-DAP
ATZM
urine
Atrazine mercapturate
6-Chemical
AZM
other
Azinphosmethyl
6-Chemical
AZMPH
other
Azinphosmethyl+Phosmet
6-Chemical
CHLR
other
Chlorpyrifos
3-DAP Sumc
DAP1
urine
DMP+DMTP+DMDTP+DEP+DETP+DEDTP
4-DAP Detect
DAP2
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one detectable measurement)
5-DAP High
DAP3
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one high measurement)11
2-DAP
DEDTP
urine
Diethyldithiophosphate (DEDTP)
2-DAP
DEP
urine
Diethylphosphate (DEP)
2-DAP
DETP
urine
Diethylthiophosphate (DETP)
2-DAP
DMDTP
urine
Dimethyldithiophosphate (DMDTP)
2-DAP
DMP
urine
Dimethylphosphate (DMP)
2-DAP
DMTP
urine
Dimethylthiophosphate (DMTP)
6-Chemical
EPAR
other
Ethyl parathion
3-DAP Sum
ETHL1
urine
DEP+DETP
3-DAP Sum
ETHL2
urine
DEP+DETP+DEDTP
4-DAP Detect
ETHL3
urine
DEP, DETP, DEDTP
(at least one detectable measurement)
6-Chemical
MAL
other
Malathion
B-4
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Grouping3
Code
Medium
Description
1-Non-DAP
MDA
urine
Malathion dicarboxylic acid
3-DAP Sum
MTHL1
urine
DMTP+DMDTP
3-DAP Sum
MTHL2
urine
DMP+DMTP+DMDTP
4-DAP Detect
MTHL3
urine
DMTP (detectable measurement)
4-DAP Detect
MTHL4
urine
DMP, DMTP
(at least one detectable measurement)
5-DAP High
MTHL5
urine
DMP, DMTP
(at least one high measurement)11
7-M eta bo lite NA
NA
urine
NA (not available or not specified)
6-Chemical
OPSUM
other
OP sum®
6-Chemical
PHSM
other
Phosmet
1-Non-DAP
TCPY
urine
3,5,6-T richloro-2-pyridinol
a The number preceding the group name indicates the order of the group as it appears in the overview tables.
b Medium is urine and other (any other medium measured).
c Sums are molar-weighted unless otherwise specified.
d See definition of high measurement in Azaroff (1999).
e OP Sum = azinphosmethyl, chlorpyrifos, malathion, and phosmet
Table B.2.3 Significance Indicator Reference Table
Code
Description
MS
Relationship is marginally significant based on critical value used in publication
NA
Significance level is not available in publication
NS
Relationship is not significant based on critical value used in publication
S
Relationship is significant based on critical value used in publication
B-5
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table B.2.4 provides a cross-reference between the relationship summary tables in the Results section and the tables in Appendices B, C, and D.
Table B.2.4 Table Numbers Cross-Referenced between Results Section and Appendices A, B, and C, by Category Group
Category
Section #
Table #a
Overview
Table #
Detailed
Table #
Comment
Table #
Group
#
Description
Results
Results
Appendix B
Appendix C
Appendix D
Source
1
Residential pesticide use
4.2.4.1
4.2.6.x
B.3.1.1
C.3.1.1
D.3.1.1
Source
2
Household characteristics
4.2.4.2
4.2.7.x
B.3.1.2
C.3.1.2
D.3.1.2
Source
3
Residential sources
(environmental measures)
4.2.4.3
4.2.8.x
B.3.1.3
C.3.1.3
D.3.1.3
Source
4
Household occupation
4.2.4.4
4.2.9.x
B.3.1.4
C.3.1.4
D.3.1.4
Source
5
Residential proximity to
agricultural fields
4.2.4.5
4.2.10.x
B.3.1.5
C.3.1.5
D.3.1.5
Source
6
Residential location
4.2.4.6
4.2.11.x
B.3.1.6
C.3.1.6
D.3.1.6
Behavior
7
Subject's personal
characteristics
4.2.5.1
4.2.13.x
B.3.2.1
C.3.2.1
D.3.2.1
Behavior
8
Child's behaviors
4.2.5.2
4.2.14.x
B.3.2.2
C.3.2.2
D.3.2.2
Behavior
9
Dietary behaviors
4.2.5.3
4.2.15.x
B.3.2.3
C.3.2.3
D.3.2.3
Behavior
10
Family hygiene practices
4.2.5.4
4.2.16.x
B.3.2.4
C.3.2.4
D.3.2.4
Behavior
11
Smoking-related activities
4.2.5.5
4.2.17.x
B.3.2.5
C.3.2.5
D.3.2.5
Behavior
12
Work exposure/practices
4.2.5.6
4.2.18.x
B.3.2.6
C.3.2.6
D.3.2.6
Other
13
Related exposure levels
4.2.6.1
4.2.20.x
B.3.3.1
C.3.3.1
D.3.3.1
Other
14
Health
4.2.6.2
4.2.21.x
B.3.3.2
C.3.3.2
D.3.3.2
a x in this column refers to the three table types, a, b, and c, described above.
B-6
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3 Overview Tables
B.3.1 Source Relationships
B.3.1.1 Category 1: Residential Pesticide Use
Table B.3.1.1 Overview of Relationships for Questions in Category 1: Residential Pesticide Use - Grouped by Medium and Sorted by Question and Citation
a) Urine, part 1
b) Urine, part 2
c) Dust
d) Indoor Air
e) Outdoor Air
f) Personal Air
g) Solid Food
Table B.3.1.1.a Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Urine Measurements, part 1
Q#
Description
Citation
1NAP
4NITR
ATZM
MDA
TCPY
DEP
DETP
DEDTP
DMP
DMDTP
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
101
Pesticide Use
Krinsley 1998
1
101
Pesticide Use
Royster 2002
102
Inside Treated
Krinsley 1998
1
102
Inside Treated
Lu 2001
102
Inside Treated
Sexton 2003
1
1
103
Inside Treated -
Bathroom
Krinsley 1998
1
104
Inside Treated - Bedroom
Krinsley 1998
3
B-7
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
1NAP
4NITR
ATZM
MDA
TCPY
DEP
DETP
DEDTP
DMP
DMDTP
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
105
Inside Treated - Cabinets
Krinsley 1998
1
106
Inside Treated - Closets
Krinsley 1998
1
107
Inside Treated -
Cupboards With Dishes
Krinsley 1998
1
108
Inside Treated - Dining
Room
Krinsley 1998
1
109
Inside Treated - Family
Room
Krinsley 1998
1
110
Inside Treated - Kitchen
Krinsley 1998
1
111
Inside Treated - Living
Room
Krinsley 1998
1
112
Inside Treated - on
Baseboards
Krinsley 1998
1
113
Inside Treated - on
Ceiling
Krinsley 1998
1
114
Inside Treated - on Floor
Krinsley 1998
1
115
Inside Treated - on
Lower Walls
Krinsley 1998
1
116
Inside Treated - on
Upper Walls
Krinsley 1998
1
117
Inside Treated - Other
Room
Krinsley 1998
2
B-8
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
1NAP
4NITR
ATZM
MDA
TCPY
DEP
DETP
DEDTP
DMP
DMDTP
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
118
Pets Treated
Lu 2000
118
Pets Treated
Lu 2001
119
Outside Treated
Krinsley 1998
1
119
Outside Treated
Sexton 2003
1
1
120
Garden Treated
Fenske 2002
1
1
120
Garden Treated
Lu 2000
120
Garden Treated
Lu 2001
121
Lawn/Yard Treated
Lu 2000
121
Lawn/Yard Treated
Lu 2001
121
Lawn/Yard Treated
Sexton 2003
1
122
Inside or Outside Treated
Ad gate 2001
1
1
1
122
Inside or Outside Treated
Aprea 2000
2
1
2
2
1
123
Previous Treatment
Lu 2000
124
Level of Pesticide Use
Ad gate 2001
1
1
1
124
Level of Pesticide Use
Krinsley 1998
2
124
Level of Pesticide Use
Sexton 2003
2
B-9
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
1NAP
4NITR
ATZM
MDA
TCPY
DEP
DETP
DEDTP
DMP
DMDTP
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
125
Frequency Personal
Application Inside
Krinsley 1998
1
126
Frequency Personal
Application Outside
Krinsley 1998
3
127
Inside/Outside Treated
by Family Member
Azaroff 1999
128
Frequency Professional
Application Inside
Krinsley 1998
1
129
Frequency Professional
Application Outside
Krinsley 1998
1
130
Personally Mixed
Pesticide Inside
Krinsley 1998
1
131
Personally Mixed
Pesticide Outside
Krinsley 1998
1
132
Presence During Mixing
Ad gate 2001
1
1
1
1
B-10
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table B.3.1.1.b Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Urine Measurements, part 2
Q#
Description
Citation
ETHL1
ETHL2
MTHL1
MTHL2
ETHL3
MTHL3
MTHL4
DAP2
DAP3
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
101
Pesticide Use
Krinsley 1998
101
Pesticide Use
Royster 2002
2
2
102
Inside Treated
Krinsley 1998
102
Inside Treated
Lu 2001
1
1
102
Inside Treated
Sexton 2003
103
Inside Treated -
Bathroom
Krinsley 1998
104
Inside Treated - Bedroom
Krinsley 1998
105
Inside Treated - Cabinets
Krinsley 1998
106
Inside Treated - Closets
Krinsley 1998
107
Inside Treated -
Cupboards With Dishes
Krinsley 1998
108
Inside Treated - Dining
Room
Krinsley 1998
109
Inside Treated - Family
Room
Krinsley 1998
110
Inside Treated - Kitchen
Krinsley 1998
111
Inside Treated - Living
Room
Krinsley 1998
112
Inside Treated - on
Baseboards
Krinsley 1998
113
Inside Treated - on
Ceiling
Krinsley 1998
B-ll
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
ETHL1
ETHL2
MTHL1
MTHL2
ETHL3
MTHL3
MTHL4
DAP2
DAP3
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
114
Inside Treated - on Floor
Krinsley 1998
115
Inside Treated - on
Lower Walls
Krinsley 1998
116
Inside Treated - on
Upper Walls
Krinsley 1998
117
Inside Treated - Other
Room
Krinsley 1998
118
Pets Treated
Lu 2000
1
118
Pets Treated
Lu 2001
1
1
119
Outside Treated
Krinsley 1998
119
Outside Treated
Sexton 2003
120
Garden Treated
Fenske 2002
120
Garden Treated
Lu 2000
1
120
Garden Treated
Lu 2001
1
1
121
Lawn/Yard Treated
Lu 2000
1
121
Lawn/Yard Treated
Lu 2001
1
1
121
Lawn/Yard Treated
Sexton 2003
122
Inside or Outside Treated
Ad gate 2001
122
Inside or Outside Treated
Aprea 2000
2
123
Previous Treatment
Lu 2000
1
124
Level of Pesticide Use
Ad gate 2001
124
Level of Pesticide Use
Krinsley 1998
124
Level of Pesticide Use
Sexton 2003
B-12
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
ETHL1
ETHL2
MTHL1
MTHL2
ETHL3
MTHL3
MTHL4
DAP2
DAP3
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
125
Frequency Personal
Application Inside
Krinsley 1998
126
Frequency Personal
Application Outside
Krinsley 1998
127
Inside/Outside Treated
By Family Member
Azaroff 1999
1
1
1
1
1
128
Frequency Professional
Application Inside
Krinsley 1998
129
Frequency Professional
Application Outside
Krinsley 1998
130
Personally Mixed
Pesticide Inside
Krinsley 1998
131
Personally Mixed
Pesticide Outside
Krinsley 1998
132
Presence During Mixing
Ad gate 2001
B-13
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table B.3.1.1.C Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Dust Measurements
Q#
Description
Citation
AZM
AZMPH
CHLR
OPSUM
S
NS
S
NS
S
NS
s
NS
101
Pesticide Use
McCauley 2001a
1
101
Pesticide Use
McCauley 2003
1
102
Inside Treated
Sexton 2003
1
118
Pets Treated
Lu 2000
1
119
Outside Treated
Sexton 2003
1
120
Garden Treated
Lu 2000
1
121
Lawn/Yard Treated
Lu 2000
1
123
Previous Treatment
Lu 2000
1
Table B.3.1.1.d Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Indoor Air Measurements
Q#
Description
Citation
CHLR
MAL
S
NS
S
NS
102
Inside Treated
Sexton 2003
1
1
Table B.3.1.1.e Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Outdoor Air Measurements
Q#
Description
Citation
CHLR
MAL
S
NS
S
NS
102
Inside Treated
Sexton 2003
1
1
B-14
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table B.3.1.1.f Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Personal Air Measurements
Q#
Description
Citation
ATZ
CHLR
MAL
S
NS
S
NS
S
NS
102
Inside Treated
Sexton 2003
1
1
124
Level of
Pesticide Use
Sexton 2003
2
Table B.3.1.1.g Overview of Relationships for Questions in Category 1: Residential Pesticide Use with Solid Food Measurements
Q#
Description
Citation
CHLR
MAL
S
NS
S
NS
102
Inside Treated
Sexton 2003
1
1
119
Outside
T reated
Sexton 2003
1
B-15
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.1.2 Category 2: Household Characteristics
Table B.3.1.2 Overview of Relationships for Questions in Category 2: Household Characteristics - Grouped by Medium and Sorted by Question and Citation
a) Urine
b) Dust
c) Indoor Air
Table B.3.1.2.a Overview of Relationships for Questions in Category 2: Household Characteristics with Urine Measurements, part 1
Q#
Description
Citation
MDA
TCPY
DEP
DETP
DEDTP
DMP
DMTP
DMDTP
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
201
Housing Type
Lu 2001
203
Age of House
> 10 Years
Krinsley 1998
1
204
Age of House
> 20 Years
Krinsley 1998
1
205
Having Air
Conditioning
Krinsley 1998
1
206
Having Central
Heating
Krinsley 1998
1
207
Having
Evaporative
Cooling
Krinsley 1998
1
208
Pets in House
Aprea 2000
2
2
2
2
1
1
208
Pets in House
Lu 2001
209
Pets Inside/
Outside House
Sexton 2003
1
B-16
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
MDA
TCPY
DEP
DETP
DEDTP
DMP
DMTP
DMDTP
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
211
Existence of
Garden or
Vegetable
Garden
Aprea 2000
2
2
2
2
1
211
Existence of
Garden or
Vegetable
Garden
Lu 2001
211
Existence of
Garden or
Vegetable
Garden
Sexton 2003
1
212
Ornamental
Plants or Cut
Flowers
Aprea 2000
2
2
2
2
1
1
Table B.3.1.2.b Overview of Relationships for Questions in Category 2: Household Characteristics with Urine Measurements, part 2
Q#
Description
Citation
ETHL1
ETHL2
MTHL2
DAP1
S
NS
S
NS
S
NS
S
NS
201
Housing Type
Lu 2001
1
1
203
Age of House > 10 Years
Krinsley 1998
204
Age of House > 20 Years
Krinsley 1998
205
Having Air Conditioning
Krinsley 1998
206
Having Central Heating
Krinsley 1998
207
Having Evaporative Cooling
Krinsley 1998
B-17
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
ETHL1
ETHL2
MTHL2
DAP1
S
NS
S
NS
S
NS
S
NS
208
Pets in House
Aprea 2000
2
1
1
208
Pets in House
Lu 2001
1
1
209
Pets Inside/Outside House
Sexton 2003
211
Existence of Garden or
Vegetable Garden
Aprea 2000
2
1
1
211
Existence of Garden or
Vegetable Garden
Lu 2001
1
1
211
Existence of Garden or
Vegetable Garden
Sexton 2003
212
Ornamental Plants or Cut
Flowers
Aprea 2000
2
1
1
Table B.3.1.2.C Overview of Relationships for Questions in Category 2: Household Characteristics with Dust Measurements
Q#
Description
Citation
AZM
CHLR
EPAR
OPSUM
PHSM
S
NS
S
NS
S
NS
s
NS
S
NS
201
Housing Type
McCauley 2001a
1
202
Property Used as a Farm
Sexton 2003
1
210
Pet Inside to Outside
Simcox 1995
2
2
2
2
213
Size of Household
McCauley 2001a
1
213
Size of Household
McCauley 2003
1
214
Location of Play Area
McCauley 2003
1
215
Age of House (Years)
McCauley 2003
1
1
B-18
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
AZM
CHLR
EPAR
OPSUM
PHSM
S
NS
S
NS
S
NS
s
NS
S
NS
216
Size of Home (Sq Ft)
McCauley 2003
1
1
217
Number of Pets In House
McCauley 2003
1
Table B.3.1.2.d Overview of Relationships for Questions in Category 2: Household Characteristics with Indoor Air Measurements
Q#
Description
Citation
CHLR
S
NS
202
Property Used as a Farm
Sexton 2003
1
B-19
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.1.3 Category 3: Residential Sources (Environmental Measures)
Table B.3.1.3 Overview of Relationships for Questions in Category 3: Residential Sources (Environmental Measures) - Grouped by Medium and Sorted by Question
and Citation
a) Urine
b) Dust
Table B.3.1.3.a Overview of Relationships for Questions in Category 3: Residential Sources (Environmental Measures) with Urine Measurements
Q#
Description
Citation
MTHL2
DAP1
NA
S
NS
S
NS
S
NS
301
Household
Dust
Curl 2002
2
301
Household
Dust
Lu 2000
2
302
Loading from
Household
Floor Dust
Shalat 2003
1
Table B.3.1.3.b Overview of Relationships for Questions in Category 3: Residential Sources (Environmental Measures) with Dust Measurements
Q#
Description
Citation
AZM
CHLR
EPAR
PHSM
S
NS
S
NS
S
NS
S
NS
303
Outdoor Soil
Simcox 1995
1
1
1
1
2
1
1
B-20
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.1.4 Category 4: Household Occupation
Table B.3.1.4 Overview of Relationships for Questions in Category 4: Household Occupation - Grouped by Medium and Sorted by Question and Citation
a) Urine, part 1
b) Urine, Part 2
c) Dust
d) Soil
Table B.3.1.4.a Overview of Relationships for Questions in Category 4: Household Occupation with Urine Measurements, part 1
Q#
Description
Citation
4NITR
TCPY
DMTP
DMDTP
ETHL2
MTHL1
MTHL2
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
402
Household Member
Spraying Fields
Azaroff 1999
403
Recent Fieldwork
Azaroff 1999
404
Applicator vs Farmworker
Fenske 2002
1
1
404
Applicator vs Farmworker
Lu 2000
1
1
1
406
Applicator vs Reference
Loewenherz
1997
5
3
407
Ap p I icato r+ Fa rmwo rke r
vs Reference
Fenske 2002
2
2
407
Ap p I icato r+ Fa rmwo rke r
vs Reference
Lu 2000
1
1
1
411
Farmworker vs Others
Koch 2002
2
2
413
Expected Occupational
Exposure
Koch 1999
3
3
414
Occupational Pesticide
Exposure
Royster 2002
2
2
B-21
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table B.3.1.4.b Overview of Relationships for Questions in Category 4: Household Occupation with Urine Measurements, part 2
Q#
Description
Citation
ETHL3
MTHL3
MTHL4
DAP2
MTHL5
DAP3
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
402
Household Member
Spraying Fields
Azaroff 1999
2
1
2
1
1
1
403
Recent Fieldwork
Azaroff 1999
2
1
1
1
1
404
Applicator vs Farmworker
Fenske 2002
404
Applicator vs Farmworker
Lu 2000
406
Applicator vs Reference
Loewenherz
1997
407
Ap p I icato r+ Fa rmwo rke r
vs Reference
Fenske 2002
407
Ap p I icato r+ Fa rmwo rke r
vs Reference
Lu 2000
411
Farmworker vs Others
Koch 2002
413
Expected Occupational
Exposure
Koch 1999
414
Occupational Pesticide
Exposure
Royster 2002
B-22
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table B.3.1.4.C Overview of Relationships for Questions in Category 4: Household Occupation with Dust Measurements
Q#
Description
Citation
AZM
AZMPH
CHLR
EPAR
OPSUM
PHSM
S
NS
S
NS
S
NS
S
NS
s
NS
S
NS
401
Agricultural Workers in
Household
McCauley
2001a
2
404
Applicator vs Farmworker
Fenske 2002
1
1
404
Applicator vs Farmworker
Lu 2000
1
1
1
405
Applicator vs Non-
Applicator
Simcox 1995
3
2
1
4
1
3
407
Ap p I icato r+ Fa rmwo rke r
vs Reference
Fenske 2002
3
2
1
407
Ap p I icato r+ Fa rmwo rke r
vs Reference
Lu 2000
1
1
1
408
Farmer vs Farmworker
Simcox 1995
3
3
2
3
3
409
Farmer+Farmworker vs
Reference
Simcox 1995
1
1
1
1
1
1
2
410
Farmworker vs Grower
McCauley
2001a
1
1
412
Field Worker vs Pesticide
Handler
Grossman
2001
1
413
Expected Occupational
Exposure
Grossman
2001
1
1
415
Tree Thinning
McCauley 2003
1
416
Number in Household
with High Pesticide
Contact
McCauley 2003
1
B-23
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table B.3.1.4.d Overview of Relationships for Questions in Category 4: Household Occupation with Soil Measurements
Q#
Description
Citation
AZM
CHLR
EPAR
PHSM
S
NS
S
NS
S
NS
S
NS
408
Farmer vs Farmworker
Simcox 1995
1
1
1
1
409
Farmer+Farmworker vs
Reference
Simcox 1995
1
1
1
1
B-24
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.1.5 Category 5: Residential Proximity to Agricultural Fields
Table B.3.1.5 Overview of Relationships for Questions in Category 5: Residential Proximity to Agricultural Fields - Grouped by Medium and Sorted by Question
and Citation
a) Urine
b) Dust
Table B.3.1.5.a Overview of Relationships for Questions in Category 5: Residential Proximity to Agricultural Fields with Urine Measurements
Q#
Description
Citation
4NITR
TCPY
DMTP
DMDTP
ETHL2
MTHL1
MTHL2
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
Curl 2002
4
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
Fenske 2002
1
1
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
Koch 2002
2
2
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
Loewenherz
1997
2
1
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
Lu 2000
1
1
1
2
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
Royster 2002
4
4
502
Living near Multiple Fields
Royster 2002
2
2
B-25
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table B.3.1.5.b Overview of Relationships for Questions in Category 5: Residential Proximity to Agricultural Fields with Dust Measurements
Q#
Description
Citation
AZM
AZMPH
CHLR
EPAR
OPSUM
PHSM
S
NS
S
NS
S
NS
S
NS
s
NS
S
NS
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
Curl 2002
2
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
Fenske 2002
2
1
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
Grossman
2001
2
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
Lu 2000
1
4
1
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
McCauley
2001a
1
1
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
McCauley
2003
1
501
Proximity of Home to Pesticide-
Treated Farmland/Orchard
Simcox 1995
2
4
1
5
5
3
6
B-26
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.1.6 Category 6: Residential Location
Table B.3.1.6 Overview of Relationships for Questions in Category 6: Residential Location - Grouped by Medium and Sorted by Question and Citation
a) Urine
b) Dust
Table B.3.1.6.a Overview of Relationships for Questions in Category 6: Residential Location with Urine Measurements
Q#
Description
Citation
1NAP
MDA
TCPY
ETHL1
MTHL2
S
NS
S
NS
S
NS
S
NS
S
NS
601
Urban vs Non-Urban
Ad gate 2001
3
3
3
602
Urban vs Rural
Krinsley 1998
1
603
Border vs. Non-Border
Krinsley 1998
1
604
Community
Lu 2001
1
1
Table B.3.1.6.b Overview of Relationships for Questions in Category 6: Residential Location with Dust Measurements
Q#
Description
Citation
AZM
S
NS
605
Vehicle vs House
Curl 2002
1
B-27
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.2 Behavior Relationships
B.3.2.1 Category 7: Subject's Personal Characteristics
Table B.3.2.1 Overview of Relationships for Questions in Category 7: Subject's Personal Characteristics - Grouped by Medium and Sorted by Question and Citation
a) Urine, part 1
b) Urine, part 2
Table B.3.2.1.a Overview of Relationships for Questions in Category 7: Subject's Personal Characteristics with Urine Measurements, part 1
Q#
Description
Citation
1NAP
MDA
TCPY
DEP
DETP
DEDTP
DMP
DMTP
DMDTP
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
701
Sex
Ad gate 2001
1
1
1
701
Sex
Aprea 2000
2
2
2
2
2
2
701
Sex
Koch 1999
701
Sex
Koch 2002
701
Sex
Krinsley 1998
1
701
Sex
Lu 2001
701
Sex
Shalat 2003
702
Age
Ad gate 2001
1
1
1
702
Age
Curl 2002
702
Age
Koch 1999
702
Age
Koch 2002
702
Age
Krinsley 1998
1
1
702
Age
Loewenherz
1997
4
10
702
Age
Lu 2001
B-28
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
1NAP
MDA
TCPY
DEP
DETP
DEDTP
DMP
DMTP
DMDTP
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
702
Age
Shalat 2003
703
Ethnicity
Ad gate 2001
1
1
703
Ethnicity
Krinsley 1998
1
704
Education
Level
Krinsley 1998
1
705
Income
Ad gate 2001
1
3
2
705
Income
Krinsley 1998
1
705
Income
Lu 2001
707a
Hand's Surface
Area
Shalat 2003
a There is no question grouping for the number 706.
Table B.3.2.1.b Overview of Relationships for Questions in Category 7: Subject's Personal Characteristics with Urine Measurements, part 2
Q#
Description
Citation
ETHL1
ETHL2
MTHL2
DAP1
S
NS
S
NS
S
NS
S
NS
701
Sex
Ad gate 2001
701
Sex
Aprea 2000
2
2
2
701
Sex
Koch 1999
1
1
701
Sex
Koch 2002
1
1
701
Sex
Krinsley 1998
701
Sex
Lu 2001
1
1
B-29
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
ETHL1
ETHL2
MTHL2
DAP1
S
NS
S
NS
S
NS
S
NS
701
Sex
Shalat 2003
2
702
Age
Ad gate 2001
702
Age
Curl 2002
2
4
702
Age
Koch 1999
1
1
702
Age
Koch 2002
1
1
702
Age
Krinsley 1998
702
Age
Loewenherz
1997
702
Age
Lu 2001
1
1
702
Age
Shalat 2003
2
703
Ethnicity
Ad gate 2001
703
Ethnicity
Krinsley 1998
704
Education
Level
Krinsley 1998
705
Income
Ad gate 2001
705
Income
Krinsley 1998
705
Income
Lu 2001
1
1
707a
Hand's Surface
Area
Shalat 2003
2
a There is no question grouping for the number 706.
B-30
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.2.2 Category 8: Child's Behaviors
Table B.3.2.2 Overview of Relationships for Questions in Category 8: Child's Behaviors - Grouped by Medium and Sorted by Question and Citation
a) Urine
Table B.3.2.2.a Overview of Relationships for Questions in Category 8: Child's Behaviors with Urine Measurements
Q#
Description
Citation
4NITR
TCPY
ETHL1
MTHL1
MTHL2
DAP1
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
801
Hand-to-Mouth Activity
Fenske 2002
1
1
801
Hand-to-Mouth Activity
Lu 2000
1
801
Hand-to-Mouth Activity
Lu 2001
1
1
802
Thumb-Sucking
Fenske 2002
1
1
802
Thumb-Sucking
Lu 2000
1
802
Thumb-Sucking
Lu 2001
1
1
803
Hand Washing Before Meals
Fenske 2002
1
1
803
Hand Washing Before Meals
Lu 2000
1
804
Frequency of Handwashing
Lu 2001
1
1
805
Time Spent Outdoors
Fenske 2002
1
1
805
Time Spent Outdoors
Lu 2000
1
806
Loading From Hand Wipe
Shalat 2003
2
B-31
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.2.3 Category 9: Dietary Behaviors
Table B.3.2.3 Overview of Relationships for Questions in Category 9: Dietary Behaviors - Grouped by Medium and Sorted by Question and Citation
a) Urine
Table B.3.2.3.a Overview of Relationships for Questions in Category 9: Dietary Behaviors with Urine Measurements
Q#
Description
Citation
TCPY
DEP
DETP
DEDTP
DMP
ETHL2
MTHL2
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
901
Type of Drinking Water
Krinsley 1998
1
902
Consumption of Homegrown
Fresh Vegetables
Krinsley 1998
1
903
Ate Lunch at School
Aprea 2000
2
2
2
2
2
904
Organic Diet
Curl 2003
2
2
B-32
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.2.4 Category 10: Family Hygiene Practices
Table B.3.2.4 Overview of Relationships for Questions in Category 10: Family Hygiene Practices - Grouped by Medium and Sorted by Question and Citation
a) Urine
b) Dust
Table B.3.2.4.a Overview of Relationships for Questions in Category 10: Family Hygiene Practices with Urine Measurements
Q#
Description
Citation
4NITR
TCPY
DMTP
ETHL1
MTHL1
MTHL2
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
1001
Shoes Removed at Door
Carrel 1996
2
2
1001
Shoes Removed at Door
Lu 2000
1
1002
Presence of Doormats
Fenske 2002
1
1
1002
Presence of Doormats
Lu 2000
1
1003
Presence of Floor Mats
Lu 2001
1
1
1004
Vacuuming Frequency
Fenske 2002
1
1
1004
Vacuuming Frequency
Lu 2000
1
1004
Vacuuming Frequency
Lu 2001
1
1
1006
Work Clothes Worn
Indoors
Carrel 1996
4
1006
Work Clothes Worn
Indoors
Fenske 2002
1
1
1006
Work Clothes Worn
Indoors
Lu 2000
1
1007
Work Clothes Mixed with
Laundry
Lu 2000
1
1008
Laundering Practices
Fenske 2002
1
1
B-33
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
4NITR
TCPY
DMTP
ETHL1
MTHL1
MTHL2
S
NS
S
NS
S
NS
S
NS
S
NS
S
NS
1008
Laundering Practices
Carrel 1996
4
1010
Shower Soon After Work
Carrel 1996
4
Table B.3.2.4.b Overview of Relationships for Questions in Category 10: Family Hygiene Practices with Dust Measurements
Q#
Description
Citation
AZM
AZMPH
CHLR
EPAR
OPSUM
PHSM
S
NS
S
NS
S
NS
S
NS
s
NS
S
NS
1001
Shoes Removed at Door
Grossman 2001
1
1001
Shoes Removed at Door
Lu 2000
1
1001
Shoes Removed at Door
McCauley 2003
1
1
1001
Shoes Removed at Door
Simcox 1995
2
2
2
2
1002
Presence of Doormats
Fenske 2002
1
1
1002
Presence of Doormats
Lu 2000
1
1002
Presence of Doormats
Simcox 1995
2
2
2
2
1004
Vacuuming Frequency
Fenske 2002
1
1
1004
Vacuuming Frequency
Lu 2000
1
1005
Vacuuming Indoor Play Areas
Simcox 1995
2
2
2
2
1006
Work Clothes Worn Indoors
Fenske 2002
1
1
1006
Work Clothes Worn Indoors
Lu 2000
1
1006
Work Clothes Worn Indoors
McCauley 2003
1
1
1007
Work Clothes Mixed with Laundry
Lu 2000
1
B-34
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q#
Description
Citation
AZM
AZMPH
CHLR
EPAR
OPSUM
PHSM
S
NS
S
NS
S
NS
S
NS
s
NS
S
NS
1008
Laundering Practices
Fenske 2002
1
1
1009
Number of Days Since Last
Vacuuming
McCauley 2003
1
1010
Shower Soon After Work
McCauley 2003
1
1
1010
Shower Soon After Work
Grossman 2001
1
1012
After Work Hygiene Index
McCauley 2003
1
1
B-35
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.2.5 Category 11: Smoking-Related Activities
Table B.3.2.5 Overview of Relationships for Questions in Category 11: Smoking-Related Activities - Grouped by Medium and Sorted by Question and Citation
a) Urine
Table B.3.2.5.a Overview of Relationships for Questions in Category 11: Smoking-Related Activities with Urine Measurements
Q#
Description
Citation
TCPY
S
NS
1101
Current Smoker
Krinsley 1998
1
1102
Subject Smoked
Krinsley 1998
2
1103
Exposure to Second Hand Smoke
Krinsley 1998
1
B-36
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.2.6 Category 12: Work Exposure/Practices
Table B.3.2.6 Overview of Relationships for Questions in Category 12: Work Exposure/Practices - Grouped by Medium and Sorted by Question and Citation
a) Urine
b) Dust
Table B.3.2.6.a Overview of Relationships for Questions in Category 12: Work Exposure/Practices with Urine Measurements
Q#
Description
Citation
TCPY
S
NS
1201
Pesticide Exposure at Work in Past 6 Mo
Krinsley 1998
1
Table B.3.2.6.b Overview of Relationships for Questions in Category 12: Work Exposure/Practices with Dust Measurements
Q#
Description
Citation
AZM
S
NS
1202
Wear Boots While Doing Fieldwork?
Grossman 2001
1
1203
Wear Gloves While Doing Fieldwork?
Grossman 2001
1
1204
Wear Hat While Doing Fieldwork?
Grossman 2001
1
B-37
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.3 Other Relationships
B.3.3.1 Category 13: Related Exposure Levels
Table B.3.3.1 Overview of Relationships for Questions in Category 13: Related Exposure Levels - Grouped by Medium and Sorted by Question and Citation
a) Urine
Table B.3.3.1.a Overview of Relationships for Questions in Category 13: Related Exposure Levels with Urine Measurements
Q#
Description
Citation
MTHL4
DAP2
DAP3
S
NS
S
NS
S
NS
1301
Detectable Levels in Adult
Household Members
Azaroff 1999
1
1
1302
High Levels in Adult
Household Members
Azaroff 1999
1
1
1
B-38
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
B.3.3.2 Category 14: Health
Table B.3.3.2 Overview of Relationships for Questions in Category 14: Health - Grouped by Medium and Sorted by Question and Citation
a) Urine
Table B.3.3.2.a Overview of Relationships for Questions in Category 14: Health with Urine Measurements
Q#
Description
Citation
TCPY
S
NS
1401
Health Status
Krinsley 1998
1
1402
Asthma and Allergies
Krinsley 1998
1
1403
Bowel Disease
Krinsley 1998
1
1404
Diabetes
Krinsley 1998
1
1405
Intestinal Disease
Krinsley 1998
3
1406
Ulcer
Krinsley 1998
1
B-39
August 2005
-------
-------
Appendix C
Detail Tables for Relationships from Literature Review
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendix C
Detail Tables for Relationships from Literature Review
Contents
C.l Description C-l
C.2 Reference Information C-4
C.3 Detail Tables C-l3
C.3.1 Source Relationships C-13
C.3.1.1 Category 1 - Residential Pesticide Use C-13
C.3.1.2 Category 2 - Household Characteristics C-25
C.3.1.3 Category 3 - Residential Sources (Environmental Measures) C-32
C.3.1.4 Category 4 - Household Occupation C-34
C.3.1.5 Category 5 - Residential Proximity to Agricultural Fields C-46
C.3.1.6 Category 6 - Residential Location C-53
C.3.2 Behavior Relationships C-55
C.3.2.1 Category 7 - Subject's Personal Characteristics C-55
C.3.2.2 Category 8 - Child's Behaviors C-64
C.3.2.3 Category 9 - Dietary Behaviors C-66
C.3.2.4 Category 10 - Family Hygiene Practices C-68
C.3.2.5 Category 11 - Smoking-Related Activities C-74
C.3.2.6 Category 12 - Work Exposure/Practices C-75
C.3.3 Other Relationships C-77
C.3.3.1 Category 13 - Related Exposure Levels C-77
C.3.3.2 Category 14-Health C-78
C-ii August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Tables
Table C. 1 Example of Relationship Detail Table for Question Category: Residential Pesticide Use (Overview Table - Table B. 1
Appendix B) C-l
Table C.2.1 List of Columns and Associated Reference Tables in Detail Tables C-5
Table C.2.2 Chemical/Metabolite Reference Table C-6
Table C.2.3 Medium Reference Table C-8
Table C.2.4 Type of Measurement Reference Table C-8
Table C.2.5 Statistical Analysis Reference Table C-9
Table C.2.6 Log Transformation Reference Table C-10
Table C.2.7 Probability for Model Reference Table C-10
Table C.2.8 Group Statistics Reference Table C-l 1
Table C.2.9 Analysis Statistics Reference Table C-l 1
Table C.2.10. Table Numbers Cross-Referenced between Results Section and Appendices A, B, and C, by Category Group C-l 1
Table C.3.1.1 Relationship Details for Questions in Category 1: Residential Pesticide Use - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis C-l3
Table C.3.1.2 Relationship Details for Questions in Category 2: Household Characteristics - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis C-25
Table C.3.1.3 Relationship Details for Questions in Category 3: Residential Sources (Environmental Measures) - Grouped by Question and
Sorted by Medium, Chemical, Citation and Analysis C-32
Table C.3.1.4 Relationship Details for Questions in Category 4: Household Occupation - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis C-34
Table C.3.1.5 Relationship Details for Questions in Category 5: Residential Proximity to Agricultural Fields - Grouped by Question and Sorted
by Medium, Chemical, Citation and Analysis C-46
Table C.3.1.6 Relationship Details for Questions in Category 6: Residential Location - Grouped by Question and Sorted by Medium, Chemical,
Citation and Analysis C-53
Table C.3.2.1 Relationship Details for Questions in Category 7: Subject's Personal Characteristics - Grouped by Question and Sorted by
Medium, Chemical, Citation and Analysis C-55
C-iii
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table C.3.2.2 Relationship Details for Questions in Category 8: Child's Behaviors - Grouped by Question and Sorted by Medium, Chemical,
Citation and Analysis C-64
Table C.3.2.3 Relationship Details for Questions in Category 9: Dietary Behaviors - Grouped by Question and Sorted by Medium, Chemical,
Citation and Analysis C-66
Table C.3.2.4 Relationship Details for Questions in Category 10: Family Hygiene Practices - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis C-68
Table C.3.2.5 Relationship Details for Questions in Category 11: Smoking-Related Activities - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis C-74
Table C.3.2.6 Relationship Details for Questions in Category 12: Work Exposure/Practices - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis C-75
Table C.3.3.1 Relationship Details for Questions in Category 13: Related Exposure Levels - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis C-77
Table C.3.3.2 Relationship Details for Questions in Category 14: Health - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis C-78
C-iv
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendices B, C, and D provide specific information about the relationships extracted from the literature review and summarized in Results sections
4.2.4, 4.2.5, and 4.2.6. The information is presented as overview, detail, and comment tables. Each appendix includes one type of table for all the
question categories and relationships. This appendix presents the detail tables.
C.l Description
Table C.l is an example of a detail table which provides the detailed statistical analysis and descriptive information for each relationship counted in
the associated overview table (Table B. 1 in Appendix B), and for each relationship in Table D. 1 in Appendix D.
Table C.l Example of Relationship Detail Table for Question Category: Residential Pesticide Use (Overview Table - Table B.1 Appendix B)
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q101-Pesticide Use
484
urine
TCPY
A
Krinsley
1998
SLR
0.77
Y
N
yes vs no
ug'g
Cre
166
0.0010
484
urine
TCPY
A
Krinsley
1998
SLR
0.77
Y
N
yes
ug/g
Cre
8.200
10.540
126
484
urine
TCPY
A
Krinsley
1998
SLR
0.77
Y
N
no
ug/g
Cre
7.490
7.310
40
814
urine
ETHL2
C
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/L
815
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/g
Cre
812
urine
MTHL2
C
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/L
813
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/g
Cre
633
dust
AZM
C
McCauley
2001a
TNR
> 0.05
N
N
Not Available
ppm
848
dust
OPSUM
C
McCauley
2003
WTWS
0.39
N
N
yes vs no
ppm
24
C-l
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q102~lnside Treated
563
urine
MDA
C
Sexton
2003
BSLR-5
0.1
Y
N
yes vs no
ug/L
-0.220
0.1200
557
urine
MDA
c
Sexton
2003
LGRG
0.174
N
N
high score vs
not high
score
ug/L
0.550
(0.23, 1.3)
485
urine
TCPY
A
Krinsley
1998
SLR
0.93
Y
N
yes vs no
ug/g
Cre
166
0.0001
485
urine
TCPY
A
Krinsley
1998
SLR
0.93
Y
N
yes
ug/g
Cre
7.690
8.270
90
485
urine
TCPY
A
Krinsley
1998
SLR
0.93
Y
N
no
ug/g
Cre
7.640
8.120
76
164
urine
ETHL1
C
Lu 2001
MWU
0.27
N
N
yes
umol/L
0.030
0*5
164
urine
ETHL1
C
Lu 2001
MWU
0.27
N
N
no
umol/L
0.040
73
165
urine
MTHL2
C
Lu 2001
MWU
0.35
N
N
yes
umol/L
0.110
0*5
Zo
165
urine
MTHL2
C
Lu 2001
MWU
0.35
N
N
no
umol/L
0.110
73
561
dust
CHLR
L
Sexton
2003
LGRG
0.436
N
N
yes vs no
ng/cm
0.710
(0.30, 1.67)
558
indair
CHLR
C
Sexton
2003
LGRG
0.296
N
N
yes vs no
ng/m3
0.310
(0.04, 2.75)
554
indair
MAL
C
Sexton
2003
LGRG
0.369
N
N
yes vs no
ng/m
0.641
(0.24, 1.69)
559
outd-
air
CHLR
C
Sexton
2003
LGRG
0.715
N
N
yes vs no
ng/m3
0.700
(0.11, 4.66)
555
outd-
air
MAL
C
Sexton
2003
LGRG
0.373
N
N
yes vs no
ng/m
2.760
(0.30. 25.7)
562
pers-
air
CHLR
C
Sexton
2003
BSLR-1
0.04
Y
N
yes vs no
ng/m
-0.820
0.0700
553
pers-
air
MAL
C
Sexton
2003
LGRG
0.073
N
N
yes vs no
ng/m
0.377
(0.13, 1.09)
560
sldfoo
d
CHLR
I
Sexton
2003
LGRG
0.38
N
N
yes vs no
ug/day
1.460
(0.63, 3.37)
C-2
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
556
sldfoo
d
MAL
I
Sexton
2003
LGRG
0.06
N
N
yes vs no
ug/day
0.442
(0.19, 1.04)
Q103~lnside Treated—Bathroom
498
urine
TCPY
A
Krinsley
1998
SLR
0.36
Y
N
yes vs no
ug.'g
Cre
167
0.0050
498
urine
TCPY
A
Krinsley
1998
SLR
0.36
Y
N
yes
ug/g
Cre
8.380
9.000
70
498
urine
TCPY
A
Krinsley
1998
SLR
0.36
Y
N
no
ug/g
Cre
7.150
7.480
97
Table C.l includes all the relationships for the Residential Pesticide Use category counted in Table B.l (Appendix B). The highest level of
organization in this table is based on the Q# for the questions within the category (Appendix E) instead of the medium as in the Appendix B overview
tables. Thus each section of the table describes the relationships for one of the questions appearing in the associated overview table, and each row
describes one aspect of one of the relationships. The rows within a question's section are sorted by medium, chemical/metabolite, citation, and
analysis type. The chemicals/metabolites are presented by the following groupings: (urinary metabolites) non-DAPs (dialkyl phosphates), single
DAPs, DAP sums, detectable DAPs, high DAPs, and chemicals for other mediums (see Table C.2.2).
In the section of Table C.l for "Q103 - Inside treated - bathroom," there are three rows for the urinary metabolite "TCPY," and the citation "Krinsley
1998." Each relationship described in a detail table has at least one row, and each row shows the ID number assigned to the relationship as described
in section 4.2.2.2. Two types of rows are used to describe relationships: analysis level and group level rows. In most cases, a relationship is
described by one type of row or the other. If there is more than one row for a relationship, the columns ID# through PM, and UNITS, have the same
values for all of the relationships.
Several types of relationship situations can be found in the detail tables:
(1) If a relationship is described by only one row, the information included in the row is at the analysis level. In addition to the header
information (columns ID# through PM), the row contains information in columns N through R2 as available in the publication. These relationships
are either for an analysis that is not a group comparison like a regression analysis, or for an analysis comparing groups, where no group level
C-3
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
statistics are provided. In the latter case, a description of the groups compared is included in the Groups Compared column, e.g., "yes vs no" for
Q101 and ID# 814.
(2) When statistical information was available for the groups compared in the analysis, a relationship is described by more than one row. In
addition to the header information, these rows include information in columns GMEAN through N as available, and there is a row for each group
compared in the analysis. Rows with names in the Group Name column are group level rows, e.g., Q102 and ID# 164.
(3) In a few relationships, analysis level and group level information is provided, e.g., Q101 and ID# 484. In this case group level statistics
(mean and standard deviation) were included, but the analysis was a regression analysis, not a test of the means.
C.2 Reference Information
To make the detail tables more compact, it was necessary to use abbreviations or codes in both the column names and contents. Table C.2.1
describes each column name used in the detail tables. The column Reference Table identifies the number of a subsequent table containing
information about the codes used. For example, the column LG includes codes described in Table C.2.6.
C-4
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table C.2.1 List of Columns and Associated Reference Tables in Detail Tables
Column Type or
Name
Description
Applies
toa
Reference Tableb
ID#
Number assigned to each relationship
NA
Medium
Sample medium
Appendix C - Table C.2.3
Chemical
Chemical, metabolite, or molar-weighted sum
a
Appendix C - Table C.2.2
MT
Type of measurement
a
Appendix C - Table C.2.4
Citation
Citation reference
Appendix A - Table A.1
Analysis
Type of statistical analysis performed
a
Appendix C - Table C.2.5
p-value
Probability value associated with statistical analysis
NA
LG
Log transformation
a
Appendix C - Table C.2.6
PM
p-value is associated with model rather than predictor
Appendix C - Table C.2.7
Groups Compared
Predictor groups compared in analysis separated by
"vs"; otherwise assume predictor is continuous
b
NA
Group Name
Name of analyzed group described by group statistics
b
NA
Units
Units for chemical measurement
a
NA
GMean
Group geometric mean
a
Appendix C - Table C.2.8
GSD
Group geometric standard deviation
a
Appendix C - Table C.2.8
Median
Group median
a
Appendix C - Table C.2.8
Mean
Group mean
a
Appendix C - Table C.2.8
StDev
Group standard deviation
a
Appendix C - Table C.2.8
PctD
Group percent of measurements above LOD (limit of
detection)
a
NA
N
Number of participants in group or analysis
a,b
NA
OR
Odds ratio for predictor (logistic regression)
b
Appendix C - Table C.2.9
C-5
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Column Type or
Name
Description
Applies
toa
Reference Tableb
CI
Confidence interval (95% assumed) for OR or Beta
depending on which is included
a
NA
Beta
Regression coefficient for predictor
b
Appendix C - Table C.2.9
R2
R2 from regression analysis of one or more predictors
b
NA
a The entry "a" is a dependent variable, in this case a chemical analytical measurement. The entry "b" is an independent variable or predictor, usually a question.
b NA - Not applicable
Table C.2.2 Chemical/Metabolite Reference Table
Grouping3
Code
Medium
Description
1-Non-DAP
1NAP
urine
1-Naphthol
1-Non-DAP
4NITR
urine
4-Nitrophenol
6-Chemical
ATZ
otherb
Atrazine
1-Non-DAP
ATZM
urine
Atrazine mercapturate
6-Chemical
AZM
other
Azinphosmethyl
6-Chemical
AZMPH
other
Azinphosmethyl+Phosmet
6-Chemical
CHLR
other
Chlorpyrifos
3-DAP Sumc
DAP1
urine
DMP+DMTP+DMDTP+DEP+DETP+DEDTP
4-DAP Detect
DAP2
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one detectable measurement)
5-DAP High
DAP3
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one high measurement)11
2-DAP
DEDTP
urine
Diethyldithiophosphate (DEDTP)
2-DAP
DEP
urine
Diethylphosphate (DEP)
C-6
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Grouping3
Code
Medium
Description
2-DAP
DETP
urine
Diethylthiophosphate (DETP)
2-DAP
DMDTP
urine
Dimethyldithiophosphate (DMDTP)
2-DAP
DMP
urine
Dimethylphosphate (DMP)
2-DAP
DMTP
urine
Dimethylthiophosphate (DMTP)
6-Chemical
EPAR
other
Ethyl parathion
3-DAP Sum
ETHL1
urine
DEP+DETP
3-DAP Sum
ETHL2
urine
DEP+DETP+DEDTP
4-DAP Detect
ETHL3
urine
DEP, DETP, DEDTP
(at least one detectable measurement)
6-Chemical
MAL
other
Malathion
1-Non-DAP
MDA
urine
Malathion dicarboxylic acid
3-DAP Sum
MTHL1
urine
DMTP+DMDTP
3-DAP Sum
MTHL2
urine
DMP+DMTP+DMDTP
4-DAP Detect
MTHL3
urine
DMTP (detectable measurement)
4-DAP Detect
MTHL4
urine
DMP, DMTP
(at least one detectable measurement)
5-DAP High
MTHL5
urine
DMP, DMTP
(at least one high measurement)11
7-M eta bo lite NA
NA
urine
NA (not available or not specified)
6-Chemical
OPSUM
other
OP sum®
6-Chemical
PHSM
other
Phosmet
1-Non-DAP
TCPY
urine
3,5,6-T richloro-2-pyridinol
a The number preceding the group name indicates the order of the group as it appears in the overview tables.
b Medium is urine or other (any other medium measured).
c Sums are molar-weighted unless otherwise specified.
C-7
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
d See definition of high measurement in Azaroff (1999).
e OP Sum = azinphosmethyl, chlorpyrifos, malathion, and phosmet.
Table C.2.3 Medium Reference Table
Code
Description
dust
dust
indair
indoor air
outdair
outdoor air
persair
personal air
sldfood
solid food
soil
soil
urine
urine
Table C.2.4 Type of Measurement Reference Table
Code
Description
A
Adjusted concentration (urine concentration adjusted by creatinine)
C
Concentration
I
Daily intake (food)
L
Loading (dust or dermal)
C-8
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table C.2.5 Statistical Analysis Reference Table
Code
Description
BSLR-#xa
Backwards Stepwise Linear Regression #x
BDPH
Bonferroni/Dunn Post Hoc Test
CHSQ
Chi-Square Test
CORR
Correlation
FISH
Fisher Exact Test
FSLR
Forward Selection Linear Regression
GLM
General Linear Model ANOVA
GLM-#x
General Linear Model ANOVA #x
KWAN
Kruskal-Wallis One-Way ANOVA
LGRG
Logistic Regression
MWU
Mann-Whitney U Test
MLR
Multiple Linear Regression
MLR-#xa
Multiple Linear Regression #x
MLGR-#xa
Multiple Logistic Regression #x
MVRG-#xa
Multivariate Regression #x
NAN
Not analyzed
OWAN
One-Way ANOVA
SLR
Simple Linear Regression
SLGR
Simple Logistic Regression
SPCR
Spearman Rank Correlation
TTST
t-test
TWAN-#xa
Two-Way ANOVA #x
C-9
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Code
Description
TNR
Type of Analysis Not Reported
WTAN
Weighted ANOVA
WSRK
Wilcoxon Signed Rank Test
WTWS
Wilcoxon Two-Sample Test
a In some analyses where more than one predictor was analyzed in a relationship, the predictor questions will likely appear in different question category sections. The user can
identify the predictors that were analyzed in the same relationship by looking for the same analysis code for the citation. For example, if a multiple linear regression was performed
with three predictors on two metabolites, there would be two analysis types: < MLR-#1 and MLR-#2. The analysis type MLR-#1 would be used as the analysis type for the three
relationships describing the three predictor questions. Aprea 2000 contains examples of this type of analysis code.
Table C.2.6 Log Transformation Reference Table
Code
Description
Y
Measurements were log-transformed before analysis.
N
Original measurement values were used in the analysis.
Table C.2.7 Probability for Model Reference Table
Code
Description
Y
p-value applies to a model which includes more than one predictor (e.g., regression analysis F-test).
N
p-value applies to one predictor in a single-predictor or multi-predictor analysis (e.g., coefficient in regression analysis).
C-10
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table C.2.8 Group Statistics Reference Table
Code3
Description
>
Group has statistic with higher value (e.g., Gmean, Median).
<
Group has statistic with lower value (e.g., Gmean, Median).
a Codes appear when publication provides only relative indicators for group statistics.
Table C.2.9 Analysis Statistics Reference Table
Code3
Description
T
Statistic (e.g. regression coefficient) > 0, that is, there is a positive association between the measurement and predictor.
1
Statistic (e.g. regression coefficient) < 0, that is, there is an inverse association between the measurement and predictor.
a Codes appear when publication provides only the direction of relationship.
Table C.2.10 provides a cross-reference between the relationship summary tables in the Results section and the tables in Appendices B, C, D.
Table C.2.10. Table Numbers Cross-Referenced between Results Section and Appendices A, B, and C, by Category Group
Category
Section #
Table #3
Overview
Table #
Detailed
Table #
Comment
Table #
Group
#
Description
Results
Results
Appendix B
Appendix C
Appendix D
Source
1
Residential pesticide use
4.2.4.1
4.2.6.x
B.3.1.1
C.3.1.1
D.3.1.1
Source
2
Household characteristics
4.2.4.2
4.2.7.x
B.3.1.2
C.3.1.2
D.3.1.2
C-ll
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Category
Section #
Table #a
Overview
Table #
Detailed
Table #
Comment
Table #
Group
#
Description
Results
Results
Appendix B
Appendix C
Appendix D
Source
3
Residential sources
(environmental measures)
4.2.4.3
4.2.8.x
B.3.1.3
C.3.1.3
D.3.1.3
Source
4
Household occupation
4.2.4.4
4.2.9.x
B.3.1.4
C.3.1.4
D.3.1.4
Source
5
Residential proximity to
agricultural fields
4.2.4.5
4.2.10.x
B.3.1.5
C.3.1.5
D.3.1.5
Source
6
Residential location
4.2.4.6
4.2.11.x
B.3.1.6
C.3.1.6
D.3.1.6
Behavior
7
Subject's personal
characteristics
4.2.5.1
4.2.13.x
B.3.2.1
C.3.2.1
D.3.2.1
Behavior
8
Child's behaviors
4.2.5.2
4.2.14.x
B.3.2.2
C.3.2.2
D.3.2.2
Behavior
9
Dietary behaviors
4.2.5.3
4.2.15.x
B.3.2.3
C.3.2.3
D.3.2.3
Behavior
10
Family hygiene practices
4.2.5.4
4.2.16.x
B.3.2.4
C.3.2.4
D.3.2.4
Behavior
11
Smoking-related activities
4.2.5.5
4.2.17.x
B.3.2.5
C.3.2.5
D.3.2.5
Behavior
12
Work exposure/practices
4.2.5.6
4.2.18.x
B.3.2.6
C.3.2.6
D.3.2.6
Other
13
Related exposure levels
4.2.6.1
4.2.20.x
B.3.3.1
C.3.3.1
D.3.3.1
Other
14
Health
4.2.6.2
4.2.21.x
B.3.3.2
C.3.3.2
D.3.3.2
a x in this column refers to the three table types, a, b, and c, described above.
C-12
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3 Detail Tables
C.3.1 Source Relationships
C.3.1.1 Category 1 - Residential Pesticide Use
Table C.3.1.1 Relationship Details for Questions in Category 1: Residential Pesticide Use - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q101-Pesticide Use
484
urine
TCPY
A
Krinsley
1998
SLR
0.77
Y
N
yes vs no
ug/g
Cre
166
0.0010
484
urine
TCPY
A
Krinsley
1998
SLR
0.77
Y
N
yes
ug/g
Cre
8.200
10.540
126
484
urine
TCPY
A
Krinsley
1998
SLR
0.77
Y
N
no
ug/g
Cre
7.490
7.310
40
814
urine
ETHL2
C
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/L
815
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/g
Cre
812
urine
MTHL2
C
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/L
813
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/g
Cre
633
dust
AZM
C
McCauley
2001a
TNR
> 0.05
N
N
Not Available
ppm
848
dust
OPSUM
C
McCauley
2003
WTWS
0.39
N
N
yes vs no
ppm
24
Q102~lnside Treated
563
urine
MDA
C
Sexton
2003
BSLR-5
0.1
Y
N
yes vs no
ug/L
-0.220
0.1200
C-13
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
557
urine
MDA
C
Sexton
2003
LGRG
0.174
N
N
high score vs
not high
score
ug/L
0.550
(0.23, 1.3)
485
urine
TCPY
A
Krinsley
1998
SLR
0.93
Y
N
yes vs no
ug/g
Cre
166
0.0001
485
urine
TCPY
A
Krinsley
1998
SLR
0.93
Y
N
yes
ug/g
Cre
7.690
8.270
90
485
urine
TCPY
A
Krinsley
1998
SLR
0.93
Y
N
no
ug/g
Cre
7.640
8.120
76
164
urine
ETHL1
C
Lu 2001
MWU
0.27
N
N
yes
umol/L
0.030
23
164
urine
ETHL1
C
Lu 2001
MWU
0.27
N
N
no
umol/L
0.040
73
165
urine
MTHL2
C
Lu 2001
MWU
0.35
N
N
yes
umol/L
0.110
23
165
urine
MTHL2
C
Lu 2001
MWU
0.35
N
N
no
umol/L
0.110
73
561
dust
CHLR
L
Sexton
2003
LGRG
0.436
N
N
yes vs no
ng/cm2
0.710
(0.30, 1.67)
558
indair
CHLR
C
Sexton
2003
LGRG
0.296
N
N
yes vs no
ng/m3
0.310
(0.04, 2.75)
554
indair
MAL
C
Sexton
2003
LGRG
0.369
N
N
yes vs no
ng/m3
0.641
(0.24, 1.69)
559
outd-
air
CHLR
C
Sexton
2003
LGRG
0.715
N
N
yes vs no
ng/m3
0.700
(0.11, 4.66)
555
outd-
air
MAL
C
Sexton
2003
LGRG
0.373
N
N
yes vs no
ng/m3
2.760
(0.30, 25.7)
562
pers-
air
CHLR
C
Sexton
2003
BSLR-1
0.04
Y
N
yes vs no
ng/m3
-0.820
0.0700
553
pers-
air
MAL
C
Sexton
2003
LGRG
0.073
N
N
yes vs no
ng/m3
0.377
(0.13, 1.09)
560
sldfoo
d
CHLR
I
Sexton
2003
LGRG
0.38
N
N
yes vs no
ug/day
1.460
(0.63, 3.37)
556
sldfoo
d
MAL
I
Sexton
2003
LGRG
0.06
N
N
yes vs no
ug/day
0.442
(0.19, 1.04)
C-14
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q103~lnside Treated-Bathroom
498
urine
TCPY
A
Krinsley
1998
SLR
0.36
Y
N
yes vs no
ug/g
Cre
167
0.0050
498
urine
TCPY
A
Krinsley
1998
SLR
0.36
Y
N
yes
ug/g
Cre
8.380
9.000
70
498
urine
TCPY
A
Krinsley
1998
SLR
0.36
Y
N
no
ug/g
Cre
7.150
7.480
97
Q104~lnside Treated-Bedroom
767
urine
TCPY
A
Krinsley
1998
FSLR#2
< 0.001
Y
Y
yes vs no
ug/g
Cre
166
0.125
0.2100
772
urine
TCPY
A
Krinsley
1998
FSLR#3
< 0.001
Y
Y
yes vs no
ug/g
Cre
0.246
0.3500
497
urine
TCPY
A
Krinsley
1998
SLR
0.02
Y
N
yes vs no
ug/g
Cre
167
0.0300
497
urine
TCPY
A
Krinsley
1998
SLR
0.02
Y
N
yes
ug/g
Cre
9.080
9.180
65
497
urine
TCPY
A
Krinsley
1998
SLR
0.02
Y
N
no
ug/g
Cre
6.760
7.320
102
Q105~lnside Treated-Cabinets
506
urine
TCPY
A
Krinsley
1998
SLR
0.15
Y
N
yes vs no
ug/g
Cre
167
0.0100
506
urine
TCPY
A
Krinsley
1998
SLR
0.15
Y
N
yes
ug/g
Cre
10.560
12.210
21
506
urine
TCPY
A
Krinsley
1998
SLR
0.15
Y
N
no
ug/g
Cre
7.250
7.350
146
Q106~lnside Treated-Closets
507
urine
TCPY
A
Krinsley
1998
SLR
0.04
Y
N
yes vs no
ug/g
Cre
167
0.0300
507
urine
TCPY
A
Krinsley
1998
SLR
0.04
Y
N
yes
ug/g
Cre
11.710
13.000
26
507
urine
TCPY
A
Krinsley
1998
SLR
0.04
Y
N
no
ug/g
Cre
6.920
6.700
141
C-15
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q107~lnside Treated-Cupboards with Dishes
505
urine
TCPY
A
Krinsley
1998
SLR
0.52
Y
N
yes vs no
ug/g
Cre
167
0.0100
505
urine
TCPY
A
Krinsley
1998
SLR
0.52
Y
N
yes
ug/g
Cre
9.210
11.800
11
505
urine
TCPY
A
Krinsley
1998
SLR
0.52
Y
N
no
ug/g
Cre
7.560
7.870
156
Q108~lnside Treated-Dining Room
496
urine
TCPY
A
Krinsley
1998
SLR
0.12
Y
N
yes vs no
ug/g
Cre
167
0.0200
496
urine
TCPY
A
Krinsley
1998
SLR
0.12
Y
N
yes
ug/g
Cre
8.500
8.370
57
496
urine
TCPY
A
Krinsley
1998
SLR
0.12
Y
N
no
ug/g
Cre
7.230
8.040
110
Q109~lnside Treated-Family Room
494
urine
TCPY
A
Krinsley
1998
SLR
0.38
Y
N
yes vs no
ug/g
Cre
167
0.0050
494
urine
TCPY
A
Krinsley
1998
SLR
0.38
Y
N
yes
ug/g
Cre
7.780
7.510
494
urine
TCPY
A
Krinsley
1998
SLR
0.38
Y
N
no
ug/g
Cre
7.630
8.370
Q110~lnside Treated-Kitchen
493
urine
TCPY
A
Krinsley
1998
SLR
0.89
Y
N
yes vs no
ug/g
Cre
167
0.0010
493
urine
TCPY
A
Krinsley
1998
SLR
0.89
Y
N
yes
ug/g
Cre
7.820
8.580
80
493
urine
TCPY
A
Krinsley
1998
SLR
0.89
Y
N
no
ug/g
Cre
7.520
7.780
87
Q111-Inside Treated-Living Room
495
urine
TCPY
A
Krinsley
1998
SLR
0.08
Y
N
yes vs no
ug/g
Cre
167
0.0200
495
urine
TCPY
A
Krinsley
1998
SLR
0.08
Y
N
yes
ug/g
Cre
8.700
9.140
66
C-16
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
495
urine
TCPY
A
Krinsley
1998
SLR
0.08
Y
N
no
ug/g
Cre
6.990
7.400
101
Q112~lnside Treated-On Baseboards
501
urine
TCPY
A
Krinsley
1998
SLR
0.51
Y
N
yes vs no
ug/g
Cre
167
0.0030
501
urine
TCPY
A
Krinsley
1998
SLR
0.51
Y
N
yes
ug/g
Cre
8.300
9.060
69
501
urine
TCPY
A
Krinsley
1998
SLR
0.51
Y
N
no
ug/g
Cre
7.220
7.450
98
Q113~lnside Treated-On Ceiling
504
urine
TCPY
A
Krinsley
1998
SLR
0.58
Y
N
yes vs no
ug/g
Cre
167
0.0020
504
urine
TCPY
A
Krinsley
1998
SLR
0.58
Y
N
yes
ug/g
Cre
9.030
8.090
8
504
urine
TCPY
A
Krinsley
1998
SLR
0.58
Y
N
no
ug/g
Cre
7.600
8.170
159
Q114~lnside Treated-On Floor
500
urine
TCPY
A
Krinsley
1998
SLR
0.27
Y
N
yes vs no
ug/g
Cre
167
0.0070
500
urine
TCPY
A
Krinsley
1998
SLR
0.27
Y
N
yes
ug/g
Cre
6.850
7.700
38
500
urine
TCPY
A
Krinsley
1998
SLR
0.27
Y
N
no
ug/g
Cre
7.900
8.290
129
Q115~lnside Treated-On Lower Walls
502
urine
TCPY
A
Krinsley
1998
SLR
0.65
Y
N
yes vs no
ug/g
Cre
167
0.0010
502
urine
TCPY
A
Krinsley
1998
SLR
0.65
Y
N
yes
ug/g
Cre
9.710
11.230
16
502
urine
TCPY
A
Krinsley
1998
SLR
0.65
Y
N
no
ug/g
Cre
7.430
7.770
151
Q116~lnside Treated-On Upper Walls
503
urine
TCPY
A
Krinsley
1998
SLR
0.2
Y
N
yes vs no
ug/g
Cre
167
0.0100
C-17
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
503
urine
TCPY
A
Krinsley
1998
SLR
0.2
Y
N
yes
ug/g
Cre
13.100
14.820
8
503
urine
TCPY
A
Krinsley
1998
SLR
0.2
Y
N
no
ug/g
Cre
7.390
7.650
159
Q117~lnside Treated-Other Room
773
urine
TCPY
A
Krinsley
1998
FSLR#3
< 0.001
Y
Y
yes vs no
ug/g
Cre
0.172
0.3500
499
urine
TCPY
A
Krinsley
1998
SLR
0.05
Y
N
yes vs no
ug/g
Cre
167
0.0200
499
urine
TCPY
A
Krinsley
1998
SLR
0.05
Y
N
yes
ug/g
Cre
12.630
14.190
21
499
urine
TCPY
A
Krinsley
1998
SLR
0.05
Y
N
no
ug/g
Cre
6.950
6.640
146
Q118—Pets Treated
160
urine
ETHL1
C
Lu 2001
MWU
0.14
N
N
yes
umol/L
0.040
18
160
urine
ETHL1
C
Lu 2001
MWU
0.14
N
N
no
umol/L
0.030
18
335
urine
MTHL1
C
Lu 2000
MWU
0.6
N
N
yes
ug/ml
0.070
335
urine
MTHL1
C
Lu 2000
MWU
0.6
N
N
no
ug/ml
0.050
161
urine
MTHL2
C
Lu 2001
MWU
0.8
N
N
yes
umol/L
0.150
18
161
urine
MTHL2
C
Lu 2001
MWU
0.8
N
N
no
umol/L
0.180
18
331
dust
AZMPH
C
Lu 2000
MWU
0.1
N
N
yes
ug/g
0.700
331
dust
AZMPH
C
Lu 2000
MWU
0.1
N
N
no
ug/g
2.100
Q119~Outside Treated
567
urine
MDA
C
Sexton
2003
BSLR-5
0.03
Y
N
yes vs no
ug/L
-0.330
0.1200
489
urine
TCPY
A
Krinsley
1998
SLR
0.11
Y
N
yes vs no
ug/g
Cre
166
0.0200
489
urine
TCPY
A
Krinsley
1998
SLR
0.11
Y
N
yes
ug/g
Cre
7.960
7.690
107
489
urine
TCPY
A
Krinsley
1998
SLR
0.11
Y
N
no
ug/g
Cre
7.130
9.020
59
C-18
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
570
urine
TCPY
C
Sexton
2003
BSLR-6
0.09
Y
N
yes vs no
ug/L
-0.290
0.0800
566
dust
CHLR
L
Sexton
2003
BSLR-4
0.01
Y
N
yes vs no
ng/cm2
-0.028
0.0800
565
sld-
food
CHLR
I
Sexton
2003
BSLR-3
0.06
Y
N
yes vs no
ug/day
-0.800
0.0400
Q120-Garden Treated
249
urine
4NITR
C
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/L
248
urine
TCPY
C
Fenske
2002
MWU
0.02
N
N
yes
ug/L
8.300
248
urine
TCPY
C
Fenske
2002
MWU
0.02
N
N
no
ug/L
2.400
156
urine
ETHL1
C
Lu 2001
MWU
0.02
N
N
yes
umol/L
0.040
27
156
urine
ETHL1
C
Lu 2001
MWU
0.02
N
N
no
umol/L
0.030
22
338
urine
MTHL1
C
Lu 2000
MWU
0.9
N
N
yes
ug/ml
0.080
338
urine
MTHL1
C
Lu 2000
MWU
0.9
N
N
no
ug/ml
0.050
157
urine
MTHL2
C
Lu 2001
MWU
0.05
N
N
yes
umol/L
0.190
27
157
urine
MTHL2
C
Lu 2001
MWU
0.05
N
N
no
umol/L
0.090
22
334
dust
AZMPH
C
Lu 2000
MWU
0.8
N
N
yes
ug/g
2.100
334
dust
AZMPH
C
Lu 2000
MWU
0.8
N
N
no
ug/g
1.900
Q121 -Lawn/Yard Treated
571
urine
TCPY
c
Sexton
2003
BSLR-6
0.09
Y
N
yes vs no
ug/L
-0.260
0.0800
162
urine
ETHL1
c
Lu 2001
MWU
0.68
N
N
yes
umol/L
0.040
45
162
urine
ETHL1
c
Lu 2001
MWU
0.68
N
N
no
umol/L
0.040
48
337
urine
MTHL1
c
Lu 2000
MWU
0.7
N
N
yes
ug/ml
0.060
337
urine
MTHL1
c
Lu 2000
MWU
0.7
N
N
no
ug/ml
0.050
163
urine
MTHL2
c
Lu 2001
MWU
0.13
N
N
yes
umol/L
0.140
45
163
urine
MTHL2
c
Lu 2001
MWU
0.13
N
N
no
umol/L
0.090
48
C-19
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
333
dust
AZMPH
C
Lu 2000
MWU
0.7
N
N
yes
ug/g
2.600
333
dust
AZMPH
C
Lu 2000
MWU
0.7
N
N
no
ug/g
1.800
Q122~lnside or Outside Treated
132
urine
1NAP
c
Adgate
2001
WTAN
> 0.05
Y
N
yes vs no
ug/L
133
urine
MDA
c
Adgate
2001
WTAN
> 0.05
Y
N
yes vs no
ug/L
134
urine
TCPY
c
Adgate
2001
WTAN
> 0.05
Y
N
yes vs no
ug/L
71
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
36.400
2.400
79
166
71
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
32.700
2.400
75
29
62
urine
DEP
A
Aprea 2000
MLR-5
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0490
72
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
24.300
2.700
62
166
72
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
14.900
2.900
46
29
73
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
9.900
2.400
21
166
73
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
7.400
2.100
10
29
64
urine
DEDTP
A
Aprea 2000
MLR-7
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0620
70
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
136.900
3.000
96
166
70
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
113.500
2.500
96
29
61
urine
DMP
A
Aprea 2000
MLR-1
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0440
67
urine
DM DTP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
19.700
3.800
41
166
C-20
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
67
urine
DM DTP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
13.300
2.800
33
29
74
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
81.500
2.200
90
166
74
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
62.200
2.200
82
29
65
urine
ETHL2
A
Aprea 2000
MLR-8
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0560
Q123~Previous Treatment
336
urine
MTHL1
C
Lu 2000
MWU
0.6
N
N
yes
ug/ml
0.050
336
urine
MTHL1
C
Lu 2000
MWU
0.6
N
N
no
ug/ml
0.050
332
dust
AZMPH
C
Lu 2000
MWU
0.3
N
N
no
ug/g
2.100
332
dust
AZMPH
C
Lu 2000
MWU
0.3
N
N
yes
ug/g
1.100
Q124-Level of Pesticide Use
136
urine
1NAP
C
Adgate
2001
WTAN
> 0.05
Y
N
high score vs
not high
score
ug/L
137
urine
MDA
C
Adgate
2001
WTAN
> 0.05
Y
N
yes vs no
ug/L
551
urine
MDA
C
Sexton
2003
SLR
0.033
Y
N
high score vs
not high
score
ug/L
549
urine
MDA
C
Sexton
2003
WTWS
0.04
N
N
high score vs
not high
score
ug/L
(1.06, 5.8)
138
urine
TCPY
C
Adgate
2001
WTAN
> 0.05
Y
N
high score vs
not high
score
ug/L
762
urine
TCPY
A
Krinsley
1998
FSLR#1
< 0.001
Y
Y
PUI index
scores
ug/g
Cre
0.003
0.1800
476
urine
TCPY
A
Krinsley
1998
SLR
< 0.004
Y
N
PUI index
scores
ug/g
Cre
0.0500
C-21
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
550
pers-
air
ATZ
C
Sexton
2003
LGRG
0.028
N
N
high score vs
not high
score
ng/m3
1.220
(1.05, 1.45)
552
pers-
air
ATZ
C
Sexton
2003
LGRG
0.02
Y
N
high score vs
not high
score
ng/m3
Q125~Frequency Personal Application Inside
486
urine
TCPY
A
Krinsley
1998
SLR
0.07
Y
N
number of
times
ug/g
Cre
90
0.0360
486
urine
TCPY
A
Krinsley
1998
SLR
0.07
Y
N
yes
ug/g
Cre
9.060
11.070
42
486
urine
TCPY
A
Krinsley
1998
SLR
0.07
Y
N
no
ug/g
Cre
6.490
4.400
48
Q126~Frequency Personal Application Outside
766
urine
TCPY
A
Krinsley
1998
FSLR#2
< 0.001
Y
Y
number of
times
ug/g
Cre
166
0.270
0.2100
771
urine
TCPY
A
Krinsley
1998
FSLR#3
< 0.001
Y
Y
number of
times
ug/g
Cre
0.020
0.3500
490
urine
TCPY
A
Krinsley
1998
SLR
0.003
Y
N
number of
times
ug/g
Cre
107
0.0800
490
urine
TCPY
A
Krinsley
1998
SLR
0.003
Y
N
yes
ug/g
Cre
6.860
4.180
76
490
urine
TCPY
A
Krinsley
1998
SLR
0.003
Y
N
no
ug/g
Cre
10.660
12.440
31
Q127~lnside/Outside Treated by Family Member
275
urine
ETHL3
C
Azaroff
1999
MLGR-6
< 0.05
N
N
yes vs no
ug/L
273
3.000
(1 -2, 8.2)
591
urine
MTHL3
C
Azaroff
1999
MLGR-7
< 0.01
N
N
yes vs no
ug/L
274
12.000
(3.8, 4.1)
276
urine
MTHL4
C
Azaroff
1999
MLGR-3
< 0.01
N
N
yes vs no
ug/L
274
6.700
(2.0, 2.6)
273
urine
DAP2
C
Azaroff
1999
MLGR-1
< 0.05
N
N
yes vs no
ug/L
274
1.800
(1 -0, 3.0)
C-22
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
274
urine
DAP3
C
Azaroff
1999
MLGR-2
< 0.10
N
N
yes vs no
ug/L
274
2.000
(1.0, 4.6)
Q128~Frequency Professional Application Inside
487
urine
TCPY
A
Krinsley
1998
SLR
0.62
Y
N
yes vs no
ug/g
Cre
88
0.0030
Q129~Frequency Professional Application Outside
491
urine
TCPY
A
Krinsley
1998
SLR
0.96
Y
N
number of
times
ug/g
Cre
107
0.0003
491
urine
TCPY
A
Krinsley
1998
SLR
0.96
Y
N
yes
ug/g
Cre
6.860
4.180
76
491
urine
TCPY
A
Krinsley
1998
SLR
0.96
Y
N
no
ug/g
Cre
10.660
12.440
31
Q130-Personally Mixed Pesticide Inside
488
urine
TCPY
A
Krinsley
1998
SLR
0.18
Y
N
yes vs no
ug/g
Cre
23
0.0840
488
urine
TCPY
A
Krinsley
1998
SLR
0.18
Y
N
yes
ug/g
Cre
14.900
20.090
4
488
urine
TCPY
A
Krinsley
1998
SLR
0.18
Y
N
no
ug/g
Cre
5.600
3.960
19
Q131-Personally Mixed Pesticide Outside
492
urine
TCPY
A
Krinsley
1998
SLR
0.46
Y
N
yes vs no
ug/g
Cre
30
0.0190
492
urine
TCPY
A
Krinsley
1998
SLR
0.46
Y
N
yes
ug/g
Cre
9.430
12.160
11
492
urine
TCPY
A
Krinsley
1998
SLR
0.46
Y
N
no
ug/g
Cre
6.630
4.960
19
Q132~Presence During Mixing
680
urine
1NAP
C
Adgate
2001
WTAN
> 0.05
Y
N
yes vs no
ug/L
683
urine
ATZM
C
Adgate
2001
NAN
> 0.05
Y
N
yes vs no
ug/L
681
urine
MDA
C
Adgate
2001
WTAN
> 0.05
Y
N
yes vs no
ug/L
C-23
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
682
urine
TCPY
C
Adgate
2001
WTAN
> 0.05
Y
N
yes vs no
ug/L
a See section 4.2.2.2 and the paragraph immediately following Table 4.2.3 regarding relationships from Sexton (2003).
C-24
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.1.2 Category 2 - Household Characteristics
Table C.3.1.2 Relationship Details for Questions in Category 2: Household Characteristics - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q201~Housing Type
168
urine
ETHL1
C
Lu 2001
TNR
> 0.05
N
N
single-family
home vs
multiunit
building
umol/L
169
urine
MTHL2
C
Lu 2001
TNR
> 0.05
N
N
single-family
home vs
multiunit
building
umol/L
632
dust
AZM
c
McCauley
2001a
TNR
> 0.05
N
N
Not Available
ppm
Q202-Property Used as a Farm
690
dust
CHLR
L
Sexton
2003
BSLR-4
0.06
Y
N
yes vs no
ng/cm2
-0.660
0.0800
564
indair
CHLR
C
Sexton
2003
BSLR-2
0.01
Y
N
yes vs no
ng/m3
-1.410
0.1200
Q203~Age of House >10 Years
483
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
Y
N
yes vs no
ug/g
Cre
Q204~Age of House >20 Years
670
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
Y
N
yes vs no
ug/g
Cre
Q205~Having Air Conditioning
480
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
Y
Y
yes vs no
ug/g
Cre
Q206~Having Central Heating
482
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
Y
N
yes vs no
ug/g
Cre
C-25
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q207~Having Evaporative Cooling
481
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
Y
N
yes vs no
ug/g
Cre
Q208~Pets in House
53
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
25.100
2.400
53
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
36.200
2.400
44
urine
DEP
A
Aprea 2000
MLR-5
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0490
54
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
15.800
2.800
54
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
16.100
2.900
45
urine
DETP
A
Aprea 2000
MLR-6
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0550
55
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
6.600
1.900
55
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
8.100
2.200
46
urine
DEDTP
A
Aprea 2000
MLR-7
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0620
52
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
99.700
2.400
52
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
122.600
2.600
43
urine
DMP
A
Aprea 2000
MLR-1
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0440
48
urine
DMTP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
98.900
2.700
48
urine
DMTP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
106.000
2.800
49
urine
DM DTP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
11.900
3.100
C-26
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
49
urine
DM DTP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
14.900
2.600
158
urine
ETHL1
C
Lu 2001
MWU
0.4
N
N
yes
umol/L
0.040
40
158
urine
ETHL1
C
Lu 2001
MWU
0.4
N
N
no
umol/L
0.040
56
56
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
54.400
2.100
56
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
68.400
2.200
47
urine
ETHL2
A
Aprea 2000
MLR-8
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0560
50
urine
MTHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
238.800
2.100
50
urine
MTHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
276.600
2.300
159
urine
MTHL2
C
Lu 2001
MWU
0.04
N
N
yes
umol/L
0.160
40
159
urine
MTHL2
C
Lu 2001
MWU
0.04
N
N
no
umol/L
0.090
56
51
urine
DAP1
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
301.900
2.000
51
urine
DAP1
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
359.300
2.200
Q209~Pets Inside/Outside House
569
urine
MDA
C
Sexton
2003
BSLR-5
0.08
Y
N
yes vs no
ug/L
-0.260
0.1200
Q210~Pet Inside to Outside
736
dust
AZM
L
Simcox
1995
MLR-3
> 0.05
N
N
yes vs no
ug/m2
436
dust
AZM
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
438
dust
CHLR
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
738
dust
CHLR
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
C-27
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
439
dust
EPAR
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
739
dust
EPAR
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
737
dust
PHSM
L
Simcox
1995
MLR-4
> 0.05
N
N
yes vs no
ug/m2
437
dust
PHSM
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
Q211-Existence of Garden or Vegetable Garden
568
urine
MDA
C
Sexton
2003
BSLR-5
0.04
Y
N
yes vs no
ug/L
0.310
0.1200
17
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
33.300
2.500
17
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
33.000
2.300
8
urine
DEP
A
Aprea 2000
MLR-5
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0490
9
urine
DETP
A
Aprea 2000
MLR-6
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0550
18
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
15.700
2.700
18
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
16.300
3.000
19
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
8.100
2.200
19
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
7.300
2.000
10
urine
DEDTP
A
Aprea 2000
MLR-7
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0620
16
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
115.500
2.300
16
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
118.500
2.700
C-28
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
7
urine
DMP
A
Aprea 2000
MLR-1
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0440
12
urine
DMTP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
114.200
2.800
12
urine
DMTP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
91.500
2.800
154
urine
ETHL1
C
Lu 2001
MWU
0.04
N
N
yes
umol/L
0.040
49
154
urine
ETHL1
C
Lu 2001
MWU
0.04
N
N
no
umol/L
0.030
46
20
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
66.600
2.300
20
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
62.200
2.100
11
urine
ETHL2
A
Aprea 2000
MLR-8
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0560
14
urine
MTHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
247.800
2.100
14
urine
MTHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
281.500
2.400
155
urine
MTHL2
C
Lu 2001
MWU
0.11
N
N
yes
umol/L
0.140
49
155
urine
MTHL2
C
Lu 2001
MWU
0.11
N
N
no
umol/L
0.080
46
15
urine
DAP1
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
362.100
2.200
15
urine
DAP1
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
321.400
2.000
Q212~Ornamental Plants or Cut Flowers
35
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
32.800
2.400
35
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
35.700
2.300
26
urine
DEP
A
Aprea 2000
MLR-5
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0490
36
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
16.300
2.900
C-29
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
36
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
14.500
2.700
27
urine
DETP
A
Aprea 2000
MLR-6
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0550
37
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
7.900
2.200
37
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
6.800
1.800
28
urine
DEDTP
A
Aprea 2000
MLR-7
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0620
34
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
121.100
2.600
34
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
90.600
2.400
25
urine
DMP
A
Aprea 2000
MLR-1
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0440
30
urine
DMTP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
103.500
2.800
30
urine
DMTP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
109.700
2.700
31
urine
DM DTP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
14.000
3.000
31
urine
DM DTP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
14.900
2.900
38
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
64.600
2.300
38
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
66.100
1.900
29
urine
ETHL2
A
Aprea 2000
MLR-8
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0560
32
urine
MTHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
270.000
2.300
32
urine
MTHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
248.700
2.100
C-30
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation3
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
33
urine
DAP1
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
327.900
1.900
33
urine
DAP1
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
347.400
2.200
Q213~Size of Household
631
dust
AZM
C
McCauley
2001a
SLR
> 0.05
Y
N
ppm
22
t
847
dust
OPSUM
C
McCauley
2003
CORR
0.29
Y
N
number of
individuals
ppm
24
0.0480
Q214~Location of Play Area
832
dust
OPSUM
C
McCauley
2003
WTWS
0.66
N
N
common
ppm
1.190
832
dust
OPSUM
C
McCauley
2003
WTWS
0.66
N
N
isolated
ppm
0.840
Q215~Age of House (Years)
843
dust
AZM
C
McCauley
2003
CORR
0.22
Y
N
years
ppm
24
0.0676
842
dust
OPSUM
C
McCauley
2003
CORR
0.25
Y
N
years
ppm
24
0.0625
Q216-Size of Home (sq ft)
844
dust
OPSUM
C
McCauley
2003
CORR
0.08
Y
N
size in sq ft
ppm
24
0.0160
845
dust
OPSUM
C
McCauley
2003
MLR
0.16
Y
N
size in sq ft
ppm
24
Q217~Number of Pets in House
849
dust
OPSUM
C
McCauley
2003
CORR
0.95
Y
N
number of
animals
ppm
24
0.0004
a See section 4.2.2.2 and the paragraph immediately following Table 4.2.3 regarding relationships from Sexton (2003).
C-31
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.1.3 Category 3 - Residential Sources (Environmental Measures)
Table C.3.1.3 Relationship Details for Questions in Category 3: Residential Sources (Environmental Measures) - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q301 -Household Dust
147
urine
MTHL2
C
Curl 2002
SLR
< 0.001
Y
N
measure-
ment
umol/L
0.1400
586
urine
MTHL2
A
Curl 2002
SLR
< 0.001
Y
N
measure-
ment
umol/g
Cre
0.1500
328
urine
NA
C
Lu 2000
SPCR
< 0.10
N
N
measure-
ment
ug/ml
329
urine
NA
C
Lu 2000
SPCR
0.09
N
N
measure-
ment
ug/ml
0.0120
Q302~Loading from Household Floor Dust
644
urine
DAP1
A
Shalat
2003
MVRG-2
> 0.05
N
Y
measure-
ment
nmol/
mo I
Cre
41
0.2600
Q303~Outdoor Soil
403
dust
AZM
C
Simcox
1995
SPCR
0.001
N
N
indoor HH
dust vs
outdoor surf
soil
ng/g
48
407
dust
AZM
C
Simcox
1995
SPCR
0.87
N
N
indoor HH
dust vs
outdoor surf
soil
ng/g
11
405
dust
CHLR
C
Simcox
1995
SPCR
< 0.001
N
N
indoor HH
dust vs
outdoor surf
soil
ng/g
48
409
dust
CHLR
C
Simcox
1995
SPCR
0.21
N
N
indoor HH
dust vs
outdoor surf
soil
ng/g
11
C-32
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
406
dust
EPAR
C
Simcox
1995
SPCR
0.02
N
N
indoor HH
dust vs
outdoor surf
soil
ng/g
48
410
dust
EPAR
C
Simcox
1995
SPCR
0.01
N
N
indoor HH
dust vs
outdoor surf
soil
ng/g
11
404
dust
PHSM
c
Simcox
1995
SPCR
< 0.001
N
N
indoor HH
dust vs
outdoor surf
soil
ng/g
48
408
dust
PHSM
c
Simcox
1995
SPCR
0.48
N
N
indoor HH
dust vs
outdoor surf
soil
ng/g
11
C-33
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.1.4 Category 4 - Household Occupation
Table C.3.1.4 Relationship Details for Questions in Category 4: Household Occupation - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q401-Agricultural Workers In Household
572
dust
AZM
C
McCauley
2001a
SLR
0.002
Y
N
number of
workers
ppm
25
t
634
dust
AZM
C
McCauley
2001a
SLR
< 0.05
Y
N
number of
workers
ppm
25
Q402~Household Member Spraying Fields
282
urine
ETHL3
c
Azaroff
1999
MLGR-5
< 0.05
N
N
yes vs no
ug/L
273
2.500
(1 -2, 5.3)
283
urine
ETHL3
c
Azaroff
1999
MLGR-6
< 0.10
N
N
yes vs no
ug/L
273
2.100
(0.9, 4.4)
590
urine
MTHL3
c
Azaroff
1999
MLGR-7
< 0.01
N
N
yes vs no
ug/L
274
2.800
(1 -4, 5.8)
279
urine
MTHL4
c
Azaroff
1999
MLGR-3
< 0.01
N
N
yes vs no
ug/L
274
3.200
(1 -8, 5.7)
281
urine
MTHL4
c
Azaroff
1999
MLGR-3
< 0.05
N
N
yes vs no
ug/L
274
3.200
(1 -2, 9.6)
277
urine
DAP2
c
Azaroff
1999
MLGR-1
< 0.05
N
N
yes vs no
ug/L
274
1.900
(1.1, 3.3)
280
urine
MTHL5
c
Azaroff
1999
MLGR-4
< 0.01
N
N
yes vs no
ug/L
274
3.900
(1.9, 8.0)
278
urine
DAP3
c
Azaroff
1999
MLGR-2
< 0.10
N
N
yes vs no
ug/L
274
2.100
(1.0, 4.9)
Q403~Recent Fieldwork
271
urine
ETHL3
c
Azaroff
1999
MLGR-5
> 0.10
N
N
yes vs no
ug/L
273
1.700
(0.8, 3.9)
272
urine
ETHL3
c
Azaroff
1999
MLGR-6
> 0.10
N
N
yes vs no
ug/L
273
1.800
(0.8, 3.9)
C-34
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
269
urine
MTHL4
C
Azaroff
1999
MLGR-3
< 0.01
N
N
yes vs no
ug/L
274
3.100
(1 -8, 5.5)
267
urine
DAP2
C
Azaroff
1999
MLGR-1
< 0.01
N
N
yes vs no
ug/L
274
3.100
(1 -8, 5.2)
270
urine
MTHL5
c
Azaroff
1999
MLGR-4
< 0.05
N
N
yes vs no
ug/L
274
2.600
(1 -2, 6.1)
268
urine
DAP3
c
Azaroff
1999
MLGR-2
< 0.05
N
N
yes vs no
ug/L
274
2.400
(1.1, 5.3)
Q404-Applicator vs Farmworker
239
urine
4NITR
c
Fenske
2002
MWU
> 0.05
N
N
applicator
ug/L
0.000
1.100
5.100
49
239
urine
4NITR
c
Fenske
2002
MWU
> 0.05
N
N
farm-worker
ug/L
0.000
121.000
419.000
12
238
urine
TCPY
c
Fenske
2002
MWU
> 0.05
N
N
applicator
ug/L
0.000
4.500
15.000
49
238
urine
TCPY
c
Fenske
2002
MWU
> 0.05
N
N
farm-worker
ug/L
0.000
6.400
15.000
12
319
urine
DMTP
c
Lu 2000
MWU
>0.10
N
N
applicator
ug/ml
0.030
0.040
0.050
49
319
urine
DMTP
c
Lu 2000
MWU
>0.10
N
N
farm-worker
ug/ml
0.020
0.030
0.040
13
320
urine
DM DTP
c
Lu 2000
MWU
>0.10
N
N
applicator
ug/ml
0.000
0.005
0.010
49
320
urine
DM DTP
c
Lu 2000
MWU
>0.10
N
N
farm-worker
ug/ml
0.000
0.002
0.003
13
321
urine
MTHL1
c
Lu 2000
MWU
>0.10
N
N
applicator
ug/ml
0.060
0.100
0.100
49
321
urine
MTHL1
c
Lu 2000
MWU
>0.10
N
N
farm-worker
ug/ml
0.050
0.070
0.080
13
316
dust
AZM
c
Lu 2000
MWU
>0.10
N
N
applicator
ug/g
1.060
2.060
2.300
49
316
dust
AZM
c
Lu 2000
MWU
>0.10
N
N
farm-worker
ug/g
0.750
1.470
1.500
13
318
dust
AZMPH
c
Lu 2000
MWU
0.07
N
N
applicator
ug/g
2.360
3.290
3.200
49
318
dust
AZMPH
c
Lu 2000
MWU
0.07
N
N
farm-worker
ug/g
0.920
1.610
1.600
13
232
dust
CHLR
c
Fenske
2002
MWU
> 0.05
N
N
applicator
ug/g
0.370
0.550
0.580
49
232
dust
CHLR
c
Fenske
2002
MWU
> 0.05
N
N
farm-worker
ug/g
0.250
0.270
0.180
12
C-35
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
235
dust
EPAR
C
Fenske
2002
MWU
0.03
N
N
applicator
ug/g
0.010
0.070
0.160
49
235
dust
EPAR
C
Fenske
2002
MWU
0.03
N
N
farm-worker
ug/g
0.000
0.020
0.080
12
317
dust
PHSM
c
Lu 2000
MWU
>0.10
N
N
applicator
ug/g
0.150
1.230
2.500
49
317
dust
PHSM
c
Lu 2000
MWU
>0.10
N
N
farm-worker
ug/g
0.110
0.140
0.100
13
Q405~Applicator vs Non-applicator
387
dust
AZM
c
Simcox
1995
MWU
> 0.05
N
N
applicator
ng/g
1225.000
1955.000
28
387
dust
AZM
c
Simcox
1995
MWU
> 0.05
N
N
non-
applicator
ng/g
769.000
1758.000
20
391
dust
AZM
L
Simcox
1995
MWU
> 0.05
N
N
non-
applicator
ug/m2
5.800
13.700
20
391
dust
AZM
L
Simcox
1995
MWU
> 0.05
N
N
applicator
ug/m2
14.400
19.300
28
425
dust
AZM
C
Simcox
1995
OWAN
> 0.05
Y
N
applicator vs
non-
applicator
ng/g
389
dust
CHLR
C
Simcox
1995
MWU
0.02
N
N
applicator
ng/g
395.000
514.000
28
389
dust
CHLR
C
Simcox
1995
MWU
0.02
N
N
non-
applicator
ng/g
156.000
318.000
20
393
dust
CHLR
L
Simcox
1995
MWU
0.04
N
N
applicator
ug/m2
2.700
5.700
28
393
dust
CHLR
L
Simcox
1995
MWU
0.04
N
N
non-
applicator
ug/m2
1.200
3.500
20
427
dust
CHLR
C
Simcox
1995
OWAN
> 0.05
Y
N
applicator vs
non-
applicator
ng/g
390
dust
EPAR
C
Simcox
1995
MWU
< 0.001
N
N
applicator
ng/g
273.000
516.000
28
390
dust
EPAR
C
Simcox
1995
MWU
< 0.001
N
N
non-
applicator
ng/g
<11
161.000
20
C-36
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
394
dust
EPAR
L
Simcox
1995
MWU
0.002
N
N
applicator
ug/m2
2.700
5.100
28
394
dust
EPAR
L
Simcox
1995
MWU
0.002
N
N
non-
applicator
ug/m2
0.050
2.200
20
428
dust
EPAR
C
Simcox
1995
OWAN
0.001
Y
N
applicator vs
non-
applicator
ng/g
429
dust
EPAR
C
Simcox
1995
TWAN-1
0.002
Y
N
applicator vs
non-
applicator
ng/g
430
dust
EPAR
C
Simcox
1995
TWAN-2
> 0.05
Y
N
applicator vs
non-
applicator
ng/g
388
dust
PHSM
C
Simcox
1995
MWU
> 0.05
N
N
applicator
ng/g
523.000
2108.000
28
388
dust
PHSM
C
Simcox
1995
MWU
> 0.05
N
N
non-
applicator
ng/g
523.000
2137.000
20
392
dust
PHSM
L
Simcox
1995
MWU
> 0.05
N
N
applicator
ug/m2
5.200
28.000
28
392
dust
PHSM
L
Simcox
1995
MWU
> 0.05
N
N
non-
applicator
ug/m2
2.500
27.500
20
426
dust
PHSM
C
Simcox
1995
OWAN
> 0.05
Y
N
applicator vs
non-
applicator
ng/g
Q406~Applicator vs Reference
201
urine
DMTP
C
Loewen-
herz 1997
CHSQ
> 0.10
N
N
applicator vs
reference
ug/ml
202
urine
DMTP
C
Loewen-
herz 1997
CHSQ
0.022
N
N
applicator
ug/ml
67
202
urine
DMTP
C
Loewen-
herz 1997
CHSQ
0.022
N
N
reference
ug/ml
40
176
urine
DMTP
C
Loewen-
herz 1997
MWU
> 0.10
N
N
applicator
ug/ml
0.015
0.033
46
176
urine
DMTP
C
Loewen-
herz 1997
MWU
> 0.10
N
N
reference
ug/ml
0.000
0.016
13
C-37
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
177
urine
DMTP
C
Loewen-
herz 1997
MWU
0.036
N
N
applicator
ug/ml
0.019
0.049
46
177
urine
DMTP
C
Loewen-
herz 1997
MWU
0.036
N
N
reference
ug/ml
0.000
0.015
13
178
urine
DMTP
c
Loewen-
herz 1997
MWU
0.015
N
N
applicator
ug/ml
0.021
0.042
90
178
urine
DMTP
c
Loewen-
herz 1997
MWU
0.015
N
N
reference
ug/ml
0.005
0.016
25
198
urine
DMTP
c
Loewen-
herz 1997
MWU
> 0.10
N
N
applicator
ug/ml
0.035
0.096
46
46
198
urine
DMTP
c
Loewen-
herz 1997
MWU
> 0.10
N
N
reference
ug/ml
0.000
0.016
23
13
199
urine
DMTP
c
Loewen-
herz 1997
MWU
0.022
N
N
applicator
ug/ml
0.019
0.049
56
43
199
urine
DMTP
c
Loewen-
herz 1997
MWU
0.022
N
N
reference
ug/ml
0.000
0.015
33
12
200
urine
DMTP
A
Loewen-
herz 1997
MWU
0.011
N
N
applicator
ug/g
Cre
0.037
0.094
51
89
200
urine
DMTP
A
Loewen-
herz 1997
MWU
0.011
N
N
reference
ug/g
Cre
0.000
0.040
28
27
Q407~Applicator+Farmworker vs Reference
241
urine
4NITR
C
Fenske
2002
KWAN
> 0.05
N
N
applicator
ug/L
0.000
1.100
5.100
49
241
urine
4NITR
C
Fenske
2002
KWAN
> 0.05
N
N
farm-worker
ug/L
0.000
121.000
419.000
12
241
urine
4NITR
C
Fenske
2002
KWAN
> 0.05
N
N
reference
ug/L
0.000
0.460
1.700
14
231
urine
4NITR
C
Fenske
2002
TNR
> 0.05
N
N
agricultural
ug/L
0.000
25.000
190.000
61
231
urine
4NITR
C
Fenske
2002
TNR
> 0.05
N
N
reference
ug/L
0.000
0.460
1.700
14
240
urine
TCPY
C
Fenske
2002
KWAN
> 0.05
N
N
applicator
ug/L
0.000
4.500
15.000
49
C-38
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
240
urine
TCPY
C
Fenske
2002
KWAN
> 0.05
N
N
farm-worker
ug/L
0.000
6.400
15.000
12
240
urine
TCPY
C
Fenske
2002
KWAN
> 0.05
N
N
reference
ug/L
0.000
4.600
9.200
14
230
urine
TCPY
c
Fenske
2002
TNR
> 0.05
N
N
agricultural
ug/L
0.000
4.900
15.000
61
230
urine
TCPY
c
Fenske
2002
TNR
> 0.05
N
N
reference
ug/L
0.000
4.600
9.200
14
325
urine
DMTP
c
Lu 2000
MWU
0.07
N
N
agricultural
ug/ml
0.020
0.040
0.040
62
325
urine
DMTP
c
Lu 2000
MWU
0.07
N
N
reference
ug/ml
0.005
0.020
0.040
14
326
urine
DM DTP
c
Lu 2000
MWU
>0.10
N
N
agricultural
ug/ml
0.000
0.004
0.009
62
326
urine
DM DTP
c
Lu 2000
MWU
>0.10
N
N
reference
ug/ml
0.000
0.003
0.005
14
327
urine
MTHL1
c
Lu 2000
MWU
0.09
N
N
agricultural
ug/ml
0.050
0.090
0.110
62
327
urine
MTHL1
c
Lu 2000
MWU
0.09
N
N
reference
ug/ml
0.010
0.060
0.090
14
322
dust
AZM
c
Lu 2000
MWU
< 0.001
N
N
agricultural
ug/g
1.000
1.940
2.190
62
322
dust
AZM
c
Lu 2000
MWU
< 0.001
N
N
reference
ug/g
0.150
0.290
0.350
14
324
dust
AZMPH
c
Lu 2000
MWU
< 0.001
N
N
agricultural
ug/g
1.920
2.950
3.000
62
324
dust
AZMPH
c
Lu 2000
MWU
< 0.001
N
N
reference
ug/g
0.270
0.370
0.370
14
233
dust
CHLR
c
Fenske
2002
KWAN
< 0.001
N
N
applicator
ug/g
0.370
0.550
0.580
49
233
dust
CHLR
c
Fenske
2002
KWAN
< 0.001
N
N
farm-worker
ug/g
0.250
0.270
0.180
12
233
dust
CHLR
c
Fenske
2002
KWAN
< 0.001
N
N
reference
ug/g
0.070
0.090
0.090
14
234
dust
CHLR
c
Fenske
2002
MWU
< 0.001
N
N
agricultural
ug/g
0.340
0.500
0.540
61
234
dust
CHLR
c
Fenske
2002
MWU
< 0.001
N
N
reference
ug/g
0.070
0.090
0.090
14
244
dust
CHLR
c
Fenske
2002
MWU
< 0.01
N
N
(applicator +
farm-worker)
vs reference
ug/g
C-39
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
236
dust
EPAR
C
Fenske
2002
KWAN
< 0.01
N
N
applicator
ug/g
0.010
0.070
0.160
49
236
dust
EPAR
C
Fenske
2002
KWAN
< 0.01
N
N
farm-worker
ug/g
0.000
0.020
0.080
12
236
dust
EPAR
c
Fenske
2002
KWAN
< 0.01
N
N
reference
ug/g
0.000
0.003
0.010
14
237
dust
EPAR
c
Fenske
2002
MWU
0.02
N
N
agricultural
ug/g
0.000
0.060
0.140
61
237
dust
EPAR
c
Fenske
2002
MWU
0.02
N
N
reference
ug/g
0.000
0.003
0.010
14
245
dust
EPAR
c
Fenske
2002
MWU
> 0.05
N
N
(applicator +
farm-worker)
vs reference
ug/g
323
dust
PHSM
c
Lu 2000
MWU
0.02
N
N
agricultural
ug/g
0.140
1.010
2.270
62
323
dust
PHSM
c
Lu 2000
MWU
0.02
N
N
reference
ug/g
0.090
0.090
0.040
14
Q408~Farmer vs Farmworker
379
dust
AZM
c
Simcox
1995
MWU
> 0.05
N
N
farmer vs
farm-worker
ng/g
383
dust
AZM
L
Simcox
1995
MWU
> 0.05
N
N
farmer
ug/m2
10.700
16.600
26
383
dust
AZM
L
Simcox
1995
MWU
> 0.05
N
N
farm-worker
ug/m2
8.000
16.700
22
650
dust
AZM
C
Simcox
1995
OWAN
> 0.05
Y
N
farmer vs
farm-worker
ng/g
381
dust
CHLR
C
Simcox
1995
MWU
> 0.05
N
N
farmer vs
farm-worker
ng/g
385
dust
CHLR
L
Simcox
1995
MWU
> 0.05
N
N
farmer
ug/m2
1.620
4.100
26
385
dust
CHLR
L
Simcox
1995
MWU
> 0.05
N
N
farm-worker
ug/m2
2.000
5.400
22
654
dust
CHLR
C
Simcox
1995
OWAN
> 0.05
Y
N
farmer vs
farm-worker
ng/g
382
dust
EPAR
C
Simcox
1995
MWU
< 0.001
N
N
farmer vs
farm-worker
ng/g
C-40
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
386
dust
EPAR
L
Simcox
1995
MWU
> 0.05
N
N
farmer
ug/m2
2.500
5.200
26
386
dust
EPAR
L
Simcox
1995
MWU
> 0.05
N
N
farm-worker
ug/m2
0.570
2.400
22
656
dust
EPAR
C
Simcox
1995
OWAN
0.001
Y
N
farmer vs
farm-worker
ng/g
659
dust
EPAR
C
Simcox
1995
TWAN-2
> 0.05
Y
N
farmer vs
farm-worker
ng/g
431
dust
EPAR
C
Simcox
1995
TWAN-3
> 0.05
Y
N
farmer vs
farm-worker
ng/g
380
dust
PHSM
C
Simcox
1995
MWU
> 0.05
N
N
farmer vs
farm-worker
ng/g
384
dust
PHSM
L
Simcox
1995
MWU
> 0.05
N
N
farmer
ug/m2
2.100
18.400
26
384
dust
PHSM
L
Simcox
1995
MWU
> 0.05
N
N
farm-worker
ug/m2
8.400
36.100
22
652
dust
PHSM
C
Simcox
1995
OWAN
> 0.05
Y
N
farmer vs
farm-worker
ng/g
367
soil
AZM
C
Simcox
1995
MWU
> 0.05
N
N
farmer
ng/g
<32
84.000
50
26
367
soil
AZM
C
Simcox
1995
MWU
> 0.05
N
N
farm-worker
ng/g
<32
<32
32
22
369
soil
CHLR
C
Simcox
1995
MWU
> 0.05
N
N
farmer
ng/g
<11
18.000
23
26
369
soil
CHLR
C
Simcox
1995
MWU
> 0.05
N
N
farm-worker
ng/g
<11
14.000
23
22
370
soil
EPAR
C
Simcox
1995
MWU
> 0.05
N
N
farmer
ng/g
<34
46.000
4
26
370
soil
EPAR
C
Simcox
1995
MWU
> 0.05
N
N
farm-worker
ng/g
<34
<34
0
22
368
soil
PHSM
C
Simcox
1995
MWU
> 0.05
N
N
farmer
ng/g
<7
38.000
19
26
368
soil
PHSM
C
Simcox
1995
MWU
> 0.05
N
N
farm-worker
ng/g
<7
11.000
14
22
C-41
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q409~Farmer+Farmworker vs Reference
371
dust
AZM
C
Simcox
1995
MWU
0.001
N
N
agricultural
ng/g
1100.000
1870.000
100
48
371
dust
AZM
C
Simcox
1995
MWU
0.001
N
N
reference
ng/g
283.000
330.000
100
11
375
dust
AZM
L
Simcox
1995
MWU
> 0.05
N
N
agricultural
ug/m2
9.900
16.600
48
375
dust
AZM
L
Simcox
1995
MWU
> 0.05
N
N
reference
ug/m2
0.830
14.000
11
373
dust
CHLR
C
Simcox
1995
MWU
0.01
N
N
agricultural
ng/g
267.000
429.000
98
48
373
dust
CHLR
C
Simcox
1995
MWU
0.01
N
N
reference
ng/g
53.000
168.000
82
11
377
dust
CHLR
L
Simcox
1995
MWU
> 0.05
N
N
agricultural
ug/m2
1.900
4.800
48
377
dust
CHLR
L
Simcox
1995
MWU
> 0.05
N
N
reference
ug/m2
0.470
0.590
11
374
dust
EPAR
C
Simcox
1995
MWU
0.02
N
N
agricultural
ng/g
154.000
365.000
69
48
374
dust
EPAR
C
Simcox
1995
MWU
0.02
N
N
reference
ng/g
<11
76.000
27
11
378
dust
EPAR
L
Simcox
1995
MWU
> 0.05
N
N
agricultural
ug/m2
1.200
3.900
48
378
dust
EPAR
L
Simcox
1995
MWU
> 0.05
N
N
reference
ug/m2
0.05
N
N
agricultural
ug/m2
3.000
27.100
48
376
dust
PHSM
L
Simcox
1995
MWU
> 0.05
N
N
reference
ug/m2
0.940
0.910
11
C-42
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
363
soil
AZM
C
Simcox
1995
MWU
0.04
N
N
agricultural
ng/g
<32
60.000
42
48
363
soil
AZM
C
Simcox
1995
MWU
0.04
N
N
reference
ng/g
<32
<32
0
11
365
soil
CHLR
c
Simcox
1995
MWU
> 0.05
N
N
agricultural
ng/g
<11
17.000
23
48
365
soil
CHLR
c
Simcox
1995
MWU
> 0.05
N
N
reference
ng/g
<11
11.000
18
11
366
soil
EPAR
c
Simcox
1995
MWU
> 0.05
N
N
agricultural
ng/g
<34
<34
2
48
366
soil
EPAR
c
Simcox
1995
MWU
> 0.05
N
N
reference
ng/g
<34
<34
0
11
364
soil
PHSM
c
Simcox
1995
MWU
> 0.05
N
N
agricultural
ng/g
<7
26.000
17
48
364
soil
PHSM
c
Simcox
1995
MWU
> 0.05
N
N
reference
ng/g
<7
<7
0
11
Q410-Farmworker vs Grower
575
dust
AZM
c
McCauley
2001a
WTWS
0.02
N
N
farm-worker
homes
ppm
0.710
25
575
dust
AZM
c
McCauley
2001a
WTWS
0.02
N
N
grower
homes
ppm
1.450
24
630
dust
AZM
c
McCauley
2001a
WTWS
> 0.05
N
N
farm-worker
homes
ppm
1.450
25
630
dust
AZM
c
McCauley
2001a
WTWS
> 0.05
N
N
grower
homes
ppm
1.640
24
Q411 -Farmworker vs Others
289
urine
ETHL2
c
Koch 2002
GLM
> 0.05
Y
N
agricultural
umol/L
0.036
1.570
621
289
urine
ETHL2
c
Koch 2002
GLM
> 0.05
Y
N
non-
agricultural
umol/L
0.036
1.550
351
291
urine
ETHL2
c
Koch 2002
GLM
> 0.05
Y
N
agricultural
umol/L
0.052
1.950
74
291
urine
ETHL2
c
Koch 2002
GLM
> 0.05
Y
N
non-
agricultural
umol/L
0.051
1.950
33
288
urine
MTHL2
c
Koch 2002
GLM
> 0.05
Y
N
agricultural
umol/L
0.079
2.490
621
C-43
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
288
urine
MTHL2
C
Koch 2002
GLM
> 0.05
Y
N
non-
agricultural
umol/L
0.081
2.510
351
290
urine
MTHL2
C
Koch 2002
GLM
> 0.05
Y
N
agricultural
umol/L
0.108
2.980
117
290
urine
MTHL2
c
Koch 2002
GLM
> 0.05
Y
N
non-
agricultural
umol/L
0.124
3.150
56
Q412~Field Worker vs Pesticide Handler
453
dust
AZM
L
Grossman
2001
MLR-7
0.011
Y
N
field worker
vs pesticide
handler
ug/m2
(0.162,
0.804)
0.361
453
dust
AZM
L
Grossman
2001
MLR-7
0.011
Y
N
field worker
ug/m2
9.630
5.370
89
453
dust
AZM
L
Grossman
2001
MLR-7
0.011
Y
N
pesticide
handler
ug/m2
3.880
6.730
23
Q413~Expected Occupational Exposure
605
urine
ETHL2
C
Koch 1999
KWAN
0.878
N
N
low vs
medium vs
high
umol/L
607
urine
ETHL2
C
Koch 1999
KWAN
0.351
N
N
low vs
medium vs
high
umol/L
609
urine
ETHL2
C
Koch 1999
KWAN
0.85
N
N
low vs
medium vs
high
umol/L
604
urine
MTHL2
C
Koch 1999
KWAN
0.93
N
N
low vs
medium vs
high
umol/L
606
urine
MTHL2
C
Koch 1999
KWAN
0.851
N
N
low vs
medium vs
high
umol/L
608
urine
MTHL2
C
Koch 1999
KWAN
0.387
N
N
low vs
medium vs
high
umol/L
448
dust
AZM
L
Grossman
2001
MLR-6
< 0.001
Y
N
low vs high
ug/m2
(3.53, 20.1)
8.410
448
dust
AZM
L
Grossman
2001
MLR-6
< 0.001
Y
N
low
ug/m2
0.984
6.670
20
C-44
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
448
dust
AZM
L
Grossman
2001
MLR-6
< 0.001
Y
N
high
ug/m2
8.000
5.810
112
449
dust
AZM
L
Grossman
2001
MLR-6
0.084
Y
N
low vs
moderate
ug/m2
(0.862,
10.2)
2.970
449
dust
AZM
L
Grossman
2001
MLR-6
0.084
Y
N
low
ug/m2
0.984
6.670
20
449
dust
AZM
L
Grossman
2001
MLR-6
0.084
Y
N
moderate
ug/m2
2.320
7.310
16
Q414~Occupational Pesticide Exposure
810
urine
ETHL2
C
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/L
811
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/g
Cre
808
urine
MTHL2
C
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/L
809
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/g
Cre
Q415~Tree Thinning
831
dust
OPSUM
C
McCauley
2003
WTWS
0.06
N
N
yes
ppm
<
831
dust
OPSUM
C
McCauley
2003
WTWS
0.06
N
N
no
ppm
>
Q416~Number in Household with High Pesticide Contact
830
dust
OPSUM
C
McCauley
2003
WTWS
0.007
N
N
1 HH
member
ppm
<
830
dust
OPSUM
C
McCauley
2003
WTWS
0.007
N
N
2 HH
members
ppm
>
C-45
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.1.5 Category 5 - Residential Proximity to Agricultural Fields
Table C.3.1.5 Relationship Details for Questions in Category 5: Residential Proximity to Agricultural Fields - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q501-Proximity of Home to Pesticide-Treated Farmland/Orchard
257
urine
4NITR
C
Fenske
2002
MWU
> 0.05
N
N
< 200 ft
ug/L
0.000
33.000
210.000
46
257
urine
4NITR
C
Fenske
2002
MWU
> 0.05
N
N
> 200 ft
ug/L
0.000
0.000
0.000
15
256
urine
TCPY
c
Fenske
2002
MWU
> 0.05
N
N
< 200 ft
ug/L
0.000
6.000
17.000
46
256
urine
TCPY
c
Fenske
2002
MWU
> 0.05
N
N
> 200 ft
ug/L
0.000
1.300
4.900
15
204
urine
DM TP
c
Loewen-
herz 1997
FISH
0.036
N
N
< 200 ft
ug/ml
-999
204
urine
DM TP
c
Loewen-
herz 1997
FISH
0.036
N
N
> 200 ft
ug/ml
-888
193
urine
DM TP
c
Loewen-
herz 1997
MWU
> 0.10
N
N
< 200 ft
ug/ml
0.015
0.034
44
36
193
urine
DM TP
c
Loewen-
herz 1997
MWU
> 0.10
N
N
> 200 ft
ug/ml
0.019
0.029
40
10
194
urine
DM TP
c
Loewen-
herz 1997
MWU
0.062
N
N
< 200 ft
ug/ml
0.023
0.056
58
36
194
urine
DM TP
c
Loewen-
herz 1997
MWU
0.062
N
N
> 200 ft
ug/ml
0.000
0.022
67
9
342
urine
DM TP
c
Lu 2000
MWU
0.009
N
N
< 200 ft
ug/ml
0.030
0.040
0.050
47
342
urine
DM TP
c
Lu 2000
MWU
0.009
N
N
> 200 ft
ug/ml
0.010
0.020
0.030
15
343
urine
DM DTP
c
Lu 2000
MWU
>0.10
N
N
< 200 ft
ug/ml
0.000
0.005
0.010
47
343
urine
DM DTP
c
Lu 2000
MWU
>0.10
N
N
> 200 ft
ug/ml
0.000
0.002
0.004
15
299
urine
ETHL2
c
Koch 2002
GLM
> 0.05
Y
N
< 200 ft
umol/L
0.033
1.440
104
299
urine
ETHL2
c
Koch 2002
GLM
> 0.05
Y
N
> 200 ft
umol/L
0.036
1.570
868
C-46
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
301
urine
ETHL2
C
Koch 2002
GLM
> 0.05
Y
N
< 200 ft
umol/L
0.035
1.460
8
301
urine
ETHL2
C
Koch 2002
GLM
> 0.05
Y
N
> 200 ft
umol/L
0.051
2.010
109
804
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
< 0.25 mile
ug/g
Cre
12.860
18.880
14.550
5
804
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
> 0.25 mile
ug/g
Cre
10.150
27.020
41.110
9
805
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
< 0.50 mile
ug/g
Cre
12.840
30.100
44.140
8
805
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
> 0.50 mile
ug/g
Cre
15.420
16.110
7.370
6
806
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
< 0.25 mile
ug/g
Cre
11.250
14.970
13.030
7
806
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
> 0.25 mile
ug/g
Cre
11.790
19.090
20.010
10
807
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
< 0.50 mile
ug/g
Cre
11.990
17.650
19.000
12
807
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
> 0.50 mile
ug/g
Cre
10.860
16.790
13.390
5
344
urine
MTHL1
C
Lu 2000
MWU
0.01
N
N
< 200 ft
ug/ml
0.070
0.100
0.110
47
344
urine
MTHL1
C
Lu 2000
MWU
0.01
N
N
> 200 ft
ug/ml
0.020
0.040
0.070
15
346
urine
MTHL1
C
Lu 2000
SLR
0.1
N
N
4 distance
categories
ug/ml
-0.200
348
urine
MTHL1
C
Lu 2000
SLR
0.06
N
N
5 distance
categories
ug/ml
A
140
urine
MTHL2
C
Curl 2002
OWAN
0.34
Y
N
6 distance
categories
umol/L
216
142
urine
MTHL2
C
Curl 2002
OWAN
0.3
Y
N
< 200 ft vs >
200 ft
umol/L
216
584
urine
MTHL2
A
Curl 2002
OWAN
0.3
Y
N
6 distance
categories
umol/g
Cre
216
585
urine
MTHL2
A
Curl 2002
OWAN
0.4
Y
N
< 200 ft vs >
200 ft
umol/g
Cre
216
298
urine
MTHL2
C
Koch 2002
GLM
> 0.05
Y
N
< 200 ft
umol/L
0.079
2.450
104
C-47
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
298
urine
MTHL2
C
Koch 2002
GLM
> 0.05
Y
N
> 200 ft
umol/L
0.080
2.510
868
300
urine
MTHL2
C
Koch 2002
GLM
> 0.05
Y
N
< 200 ft
umol/L
0.137
3.560
21
300
urine
MTHL2
c
Koch 2002
GLM
> 0.05
Y
N
> 200 ft
umol/L
0.110
2.970
152
800
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
< 0.25 mile
ug/g
Cre
22.200
122.900
220.850
5
800
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
> 0.25 mile
ug/g
Cre
31.700
44.130
48.620
9
801
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
< 0.50 mile
ug/g
Cre
41.600
105.710
172.620
8
801
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
> 0.50 mile
ug/g
Cre
19.600
27.670
29.970
6
802
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
< 0.25 mile
ug/g
Cre
34.750
68.770
90.260
7
802
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
> 0.25 mile
ug/g
Cre
33.250
99.860
149.190
10
803
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
< 0.50 mile
ug/g
Cre
40.800
73.730
90.550
12
803
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
> 0.50 mile
ug/g
Cre
9.500
120.040
197.050
5
141
dust
AZM
C
Curl 2002
OWAN
0.58
Y
N
6 distance
categories
ug/g
143
dust
AZM
C
Curl 2002
OWAN
0.58
Y
N
< 200 ft vs >
200 ft
ug/g
216
454
dust
AZM
L
Grossman
2001
MLR-8
> 0.05
Y
N
< 2 blocks vs
2-8 blocks
ug/m2
(0.282,
2.12)
0.773
454
dust
AZM
L
Grossman
2001
MLR-8
> 0.05
Y
N
< 2 blocks
ug/m2
4.640
7.070
72
454
dust
AZM
L
Grossman
2001
MLR-8
> 0.05
Y
N
2-8 blocks
ug/m2
3.590
6.350
18
455
dust
AZM
L
Grossman
2001
MLR-8
> 0.05
Y
N
< 2 blocks vs
> 8 blocks
ug/m2
(0.801,
3.11)
1.580
455
dust
AZM
L
Grossman
2001
MLR-8
> 0.05
Y
N
< 2 blocks
ug/m2
4.640
7.070
72
C-48
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
455
dust
AZM
L
Grossman
2001
MLR-8
> 0.05
Y
N
> 8 blocks
ug/m2
7.330
6.960
57
339
dust
AZM
C
Lu 2000
MWU
0.008
N
N
< 200 ft
ug/g
1.300
2.200
2.200
45
339
dust
AZM
C
Lu 2000
MWU
0.008
N
N
> 200 ft
ug/g
0.490
1.300
2.100
15
573
dust
AZM
C
McCauley
2001a
SLR
0.32
Y
N
distance
(meters)
ppm
25
574
dust
AZM
C
McCauley
2001a
SLR
0.04
Y
N
distance
(meters)
ppm
22
A
411
dust
AZM
C
Simcox
1995
KWAN
> 0.05
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ng/g
419
dust
AZM
C
Simcox
1995
KWAN
< 0.001
N
N
< 50 ft vs >
50 ft vs >
0.25 mi
ng/g
432
dust
AZM
L
Simcox
1995
KWAN
> 0.05
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ug/m2
732
dust
AZM
L
Simcox
1995
MLR-2
> 0.05
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ug/m2
415
dust
AZM
C
Simcox
1995
MWU
0.04
N
N
< 50 ft
ng/g
>
415
dust
AZM
C
Simcox
1995
MWU
0.04
N
N
> 50 ft
ng/g
<
651
dust
AZM
C
Simcox
1995
OWAN
> 0.05
Y
N
< 50 ft vs >
50 ft
ng/g
341
dust
AZMPH
C
Lu 2000
MWU
0.014
N
N
< 200 ft
ug/g
2.600
3.400
3.100
45
341
dust
AZMPH
C
Lu 2000
MWU
0.014
N
N
> 200 ft
ug/g
0.870
1.700
2.200
15
349
dust
AZMPH
C
Lu 2000
MWU
0.02
N
N
agricultural
ug/g
>
11
349
dust
AZMPH
C
Lu 2000
MWU
0.02
N
N
reference
ug/g
<
14
345
dust
AZMPH
C
Lu 2000
SLR
0.04
N
N
4 distance
categories
ug/g
-0.680
347
dust
AZMPH
C
Lu 2000
SLR
< 0.01
N
N
5 distance
categories
ug/g
A
C-49
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
254
dust
CHLR
C
Fenske
2002
MWU
< 0.01
N
N
< 200 ft
ug/g
0.400
0.590
0.590
46
254
dust
CHLR
C
Fenske
2002
MWU
< 0.01
N
N
> 200 ft
ug/g
0.150
0.220
0.180
15
258
dust
CHLR
c
Fenske
2002
SLR
< 0.001
N
N
4 distance
categories
ug/g
61
-0.160
413
dust
CHLR
c
Simcox
1995
KWAN
> 0.05
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ng/g
421
dust
CHLR
c
Simcox
1995
KWAN
0.02
N
N
< 50 ft vs >
50 ft vs >
0.25 mi
ng/g
434
dust
CHLR
L
Simcox
1995
KWAN
> 0.05
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ug/m2
734
dust
CHLR
L
Simcox
1995
MLR-1
> 0.05
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ug/m2
417
dust
CHLR
C
Simcox
1995
MWU
> 0.05
N
N
> 50 ft
ng/g
<
417
dust
CHLR
C
Simcox
1995
MWU
> 0.05
N
N
< 50 ft
ng/g
>
655
dust
CHLR
C
Simcox
1995
OWAN
> 0.05
Y
N
< 50 ft vs >
50 ft
ng/g
255
dust
EPAR
C
Fenske
2002
MWU
> 0.05
N
N
< 200 ft
ug/g
0.010
0.050
0.100
46
255
dust
EPAR
C
Fenske
2002
MWU
> 0.05
N
N
> 200 ft
ug/g
0.000
0.080
0.240
15
414
dust
EPAR
C
Simcox
1995
KWAN
0.005
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ng/g
422
dust
EPAR
C
Simcox
1995
KWAN
0.001
N
N
< 50 ft vs >
50 ft vs >
0.25 mi
ng/g
C-50
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
435
dust
EPAR
L
Simcox
1995
KWAN
> 0.05
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ug/m2
735
dust
EPAR
L
Simcox
1995
MLR-2
> 0.05
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ug/m2
418
dust
EPAR
C
Simcox
1995
MWU
0.005
N
N
> 50 ft
ng/g
<
418
dust
EPAR
C
Simcox
1995
MWU
0.005
N
N
< 50 ft
ng/g
>
657
dust
EPAR
C
Simcox
1995
OWAN
0.001
Y
N
< 50 ft vs >
50 ft
ng/g
658
dust
EPAR
C
Simcox
1995
TWAN-1
0.004
Y
N
< 50 ft vs >
50 ft
ng/g
660
dust
EPAR
C
Simcox
1995
TWAN-3
> 0.05
Y
N
< 50 ft vs >
50 ft
ng/g
340
dust
PHSM
C
Lu 2000
MWU
>0.10
N
N
< 200 ft
ug/g
1.140
1.200
2.600
45
340
dust
PHSM
C
Lu 2000
MWU
>0.10
N
N
> 200 ft
ug/g
0.120
0.450
0.600
15
412
dust
PHSM
C
Simcox
1995
KWAN
> 0.05
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ng/g
420
dust
PHSM
C
Simcox
1995
KWAN
> 0.05
N
N
< 50 ft vs >
50 ft vs >
0.25 mi
ng/g
433
dust
PHSM
L
Simcox
1995
KWAN
> 0.05
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ug/m2
733
dust
PHSM
L
Simcox
1995
MLR-3
> 0.05
N
N
< 50 ft vs 50-
200 ft vs >
200 ft
ug/m2
416
dust
PHSM
C
Simcox
1995
MWU
> 0.05
N
N
< 50 ft
ng/g
>
416
dust
PHSM
C
Simcox
1995
MWU
> 0.05
N
N
> 50 ft
ng/g
<
653
dust
PHSM
C
Simcox
1995
OWAN
> 0.05
Y
N
< 50 ft vs >
50 ft
ng/g
C-51
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
846
dust
OPSUM
C
McCauley
2003
CORR
0.5
Y
N
distance
(feet)
ppm
24
0.0080
Q502~Living Near Multiple Fields
818
urine
ETHL2
C
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/L
819
urine
ETHL2
A
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/g
Cre
816
urine
MTHL2
C
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/L
817
urine
MTHL2
A
Royster
2002
MWU
> 0.05
N
N
yes vs no
ug/g
Cre
C-52
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.1.6 Category 6 - Residential Location
Table C.3.1.6 Relationship Details for Questions in Category 6: Residential Location - Grouped by Question and Sorted by Medium, Chemical, Citation and Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q601-Urban vs Non-Urban
684
urine
1NAP
A
Adgate
2001
WTAN
0.097
Y
N
urban
ug/g
Cre
<
58
684
urine
1NAP
A
Adgate
2001
WTAN
0.097
Y
N
non-urban
ug/g
Cre
>
22
750
urine
1NAP
C
Adgate
2001
WTAN
0.13
Y
N
urban
ug/L
1.700
58
750
urine
1NAP
C
Adgate
2001
WTAN
0.13
Y
N
non-urban
ug/L
1.200
22
751
urine
1NAP
C
Adgate
2001
WTAN
0.1
Y
N
urban
ug/L
1.700
58
751
urine
1NAP
C
Adgate
2001
WTAN
0.1
Y
N
non-urban
ug/L
1.200
22
685
urine
MDA
A
Adgate
2001
WTAN
0.16
Y
N
urban
ug/g
Cre
>
58
685
urine
MDA
A
Adgate
2001
WTAN
0.16
Y
N
non-urban
ug/g
Cre
<
25
752
urine
MDA
C
Adgate
2001
WTAN
0.099
Y
N
urban
ug/L
0.770
58
752
urine
MDA
C
Adgate
2001
WTAN
0.099
Y
N
non-urban
ug/L
0.610
25
753
urine
MDA
C
Adgate
2001
WTAN
0.16
Y
N
urban
ug/L
0.770
58
753
urine
MDA
C
Adgate
2001
WTAN
0.16
Y
N
non-urban
ug/L
0.610
25
686
urine
TCPY
A
Adgate
2001
WTAN
0.019
Y
N
urban
ug/g
Cre
>
60
686
urine
TCPY
A
Adgate
2001
WTAN
0.019
Y
N
non-urban
ug/g
Cre
<
23
C-53
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
754
urine
TCPY
C
Adgate
2001
WTAN
0.036
Y
N
urban
ug/L
7.200
60
754
urine
TCPY
C
Adgate
2001
WTAN
0.036
Y
N
non-urban
ug/L
4.700
23
755
urine
TCPY
c
Adgate
2001
WTAN
0.02
Y
N
urban
ug/L
7.200
60
755
urine
TCPY
c
Adgate
2001
WTAN
0.02
Y
N
non-urban
ug/L
4.700
23
Q602-Urban vs Rural
464
urine
TCPY
A
Krinsley
1998
SLR
0.62
Y
Y
rural vs
urban
ug/g
Cre
170
0.0014
464
urine
TCPY
A
Krinsley
1998
SLR
0.62
Y
Y
rural
ug/g
Cre
7.360
8.900
24
464
urine
TCPY
A
Krinsley
1998
SLR
0.62
Y
Y
urban
ug/g
Cre
7.780
8.050
144
Q603~Border vs Non-Border
463
urine
TCPY
A
Krinsley
1998
SLR
0.86
Y
Y
border vs
non-border
ug/g
Cre
167
0.0002
463
urine
TCPY
A
Krinsley
1998
SLR
0.86
Y
Y
border
ug/g
Cre
7.810
8.970
22
463
urine
TCPY
A
Krinsley
1998
SLR
0.86
Y
Y
non-border
ug/g
Cre
7.680
8.030
149
Q604~Community
150
urine
ETHL1
C
Lu 2001
MWU
> 0.05
N
N
community 1
umol/L
0.030
0.040
50
150
urine
ETHL1
C
Lu 2001
MWU
> 0.05
N
N
community 2
umol/L
0.040
0.050
46
151
urine
MTHL2
C
Lu 2001
MWU
> 0.05
N
N
community 1
umol/L
0.100
0.170
50
151
urine
MTHL2
C
Lu 2001
MWU
> 0.05
N
N
community 2
umol/L
0.110
0.200
46
Q605-Vehicle vs House
145
dust
AZM
C
Curl 2002
SLR
< 0.001
Y
N
vehicle vs
house
ug/g
145
0.4100
C-54
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.2 Behavior Relationships
C.3.2.1 Category 7 - Subject's Personal Characteristics
Table C.3.2.1 Relationship Details for Questions in Category 7: Subject's Personal Characteristics - Grouped by Question and Sorted by Medium, Chemical, Citation
and Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q701~Sex
112
urine
1 NAP
C
Adgate
2001
WTAN
> 0.05
Y
N
male vs
female
ug/L
113
urine
MDA
C
Adgate
2001
WTAN
> 0.05
Y
N
male vs
female
ug/L
114
urine
TCPY
c
Adgate
2001
WTAN
> 0.05
Y
N
male vs
female
ug/L
460
urine
TCPY
A
Krinsley
1998
SLR
0.59
Y
Y
male vs
female
ug/g
Cre
167
0.0017
460
urine
TCPY
A
Krinsley
1998
SLR
0.59
Y
Y
male
ug/g
Cre
7.060
7.760
66
460
urine
TCPY
A
Krinsley
1998
SLR
0.59
Y
Y
female
ug/g
Cre
8.110
8.430
100
108
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
male
nmol/g
Cre
32.600
2.600
108
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
female
nmol/g
Cre
33.800
2.200
99
urine
DEP
A
Aprea 2000
MLR-5
> 0.05
Y
Y
male vs
female
nmol/g
Cre
0.0490
109
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
male
nmol/g
Cre
17.300
2.900
109
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
female
nmol/g
Cre
15.000
2.800
100
urine
DETP
A
Aprea 2000
MLR-6
> 0.05
Y
Y
male vs
female
nmol/g
Cre
0.0550
C-55
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
110
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
male
nmol/g
Cre
7.800
2.300
110
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
female
nmol/g
Cre
7.700
2.100
101
urine
DEDTP
A
Aprea 2000
MLR-7
> 0.05
Y
Y
male vs
female
nmol/g
Cre
0.0620
107
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
male
nmol/g
Cre
111.100
2.700
107
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
female
nmol/g
Cre
122.000
2.400
98
urine
DMP
A
Aprea 2000
MLR-1
> 0.05
Y
Y
male vs
female
nmol/g
Cre
0.0440
103
urine
DM TP
A
Aprea 2000
BDPH
> 0.05
Y
N
male
nmol/g
Cre
111.700
3.000
103
urine
DM TP
A
Aprea 2000
BDPH
> 0.05
Y
N
female
nmol/g
Cre
98.100
2.500
94
urine
DM TP
A
Aprea 2000
MLR-2
< 0.05
Y
Y
male vs
female
nmol/g
Cre
0.0790
104
urine
DM DTP
A
Aprea 2000
BDPH
> 0.05
Y
N
male
nmol/g
Cre
16.400
3.400
104
urine
DM DTP
A
Aprea 2000
BDPH
> 0.05
Y
N
female
nmol/g
Cre
12.300
2.500
95
urine
DM DTP
A
Aprea 2000
MLR-3
< 0.05
Y
Y
male vs
female
nmol/g
Cre
0.0800
152
urine
ETHL1
C
Lu 2001
MWU
> 0.05
N
N
male
umol/L
0.040
0.050
49
152
urine
ETHL1
C
Lu 2001
MWU
> 0.05
N
N
female
umol/L
0.040
0.040
47
111
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
male
nmol/g
Cre
66.300
2.300
111
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
female
nmol/g
Cre
63.500
2.100
102
urine
ETHL2
A
Aprea 2000
MLR-8
> 0.05
Y
Y
male vs
female
nmol/g
Cre
0.0560
601
urine
ETHL2
C
Koch 1999
MWU
0.411
N
N
male vs
female
umol/L
C-56
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
285
urine
ETHL2
C
Koch 2002
GLM-2
0.046
Y
N
male
umol/L
0.037
1.590
351
285
urine
ETHL2
C
Koch 2002
GLM-2
0.046
Y
N
female
umol/L
0.036
1.470
621
105
urine
MTHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
male
nmol/g
Cre
277.000
2.500
105
urine
MTHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
female
nmol/g
Cre
258.600
2.100
96
urine
MTHL2
A
Aprea 2000
MLR-4
< 0.05
Y
Y
male vs
female
nmol/g
Cre
0.0670
600
urine
MTHL2
C
Koch 1999
MWU
0.097
N
N
male vs
female
umol/L
284
urine
MTHL2
C
Koch 2002
GLM-1
0.005
Y
N
male
umol/L
0.085
2.450
351
284
urine
MTHL2
C
Koch 2002
GLM-1
0.005
Y
N
female
umol/L
0.078
2.290
621
153
urine
MTHL2
C
Lu 2001
MWU
> 0.05
N
N
male
umol/L
0.100
0.190
49
153
urine
MTHL2
C
Lu 2001
MWU
> 0.05
N
N
female
umol/L
0.110
0.180
47
106
urine
DAP1
A
Aprea 2000
BDPH
> 0.05
Y
N
male
nmol/g
Cre
356.400
2.300
106
urine
DAP1
A
Aprea 2000
BDPH
> 0.05
Y
N
female
nmol/g
Cre
334.800
2.000
97
urine
DAP1
A
Aprea 2000
MLR-9
< 0.05
Y
Y
male vs
female
nmol/g
Cre
0.0720
533
urine
DAP1
A
Shalat
2003
MVRG-1
0.310
N
N
male vs
female
nmol/m
ol Cre
41
(-44.02,
14.38)
-14.820
0.2800
641
urine
DAP1
A
Shalat
2003
MVRG-2
> 0.05
N
Y
male vs
female
nmol/m
ol Cre
41
0.2600
Q702—Age
116
urine
1 NAP
C
Adgate
2001
WTAN
> 0.05
Y
N
< 6 years old
vs > 6 years
old
ug/L
117
urine
MDA
C
Adgate
2001
WTAN
> 0.05
Y
N
< 6 years old
vs > 6 years
old
ug/L
C-57
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
118
urine
TCPY
C
Adgate
2001
WTAN
> 0.05
Y
N
< 6 years old
vs > 6 years
old
ug/L
461
urine
TCPY
A
Krinsley
1998
SLR
0.75
Y
Y
< 18 years
old vs 18-59
years old vs
> 60 years
old
ug/g
Cre
166
0.0006
461
urine
TCPY
A
Krinsley
1998
SLR
0.75
Y
Y
< 18 years
old
ug/g
Cre
7.130
4.670
31
461
urine
TCPY
A
Krinsley
1998
SLR
0.75
Y
Y
18-59 years
old
ug/g
Cre
7.820
9.060
101
461
urine
TCPY
A
Krinsley
1998
SLR
0.75
Y
Y
> 60 years
old
ug/g
Cre
7.680
7.940
35
671
urine
TCPY
C
Krinsley
1998
SLR
< 0.05
Y
N
<18 years
old vs 18-59
years old vs
> 60 years
old
ug/L
166
0.0006
671
urine
TCPY
C
Krinsley
1998
SLR
< 0.05
Y
N
<18 years
old
ug/L
11.530
31
671
urine
TCPY
C
Krinsley
1998
SLR
< 0.05
Y
N
18-59 years
old
ug/L
7.600
101
671
urine
TCPY
C
Krinsley
1998
SLR
< 0.05
Y
N
> 60 years
old
ug/L
7.270
35
181
urine
DM TP
C
Loewen-
herz 1997
MWU
> 0.10
N
N
0-2 years old
ug/ml
0.015
0.028
19
181
urine
DM TP
C
Loewen-
herz 1997
MWU
> 0.10
N
N
3-4 years old
ug/ml
0.009
0.029
25
182
urine
DM TP
C
Loewen-
herz 1997
MWU
> 0.10
N
N
0-2 years old
ug/ml
0.015
0.028
19
182
urine
DM TP
C
Loewen-
herz 1997
MWU
> 0.10
N
N
5-6 years old
ug/ml
0.009
0.025
19
183
urine
DM TP
C
Loewen-
herz 1997
MWU
> 0.10
N
N
3-4 years old
ug/ml
0.009
0.029
25
C-58
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
183
urine
DM TP
C
Loewen-
herz 1997
MWU
> 0.10
N
N
5-6 years old
ug/ml
0.009
0.025
19
184
urine
DM TP
C
Loewen-
herz 1997
MWU
> 0.10
N
N
0-2 years old
ug/ml
0.045
0.034
20
184
urine
DM TP
c
Loewen-
herz 1997
MWU
> 0.10
N
N
3-4 years old
ug/ml
0.033
0.059
25
185
urine
DM TP
c
Loewen-
herz 1997
MWU
> 0.10
N
N
0-2 years old
ug/ml
0.034
0.045
20
185
urine
DM TP
c
Loewen-
herz 1997
MWU
> 0.10
N
N
5-6 years old
ug/ml
0.009
0.035
20
186
urine
DM TP
c
Loewen-
herz 1997
MWU
0.06
N
N
3-4 years old
ug/ml
0.033
0.059
25
186
urine
DM TP
c
Loewen-
herz 1997
MWU
0.06
N
N
5-6 years old
ug/ml
0.009
0.035
20
187
urine
DM TP
A
Loewen-
herz 1997
MWU
> 0.10
N
N
0-2 years old
ug/g
Cre
0.042
0.099
41
17
187
urine
DM TP
A
Loewen-
herz 1997
MWU
> 0.10
N
N
3-4 years old
ug/g
Cre
0.013
0.089
36
25
188
urine
DM TP
A
Loewen-
herz 1997
MWU
> 0.10
N
N
0-2 years old
ug/g
Cre
0.042
0.099
41
17
188
urine
DM TP
A
Loewen-
herz 1997
MWU
> 0.10
N
N
5-6 years old
ug/g
Cre
0.011
0.037
37
19
189
urine
DM TP
A
Loewen-
herz 1997
MWU
> 0.10
N
N
3-4 years old
ug/g
Cre
0.013
0.089
36
25
189
urine
DM TP
A
Loewen-
herz 1997
MWU
> 0.10
N
N
5-6 years old
ug/g
Cre
0.011
0.037
37
19
190
urine
DM TP
A
Loewen-
herz 1997
MWU
> 0.10
N
N
0-2 years old
ug/g
Cre
0.061
0.223
63
19
190
urine
DM TP
A
Loewen-
herz 1997
MWU
> 0.10
N
N
3-4 years old
ug/g
Cre
0.062
0.088
63
24
191
urine
DM TP
A
Loewen-
herz 1997
MWU
0.038
N
N
0-2 years old
ug/g
Cre
0.061
0.223
63
19
191
urine
DM TP
A
Loewen-
herz 1997
MWU
0.038
N
N
5-6 years old
ug/g
Cre
0.012
0.043
47
17
C-59
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
192
urine
DM TP
A
Loewen-
herz 1997
MWU
0.083
N
N
3-4 years old
ug/g
Cre
0.062
0.088
63
24
192
urine
DM TP
A
Loewen-
herz 1997
MWU
0.083
N
N
5-6 years old
ug/g
Cre
0.012
0.043
47
17
179
urine
DM TP
C
Loewen-
herz 1997
WSRK
> 0.10
N
N
younger vs
older
ug/ml
21
180
urine
DM TP
C
Loewen-
herz 1997
WSRK
0.04
N
N
younger
ug/ml
>
180
urine
DM TP
C
Loewen-
herz 1997
WSRK
0.04
N
N
older
ug/ml
<
581
urine
ETHL1
C
Curl 2002
OWAN
> 0.05
Y
N
adult vs child
umol/L
582
urine
ETHL1
A
Curl 2002
OWAN
> 0.05
Y
N
adult vs child
umol/g
Cre
170
urine
ETHL1
C
Lu 2001
KWAN
0.64
N
N
2, 3, 4, 5
years old
umol/L
603
urine
ETHL2
C
Koch 1999
MWU
0.014
N
N
2-4 years old
umol/L
>
25
603
urine
ETHL2
C
Koch 1999
MWU
0.014
N
N
5-6 years old
umol/L
<
13
287
urine
ETHL2
C
Koch 2002
GLM-2
0.27
Y
N
5 age
categories
umol/L
144
urine
MTHL2
A
Curl 2002
OWAN
0.001
Y
N
adult
umol/g
Cre
0.090
7.200
213
144
urine
MTHL2
A
Curl 2002
OWAN
0.001
Y
N
child
umol/g
Cre
0.140
3.200
211
580
urine
MTHL2
C
Curl 2002
OWAN
0.01
Y
N
adult
umol/L
>
580
urine
MTHL2
C
Curl 2002
OWAN
0.01
Y
N
child
umol/L
<
146
urine
MTHL2
C
Curl 2002
SLR
< 0.001
Y
N
adult vs child
umol/L
206
0.1800
583
urine
MTHL2
A
Curl 2002
SLR
< 0.001
Y
N
adult vs child
umol/g
Cre
206
0.1500
602
urine
MTHL2
C
Koch 1999
MWU
0.295
N
N
2-4 years old
vs 5-6 years
old
umol/L
286
urine
MTHL2
C
Koch 2002
GLM-1
0.16
Y
N
5 age
categories
umol/L
C-60
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
171
urine
MTHL2
C
Lu 2001
KWAN
0.36
N
N
2, 3, 4, 5
years old
umol/L
532
urine
DAP1
A
Shalat
2003
MVRG-1
0.007
N
N
age in
months
nmol/
mol
Cre
41
(-3.6, -0.61)
-2.110
0.2800
640
urine
DAP1
A
Shalat
2003
MVRG-2
< 0.05
N
Y
age in
months
nmol/
mol
Cre
41
0.2600
Q703—Ethnicity
124
urine
1 NAP
C
Adgate
2001
WTAN
0.009
Y
N
white
ug/L
>
124
urine
1 NAP
C
Adgate
2001
WTAN
0.009
Y
N
non-white
ug/L
<
125
urine
MDA
C
Adgate
2001
WTAN
0.035
Y
N
white
ug/L
<
125
urine
MDA
C
Adgate
2001
WTAN
0.035
Y
N
non-white
ug/L
>
462
urine
TCPY
A
Krinsley
1998
SLR
0.99
Y
Y
Hispanic vs
non-Hispanic
ug/g
Cre
168
0.0003
462
urine
TCPY
A
Krinsley
1998
SLR
0.99
Y
Y
Hispanic
ug/g
Cre
8.470
10.510
52
462
urine
TCPY
A
Krinsley
1998
SLR
0.99
Y
Y
non-Hispanic
ug/g
Cre
7.390
6.870
116
Q704-Education Level
466
urine
TCPY
A
Krinsley
1998
SLR
0.44
Y
Y
no HS
diploma vs
HS diploma
+
ug/g
Cre
167
0.0036
466
urine
TCPY
A
Krinsley
1998
SLR
0.44
Y
Y
no HS
diploma
ug/g
Cre
9.340
7.510
53
466
urine
TCPY
A
Krinsley
1998
SLR
0.44
Y
Y
HS diploma+
ug/g
Cre
7.760
5.680
114
Q705~lncome
128
urine
1 NAP
C
Adgate
2001
WTAN
0.025
Y
N
$30K-50K
ug/L
>
C-61
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
128
urine
1 NAP
C
Adgate
2001
WTAN
0.025
Y
N
> $75K
ug/L
<
756
urine
MDA
C
Adgate
2001
WTAN
0.047
Y
N
$30K-50K
ug/L
<
756
urine
MDA
c
Adgate
2001
WTAN
0.047
Y
N
< $30K
ug/L
>
757
urine
MDA
c
Adgate
2001
WTAN
0.07
Y
N
$30K-50K
ug/L
<
757
urine
MDA
c
Adgate
2001
WTAN
0.07
Y
N
$50K-75K
ug/L
>
758
urine
MDA
c
Adgate
2001
WTAN
0.009
Y
N
$30K-50K
ug/L
<
758
urine
MDA
c
Adgate
2001
WTAN
0.009
Y
N
< $30K
ug/L
>
759
urine
TCPY
c
Adgate
2001
WTAN
0.012
Y
N
$50K-75K
ug/L
<
759
urine
TCPY
c
Adgate
2001
WTAN
0.012
Y
N
< $30 K
ug/L
>
760
urine
TCPY
c
Adgate
2001
WTAN
0.012
Y
N
$50K-75K
ug/L
<
760
urine
TCPY
c
Adgate
2001
WTAN
0.012
Y
N
$30K-50K
ug/L
>
465
urine
TCPY
A
Krinsley
1998
SLR
0.32
Y
Y
< $20K vs >
$20K
ug/g
Cre
162
0.0062
465
urine
TCPY
A
Krinsley
1998
SLR
0.32
Y
Y
< $20K
ug/g
Cre
10.240
13.500
35
465
urine
TCPY
A
Krinsley
1998
SLR
0.32
Y
Y
> $20K
ug/g
Cre
6.940
5.970
127
166
urine
ETHL1
C
Lu 2001
TNR
> 0.05
N
N
> $35K vs <
$35K
umol/L
167
urine
MTHL2
C
Lu 2001
TNR
> 0.05
N
N
> $35K vs <
$35K
umol/L
Q706-Loading From Hand Wipes3
C-62
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q707~Hand's Surface Area
534
urine
DAP1
A
Shalat
2003
MVRG-1
0.49
N
N
measure-
ment
nmol/
mol
Cre
41
(0.28, 2.28)
1.270
0.2800
642
urine
DAP1
A
Shalat
2003
MVRG-2
> 0.05
N
Y
measure-
ment
nmol
/mol
Cre
41
0.2600
a There is no question grouping for number 706.
C-63
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.2.2 Category 8 - Child's Behaviors
Table C.3.2.2 Relationship Details for Questions in Category 8: Child's Behaviors - Grouped by Question and Sorted by Medium, Chemical, Citation and Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q801--Hand-to-Mouth Activity
212
urine
4NITR
C
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/L
208
urine
TCPY
C
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/L
624
urine
ETHL1
c
Lu 2001
MWU
> 0.05
N
N
yes vs no
umol/L
304
urine
MTHL1
c
Lu 2000
MWU
0.6
N
N
yes
ug/ml
0.050
304
urine
MTHL1
c
Lu 2000
MWU
0.6
N
N
no
ug/ml
0.060
625
urine
MTHL2
c
Lu 2001
MWU
> 0.05
N
N
yes vs no
umol/L
Q802~Thumb Sucking
213
urine
4NITR
c
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/L
209
urine
TCPY
c
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/L
626
urine
ETHL1
c
Lu 2001
MWU
> 0.05
N
N
yes vs no
umol/L
305
urine
MTHL1
c
Lu 2000
MWU
0.6
N
N
yes
ug/ml
0.090
305
urine
MTHL1
c
Lu 2000
MWU
0.6
N
N
no
ug/ml
0.050
627
urine
MTHL2
c
Lu 2001
MWU
> 0.05
N
N
yes vs no
umol/L
Q803~Hand Washing before Meals
211
urine
4NITR
c
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/L
207
urine
TCPY
c
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/L
303
urine
MTHL1
c
Lu 2000
MWU
0.2
N
N
yes
ug/ml
0.090
303
urine
MTHL1
c
Lu 2000
MWU
0.2
N
N
no
ug/ml
0.050
Q804~Frequency of Handwashing
C-64
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
172
urine
ETHL1
C
Lu 2001
MWU
> 0.05
N
N
yes vs no
umol/L
173
urine
MTHL2
C
Lu 2001
MWU
> 0.05
N
N
yes vs no
umol/L
Q805~Time Spent Outdoors
210
urine
4NITR
c
Fenske
2002
KWAN
> 0.05
N
N
3 time
categories
ug/L
206
urine
TCPY
c
Fenske
2002
KWAN
> 0.05
N
N
3 time
categories
ug/L
302
urine
MTHL1
c
Lu 2000
KWAN
0.8
N
N
< 1 hr
ug/ml
0.050
302
urine
MTHL1
c
Lu 2000
KWAN
0.8
N
N
1-4 hr
ug/ml
0.050
302
urine
MTHL1
c
Lu 2000
KWAN
0.8
N
N
> 4 hr
ug/ml
0.060
Q806-Loadinq from Hand Wipe
535
urine
DAP1
A
Shalat
2003
MVRG-1
0.022
N
N
measure-
ment
nmol/
mo I
Cre
41
(0.98,
11.80)
6.390
0.2800
643
urine
DAP1
A
Shalat
2003
MVRG-2
< 0.05
N
Y
measure-
ment
nmol/
mo I
Cre
41
0.2600
C-65
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.2.3 Category 9 - Dietary Behaviors
Table C.3.2.3 Relationship Details for Questions in Category 9: Dietary Behaviors - Grouped by Question and Sorted by Medium, Chemical, Citation and Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q901~Type of Drinking Water
459
urine
TCPY
A
Krinsley
1998
SLR
< 0.19
Y
N
tap
ug/g
Cre
<
459
urine
TCPY
A
Krinsley
1998
SLR
< 0.19
Y
N
bottled
ug/g
Cre
>
Q902-Consumption of Homegrown Fresh Vegetables
457
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
Y
N
yes vs no
ug/g
Cre
Q903~Ate Lunch at School
90
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
32.900
2.400
90
urine
DEP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
33.400
2.400
81
urine
DEP
A
Aprea 2000
MLR-5
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0490
91
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
18.700
3.300
91
urine
DETP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
14.700
2.600
82
urine
DETP
A
Aprea 2000
MLR-6
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0550
92
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
7.000
2.100
92
urine
DEDTP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
8.200
2.200
83
urine
DEDTP
A
Aprea 2000
MLR-7
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0620
89
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
125.400
2.700
C-66
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
89
urine
DMP
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
111.900
2.500
80
urine
DMP
A
Aprea 2000
MLR-1
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0440
93
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
yes
nmol/g
Cre
67.400
2.300
93
urine
ETHL2
A
Aprea 2000
BDPH
> 0.05
Y
N
no
nmol/g
Cre
63.300
2.200
84
urine
ETHL2
A
Aprea 2000
MLR-8
> 0.05
Y
Y
yes vs no
nmol/g
Cre
0.0560
Q904~Organic Diet
822
urine
ETHL2
C
Curl 2003
MWU
0.13
N
N
conventional
vs organic
umol/L
39
823
urine
ETHL2
C
Curl 2003
MWU
> 0.05
N
N
conventional
vs organic
umol/L
39
820
urine
MTHL2
C
Curl 2003
MWU
< 0.001
N
N
conventional
umol/L
0.170
0.340
21
820
urine
MTHL2
C
Curl 2003
MWU
< 0.001
N
N
organic
umol/L
0.030
0.040
18
821
urine
MTHL2
C
Curl 2003
MWU
< 0.05
N
N
conventional
vs organic
umol/L
39
C-67
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.2.4 Category 10 - Family Hygiene Practices
Table C.3.2.4 Relationship Details for Questions in Category 10: Family Hygiene Practices - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q1001-Shoes Removed at Door
616
urine
DM TP
C
Carrel 1996
CHSQ
> 0.10
N
N
yes
ug/ml
40
20
616
urine
DM TP
C
Carrel 1996
CHSQ
> 0.10
N
N
no
ug/ml
36
31
617
urine
DM TP
c
Carrel 1996
CHSQ
0.083
N
N
yes
ug/ml
45
20
617
urine
DM TP
c
Carrel 1996
CHSQ
0.083
N
N
no
ug/ml
74
31
612
urine
DM TP
c
Carrel 1996
MWU
> 0.10
N
N
yes
ug/ml
0.015
0.033
20
612
urine
DM TP
c
Carrel 1996
MWU
> 0.10
N
N
no
ug/ml
0.009
0.025
31
613
urine
DM TP
c
Carrel 1996
MWU
0.096
N
N
yes
ug/ml
0.015
0.037
20
613
urine
DM TP
c
Carrel 1996
MWU
0.096
N
N
no
ug/ml
0.037
0.063
31
311
urine
MTHL1
c
Lu 2000
MWU
0.2
N
N
yes
ug/ml
0.040
311
urine
MTHL1
c
Lu 2000
MWU
0.2
N
N
no
ug/ml
0.070
440
dust
AZM
L
Grossman
2001
MLR-1
> 0.05
Y
N
always/
usually vs
sometimes/
rarely/never
ug/m2
(0.605,
2.66)
1.670
440
dust
AZM
L
Grossman
2001
MLR-1
> 0.05
Y
N
always/
usually
ug/m2
9.860
4.970
52
440
dust
AZM
L
Grossman
2001
MLR-1
> 0.05
Y
N
sometimes to
never
ug/m2
9.320
6.090
37
835
dust
AZM
C
McCauley
2003
WTWS
0.46
N
N
yes vs no
ppm
720
dust
AZM
L
Simcox
1995
MLR-1
> 0.05
N
N
yes vs no
ug/m2
351
dust
AZM
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
306
dust
AZMPH
C
Lu 2000
MWU
0.8
N
N
yes
ug/g
1.500
C-68
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
306
dust
AZMPH
C
Lu 2000
MWU
0.8
N
N
no
ug/g
2.100
722
dust
CHLR
L
Simcox
1995
MLR-3
> 0.05
N
N
yes vs no
ug/m2
353
dust
CHLR
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
723
dust
EPAR
L
Simcox
1995
MLR-4
> 0.05
N
N
yes vs no
ug/m2
354
dust
EPAR
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
721
dust
PHSM
L
Simcox
1995
MLR-2
> 0.05
N
N
yes vs no
ug/m2
352
dust
PHSM
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
834
dust
OPSUM
C
McCauley
2003
WTWS
0.36
N
N
yes vs no
ppm
Q1002~Presence of Doormats
218
urine
4NITR
C
Fenske
2002
TNR
> 0.05
N
N
yes vs no
ug/L
214
urine
TCPY
C
Fenske
2002
TNR
> 0.05
N
N
yes vs no
ug/L
312
urine
MTHL1
C
Lu 2000
MWU
0.3
N
N
yes
ug/ml
0.070
312
urine
MTHL1
C
Lu 2000
MWU
0.3
N
N
no
ug/ml
0.030
724
dust
AZM
L
Simcox
1995
MLR-1
> 0.05
N
N
yes vs no
ug/m2
355
dust
AZM
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
307
dust
AZMPH
C
Lu 2000
MWU
0.6
N
N
yes
ug/g
1.800
307
dust
AZMPH
C
Lu 2000
MWU
0.6
N
N
no
ug/g
2.900
222
dust
CHLR
C
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/g
726
dust
CHLR
L
Simcox
1995
MLR-3
> 0.05
N
N
yes vs no
ug/m2
C-69
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
357
dust
CHLR
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
226
dust
EPAR
C
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/g
727
dust
EPAR
L
Simcox
1995
MLR-4
> 0.05
N
N
yes vs no
ug/m2
358
dust
EPAR
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
725
dust
PHSM
L
Simcox
1995
MLR-2
> 0.05
N
N
yes vs no
ug/m2
356
dust
PHSM
L
Simcox
1995
MWU
> 0.05
N
N
yes vs no
ug/m2
Q1003~Presence of Floor Mats
174
urine
ETHL1
C
Lu 2001
MWU
> 0.05
N
N
yes vs no
umol/L
175
urine
MTHL2
C
Lu 2001
MWU
> 0.05
N
N
yes vs no
umol/L
Q1004~Vacuuming Frequency
221
urine
4NITR
C
Fenske
2002
MWU
> 0.05
N
N
< 1/week vs
> 1/week
ug/L
217
urine
TCPY
C
Fenske
2002
MWU
> 0.05
N
N
< 1/week vs
> 1/week
ug/L
622
urine
ETHL1
C
Lu 2001
MWU
> 0.05
N
N
< 1/week vs
> 1/week
umol/L
315
urine
MTHL1
C
Lu 2000
MWU
0.3
N
N
> 1/week
ug/ml
0.070
315
urine
MTHL1
C
Lu 2000
MWU
0.3
N
N
< 1/week
ug/ml
0.050
315
urine
MTHL1
C
Lu 2000
MWU
0.3
N
N
no answer
ug/ml
0.080
623
urine
MTHL2
C
Lu 2001
MWU
> 0.05
N
N
< 1/week vs
> 1/week
umol/L
310
dust
AZMPH
C
Lu 2000
MWU
0.6
N
N
> 1/week
ug/g
1.500
310
dust
AZMPH
C
Lu 2000
MWU
0.6
N
N
< 1/week
ug/g
2.600
310
dust
AZMPH
C
Lu 2000
MWU
0.6
N
N
no answer
ug/g
2.300
225
dust
CHLR
C
Fenske
2002
MWU
> 0.05
N
N
< 1/week vs
> 1/week
ug/g
C-70
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
229
dust
EPAR
C
Fenske
2002
MWU
> 0.05
N
N
< 1/week vs
> 1/week
ug/g
Q1005~Vacuuming Indoor Play Areas
359
dust
AZM
L
Simcox
1995
KWAN
> 0.05
N
N
< weekly vs
weekly vs >
weekly
ug/m2
728
dust
AZM
L
Simcox
1995
MLR-1
> 0.05
N
N
< weekly vs
weekly vs >
weekly
ug/m2
361
dust
CHLR
L
Simcox
1995
KWAN
> 0.05
N
N
< weekly vs
weekly vs >
weekly
ug/m2
730
dust
CHLR
L
Simcox
1995
MLR-3
> 0.05
N
N
< weekly vs
weekly vs >
weekly
ug/m2
362
dust
EPAR
L
Simcox
1995
KWAN
> 0.05
N
N
< weekly vs
weekly vs >
weekly
ug/m2
731
dust
EPAR
L
Simcox
1995
MLR-1
> 0.05
N
N
< weekly vs
weekly vs >
weekly
ug/m2
360
dust
PHSM
L
Simcox
1995
KWAN
> 0.05
N
N
< weekly vs
weekly vs >
weekly
ug/m2
729
dust
PHSM
L
Simcox
1995
MLR-2
> 0.05
N
N
< weekly vs
weekly vs >
weekly
ug/m2
Q1006~Work Clothes Worn Indoors
219
urine
4NITR
C
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/L
215
urine
TCPY
C
Fenske
2002
TNR
> 0.05
N
N
yes vs no
ug/L
716
urine
DM TP
C
Carrel 1996
CHSQ
> 0.10
N
N
yes vs no
ug/ml
717
urine
DM TP
C
Carrel 1996
CHSQ
> 0.10
N
N
yes vs no
ug/ml
712
urine
DM TP
C
Carrel 1996
MWU
> 0.10
N
N
yes vs no
ug/ml
C-71
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
713
urine
DM TP
C
Carrel 1996
MWU
> 0.10
N
N
yes vs no
ug/ml
313
urine
MTHL1
C
Lu 2000
MWU
0.2
N
N
yes
ug/ml
0.070
313
urine
MTHL1
c
Lu 2000
MWU
0.2
N
N
no
ug/ml
0.050
837
dust
AZM
c
McCauley
2003
TTST
< 0.01
Y
N
< 2 hours
ppm
0.530
0.920
18
837
dust
AZM
c
McCauley
2003
TTST
< 0.01
Y
N
> 2 hours
ppm
5.900
3.960
5
308
dust
AZMPH
c
Lu 2000
MWU
0.2
N
N
yes
ug/g
2.700
308
dust
AZMPH
c
Lu 2000
MWU
0.2
N
N
no
ug/g
1.500
223
dust
CHLR
c
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/g
227
dust
EPAR
c
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/g
836
dust
OPSUM
c
McCauley
2003
TTST
< 0.01
Y
N
< 2 hours
ppm
0.950
1.140
18
836
dust
OPSUM
c
McCauley
2003
TTST
< 0.01
Y
N
> 2 hours
ppm
6.180
4.950
5
Q1007~Work Clothes Mixed with Laundry
314
urine
MTHL1
c
Lu 2000
MWU
0.8
N
N
yes
ug/ml
0.050
314
urine
MTHL1
c
Lu 2000
MWU
0.8
N
N
no
ug/ml
0.080
309
dust
AZMPH
c
Lu 2000
MWU
0.4
N
N
yes
ug/g
1.400
309
dust
AZMPH
c
Lu 2000
MWU
0.4
N
N
no
ug/g
3.100
Q1008~Laundering Practices
220
urine
4NITR
c
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/L
216
urine
TCPY
c
Fenske
2002
TNR
> 0.05
N
N
yes vs no
ug/L
706
urine
DM TP
c
Carrel 1996
CHSQ
> 0.10
N
N
yes vs no
ug/ml
707
urine
DM TP
c
Carrel 1996
CHSQ
> 0.10
N
N
yes vs no
ug/ml
702
urine
DM TP
c
Carrel 1996
MWU
> 0.10
N
N
yes vs no
ug/ml
703
urine
DM TP
c
Carrel 1996
MWU
> 0.10
N
N
yes vs no
ug/ml
C-72
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
224
dust
CHLR
C
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/g
228
dust
EPAR
C
Fenske
2002
MWU
> 0.05
N
N
yes vs no
ug/g
Q1009-Number of Weeks Since Last Vacuuming
833
dust
OPSUM
c
McCauley
2003
MLR
0.03
Y
N
# weeks
since last
cleaning
ppm
(0.1, 2.2)
1.200
Q1010~Shower Soon After Work
866
urine
DM TP
c
Carrel 1996
CHSQ
> 0.10
N
N
yes vs no
ug/ml
867
urine
DM TP
c
Carrel 1996
CHSQ
> 0.10
N
N
yes vs no
ug/ml
862
urine
DM TP
c
Carrel 1996
MWU
> 0.10
N
N
yes vs no
ug/ml
863
urine
DM TP
c
Carrel 1996
MWU
> 0.10
N
N
yes vs no
ug/ml
441
dust
AZM
L
Grossman
2001
MLR-2
> 0.05
Y
N
< hrvs > 1
hr
ug/m2
(0.559,
2.39)
1.160
441
dust
AZM
L
Grossman
2001
MLR-2
> 0.05
Y
N
< 1 hr
ug/m2
10.400
7.310
41
441
dust
AZM
L
Grossman
2001
MLR-2
> 0.05
Y
N
> 1 hr
ug/m2
9.170
3.990
48
839
dust
AZM
C
McCauley
2003
TTST
0.89
Y
N
< 30 minutes
vs > 30
minutes
ppm
838
dust
OPSUM
C
McCauley
2003
TTST
0.63
Y
N
< 30 minutes
vs > 30
minutes
ppm
Q1011a
C-73
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q1012~After Work Hygiene Index
841
dust
AZM
C
McCauley
2003
CORR
0.43
Y
N
5 index
categories
ppm
24
0.0290
840
dust
OPSUM
C
McCauley
2003
CORR
0.8
Y
N
5 index
categories
ppm
24
0.0025
a No questions associated with this Q#.
C.3.2.5 Category 11 - Smoking-Related Activities
Table C.3.2.5 Relationship Details for Questions in Category 11: Smoking-Related Activities - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q1101 -Current Smoker
467
urine
TCPY
A
Krinsley
1998
SLR
0.009
Y
N
yes vs no
ug/g
Cre
167
0.0400
467
urine
TCPY
A
Krinsley
1998
SLR
0.009
Y
N
yes
ug/g
Cre
5.700
7.460
35
467
urine
TCPY
A
Krinsley
1998
SLR
0.009
Y
N
no
ug/g
Cre
8.190
8.270
132
Q1102-Subject Smoked
764
urine
TCPY
A
Krinsley
1998
FSLR#1
< 0.001
Y
Y
yes vs no
ug/g
Cre
-0.170
0.1800
769
urine
TCPY
A
Krinsley
1998
FSLR#2
< 0.001
Y
Y
yes vs no
ug/g
Cre
166
0.169
0.2100
Q1103-Exposure to Second Hand Smoke
474
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
Y
N
yes vs no
ug/g
Cre
C-74
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.2.6 Category 12 - Work Exposure/Practices
Table C.3.2.6 Relationship Details for Questions in Category 12: Work Exposure/Practices - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q1201-Pesticide Exposure at Work in Past 6 Mo
531
urine
TCPY
A
Krinsley
1998
SLR
0.37
Y
N
yes vs no
ug/g
Cre
0.0050
531
urine
TCPY
A
Krinsley
1998
SLR
0.37
Y
N
yes
ug/g
Cre
5.420
3.410
11
531
urine
TCPY
A
Krinsley
1998
SLR
0.37
Y
N
no
ug/g
Cre
7.020
5.740
65
Q1202-Wear Boots While Doing Fieldwork?
446
dust
AZM
L
Grossman
2001
MLR-4
> 0.05
Y
N
always
/usually vs
sometimes/
rarely/never
ug/m2
(0.423,
1.83)
0.880
446
dust
AZM
L
Grossman
2001
MLR-4
> 0.05
Y
N
always/
usually
ug/m2
8.960
5.510
36
446
dust
AZM
L
Grossman
2001
MLR-4
> 0.05
Y
N
sometimes to
never
ug/m2
10.100
5.360
53
Q1203-Wear Gloves While Doing Fieldwork?
447
dust
AZM
L
Grossman
2001
MLR-5
> 0.05
Y
N
always/
usually vs
sometimes/
rarely/never
ug/m2
(0.313,
1.64)
0.717
447
dust
AZM
L
Grossman
2001
MLR-5
> 0.05
Y
N
always/
usually
ug/m2
6.690
4.840
25
447
dust
AZM
L
Grossman
2001
MLR-5
> 0.05
Y
N
sometimes to
never
ug/m2
11.100
5.540
64
C-75
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q1204-Wear Hat While Doing Fieldwork?
445
dust
AZM
L
Grossman
2001
MLR-3
> 0.05
Y
N
always/
usually vs
sometimes/
rarely/never
ug/m2
(0.651,
4.20)
1.650
445
dust
AZM
L
Grossman
2001
MLR-3
> 0.05
Y
N
always/
usually
ug/m2
10.200
5.280
74
445
dust
AZM
L
Grossman
2001
MLR-3
> 0.05
Y
N
sometimes to
never
ug/m2
7.340
6.030
15
C-76
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.3 Other Relationships
C.3.3.1 Category 13 - Related Exposure Levels
Table C.3.3.1 Relationship Details for Questions in Category 13: Related Exposure Levels - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q1301 -Detectable Levels In Adult Household Members
262
urine
DAP2
C
Azaroff
1999
SLGR
< 0.01
N
N
number of
adults
ug/L
135
2.000
(1.4, 3.0)
265
urine
DAP2
C
Azaroff
1999
SLGR
0.1
N
N
yes vs no
ug/L
4.400
(0.9, 2.2)
Q1302~High Levels In Adult Household Members
266
urine
MTHL4
c
Azaroff
1999
SLGR
< 0.01
N
N
yes
ug/L
>
12
266
urine
MTHL4
c
Azaroff
1999
SLGR
< 0.01
N
N
no
ug/L
<
30
263
urine
DAP2
c
Azaroff
1999
SLGR
< 0.01
N
N
number of
adults
ug/L
135
2.100
(1.3, 3.6)
264
urine
DAP3
c
Azaroff
1999
SLGR
< 0.01
N
N
number of
adults
ug/L
135
2.200
(1.2, 4.0)
C-77
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
C.3.3.2 Category 14 - Health
Table C.3.3.2 Relationship Details for Questions in Category 14: Health - Grouped by Question and Sorted by Medium, Chemical, Citation and Analysis
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
Q1401 -Health Status
472
urine
TCPY
A
Krinsley
1998
SLR
< 0.66
Y
N
good vs fair-
poor
ug/g
Cre
Q1402-Asthma and Allergies
470
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
Y
N
yes vs no
ug/g
Cre
Q1403-Bowel Disease
761
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
Y
Y
yes vs no
ug/g
Cre
-0.380
0.1800
Q1404-Dlabetes
471
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
Y
N
yes vs no
ug/g
Cre
Q1405-lntestlnal Disease
765
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
Y
Y
yes vs no
ug/g
Cre
166
0.326
0.2100
770
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
Y
Y
yes vs no
ug/g
Cre
71
-0.305
0.3500
469
urine
TCPY
A
Krinsley
1998
SLR
0.004
Y
N
yes vs no
ug/g
Cre
166
0.0500
469
urine
TCPY
A
Krinsley
1998
SLR
0.004
Y
N
yes
ug/g
Cre
4.470
4.400
13
469
urine
TCPY
A
Krinsley
1998
SLR
0.004
Y
N
no
ug/g
Cre
7.950
8.370
153
Q1406—Ulcer
468
urine
TCPY
A
Krinsley
1998
SLR
0.02
Y
N
yes vs no
ug/g
Cre
167
0.0300
468
urine
TCPY
A
Krinsley
1998
SLR
0.02
Y
N
yes
ug/g
Cre
5.380
4.520
22
C-78
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID#
Me-
dium
Chemi-
cal
M
T
Citation
Analy-
sis
p-value
L
G
P
M
Groups
Compared
Group Name
Units
Gmean
GSD
Median
Mean
StDev
PctD
N
OR
CI
Beta
R2
468
urine
TCPY
A
Krinsley
1998
SLR
0.02
Y
N
no
ug/g
Cre
8.010
8.520
145
C-79
August 2005
-------
-------
Appendix D
Comment Tables for Relationships from Literature Review
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendix D
Comment Tables for Relationship from Literature Review
Contents
D. 1 Description D-l
D.2 Reference Information D-3
D.3 Comment Tables D-10
D.3.1 Source Relationships D-10
D.3.1.1 Category 1: Residential Pesticide Use D-10
D.3.1.2 Category 2: Household Characteristics D-22
D.3.1.3 Category 3: Residential Sources (Environmental Measures) D-29
D.3.1.4 Category 4: Household Occupation D-31
D.3.1.5 Category 5: Residential Proximity to Agricultural Fields D-41
D.3.1.6 Category 6: Residential Location D-48
D.3.2 Behavior Relationships D-50
D.3.2.1 Category 7: Subject's Personal Characteristics D-50
D.3.2.2 Category 8: Child's Behaviors D-56
D.3.2.3 Category 9: Dietary Behaviors D-58
D.3.2.4 Category 10: Family Hygiene Practices D-60
D.3.2.5 Category 11: Smoking-Related Activities D-68
D.3.2.6 Category 12: Work Exposure/Practices D-69
D.3.3 Other Relationships D-70
D.3.3.1 Category 13: Related Exposure Levels D-70
D.3.3.2 Category 14: Health D-71
D-ii August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Tables
Table D. 1 Example of Relationship Comment Table for Question Category: Residential Pesticide Use (Detail Table - Table C. 1 Appendix
C) D-l
Table D.2.1 List of Columns and Associated Reference Tables for Comment Tables D-3
Table D.2.2 Chemical/Metabolite Reference Table D-4
Table D.2.3 Medium Reference Table D-6
Table D.2.4 Type of Measurement Reference Table D-7
Table D.2.5 Statistical Analysis Reference Table D-7
Table D.2.6 Table Numbers Cross-Referenced between Results Section and Appendices A, B, and C, by Category Group D-9
Table D.3.1.1 Relationship Comments for Questions in Category 1: Residential Pesticide Use - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis D-10
Table D.3.1.2 Relationship Comments for Questions in Category 2: Household Characteristics - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis D-22
Table D.3.1.3 Relationship Comments for Questions in Category 3: Residential Sources (Environmental Measures) - Grouped by Question and
Sorted by Medium, Chemical, Citation and Analysis D-29
Table D.3.1.4 Relationship Comments for Questions in Category 4: Household Occupation - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis D-31
Table D.3.1.5 Relationship Comments for Questions in Category 5: Residential Proximity to Agricultural Fields - Grouped by Question and
Sorted by Medium, Chemical, Citation and Analysis D-41
Table D.3.1.6 Relationship Comments for Questions in Category 6: Residential Location - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis D-48
Table D.3.2.1 Relationship Comments for Questions in Category 7: Subject's Personal Characteristics - Grouped by Question and Sorted by
Medium, Chemical, Citation and Analysis D-50
Table D.3.2.2 Relationship Comments for Questions in Category 8: Child's Behaviors - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis D-56
Table D.3.2.3 Relationship Comments for Questions in Category 9: Dietary Behaviors - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis D-58
D-iii
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table D.3.2.4 Relationship Comments for Questions in Category 10: Family Hygiene Practices - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis D-60
Table D.3.2.5 Relationship Comments for Questions in Category 11: Smoking-Related Activities - Grouped by Question and Sorted by
Medium, Chemical, Citation and Analysis D-68
Table D.3.2.6 Relationship Comments for Questions in Category 12: Work Exposure/Practices - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis D-69
Table D.3.3.1 Relationship Comments for Questions in Category 13: Related Exposure Levels - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis D-70
Table D.3.3.2 Relationship Comments for Questions in Category 14: Health - Grouped by Question and Sorted by Medium, Chemical, Citation
and Analysis D-71
D-iv
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendices B, C, and D provide specific information about the relationships extracted from the literature review and summarized in Results sections
4.2.4, 4.2.5, and 4.2.6. The information is presented as overview, detail, and comment tables. Each appendix includes one type of table for all the
question categories and relationships. This appendix presents the comment tables.
D.l Description
Information about each relationship with respect to the subpopulation analyzed, the chemical measurement, and the analysis are included in the
comment tables. Table D. 1 is an example of the comment table associated with Table C. 1 in Appendix C. The sections in Table D. 1 are organized
by question as in Table C.l, and the rows within each question section are sorted by medium, chemical groupings (Table D.2.2), citation, and
analysis type. The comment tables, however, contain only one row for each relationship, and the information for the relationship can be matched to
the information in the detail table by the relationship ID number in the first column. The columns ID through MT match in both tables for a
particular ID#. The Original Question phrasing is the description provided in the publication. In most cases the full phrasing was not provided.
Table D.l Example of Relationship Comment Table for Question Category: Residential Pesticide Use (Detail Table - Table C.l Appendix C)
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q101--Pesticide Use
484
urine
TCPY
A
Krinsley
1998
SLR
0.77
any pesticide use
814
urine
ETHL2
C
Royster
2002
MWU
> 0.05
pesticide use
815
urine
ETHL2
A
Royster
2002
MWU
> 0.05
pesticide use
812
urine
MTHL2
C
Royster
2002
MWU
> 0.05
pesticide use
813
urine
MTHL2
A
Royster
2002
MWU
> 0.05
pesticide use
633
dust
AZM
C
McCauley
2001a
TNR
> 0.05
family use of pesticide
control products
farmworker and grower
homes in Hood River
County
D-l
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
848
dust
OPSUM
C
McCauley
2003
WTWS
0.39
family use of pesticide
control products
Q102--lnside Treated
563
urine
MDA
C
Sexton
2003
BSLR-5
0.1
In the past six months
were any chemicals for
the control of fleas,
roaches, ants, or other
insects used inside this
house/apartment?
unexpected direction of
coefficient
557
urine
MDA
c
Sexton
2003
LGRG
0.174
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
485
urine
TCPY
A
Krinsley
1998
SLR
0.93
pesticide use inside in
past 6 mo
164
urine
ETHL1
C
Lu 2001
MWU
0.27
pesticide use inside
focus children:
communities combined
average urine samples per child
165
urine
MTHL2
C
Lu 2001
MWU
0.35
pesticide use inside
focus children:
communities combined
average urine samples per child
561
dust
CHLR
L
Sexton
2003
LGRG
0.436
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable); child's play
area
558
indair
CHLR
C
Sexton
2003
LGRG
0.296
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
554
indair
MAL
C
Sexton
2003
LGRG
0.369
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
559
outdair
CHLR
C
Sexton
2003
LGRG
0.715
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
555
outdair
MAL
C
Sexton
2003
LGRG
0.373
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
D-2
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
562
persair
CHLR
C
Sexton
2003
BSLR-1
0.04
In the past six months
were any chemicals for
the control of fleas,
roaches, ants, or other
insects used inside this
house/apartment?
child's breathing zone
unexpected direction of
coefficient
553
persair
MAL
C
Sexton
2003
LGRG
0.073
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable); child's
breathing zone
560
sldfood
CHLR
I
Sexton
2003
LGRG
0.38
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
556
sldfood
MAL
I
Sexton
2003
LGRG
0.06
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
Q103--lnside Treated-Bath room
498
urine
TCPY
A
Krinsley
1998
SLR
0.36
pesticide used in
bathroom in past 6 mo
D.2 Reference Information
To make the comment tables more compact, it was necessary to use abbreviations or codes in both the column names and contents. Table D.2.1
describes each column used in the comment tables. The column Reference Table identifies the number of a subsequent table with information about
the codes used. For example, the column MT includes codes described in Table D.2.5.
Table D.2.1 List of Columns and Associated Reference Tables for Comment Tables
Column Type or
Name
Description
Applies
toa
Reference Tableb
ID#
Number assigned to each relationship
NA
D-3
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Column Type or
Name
Description
Applies
toa
Reference Tableb
Medium
Sample medium
Appendix D - Table D.2.3
Chemical
Chemical, metabolite, or molar-weighted sum
a
Appendix D - Table D.2.2
MT
Type of measurement
a
Appendix D - Table D.2.4
Citation
Citation reference
App A - Table A.1
Analysis
Type of statistical analysis performed
a
Appendix D - Table D.2.5
p-value
Probability value associated with statistical analysis
NA
Original Question
Question description as included in the publication
b
NA
Subpopulation
Analyzed
Description of particular study's subpopulation
included in this analysis
NA
Notes on
Measurement
Additional information on analytical measurements
used in the analysis
a, b
NA
Notes on Analysis
Additional information regarding the statistical
analysis
NA
a The entry "a" is a dependent variable, in this case a chemical analytical measurement. The entry "b" is an independent variable or predictor, usually a question.
b NA- Not applicable
Table D.2.2 Chemical/Metabolite Reference Table
Grouping3
Code
Medium
Description
1-Non-DAP
1NAP
urine
1-Naphthol
1-Non-DAP
4NITR
urine
4-Nitrophenol
6-Chemical
ATZ
otherb
Atrazine
1-Non-DAP
ATZM
urine
Atrazine mercapturate
D-4
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Grouping3
Code
Medium
Description
6-Chemical
AZM
other
Azinphosmethyl
6-Chemical
AZMPH
other
Azinphosmethyl+Phosmet
6-Chemical
CHLR
other
Chlorpyrifos
3-DAP Sumc
DAP1
urine
DMP+DMTP+DMDTP+DEP+DETP+DEDTP
4-DAP Detect
DAP2
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one detectable measurement)
5-DAP High
DAP3
urine
DEP, DETP, DEDTP, DMP, DMTP
(at least one high measurement)11
2-DAP
DEDTP
urine
Diethyldithiophosphate (DEDTP)
2-DAP
DEP
urine
Diethylphosphate (DEP)
2-DAP
DETP
urine
Diethylthiophosphate (DETP)
2-DAP
DMDTP
urine
Dimethyldithiophosphate (DMDTP)
2-DAP
DMP
urine
Dimethylphosphate (DMP)
2-DAP
DMTP
urine
Dimethylthiophosphate (DMTP)
6-Chemical
EPAR
other
Ethyl parathion
3-DAP Sum
ETHL1
urine
DEP+DETP
3-DAP Sum
ETHL2
urine
DEP+DETP+DEDTP
4-DAP Detect
ETHL3
urine
DEP, DETP, DEDTP
(at least one detectable measurement)
6-Chemical
MAL
other
Malathion
1-Non-DAP
MDA
urine
Malathion dicarboxylic acid
3-DAP Sum
MTHL1
urine
DMTP+DMDTP
3-DAP Sum
MTHL2
urine
DMP+DMTP+DMDTP
4-DAP Detect
MTHL3
urine
DMTP (detectable measurement)
D-5
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Grouping3
Code
Medium
Description
4-DAP Detect
MTHL4
urine
DMP, DMTP
(at least one detectable measurement)
5-DAP High
MTHL5
urine
DMP, DMTP
(at least one high measurement)11
7-M eta bo lite NA
NA
urine
NA (not available or not specified)
6-Chemical
OPSUM
other
OP sum®
6-Chemical
PHSM
other
Phosmet
1-Non-DAP
TCPY
urine
3,5,6-T richloro-2-pyridinol
a The number preceding the group name indicates the order of the group as it appears in the overview tables.
b Medium is other than urine, e.g., air, dermal
c Sums are molar-weighted unless otherwise specified.
d See definition of high measurement in Azaroff (1999).
e OP Sum = azinphosmethyl, chlorpyrifos, malathion, and phosmet
Table D.2.3 Medium Reference Table
Code
Description
dust
dust
indair
indoor air
outdair
outdoor air
persair
personal air
sldfood
solid food
soil
soil
urine
urine
D-6
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table D.2.4 Type of Measurement Reference Table
Code
Description
A
Adjusted concentration (urine concentration adjusted by creatinine)
C
Concentration
1
Daily intake (food)
L
Loading (dust or dermal)
Table D.2.5 Statistical Analysis Reference Table
Code
Description
BSLR-#xa
Backwards Stepwise Linear Regression #x
BDPH
Bonferroni/Dunn Post Hoc Test
CHSQ
Chi-square Test
CORR
Correlation
FISH
Fisher Exact Test
FSLR
Forward Selection Linear Regression
GLM
General Linear Model ANOVA
GLM-#x
General Linear Model ANOVA #x
KWAN
Kruskal-Wallis One-Way ANOVA
LGRG
Logistic Regression
MWU
Mann-Whitney U Test
MLR
Multiple Linear Regression
MLR-#xa
Multiple Linear Regression #x
D-7
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
MLGR-#xa
Multiple Logistic Regression #x
MVRG-#xa
Multivariate Regression #x
NAN
Not analyzed
OWAN
One-Way ANOVA
SLR
Simple Linear Regression
SLGR
Simple Logistic Regression
SPCR
Spearman Rank Correlation
TTST
t-test
TWAN-#xa
Two-Way ANOVA #x
TNR
Type of Analysis Not Reported
WTAN
Weighted ANOVA
WSRK
Wilcoxon Signed Rank Test
WTWS
Wilcoxon Two-Sample Test
a In some analyses where more than one predictor was analyzed in a relationship, the predictor questions will likely appear in different question category sections. The user can
identify the predictors that were analyzed in the same relationship by looking for the same analysis code for the citation. For example, if a multiple linear regression was performed
with three predictors on two metabolites, there would be two analysis types: < MLR-#1 and MLR-#2. The analysis type MLR-#1 would be used as the analysis type for the three
relationships describing the three predictor questions. Aprea 2000 contains examples of this type of analysis code.
D-8
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table D.2.6 provides a cross-reference between the relationship summary tables in the Results section and the tables in Appendices B, C, and D.
Table D.2.6 Table Numbers Cross-Referenced between Results Section and Appendices A, B, and C, by Category Group
Category
Section #
Table #a
Overview
Table #
Detailed
Table #
Comment
Table #
Group
#
Description
Results
Results
Appendix B
Appendix C
Appendix D
Source
1
Residential pesticide use
4.2.4.1
4.2.6.x
B.3.1.1
C.3.1.1
D.3.1.1
Source
2
Household characteristics
4.2.4.2
4.2.7.x
B.3.1.2
C.3.1.2
D.3.1.2
Source
3
Residential sources
(environmental measures)
4.2.4.3
4.2.8.x
B.3.1.3
C.3.1.3
D.3.1.3
Source
4
Household occupation
4.2.4.4
4.2.9.x
B.3.1.4
C.3.1.4
D.3.1.4
Source
5
Residential proximity to
agricultural fields
4.2.4.5
4.2.10.x
B.3.1.5
C.3.1.5
D.3.1.5
Source
6
Residential location
4.2.4.6
4.2.11.x
B.3.1.6
C.3.1.6
D.3.1.6
Behavior
7
Subject's personal
characteristics
4.2.5.1
4.2.13.x
B.3.2.1
C.3.2.1
D.3.2.1
Behavior
8
Child's behaviors
4.2.5.2
4.2.14.x
B.3.2.2
C.3.2.2
D.3.2.2
Behavior
9
Dietary behaviors
4.2.5.3
4.2.15.x
B.3.2.3
C.3.2.3
D.3.2.3
Behavior
10
Family hygiene practices
4.2.5.4
4.2.16.x
B.3.2.4
C.3.2.4
D.3.2.4
Behavior
11
Smoking-related activities
4.2.5.5
4.2.17.x
B.3.2.5
C.3.2.5
D.3.2.5
Behavior
12
Work exposure/practices
4.2.5.6
4.2.18.x
B.3.2.6
C.3.2.6
D.3.2.6
Other
13
Related exposure levels
4.2.6.1
4.2.20.x
B.3.3.1
C.3.3.1
D.3.3.1
Other
14
Health
4.2.6.2
4.2.21.x
B.3.3.2
C.3.3.2
D.3.3.2
a x in this column refers to the three table types, a, b, and c, described above.
D-9
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3 Comment Tables
D.3.1 Source Relationships
D.3.1.1 Category 1: Residential Pesticide Use
Table D.3.1.1 Relationship Comments for Questions in Category 1: Residential Pesticide Use - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q101--Pesticide Use
484
urine
TCPY
A
Krinsley
1998
SLR
0.77
any pesticide use
814
urine
ETHL2
C
Royster
2002
MWU
> 0.05
pesticide use
815
urine
ETHL2
A
Royster
2002
MWU
> 0.05
pesticide use
812
urine
MTHL2
C
Royster
2002
MWU
> 0.05
pesticide use
813
urine
MTHL2
A
Royster
2002
MWU
> 0.05
pesticide use
633
dust
AZM
C
McCauley
2001a
TNR
> 0.05
family use of pesticide
control products
farmworker and grower
homes in Hood River
County
848
dust
OPSUM
C
McCauley
2003
WTWS
0.39
family use of pesticide
control products
D-10
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q102--lnside Treated
563
urine
MDA
C
Sexton
2003
BSLR-5
0.1
In the past six months
were any chemicals for
the control of fleas,
roaches, ants, or other
insects used inside this
house/apartment?
unexpected direction of
coefficient
557
urine
MDA
C
Sexton
2003
LGRG
0.174
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
485
urine
TCPY
A
Krinsley
1998
SLR
0.93
pesticide use inside in
past 6 mo
164
urine
ETHL1
C
Lu 2001
MWU
0.27
pesticide use inside
focus children:
communities combined
average urine samples per child
165
urine
MTHL2
C
Lu 2001
MWU
0.35
pesticide use inside
focus children:
communities combined
average urine samples per child
561
dust
CHLR
L
Sexton
2003
LGRG
0.436
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable) ; child's play
area
558
indair
CHLR
C
Sexton
2003
LGRG
0.296
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
554
indair
MAL
C
Sexton
2003
LGRG
0.369
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
559
outdair
CHLR
C
Sexton
2003
LGRG
0.715
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
555
outdair
MAL
C
Sexton
2003
LGRG
0.373
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
D-ll
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
562
persair
CHLR
C
Sexton
2003
BSLR-1
0.04
In the past six months
were any chemicals for
the control of fleas,
roaches, ants, or other
insects used inside this
house/apartment?
child's breathing zone
unexpected direction of
coefficient
553
persair
MAL
C
Sexton
2003
LGRG
0.073
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable); child's
breathing zone
560
sldfood
CHLR
I
Sexton
2003
LGRG
0.38
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
556
sldfood
MAL
I
Sexton
2003
LGRG
0.06
Was there indoor
pesticide application in
past 6 months?
concentration (detectable /
nondetectable)
Q103--lnside Treated-Bath room
498
urine
TCPY
A
Krinsley
1998
SLR
0.36
pesticide used in
bathroom in past 6 mo
Q104--lnside Treated--Bedroom
767
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
whether subject applied
pesticides in the
bedroom
p < 0.00001
772
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
whether subject applied
pesticides in the
bedroom
used pesticides both
inside and outside,
personally or
professionally applied
p < 0.00001
497
urine
TCPY
A
Krinsley
1998
SLR
0.02
pesticide used in
bedroom in past 6 mo
Q105--lnside Treated-Cabinets
506
urine
TCPY
A
Krinsley
1998
SLR
0.15
pesticide used in
cabinets in past 6 mo
D-12
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q106--lnside Treated--Closets
507
urine
TCPY
A
Krinsley
1998
SLR
0.04
pesticide used in
closets in past 6 mo
Q107--lnside Treated-Cupboards with Dishes
505
urine
TCPY
A
Krinsley
1998
SLR
0.52
pesticide used in
cupboards with dishes
in past 6 mo
Q108--lnside Treated-Dining Room
496
urine
TCPY
A
Krinsley
1998
SLR
0.12
pesticide used in dining
room in past 6 mo
Q109--lnside Treated-Family Room
494
urine
TCPY
A
Krinsley
1998
SLR
0.38
pesticide used in family
room in past 6 mo
Q110--lnside Treated-Kitchen
493
urine
TCPY
A
Krinsley
1998
SLR
0.89
pesticide used in
kitchen in past 6 mo
Q111--lnside Treated-Living Room
495
urine
TCPY
A
Krinsley
1998
SLR
0.08
pesticide used in living
room in past 6 mo
Q112--lnside Treated-On Baseboards
501
urine
TCPY
A
Krinsley
1998
SLR
0.51
pesticide used on
baseboards in past 6
mo
Q113--lnside Treated-On Ceiling
504
urine
TCPY
A
Krinsley
1998
SLR
0.58
pesticide used on
ceiling in past 6 mo
Q114--lnside Treated-On Floor
500
urine
TCPY
A
Krinsley
1998
SLR
0.27
pesticide used on floor
in past 6 mo
D-13
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q115--lnside Treated--On Lower Walls
502
urine
TCPY
A
Krinsley
1998
SLR
0.65
pesticide used on lower
walls in past 6 mo
Q116--lnside Treated-On Upper Walls
503
urine
TCPY
A
Krinsley
1998
SLR
0.2
pesticide used on
upper walls in past 6
mo
Q117--lnside Treated-Other Room
773
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
whether the subject
used pesticides in other
room
used pesticides both
inside and outside,
personally or
professionally applied
p < 0.00001
499
urine
TCPY
A
Krinsley
1998
SLR
0.05
pesticide used in other
room in past 6 mo
Q118-Pets Treated
160
urine
ETHL1
C
Lu 2001
MWU
0.14
pesticide used on
household pets
focus children:
communities combined
average urine samples per child
335
urine
MTHL1
C
Lu 2000
MWU
0.6
Are household pets
treated with pesticides?
(applicator + farm-
worker) families'
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
161
urine
MTHL2
C
Lu 2001
MWU
0.8
pesticide used on
household pets
focus children:
communities combined
average urine samples per child
331
dust
AZMPH
C
Lu 2000
MWU
0.1
Are household pets
treated with pesticides?
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
Q119--Outside Treated
567
urine
MDA
C
Sexton
2003
BSLR-5
0.03
In the past six months
were any chemicals for
the control of fleas,
roaches, ants, or other
insects used on the
exterior of this
house/apartment?
unexpected direction of
coefficient
D-14
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
489
urine
TCPY
A
Krinsley
1998
SLR
0.11
pesticide use outside in
past 6 mo
570
urine
TCPY
C
Sexton
2003
BSLR-6
0.09
In the past six months
were any chemicals for
the control of fleas,
roaches, ants, or other
insects used on the
exterior of this
house/apartment?
unexpected direction of
coefficient
566
dust
CHLR
L
Sexton
2003
BSLR-4
0.01
In the past six months
were any chemicals for
the control of fleas,
roaches, ants, or other
insects used on the
exterior of this
house/apartment?
from child's play area
unexpected direction of
coefficient
565
sldfood
CHLR
I
Sexton
2003
BSLR-3
0.06
In the past six months
were any chemicals for
the control of fleas,
roaches, ants, or other
insects used on the
exterior of this
house/apartment?
unexpected direction of
coefficient
Q120--Garden Treated
249
urine
4NITR
C
Fenske
2002
MWU
> 0.05
OP pesticide use in
garden
focus children
average of visit 1 and visit 2
values
248
urine
TCPY
C
Fenske
2002
MWU
0.02
OP pesticide use in
garden
focus children
average of visit 1 and visit 2
values
156
urine
ETHL1
C
Lu 2001
MWU
0.02
garden pesticide used
focus children:
communities combined
average urine samples per child
338
urine
MTHL1
C
Lu 2000
MWU
0.9
Have you ever used
OPs in your garden?
(applicator + farm-
worker) families'
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
157
urine
MTHL2
C
Lu 2001
MWU
0.05
garden pesticide used
focus children:
communities combined
average urine samples per child
D-15
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
334
dust
AZMPH
C
Lu 2000
MWU
0.8
Have you ever used
OPs in your garden?
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
Q121~Lawn/Yard Treated
571
urine
TCPY
C
Sexton
2003
BSLR-6
0.09
In the past six months
have there been any
regular treatments by
anyone on the lawn or
yard outside of this
house/apartment?
unexpected direction of
coefficient
162
urine
ETHL1
c
Lu 2001
MWU
0.68
pesticide use on lawn
focus children:
communities combined
average urine samples per child
337
urine
MTHL1
c
Lu 2000
MWU
0.7
Has your lawn ever
been treated with OPs?
(applicator + farm-
worker) families'
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
163
urine
MTHL2
c
Lu 2001
MWU
0.13
pesticide use on lawn
focus children:
communities combined
average urine samples per child
333
dust
AZMPH
c
Lu 2000
MWU
0.7
Has your lawn ever
been treated with OPs?
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
Q122--lnside or Outside Treated
132
urine
1 NAP
c
Adgate
2001
WTAN
> 0.05
recent pesticide use
indoor or outdoor
weighted intra-child means
133
urine
MDA
c
Adgate
2001
WTAN
> 0.05
recent pesticide use
indoor or outdoor
weighted intra-child means
134
urine
TCPY
c
Adgate
2001
WTAN
> 0.05
recent pesticide use
indoor or outdoor
weighted intra-child means
71
urine
DEP
A
Aprea
2000
BDPH
> 0.05
use of pesticides inside
or outside
multiple comparison test w/1
independent variable
62
urine
DEP
A
Aprea
2000
MLR-5
> 0.05
use of pesticides inside
or outside
72
urine
DETP
A
Aprea
2000
BDPH
> 0.05
use of pesticides inside
or outside
multiple comparison test w/1
independent variable
D-16
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
73
urine
DEDTP
A
Aprea
2000
BDPH
> 0.05
use of pesticides inside
or outside
multiple comparison test w/1
independent variable
64
urine
DEDTP
A
Aprea
2000
MLR-7
> 0.05
use of pesticides inside
or outside
70
urine
DMP
A
Aprea
2000
BDPH
> 0.05
use of pesticides inside
or outside
multiple comparison test w/1
independent variable
61
urine
DMP
A
Aprea
2000
MLR-1
> 0.05
use of pesticides inside
or outside
67
urine
DMDTP
A
Aprea
2000
BDPH
> 0.05
use of pesticides inside
or outside
multiple comparison test w/1
independent variable
74
urine
ETHL2
A
Aprea
2000
BDPH
> 0.05
use of pesticides inside
or outside
multiple comparison test w/1
independent variable
65
urine
ETHL2
A
Aprea
2000
MLR-8
> 0.05
use of pesticides inside
or outside
Q123~Previous Treatment
336
urine
MTHL1
C
Lu 2000
MWU
0.6
Has your house been
treated with OPs since
January 1995?(within 6
months)
(applicator + farm-
worker) families'
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
332
dust
AZMPH
C
Lu 2000
MWU
0.3
Has your house been
treated with OPs since
January 1995?(within 6
months)
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
Q124--Level of Pesticide Use
136
urine
1 NAP
C
Adgate
2001
WTAN
> 0.05
level of pesticide use
weighted intra-child means
137
urine
MDA
C
Adgate
2001
WTAN
> 0.05
level of pesticide use
weighted intra-child means
D-17
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
551
urine
MDA
C
Sexton
2003
SLR
0.033
level of pesticide use
household pesticide-use
screening score was
subjectively assigned, to reflect
the household's potential for
pesticide exposure and was
based on questionnaire
responses and a pesticide
inventory
549
urine
MDA
C
Sexton
2003
WTWS
0.04
level of pesticide use
household pesticide-use
screening score was
subjectively assigned, to reflect
the household's potential for
pesticide exposure and was
based on questionnaire
responses and a pesticide
inventory
138
urine
TCPY
c
Adgate
2001
WTAN
> 0.05
level of pesticide use
weighted intra-child means
762
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
pesticide use index
(PUI)
PUI was constructed from
pesticide use variables: p <
0.00001
476
urine
TCPY
A
Krinsley
1998
SLR
< 0.004
pesticide use index
(PUI)
PUI was constructed from
pesticide use variables: p <
0.0042
550
persair
ATZ
C
Sexton
2003
LGRG
0.028
level of pesticide use
concentration (detectable /
nondetectable); child's
breathing zone
household pesticide-use
screening score was
subjectively assigned, to reflect
the household's potential for
pesticide exposure and was
based on questionnaire
responses and a pesticide
inventory
D-18
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
552
persair
ATI
C
Sexton
2003
LGRG
0.02
level of pesticide use
child's breathing zone
household pesticide-use
screening score was
subjectively assigned, to reflect
the household's potential for
pesticide exposure and was
based on questionnaire
responses and a pesticide
inventory
Q125--Frequency Personal Application Inside
486
urine
TCPY
A
Krinsley
1998
SLR
0.07
number of times
personally applied
pesticide inside in past
6 mo
pesticide users
Q126--Frequency Personal Application Outside
766
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
number of times the
subject personally
applied pesticides
outside in past 6 mo
p < 0.00001
771
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
number of times the
subject personally
applied pesticides
outside in past 6 mo
used pesticides both
inside and outside,
personally or
professionally applied
p < 0.00001
490
urine
TCPY
A
Krinsley
1998
SLR
0.003
number of times
personally applied
pesticide outside in
past 6 mo
pesticide users
Q127--lnside/Outside Treated by Family Member
Z / O
urine
ETHL3
C
Azaroff
1999
MLGR-6
< 0.05
OP applied in house or
yard by household
mother
detectable ethylated AP
metabolites - 3 samples
combined - second model
591
urine
MTHL3
C
Azaroff
1999
MLGR-7
< 0.01
methamidophos
applied in house or
yard by household
mother
detectable levels of DMTP - 3
samples combined
controlled for fieldwork -
variable included as predictor
D-19
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
276
urine
MTHL4
C
Azaroff
1999
MLGR-3
< 0.01
methamidophos
applied in house or
yard by household
mother
detectable methylated AP
metabolites - 3 samples
combined
273
urine
DAP2
C
Azaroff
1999
MLGR-1
< 0.05
OP applied in house or
yard by household
mother
detectable AP metabolites - 3
samples combined
274
urine
DAP3
c
Azaroff
1999
MLGR-2
< 0.10
OP applied in house or
yard by household
mother
high or very high level of an AP
metabolite - 3 samples
combined
Q128--Frequency Professional Application Inside
487
urine
TCPY
A
Krinsley
1998
SLR
0.62
number of times
professionally applied
pesticide inside in past
6 mo
pesticide users
Q129--Frequency Professional Application Outside
491
urine
TCPY
A
Krinsley
1998
SLR
0.96
number of times
professionally applied
pesticide outside in
past 6 mo
pesticide users
Q130--Personally Mixed Pesticide Inside
488
urine
TCPY
A
Krinsley
1998
SLR
0.18
personally mixed inside
pesticide in past 6 mo
pesticide users
Q131~Personally Mixed Pesticide Outside
492
urine
TCPY
A
Krinsley
1998
SLR
0.46
personally mixed
outside pesticide in
past 6 mo
Q132~Presence During Mixing
680
urine
1 NAP
C
Adgate
2001
WTAN
> 0.05
child present during
pesticide mixing
children with 3 urine
samples
weighted intra-child means
683
urine
ATZM
C
Adgate
2001
NAN
> 0.05
child present during
pesticide mixing
children with 3 urine
samples
weighted intra-child means
D-20
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
681
urine
MDA
C
Adgate
2001
WTAN
> 0.05
child present during
pesticide mixing
children with 3 urine
samples
weighted intra-child means
682
urine
TCPY
C
Adgate
2001
WTAN
> 0.05
child present during
pesticide mixing
children with 3 urine
samples
weighted intra-child means
a See section 4.2.2.2 and the paragraph immediately following Table 4.2.3 regarding relationships from Sexton (2003).
D-21
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.1.2 Category 2: Household Characteristics
Table D.3.1.2 Relationship Comments for Questions in Category 2: Household Characteristics - Grouped by Question and Sorted by Medium, Chemical, Citation
and Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q201--Housing Type
168
urine
ETHL1
C
Lu 2001
TNR
> 0.05
type of housing
focus children:
communities combined
average urine samples per child
169
urine
MTHL2
C
Lu 2001
TNR
> 0.05
type of housing
focus children:
communities combined
average urine samples per child
632
dust
AZM
c
McCauley
2001a
TNR
> 0.05
housing type
farmworker and grower
homes in Hood River
County
Q202--Property Used as a Farm
690
dust
CHLR
L
Sexton
2003
BSLR-4
0.06
Is this property used as
a farm?
unexpected direction of
coefficient
564
indair
CHLR
C
Sexton
2003
BSLR-2
0.01
Is this property used as
a farm?
unexpected direction of
coefficient
Q203--Age of House > 10 Years
483
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
age of house
Q204--Age of House > 20 Years
670
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
age of house
Q205--Having Air Conditioning
480
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
having air conditioning
D-22
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q206--Having Central Heating
482
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
having central heating
Q207--Having Evaporative Cooling
481
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
having evaporative
cooling
Q208--Pets in House
53
urine
DEP
A
Aprea
2000
BDPH
> 0.05
domestic animals in
house
multiple comparison test w/1
independent variable
44
urine
DEP
A
Aprea
2000
MLR-5
> 0.05
domestic animals in
house
54
urine
DETP
A
Aprea
2000
BDPH
> 0.05
domestic animals in
house
multiple comparison test w/1
independent variable
45
urine
DETP
A
Aprea
2000
MLR-6
> 0.05
domestic animals in
house
55
urine
DEDTP
A
Aprea
2000
BDPH
> 0.05
domestic animals in
house
multiple comparison test w/1
independent variable
46
urine
DEDTP
A
Aprea
2000
MLR-7
> 0.05
domestic animals in
house
52
urine
DMP
A
Aprea
2000
BDPH
> 0.05
domestic animals in
house
multiple comparison test w/1
independent variable
43
urine
DMP
A
Aprea
2000
MLR-1
> 0.05
domestic animals in
house
48
urine
DMTP
A
Aprea
2000
BDPH
> 0.05
domestic animals in
house
multiple comparison test w/1
independent variable
49
urine
DMDTP
A
Aprea
2000
BDPH
> 0.05
domestic animals in
house
multiple comparison test w/1
independent variable
158
urine
ETHL1
C
Lu 2001
MWU
0.4
pets in household
focus children:
communities combined
average urine samples per child
D-23
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
56
urine
ETHL2
A
Aprea
2000
BDPH
> 0.05
domestic animals in
house
multiple comparison test w/1
independent variable
47
urine
ETHL2
A
Aprea
2000
MLR-8
> 0.05
domestic animals in
house
50
urine
MTHL2
A
Aprea
2000
BDPH
> 0.05
domestic animals in
house
multiple comparison test w/1
independent variable
159
urine
MTHL2
C
Lu 2001
MWU
0.04
pets in household
focus children:
communities combined
average urine samples per child
51
urine
DAP1
A
Aprea
2000
BDPH
> 0.05
domestic animals in
house
multiple comparison test w/1
independent variable
Q209--Pets Inside/Outside House
569
urine
MDA
C
Sexton
2003
BSLR-5
0.08
Do you have pets such
as dogs, cats, gerbils,
hamsters, rabbits,
guinea pigs, birds, or
horses?
unexpected direction of
coefficient
Q210--Pet Inside to Outside
736
dust
AZM
L
Simcox
1995
MLR-3
> 0.05
Is there a pet that goes
in and out of the
house?
dust: 2 samples pooled
436
dust
AZM
L
Simcox
1995
MWU
> 0.05
Is there a pet that goes
in and out of the
house?
dust: 2 samples pooled
438
dust
CHLR
L
Simcox
1995
MWU
> 0.05
Is there a pet that goes
in and out of the
house?
dust: 2 samples pooled
738
dust
CHLR
L
Simcox
1995
MWU
> 0.05
Is there a pet that goes
in and out of the
house?
dust: 2 samples pooled
439
dust
EPAR
L
Simcox
1995
MWU
> 0.05
Is there a pet that goes
in and out of the
house?
dust: 2 samples pooled
D-24
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
739
dust
EPAR
L
Simcox
1995
MWU
> 0.05
Is there a pet that goes
in and out of the
house?
dust: 2 samples pooled
737
dust
PHSM
L
Simcox
1995
MLR-4
> 0.05
Is there a pet that goes
in and out of the
house?
dust: 2 samples pooled
437
dust
PHSM
L
Simcox
1995
MWU
> 0.05
Is there a pet that goes
in and out of the
house?
dust: 2 samples pooled
Q211-Existence of Garden or Vegetable Garden
568
urine
MDA
C
Sexton
2003
BSLR-5
0.04
Do you have a flower,
vegetable, or fruit
garden to which you
apply chemicals?
17
urine
DEP
A
Aprea
2000
BDPH
> 0.05
existence of garden or
vegetable garden
multiple comparison test w/1
independent variable
8
urine
DEP
A
Aprea
2000
MLR-5
> 0.05
existence of garden or
vegetable garden
18
urine
DETP
A
Aprea
2000
BDPH
> 0.05
existence of garden or
vegetable garden
multiple comparison test w/1
independent variable
9
urine
DETP
A
Aprea
2000
MLR-6
> 0.05
existence of garden or
vegetable garden
19
urine
DEDTP
A
Aprea
2000
BDPH
> 0.05
existence of garden or
vegetable garden
multiple comparison test w/1
independent variable
10
urine
DEDTP
A
Aprea
2000
MLR-7
> 0.05
existence of garden or
vegetable garden
16
urine
DMP
A
Aprea
2000
BDPH
> 0.05
existence of garden or
vegetable garden
multiple comparison test w/1
independent variable
7
urine
DMP
A
Aprea
2000
MLR-1
> 0.05
existence of garden or
vegetable garden
12
urine
DMTP
A
Aprea
2000
BDPH
> 0.05
existence of garden or
vegetable garden
multiple comparison test w/1
independent variable
D-25
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
154
urine
ETHL1
C
Lu 2001
MWU
0.04
garden exists
focus children:
communities combined
average urine samples per child
20
urine
ETHL2
A
Aprea
2000
BDPH
> 0.05
existence of garden or
vegetable garden
multiple comparison test w/1
independent variable
11
urine
ETHL2
A
Aprea
2000
MLR-8
> 0.05
existence of garden or
vegetable garden
14
urine
MTHL2
A
Aprea
2000
BDPH
> 0.05
existence of garden or
vegetable garden
multiple comparison test w/1
independent variable
155
urine
MTHL2
C
Lu 2001
MWU
0.11
garden exists
focus children:
communities combined
average urine samples per child
15
urine
DAP1
A
Aprea
2000
BDPH
> 0.05
existence of garden or
vegetable garden
multiple comparison test w/1
independent variable
Q212--Ornamental Plants or Cut Flowers
35
urine
DEP
A
Aprea
2000
BDPH
> 0.05
ornamental plants or
cut flowers in house
multiple comparison test w/1
independent variable
26
urine
DEP
A
Aprea
2000
MLR-5
> 0.05
ornamental plants or
cut flowers in house
36
urine
DETP
A
Aprea
2000
BDPH
> 0.05
ornamental plants or
cut flowers in house
multiple comparison test w/1
independent variable
27
urine
DETP
A
Aprea
2000
MLR-6
> 0.05
ornamental plants or
cut flowers in house
37
urine
DEDTP
A
Aprea
2000
BDPH
> 0.05
ornamental plants or
cut flowers in house
multiple comparison test w/1
independent variable
28
urine
DEDTP
A
Aprea
2000
MLR-7
> 0.05
ornamental plants or
cut flowers in house
34
urine
DMP
A
Aprea
2000
BDPH
> 0.05
ornamental plants or
cut flowers in house
multiple comparison test w/1
independent variable
25
urine
DMP
A
Aprea
2000
MLR-1
> 0.05
ornamental plants or
cut flowers in house
30
urine
DMTP
A
Aprea
2000
BDPH
> 0.05
ornamental plants or
cut flowers in house
multiple comparison test w/1
independent variable
D-26
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
31
urine
DMDTP
A
Aprea
2000
BDPH
> 0.05
ornamental plants or
cut flowers in house
multiple comparison test w/1
independent variable
38
urine
ETHL2
A
Aprea
2000
BDPH
> 0.05
ornamental plants or
cut flowers in house
multiple comparison test w/1
independent variable
29
urine
ETHL2
A
Aprea
2000
MLR-8
> 0.05
ornamental plants or
cut flowers in house
32
urine
MTHL2
A
Aprea
2000
BDPH
> 0.05
ornamental plants or
cut flowers in house
multiple comparison test w/1
independent variable
33
urine
DAP1
A
Aprea
2000
BDPH
> 0.05
ornamental plants or
cut flowers in house
multiple comparison test w/1
independent variable
Q213--Size of Household
631
dust
AZM
C
McCauley
2001a
SLR
> 0.05
number of persons
living in household
homes in Hood River
County with detectable
dust samples
assumed not significant based
on comments in article
847
dust
OPSUM
C
McCauley
2003
CORR
0.29
number of individuals in
household
r = 0.22
Q214--Location of Play Area
832
dust
OPSUM
C
McCauley
2003
WTWS
0.66
location of child's play
area
Q215--Age of House (years)
843
dust
AZM
C
McCauley
2003
CORR
0.22
age of house (years)
r = 0.26
842
dust
OPSUM
C
McCauley
2003
CORR
0.25
age of house (years)
r = 0.25
Q216--Size of House (sq ft)
844
dust
OPSUM
C
McCauley
2003
CORR
0.08
size of house (sq ft)
r = 0.40
845
dust
OPSUM
C
McCauley
2003
MLR
0.16
size of house (sq ft)
partial correlation = 0.31;
analysis adjusted for age of
house
D-27
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation9
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q217~Number of Pets in House
849
dust
OPSUM
C
McCauley
2003
CORR
0.95
number of cats and
dogs living in house
r = 0.02
a See section 4.2.2.2 and the paragraph immediately following Table 4.2.3 regarding relationships from Sexton (2003).
D-28
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.1.3 Category 3: Residential Sources (Environmental Measures)
Table D.3.1.3 Relationship Comments for Questions in Category 3: Residential Sources (Environmental Measures) - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q301--Household Dust
147
urine
MTHL2
C
Curl 2002
SLR
< 0.001
household dust level of
azinphosmethyl
p < 0.0001
586
urine
MTHL2
A
Curl 2002
SLR
< 0.001
household dust level of
azinphosmethyl
328
urine
NA
C
Lu 2000
SPCR
< 0.10
measurement
(applicator + farm-
worker) families'
children
average of visit 1 and visit 2
values for urine; dust and urine
adjusted by extraction
efficiencies
dust:
azinophosmethyl+phosmet
329
urine
NA
C
Lu 2000
SPCR
0.09
measurement
all families' children
average of visit 1 and visit 2
values for urine; dust and urine
adjusted by extraction
efficiencies
dust:
azinophosmethyl+phosmet
Q302--Loading from Household Floor Dust
644
urine
DAP1
A
Shalat
2003
MVRG-2
> 0.05
household dust load
p = 0.076 for model;
significance level assumed
based on MVRG-1and
comments in article
Q303--Outdoor Soil
403
dust
AZM
C
Simcox
1995
SPCR
0.001
measurement
(farmer + farm-worker)
families
dust: 2 samples pooled; soil: 5
sample composite
r = 0.49
407
dust
AZM
C
Simcox
1995
SPCR
0.87
measurement
reference families
dust: 2 samples pooled; soil: 5
sample composite
r = 0.05
405
dust
CHLR
C
Simcox
1995
SPCR
< 0.001
measurement
(farmer + farm-worker)
families
dust: 2 samples pooled; soil: 5
sample composite
r = 0.52: p = 0.0003
D-29
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
409
dust
CHLR
C
Simcox
1995
SPCR
0.21
measurement
reference families
dust: 2 samples pooled; soil: 5
sample composite
r = 0.4
406
dust
EPAR
C
Simcox
1995
SPCR
0.02
measurement
(farmer + farm-worker)
families
dust: 2 samples pooled; soil: 5
sample composite
r = 0.35
410
dust
EPAR
c
Simcox
1995
SPCR
0.01
measurement
reference families
dust: 2 samples pooled; soil: 5
sample composite
r = 0.81
404
dust
PHSM
c
Simcox
1995
SPCR
< 0.001
measurement
(farmer + farm-worker)
families
dust: 2 samples pooled; soil: 5
sample composite
r = 0.67; p < 0.0001
408
dust
PHSM
c
Simcox
1995
SPCR
0.48
measurement
reference families
dust: 2 samples pooled; soil: 5
sample composite
r = 0.23
D-30
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.1.4 Category 4: Household Occupation
Table D.3.1.4 Relationship Comments for Questions in Category 4: Household Occupation - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q401--Agricultural Workers in Household
572
dust
AZM
C
McCauley
2001a
SLR
0.002
number of agricultural
workers residing in
home
homes in Hood River
County with detectable
dust samples
634
dust
AZM
C
McCauley
2001a
SLR
< 0.05
number of agricultural
workers residing in
home
homes in Hood River
County
assumed significant based on
comments in article in
comparison to analysis with only
detectable samples
Q402--Household Member Spraying Fields
282
urine
ETHL3
c
Azaroff
1999
MLGR-5
< 0.05
OP other than
parathion,
methamidophos,
phoxim applied to fields
during past year by
head household farmer
detectable ethylated AP
metabolites - 3 samples
combined
283
urine
ETHL3
c
Azaroff
1999
MLGR-6
< 0.10
OP other than
parathion,
methamidophos,
phoxim applied to fields
during past year by
head household farmer
detectable ethylated AP
metabolites - 3 samples
combined
590
urine
MTHL3
c
Azaroff
1999
MLGR-7
< 0.01
methyl parathion
applied to fields within
past year by head
household farmer
detectable levels of DMTP - 3
samples combined
controlled for fieldwork -
variable included as predictor
279
urine
MTHL4
c
Azaroff
1999
MLGR-3
< 0.01
methyl parathion
applied to fields within
past year by head
household farmer
detectable methylated AP
metabolites - 3 samples
combined
D-31
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
281
urine
MTHL4
C
Azaroff
1999
MLGR-3
< 0.05
malathion or omethoate
applied to fields within
past year by head
household farmer
detectable methylated AP
metabolites - 3 samples
combined
277
urine
DAP2
C
Azaroff
1999
MLGR-1
< 0.05
OP applied to fields
within past year by
head household farmer
detectable AP metabolites - 3
samples combined
280
urine
MTHL5
c
Azaroff
1999
MLGR-4
< 0.01
methyl parathion
applied to fields within
past year by head
household farmer
high or very high level of a
methylated AP metabolite - 3
samples combined
278
urine
DAP3
c
Azaroff
1999
MLGR-2
< 0.10
OP applied to fields
within past year by
head household farmer
high or very high level of an AP
metabolite - 3 samples
combined
Q403--Recent Fieldwork
271
urine
ETHL3
c
Azaroff
1999
MLGR-5
> 0.10
reporting fieldwork
within past 2 weeks
detectable ethylated AP
metabolites - 3 samples
combined
272
urine
ETHL3
c
Azaroff
1999
MLGR-6
> 0.10
reporting fieldwork
within past 2 weeks
detectable ethylated AP
metabolites - 3 samples
combined
269
urine
MTHL4
c
Azaroff
1999
MLGR-3
< 0.01
reporting fieldwork
within past 2 weeks
detectable methylated AP
metabolites - 3 samples
combined
267
urine
DAP2
c
Azaroff
1999
MLGR-1
< 0.01
reporting fieldwork
within past 2 weeks
detectable AP metabolites - 3
samples combined
270
urine
MTHL5
c
Azaroff
1999
MLGR-4
< 0.05
reporting fieldwork
within past 2 weeks
high or very high level of a
methylated AP metabolite - 3
samples combined
268
urine
DAP3
c
Azaroff
1999
MLGR-2
< 0.05
reporting fieldwork
within past 2 weeks
high or very high level of an AP
metabolite - 3 samples
combined
D-32
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q404--Applicator vs Farmworker
239
urine
4NITR
C
Fenske
2002
MWU
> 0.05
household occupation
focus children
average of visit 1 and visit 2
values
238
urine
TCPY
C
Fenske
2002
MWU
> 0.05
household occupation
focus children
average of visit 1 and visit 2
values
319
urine
DMTP
c
Lu 2000
MWU
>= 0.10
household occupation
focus children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
320
urine
DMDTP
c
Lu 2000
MWU
>= 0.10
household occupation
focus children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
321
urine
MTHL1
c
Lu 2000
MWU
>= 0.10
household occupation
focus children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
316
dust
AZM
c
Lu 2000
MWU
>= 0.10
household occupation
adjusted by extraction
efficiencies
318
dust
AZMPH
c
Lu 2000
MWU
0.07
household occupation
adjusted by extraction
efficiencies
232
dust
CHLR
c
Fenske
2002
MWU
> 0.05
household occupation
adjusted by extraction
efficiencies
235
dust
EPAR
c
Fenske
2002
MWU
0.03
household occupation
adjusted by extraction
efficiencies
317
dust
PHSM
c
Lu 2000
MWU
>= 0.10
household occupation
adjusted by extraction
efficiencies
Q405--Applicator vs Non-applicator
387
dust
AZM
c
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
391
dust
AZM
L
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
425
dust
AZM
C
Simcox
1995
OWAN
> 0.05
pesticide application
activity classification
(farmer + farm-worker)
families
dust: 2 samples pooled
D-33
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
389
dust
CHLR
C
Simcox
1995
MWU
0.02
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
393
dust
CHLR
L
Simcox
1995
MWU
0.04
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
427
dust
CHLR
C
Simcox
1995
OWAN
> 0.05
pesticide application
activity classification
(farmer + farm-worker)
families
dust: 2 samples pooled
390
dust
EPAR
C
Simcox
1995
MWU
< 0.001
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
p = 0.0003
394
dust
EPAR
L
Simcox
1995
MWU
0.002
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
428
dust
EPAR
C
Simcox
1995
OWAN
0.001
pesticide application
activity classification
(farmer + farm-worker)
families
dust: 2 samples pooled
429
dust
EPAR
C
Simcox
1995
TWAN-1
0.002
pesticide application
activity classification
(farmer + farm-worker)
families
dust: 2 samples pooled
no significant interaction
between proximity to orchards
and applicator status
430
dust
EPAR
C
Simcox
1995
TWAN-2
> 0.05
pesticide application
activity classification
(farmer + farm-worker)
families
dust: 2 samples pooled
significant interaction between
occupation (farmer vs
farmworker) and applicator
status, thus signficance level
for applicator status assigned as
NS
388
dust
PHSM
C
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
392
dust
PHSM
L
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
426
dust
PHSM
C
Simcox
1995
OWAN
> 0.05
pesticide application
activity classification
(farmer + farm-worker)
families
dust: 2 samples pooled
Q406--Applicator vs Reference
201
urine
DMTP
C
Loewen-
herz 1997
CHSQ
> 0.10
household occupation
focus applicator
children; visit 1
frequency of detectability
202
urine
DMTP
C
Loewen-
herz 1997
CHSQ
0.022
household occupation
focus applicator
children; visit 2
frequency of detectability
D-34
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
176
urine
DMTP
C
Loewen-
herz 1997
MWU
> 0.10
household occupation
focus applicator
children; visit 1
177
urine
DMTP
C
Loewen-
herz 1997
MWU
0.036
household occupation
focus applicator
children; visit 2
178
urine
DMTP
c
Loewen-
herz 1997
MWU
0.015
household occupation
focus applicator
children; visits
combined
198
urine
DMTP
c
Loewen-
herz 1997
MWU
> 0.10
household occupation
focus applicator
children; visit 1
199
urine
DMTP
c
Loewen-
herz 1997
MWU
0.022
household occupation
focus applicator
children; visit 2
200
urine
DMTP
A
Loewen-
herz 1997
MWU
0.011
household occupation
focus applicator
children; visits
combined
Q407~Applicator+Farmworker vs Reference
241
urine
4NITR
C
Fenske
2002
KWAN
> 0.05
household occupation
focus children
average of visit 1 and visit 2
values
231
urine
4NITR
C
Fenske
2002
TNR
> 0.05
household occupations
focus children
average of visit 1 and visit 2
values
240
urine
TCPY
C
Fenske
2002
KWAN
> 0.05
household occupation
focus children
average of visit 1 and visit 2
values
230
urine
TCPY
C
Fenske
2002
TNR
> 0.05
household occupations
focus children
average of visit 1 and visit 2
values
325
urine
DMTP
C
Lu 2000
MWU
0.07
household occupation
focus children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
326
urine
DMDTP
C
Lu 2000
MWU
>= 0.10
household occupation
focus children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
327
urine
MTHL1
C
Lu 2000
MWU
0.09
household occupation
focus children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
D-35
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
322
dust
AZM
C
Lu 2000
MWU
< 0.001
household occupation
household dust, adjusted by
extraction efficiencies
324
dust
AZMPH
C
Lu 2000
MWU
< 0.001
household occupation
household dust, adjusted by
extraction efficiencies
233
dust
CHLR
c
Fenske
2002
KWAN
< 0.001
household occupation
adjusted by extraction
efficiencies
234
dust
CHLR
c
Fenske
2002
MWU
< 0.001
household occupation
adjusted by extraction
efficiencies
244
dust
CHLR
c
Fenske
2002
MWU
< 0.01
household occupation
all reference families,
and applicator and
farworker families with
distance > 0.25 mi -
home to pesticide-
treated farmland
adjusted by extraction
efficiencies
236
dust
EPAR
c
Fenske
2002
KWAN
< 0.01
household occupation
adjusted by extraction
efficiencies
237
dust
EPAR
c
Fenske
2002
MWU
0.02
household occupation
adjusted by extraction
efficiencies
245
dust
EPAR
c
Fenske
2002
MWU
> 0.05
household occupation
all reference families,
and applicator and
farworker families with
distance > 0.25 mi -
home to pesticide-
treated farmland
adjusted by extraction
efficiencies
323
dust
PHSM
c
Lu 2000
MWU
0.02
household occupation
household dust, adjusted by
extraction efficiencies
Q408~Farmer vs Farmworker
379
dust
AZM
c
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
383
dust
AZM
L
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
650
dust
AZM
C
Simcox
1995
OWAN
> 0.05
occupational
classification
(farmer + farm-worker)
families
dust: 2 samples pooled
D-36
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
381
dust
CHLR
C
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
385
dust
CHLR
L
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
654
dust
CHLR
C
Simcox
1995
OWAN
> 0.05
occupational
classification
(farmer + farm-worker)
families
dust: 2 samples pooled
382
dust
EPAR
C
Simcox
1995
MWU
< 0.001
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
p = 0.0007
386
dust
EPAR
L
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
656
dust
EPAR
C
Simcox
1995
OWAN
0.001
occupational
classification
(farmer + farm-worker)
families
dust: 2 samples pooled
659
dust
EPAR
C
Simcox
1995
TWAN-2
> 0.05
farm occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
significant interaction between
occupation (farmer vs
farmworker) and applicator
status thus signficance level for
occupation assigned as NS
431
dust
EPAR
C
Simcox
1995
TWAN-3
> 0.05
farm occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
significant interaction between
occupation (farmer vs
farmworker) and proximity to
orchards thus significance level
for occupation assigned as NS
380
dust
PHSM
C
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
384
dust
PHSM
L
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
dust: 2 samples pooled
652
dust
PHSM
C
Simcox
1995
OWAN
> 0.05
occupational
classification
(farmer + farm-worker)
families
dust: 2 samples pooled
367
soil
AZM
C
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
soil: 5 sample composite
369
soil
CHLR
C
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
soil: 5 sample composite
D-37
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
370
soil
EPAR
C
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
soil: 5 sample composite
368
soil
PHSM
C
Simcox
1995
MWU
> 0.05
household occupation
(farmer + farm-worker)
families
soil: 5 sample composite
Q409~Farmer+Farmworker vs Reference
371
dust
AZM
c
Simcox
1995
MWU
0.001
household occupation
dust: 2 samples pooled
375
dust
AZM
L
Simcox
1995
MWU
> 0.05
household occupation
dust: 2 samples pooled
373
dust
CHLR
C
Simcox
1995
MWU
0.01
household occupation
dust: 2 samples pooled
377
dust
CHLR
L
Simcox
1995
MWU
> 0.05
household occupation
dust: 2 samples pooled
374
dust
EPAR
C
Simcox
1995
MWU
0.02
household occupation
dust: 2 samples pooled
378
dust
EPAR
L
Simcox
1995
MWU
> 0.05
household occupation
dust: 2 samples pooled
372
dust
PHSM
C
Simcox
1995
MWU
0.07
household occupation
dust: 2 samples pooled
376
dust
PHSM
L
Simcox
1995
MWU
> 0.05
household occupation
dust: 2 samples pooled
363
soil
AZM
C
Simcox
1995
MWU
0.04
household occupation
soil: 5 sample composite
365
soil
CHLR
C
Simcox
1995
MWU
> 0.05
household occupation
soil: 5 sample composite
366
soil
EPAR
C
Simcox
1995
MWU
> 0.05
household occupation
soil: 5 sample composite
364
soil
PHSM
C
Simcox
1995
MWU
> 0.05
household occupation
soil: 5 sample composite
D-38
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q410~Farmworker vs Grower
575
dust
AZM
C
McCauley
2001a
WTWS
0.02
open areas in
farmworker homes vs.
play areas in grower
homes
farmworker and grower
homes in Hood River
County
630
dust
AZM
C
McCauley
2001a
WTWS
> 0.05
open areas in
farmworker homes vs.
entry areas in grower
homes
farmworker and grower
homes in Hood River
County
Q411-Farmworker vs Others
289
urine
ETHL2
c
Koch
2002
GLM
> 0.05
household occupations
Includes multiple samples per
child
291
urine
ETHL2
c
Koch
2002
GLM
> 0.05
household occupations
samples from spray
months in 1998
Includes multiple samples per
child
288
urine
MTHL2
c
Koch
2002
GLM
> 0.05
household occupations
Includes multiple samples per
child
290
urine
MTHL2
c
Koch
2002
GLM
> 0.05
household occupations
samples from spray
months in 1998
Includes multiple samples per
child
Q412--Field Worker vs Pesticide Handler
453
dust
AZM
L
Grossman
2001
MLR-7
0.011
occupational category
farm-worker with high
expected exposure
analysis adjusted for
respondents' residential
proximity to treated field or
orchard
Q413--Expected Occupational Exposure
605
urine
ETHL2
C
Koch
1999
KWAN
0.878
expected occupational
exposure
median excretion value per
child
607
urine
ETHL2
C
Koch
1999
KWAN
0.351
expected occupational
exposure
samples from spray
months in 1988
median excretion value per
child
609
urine
ETHL2
C
Koch
1999
KWAN
0.85
expected occupational
exposure
samples from non-
spray months in 1988
median excretion value per
child
D-39
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
604
urine
MTHL2
C
Koch
1999
KWAN
0.93
expected occupational
exposure
median excretion value per
child
606
urine
MTHL2
C
Koch
1999
KWAN
0.851
expected occupational
exposure
samples from spray
months in 1988
median excretion value per
child
608
urine
MTHL2
c
Koch
1999
KWAN
0.387
expected occupational
exposure
samples from non-
spray months in 1988
median excretion value per
child
448
dust
AZM
L
Grossman
2001
MLR-6
< 0.001
expected occupational
exposure
farm-worker
analyses adjusted for
respondents' educational status;
p < 0.0001
449
dust
AZM
L
Grossman
2001
MLR-6
0.084
expected occupational
exposure
farm-worker
analyses adjusted for
respondents' educational status
Q414--Occupational Pesticide Exposure
810
urine
ETHL2
C
Royster
2002
MWU
> 0.05
occupational pesticide
exposure
811
urine
ETHL2
A
Royster
2002
MWU
> 0.05
occupational pesticide
exposure
808
urine
MTHL2
C
Royster
2002
MWU
> 0.05
occupational pesticide
exposure
809
urine
MTHL2
A
Royster
2002
MWU
> 0.05
occupational pesticide
exposure
Q415~Tree Thinning
831
dust
OPSUM
C
McCauley
2003
WTWS
0.06
occupation as tree
thinner
play area; carpet, rug covering
or bare floor samples
Q416--Number with High Contact Exposure in Household
830
dust
OPSUM
C
McCauley
2003
WTWS
0.007
household members
with high pesticide
contact jobs
households where
member(s) have high
pesticide contact jobs
play area; carpet, rug covering
or bare floor samples
D-40
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.1.5 Category 5: Residential Proximity to Agricultural Fields
Table D.3.1.5 Relationship Comments for Questions in Category 5: Residential Proximity to Agricultural Fields - Grouped by Question and Sorted by Medium,
Chemical, Citation and Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q501~Proximity of Home to Pesticide-Treated Farmland/Orchard
257
urine
4NITR
C
Fenske
2002
MWU
> 0.05
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families - focus
children
average of visit 1 and visit 2
values
256
urine
TCPY
C
Fenske
2002
MWU
> 0.05
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families - focus
children
average of visit 1 and visit 2
values
204
urine
DMTP
c
Loewen-
herz 1997
FISH
0.036
distance - home to
sprayed field
focus applicator
children; visit 2
counts of detects, traces, or
non-detects
193
urine
DMTP
c
Loewen-
herz 1997
MWU
> 0.10
distance - home to
sprayed field
focus applicator
children; visit 1
194
urine
DMTP
c
Loewen-
herz 1997
MWU
0.062
distance - home to
sprayed field
focus applicator
children; visit 2
342
urine
DMTP
c
Lu 2000
MWU
0.009
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families - focus
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
343
urine
DMDTP
c
Lu 2000
MWU
>= 0.10
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families - focus
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
299
urine
ETHL2
c
Koch
2002
GLM
> 0.05
residential proximity to
a fruit tree orchard
Includes multiple samples per
child
301
urine
ETHL2
c
Koch
2002
GLM
> 0.05
residential proximity to
a fruit tree orchard
samples from spray
months in 1998
Includes multiple samples per
child
804
urine
ETHL2
A
Royster
2002
MWU
> 0.05
proximity of home to
closest agricultural field
visit 1, GPS/GIS
measure of distance
D-41
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
805
urine
ETHL2
A
Royster
2002
MWU
> 0.05
proximity of home to
closest agricultural field
visit 1, GPS/GIS
measure of distance
806
urine
ETHL2
A
Royster
2002
MWU
> 0.05
proximity of home to
closest agricultural field
visit 2, GPS/GIS
measure of distance
807
urine
ETHL2
A
Royster
2002
MWU
> 0.05
proximity of home to
closest agricultural field
visit 2, GPS/GIS
measure of distance
344
urine
MTHL1
C
Lu 2000
MWU
0.01
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families - focus
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
346
urine
MTHL1
C
Lu 2000
SLR
0.1
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families'
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
distance represented by
categories: <50 ft, 50-200 ft,
200 ft-0.25 mi, >0.25 mi
348
urine
MTHL1
C
Lu 2000
SLR
0.06
distance - home to
pesticide-treated
farmland
all families' children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
distance represented by
categories: <50 ft, 50-200 ft,
200 ft-0.25 mi, >0.25 mi,
reference family
140
urine
MTHL2
C
Curl 2002
OWAN
0.34
distance - home to
sprayed field-orig
categories
142
urine
MTHL2
C
Curl 2002
OWAN
0.3
distance - home to
sprayed field-rev
categories
584
urine
MTHL2
A
Curl 2002
OWAN
0.3
distance - home to
sprayed field-orig
categories
585
urine
MTHL2
A
Curl 2002
OWAN
0.4
distance - home to
sprayed field-rev
categories
298
urine
MTHL2
C
Koch
2002
GLM
> 0.05
residential proximity to
a fruit tree orchard
Includes multiple samples per
child
300
urine
MTHL2
C
Koch
2002
GLM
> 0.05
residential proximity to
a fruit tree orchard
samples from spray
months in 1998
Includes multiple samples per
child
D-42
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
800
urine
MTHL2
A
Royster
2002
MWU
> 0.05
proximity of home to
closest agricultural field
visit 1, GPS/GIS
measure of distance
801
urine
MTHL2
A
Royster
2002
MWU
> 0.05
proximity of home to
closest agricultural field
visit 1, GPS/GIS
measure of distance
802
urine
MTHL2
A
Royster
2002
MWU
> 0.05
proximity of home to
closest agricultural field
visit 2, GPS/GIS
measure of distance
803
urine
MTHL2
A
Royster
2002
MWU
> 0.05
proximity of home to
closest agricultural field
visit 2, GPS/GIS
measure of distance
141
dust
AZM
C
Curl 2002
OWAN
0.58
distance - home to
sprayed field-orig
categories
143
dust
AZM
C
Curl 2002
OWAN
0.58
distance - home to
sprayed field-rev
categories
454
dust
AZM
L
Grossman
2001
MLR-8
> 0.05
distance - home to
pesticide-treated
farmland
farm-worker
455
dust
AZM
L
Grossman
2001
MLR-8
> 0.05
distance - home to
pesticide-treated
farmland
farm-worker
339
dust
AZM
C
Lu 2000
MWU
0.008
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
573
dust
AZM
C
McCauley
2001a
SLR
0.32
distance - home to
agricultural fields
homes in Hood River
County with dust
samples
574
dust
AZM
C
McCauley
2001a
SLR
0.04
distance - home to
agricultural fields
homes in Hood River
County with detectable
dust samples
411
dust
AZM
C
Simcox
1995
KWAN
> 0.05
distance - home to
pesticide-treated
farmland (orchard)
(farmer + farm-worker)
families
dust: 2 samples pooled
concentration decreases with
increase in distance
D-43
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
419
dust
AZM
C
Simcox
1995
KWAN
< 0.001
distance - home to
pesticide-treated
farmland (orchard)
dust: 2 samples pooled
concentration decreases with
increase in distance; p = 0.0001
432
dust
AZM
L
Simcox
1995
KWAN
> 0.05
How far is the house
from a commercial
orchard?
(farmer + farm-worker)
families
dust: 2 samples pooled
732
dust
AZM
L
Simcox
1995
MLR-2
> 0.05
How far is the house
from a commercial
orchard?
(farmer + farm-worker)
families
dust: 2 samples pooled
415
dust
AZM
C
Simcox
1995
MWU
0.04
distance - home to
pesticide-treated
farmland (orchard)
(farmer + farm-worker)
families
dust: 2 samples pooled
651
dust
AZM
C
Simcox
1995
OWAN
> 0.05
proximity to orchards
(farmer + farm-worker)
families
dust: 2 samples pooled
341
dust
AZMPH
C
Lu 2000
MWU
0.014
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
349
dust
AZMPH
C
Lu 2000
MWU
0.02
distance - home to
pesticide-treated
farmland
distance > 0.25 mi -
home to pesticide-
treated farmland
adjusted by extraction
efficiencies
345
dust
AZMPH
C
Lu 2000
SLR
0.04
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
distance represented by
categories: <50 ft, 50-200 ft,
200 ft-0.25 mi, >0.25 mi
347
dust
AZMPH
C
Lu 2000
SLR
< 0.01
distance - home to
pesticide-treated
farmland
all families
adjusted by extraction
efficiencies
distance represented by
categories: <50 ft, 50-200 ft,
200 ft-0.25 mi, >0.25 mi,
reference family
254
dust
CHLR
C
Fenske
2002
MWU
< 0.01
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
258
dust
CHLR
C
Fenske
2002
SLR
< 0.001
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
distance represented by
categories: <50 ft, 50-200 ft,
200 ft-0.25 mi, >0.25 mi
D-44
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
413
dust
CHLR
C
Simcox
1995
KWAN
> 0.05
distance - home to
pesticide-treated
farmland (orchard)
(farmer + farm-worker)
families
dust: 2 samples pooled
concentration decreases with
increase in distance
421
dust
CHLR
C
Simcox
1995
KWAN
0.02
distance - home to
pesticide-treated
farmland (orchard)
dust: 2 samples pooled
concentration decreases with
increase in distance
434
dust
CHLR
L
Simcox
1995
KWAN
> 0.05
How far is the house
from a commercial
orchard?
(farmer + farm-worker)
families
dust: 2 samples pooled
734
dust
CHLR
L
Simcox
1995
MLR-1
> 0.05
How far is the house
from a commercial
orchard?
(farmer + farm-worker)
families
dust: 2 samples pooled
417
dust
CHLR
C
Simcox
1995
MWU
> 0.05
distance - home to
pesticide-treated
farmland (orchard)
(farmer + farm-worker)
families
dust: 2 samples pooled
655
dust
CHLR
C
Simcox
1995
OWAN
> 0.05
proximity to orchards
(farmer + farm-worker)
families
dust: 2 samples pooled
255
dust
EPAR
C
Fenske
2002
MWU
> 0.05
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
414
dust
EPAR
C
Simcox
1995
KWAN
0.005
distance - home to
pesticide-treated
farmland (orchard)
(farmer + farm-worker)
families
dust: 2 samples pooled
concentration decreases with
increase in distance
422
dust
EPAR
C
Simcox
1995
KWAN
0.001
distance - home to
pesticide-treated
farmland (orchard)
dust: 2 samples pooled
concentration decreases with
increase in distance
435
dust
EPAR
L
Simcox
1995
KWAN
> 0.05
How far is the house
from a commercial
orchard?
(farmer + farm-worker)
families
dust: 2 samples pooled
735
dust
EPAR
L
Simcox
1995
MLR-2
> 0.05
How far is the house
from a commercial
orchard?
(farmer + farm-worker)
families
dust: 2 samples pooled
D-45
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
418
dust
EPAR
C
Simcox
1995
MWU
0.005
distance - home to
pesticide-treated
farmland (orchard)
(farmer + farm-worker)
families
dust: 2 samples pooled
657
dust
EPAR
C
Simcox
1995
OWAN
0.001
proximity to orchards
(farmer + farm-worker)
families
dust: 2 samples pooled
658
dust
EPAR
c
Simcox
1995
TWAN-1
0.004
proximity to orchards
(farmer + farm-worker)
families
dust: 2 samples pooled
no significant interaction
between proximity to orchards
and applicator status
660
dust
EPAR
c
Simcox
1995
TWAN-3
> 0.05
proximity to orchards
(farmer + farm-worker)
families
dust: 2 samples pooled
significant interaction between
occupation (farmer vs
farmworker) and proximity to
orchards thus significance level
for proximity assigned as NS
340
dust
PHSM
c
Lu 2000
MWU
>= 0.10
distance - home to
pesticide-treated
farmland
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
412
dust
PHSM
c
Simcox
1995
KWAN
> 0.05
distance - home to
pesticide-treated
farmland (orchard)
(farmer + farm-worker)
families
dust: 2 samples pooled
concentration decreases with
increase in distance
420
dust
PHSM
c
Simcox
1995
KWAN
> 0.05
distance - home to
pesticide-treated
farmland (orchard)
dust: 2 samples pooled
concentration decreases with
increase in distance
433
dust
PHSM
L
Simcox
1995
KWAN
> 0.05
How far is the house
from a commercial
orchard?
(farmer + farm-worker)
families
dust: 2 samples pooled
733
dust
PHSM
L
Simcox
1995
MLR-3
> 0.05
How far is the house
from a commercial
orchard?
(farmer + farm-worker)
families
dust: 2 samples pooled
416
dust
PHSM
C
Simcox
1995
MWU
> 0.05
distance - home to
pesticide-treated
farmland (orchard)
(farmer + farm-worker)
families
dust: 2 samples pooled
653
dust
PHSM
C
Simcox
1995
OWAN
> 0.05
proximity to orchards
(farmer + farm-worker)
families
dust: 2 samples pooled
D-46
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
846
dust
OPSUM
C
McCauley
2003
CORR
0.5
distance from home to
nearest active orchard
r = -0.09
Q502--Living near Multiple Fields
818
urine
ETHL2
C
Royster
2002
MWU
> 0.05
living near multiple
fields
819
urine
ETHL2
A
Royster
2002
MWU
> 0.05
living near multiple
fields
816
urine
MTHL2
C
Royster
2002
MWU
> 0.05
living near multiple
fields
817
urine
MTHL2
A
Royster
2002
MWU
> 0.05
living near multiple
fields
D-47
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.1.6 Category 6: Residential Location
Table D.3.1.6 Relationship Comments for Questions in Category 6: Residential Location - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q601--Urban vs Non-urban
684
urine
1 NAP
A
Adgate
2001
WTAN
0.097
urban vs non-urban
children with 3 urine
samples
weighted intra-child means
750
urine
1 NAP
C
Adgate
2001
WTAN
0.13
urban vs non-urban
weighted intra-child means
751
urine
1 NAP
C
Adgate
2001
WTAN
0.1
urban vs non-urban
children with 3 urine
samples
weighted intra-child means
685
urine
MDA
A
Adgate
2001
WTAN
0.16
urban vs non-urban
children with 3 urine
samples
weighted intra-child means
752
urine
MDA
C
Adgate
2001
WTAN
0.099
urban vs non-urban
weighted intra-child means
753
urine
MDA
C
Adgate
2001
WTAN
0.16
urban vs non-urban
children with 3 urine
samples
weighted intra-child means
686
urine
TCPY
A
Adgate
2001
WTAN
0.019
urban vs non-urban
children with 3 urine
samples
weighted intra-child means
754
urine
TCPY
C
Adgate
2001
WTAN
0.036
urban vs non-urban
weighted intra-child means
755
urine
TCPY
C
Adgate
2001
WTAN
0.02
urban vs non-urban
children with 3 urine
samples
weighted intra-child means
Q602--Urban vs Rural
464
urine
TCPY
A
Krinsley
1998
SLR
0.62
rural residences vs.
urban residences
Q603--Border vs Non-border
463
urine
TCPY
A
Krinsley
1998
SLR
0.86
border residences vs.
non-border residences
D-48
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q604--Community
150
urine
ETHL1
C
Lu 2001
MWU
> 0.05
community
focus children:
community 1
average urine samples per child
151
urine
MTHL2
C
Lu 2001
MWU
> 0.05
community
focus children:
community 1
average urine samples per child
Q605--\/ehicle vs House
145
dust
AZM
c
Curl 2002
SLR
< 0.001
sampling location
measurements increase
together, r >0
D-49
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.2 Behavior Relationships
D.3.2.1 Category 7: Subject's Personal Characteristics
Table D.3.2.1 Relationship Comments for Questions in Category 7: Subject's Personal Characteristics - Grouped by Question and Sorted by Medium, Chemical,
Citation and Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q701~Sex
112
urine
1 NAP
C
Adgate
2001
WTAN
> 0.05
sex
weighted intra-child means
113
urine
MDA
C
Adgate
2001
WTAN
> 0.05
sex
weighted intra-child means
114
urine
TCPY
c
Adgate
2001
WTAN
> 0.05
sex
weighted intra-child means
460
urine
TCPY
A
Krinsley
1998
SLR
0.59
sex
n based on degrees of freedom
specified for analysis
108
urine
DEP
A
Aprea
2000
BDPH
> 0.05
sex
multiple comparison test w/1
independent variable
99
urine
DEP
A
Aprea
2000
MLR-5
> 0.05
sex
109
urine
DETP
A
Aprea
2000
BDPH
> 0.05
sex
multiple comparison test w/1
independent variable
100
urine
DETP
A
Aprea
2000
MLR-6
> 0.05
sex
110
urine
DEDTP
A
Aprea
2000
BDPH
> 0.05
sex
multiple comparison test w/1
independent variable
101
urine
DEDTP
A
Aprea
2000
MLR-7
> 0.05
sex
107
urine
DMP
A
Aprea
2000
BDPH
> 0.05
sex
multiple comparison test w/1
independent variable
D-50
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
98
urine
DMP
A
Aprea
2000
MLR-1
> 0.05
sex
103
urine
DMTP
A
Aprea
2000
BDPH
> 0.05
sex
multiple comparison test w/1
independent variable
94
urine
DMTP
A
Aprea
2000
MLR-2
< 0.05
sex
104
urine
DMDTP
A
Aprea
2000
BDPH
> 0.05
sex
multiple comparison test w/1
independent variable
95
urine
DMDTP
A
Aprea
2000
MLR-3
< 0.05
sex
152
urine
ETHL1
C
Lu 2001
MWU
> 0.05
sex of child
focus children:
communities combined
average urine samples per child
111
urine
ETHL2
A
Aprea
2000
BDPH
> 0.05
sex
multiple comparison test w/1
independent variable
102
urine
ETHL2
A
Aprea
2000
MLR-8
> 0.05
sex
601
urine
ETHL2
C
Koch
1999
MWU
0.411
sex
median excretion value per
child
285
urine
ETHL2
C
Koch
2002
GLM-2
0.046
sex
Includes multiple samples per
child
model adjusted for variables
including residential pesticide
use, proximity, and household
occupations
105
urine
MTHL2
A
Aprea
2000
BDPH
> 0.05
sex
multiple comparison test w/1
independent variable
96
urine
MTHL2
A
Aprea
2000
MLR-4
< 0.05
sex
600
urine
MTHL2
C
Koch
1999
MWU
0.097
sex
median excretion value per
child
model adjusted for variables
including residential pesticide
use, proximity, and household
occupations
D-51
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
284
urine
MTHL2
C
Koch
2002
GLM-1
0.005
sex
Includes multiple samples per
child
model adjusted for variables
including residential pesticide
use, proximity, and household
occupations
153
urine
MTHL2
C
Lu 2001
MWU
> 0.05
sex of child
focus children:
communities combined
average urine samples per child
106
urine
DAP1
A
Aprea
2000
BDPH
> 0.05
sex
multiple comparison test w/1
independent variable
97
urine
DAP1
A
Aprea
2000
MLR-9
< 0.05
sex
533
urine
DAP1
A
Shalat
2003
MVRG-1
0.310
gender of child
p = 0.016 for model; male = 1,
female = 0: p =0.3101
641
urine
DAP1
A
Shalat
2003
MVRG-2
> 0.05
gender of child
p = 0.076 for model; male = 1,
female = 0; significance level
assumed based on MVRG-1
Q702~Age
116
urine
1 NAP
C
Adgate
2001
WTAN
> 0.05
age
weighted intra-child means
117
urine
MDA
C
Adgate
2001
WTAN
> 0.05
age
weighted intra-child means
118
urine
TCPY
C
Adgate
2001
WTAN
> 0.05
age
weighted intra-child means
461
urine
TCPY
A
Krinsley
1998
SLR
0.75
age as continuous
variable
n based on degrees of freedom
specified for analysis
671
urine
TCPY
C
Krinsley
1998
SLR
< 0.05
age as continuous
variable
181
urine
DMTP
C
Loewen-
herz 1997
MWU
> 0.10
age
focus applicator
children; visit 1
182
urine
DMTP
C
Loewen-
herz 1997
MWU
> 0.10
age
focus applicator
children; visit 1
D-52
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
183
urine
DMTP
C
Loewen-
herz 1997
MWU
> 0.10
age
focus applicator
children; visit 1
184
urine
DMTP
C
Loewen-
herz 1997
MWU
> 0.10
age
focus applicator
children; visit 2
185
urine
DMTP
c
Loewen-
herz 1997
MWU
> 0.10
age
focus applicator
children; visit 2
186
urine
DMTP
c
Loewen-
herz 1997
MWU
0.06
age
focus applicator
children; visit 2
187
urine
DMTP
A
Loewen-
herz 1997
MWU
> 0.10
age
focus applicator
children; visit 1
188
urine
DMTP
A
Loewen-
herz 1997
MWU
> 0.10
age
focus applicator
children; visit 1
189
urine
DMTP
A
Loewen-
herz 1997
MWU
> 0.10
age
focus applicator
children; visit 1
190
urine
DMTP
A
Loewen-
herz 1997
MWU
> 0.10
age
focus applicator
children; visit 2
191
urine
DMTP
A
Loewen-
herz 1997
MWU
0.038
age
focus applicator
children; visit 2
192
urine
DMTP
A
Loewen-
herz 1997
MWU
0.083
age
focus applicator
children; visit 2
179
urine
DMTP
C
Loewen-
herz 1997
WSRK
> 0.10
age - paired siblings
focus applicator
children; visit 1
180
urine
DMTP
C
Loewen-
herz 1997
WSRK
0.04
age - paired siblings
focus applicator
children; visit 2
581
urine
ETHL1
C
Curl 2002
OWAN
> 0.05
age
582
urine
ETHL1
A
Curl 2002
OWAN
> 0.05
age
170
urine
ETHL1
C
Lu 2001
KWAN
0.64
age
focus children:
communities combined
average urine samples per child
603
urine
ETHL2
C
Koch
1999
MWU
0.014
age
median excretion value per
child
D-53
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
287
urine
ETHL2
C
Koch
2002
GLM-2
0.27
age
Includes multiple samples per
child
model adjusted for variables
including residential pesticide
use, proximity, and household
occupations
144
urine
MTHL2
A
Curl 2002
OWAN
0.001
age
580
urine
MTHL2
C
Curl 2002
OWAN
0.01
age
146
urine
MTHL2
C
Curl 2002
SLR
< 0.001
age
583
urine
MTHL2
A
Curl 2002
SLR
< 0.001
age
602
urine
MTHL2
C
Koch
1999
MWU
0.295
age
median excretion value per
child
286
urine
MTHL2
C
Koch
2002
GLM-1
0.16
age
Includes multiple samples per
child
model adjusted for variables
including residential pesticide
use, proximity, and household
occupations
171
urine
MTHL2
C
Lu 2001
KWAN
0.36
age
focus children:
communities combined
average urine samples per child
532
urine
DAP1
A
Shalat
2003
MVRG-1
0.007
age of child in months
p = 0.016 for model
640
urine
DAP1
A
Shalat
2003
MVRG-2
< 0.05
age of child in months
p = 0.076 for model;
significance level assumed
based on MVRG-1
Q703--Ethnicity
124
urine
1 NAP
C
Adgate
2001
WTAN
0.009
race
weighted intra-child means
125
urine
MDA
C
Adgate
2001
WTAN
0.035
race
weighted intra-child means
462
urine
TCPY
A
Krinsley
1998
SLR
0.99
ethnicity
Q704--Education Level
466
urine
TCPY
A
Krinsley
1998
SLR
0.44
education level
D-54
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q705~lncome
128
urine
1 NAP
C
Adgate
2001
WTAN
0.025
income
p-values not adjusted for
multiple comparisons
756
urine
MDA
C
Adgate
2001
WTAN
0.047
income
p-values not adjusted for
multiple comparisons
757
urine
MDA
c
Adgate
2001
WTAN
0.07
income
p-values not adjusted for
multiple comparisons
758
urine
MDA
c
Adgate
2001
WTAN
0.009
income
p-values not adjusted for
multiple comparisons
759
urine
TCPY
c
Adgate
2001
WTAN
0.012
income
p-values not adjusted for
multiple comparisons
760
urine
TCPY
c
Adgate
2001
WTAN
0.012
income
p-values not adjusted for
multiple comparisons
465
urine
TCPY
A
Krinsley
1998
SLR
0.32
household income
166
urine
ETHL1
C
Lu 2001
TNR
> 0.05
income level
focus children:
communities combined
average urine samples per child
167
urine
MTHL2
C
Lu 2001
TNR
> 0.05
income level
focus children:
communities combined
average urine samples per child
Q706--Loading from Hand Wipes"
Q707--Hand's Surface Area
534
urine
DAP1
A
Shalat
2003
MVRG-1
0.49
child's hand area
p = 0.016 for model
642
urine
DAP1
A
Shalat
2003
MVRG-2
> 0.05
child's hand area
p = 0.076 for model;
significance level assumed
based on MVRG-1
a There is no question grouping for the number 706.
D-55
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.2.2 Category 8: Child's Behaviors
Table D.3.2.2 Relationship Comments for Questions in Category 8: Child's Behaviors - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q801-Hand-to-Mouth Activity
212
urine
4NITR
C
Fenske
2002
MWU
> 0.05
child's hand-to-mouth
activity
focus children
average of visit 1 and visit 2
values
208
urine
TCPY
C
Fenske
2002
MWU
> 0.05
child's hand-to-mouth
activity
focus children
average of visit 1 and visit 2
values
624
urine
ETHL1
c
Lu 2001
MWU
> 0.05
hands in mouth
focus children:
communities combined
average urine samples per child
304
urine
MTHL1
c
Lu 2000
MWU
0.6
Do children have hand-
to-mouth activity?
(applicator + farm-
worker) families - focus
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
625
urine
MTHL2
c
Lu 2001
MWU
> 0.05
hands in mouth
focus children:
communities combined
average urine samples per child
Q802-Thumb-Sucking
213
urine
4NITR
c
Fenske
2002
MWU
> 0.05
child's frequent thumb-
sucking
focus children
average of visit 1 and visit 2
values
209
urine
TCPY
c
Fenske
2002
MWU
> 0.05
child's frequent thumb-
sucking
focus children
average of visit 1 and visit 2
values
626
urine
ETHL1
c
Lu 2001
MWU
> 0.05
thumb sucking
focus children:
communities combined
average urine samples per child
305
urine
MTHL1
c
Lu 2000
MWU
0.6
Do children suck their
thumbs?
(applicator + farm-
worker) families - focus
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
627
urine
MTHL2
c
Lu 2001
MWU
> 0.05
thumb sucking
focus children:
communities combined
average urine samples per child
D-56
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q803--Hand Washing before Meals
211
urine
4NITR
C
Fenske
2002
MWU
> 0.05
child's hand washing
before each meal
focus children
average of visit 1 and visit 2
values
207
urine
TCPY
C
Fenske
2002
MWU
> 0.05
child's hand washing
before each meal
focus children
average of visit 1 and visit 2
values
303
urine
MTHL1
c
Lu 2000
MWU
0.2
Do children wash their
hands before meals?
(applicator + farm-
worker) families - focus
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
Q804--Frequency of Handwashing
172
urine
ETHL1
c
Lu 2001
MWU
> 0.05
frequency of
handwashing
focus children:
communities combined
average urine samples per child
173
urine
MTHL2
c
Lu 2001
MWU
> 0.05
frequency of
handwashing
focus children:
communities combined
average urine samples per child
Q805--Time Spent Outdoors
210
urine
4NITR
c
Fenske
2002
KWAN
> 0.05
child's time spent
outdoors
focus children
average of visit 1 and visit 2
values
206
urine
TCPY
c
Fenske
2002
KWAN
> 0.05
child's time spent
outdoors
focus children
average of visit 1 and visit 2
values
302
urine
MTHL1
c
Lu 2000
KWAN
0.8
How many hours/day
are children outdoors?
(applicator + farm-
worker) families - focus
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
Q806--Loading from Hand Wipe
535
urine
DAP1
A
Shalat
2003
MVRG-1
0.022
child's hand load
p = 0.016 for model p = 0.0219
642
urine
DAP1
A
Shalat
2003
MVRG-2
> 0.05
child's hand area
p = 0.076 for model;
significance level assumed
based on MVRG-1
D-57
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.2.3 Category 9: Dietary Behaviors
Table D.3.2.3 Relationship Comments for Questions in Category 9: Dietary Behaviors - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q901--Type of Drinking Water
459
urine
TCPY
A
Krinsley
1998
SLR
< 0.19
drinkers of bottled
water
Q902--Consumption of Homegrown Fresh Vegetables
457
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
consumption of
homegrown fresh
vegetables
Q903--Ate Lunch at School
90
urine
DEP
A
Aprea
2000
BDPH
> 0.05
ate lunch at school
multiple comparison test w/1
independent variable
81
urine
DEP
A
Aprea
2000
MLR-5
> 0.05
ate lunch at school
91
urine
DETP
A
Aprea
2000
BDPH
> 0.05
ate lunch at school
multiple comparison test w/1
independent variable
82
urine
DETP
A
Aprea
2000
MLR-6
> 0.05
ate lunch at school
92
urine
DEDTP
A
Aprea
2000
BDPH
> 0.05
ate lunch at school
multiple comparison test w/1
independent variable
83
urine
DEDTP
A
Aprea
2000
MLR-7
> 0.05
ate lunch at school
89
urine
DMP
A
Aprea
2000
BDPH
> 0.05
ate lunch at school
multiple comparison test w/1
independent variable
80
urine
DMP
A
Aprea
2000
MLR-1
> 0.05
ate lunch at school
D-58
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
93
urine
ETHL2
A
Aprea
2000
BDPH
> 0.05
ate lunch at school
multiple comparison test w/1
independent variable
84
urine
ETHL2
A
Aprea
2000
MLR-8
> 0.05
ate lunch at school
Q904--Organic Diet
822
urine
ETHL2
C
Curl 2003
MWU
0.13
organic vs conventional
diet
823
urine
ETHL2
C
Curl 2003
MWU
> 0.05
organic vs conventional
diet
no residential use of
OP pesticides
820
urine
MTHL2
C
Curl 2003
MWU
< 0.001
organic vs conventional
diet
p = 0.0003
821
urine
MTHL2
C
Curl 2003
MWU
< 0.05
organic vs conventional
diet
no residential use of
OP pesticides
D-59
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.2.4 Category 10: Family Hygiene Practices
Table D.3.2.4 Relationship Comments for Questions in Category 10: Family Hygiene Practices - Grouped by Question and Sorted by Medium, Chemical, Citation
and Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q1001-Shoes Removed at Door
616
urine
DMTP
C
Carrel
1996
CHSQ
> 0.10
Are work shoes worn
inside house? (group
names are reversed to
fit Q1001)
focus applicator
children; < 200 feet
from orchard; visit 1
comparison of % detects, %
trace, and % non-detects; group
names reversed to fit question
Q1001
617
urine
DMTP
C
Carrel
1996
CHSQ
0.083
Are work shoes worn
inside house? (group
names are reversed to
fit Q1001)
focus applicator
children; < 200 feet
from orchard; visit 2
comparison of % detects, %
trace, and % non-detects; group
names reversed to fit question
Q1001
612
urine
DMTP
c
Carrel
1996
MWU
> 0.10
Are work shoes worn
inside house? (group
names reversed to fit
Q1001)
focus applicator
children; < 200 feet
from orchard; visit 1
group names reversed to fit
question Q1001
613
urine
DMTP
c
Carrel
1996
MWU
0.096
Are work shoes worn
inside house? (group
names reversed to fit
Q1001)
focus applicator
children; < 200 feet
from orchard: visit 2
group names reversed to fit
question Q1001
311
urine
MTHL1
c
Lu 2000
MWU
0.2
Do household
members remove
shoes at the door?
(applicator + farm-
worker) families'
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
440
dust
AZM
L
Grossman
2001
MLR-1
> 0.05
Remove shoes outside
home?
fieldworker
835
dust
AZM
C
McCauley
2003
WTWS
0.46
shoes removed at door
720
dust
AZM
L
Simcox
1995
MLR-1
> 0.05
Do familiy members
remove shoes at the
door?
dust: 2 samples pooled
D-60
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
351
dust
AZM
L
Simcox
1995
MWU
> 0.05
Do familiy members
remove shoes at the
door?
dust: 2 samples pooled
306
dust
AZMPH
C
Lu 2000
MWU
0.8
Do household
members remove
shoes at the door?
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
722
dust
CHLR
L
Simcox
1995
MLR-3
> 0.05
Do familiy members
remove shoes at the
door?
dust: 2 samples pooled
353
dust
CHLR
L
Simcox
1995
MWU
> 0.05
Do familiy members
remove shoes at the
door?
dust: 2 samples pooled
723
dust
EPAR
L
Simcox
1995
MLR-4
> 0.05
Do familiy members
remove shoes at the
door?
dust: 2 samples pooled
354
dust
EPAR
L
Simcox
1995
MWU
> 0.05
Do familiy members
remove shoes at the
door?
dust: 2 samples pooled
721
dust
PHSM
L
Simcox
1995
MLR-2
> 0.05
Do familiy members
remove shoes at the
door?
dust: 2 samples pooled
352
dust
PHSM
L
Simcox
1995
MWU
> 0.05
Do familiy members
remove shoes at the
door?
dust: 2 samples pooled
834
dust
OPSUM
C
McCauley
2003
WTWS
0.36
shoes removed at door
Q1002~Presence of Doormats
218
urine
4NITR
C
Fenske
2002
TNR
> 0.05
presence of doormats
focus children
average of visit 1 and visit 2
values
214
urine
TCPY
C
Fenske
2002
TNR
> 0.05
presence of doormats
focus children
average of visit 1 and visit 2
values
D-61
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
312
urine
MTHL1
C
Lu 2000
MWU
0.3
Are there doormats
outside the main
entrance?
(applicator + farm-
worker) families'
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
724
dust
AZM
L
Simcox
1995
MLR-1
> 0.05
Are there walk-off mats
outside main entries?
dust: 2 samples pooled
355
dust
AZM
L
Simcox
1995
MWU
> 0.05
Are there walk-off mats
outside main entries?
dust: 2 samples pooled
307
dust
AZMPH
C
Lu 2000
MWU
0.6
Are there doormats
outside the main
entrance?
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
222
dust
CHLR
C
Fenske
2002
MWU
> 0.05
presence of doormats
adjusted by extraction
efficiencies
726
dust
CHLR
L
Simcox
1995
MLR-3
> 0.05
Are there walk-off mats
outside main entries?
dust: 2 samples pooled
357
dust
CHLR
L
Simcox
1995
MWU
> 0.05
Are there walk-off mats
outside main entries?
dust: 2 samples pooled
226
dust
EPAR
C
Fenske
2002
MWU
> 0.05
presence of doormats
adjusted by extraction
efficiencies
727
dust
EPAR
L
Simcox
1995
MLR-4
> 0.05
Are there walk-off mats
outside main entries?
dust: 2 samples pooled
358
dust
EPAR
L
Simcox
1995
MWU
> 0.05
Are there walk-off mats
outside main entries?
dust: 2 samples pooled
725
dust
PHSM
L
Simcox
1995
MLR-2
> 0.05
Are there walk-off mats
outside main entries?
dust: 2 samples pooled
356
dust
PHSM
L
Simcox
1995
MWU
> 0.05
Are there walk-off mats
outside main entries?
dust: 2 samples pooled
Q1003--Presence of Floor Mats
174
urine
ETHL1
C
Lu 2001
MWU
> 0.05
presence of floor mats
focus children:
communities combined
average urine samples per child
175
urine
MTHL2
C
Lu 2001
MWU
> 0.05
presence of floor mats
focus children:
communities combined
average urine samples per child
D-62
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q1004--Vacuuming Frequency
221
urine
4NITR
C
Fenske
2002
MWU
> 0.05
vacuuming frequency
focus children
average of visit 1 and visit 2
values
217
urine
TCPY
C
Fenske
2002
MWU
> 0.05
vacuuming frequency
focus children
average of visit 1 and visit 2
values
622
urine
ETHL1
c
Lu 2001
MWU
> 0.05
frequency of
vacuuming
focus children:
communities combined
average urine samples per child
315
urine
MTHL1
c
Lu 2000
MWU
0.3
How frequently is the
carpet vacuumed?
(applicator + farm-
worker) families'
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
623
urine
MTHL2
c
Lu 2001
MWU
> 0.05
frequency of
vacuuming
focus children:
communities combined
average urine samples per child
310
dust
AZMPH
c
Lu 2000
MWU
0.6
How frequently is the
carpet vacuumed?
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
225
dust
CHLR
c
Fenske
2002
MWU
> 0.05
vacuuming frequency
adjusted by extraction
efficiencies
229
dust
EPAR
c
Fenske
2002
MWU
> 0.05
vacuuming frequency
adjusted by extraction
efficiencies
Q1005--\/acuuming Indoor Play Areas
359
dust
AZM
L
Simcox
1995
KWAN
> 0.05
How frequently are
children's indoor play
areas vacuumed?
dust: 2 samples pooled
728
dust
AZM
L
Simcox
1995
MLR-1
> 0.05
How frequently are
children's indoor play
areas vacuumed?
dust: 2 samples pooled
361
dust
CHLR
L
Simcox
1995
KWAN
> 0.05
How frequently are
children's indoor play
areas vacuumed?
dust: 2 samples pooled
730
dust
CHLR
L
Simcox
1995
MLR-3
> 0.05
How frequently are
children's indoor play
areas vacuumed?
dust: 2 samples pooled
D-63
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
362
dust
EPAR
L
Simcox
1995
KWAN
> 0.05
How frequently are
children's indoor play
areas vacuumed?
dust: 2 samples pooled
731
dust
EPAR
L
Simcox
1995
MLR-1
> 0.05
How frequently are
children's indoor play
areas vacuumed?
dust: 2 samples pooled
360
dust
PHSM
L
Simcox
1995
KWAN
> 0.05
How frequently are
children's indoor play
areas vacuumed?
dust: 2 samples pooled
729
dust
PHSM
L
Simcox
1995
MLR-2
> 0.05
How frequently are
children's indoor play
areas vacuumed?
dust: 2 samples pooled
Q1006--Work Clothes Worn Indoors
219
urine
4NITR
C
Fenske
2002
MWU
> 0.05
wearing of work shoes
and work clothes in the
house
focus children
average of visit 1 and visit 2
values
215
urine
TCPY
C
Fenske
2002
TNR
> 0.05
wearing of work shoes
and work clothes in the
house
focus children
average of visit 1 and visit 2
values
716
urine
DMTP
C
Carrel
1996
CHSQ
> 0.10
clothes changing
focus applicator
children; < 200 feet
from orchard; visit 1
comparison of % detects, %
trace, and % non-detects
717
urine
DMTP
C
Carrel
1996
CHSQ
> 0.10
clothes changing
focus applicator
children; < 200 feet
from orchard; visit 2
comparison of % detects, %
trace, and % non-detects
712
urine
DMTP
C
Carrel
1996
MWU
> 0.10
clothes changing
focus applicator
children; < 200 feet
from orchard; visit 1
713
urine
DMTP
C
Carrel
1996
MWU
> 0.10
clothes changing
focus applicator
children; < 200 feet
from orchard: visit 2
D-64
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
313
urine
MTHL1
C
Lu 2000
MWU
0.2
Do household
members wear work
clothes in the house?
(applicator + farm-
worker) families'
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
837
dust
AZM
C
McCauley
2003
TTST
< 0.01
amount of time until
work clothes are
changed
308
dust
AZMPH
c
Lu 2000
MWU
0.2
Do household
members wear work
clothes in the house?
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
223
dust
CHLR
c
Fenske
2002
MWU
> 0.05
wearing of work shoes
and work clothes in the
house
adjusted by extraction
efficiencies
227
dust
EPAR
c
Fenske
2002
MWU
> 0.05
wearing of work shoes
and work clothes in the
house
adjusted by extraction
efficiencies
836
dust
OPSUM
c
McCauley
2003
TTST
< 0.01
amount of time until
work clothes are
changed
Q1007--Work Clothes Mixed with Laundry
314
urine
MTHL1
c
Lu 2000
MWU
0.8
Do work clothes mix
with family laundry?
(applicator + farm-
worker) families'
children
average of visit 1 and visit 2
values, adjusted by extraction
efficiencies
309
dust
AZMPH
c
Lu 2000
MWU
0.4
Do work clothes mix
with family laundry?
(applicator + farm-
worker) families
adjusted by extraction
efficiencies
Q1008--Laundering Practices
220
urine
4NITR
c
Fenske
2002
MWU
> 0.05
laundering practices
(work clothes with
laundry)
focus children
average of visit 1 and visit 2
values
216
urine
TCPY
c
Fenske
2002
TNR
> 0.05
laundering practices
(work clothes with
laundry)
focus children
average of visit 1 and visit 2
values
D-65
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
706
urine
DMTP
C
Carrel
1996
CHSQ
> 0.10
washing practices
focus applicator
children; < 200 feet
from orchard; visit 1
comparison of % detects, %
trace, and % non-detects
707
urine
DMTP
C
Carrel
1996
CHSQ
> 0.10
washing practices
focus applicator
children; < 200 feet
from orchard; visit 2
comparison of % detects, %
trace, and % non-detects
702
urine
DMTP
c
Carrel
1996
MWU
> 0.10
washing practices
focus applicator
children; < 200 feet
from orchard; visit 1
703
urine
DMTP
c
Carrel
1996
MWU
> 0.10
washing practices
focus applicator
children; < 200 feet
from orchard: visit 2
224
dust
CHLR
c
Fenske
2002
MWU
> 0.05
laundering practices
(work clothes with
laundry)
adjusted by extraction
efficiencies
228
dust
EPAR
c
Fenske
2002
MWU
> 0.05
laundering practices
(work clothes with
laundry)
adjusted by extraction
efficiencies
Q1009--Number of Days Since Last Vacuuming
833
dust
OPSUM
c
McCauley
2003
MLR
0.03
number of days since
last cleaning of area
where sample was
taken
partial correlation = 0.45;
analysis adjusted for location of
sampled area
Q1010--Shower Soon After Work
866
urine
DMTP
c
Carrel
1996
CHSQ
> 0.10
showering
focus applicator
children; < 200 feet
from orchard; visit 1
comparison of % detects, %
trace, and % non-detects
867
urine
DMTP
c
Carrel
1996
CHSQ
> 0.10
showering
focus applicator
children; < 200 feet
from orchard; visit 2
comparison of % detects, %
trace, and % non-detects
862
urine
DMTP
c
Carrel
1996
MWU
> 0.10
showering
focus applicator
children; < 200 feet
from orchard; visit 1
D-66
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
863
urine
DMTP
C
Carrel
1996
MWU
> 0.10
showering
focus applicator
children; < 200 feet
from orchard: visit 2
441
dust
AZM
L
Grossman
2001
MLR-2
> 0.05
Shower after work
within 1 hour?
fieldworker
839
dust
AZM
C
McCauley
2003
TTST
0.89
washed immediately
upon arriving home (<
30 min)
838
dust
OPSUM
C
McCauley
2003
TTST
0.63
washed immediately
upon arriving home (<
30 min)
Q1011'
Q1012--After Work Hygiene Index
841
dust
AZM
C
McCauley
2003
CORR
0.43
aggregate measure of
after work hygiene
practices (removing
shoes, time before
clothes change,
wearing clothes
indoors, time before
washing)
r = 0.17
840
dust
OPSUM
C
McCauley
2003
CORR
0.8
aggregate measure of
after work hygiene
practices (removing
shoes, time before
clothes change,
wearing clothes
indoors, time before
washing)
r = 0.05
a No questions associated with this Q#.
D-67
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.2.5 Category 11: Smoking-Related Activities
Table D.3.2.5 Relationship Comments for Questions in Category 11: Smoking-Related Activities - Grouped by Question and Sorted by Medium, Chemical, Citation
and Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q1101--Current Smoker
467
urine
TCPY
A
Krinsley
1998
SLR
0.009
current smoker
Q1102--Subject Smoked
764
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
whether subject
smoked
PUI was constructed from
pesticide use variables: p <
0.00001
769
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
whether subject
smoked
p < 0.00001
Q1103--Exposure to Second Hand Smoke
474
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
exposure to second
hand smoke
D-68
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.2.6 Category 12: Work Exposure/Practices
Table D.3.2.6 Relationship Comments for Questions in Category 12: Work Exposure/Practices - Grouped by Question and Sorted by Medium, Chemical, Citation
and Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q1201-Pesticide Exposure at Work in Past 6 mo
531
urine
TCPY
A
Krinsley
1998
SLR
0.37
pesticide exposure at
work in past 6 mo
Q1202--Wear Boots while Doing Fieldwork?
446
dust
AZM
L
Grossman
2001
MLR-4
> 0.05
Wear boots while doing
fieldwork?
fieldworker
Q1203--Wear Gloves while Doing Fieldwork?
447
dust
AZM
L
Grossman
2001
MLR-5
> 0.05
Wear gloves while
doing fieldwork?
fieldworker
Q1204--Wear Hat while Doing Fieldwork?
445
dust
AZM
L
Grossman
2001
MLR-3
> 0.05
Wear hat while doing
fieldwork?
fieldworker
D-69
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.3 Other Relationships
D.3.3.1 Category 13: Related Exposure Levels
Table D.3.3.1 Relationship Comments for Questions in Category 13: Related Exposure Levels - Grouped by Question and Sorted by Medium, Chemical, Citation and
Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q1301--Detectable Levels in Adult Household Members
262
urine
DAP2
C
Azaroff
1999
SLGR
< 0.01
numbers of household
members > 17 years
old excreting
detectable AP
metabolites
children 8-17 years old
child's detectable AP
metabolites - 3 samples
combined
265
urine
DAP2
C
Azaroff
1999
SLGR
0.1
numbers of household
members > 17 years
old excreting
detectable AP
metabolites
children 8-17 years old,
reporting no fieldwork
within past 2 weeks
child's detectable AP
metabolites - 3 samples
combined
Q1302--High Levels in Adult Household Members
266
urine
MTHL4
c
Azaroff
1999
SLGR
< 0.01
numbers of household
members > 17 years
old excreting high level
AP metabolites
children 8-17 years old,
reporting no fieldwork
of any kind
child's detectable methylated
AP metabolites - 3 samples
combined
263
urine
DAP2
c
Azaroff
1999
SLGR
< 0.01
numbers of household
members > 17 years
old excreting high or
very high level AP
metabolites
children 8-17 years old
child's detectable AP
metabolites - 3 samples
combined
264
urine
DAP3
c
Azaroff
1999
SLGR
< 0.01
numbers of household
members > 17 years
old excreting high or
very high level AP
metabolites
children 8-17 years old
child's high or very high level of
an AP metabolite - 3 samples
combined
D-70
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
D.3.3.2 Category 14: Health
Table D.3.3.2 Relationship Comments for Questions in Category 14: Health - Grouped by Question and Sorted by Medium, Chemical, Citation and Analysis
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q1401 -Health Status
472
urine
TCPY
A
Krinsley
1998
SLR
< 0.66
self-reported health
status
Q1402--Asthma and Allergies
470
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
ever having had
asthma and allergies
Q1403--Bowel Disease
761
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
ever having had bowel
disease
PUI was constructed from
pesticide use variables: p <
0.00001
Q1404--Diabetes
471
urine
TCPY
A
Krinsley
1998
SLR
> 0.20
ever having had
diabetes
Q1405--lntestinal Disease
765
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
whether the subject
ever had intestinal
disease
p < 0.00001
770
urine
TCPY
A
Krinsley
1998
FSLR
< 0.001
whether the subject
ever had intestinal
disease
used pesticides both
inside and outside,
personally or
professionally applied
p < 0.00001
469
urine
TCPY
A
Krinsley
1998
SLR
0.004
ever having had
intestinal disease
D-71
August 2005
-------
Relationship Between Questionnaire Responses and Children's Pesticide Exposure Measurements
ID
#
Medium
Chemi-
cal
M
T
Citation
Analysis
p-value
Original Question
Subpopulation
Analyzed
Notes on Measurement
Notes on Analysis
Q1406~Ulcer
468
urine
TCPY
A
Krinsley
1998
SLR
0.02
ever having had an
ulcer
D-72
August 2005
-------
Appendix E
Questions Tracked in the Literature Review
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Contents
E. 1 Introducti on E-1
E.2 Questions Tracked for Relationships E-l
Tables
Table E. 1 Distribution of Questions Tracked in Literature Review by Risk Factor and
Question Category E-l
Table E.2 Questions Tracked in Literature Review by Risk Factor and Question Category .E-2
E-ii August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendix E
Questions Tracked in the Literature Review
E.l Introduction
Sections 4.2.1 and 4.2.2 describe the process for extracting relationship information from
publications identified in the literature review. Sections 4.2.4, 4.2.5, 4.2.6, and
Appendices B, C, and D describe the relationships of pesticide or pesticide metabolite
concentrations in various medium measurements with questions regarding environment
or behavior. This Appendix describes the breadth of the questions extracted from the
literature review (section 4.2).
E.2 Questions Tracked for Relationships
As the publications were reviewed for relationships, a level of organization for the
relationships was introduced at the question level. As described in section 4.2.2.2,
individual question phrasings found in the publications that implied the same question
were grouped together. The group of questions was then assigned an abbreviated
question phrasing or description. For example, the abbreviated question phrasing or
description "inside treated" includes the following questions:
• pesticide use inside
• pesticide used inside in past 6 months
• Was there indoor pesticide application in past 6 months?
• In the past 6 months were any chemicals for the control of fleas, roaches, ants or
other insects used inside this house/apartment?
A question #, e.g., Q102, is assigned to each question phrasing for ease of refernce in
other tables. The question groupings were then organized into 14 question categories
under three risk factors for presentation and discussion purposes. Table E. 1 shows the
risk factors and question categories used and the number of different abbreviated question
descriptions or question groupings included in each.
Table E.l Distribution of Questions Tracked in Literature Review by Risk Factor and Question
Category
Risk Factor
Category
# Question Groupings
#
Name
#
Name
1
Source
1
Residential pesticide use
32
1
Source
2
Household characteristics
17
1
Source
3
Residential sources (environmental
measurements)
3
1
Source
4
Household occupation
16
1
Source
5
Residential proximity to agricultural fields
2
1
Source
6
Residential location
5
E-l
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Ris
k Factor
Category
# Question Groupings
2
Behavior
7
Subject's personal characteristics
6
2
Behavior
8
Child's behaviors
6
2
Behavior
9
Dietary behaviors
4
2
Behavior
10
Family hygienic practices
11
2
Behavior
11
Smoking-related activities
3
2
Behavior
12
Work exposure/practices
4
3
Other
13
Related exposure levels
2
3
Other
14
Health
5
Table E.2 shows the abbreviated question phrasing or description for each of the question
groupings by risk factor and question category. The detailed question phrasing for a
specific relationship, as noted in its publication, can be found in Appendix D.
Table E.2 Questions Tracked in Literature Review by Risk Factor and Question Category
Risk Factor
Category
Question Grouping
#
Name
#
Name
#
Name
1
Source
1
Residential pesticide use
101
pesticide use
102
inside treated
103
inside treated - bathroom
104
inside treated - bedroom
105
inside treated - cabinets
106
inside treated - closets
107
inside treated - cupboards with dishes
108
inside treated - dining room
109
inside treated - family room
110
inside treated - kitchen
111
inside treated - living room
112
inside treated - on baseboards
113
inside treated - on ceiling
114
inside treated - on floor
115
inside treated - on lower walls
E-2
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Risk Factor
Category
Question Grouping
#
Name
#
Name
#
Name
116
inside treated - on upper walls
117
inside treated - other room
118
pets treated
119
outside treated
120
garden treated
121
lawn/yard treated
122
inside or outside treated
123
previous treatment
124
level of pesticide use
125
frequency personal application inside
126
frequency personal application outside
127
inside/outside treated by family member
128
frequency professional application inside
129
frequency professional application outside
130
personally mixed pesticide inside
131
personally mixed pesticide outside
132
presence during mixing
1
Source
2
Household characteristics
201
housing type
202
property used as a farm
203
age of house > 10 years
204
age of house > 20 years
205
having air conditioning
206
having central heating
207
having evaporative cooling
208
pets in house
209
pets inside/outside house
210
pet inside to outside
211
existence of garden or vegetable garden
212
ornamental plants or cut flowers
213
size of household
E-3
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Risk Factor
Category
Question Grouping
#
Name
#
Name
#
Name
214
location of play area
215
age of house (years)
216
size of home (sq ft)
217
number of pets in house
1
Source
3
Residential sources
(environmental
measurements)
301
household dust
302
loading from household floor dust
303
outdoor soil
1
Source
4
Household occupation
401
agricultural workers in household
402
household member spraying fields
403
recent fieldwork
404
applicator vs farmworker
405
applicator vs non-applicator
406
applicator vs reference
407
applicator+farmworker vs reference
408
farmer vs farmworker
409
farmer+farmworker vs reference
410
farmworker vs grower
411
farmworker vs others
412
field worker vs pesticide handler
413
expected occupational pesticide exposure
414
occupational pesticide exposure
415
tree thinning
416
number in household with high pesticide
contact
1
Source
5
Residential proximity to
agricultural fields
501
proximity of home to pesticide-treated
farmland/orchard
502
living near multiple fields
E-4
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Risk Factor
Category
Question Grouping
#
Name
#
Name
#
Name
1
Source
6
Residential location
601
urban vs non-urban
602
urban vs rural
603
border vs. non-border
604
community
605
vehicle vs house
2
Behavior
7
Subject's personal
characteristics
701
sex
702
age
703
ethnicity
704
education level
705
income
707a
hand's surface area
2
Behavior
8
Child's behaviors
801
hand-to-mouth activity
802
thumb-sucking
803
hand washing before meals
804
frequency of handwashing
805
time spent outdoors
806
loading from hand wipe
2
Behavior
9
Dietary behaviors
901
type of drinking water
902
consumption of homegrown fresh
vegetables
903
ate lunch at school
904
organic diet
2
Behavior
10
Family hygienic practices
1001
shoes removed at door
1002
presence of doormats
1003
presence of floor mats
1004
vacuuming frequency
1005
vacuuming indoor play areas
E-5
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Risk Factor
Category
Question Grouping
#
Name
#
Name
#
Name
1006
work clothes worn indoors
1007
work clothes mixed with laundry
1008
laundering practices
1009
number of weeks since last vacuuming
1010
shower soon after work
1012b
after work hygiene index
2
Behavior
11
Smoking-related activities
1101
current smoker
1102
subject smoked
1103
exposure to second hand smoke
2
Behavior
12
Work exposure/practices
1201
pesticide exposure at work in past 6 mo
1202
wear boots while doing fieldwork
1203
wear gloves while doing fieldwork
1204
wear hat while doing fieldwork
3
Other
13
Related exposure levels
1301
detectable levels in adult household
members
1302
high levels in adult household members
3
Other
14
Health
1401
health status
1402
asthma and allergies
1403
bowel disease
1404
diabetes
1405
intestinal disease
1406
ulcer
a There is no question grouping with the number 706.
b There is no question grouping with the number 1011.
E-6
August 2005
-------
Appendix F
Definition of Chemical Measurement Variables Used in the
Analysis of the Yuma Study
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Contents
F. 1 Introducti on F-1
F.2 Chemicals/Metabolites Measured F-l
F.3 Chemicals/Metabolites Used in Statistical Analyses F-2
F.3.1 Molar-Equivalent Concentrations F-2
F.3.2 Molar-Weighted Concentration Sums F-2
Tables
Table F. 1 Metabolites Measured in Yuma Study Urine Samples F-l
Table F.2 Pesticides Measured in Yuma Study Household and School Dust Samples F-l
Table F.3 Molar-Weighted Sums in the Yuma Study Data Mining Analyses F-3
Table F.4 Pesticides Included in Molar-Weighted Dust Sums Used in the Yuma Study
Data Mining Analyses F-4
F-ii
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendix F
Definition of Chemical Measurement Variables Used in the Analysis of the Yuma Study
F.l Introduction
The Children's Pesticide Exposure Study (Yuma Study) described in Section 4.3 included
measurements of pesticide and pesticide metabolite concentrations in samples of household
and school dust, and children's urine, respectively. The statistical analyses of the Yuma
Study data were performed on molar-weighted sums of the concentrations because the levels
of individual chemicals/metabolites were very small or below detection level (BLD). This
appendix describes the chemicals/metabolites measured in the study and the molar-weighted
sums used in the statistical analyses.
F.2 Chemicals/Metabolites Measured
The urine samples were analyzed for the dialkylphosphates (DAPs) shown in Table F. 1.
Table F.l Metabolites Measured in Yuma Study Urine Samples
Name
Description
DEP
diethylphosphate
DETP
diethylthiophosphate
DEDTP
diethyldithiophosphate
DMP
dimethylphosphate
DMTP
dimethylthiophosphate
DMDTP
dimethyldithiophosphate
The household dust and school dust samples were analyzed for pesticides in the classes
organophoshates, organochlorines, permethrins, and miscellaneous (Table F.2).
Table F.2 Pesticides Measured in Yuma Study Household and School Dust Samples
atrazine
4,4-' DDT
methyl parathion3
azinphos-methyl3
diazinon3
methoxychlor
bendiocarb
dichlorvos3
metolachlor
bensulide
dicofol
pendimethalin
benzamide
dieldrin
cis-permethrin
captan
disulfoton3
trans-permethrin
carbaryl
endosulfan 1
o-phenylphenol
carbofuran
endosulfan 2
phorate3
alpha-chlordane
ethyl parathion3
prometryn
F-l
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
gamma-chlordane
folpet
propoxur
chlorpyrifos3
fonophos3
simazine
chlorthal-dimethyl
heptachlor
terbufos3
cypermethrin
hexachlorobenzene
trifluralin
4,4-' DDD
lindane
4,4-' DDE
malathion3
a Organophosphorous (OP) pesticides
F.3 Chemicals/Metabolites Used in Statistical Analyses
In many cases, the concentration of the pesticides and/or metabolites found in the samples
was BLD. Also, it was not as important for this study to identify relationships with the
individual pesticides. Thus, for comparability of results across DAPs and to identify trends
by categories of pesticides, two types of additional measurement variables were created for
the statistical analyses: the molar equivalent of the chemical or metabolite concentration, and
molar-weighted sums of chemical or metabolite concentrations.
F.3.1 Molar-Equivalent Concentrations
A molar-equivalent concentration (MEC) is the ratio of the concentration to the molecular
weight of the chemical or metabolite. For example,
DEPMEc(nmoles/L) = DEP concentration (ng/L)/154.103 g/mole DEP or
CarbarylMEc(nmoles/g) = Carbaryl concentration (ng/g) /201.22 g/mole carbaryl,
depending on the units involved.
F.3.2 Molar-Weighted Concentration Sums
Molar-weighted sums were created for the ethylated and methylated DAPs under both
approaches (sections 4.3.2 and 4.3.3). Examples of these sums for the concentrations follow.
Ethylated DAP Sum MEc(nmoles/L) = DEPMec + DETPmec + DEDTPmec
Methylated DAP Sum MEc(nmoles/L) = DMPmec + DMTPmec + DMDTPmec
Sums were created for the concentrations adjusted, and not adjusted, for creatinine. The
Yuma Study report (CDC 2002) identified the sums of the ethylated and methylated DAPs as
DEOP and DMOP, respectively. For the data mining approach, the log of each sum was
used for the analyses. The sums of the ethylated and methylated DAPs were identified as
LWETHSUM and LWMETHSM, respectively.
F-2
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Under the data mining approach, sums were also created for the concentrations measured in
the household and school dust samples (Tables F.3 and F.4). Pesticides that had large
numbers of non-detectable measurements or that were similar (e.g., chlordanes) were
grouped together into sums.
Table F.3 Molar-Weighted Sums in the Yuma Study Data Mining Analyses
Name
Description3
Household Dust
WCHDNSUM
Weighted sum of alpha-chlordane and gamma-chlordane
WCHLPYRF
Weighted chlorpyrifos
WCYPERMETb
Weighted cypermethrin
WDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and 4,4'DDT
WDIAZNON
Weighted diazonin
WDUSTBAL
Weighted sum of dust analytes except OP pesticides
WDUSTSUM
Weighted sum of all dust analytes
WOPBAL
Weighted sum of OP pesticides except chlorpyrifos and diazinon
WOPHNYLPb
Weighted o-phenylphenol
WOPSUM
Weighted sum of OP pesticides
WPERMSUM
Weighted sum of cispermethrin and transpermethrin
School Dust
SWCHDNSM
Weighted sum of alpha-chlordane and gamma-chlordane
SWCHLPYR
Weighted chlorpyrifos
SWCYPRMEb
Weighted cypermethrin
SWDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and 4,4'DDT
SWDIAZNO
Weighted diazonin
SWDSTBAL
Weighted sum of dust analytes except OP pesticides
SWDUSTSM
Weighted sum of all dust analytes
SWOPBAL
Weighted sum of OP pesticides except chlorpyrifos and diazinon
SWOPHNYLb
Weighted o-phenylphenol
SWOPSUM
Weighted sum of OP pesticides
SWPERMSM
Weighted sum of cispermethrin and transpermethrin
a See Table F.4 for chemicals/metabolites included in sums.
b Weighted sums included in Principal Component analyses, but not in CART analyses.
F-3
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table F.4 Pesticides Included in Molar-Weighted Dust Sums for the Yuma Study Data Mining Analyses
Pesticide
Household Dust
School Dust
W
w
W
w
w
W
w
w
w
W
w
S
S
S
S
s
S
s
s
S
S
s
C
c
c
D
D
D
D
0
0
0
p
w
w
W
w
w
w
w
w
w
W
w
H
H
Y
D
1
U
U
p
p
P
E
c
c
c
D
D
D
D
0
0
0
p
D
L
P
S
A
S
s
B
H
S
R
H
H
Y
D
1
S
U
p
p
p
E
N
P
E
U
Z
T
T
A
N
U
M
D
L
P
S
A
T
s
B
H
s
R
S
Y
R
M
N
B
s
L
Y
Ma
S
N
P
R
U
z
B
T
A
N
u
M
U
R
M
0
A
u
L
u
S
Y
M
M
N
A
s
L
Y
Ma
S
M
F
E
N
L
M
P
M
U
R
E
0
L
u
L
M
T
M
M
atrazine
X
X
X
X
azinphos-methyl
X
X
X
X
X
X
X
X
bendiocarb
X
X
X
X
bensulide
X
X
X
X
benzamide
X
X
X
X
captan
X
X
X
X
carbaryl
X
X
X
X
carbofuran
X
X
X
X
alpha-chlordane
X
X
X
X
X
X
gamma-chlordane
X
X
X
X
X
X
chlorpyrifos
X
X
X
X
X
X
chlorthal-dimethyl
X
X
X
X
cypermethrin
X
X
X
X
4,4-' DDD
X
X
X
X
X
X
4,4-' DDE
X
X
X
X
X
X
4,4-' DDT
X
X
X
X
X
X
diazinon
X
X
X
X
X
X
dichlorvos
X
X
X
X
X
X
X
X
F-4
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Pesticide
Household Dust
School Dust
W
W
W
W
W
W
w
w
w
w
w
S
S
S
S
S
S
s
s
s
S
s
C
C
C
D
D
D
D
0
0
0
p
w
w
W
w
w
w
w
w
w
w
w
H
H
Y
D
I
U
U
p
p
p
E
c
c
c
D
D
D
D
0
0
0
p
D
L
P
S
A
S
s
B
H
s
R
H
H
Y
D
I
S
U
p
p
p
E
N
P
E
U
Z
T
T
A
N
u
M
D
L
P
S
A
T
s
B
H
s
R
S
Y
R
M
N
B
s
L
Y
Ma
S
N
P
R
U
Z
B
T
A
N
u
M
U
R
M
0
A
u
L
u
S
Y
M
M
N
A
s
L
Y
Ma
S
M
F
E
N
L
M
P
M
U
R
E
0
L
u
L
M
T
M
M
dicofol
X
X
X
X
dieldrin
X
X
X
X
disulfoton
X
X
X
X
X
X
X
X
endosulfan 1
X
X
X
X
endosulfan 2
X
X
X
X
ethyl parathion
X
X
X
X
X
X
X
X
folpet
X
X
X
X
fonophos
X
X
X
X
X
X
heptachlor
X
X
X
X
hexachlorobenzene
X
X
X
X
lindane
X
X
X
X
malathion
X
X
X
X
X
X
X
X
methyl parathion
X
X
X
X
X
X
X
X
methoxychlor
X
X
X
X
metolachlor
X
X
X
X
pendimethalin
X
X
X
X
cis-permethrin
X
X
X
X
trans-permethrin
X
X
X
X
o-phenylphenol
X
X
X
X
phorate
X
X
X
X
X
X
X
X
F-5
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Pesticide
Household Dust
School Dust
W
W
W
W
W
W
w
w
w
W
w
S
S
S
S
S
S
S
s
S
S
s
C
C
C
D
D
D
D
0
0
0
p
w
w
W
w
w
w
w
w
w
W
w
H
H
Y
D
1
U
U
p
p
P
E
c
c
c
D
D
D
D
0
0
0
p
D
L
P
S
A
S
S
B
H
S
R
H
H
Y
D
1
S
U
p
p
p
E
N
P
E
U
Z
T
T
A
N
U
M
D
L
P
S
A
T
S
B
H
s
R
S
Y
R
M
N
B
S
L
Y
Ma
S
N
P
R
U
Z
B
T
A
N
u
M
U
R
M
0
A
u
L
u
S
Y
M
M
N
A
S
L
Y
Ma
S
M
F
E
N
L
M
P
M
U
R
E
0
L
u
L
M
T
M
M
prometryn
X
X
X
X
propoxur
X
X
X
X
simazine
X
X
X
X
terbufos
X
X
X
X
X
X
X
X
trifluralin
X
X
X
X
a Fonophos was inadvertently excluded from the OP sum for household and school dust samples; however, most of its measurements were below detection limit. Thus, analysis results should
not have been significantly affected.
X indicates that molar-weighted pesticide concentration was included in sum.
F-6
August 2005
-------
Appendix G
Data Mining Methodology and Results for the Yuma Study
-------
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Appendix G
Data Mining Methodology and Results for the Yuma Study
Contents
G. 1 Introduction G-1
G.2 Methodology for the Data Mining Approach G-1
G.2.1 Stage 1 - Data Preparation G-l
G.2.1.1 Data Format and Code Value Assignment G-l
G.2.1.2 Additional Changes in Questionnaire Variables G-2
G.2.1.3 Conditional Questions G-5
G.2.1.4 Non-Response Categories G-6
G.2.1.5 Changes in Analytical Measurement Variables G-l
G.2.2 Stage 2 - Review of Basic Relationships G-9
G.2.2.1 Simple Indicators of Exposure Levels G-9
G.2.2.2 An Underlying Structure G-10
G.2.3 Subpopulations Selected for Analysis G-l 1
G.2.4 Stage 3 - Classification Approach G-l 1
G.2.4.1 Classification Techniques G-l 1
G.2.4.2 Sample Classification Tree Output G-12
G.3 Results from Data Mining Approach - CART Analyses G-20
G.3.1 Simple Indicators of Exposure Levels G-20
G.3.2 Underlying Structure - Principal Component Analysis G-21
G.3.3 Classification Approach - CART Analyses G-25
G.3.4 Comparison of Questionnaire Responses for High- and Low-end
Measurements G-91
G-i August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Tables
Table G.2.1 Questionnaire Variables from the Yuma Study Used in Data Mining
Analyses, Sorted Alphabetically G-2
Table G.2.2 Example of Questionnaire Variable CHEMINHS with Code Values
Assigned for "No Response" and "Not Applicable" Where Exposure
Impact Is Less Likely with the "No" Response than with the "Yes"
Response G-7
Table G.2.3 Example of Questionnaire Variable VEGGIES with Code Values
Assigned for "No Response" and "Do Not Know" Where Exposure
Impact Is Less Likely When Child Eats Locally Grown Fresh Fruit
Fewer Times Per Year G-7
Table G.2.4 Analytical Measurement Variables Used in the Yuma Study Data
Mining Analyses G-8
Table G.3.1 An Underlying Structure of the Yuma Study Questionnaire and
Measurement Variables Based on Principal Component Analysis
Under Two Scenarios G-21
Table G.3.2 Description of and Cross-Reference for CART Analyses Performed on
Yuma Study Data G-26
Table G.3.3 Code Values Assigned to Ordinal and Categorical Questionnaire
Variables G-27
Table G.3.4 Results of Classifying Yuma Study Children's Measurements of
LWETHSUM [Log(WETHSUM)] for Six Scenarios of Predictors G-71
Table G.3.5 Results of Classifying Yuma Study Children's Measurements of
LWMETHSM [Log(WMETHSUM)] for Six Scenarios of Predictors. G-78
Table G.3.6 Results of Classifying Yuma Study Children's Measurements of
LWETHSUM [Log(WETHSUM)] for Six Scenarios of Predictors
Including and Excluding CHLDTM3 (Questions Sorted
Alphabetically) G-84
Table G.3.7 Predictors Selected for Classifying Yuma Study Children's
Measurements of LWETHSUM [Log(WETHSUM)] for Six Scenarios
of Predictors Including and Excluding CHLDTM3 G-90
Table G.3.8 Questions and Weights Used to Create the Exposure Weighted Sum
for Comparing High and Low End Measurements G-92
Table G.3.9 Results from Non-statistical Comparison of Questionnaire Responses
between High and Low End Measurements G-96
G-ii
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Figures
Figure G.2.1 Example CART Analysis of LWETHSUM [LOG(WETHSUM)] with
All Questions and House and School Dust Measurements for Yuma
Study (131 participants): CART Tree G-14
Figure G.2.2 Example CART Analysis of LWETHSUM[LOG(WETHSUM)] with
All Questions and House and School Dust Measurements Yuma Study
(131 Participants): Summary Statistics for Nodes in CART Tree
(Figure G.2.1) G-16
Figure G.3 .1.a CART Analysis of LWETHSUM [LOG(WETHSUM)] with All
Questions for 130 Yuma Study Participants: CART Tree G-35
Figure G.3 .1.b CART Analysis of LWETHSUM [LOG(WETHSUM)] with All
Questions for 130 Yuma Study Participants: Summary Statistics for
Nodes in CART Tree (Figure G.3.1.a) G-37
Figure G.3 .2.a CART Analysis of LWETHSUM [LOG(WETHSUM)] with Limited
Questions for 130 Yuma Study Participants: CART Tree G-38
Figure G.3 .2.b CART Analysis of LWETHSUM [LOG(WETHSUM)] with Limited
Questions for 130 Yuma Study Participants: Summary Statistics for
Nodes in CART Tree (Figure G.3.2.a) G-40
Figure G.3 .3 .a CART Analysis of LWETHSUM [LOG(WETHSUM)] with All
Questions and House Dust Measurements for 130 Yuma Study
Particpants: CART Tree G-41
Figure G.3 .3 .b CART Analysis of LWETHSUM [LOG(WETHSUM)] with All
Questions and House Dust Measurements for 130 Yuma Study
Participants: Summary Statistics for Nodes in CART Tree (Figure
G.3.3.a) G-43
Figure G.3 .4.a CART Analysis of LWETHSUM [LOG(WETHSUM)] with Limited
Questions and House Dust Measurements for 130 Yuma Study
Participants: CART Tree G-44
Figure G.3.4.b CART Analysis of LWETHSUM [LOG(WETHSUM)] with Limited
Questions and House Dust Measurements for 130 Yuma Study
Participants: Summary Statistics for Nodes in CART Tree (Figure
G.3.4.a) G-46
Figure G.3 .5 .a CART Analysis of LWETHSUM [LOG(WETHSUM)] with All
Questions and House and School Dust Measurements for 130 Yuma
Study Participants: CART Tree G-47
Figure G.3 .5 .b CART Analysis of LWETHSUM [LOG(WETHSUM)] with All
Questions and House and School Dust Measurements for 130 Yuma
Study Participants: Summary Statistics for Nodes in CART Tree
(Figure G.3.5.a) G-49
Figure G.3 .6.a CART Analysis of LWETHSUM [LOG(WETHSUM)] with Limited
Questions and House and School Dust Measurements for 130 Yuma
Study Participants: CART Tree G-50
Figure G.3 .6.b CART Analysis of LWETHSUM [LOG(WETHSUM)] with Limited
Questions and House and School Dust Measurements for 130 Yuma
G-iii
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Study Participants: Summary Statistics for Nodes in CART Tree
(Figure G.3.6.a) G-52
Figure G.3.7.a CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All
Questions for 130 Yuma Study Participants: CART Tree G-53
Figure G.3.7.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All
Questions for 130 Yuma Study Participants: Summary Statistics for
Nodes in CART Tree (Figure G.3.7.a) G-55
Figure G.3.8.a CART Analysis of LWMETHSM [LOG(WMETHSUM)] with
Limited Questions for 130 Yuma Study Participants: CART Tree G-56
Figure G.3.8.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with
Limited Questions for 130 Yuma Study Participants: Summary
Statistics for Nodes in CART Tree (Figure G.3.8.a) G-58
Figure G.3.9.a CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All
Questions and House Dust Measurements for 130 Yuma Study
Participants: CART Tree G-59
Figure G.3.9.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All
Questions and House Dust Measurements for 130 Yuma Study
Participants: Summary Statistics for Nodes in CART Tree (Figure
G.3.9.a) G-61
Figure G.3.10.a CART Analysis of LWMETHSM [LOG(WMETHSUM)] with
Limited Questions and House Dust Measurements for 130 Yuma
Study Participants: CART Tree G-62
Figure G.3.10.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with
Limited Questions and House Dust Measurements for 130 Yuma
Study Participants: Summary Statistics for Nodes in CART Tree
(Figure G.3.10.a) G-64
Figure G.3.11.a CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All
Questions and House and School Dust Measurements for 130 Yuma
Study Participants: CART Tree G-65
Figure G.3.1 l.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All
Questions and House and School Dust Measurements for 130 Yuma
Study Participants: Summary Statistics for Nodes in CART Tree
(Figure G.3.11.a) G-67
Figure G.3 .12.a CART Analysis of LWMETHSM [LOG(WMETHSUM)] with
Limited Questions and House and School Dust Measurements for 130
Yuma Study Participants: CART Tree G-68
Figure G.3 .12.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with
Limited Questions and House and School Dust Measurements for 130
Yuma Study Participants: Summary Statistics for Nodes in CART
Tree (Figure G.3.12.a) G-70
G-iv
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
G.l Introduction
The second approach performed to evaluate the effectiveness of questions for predicting a
child's pesticide exposure level was an analysis of a recent children's pesticide exposure
study in Yuma, Arizona, which included questionnaires and measurements. Some aspects of
the Yuma Study are described in section 4.3.1, and a report evaluating relationships that
address a priori hypotheses has been published (CDC 2002).
As a supplement to the initial findings in the Yuma Study report, the study's data were also
evaluated using data mining. Data mining describes an analysis approach that searches
through data for relationships that may or may not be defined a priori. This technique is
exploratory in nature, in comparison to a confirmatory analysis that is interested in
determining whether a proposed relationship adequately explains the observed set of data
(Hand 1999). In this appendix the methodology for processing and analyzing the Yuma
Study data under the data mining approach is described, and detailed results from the
analyses which are summarized in section 4.3.2 are presented.
G.2 Methodology for the Data Mining Approach
The data mining approach focused on identifying relationships that would be useful in
classifying children by their organophosphate (OP) pesticide exposure level, and would at
least have a higher likelihood of being able to identify children with high or low exposure
levels. The first stage of this approach prepared the data for analysis, the second stage
reviewed basic relationships in the data, and the third stage performed classification type
analyses. The data manipulation and analysis steps were carried out with SPSS versions 11.5
and 12.0 (SPSS, Inc., Chicago, IL), and S-Plus version 6 (Insightful, Inc., Seattle WA).
G.2.1 Stage 1 - Data Preparation
Data from the Yuma Study were reviewed to determine the types of analyses to be
performed. Adjustments were made to the data only to facilitate analyses, and not to change
the intent of any responses. These adjustments included changes in data formats, the addition
of code values to describe certain situations, the creation of additional variables based on the
original data, and the identification of subgroups within the study to be used for the analyses.
Steps were taken to ensure the quality of any changes made to the data and for any additional
variables created.
G.2.1.1 Data Format and Code Value Assignment
Questionnaires usually include questions with defined sets of responses, such as Yes and No,
and questions with no predefined set of responses (open-ended responses), such as brands of
pesticide used. Several questions from the Yuma Study allowed for the latter type of
responses. The unique responses given for these questions were assigned code values in new
variables, although subsequently these questions were not included in the analyses. For all
questions, code values were also assigned when some form of a no-response was given (e.g.,
don't know or response missing). The code value assigned in the Yuma Study to identify
G-l
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
urine and dust analytical measurements below the detection limit (BDL) was changed to
better fit with the types of data analyses performed for this report. A measurement value of
O.Olwas assigned for a BDL dust concentration, and a value of 0.001 was assigned for a BDL
urinary metabolite concentration. These values were assigned before creating the molar-
weighted concentrations described in Appendix F.
G.2.1.2 Additional Changes in Questionnaire Variables
To facilitate the analyses in Stages 2 and 3, additional variables based on the original data in
the study were created (Table G.2.1). For example, questions about where the child routinely
spent their time, and where medical care was received, allowed for multiple responses to be
selected from a predefined list. In these check-all-that-apply questions, each response
checked was included in a separate variable. For example, if a person checked responses B
and D, response B would be noted under the first variable, and response D under the second.
If a person checked responses A, B, and E, response A would be noted under the first
variable, response B under the second variable, and response E under the third. Based on
these two examples, response B could be found under more than one variable, a situation that
did not allow for an easy analysis of the responses based on this project's objectives. Thus
an additional set of variables was created where each option in the predefined list was
associated with one variable. A value of 1 was assigned if the response was selected by the
household, and a value of 0 was assigned if the response was not selected. For example, the
set of variables CHLDTM1 to CHLDTM7 was created to capture the seven possible
responses to where the child routinely spent their time.
In some instances, summaries or revised definitions of the original variables were created.
For example, a variable was created to describe the number of rooms sprayed or treated with
pesticides (NRMSPRYD) based on the rooms included in the questionnaire. Questions
relating to the mother's and father's occupation and whether pesticides were used on the job
were reviewed, and additional variables, DADCON2 and MOMCON2, were created to
describe whether the parent's job was indoors with or without pesticide use, or outdoors with
or without pesticide use.
Table G.2.1 Questionnaire Variables from the Yuma Study Used in Data Mining Analyses, Sorted
Alphabetically
Type3
Name
Brief Description
Extended Description
Original
ADLTPEST
Non-parent in home works where
pesticides used?
Is there another person living in the house
(other than parent) who works in a place
where pesticides are used?
Additional
ADTPSWK"
Any adult works where pesticide used?
Any adult in household works where
pesticides used?
Original
AGEb
Age of principal child
Age of principal child calculated from date of
birth
Original
BASEMENT
Basement treated with pesticides?
Was basement treated with pesticides?
Original
BATHROOM
Bathroom treated with pesticides?
Was bathroom treated with pesticides?
Original
BEDROOM
Bedroom treated with pesticides?
Was bedroom treated with pesticides?
G-2
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Type3
Name
Brief Description
Extended Description
Original
CHEMINHS
Pesticides used inside home last
month?
Were chemicals to control insects used
inside the house during the last month?
Original
CHEMOUTH
Pesticides used outside home last
month?
Were chemicals to control insects used on
the exterior or foundation of the house
during the last month?
Original
CHILDBED
Child's bedroom treated with
pesticides?
Was child's bedroom treated with
pesticides?
Original
CHILDFLD
Child worked in fields last month?
Has principal child been to the work field(s)
during past month?
Additional
CHLDTM1
Child spends time in another home?
Principal child routinely spends time away
from home - in another home
Additional
CHLDTM2
Child spends time at day care center?
Principal child routinely spends time away
from home - at day care center
Additional
CHLDTM3
Child spends time at school?
Principal child routinely spends time away
from home - at school
Additional
CHLDTM4
Child spends time at sport event?
Principal child routinely spends time away
from home - at sport event
Additional
CHLDTM5
Child spends time playing in field?
Principal child routinely spends time away
from home - playing in field
Additional
CHLDTM6
Child spends time playing in irrigation
water?
Principal child routinely spends time away
from home - playing in irrigation water
Additional
CHLDTM7
Child spends time playing outside?
Principal child routinely spends time away
from home - playing outside
Original
CLOSEAPP
Distance between home and nearest
application of pesticides
In past month, how close to participant's
home was the nearest application of
agricultural or gardening chemicals?
Additional
DADCON2
Father's occupation - location and
pesticide use
Does father work indoors or outdoors and
with or without pesticides?
Additional
DADPESTC
Are pesticides used where father
works?
Are pesticides used where father works? -
categories
Original
DADWORK0
Is the father currently employed?
Is the father currently employed?
Original
DININGRM
Dining room treated with pesticides?
Was dining room treated with pesticides?
Original
ETHNIC
Child's ethnic and racial background
Child's ethnic and racial background
Original
FAMILYRM
Family room treated with pesticides?
Was family room treated with pesticides?
Original
FARFIELD
Distance between home and
agricultural field
How far is participant's home from a field
where crops are grown?
Original
GPSb
Distance between home and field using
GPS
Distance from home to field using GPS
measurement categories
Original
GRADE
Child's grade
What grade is the principal child in?
Original
HEIGHT
Child's height (inches)
Measurement of principal child's height
without shoes (inches)
Original
HOURAWAY
Number hours/wk child not at home
During school year, about how many hours
per week does principal child spend away
from home?
Original
HOWCHEMO
How pesticides were applied to fields
How were agricultural chemicals applied to
field close to participant's home?
Original
HOWCHILD
Child's health in general
Description of principal child's health in
general
Original
INSURED
Is child covered by medical insurance?
Is principal child covered by medical
G-3
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Type3
Name
Brief Description
Extended Description
insurance?
Original
KITCHEN
Kitchen treated with pesticides?
Was kitchen treated with pesticides?
Original
LICE
Child treated for head lice past six
months?
Has principal child been treated for head lice
in past six months?
Original
LIVEAREA"
Children/respondent live in area part-
time
Children/respondent live in area < 10
months/year
Original
liveyear"
Number of years child lived at this
address
Number of years child lived at this address
Original
LIVINGRM
Living room treated with pesticides?
Was living room treated with pesticides?
Additional
MOMCON2
Mother's occupation - location and
pesticide use
Does mother work indoors or outdoors and
with or without pesticides?
Additional
MOMPESTc
Are pesticides used where mother
works?
Are pesticides used where mother works? -
categories
Original
MOMWORK0
Mother now employed (not as
housewife)?
Is the mother currently employed?
Original
NCATWRKD
Father's occupation - categories
Father's occupation - categories
Original
NCATWRKM
Mother's occupation - categories
Mother's occupation - categories
Additional
NRMSPRYD
Number of rooms sprayed last month
Number of rooms in house sprayed with
pesticides in past month
Original
NUMADLTS"
Number of additional adults in home
Number of non-parent adults in home
working with pesticides
Original
OFTCHEMI
How often is home treated for pests?
How often is participant's home treated for
pests?
Original
OTHERRM
Other rooms treated with pesticides?
Were other rooms in the house treated with
pesticide?
Original
PEOPLIVE"
Number people in household including
participant
Number people in household including
participant
Original
POISON
Anyone treated for pesticide poison?
Has anyone in the household been treated
for pesticide poisoning in past year?
Original
SCHOOL
Child's school
School where principal child attends
Original
SEX
Child's gender
Gender of principal child
Original
SPRAYFLD
Child in yard when fields sprayed or
dusted?
Does principal child play outside in the yard
when the fields are sprayed or dusted?
Original
VEGGIES
How often child eats local fresh
fruit/veg?
During the year, how often does principal
child eat locally grown fresh fruits or
vegetables?
Original
WASHVEGI
How often wash local fresh fruit/veg
before eating?
How often are the locally grown fresh fruits
and vegetables washed before they are
eaten?
Original
WATERSR1
Drinking water source -
public/commercial
Source of drinking water in participant's
home is public/commercial
Original
WATERSR2
Drinking water source - private well
Source of drinking water in participant's
home is private well
Original
WATERSR3
Drinking water source - cistern
Source of drinking water in participant's
home is cistern
Original
WEIGHT
Child's weight (lbs)
Measurement of principal child's weight
without shoes or other heavy articles (lbs)
Original
WHEEL
Distance between home and field -
Distance from home to field measured with
G-4
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Type3
Name
Brief Description
Extended Description
rotary wheel
rotary wheel - categories
Original
WHENFILD
Last time child was in work field
When was the last time principal child was in
the work field?
Additional
WHERMD1
Family med care at private medical
clinic
Where principal child's family receives
medical care - private medical clinic
Additional
WHERMD2
Family med care at health dept clinic
Where principal child's family receives
medical care - local health department clinic
Additional
WHERMD3
Family med care at other med clinic
Where principal child's family receives
medical care - other medical clinic
Additional
WHERMD4
Family med care in Mexico
Where principal child's family receives
medical care - Mexico
Additional
WHERMD5
No access to medical care
Where principal child's family receives
medical care - no access
Additional
WHERMD6
Family med care at other place
Where principal child's family receives
medical care - at other facility
Additional
WHERMD7
Family med care - do not know
Where principal child's family receives
medical care - do not know
Original
WHERTIME
Room where child spends most awake
time
Room where principal child spends most of
their awake time
Original
WHNCHEMO
Last time field treated with pesticides?
When was the last time the field was
sprayed or treated with pesticides?
Original
WHOCHEMI
Who applied pesticides inside the
house?
Who applied chemicals inside the house?
Original
WHOCHEMO
Who applied pesticides outside house?
Who applied chemicals outside house?
Original
Y0UNGSIB3
Number of children in household < 11
years old
Number of additional children in household
< 11 years old
a Original variables existed in the data set provided from the Yuma Study. Additional variables were created based on the
original variables.
b Questionnaire variable included in the Principal Component Analysis (section G.2.2.2) but not in the CART analyses (section
G.2.4.1).
b Questionnaire variable included in the CART analyses (section G.2.4.1) but not in the Principal Component Analysis (section
G.2.2.2).
G.2.1.3 Conditional Questions
Most questionnaires use conditional questions, that is, questions that are or are not asked of
participants based on their response to a previous question. These questions are part of skip
patterns in a questionnaire's administration. An example of a conditional pairing in the
Yuma Study is the question CHILDFLD, Has the child been to work in the fields in the past
month? and WHENFILD, When was the last time your child was in the work field? If the
response to CHILDFLD (the condition question) is "No", then WHENFILD (the conditional
question) was not asked. To ensure that responses on conditional questions accurately and
consistently reflected the response to the condition question, responses to the conditional
questions that were skipped were coded with a "Not Applicable (NA)" response. The
consistent coding between the condition and conditional questions maintains the relatedness
between the paired questions: however, for condition questions with a large number of "No"
responses, the distribution for the conditional question will be heavily weighted with NA
responses. This type of recoding has an impact on the analysis that needs to be recognized.
G-5
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
G.2.1.4 Non-Response Categories
Questionnaires usually include non-response categories to handle response outcomes such as
"Missing," "Refused," "Not Applicable," and "Don't Know." The Yuma Study allowed for
Don't Know, Refused-to-Answer, and Missing as non-response categories. The very small
number of Refused-to-Answer responses was combined with the Missing responses into a No
Response category. Even though the distinctions are somewhat gray, the Don't Know and
No Response categories were assigned distinct code values in order to evaluate whether they
represented groups with different exposure levels.
Data with non-responses offer analysis challenges. When analyzing one questionnaire
variable with non-responses, the cases with No Response can be excluded from the analysis,
or the cases with responses can be compared to those with No Response to determine if they
have different measurement values. The chosen solution becomes more complex when the
analysis includes several questionnaire variables simultaneously. If only the cases with
responses to all questions are analyzed, the number of cases can decrease to levels that are
not necessarily representative of the population measured. If variables with some non-
responses are excluded from the analysis, potentially useful information from the responses
is discarded. Imputation is sometimes used as a solution for analyzing such incomplete data
sets; however, because these methods usually require an a priori knowledge of relationships
between the variables, it was not considered a feasible option for this project.
An approach for handling non-responses in the statistical analyses was designed. Numeric
codes were assigned to the non-response categories to preserve the sample size in the
analysis, and to allow the investigation of differences between respondents and non-
respondents, if desired. Variables that had nominal categories were assigned numeric codes
for analysis (e.g., l=yes, 2=no). The basis for the coding approach considers a variable's
categories or values as a continuum of relative impact to exposure. This continuum could be
applied to categories with an underlying ordinality, or to numeric values. Code values for the
"Not Applicable," "Don't Know," and "No Response" categories were assigned to be
consistent with the question's continuum of exposure impact as follows:
• category with most impact on exposure level
•
• category with least impact on exposure level
• Don't Know (assumes no potential impact)
• Not Applicable (implies no potential impact)
• No Response (assumes no potential impact).
The values assigned to the non-response categories depended on which categories actually
had responses and the values already assigned to the response categories. Tables G.2.2 and
G.2.3 illustrate this coding scheme with two examples.
G-6
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table G.2.2 Example of Questionnaire Variable CHEMINHS with Code Values Assigned for "No
Response" and "Not Applicable" Where Exposure Impact Is Less Likely with the "No"
Response than with the "Yes" Response
Code
Description
1
Yes
2
No
3
Not Applicable
4
No Response
Table G.2.3 Example of Questionnaire Variable VEGGIES with Code Values Assigned for "No
Response" and "Do Not Know" Where Exposure Impact Is Less Likely When Child Eats
Locally Grown Fresh Fruit Fewer Times Per Year
Code
Description
-1
No Response
0
Do Not Know
1
Never
2
About once a year
3
About once a month
4
About once a week
5
About once a day
The number of non-responses in the Yuma Study was small for the variables analyzed. Thus
analysis comparing differences between respondents and non-respondents on a particular
question was not pursued. The coding scheme for non-responses has limitations; however, it
provided an underlying ordinality for the variables, where needed, and facilitated subsequent
analyses.
G.2.1.5 Changes in Analytical Measurement Variables
The analytical measurement (AM) data were reviewed with preliminary analyses, and
additional AM variables were created based on the original measurement variables.
Variables were created for each chemical to indicate if the concentration value was above or
below predefined cut-points. The cut-point used for the urinary metabolites was the 90th
percentile, based on EPA's definition of a high-exposure level (USEPA 1992). Many
pesticides in the dust samples had a large proportion of concentrations that were below the
limit of detection (BLD). The 80th percentile was used as the cutpoint for all dust chemicals
to broaden the range of higher concentration values.
G-7
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The low number of detectable measurements for many of the dust chemicals and urine
metabolites suggested the need for a second type of variable, the sum of molar-weighted
measurements for groups of related chemicals or metabolites. Several chemicals with more
detectable measurements such as chlorpyrifos, diazonin, cypermet, and o-phenylphenol were
included in some of the weighted sums and were defined as individual molar-weighted
variables. Appendix F describes the molar-weighting process and the chemicals/metabolites
included in the molar-weighted sums. Some examples of the summed measurements are the
sum of ethylated DAPs (DEP + DETP + DEDTP), all dust chemicals, and all available OP
dust chemicals. Table G.2.4 shows the list of AM variables used in the Stage 2 and Stage 3
analyses. Based on their distributions, the AM data were log-transformed, where appropriate
for the statistical analysis performed.
Table G.2.4 Analytical Measurement Variables Used in the Yuma Study Data Mining Analyses
Name
Description3
Urine from Child
LWETHSUM
Log of weighted sum of DEP, DETP, and DEDTP (adjusted for creatinine)13
LWMETHSM
Log of weighted sum of DMP, DMTP, and DMDTP (adjusted for creatinine)0
Household Dust
WCHDNSUM
Weighted sum of alpha-chlordane and gamma-chlordane
WCHLPYRF
Weighted chlorpyrifos
WCYPERMETd
Weighted cy-permethrin
WDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and 4,4'DDT
WDIAZNON
Weighted diazinon
WDUSTBAL
Weighted sum of dust analytes except OP pesticides
WDUSTSUM
Weighted sum of all dust analytes
WOPBAL
Weighted sum of OP pesticides except chlorpyrifos, diazinon, permethrins, and
o-phenylphenol
WOPHNYLPd
Weighted o-phenylphenol
WOPSUM
Weighted sum of OP pesticides
WPERMSUM
Weighted sum of cis-permethrin and trans-permethrin
School Dust
SWCHDNSM
Weighted sum of alpha-chlordane and gamma-chlordane
SWCHLPYR
Weighted chlorpyrifos
SWCYPRMEd
Weighted cy-permethrin
SWDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and 4,4'DDT
SWDIAZNO
Weighted diazinon
G-8
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Name
Description3
SWDSTBAL
Weighted sum of dust analytes except OP pesticides
SWDUSTSM
Weighted sum of all dust analytes
SWOPBAL
Weighted sum of OP pesticides except chlorpyrifos, diazinon, permethrins, and
o-phenylphenol
SWOPHNYLd
Weighted o-phenylphenol
SWOPSUM
Weighted sum of OP pesticides
SWPERMSM
Weighted sum of cis-permethrin and trans-permethrin
a See Appendix F for detailed descriptions.
b DEP = diethylphosphate, DETP = diethylthiophosphate, DEDTP = diethyldithiophosphate
c DMP = dimethylphosphate, DMTP = dimethylthiophosphate, DMDTP = dimethyldithiophosphate.
d Measurement variable included in Principal Component Analysis (section G.2.2.2) but not in CART analyses
(section G.2.4.1).
Twenty-five school dust samples were taken at six of the participating schools. These
measurements from a specific school and grade were added to the records of all principal
participants in that school and grade. If more than one room was sampled within a school for
a particular grade level, the school dust measurements for the school/grade combination were
averaged before being merged with the principal participant's information.
G.2.2 Stage 2 - Review of Basic Relationships
G.2.2.1 Simple Indicators of Exposure Levels
As an initial evaluation of potential predictors of high exposure levels, three types of
analyses were performed between the chemicals/metabolites and the questionnaire or
grouping variables: contingency table analysis, Kruskal-Wallis (non-parametric one-way
analysis of variance), and the median test. The two-level AM variables (section G.2.1.5)
used in the contingency table analyses described the concentrations that were "high" and not
high. The actual AM measurement values were used for the other analyses. Forty-nine of the
original questionnaire variables in the categories of demographics, residential pesticide use,
local pesticide use, activities of the principal participant, drinking water, parents' work
environment, and medical care, were used as grouping variables for the analyses. An
example of a grouping variable in the Kruskal-Wallis analysis is the variable that divides the
data into the subgroups (e.g., male and female) whose measurements are compared in the
analysis. Since house dust measurements are potential indicators of exposure, rank
correlations between the dust measurements and the urine measurements were also
calculated. These analyses were performed using the questionnaire responses and available
measurements from the 152 principal participants/households. Results from these analyses
are summarized in section G.3.1.
G-9
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
G.2.2.2 An Underlying Structure
Within a set of variables like the questionnaire data, the questions usually cluster into groups
by subject matter. The groups may be predefined based on the questionnaire design, or based
on the relationships that exist between the actual responses. Principal Component Analysis
(PC A) was performed on the Yuma Study data to look at the type of question groupings, or
dimensions, existing in the data set. Looking at this underlying structure in the data helps to
understand relationships that appear or do not appear in the Stage 3 analyses. Knowing
which questions may be measuring similar information can also help reduce the number of
questions used in future studies, or may offer options for surrogate questions to be used..
Note that no one question in a PCA dimension represents the dimension.
PCA is one of the oldest and best known of the multivariate analysis techniques (Hotelling
1933, Jackson 1991, Jolliffe 1986, and Jolliffe 2002). Its central idea is "to reduce the
dimensionality of a data set in which there are a large number of interrelated variables, while
retaining as much as possible of the variation present in the data set" (Jolliffe 1986). PCA
determines principal components (PCs) as linear combinations of the questionnaire variables
that maximally discriminate among the cases. PCs are also known as latent or underlying
variables or dimensions.
PCA derives the PCs in an order based on the magnitude of the eigenvalue, a measure of the
variability accounted for by a PC. The PCs are designed to account for as high a percentage
of variation among the questionnaire variables with as few PCs as possible. The dimensions
in the data described by the PCs are orthogonal to, or uncorrected with, each other. The
result of a PCA is a matrix of loading values for each variable, with respect to each PC that
has been identified. If the data are standardized, that is, the PCA is generated from a
correlation matrix of the variables (as in these analyses) rather than a covariance matrix, the
loading value represents the variable's relative weight in, or importance to, that PC.
Subsequently the loading values range from -1.0 to 1.0, and represents a variable's
correlation with the PC.
Traditionally PCA is performed on continuous or at least ordinal type variables; however,
most of the questionnaire variables in the Yuma Study are categorical. Jolliffe (1986, 2002)
confirms that when using PCA as a descriptive, rather than inferential, technique,
assumptions about the type of data included are not required. He also notes that although
linear functions of nominal or non-continuous variables may be harder to interpret, "the basic
objective of PCA, to summarize most of the 'variation' which is present in the original set of
p variables, using a smaller number of composite variables (i.e., PC scores) can be achieved
regardless of the nature of the original variables."
A varimax rotation is used with the PCA to produce PCs which are more parsimonious, and
PCs with eigenvalues > 0.7 rather than >1.0 (Jolliffe 1986) are reviewed. Variables are
associated with a specific rotated PC based on their correlation with it. Loading values from
the PC As range from -1.0 to 1.0, and variables with an absolute loading value > 0.6 for a PC
were considered descriptive of that component, that is, having a medium range of correlation
with the PC. Additional details for a similar use of PCA can be found in USEPA (2004a).
G-10
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
G.2.3 Subpopulations Selected for Analysis
The analyses planned for Stages 2 and 3 were different than those performed in CDC (2002),
because these analysis techniques were not amenable to handling within-household
variability, and the objectives of the analyses were different. The Yuma Study included
principal participants and siblings, however, questionnaire responses were collected only for
the principal participants, and only demographic data, such as gender, age, height, and
weight, were collected for the siblings. Although certain household questionnaire responses
would apply to both principal participants and siblings (e.g.., questions analyzed in
CDC(2002)), the responses for other household questions might not be the same for all of its
participating children. Since several questionnaire responses and the school dust
measurements were not collected for siblings, the analyses reported here were performed
only on data from principal participants. To further limit the impact of factors relating to
children not defined in the initial study design, the principal participants from kindergarten
and first grade, and from the initial eight schools, were selected as the core set of participants
for the data mining analysis (Stage 3). This approach reduced potential confounding factors
that might affect the analysis results. Additionally, only those principal participants (130)
with available and non-suspect urine measurement data were included.
G.2.4 Stage 3 - Classification Approach
In order for future study designs to be able to produce either an enriched population (i.e., a
larger percentage of individuals with higher exposure levels), or to eliminate from further
processing individuals with lower exposure levels, a screening tool must be able to identify
the characteristics of potential study participants that classify them into one category of
interest or the other. The tool does not, however, need to predict the participant's exposure
measurement, as much as an exposure level.
The selected data mining approach creates groups of participants, described by questionnaire
responses and/or measurements, with indicators of the likelihood of their group's exposure
level. Researchers can then identify the groups from the data mining analysis that best fit
their population of interest, and can then compare potential participants to the groups'
descriptors in the screening process. The accuracy of the classification of potential
participants depends on similarities in the populations, the number of respondents used in the
data mining analysis, and the level of classification error in the data mining model. These
caveats are similar to caveats put forth in any predictive modeling situation.
G.2.4.1 Classification Techniques
The data mining technique Classification and Regression Trees (CART) was selected as the
primary type of analysis in this stage. CART is a method of defining subsets of the
population of interest, in this case the Yuma Study principal participant children, where the
between-subset variability of the target or dependent variable is maximized, and the within-
subset variability is minimized. In these analyses, the target or dependent variables are
LWETHSUM and LWMTHSM (Table G.2.4). The subsets are defined in terms of predictor
G-ll
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
or independent variables (questions or other concentrations) that can be nominal, ordinal or
continuous in nature. One of the benefits of CART is that the distributional assumptions for
the target and predictor variables are much less restrictive than for traditional types of
analyses like regression or discriminant analysis that might be considered as analysis options.
The output from CART describes the population subsets in terms of categories or ranges of
values for the predictor variables.
Stage 3 looks at the relationships between a dependent variable and a set of independent
variables to identify a model of predictors and their interactions that optimally classify the
principal participants by their exposure level. The term exposure level is used generically
here to describe the dependent variable of the model whether it is a concentration in
household dust or a personal sample such as urine. A classification map or scheme defines
subsets of the sampled population that have different levels of the dependent variable and
provides characteristics of those subsets in terms of the predictors' values. An example of a
classification map is included below. Data mining techniques like CART are used
considerably, though not exclusively, in consumer preference and health studies (Magidson
1993, Weitlisbath 1999, The Measurement Group Website). Two examples of CART from
the exposure assessment field are USEPA (2004a) and Roy (2003). The mapping or outcome
from CART is called a tree and is similar to a decision tree diagram (Two Crows 1999).
Breiman (1984) describes such mappings as classification trees or regression trees,
depending on whether the dependent variable is categorical or continuous, respectively. The
CART mapping partitions the data into subsets through an iterative and sequential process.
Potential predictors are considered for each subset of the data locally, that is, independent of
what fits for the other subsets. This aspect is different than interactions in traditional
modeling analyses which are defined globally across the data set. The goal of the
classification technique is to define the best set of rules or characteristics for identifying the
class to which a case belongs (Breiman 1984).
CART performs the analysis in a sequential manner that results in the tree output. For a
subset of participants or node of the tree, that is, being considered for further splitting or
subdividing, CART looks across all the predictors, and at all categories or values in each
predictor, to identify the predictor and values that create the next best split into two
additional nodes. The best split is defined as the one producing the largest reduction in
variance for the tree as a whole where variance for the tree is defined as the sum of the
variance of the dependent variable within all of the nodes. Another technique, Chi-Square
Automatic Interaction Detection, CHAID, (Magidson 1993), performs an analysis similar to
CART, but can split each node into more than two subsets. The S-Plus version of CART was
used for the results described below. Pruning and shrinking trees are refining techniques
available for CART and are used to identify an overall "best" model for each target variable.
These techniques, however, were not performed for this study.
G.2.4.2 Sample Classification Tree Output
The tree in Figure G.2.1 is an example of CART output from S-Plus. The tree is based on
131 principal participants with log measurements of the molar-weighted sum of ethylated
DAPs, LWETHSUM. Note that this example is not one of the final analyses described in
G-12
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
section G.3.3. Node 1 is the starting point and includes all 131 participants. Some of the
node numbers in Figure G.2.1 are shown in a larger bolded font and correspond to the node
numbers in Figure G.2.2. Figure G.2.2 lists the characteristics of the splits (predictors and
values) at each node and the mean and standard deviation of the dependent variable for each
node. More details on the tree output follow Figure G.2.2.
G-13
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WETHSUM
CHLDTM3<1.5
1
WEIGH!
<45.685
SCHOOL<2.5
-3.797 -4.684
131
2
WDUSTSU
NCATWI*KD<7.5
SWDSTBA .<4.01061
DADCC
5
SWOPSUIV
9
WPERMSU
N2<1.5
WHERT
HOURA^
ME<6.5
OTHER
m<2.5
-2.295
3
<0.563682
WHNCHf MO<6.5
- EN<2
| | -2.434
-3.012 -3.646 19
-2.098 -3.227
10 11
VAY<41
GRACE<3WOPBAL<( .0002929K
-4.478
KITCh
I
-3.351 -4.017
WOPSUM
-4.196 -3.522 -2.971 -3.555 "3.555 I
0.815759
-2.411 -3.064
534
Figure G.2.1 Example CART Analysis of LWETHSUM [LOG(WETHSUM)] with All Questions and House and School Dust Measurements for
Yuma Study (131 participants): CART Tree
G-14
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
1) root 131 91.37000 -3.352
2) CHLDTM3C1.5 123 78.88000 -3.421
4) WDUSTSUM<57.2905 111 63.72000 -3.503
8) NCATWRKD<7.5 91 51.88000 -3.587
16) SWDSTBAL<4.01061 72 37.41000 -3.462
32) DADCON2C1.5 19 7.97400 -3.847
64) WEIGHT<45.685 8 1.21300 -3.360 *
65) WEIGHT>45.685 11 3.49400 -4.200
130) SCHOOL<2.5 6 1.16900 -3.797 *
131) SCHOOL>2.5 5 0.17690 -4.684 *
33) DADCON2>l.5 53 25.62000 -3.324
66) WHERTIME<6.5 46 21.38000 -3.229
132) HOURAWAY<41 26 8.68300 -3.468
264) WDUSTSUM<5.13875 11 2.24800 -3.828
528) GRADEC1.5 5 0.05032 -4.196 *
529) GRADE>1.5 6 0.96070 -3.522 *
265) WDUSTSUM>5.13875 15 3.96500 -3.205
530) SWOPBALCO.000292916 9 1.55700 -2.971
531) SWOPBAL>0.000292916 6 1.18200 -3.555
133) H OU RAWAY >41 20 9.26200 -2.918
266) CHILDFLD<1.5 6 2.02800 -3.555 *
267) CHILDFLD>1.5 14 3.75200 -2.644
534) WOPSUM<0.815759 9 1.04500 -2.411 *
535) WOPSUM>0.815759 5 1.33600 -3.064 *
67) WHERTIME>6.5 7 1.07700 -3.951 *
17) SWDSTBAL>4.01061 19 9.08600 -4.060
34) WHNCHEMCK6.5 9 2.34400 -4.478 *
35) WHNCHEMO>6.5 10 3.75400 -3.684
70) KITCHEN<2 5 0.48550 -3.351 *
71) KITCHEN>2 5 2.16000 -4.017 *
9) NCATWRKD>7.5 20 8.28100 -3.121
18) WPERMSUM<7.12771 15 4.09700 -3.350
36) OTHERRM<2.5 7 0.17600 -3.012 *
37) OTHERRM>2.5 8 2.41900 -3.646 *
19) WPERMSUM>7.12771 5 1.04100 -2.434 *
5) WDUSTSUM>57.2905 12 7.50400 -2.662
10) SWOPSUM<0.563682 6 1.89600 -2.098 *
11) SWOPSUM>0.563682 6 1.78500 -3.227 *
3) CHLDTM3XL.5 8 2.97500 -2.295 *
G-15
August 2005
-------
Relationships Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WETHSUM) are 10, 3, 534, and 19. Node 131 has the lowest average level of
Log(WETHSUM). The characteristics of participants in these nodes are described below. Table G.3.3 describes the code values for the
questionnaire predictors. Table G.2.4 describes the dust measurement predictors. The nodes are numbered in bold on Figure G.2.1 and
the final split characteristics are bolded in the above tree description.
Node 10 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM > 57.2905, and SWOPSUM < 0.563682. The average level
of WETHSUM for these participants was 0.1227 nmoles/g (Log(WETHSUM) = -2.098).
Node 3 is characterized by participants with CHLDTM3 responses > 1.5, that is, the child did not spend time in school.
These cases had an average level of 0.1008 nmoles/g WETHSUM (Log(WETHSUM) = -2.295). These may be participants who spent
"additional" time in school. See discussion surrounding Table 4.3.17 in Results.
Node 534 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, NCATWRKD < 7.5, SWDSTBAL < 4.1061,
DADCON2 > 1.5, WHERTIME < 6.5, HOURAWAY > 41, CHLDFLD > 1.5, and WOPSUM < 0.815759. The average level of WETHSUM for these
participants was 0.0897 nmoles/g (Log(WETHSUM) = -2.411).
Node 19 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, NCATWRKD > 7.5, and WPERMSUM > 7.12771.
The average level of WETHSUM for these participants was 0.0877 nmoles/g (Log(WETHSUM) = -2.434).
Node 131 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, NCATWRKD < 7.5, SWDSTBAL < 4.1061,
DADCON2 < 1.5, WEIGHT < 45.685 (pounds), and SCHOOL > 2.5. The average level of WETHSUM for these participants was 0.0092
nmoles/g (Log(WETHSUM) = -4.684).
Figure G.2.2 Example CART Analysis of LWETHSUM[LOG(WETHSUM)] with All Questions and House and School Dust Measurements Yuma
Study (131 Participants): Summary Statistics for Nodes in CART Tree (Figure G.2.1)
G-16
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Sixty-three of the questions in Table G.2.1 and 18 of the measurements in Table G.2.4 were
used as potential predictors for this analysis. The tree is grown at the first level by identifying
the predictor (question or dust measurement) that splits the 131 cases into the two most
distinct subsets of participants, that is, the subsets that have the largest mean difference for
LWETHSUM. The first predictor selected is CHLDTM3, the child spends time at school.
The groups are divided at the CHLDTM3 value of 1.5,as noted above the node's two
branches (Figure G.3.3.a). If the case has a CHLDTM3 value <1.5 (Node 2), it falls in the
subset with the lower mean (-3.421) for LWETHSUM. If the case has a CHLDTM3 value >
1.5 (Node 3), it falls in the subset with the higher mean (-2.295) for LWETHSUM.
The process of identifying another predictor that creates the most "distinct" subsets is
performed for each node, independent of the other nodes. Node 3 cannot be further
subdivided given the available predictors, and the splitting options defined for the CART
analysis. (Default options for the S-Plus implementation of CART (Venables 1994) were
used. This node is an example of a terminal node or final subset of the tree. Node 2 is split
by WDUSTSUM, the weighted sum of all dust analytes, at the value of 57.0295 nmoles/g.
Nodes 10 and 11 split Node 5 by values of SWOPSUM, the weighted sum of OPs in school
dust, at the value of 0.563682 nmoles/g. Nodes 8 and 9 split Node 4 by the categories of the
father's occupation, NCATWRKD (Table G.3.3). The number of significant digits shown
for the split points of the measurement variables is determined by the S-Plus program, and is
not indicative of the precision available in the measurement data.
In some cases, a node may be split in an unexpected manner. For example, one might expect
that lower levels of SWOPSUM (Node 10) would be associated with lower levels of
LWETHSUM. But Node 11, with the higher level of SWOPSUM, has the lower mean value.
This seeming inconsistency may be due to one of several factors: the level at which the tree's
growth was stopped, whether the predictor is a surrogate for one that is not available in the
list of questions or measurements, or whether this relationship applies only to this subset, and
is not indicative of the relationship in other subsets. A node is considered terminal if the
CART algorithms cannot split it further given the available predictors and the splitting rules
selected. Nodes 3, 10, 11, 19, 534, and 131 are examples of terminal nodes.
Figure G.2.2 describes each intermediate and terminal node in Figure G.2.1 with additional
information. The table in Figure G.2.2 shows levels of indentation for contiguous node
splits. Each node in the tree (Figure G.2.1) is represented in the table (Figure G.2.2);
however, not all of the node numbers in the tree are shown in larger bolded font. Node 1 is
considered the root node in the table. At the bottom of Figure G.2.2, Node 10, for example,
shows the splitting value from node 5 to be SWOPSUM < 0.563682 nmoles/g. Node 10 has
six cases, a standard deviation for LWETHSUM of 1.896, and a mean value of LWETHSUM
of -2.098. The * indicates that Node 10 is a terminal node. To identify the rest of Node 10's
characteristics, follow the rows above node 10 with less indentation levels. Node 5 is the
next left-most indentation, and indicates the participants had WDUSTSUM values > 57.2905
nmoles/g. The next left-most indentation from Node 5 is Node 2 (at the top of the list),
which describes the characteristic that CHLDTM3 <1.5 (YES). Thus the complete
description of participants in Node 10 is: CHLDTM3 <1.5, WDUSTSUM > 57.2905
nmoles/g, and SWOPSUM < 0.563682 nmoles/g.
G-17
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
At this point, pruning or shrinking techniques could be used to define a best model since not
every division of nodes makes a significantly better model. These techniques would be
comparable to the deletion of predictors in a stepwise regression analysis after they have
been included in the model: however, for this project, pruning and shrinking were not
performed. The tree produced from the initial CART analysis was accepted as an indicator
of the predictors useful in classifying a child's exposure level. Only the predictors selected
and not the specific split points for the selected predictors were considered important for this
project's broad objective which is to identify useful predictors rather than to create a
predictive model.
Several factors impact how well the resulting tree is able to classify the subsets and which
predictors are selected for classification. If the distribution of the dependent variable values
covers a small range of close-knit measurements, it may be difficult for the technique to
identify distinctions among the values. In this case, the technique may pick up minute
distinctions, and the resulting predictors may seem inconsistent with expectations. The level
of measurement values can also affect the interpretation or validity of the analysis. If the
majority of values are not of a practical significance in measuring a child's pesticide
exposure level, that is, they are BDL or less than some predefined threshold, the predictors
resulting from the analysis may be spurious and not useful. The number of cases and the
predictors included in an analysis can also impact the predictors selected for the tree. The
smaller the number of cases analyzed, the less information is available to the algorithms to
ground the relationships between predictors and the target or dependent variable.
When comparing predictors selected under different scenarios, it is also important to consider
which predictors were available for the analysis. For example, the analyses of questions with
LWETHSUM as the dependent variable will show differences because one scenario (ALL)
uses 63 questions, and another scenario (LTD) uses only 29 of the 63 questions. Thus
predictors selected under the LTD scenario may be replaced by other predictors in the ALL
scenario. These exchangeable predictors can be considered potential surrogate questions.
Some questions may appear under both scenarios, and can be considered more globally or
universally useful.
The CART tree output will be used to identify questions or groups of questions that classify
the principal participants by exposure level. They will not be used to predict exposure levels
as with a regression equation. Instead a user might look at Figure G.2.2 and see that the node
with the highest mean exposure level is node 10 (LWETHSUM = -2.098, WETHSUM
=0.1227 nmoles/g). The group of participants with the second highest exposure level is node
3 (LWETHSUM = -2.295, WETHSUM = 0.1008 nmoles/g). Both nodes are based on a
small number of measurements, because of the splitting rules used and the number of cases
with higher LWETHSUM values out of the 131 cases. The characteristics of participants in
Nodes 10 and 3 might be used as screening characteristics to identify children with higher
pesticide exposure levels. The group with the lowest exposure level is node 131
(LWETHSUM = -4.684, WETHSUM = 0.0092 nmoles/g). The characteristics of
participants in Node 131, 34, and 71, for example, might be used as screening characteristics
to exclude children with lower pesticide exposure levels. It is important to recognize,
G-18
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
however, that the splitting values for the predictors WDUSTSUM and SWOPSUM are
particular to this study group, and the same values may not be as useful in splitting for other
study populations. With a larger number of cases to analyze, a cross-validation could better
substantiate the split values for the population of interest. This does not take away from the
potential usefulness that household pesticide dust measurements and OP dust measurements
in the school may have in the screening process.
Lastly, in comparing the means of the nodes, the researcher needs to consider whether the
means of the higher exposure level nodes are high in terms of potential exposure impact, or
whether they are not much different than the means in the other nodes. For example, the
difference between the mean of Nodes 10 (highest mean) and 131 (lowest mean) is about
than 0.1135 nmoles/g Creatinine. The researcher might question whether that is a practical
difference for the adjusted concentration level, and whether the mean of 0.1227 nmoles/g
Creatinine for Node 10 is the exposure level of interest, e.g., high enough for a potential
health effect. Data mining techniques operate on mathematical relationships and judgments;
the latter evaluation is a judgment of practical significance.
G-19
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
G.3 Results from Data Mining Approach - CART Analyses
The analyses of interest for the data mining approach on the Yuma Study data focused on
identifying predictor questions or dust measurements that could be useful in classifying a
child's pesticide exposure level as measured by the DAP urinary metabolites. A review of
the available study questions from Table G.2.1 identified a subset of questions that were of
higher interest or that were viewed as having a potentially stronger relationship with the
exposure level. These questions are identified as the limited set of questions (LTD). The
classification analyses were performed on two molar-weighted sums of DAP metabolites
under several scenarios including questions (all or limited), and house and school dust
measurements of pesticides.
G.3.1 Simple Indicators of Exposure Levels
The analyses in this stage were performed to help understand some of the basic relationships
in the Yuma Study to potentially refine the analyses performed in Stage 3. Tests of bivariate
relationships were conducted on the six individual DAPs (DMP, DMTP, DMDTP, DEP,
DETP, and DEDTP) and the 51 original questionnaire variables in Table G.2.1. The
questions covered demographics, residential pesticide use, local pesticide use, activities of
the principal participant, source of drinking water, parents' work environment, and medical
care.
The bivariate analyses were performed on the principal participants with usable urinary
metabolite measurements (approximately 148 children) and before the recoding of
conditional questions and non-responses. Because the questions were used as grouping
variables (section G.2.2.1), the lack of recoding did not affect the evaluation of the
relationships. Questionnaire variables that indicated some differences in levels for at least
three of the six chemicals were:
Since house dust measurements are potential indicators of exposure, rank correlations
between the dust measurements and the urine measurements were performed. Fourteen of
the forty-four dust chemicals from Table 4.3.3 showed some correlation with the DAP
measurements. Chlorpyrifos, diazinon, endosulfan I, endosulfan II, pendimethrin, trifuralin,
and terbufos showed some correlation with more than one of the urinary metabolites.
Variable Name
Variable Description
Pesticides used inside home last month?
Pesticides used outside home last month?
Distance between home and nearest application of pesticides
How often wash local fruit/veg before eating?
Mother now employed (not as housewife)?
Is child covered by medical insurance?
cheminhs
chemouth
closeapp
washvegi
momwork
insured
G-20
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
G.3.2 Underlying Structure - Principal Component Analysis
Principal component analysis (PCA) was performed to help understand relationships between
questions and to identify those that best described the most important dimensions in the data,
that is, those explaining the most variability in the data. PCA was performed on data from
131 core participants after recoding for conditional questions and non-responses, and after
creating the additional questionnaire variables included in Table G.2.1. Sixty-seven of the
questions, as noted in the table and 22 of the house and school dust measurements from
Table G.2.4 were used for the PCA runs. Three scenarios were run: questionnaires only,
questionnaires and house dust measurements, and questionnaires and house and school dust
measurements. These scenarios represent situations with increasing measurement burden
which is a cost/effectiveness concern in study design. Because of the limited availability of
school dust measurements, the PCA with the dust measurements was run with only 107 core
participants.
Table G.3.1 shows the dimensions from two of the PCA runs, and the questionnaire variables
most correlated with each dimension as described in section G.2.2.2. The two PCA scenarios
include questionnaire responses only, and questionnaire responses and all dust
measurements. Table G.3.1 shows the principal component (PC) number for each question
or measurement under the two scenarios. The PC numbers indicate the order in which the
PCs were extracted based on the amount of variability in the data explained, that is, low PC
numbers explain more of the variability. Questions and measurements were assigned to the
PC with which they were most correlated. Thus a question/measurement was assigned to a
PC if the absolute value of the loading between the PC and variable was > 0.6 (section
G.2.3).
Table G.3.1 An Underlying Structure of the Yuma Study Questionnaire and Measurement Variables
Based on Principal Component Analysis Under Two Scenarios
Principal Component Number3
Questions and Measurements
Questions-
only
Scenario
Questions and Dust
(House and School)
Scenario
Name
Brief Description13
1
1
BASEMENT
Basement treated with pesticides?
1
1
BATHROOM
Bathroom treated with pesticides?
1
1
BEDROOM
Bedroom treated with pesticides?
1
1
CHEMINHS
Pesticides used inside home last month?
1
1
CHILDBED
Child's bedroom treated with pesticides?
1
1
DININGRM
Dining room treated with pesticides?
1
1
FAMILYRM
Family room treated with pesticides?
1
1
KITCHEN
Was kitchen treated with pesticides?
1
1
LIVINGRM
Living room treated with pesticides?
1
1
NRMSPRYD
Number of rooms sprayed last month
1
1
OTHERRM
Other rooms treated with pesticides?
G-21
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Principal Component Number3
Questions and Measurements
Questions-
only
Scenario
Questions and Dust
(House and School)
Scenario
Name
Brief Description13
1
1
WHOCHEMI
Who applied pesticides inside the house?
0
2
SCHOOL
Child's school
2
SWCHDNSM
Weighted sum of alpha-chlordane and gamma-
chlordane (School dust)
2
SWCHLPYR
Weighted chlorpyrifos (School dust)
2
SWDSTBAL
Weighted sum of dust analytes except OP pesticides
(School dust)
2
SWOPBAL
Weighted sum of OP pesticides except chlorpyrifos,
diazinon, permethrins, and o-phenylphenol (School
dust)
2
SWOPSUM
Weighted sum of OP pesticides (School dust)
2
3
CHILDFLD
Child worked in fields last month?
2
3
WHENFILD
Last time child was in work field
4
WCHLPYRF
Weighted chlorpyrifos (Household dust)
4
WDUSTBAL
Weighted sum of dust analytes except OP pesticides
(Household dust)
4
WOPSUM
Weighted sum of OP pesticides (Household dust)
11
5
AGE
Age of principal participant
11
5
GRADE
Child's grade
5
SWDUSTSM
Weighted sum of all dust analytes (School dust)
5
SWPERMSM
Weighted sum of cis-permethrin and trans-permethrin
(School dust)
6
6
WATERSR1
Drinking water source - public/commercial
6
6
WATERSR3
Drinking water source - cistern
5
7
ADLTPEST
Non-parent in home works where pesticides used?
5
7
NUMADLTS
Number of additional adults in home
4
8
CLOSEAPP
Distance between home and nearest application of
pesticides
4
8
HOWCHEMO
How pesticides were applied to fields
3
9
FARFIELD
Distance between home and agricultural field
3
9
GPS
Category distance of home from field - GPS
measurement
3
9
WHEEL
Distance between home and field - rotary wheel
10
WDUSTSUM
Weighted sum of all dust analytes (Household dust)
10
WPERMSUM
Weighted sum of cis-permethrin and trans-permethrin
(Household dust)
G-22
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Principal Component Number3
Questions and Measurements
Questions-
only
Scenario
Questions and Dust
(House and School)
Scenario
Name
Brief Description13
8
11
CHEMOUTH
Pesticides used outside home last month?
8
11
WHOCHEMO
Who applied pesticides outside house?
10
12
HEIGHT
Child's height (inches)
10
12
WEIGHT
Child's weight (lbs)
9
13
MOMCON2
Mother's occupation — location and pesticide use
9
13
NCATWRKM
Mother's occupation — categories
14
WCHDNSUM
Weighted sum of alpha-chlordane and gamma-
chlordane (Household dust)
18
14
WHERMD3
Family med care at other med clinic
7
15
PEOPLIVE
Number people, including participating children, in
household
7
15
YOUNGSIB
Number children < 11 years old in household
0
16
WHERMD1
Family med care at private medical clinic
13
16
WHERMD4
Family med care in Mexico
17
17
DADCON2
Father's occupation - location and pesticide use
17
17
NCATWRKD
Father's occupation — categories
23
18
CHLDTM7
Child spends time playing outside?
22
19
CHLDTM3
Child spends time at school?
14
20
CHLDTM5
Child spends time playing in field?
14
20
WHERMD7
Family med care - do not know
21
WDIAZNON
Weighted diazinon (Household dust)
26
22
WASHVEGI
How often wash local fresh fruit/veg before eating?
27
23
HOWCHILD
Child's health in general
15
24
WHERMD5
No access to medical care
0
25
LIVEYEAR
Number years child lived at this address?
16
26
CHLDTM1
Child spends time in another home?
21
27
LIVEAREA
Children, respondent live < 10 months/year in area?
28
28
CHLDTM2
Child spends time at day care center?
20
29
WATERSR2
Drinking water source - private well
0
30
CHLDTM6
Child spends time playing in irrigation water?
12
31
WHERMD6
Family med care at other place
32
SWCYPRME
Weighted cy-permethrin (School dust)
33
WDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and
G-23
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Principal Component Number3
Questions and Measurements
Questions-
only
Scenario
Questions and Dust
(House and School)
Scenario
Name
Brief Description13
4,4'DDT(Household dust)
24
34
SPRAYFLD
Child in yard when fields sprayed or dusted?
29
35
WHERTIME
Room where child spends most awake time
0
0
ADTPSTWK
Any adult works where pesticides used?
19
0
CHLDTM4
Child spends time at sport event?
16
0
ETHNIC
Child's ethnic and racial background
0
0
HOURAWAY
Number hours/wk child not at home
0
0
INSURED
Is child covered by medical insurance?
0
0
LICE
Child treated for head lice past six months?
0
0
OFTCHEMI
How often is home treated for pests?
12
0
POISON
Anyone treated for pesticide poison?
0
0
SEX
Child's gender
0
SWDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and 4,4'DDT
(School dust)
0
SWDIAZNO
Weighted diazinon (School dust)
0
SWOPHNYL
Weighted o-phenylphenol (School dust)
0
0
VEGGIES
How often child eats local fresh fruit/veg?
0
WCYPRMET
Weighted cy-permethrin (Household dust)
25
0
WHERMD2
Family med care at health dept clinic
0
0
WHNCHEMO
Last time field treated with pesticides?
0
WOPBAL
Weighted sum of OP pesticides except chlorpyrifos,
diazinon, permethrins, and o-phenylphenol (Household
dust)
0
WOPHNYLP
Weighted o-phenylphenol (Household dust)
a 0 indicates the variable did not have an absolute loading value > 0.6. Blank indicates the variable was not
included in analysis scenario.
b See Table G.2.1 for extended descriptions
The 29 dimensions for the questions-only scenario accounted for 86% of the variability in the
questionnaire responses with the first dimension accounting for 18%. The 35 dimensions for
the question and dust measurement scenario accounted for 89% of the variability in the
questionnaire responses with the first dimension accounting for 14%. The first dimension
under both scenarios represents spraying of pesticides inside the home, in general and by
room. Subsequent dimensions each account for a significantly smaller amount of the
variability. Note that in many PCA analyses, only the first few dimensions are considered
G-24
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
for further, and different, uses. In this instance, more dimensions were reviewed to look at
the relationships between questions only.
Some items to note regarding the results from the PCA runs based on Table G.3.1:
1. The dimensions identified made sense. Condition and conditional questions (section
G.2.1.3) grouped into the same PC.
2. The components of the PCs explaining the most variability did not change between
the two scenarios, although the PC number (order of extraction) may have changed
slightly.
3. The dimensions explaining the most variability across the two scenarios were:
a. Pesticide sprayed inside house
b. School and school dust measurements
c. Child working in agricultural field
d. Relationship of home to agricultural fields
e. House dust measurements—OPs
f. Adults in household working with pesticides
Although these dimensions were not analyzed with respect to the urine
measurements, they are consistent with the findings in Stage 3 which were so
analyzed.
4. The results from the two PCAs showed a consistency of grouping questions into the
same dimensions with slight variations in the later or smaller components. Although
some of the correlations between the questionnaire variables and dimensions (PCs)
are stronger because of the coding for conditional questions, the fact that other
questions do not have strong correlations with the dimension speaks to cross-question
consistency in the responses.
5. There were only a few instances where two different types of questions were
correlated with the same dimension, e.g., CHLDTM1 (Child spends time in another
home) and ETHNIC (Child's ethnic and racial background). This is an example of
one question being a potential surrogate for the other.
G.3.3 Classification Approach - CART Analyses
The technique Classification and Regression Trees (CART) was selected as the primary type
of data mining analysis. Details of the technique can be found in section G.2.4.1. The
principal participants included in these analyses was limited to 130 children, those in
kindergarten or first grade, from the initial eight study schools, and with available and non-
suspect urine measurement data (section G.2.3). Twelve CART analyses were performed.
As shown in Table G.3.2, six of the analyses were run for the log of the molar-weighted sum
of ethylated DAPs (LWETHSUM), and six were run for the log of the molar-weighted sum
of the methylated DAPs (LWMETHSM). These scenarios were run to understand the impact
of increasing measurement burden which is often considered in the cost/effectiveness aspect
of study design. CART analyses can handle independent variables with missing values; thus
scenarios including school dust measurements did not have to be run with a smaller number
of cases as for the PCA (section 4.3.2.3).
G-25
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table G.3.2 Description of and Cross-Reference for CART Analyses Performed on Yuma Study Data
Predictors Included
Dependent
Variable3
Question
Groupb
House
Dust
School
Dust
Summary
Table
CART Details-
Figures in Appendix G
LWETHSUM
ALL
No
No
G.3.4 c
G.2.1.a, G.2.1 .b
LWETHSUM
LTD
No
No
G.3.4 c
G.2.2.a, G.2.2.b
LWETHSUM
ALL
Yes
No
G.3.4 c
G.2.3.a, G.2.3.b
LWETHSUM
LTD
Yes
No
G.3.4 c
G.2.4.a, G.2.4.b
LWETHSUM
ALL
Yes
Yes
G.3.4 c
G.2.5.a, G.2.5.b
LWETHSUM
LTD
Yes
Yes
G.3.4 c
G.2.6.a, G.2.6.b
LWMETHSM
ALL
No
No
G.3.5
G.2.7.a, G.2.7.b
LWMETHSM
LTD
No
No
G.3.5
G.2.8.a, G.2.8.b
LWMETHSM
ALL
Yes
No
G.3.5
G.2.9.a, G.2.9.b
LWMETHSM
LTD
Yes
No
G.3.5
G.2.10.a, G.2.10.b
LWMETHSM
ALL
Yes
Yes
G.3.5
G.2.11.a, G.2.11 .b
LWMETHSM
LTD
Yes
Yes
G.3.5
G.2.12.a, G.2.12.b
a LWETHSUM is log (molar-weighted sum of ethylated DAPs adjusted for creatinine); LWMETHSM is log
(molar-weighted sum of methylated DAPs adjusted for creatinine). See Appendix F for more details.
b ALL represents analyses with all 67 questions used with CART. LTD represents analyses with 29 of the 67
questions considered to be more likely predictors.
c See also Table G.3.6 for comparisons of CART analyses for LWETHSUM with and without CHLDTM3.
Section G.2.4.2 is an example of the output that will be provided for each of the scenarios
noted above. The output references the characteristics of the subsets, based on split points
for the variables, e.g., CHLDTM1 < 1.5, but does not describe what the subsets CHLDTM1
<1.5 and CHLDTM1 >1.5 represent. Table G.3.3 provides the translation between values
and descriptions, and provides the variable description for the questionnaire variables
included in the CART analyses, in alphabetical order by the variable names. By using Table
G.3.3, the reader can see that CHLDTM1 <1.5 represents a "Yes" response to whether the
child spends time in another home. CHLDTM1 >1.5 represents a "No" response. Five
questionnaire variables are not included in Table G.3.3 because they are more continuous in
nature, or as in the case of SCHOOL, is purely a way to anonymously labels the schools:
V ariable Name V ariable Description
HEIGHT Child's height (inches)
HOURAWAY Number hours/wk child not at home
NRMSPRYD Number of rooms sprayed last month
SCHOOL Child's school
WEIGHT Child's weight (lbs)
G-26
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table G.3.3 Code Values Assigned to Ordinal and Categorical Questionnaire Variables
Q Name
Q Description''
Value
Value Label
adltpest
Non-parent in home works where pesticides used?
1
Yes
2
No
basement
Basement treated with pesticides?
1
Yes
2
No
3
House not treated
bathroom
Bathroom treated with pesticides?
1
Yes
2
No
3
House not treated
bedroom
Bedroom treated with pesticides?
1
Yes
2
No
3
House not treated
cheminhs
Pesticides used inside home last month?
1
Yes
2
No
chemouth
Pesticides used outside home last month?
1
Yes
2
No
3
Do not know
childbed
Child's bedroom treated with pesticides?
1
Yes
2
No
3
House not treated
childfld
Child worked in fields in last month?
1
Yes
2
No
3
Do not know
4
No Response
chldtml
Child spends time in another home?
1
Yes
2
No
chldtm2
Child spends time at day care center?
1
Yes
2
No
G-27
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q Name
Q Description''
Value
Value Label
chldtm3
Child spends time at school?
1
Yes
2
No
chldtm4
Child spends time at sport event?
1
Yes
2
No
chldtm5
Child spends time playing in field?
1
Yes
2
No
chldtm6
Child spends time playing in irrigation water?
1
Yes
2
No
chldtm7
Child spends time playing outside?
1
Yes
2
No
closeapp
Distance between home and nearest application of
pesticides
1
In your yard/garden
2
In neighbor's yard
3
Further away
4
Not used near home
5
Do not know
6
No Response
dadcon2
Father's occupation—location and pesticide use
1
Works Inside, no
pesticides assumed
2
Works Outside, no
pesticides assumed
3
Works Inside, pesticides
assumed
4
Works Outside, pesticides
assumed
5
Dad doesn't work
6
No job response
dadpest
Are pesticides used where father works?
1
Yes
2
No
3
Not Applicable
4
No Response
dadwork
Is the father currently employed?
G-28
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q Name
Q Description''
Value
Value Label
1
Yes
2
No
3
Not Applicable
diningrm
Dining room treated with pesticides?
1
Yes
2
No
3
House not treated
ethnic
Child's ethnic and racial background
1
Hispanic
2
Non-Hispanic white
6
Other, specify
7
No Response
familyrm
Family room treated with pesticides?
1
Yes
2
No
3
House not treated
farfield
Distance between home and agricultural field
1
250 feet or less
2
Over 250 feet
3
Do not know
4
No Response
grade
Child's grade
1
Kindergarden
2
First Grade
howchemo
How pesticides were applied to fields
1
By airplane
2
Mechanized spraying
3
Hand application
4
Other (Specify)
5
Not used near home
6
Do not know
7
No Response
howchild
Child's health in general
1
Excellent
2
Very Good
3
Good
4
Fair
insured
Is child covered by medical insurance?
1
Yes
G-29
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q Name
Q Description''
Value
Value Label
2
No
3
Do not know
4
No Response
kitchen
Was kitchen treated with pesticides?
1
Yes
2
No
3
House not treated
lice
Child treated for head lice past six months?
1
Yes
2
No
3
No Response
livingrm
Living room treated with pesticides?
1
Yes
2
No
3
House not treated
momcon2
Mother's occupation-location and pesticide use
1
Works Inside, no
pesticides assumed
2
Works Outside, no
pesticides assumed
3
Works Inside, pesticides
assumed
4
Works Outside, pesticides
assumed
5
Mom doesn't work
6
No job response
mompest
Are pesticides used where mother works?
1
Yes
2
No
3
Not Applicable
4
No Response
momwork
Mother now employed (not as housewife)?
1
Yes
2
No
3
Not Applicable
ncatwrkd
Father's occupation-categories
1
Agriculture
2
Laborer
3
Repair
G-30
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q Name
Q Description''
Value
Value Label
4
Service
5
Sales
6
Professional
7
Other
8
Dad doesn't work
9
No response
ncatwrkm
Mother's occupation-categories
1
Agriculture
4
Service
5
Sales
6
Professional
7
Other
8
Mom doesn't work
9
No response
oftchemi
How often is home treated for pests?
1
About once a week
2
About once a month
3
Several times a year
4
About once a year
5
Infrequently
6
Never or not yet
7
Do not know
8
No Response
otherrm
Other rooms treated with pesticides?
1
Yes
2
No
3
House not treated
poison
Anyone treated for pesticide poison?
1
Yes
2
No
3
Do not know
4
No Response
sex
Child's gender
1
Female
2
Male
sprayfld
Child in yard when fields sprayed or dusted?
1
Yes
2
No
3
Do not know
G-31
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q Name
Q Description''
Value
Value Label
4
No Response
veggies
How often child eats local fruit/veg?
-1
No Response
0
Do not know
1
Never
2
About once a year
3
About once a month
4
About once a week
5
About once a day
washvegi
How often wash local fruit/veg before eating?
1
Always
2
Usually
3
Sometimes
4
Never
watersrl
Drinking water source-public/commercial
1
Yes
2
No
waters r2
Drinking water source-private well
1
Yes
2
No
waters r3
Drinking water source-cistern
1
Yes
2
No
wheel
Distance between home and field-rotary wheel
1
< 250 feet
2
> 250 and < 500 ft
3
> 500 feet
whenfild
Last time child was in work field
1
Today
2
Yesterday
3
> 2 days ago
4
A week ago
5
> a week ago
6
Do not know
7
Child not in field
8
No Response
whermdl
Family med care at private medical clinic
1
Used
G-32
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q Name
Q Description''
Value
Value Label
2
Not used or Not
applicable
whermd2
Family med care at health department clinic
1
Used
2
Not used or Not
applicable
whermd3
Family med care at other med clinic
1
Used
2
Not used or Not
applicable
whermd4
Family med care in Mexico
1
Used
2
Not used or Not
applicable
whermd5
No access to medical care
1
Used
2
Not used or Not
applicable
whermd6
Family med care at other place
1
Used
2
Not used or Not
applicable
whermd7
Family med care - do not know
1
Used
2
Not used or Not
applicable
whertime
Room where child spend most awake time
1
Living room
2
Family room
3
Dining room
6
Bedroom
7
Other Location
whnchemo
Last time field treated with pesticides?
1
Today
2
Yesterday
3
> 2 days ago
4
A week ago
5
> a week ago
6
Other
7
Do not know
8
Not applicable
G-33
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q Name
Q Description''
Value
Value Label
9
No Response
whochemi
Who applied pesticides inside the house?
1
Self
2
Professional service
3
Family member+other
4
House not treated-DK
whochemo
Who applied pesticides outside house?
1
Self
2
Professional service
3
Family member+other
4
Outside not treated-DK
a See Table G.2.1 for extended descriptions
The scenarios for the following figures are described in Table G.3.2. The contents of the
figures are described in the example CART output found in section G.2.4.2. Note that the
target or dependent variables in the following CART output are logs of molar-weighted sums
of DAPs, adjusted for creatinine.
G-34
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WETHSUM
, chldtm3<1.5
otherr n<1.5
whnche no<6.5
-2.620
4
dadwo
closea >p< 1.5
-2.295
3
k<1.5
wherti
houraw;
-3.175
-2.994
whoche
y<47.5
n2<4.5
wherm
fa rf ie I
-4.219
J < 1.5
i 1 < 1.5
height<
sex-1.5
-4.603
3.892 -3.132 166
wherm
me<4
mi<3.5
48.095 momco
J1 <1.5
weight :46.41
n2<3.5
-3.317 „
-2.923 -2.457 -3.832
sprayf
177
-3.274 -4.202
d<1.5
height- 43.72
-4.276 -3.328
-3.828 | houraw; y<37.5
-2.591 I I
366
-2.992 -3.545
Figure G.3.1.a CART Analysis of LWETHSUM [LOG(WETHSUM)] with All Questions for 130 Yuma Study Participants: CART Tree
G-35
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
* denotes terminal node
1) ROOT 130 91.3700 -3.352
2) CHLDTM3C1.5 122 78.8700 -3.421
4) OTHERRMl.5 114 67.8500 -3.478
10) WHNCHEMO<6.5 50 29.2300 -3.686
20) DADWORKC1.5 45 26.3000 -3.763
40) CLOSEAPPC1.5 7 4.6350 -3.175 *
41) CLOSEAPP>l.5 38 18.8000 -3.871
82) HOURAWAY<4 7.5 21 7.27 90 -3.660
164) MOMCON2<4.5 6 0.5516 -4.219 *
165) MOMCON2>4.5 15 4.0970 -3.436
330) FARFIELD<1.5 6 0.7633 -3.892 *
331) FARFIELD>1.5 9 1.2510 -3.132 *
83) HOURAWAY>47.5 17 9.4320 -4.132
166) WHERMDKl . 5 7 2.4450 -4.603 *
167) WHERMD1>1.5 10 4.3430 -3.802
334) SEX<1.5 5 0.4823 -4.276 *
335) SEX>1.5 5 1.6120 -3.328 *
21) DADWORK>l.5 5 0.2675 -2.994 *
11) WHNCHEMO>6.5 64 34.7700 -3.315
22) WHERTIME<4 50 25.4300 -3.178
44) WHOCHEMK3. 5 21 5.1600 -2.880
88) HEIGHT<4 8.095 15 2.6180 -2.705
176) WHERMDKl.5 8 0.9439 -2.923 *
177) WHERMD1XL.5 7 0.8653 -2.457 *
89) HEIGHT>4 8.095 6 0.9392 -3.317 *
45) WHOCHEMI>3.5 29 17.0500 -3.394
90) MOMCON2<3.5 8 4.9110 -3.832 *
91) MOMCON2>3.5 21 10.0200 -3.227
182) SPRAYFLDC1.5 5 2.1590 -3.828 *
183) SPRAYFLD>1.5 16 5.4960 -3.040
366) HEIGHT<43.72 5 2.1700 -2.591 *
367) HEIGHT>43.72 11 1.8630 -3.244
734) HOURAWAY<37.5 6 0.4431 -2.992
735) HOURAWAY>37.5 5 0.5861 -3.545
23) WHERTIME>4 14 5.0550 -3.804
46) WEIGHT<4 6.41 6 1.4650 -3.274 *
47) WEIGHT>4 6.41 8 0.6442 -4.202 *
3) CHLDTM3>1.5 8 2.9750 -2.295 *
G-36
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WETHSUM) are 3, 177, 366, and 4. Node 166 has the lowest average level of Log(WETHSUM). The
nodes are numbered in bold on Figure G.3.1.a and the final split characteristics are bolded in the above tree description.
Node 3 is characterized by participants with CHLDTM3 responses > 1.5, that is, the child did not spend time in school. These cases had an
average level of 0.1008 nmoles/g Creatinine WETHSUM (Log(WETHSUM) = -2.295) . These may be participants who spent "additional" time in
school. See discussion surrounding Table G.3.6.
Node 177 is characterized by participants with CHLDTM3 < 1.5, OTHERRM > 1.5, WHNCHEMO > 6.5, WHERTIME < 4, WHOCHEMI < 3.5, HEIGHT < 48.095
(inches), and WHERMD1 < 1.5. The average level of WETHSUM for these participants was 0.0857 nmoles/g Creatinine (Log(WETHSUM) = -2.457).
Node 366 is characterized by participants with CHLDTM3 < 1.5, OTHERRM > 1.5, WHNCHEMO > 6.5, WHERTIME < 4, WHOCHEMI < 3.5, MOMCON2 > 3.5,
SPRAYFLD > 1.5, and HEIGHT < 43.72 (inches). The average level of WETHSUM for these participants was 0.0749 nmoles/g Creatinine
(Log(WETHSUM) = -2.591).
Node 4 is characterized by participants with CHLDTM3 < 1.5, and OTHERRM < 1.5. The average level of WETHSUM for these participants was
0.0728 nmoles/g Creatinine (Log(WETHSUM) = -2.620).
Node 166 is characterized by participants with CHLDTM3 < 1.5, OTHERRM > 1.5, WHNCHEMO < 6.5, DADWORK < 1.5, CLOSEAPP > 1.5, HOURAWAY >
47.5, and WHERMD1 < 1.5. The average level of WETHSUM for these participants was 0.0100 nmoles/g Creatinine (Log(WETHSUM) = -4.603) .
Figure G.3.1.b CART Analysis of LWETHSUM [LOG(WETHSUM)] with All Questions for 130 Yuma Study Participants: Summary Statistics for Nodes in CART
Tree (Figure G.3.1.a)
G-37
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WETHSUM
, chldtm3<1.5
weight'
whertirie<1.5
47.025
-3.447
-2.155 -3.103
16
height'
42.94
dadwo
height<
sex':1.5
grad^<1.5
-3.479 -4.103
height
height<
houraw; y<55.5
-2.295
3
k<1.5
howch
47.125
<45.9
whertirie<1.5
45.345
schoc
grade
Id < 2.5
<1.5
weight<|47.835 I
I I -2.837
-3.610 -3.014
-4.335
wherm
I
J2<1.5
n
houraw; y<42.5
I
|<4 5 -3.063 -2.702
~~I 165
-3.406 -3.831
-3.506
w
howch
I
eightfr45.97 ^ 935 "3-682 -4.388
163
ld< 2.5
I
-3.496 -3.907
-2.714
643
Figure G.3.2.a CART Analysis of LWETHSUM [LOG(WETHSUM)] with Limited Questions for 130 Yuma Study Participants: CART Tree
G-38
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
* denotes terminal node
1) ROOT 130 91.3700 -3.352
2) CHLDTM3C1.5 122 78.8700 -3.421
4) HEIGHT<42.94 16 10.4700 -2.914
8) WEIGHT<47.025 11 5.4890 -2.672
16) WHERTIME<1.5 5 0.9163 -2.155 *
17) WHERTIME>1.5 6 2.1230 -3.103 *
9) WEIGHT>47.025 5 2.9150 -3.447 *
5) HEIGHT>42.94 106 63.6700 -3.498
10) DADWORKC1.5 88 52.7 900 -3.57 9
20) HOURAWAY<55.5 69 39.3300 -3.475
40) HEIGHT<47.125 48 24.0100 -3.610
80) HEIGHT<45.9 36 15.6900 -3.448
160) HEIGHT<45.345 30 11.7100 -3.551
320) SEXC1.5 14 4.6710 -3.746
640) GRADEC1.5 8 3.0440 -3.479 *
641) GRADE>1.5 6 0.2934 -4.103 *
321) SEX>1.5 16 6.0410 -3.380
642) WEIGHT<45.97 11 1.9940 -3.683
1284) HOWCHILD<2.5 6 0.7926 -3.496
1285) HOWCHILD>2.5 5 0.7408 -3.907
643) WEIGHT>45.97 5 0.8216 -2.714 *
161) HEIGHT>45.345 6 2.0860 -2.935 *
81) HEIGHT>45.9 12 4.5690 -4.094
162) SCHOOL<4.5 5 2.5550 -3.682 *
163) SCHOOL>4.5 7 0.5605 -4.388 *
41) HEIGHT>47.125 21 12.4500 -3.167
82) WHERTIMEC1.5 12 5.9120 -2.913
164) HOURAWAY<42.5 7 4.9760 -3.063 *
165) HOURAWAY>42.5 5 0.5563 -2.702 *
83) WHERTIME>1.5 9 4.7270 -3.506 *
21) HOURAWAY>55.5 19 9.97 90 -3.958
42) GRADEC1.5 9 6.2200 -4.335 *
43) GRADE>1.5 10 1.3240 -3.619
86) WHERMD2C1.5 5 0.2202 -3.406 *
87) WHERMD2>1.5 5 0.6513 -3.831 *
11) DADWORK>l.5 18 7.4670 -3.101
22) HOWCHILD<2.5 10 2.7910 -3.312
44) WEIGHT<47.835 5 0.7780 -3.610 *
45) WEIGHT>47.835 5 1.1260 -3.014 *
23) HOWCHILD>2.5 8 3.6720 -2.837 *
3) CHLDTM3XL.5 8 2.9750 -2.295 *
G-39
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WETHSUM) are 16, 3, 165, and 643. Node 163 has the lowest average level of Log(WETHSUM). The
nodes are numbered in bold on Figure G.3.2.a and the final split characteristics are bolded in the above tree description.
Node 16 is characterized by participants with CHLDTM3 < 1.5, HEIGHT < 42.94 (inches), WEIGHT < 47.025 (pounds), and WHERTIME < 1.5. The
average level of WETHSUM for these participants was 0.1159 nmoles/g Creatinine (Log(WETHSUM) = -2.155) .
Node 3 is characterized by participants with CHLDTM3 responses > 1.5, that is, the child did not spend time in school. These cases had an
average level of 0.1008 nmoles/g Creatinine WETHSUM (Log(WETHSUM) = -2.295) . These may be participants who spent "additional" time in
school. See discussion surrounding Table G.3.6.
Node 165 is characterized by participants with CHLDTM3 < 1.5, HEIGHT > 42.94 (inches), DADWORK < 1.5, HOURAWAY < 55.5, HEIGHT > 47.125
(inches), WHERTIME < 1.5, and HOURAWAY > 42.5. The average level of WETHSUM for these participants was 0.0677 nmoles/g Creatinine
(Log(WETHSUM) = -2.702).
Node 643 is characterized by participants with CHLDTM3 < 1.5, HEIGHT > 42.94 (inches), DADWORK < 1.5, HOURAWAY < 55.5, HEIGHT < 47.125
(inches), HEIGHT < 45.9 (inches), HEIGHT < 45.345 (inches), SEX > 1.5, and WEIGHT > 45.97 (pounds). The average level of WETHSUM for
these participants was 0.0663 nmoles/g Creatinine (Log(WETHSUM) = -2.714).
Node 163 is characterized by participants with CHLDTM3 < 1.5, HEIGHT > 42.94 (inches), DADWORK < 1.5, HOURAWAY < 55.5, HEIGHT < 47.125
(inches), HEIGHT > 45.9 (inches), and SCHOOL < 4.5. The average level of WETHSUM for these participants was 0.0124 nmoles/g Creatinine
(Log(WETHSUM) = -4.388).
Figure G.3.2.b CART Analysis of LWETHSUM [LOG(WETHSUM)] with Limited Questions for 130 Yuma Study Participants: Summary Statistics for Nodes in
CART Tree (Figure G.3.2.a)
G-40
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
-3.011
Log WETHSUM
. ch Idtm3< 1.5
wdustsum <57.2905
ncatwr cd<7.5
schoc
fa rfiel
height<
l<6.5
J <2.5
42.815
wopsum
wdustsum <26.0648
height
sprayf
wchlpy rf*
livingr n< 1.5
wpermsurr
hourawry<37.5
<7.12771
-2.295
3
weight
47.62
familyrn<2.5
I | -2.434
-3.012 -3.646 19
-2.167 -3.158
10
2.48248
1.28783
-3.1 45
-2.880 _3 370 | |
33 -3.944 -4.655
71
d < 1.5
-4.266
= 45.5
houraw
-4.041 -3.78^vdustsum=5.11172
y<42 .5
-4.228
-2.926
-3.793 -3.230
Figure G.3.3.a CART Analysis of LWETHSUM [LOG(WETHSUM)] with All Questions and House Dust Measurements for 130 Yuma Study Particpants: CART
Tree
G-41
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
* denotes terminal node
1) ROOT 130 91.3700 -3.352
2) CHLDTM3C1.5 122 78.8700 -3.421
4) WDUSTSUM<57.2905 110 63.7000 -3.504
8) NCATWRKD<7.5 90 51.8300 -3.589
16) SCHOOL<6.5 71 39.8000 -3.483
32) FARFIELD<2.5 64 36.1000 -3.549
64) HEIGHT<42.815 8 5.8330 -3.011 *
65) HEIGHT>42.815 56 27.6300 -3.626
130) WOPSUM<2.48248 47 22.1100 -3.718
260) WDUSTSUM<26.0648 38 15.4000 -3.588
520) WCHLPYRFCl.28783 33 11.7900 -3.491
1040) SPRAYFLDC1.5 10 1.1690 -3.915
2080) HEIGHT<45.5 5 0.5299 -4.041 *
2081) HEIGHT>45.5 5 0.4783 -3.788 *
1041) SPRAYFLD>1.5 23 8.0480 -3.307
2082) HOURAWAY<4 2.5 14 2.9240 -3.552
4164) WDUSTSUM<5.11172 8 0.9874 -3.793
4165) WDUSTSUM>5.11172 6 0.8485 -3.230
2083) HOURAWAY>42.5 9 2.9810 -2.926 *
521) WCHLPYRF>1.28783 5 1.2500 -4.228 *
261) WDUSTSUM>26.0648 9 3.3650 -4.266 *
131) WOPSUM>2.48248 9 3.0370 -3.145 *
33) FARFIELD>2.5 7 0.8801 -2.880 *
17) SCHOOL>6.5 19 8.2150 -3.987
34) LIVINGRMC1.5 6 1.1600 -3.370 *
35) LIVINGRMM.5 13 3.7170 -4.272
70) WEIGHT<47.62 7 1.7760 -3.944 *
71) WEIGHT>47.62 6 0.3120 -4.655 *
9) NCATWRKD>7.5 20 8.2810 -3.121
18) WPERMSUM<7.12771 15 4.0970 -3.350
36) FAMILYRM<2.5 7 0.1760 -3.012 *
37) FAMILYRM>2.5 8 2.4190 -3.646 *
19) WPERMSUM>7.12771 5 1.0410 -2.434 *
5) WDUSTSUM>57.2905 12 7.5040 -2.662
10) HOURAWAY<37.5 6 2.8740 -2.167 *
11) HOURAWAY>37.5 6 1.6850 -3.158 *
3) CHLDTM3XL.5 8 2.9750 -2.295 *
G-42
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WETHSUM) are 10, 3, 19, and 33. Node 71 has the lowest average level of Log(WETHSUM). The
nodes are numbered in bold on Figure G.3.3.a and the final split characteristics are bolded in the above tree description.
Node 10 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM > 57.2905, and HOURAWAY < 37.5. The average level of WETHSUM for
these participants was 0.1145 nmoles/g Creatinine (Log(WETHSUM) = -2.167) .
Node 3 is characterized by participants with CHLDTM3 responses > 1.5, that is, the child does not spend time in school. These cases had
an average level of 0.1008 nmoles/g Creatinine WETHSUM (Log(WETHSUM) = -2.295) . These may be participants who spent "additional" time in
school. See discussion surrounding Table G.3.6.
Node 19 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, NCATWRKD > 7.5, and WPERMSUM > 7.12771. The average level
of WETHSUM for these participants was 0.0877 nmoles/g Creatinine (Log(WETHSUM) = -2.434) .
Node 33 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, NCATWRKD < 7.5, SCHOOL < 6.5, and FARFIELD > 2.5. The
average level of WETHSUM for these participants was 0.0561 nmoles/g Creatinine (Log(WETHSUM) = -2.880) .
Node 71 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, NCATWRKD < 7.5, SCHOOL > 6.5, LIVINGRM > 1.5, and WEIGHT
> 47.62 (pounds). The average level of WETHSUM for these participants was 0.0095 nmoles/g Creatinine (Log(WETHSUM) = -4.655).
Figure G.3.3.b CART Analysis of LWETHSUM [LOG(WETHSUM)] with All Questions and House Dust Measurements for 130 Yuma Study Participants: Summary
Statistics for Nodes in CART Tree (Figure G.3.3.a)
G-43
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WETHSUM
schoc
wdustsum
wdustbal<( i. 0360587
wpermsurr
K6.5
<24.1155
<3.14345
howch
wherm i1 <1.5
-3.072
wdustbal
^eight'|:48.94 |
2.2066Wchlpyrf< 0.1 5695
wopsum<
Id <1.5
chldtm3<1.5
wdustsum <57.2905
dadwo k<1.5
wdiaznon<
3.485786
wpermsurr
wdustbal<
0.165106 |-
houraw; y<37.5
<7.12771
-2.295
3
D. 535901
"I -2.434
-2.167 -3.158
10
dustsum <13.6977 -3.678 -3.131
grade
<1.5
I I -4.538
-3.536 -4.045 ^
19
wchlpyrf<
-4.428 | |
-3.258 -3.920
3.254571
-3.428
-3.672 -4.324
"| -3.726 | |
-3.616 -3.069 -2.513 -3.182
262
Figure G.3.4.a CART Analysis of LWETHSUM [LOG(WETHSUM)] with Limited Questions and House Dust Measurements for 130 Yuma Study Participants: CART
Tree
G-44
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
* denotes terminal node
1) ROOT 130 91.37000 -3.352
2) CHLDTM3C1.5 122 78.87000 -3.421
4) WDUSTSUM<57.2905 110 63.70000 -3.504
8) DADWORKC1.5 90 51.83000 -3.589
16) SCHOOL<6.5 71 39.80000 -3.483
32) WDUSTSUM<24.1155 54 28.87000 -3.372
64) WPERMSUM<3.14345 36 17.90000 -3.540
128) WDUSTBALC0.0360587 7 5.66100 -3.072 *
129) WDUSTBAL>0.0360587 29 10.34000 -3.653
258) WHERMDKl . 5 18 5.98600 -3.858
516) WDUSTBAL<2.20667 13 3.13600 -4.023
1032) WEIGHT<4 8.94 6 1.29500 -3.672 *
1033) WEIGHT>4 8.94 7 0.46990 -4.324 *
517) WDUSTBAL>2.20667 5 1.57100 -3.428 *
259) WHERMD1>1.5 11 2.35900 -3.318
518) WCHLPYRFCO.15695 5 0.03282 -3.616 *
519) WCHLPYRF>0.15695 6 1.51000 -3.069 *
65) WPERMSUM>3.14345 18 7.91800 -3.036
130) HOWCHILDC1.5 5 1.39900 -3.726 *
131) HOWCHILD>l.5 13 3.22200 -2.770
262) WCHLPYRF<0.254571 8 0.29330 -2.513 *
263) WCHLPYRF>0.254571 5 1.55000 -3.182 *
33) WDUSTSUM>24.1155 17 8.15300 -3.835
66) WOPSUMCO.485786 5 2.12000 -4.428 *
67) WOPSUM>0.485786 12 3.54900 -3.589
134) GRADEC1.5 6 0.88550 -3.258 *
135) GRADE>1.5 6 1.35100 -3.920 *
17) SCHOOL>6.5 19 8.21500 -3.987
34) WDIAZNONCO.165106 14 5.55100 -3.791
68) WDUSTSUM<13.6977 7 2.06700 -3.536 *
69) WDUSTSUM>13.6977 7 2.57700 -4.045 *
35) WDIAZNON>0.165106 5 0.60860 -4.538 *
9) DADWORK>l.5 20 8.28100 -3.121
18) WPERMSUM<7.12771 15 4.09700 -3.350
36) WDUSTBAL<0.535901 6 0.75070 -3.678 *
37) WDUSTBAL>0.535901 9 2.26700 -3.131 *
19) WPERMSUM>7.12771 5 1.04100 -2.434 *
5) WDUSTSUM>57.2905 12 7.50400 -2.662
10) HOURAWAY<37.5 6 2.87400 -2.167 *
11) HOURAWAY>37.5 6 1.68500 -3.158 *
3) CHLDTM3XL.5 8 2.97500 -2.295 *
G-45
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WETHSUM) are 10, 3, 19, and 262. Node 35 has the lowest average level of Log(WETHSUM). The
nodes are numbered in bold on Figure G.3.4.a and the final split characteristics are bolded in the above tree description.
Node 10 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM > 57.2905, and HOURAWAY < 37.5. The average level of WETHSUM for
these participants was 0.1145 nmoles/g Creatinine (Log(WETHSUM) = -2.167) .
Node 3 is characterized by participants with CHLDTM3 responses > 1.5, that is, the child did not spend time in school. These cases had an
average level of 0.1008 nmoles/g Creatinine WETHSUM (Log(WETHSUM) = -2.295) . These may be participants who spent "additional" time in
school. See discussion surrounding Table G.3.6.
Node 19 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, DADWORK > 1.5, and, WPERMSUM > 7.12771. The average
level of WETHSUM for these participants was 0.0877 nmoles/g Creatinine (Log(WETHSUM) = -2.434) .
Node 262 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, DADWORK < 1.5, SCHOOL < 6.5, WDUSTSUM < 24.1555,
WPERMSUM > 3.14345, HOWCHILD > 1.5, and WCHLPYRF < 0.254571. The average level of WETHSUM for these participants was 0.081 nmoles/g
Creatinine (Log(WETHSUM) = -2.513).
Node 35 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, DADWORK < 1.5, SCHOOL > 6.5, and WDIAZNON > 0.165106. The
average level of WETHSUM for these participants was 0.0107 nmoles/g Creatinine (Log(WETHSUM) = -4.538).
Figure G.3.4.b CART Analysis of LWETHSUM [LOG(WETHSUM)] with Limited Questions and House Dust Measurements for 130 Yuma Study Participants:
Summary Statistics for Nodes in CART Tree (Figure G.3.4.a)
G-46
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WETHSUM
weight<
45.685
schoc
K2.5
-3.360
-3.797 -4.684
dadcor
2< 1.5
swdstbal
4.01061
whertir
houraw
wdustsum
e<6.5
ay<41
<5.13875
ch ildfl
131 gradq<1.5wopbal<0.|)002929 6
ncatwr cd<7.5
otherr n<2.5
whnche
, chldtm3<1.5
wdustsum <57.2905
wpermsurr
swopsu m <0.563682
<7.12771
no<6.5
familyrm<2.5
rn
-3.012 -3.646
-2.295
3
-2.434
19
-2.098 -3.227
10
-4.538
-4.143 -3.351
-3.951
J<1 .5
wopsum <
3.815759
1
-4.196 -3.522 -2.971 -3.555 -3.555 [
-2.411 -3.064
534
Figure G.3.5.a CART Analysis of LWETHSUM [LOG(WETHSUM)] with All Questions and House and School Dust Measurements for 130 Yuma Study Participants:
CART Tree
G-47
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
1) ROOT 130 91.37000 -3.352
2) CHLDTM3C1.5 122 78.87000 -3.421
4) WDUSTSUM<57.2905 110 63.70000 -3.504
8) NCATWRKD<7.5 90 51.83000 -3.589
16) SWDSTBAL<4.01061 72 37.41000 -3.462
32) DADCON2C1.5 19 7.97400 -3.847
64) WEIGHT<45.685 8 1.21300 -3.360 *
65) WEIGHT>45.685 11 3.49400 -4.200
130) SCHOOL<2.5 6 1.16900 -3.797 *
131) SCHOOL>2.5 5 0.17690 -4.684 *
33) DADCON2>l.5 53 25.62000 -3.324
66) WHERTIME<6.5 46 21.38000 -3.229
132) HOURAWAY<41 26 8.68300 -3.468
264) WDUSTSUM<5.13875 11 2.24800 -3.828
528) GRADEC1.5 5 0.05032 -4.196 *
529) GRADE>1.5 6 0.96070 -3.522 *
265) WDUSTSUM>5.13875 15 3.96500 -3.205
530) SWOPBALCO.000292916 9 1.55700 -2.971
531) SWOPBAL>0.000292916 6 1.18200 -3.555
133) H OU RAWAY >41 20 9.26200 -2.918
266) CHILDFLD<1.5 6 2.02800 -3.555 *
267) CHILDFLD>1.5 14 3.75200 -2.644
534) WOPSUM<0.815759 9 1.04500 -2.411 *
535) WOPSUM>0.815759 5 1.33600 -3.064 *
67) WHERTIME>6.5 7 1.07700 -3.951 *
17) SWDSTBAL>4.01061 18 8.59000 -4.098
34) OTHERRM<2.5 10 2.32500 -3.747
68) WHNCHEMO<6.5 5 0.27310 -4.143 *
69) WHNCHEMO>6.5 5 0.48550 -3.351 *
35) OTHERRM>2.5 8 3.48700 -4.538 *
9) NCATWRKD>7.5 20 8.28100 -3.121
18) WPERMSUM<7.12771 15 4.09700 -3.350
36) FAMILYRM<2.5 7 0.17600 -3.012 *
37) FAMILYRM>2.5 8 2.41900 -3.646 *
19) WPERMSUM>7.12771 5 1.04100 -2.434 *
5) WDUSTSUM>57.2905 12 7.50400 -2.662
10) SWOPSUM<0.563682 6 1.89600 -2.098 *
11) SWOPSUM>0.563682 6 1.78500 -3.227 *
3) CHLDTM3XL.5 8 2.97500 -2.295 *
G-48
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WETHSUM) are 10, 3, 534, and 19. Node 131 has the lowest average level of Log(WETHSUM). The
nodes are numbered in bold on Figure G.3.5.a and the final split characteristics are bolded in the above tree description.
Node 10 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM > 57.2905, and SWOPSUM < 0.563682. The average level of WETHSUM for
these participants was 0.1227 nmoles/g Creatinine (Log(WETHSUM) = -2.098) .
Node 3 is characterized by participants with CHLDTM3 responses > 1.5, that is, the child did not spend time in school. These cases had an
average level of 0.1008 nmoles/g Creatinine WETHSUM (Log(WETHSUM) = -2.295). These may be participants who spent "additional" time in
school. See discussion surrounding Table G.3.6.
Node 534 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, NCATWRKD < 7.5, SWDSTBAL < 4.01061, DADCON2 > 1.5,
WHERTIME < 6.5, HOURAWAY > 41, CHLDFLD > 1.5, and WOPSUM < 0.815759. The average level of WETHSUM for these participants was 0.0897
nmoles/g Creatinine (Log(WETHSUM) = -2.411).
Node 19 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, NCATWRKD > 7.5, and WPERMSUM > 7.12771. The average
level of WETHSUM for these participants was 0.0877 nmoles/g Creatinine (Log(WETHSUM) = -2.434) .
Node 131 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, NCATWRKD < 7.5, SWDSTBAL < 4.01061, DADCON2 < 1.5,
WEIGHT > 45.685 (pounds), and SCHOOL > 2.5. The average level of WETHSUM for these participants was 0.0092 nmoles/g Creatinine
(Log(WETHSUM) = -4.684).
Figure G.3.5.b CART Analysis of LWETHSUM [LOG(WETHSUM)] with All Questions and House and School Dust Measurements for 130 Yuma Study Participants:
Summary Statistics for Nodes in CART Tree (Figure G.3.5.a)
G-49
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WETHSUM
chldtm3<1.5
wdustsum <57.2905
dadwo *k<1.5
swdstbal
swddsum<8
wddsum
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
* denotes terminal node
1) ROOT 130 91.3700 -3.352
2) CHLDTM3C1.5 122 78.8700 -3.421
4) WDUSTSUM<57.2905 110 63.7000 -3.504
8) DADWORKC1.5 90 51.8300 -3.589
16) SWDSTBAL<4.01061 72 37.4100 -3.462
32) SWDDSUM<8.74071E-005 43 19.0900 -3.634
64) WDDSUMC0.0895053 37 14.8100 -3.731
128) GRADEC1.5 11 4.3610 -4.046
256) WEIGHT<4 6.945 6 1.8300 -3.636 *
257) WEIGHT>46.945 5 0.3091 -4.539 *
129) GRADE>1.5 26 8.8940 -3.597
258) WDUSTBALC0.879648 13 2.7320 -3.330
516) WDUSTSUM<2.63068 6 0.1600 -3.663
517) WDUSTSUM>2.63068 7 1.3390 -3.045
259) WDUSTBAL>0.879648 13 4.3030 -3.865
518) WDUSTBAL<3.42399 8 0.6443 -4.193
519) WDUSTBAL>3.42399 5 1.4210 -3.340
65) WDDSUM>0.0895053 6 1.7990 -3.039 *
33) SWDDSUM>8.74071E-005 29 15.1600 -3.207
66) WDUSTBAL<0.0360587 5 0.6672 -2.413 *
67) WDUSTBAL>0.0360587 24 10.6800 -3.372
134) WPERMSUM<19.5124 19 7.4790 -3.222
268) WPERMSUM<2.55757 9 1.3840 -3.725 *
269) WPERMSUM>2.55757 10 1.7570 -2.768
538) WOPSUM<0.923761 5 0.1611 -2.420 *
539) WOPSUM>0.923761 5 0.3797 -3.117 *
135) WPERMSUM>19.5124 5 1.1300 -3.946 *
17) SWDSTBAL>4.01061 18 8.5900 -4.098
34) WDIAZNONCO.14802 13 4.4740 -3.868
68) WDDSUMCO.25584 8 2.0210 -3.581 *
69) WDDSUM>0.25584 5 0.7376 -4.327 *
35) WDIAZNON>0.14802 5 1.6260 -4.698 *
9) DADWORK>l.5 20 8.2810 -3.121
18) WPERMSUM<7.12771 15 4.0970 -3.350
36) WDUSTBAL<0.535901 6 0.7507 -3.678 *
37) WDUSTBAL>0.535901 9 2.2670 -3.131 *
19) WPERMSUM>7.12771 5 1.0410 -2.434 *
5) WDUSTSUM>57.2905 12 7.5040 -2.662
10) SWOPSUM<0.563682 6 1.8960 -2.098 *
11) SWOPSUM>0.563682 6 1.7850 -3.227 *
3) CHLDTM3XL.5 8 2.9750 -2.295 *
G-51
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WETHSUM) are 10, 3, 66, and 538. Node 35 has the lowest average level of Log(WETHSUM). The
nodes are numbered in bold on Figure G.3.6.a and the final split characteristics are bolded in the above tree description.
Node 10 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM > 57.2905, and SWOPSUM < 0.563682. The average level of WETHSUM for
these participants was 0.1227 nmoles/g Creatinine (Log(WETHSUM) = -2.098) .
Node 3 is characterized by participants with CHLDTM3 responses > 1.5, that is, the child did not spend time in school. These cases had an
average level of 0.1008 nmoles/g Creatinine WETHSUM (Log(WETHSUM) = -2.295) . These may be participants who spent "additional" time in
school. See discussion surrounding Table G.3.6.
Node 66 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, DADWORK < 1.5, SWDSTBAL < 4.01061, SWDDSUM > 8.7E-05, and
WDUSTBAL < 0.0360587. The average level of WETHSUM for these participants was 0.0895 nmoles/g Creatinine (Log(WETHSUM) = -2.413) .
Node 538 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, DADWORK < 1.5, SWDSTBAL < 4.01061, SWDDSUM > 8.7E-05,
WDUSTBAL > 0.0360587, WPERMSUM < 19.5124, WPERMSUM > 2.55757, and WOPSUM < 0.923761. The average level of WETHSUM for these participants
was 0.0889 nmoles/g Creatinine (Log(WETHSUM) = -2.420).
Node 35 is characterized by participants with CHLDTM3 < 1.5, WDUSTSUM < 57.2905, DADWORK < 1.5, SWDSTBAL > 4.01061, and WDIAZNON <
0.14802. The average level of WETHSUM for these participants was 0.0091 nmoles/g Creatinine (Log(WETHSUM) = -4.698).
Figure G.3.6.b CART Analysis of LWETHSUM [LOG(WETHSUM)] with Limited Questions and House and School Dust Measurements for 130 Yuma Study
Participants: Summary Statistics for Nodes in CART Tree (Figure G.3.6.a)
G-52
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WMETHSUM
jvhnchemo<6.5
schoc l<6.5
ncatwrl m<5.5
ncatwrl
whee < 1.5
m<4.5 whoche
ch Idtm
mi<1.5
-4.2650 -3.0360
farfiel J<1.5
-1.7350
weight
-2.8690 -0.3861
17
46.74 dadcor
2<3.5
height' 44.16
7< 1.5
height<
wherm
-2.9100 -2.1560
-2.6740 -3.9230
77
-0.1 795
12
wherm 11 <1.5
11 <1.5
howch
46.595
wherti me<4
Id < 2.5
dadcor
2<3.5
| oftche i<2.5 -2.7470
-2.0180
-3.5590
-1.4340 -2.6610 -1.7460 -2.8430
26
-1.8860 -1 .0570
123
Figure G.3.7.a CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All Questions for 130 Yuma Study Participants: CART Tree
G-53
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
* denotes terminal node
1) ROOT 130 237.400 -2.2580
2) WHNCHEMO<6.5 57 100.600 -2.6210
4) SCHOOL<6.5 45 74.430 -2.3740
8) NCATWRKM<5.5 10 25.250 -1.6280
16) NCATWRKM<4.5 5 1.138 -2.8690 *
17) NCATWRKM>4.5 5 8.700 -0.3861 *
9) NCATWRKM>5.5 35 42.010 -2.5870
18) WHOCHEMICl.5 8 10.100 -1.7350 *
19) WHOCHEMI>l.5 27 24.380 -2.8400
38) FARFIELDC1.5 11 10.540 -3.3550
76) WEIGHT<4 6.74 5 2.104 -2.6740 *
77) WEIGHT>46.74 6 4.182 -3.9230 *
39) FARFIELD>1.5 16 8.915 -2.4860
78) DADCON2<3.5 7 4.271 -2.9100 *
79) DADCON2>3.5 9 2.406 -2.1560 *
5) SCHOOL>6.5 12 13.100 -3.5480
10) WHEEL<1.5 5 4.465 -4.2650 *
11) WHEEL>1.5 7 4.218 -3.0360 *
3) WHNCHEMO>6.5 73 123.400 -1.9740
6) HEIGHT<4 4.16 24 46.830 -1.2700
12) CHLDTM7<1.5 9 3.500 -0.1795 *
13) CHLDTM7>1.5 15 26.200 -1.9250
26) WHERMDKl . 5 9 14.650 -1.4340 *
27) WHERMD1>1.5 6 6.125 -2.6610 *
7) HEIGHT>4 4.16 49 58.890 -2.3190
14) HEIGHT<4 6.595 23 29.130 -2.7890
28) WHERMDKl.5 14 17.140 -2.2940
56) HOWCHILD<2.5 7 6.467 -1.7460 *
57) HOWCHILD>2.5 7 6.462 -2.8430 *
29) WHERMD1>1.5 9 3.232 -3.5590 *
15) HEIGHT>4 6.595 26 20.150 -1.9020
30) WHERTIME<4 20 12.480 -1.6490
60) DADCON2<3.5 8 3.190 -2.0180 *
61) DADCON2>3.5 12 7.467 -1.4020
122) OFTCHEMK2 . 5 5 3.535 -1.8860 *
123) OFTCHEMI>2.5 7 1.925 -1.0570 *
31) WHERTIME>4 6 2.106 -2.7470 *
G-54
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WMETHSUM) are 12, 17, 123, and 26. Node 77 has the lowest average level of Log(WMETHSUM). The
nodes are numbered in bold on Figure G.3.7.a and the final split characteristics are bolded in the above tree description.
Node 12 is characterized by participants with WHNCHEMO > 6.5, HEIGHT < 44.16 (inches), and CHLDTM7 < 1.5. The average level of WMETHSUM
for these participants was 0.8357 nmoles/g Creatinine (Log(WMETHSUM) = -0.1795).
Node 17 is characterized by participants with WHNCHEMO < 6.5, SCHOOL < 6.5, NCATWRKM < 5.5, and NCATWRKM > 4.5. The average level of
WMETHSUM for these participants was 0.6797 nmoles/g Creatinine (Log(WMETHSUM) = -0.3861) .
Node 123 is characterized by participants with WHNCHEMO > 6.5, HEIGHT > 44.16 (inches), HEIGHT > 46.595 (inches), WHERTIME < 4, DADCON2 >
3.5, and OFTCHEMI > 2.5. The average level of WMETHSUM for these participants was 0.3475 nmoles/g Creatinine (Log(WMETHSUM) = -1.0570).
Node 26 is characterized by participants with WHNCHEMO > 6.5, HEIGHT < 44.16 (inches), CHLDTM7 > 1.5, and WHERMD1 < 1.5. The average
level of WMETHSUM for these participants was 0.2384 nmoles/g Creatinine (Log(WMETHSUM) = -1.4340) .
Node 77 is characterized by participants with WHNCHEMO < 6.5, SCHOOL < 6.5, NCATWRKM < 5.5, WHOCHEMI > 1.5, FARFIELD < 1.5, and WEIGHT >
46.74 (pounds). The average level of WMETHSUM for these participants was 0.0196 nmoles/g Creatinine (Log(WMETHSUM) = -3.9320).
Figure G.3.7.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All Questions for 130 Yuma Study Participants: Summary Statistics for Nodes in CART
Tree (Figure G.3.7.a)
G-55
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WMETHSUM
height<44.1 6
chldtm 1 <1.5
wherm J2<1.5
-0.4536
4
wherti
-3.3790
schoc
height<
ne<4
l< 6.5 whertir
43.065
weight<
48.275
e<6.5 I I
| | -3.9290-3.1350
2.1720-2.9750 24
-1.5090-2.2690
-0.7541
45
wherm
height'47.1 6
J4< 1.5
howch
dadwo
weight-
weight<
k<1.5
48.055
-2.0670
Id <2.5
49.815 weight<
-1.0400
28
59.525
-1.9300-1.0570
31
rk< 1.5
-3.3840-2.2280
-1.4670
height'45.44 schoc
height<|45.1 25 I
I I
-2.3190-2.9500
K2.5
-2.4730-3.4670
Figure G.3.8.a CART Analysis of LWMETHSM [LOG(WMETHSUM)] with Limited Questions for 130 Yuma Study Participants: CART Tree
G-56
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
* denotes terminal node
1) ROOT 130 237.400 -2.2580
2) HEIGHT<4 4.16 43 101.700 -1.8360
4) CHLDTMK1 .5 7 15.870 -0.4536 *
5) CHLDTM1>1.5 36 69.880 -2.1050
10) WHERMD2C1.5 5 14.240 -3.3790 *
11) WHERMD2>1.5 31 46.210 -1.8990
22) WHERTIME<4 20 31.530 -1.5480
44) SCHOOL<6.5 15 18.990 -1.8130
88) HEIGHT<43.065 9 13.330 -1.5090 *
89) HEIGHT>43.065 6 3.588 -2.2690 *
45) SCHOOL>6.5 5 8.334 -0.7541 *
23) WHERTIME>4 11 7.744 -2.5370
46) WHERTIME<6.5 6 4.860 -2.1720 *
47) WHERTIME>6.5 5 1.126 -2.9750 *
3) HEIGHT>4 4.16 87 124.200 -2.4660
6) HEIGHT<47.16 52 61.400 -2.8390
12) WHERMD4<1.5 16 11.170 -3.5320
24) WEIGHT<48.275 8 5.485 -3.9290 *
25) WEIGHT>4 8.275 8 3.165 -3.1350 *
13) WHERMD4>1.5 36 39.130 -2.5310
26) DADWORK45.12 5 5 1.584 -2.9500 *
105) HEIGHT>4 5.4 4 6 3.607 -2.0670 *
53) WEIGHT>4 8.055 14 12.370 -3.1120
106) SCHOOL<2.5 5 7.099 -2.4730 *
107) SCHOOL>2.5 9 2.099 -3.4670 *
27) DADWORK>l.5 6 4.222 -1.4 67 0 *
7) HEIGHT>47.16 35 44.880 -1.9130
14) HOWCHILD<2.5 19 25.660 -2.2190
28) WEIGHT<49.815 5 1.886 -1.0400 *
29) WEIGHT>4 9.815 14 14.330 -2.6410
58) MOMWORKC1.5 5 1.741 -3.3840 *
59) MOMWORK>l.5 9 8.297 -2.2280 *
15) HOWCHILD>2.5 16 15.300 -1.5480
30) WEIGHT<59.525 9 3.379 -1.9300 *
31) WEIGHT>59.525 7 8.919 -1.0570 *
G-57
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WMETHSUM) are 4, 45, 28, and 31. Node 24 has the lowest average level of Log(WMETHSUM). The
nodes are numbered in bold on Figure G.3.8.a and the final split characteristics are bolded in the above tree description.
Node 4 is characterized by participants with HEIGHT < 44.16 (inches), and CHLDTM1 < 1.5. The average level of WMETHSUM for these
participants was 0.6353 nmoles/g Creatinine (Log(WMETHSUM) = -0.4536).
Node 45 is characterized by participants with HEIGHT < 44.16 (inches), CHLDTM1 > 1.5, WHERMD2 > 1.5, WHERTIME < 4, and SCHOOL < 6.5. The
average level of WMETHSUM for these participants was 0.4704 nmoles/g Creatinine (Log(WMETHSUM) = -0.7541).
Node 28 is characterized by participants with HEIGHT > 44.16 (inches), HEIGHT > 47.16 (inches), HOWCHILD < 2.5, and WEIGHT < 49.815
(pounds) . The average level of WMETHSUM for these participants was 0.3535 nmoles/g Creatinine (Log(WMETHSUM) = -1.0400) .
Node 31 is characterized by participants with HEIGHT > 44.16 (inches), HEIGHT > 47.16 (inches), HOWCHILD > 2.5, and WEIGHT > 59.525
(pounds). The average level of WMETHSUM for these participants was 0.3475 nmoles/g Creatinine (Log(WMETHSUM) = -1.0570).
Node 24 is characterized by participants with HEIGHT > 44.16 (inches), HEIGHT < 47.16 (inches), WHERMD4 < 1.5, and WEIGHT < 48.275
(inches) . The average level of WMETHSUM for these participants was 0.0197 nmoles/g Creatinine (Log(WMETHSUM) = -3.9290) .
Figure G.3.8.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with Limited Questions for 130 Yuma Study Participants: Summary Statistics for Nodes in
CART Tree (Figure G.3.8.a)
G-58
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WMETHSUM
j/hnchemo<6.5
schoc l<6.5
height<
wchlpyrf- 0.561 91
whee <1.5
wpermsuri<10.082 wdiaznon<
43.125
whoche
-1.6480
height<
n
chldtm
-4.2650-3.0360
0.114588
1
10
-3.3970-2.7270
height':44.16
7<1.5
height<
wchlpyrf
-0.1795
12
wdiaznon<
wopbal<0.C
0.60826
>.0895515
ncatw
46.595
00266113
wherti
kd<4wchlpyrf<
wdustbal<
f
3.1 41 61"$ -2130
1 60
ne<4
3.450858
sex+l .5 .2 7470
I I
-1.6450-2.3520
-2.5400-1.421 0-4.0500-3.0120
-3.0880
no<3.5
46.375
-0.2036
17
-0.7838-1.9030
52
-3.4070
-2.6020-1.6210
Figure G.3.9.a CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All Questions and House Dust Measurements for 130 Yuma Study Participants: CART
Tree
G-59
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
* denotes terminal node
1) ROOT 130 237.4000 -2.2580
2) WHNCHEMO<6.5 57 100.6000
4) SCHOOL<6.5 45 74.4300
8) WCHLPYRFC0.56191 30
16) WPERMSUMCl0.082 24
13.1000 -3.
4.4650 -4
4.2180 -3.
32) HEIGHT<43.125 5
33) HEIGHT>43.125 19
66) WHOCHEMO<3.5 10
132) HEIGHT<4 6.375 5
133) HEIGHT>4 6.375 5
67) WHOCHEMO>3.5 9
17) WPERMSUMXLO . 082 6
9) WCHLPYRF>0.56191 15
18) WDIAZNONCO.114588 7
19) WDIAZNON>0.114588 8
5) SCHOOL>6.5 12
10) WHEEL<1.5 5
11) WHEEL>1.5 7
) WHNCHEMO>6.5 73 123.4000 -1
6) HEIGHT<4 4.16 24 46.8300
12) CHLDTM7<1.5 9 3.5000
13) CHLDTM7>1.5 15 26.2000
26) WCHLPYRFCO.60826 10
52) WDIAZNON<0.0895515
53) WDIAZNON>0.0895515
27) WCHLPYRF>0.60826 5
7) HEIGHT>4 4.16 4 9 58.8900
14) HEIGHT<4 6.595 23 29.13
28) WOPBALCO.000266113 10
56) NCATWRKD<4 5 4
57) NCATWRKD>4 5 2.73
29) WOPBAL>0.000266113 13
58) WCHLPYRFCO.141618 5
59) WCHLPYRF>0.141618 8
15) HEIGHT>4 6.595 26 20.15
30) WHERTIME<4 20 12.480
60) WDUSTBAL<0.450858 8
61) WDUSTBAL>0.450858 1
122) SEXC1.5 7 2.001
123) SEX>1.5 5 3.201
31) WHERTIME>4 6
6210
3740
1000 -2.0410
0.6800 -2.5010
0470 -1.6480 *
3.0500 -2.7250
4 . 9560 -2.1110
0.9366 -2.6020 *
1.6160 -1.6210 *
0.1400 -3.4070 *
.0920 -0.2036 *
.3590 -3.0400
1.4320 -3.3970 *
1.2560 -2.7270 *
5480
2650 *
0360 *
9740
1.2700
0.1795 *
-1.9250
3.5800 -1.3430
6.0950 -0.7838
4 . 3540 -1.9030
.4690 -3.0880 *
2 . 3190
-2.7890
9.9560 -1.9810
-2.5400 *
0 -1.4210 *
7.6150 -3.4110
1.5730 -4.0500 *
2.7310 -3.0120 *
0 -1.9020
-1.6490
3.2770 -1.2130 *
6.6620 -1.9400
-1.6450 *
-2.3520 *
2.1060 -2.7470 *
0880
G-60
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WMETHSUM) are 12, 17, 52, and 60. Node 10 has the lowest average level of Log(WMETHSUM). The
nodes are numbered in bold on Figure G.3.9.a and the final split characteristics are bolded in the above tree description.
Node 12 is characterized by participants with WHNCHEMO > 6.5, HEIGHT < 44.16 (inches), and CHLDTM7 < 1.5. The average level of WMETHSUM
for these participants was 0.8357 nmoles/g Creatinine (Log(WMETHSUM) = -0.1795) .
Node 17 is characterized by participants with WHNCHEMO < 6.5, SCHOOL < 6.5, WCHLPYRF < 0.56191, and WPERMSUM > 10.082. The average level
of WMETHSUM for these participants was 0.8158 nmoles/g Creatinine (Log(WMETHSUM) = -0.2036) .
Node 60 is characterized by participants with WHNCHEMO > 6.5, HEIGHT > 44.16 (inches), HEIGHT > 46.595 (inches), WHERTIME < 4, and
WDUSTBAL < 0.450858. The average level of WMETHSUM for these participants was 0.2973 nmoles/g Creatinine (Log(WMETHSUM) = -1.2130).
Node 52 is characterized by participants with WHNCHEMO > 6.5, HEIGHT < 44.16 (inches), CHLDTM7 > 1.5, WCHLPYRF < 0.60826, and WDIAZNON <
0.0895515. The average level of WMETHSUM for these participants was 0.4567 nmoles/g Creatinine (Log(WMETHSUM) = -0.7838) .
Node 10 is characterized by participants with WHNCHEMO < 6.5, SCHOOL > 6.5, and WHEEL < 1.5. The average level of WMETHSUM for these
participants was 0.0141 nmoles/g Creatinine (Log(WMETHSUM) = -4.2650).
Figure G.3.9.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All Questions and House Dust Measurements for 130 Yuma Study Participants:
Summary Statistics for Nodes in CART Tree (Figure G.3.9.a)
G-61
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WMETHSUM
height<44.16
chldtm 1 <1.5
wherm J2<1.5
-0.4536
4
wpermsurr
-3.3790
schoc
wherm
<3.78236
whertine<4
-1.8280 I I
-2.3140-2.9940
wchlpyrf<
height<47.16
J4<1.5
3.123506
dadwo
-4.0040-3.0600
K2.5 hejghft42.94 24 wch^yrf<4
k<1.5
wpermsurr
<1.04334
schoc
.0360963
ay<29
hourav\
-2.8630
weight :55.78
weight<
K5.5
52.115 whertir
1
-3.4540
wdiaznon< ).0464268
-1.7460-0.4441
47
-1.4670 -1.0430-1.6820
120
-0.7509
e<1.5
31
-1.8320 |wpermsurr <4.94263
-3.3120
-3.0250-1.8980
-1.5690-3.1350
Figure G.3.10.a CART Analysis of LWMETHSM [LOG(WMETHSUM)] with Limited Questions and House Dust Measurements for 130 Yuma Study Participants:
CART Tree
G-62
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
* denotes terminal node
1) ROOT 130 237.4000 -2.2580
2) HEIGHT<4 4.16 43 101.7000 -1.8360
4) CHLDTMK1 .5 7 15.8700 -0.4536 *
5) CHLDTM1>1.5 36 69.8800 -2.1050
10) WHERMD2C1.5 5 14.2400 -3.3790 *
11) WHERMD2>1.5 31 46.2100 -1.8990
22) WPERMSUM<3.78236 18 15.5900 -2.4440
44) SCHOOL<2.5 5 6.7420 -1.8280 *
45) SCHOOL>2.5 13 6.2240 -2.6800
90) WHERTIME<4 6 3.2790 -2.3140 *
91) WHERTIME>4 7 1.4510 -2.9940 *
23) WPERMSUM>3.78236 13 17.9000 -1.1450
46) HEIGHT<42.94 7 6.4150 -1.7460 *
47) HEIGHT>42.94 6 6.0080 -0.4441 *
3) HEIGHT>4 4.16 87 124.2000 -2.4660
6) HEIGHT<47.16 52 61.4000 -2.8390
12) WHERMD4<1.5 16 11.1700 -3.5320
24) WCHLPYRF<0.123506 8 5.6550 -4.0040 *
25) WCHLPYRF>0.123506 8 1.9540 -3.0600 *
13) WHERMD4>1.5 36 39.1300 -2.5310
26) DADWORK0.0360963 24 19.9200 -2.5660
106) HOURAWAY<29 5 2.0940 -1.8320 *
107) HOURAWAY>2 9 19 14.4200 -2.7600
214) WDIAZNONCO.0464268 6 2.3920 -3.3120 *
215) WDIAZNON>0.0464268 13 9.3600 -2.5050
430) WPERMSUM<4.94263 7 1.4000 -3.0250 *
431) WPERMSUM>4.94263 6 3.8570 -1.8980 *
27) DADWORK>l.5 6 4.2220 -1.4670 *
7) HEIGHT>47.16 35 44.8800 -1.9130
14) WPERMSUMCl.04334 8 6.1140 -2.8630 *
15) WPERMSUM>1.04334 27 29.4000 -1.6310
30) SCHOOL<5.5 22 19.4500 -1.8310
60) WEIGHT<55.7 8 10 3.3010 -1.3620
120) WEIGHT<52.115 5 1.3320 -1.0430 *
121) WEIGHT>52.115 5 0.9485 -1.6820 *
61) WEIGHT>55.7 8 12 12.1200 -2.2220
122) WHERTIMEC1.5 7 2.9680 -1.5690 *
123) WHERTIME>1.5 5 1.9990 -3.1350 *
31) SCHOOL>5.5 5 5.1990 -0.7509 *
G-63
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WMETHSUM) are 47, 4, 31, and 120. Node 24 has the lowest average level of Log(WMETHSUM). The
nodes are numbered in bold on Figure G.3.10.a and the final split characteristics are bolded in the above tree description.
Node 47 is characterized by participants with HEIGHT < 44.16 (inches), CHLDTM1 > 1.5, WHERMD2 > 1.5, WPERMSUM > 3.78236, and HEIGHT >
42.94 (inches) . The average level of WMETHSUM for these participants was 0.6414 nmoles/g Creatinine (Log(WMETHSUM) = -0.4441) .
Node 4 is characterized by participants with HEIGHT < 44.16 (inches), and CHLDTM1 < 1.5. The average level of WMETHSUM for these
participants was 0.6353 nmoles/g Creatinine (Log(WMETHSUM) = -0.4536).
Node 31 is characterized by participants with HEIGHT > 44.16 (inches), HEIGHT > 47.16 (inches), WPERMSUM > 1.04334, and SCHOOL > 5.5. The
average level of WMETHSUM for these participants was 0.4719 nmoles/g Creatinine (Log(WMETHSUM) = -0.7509) .
Node 120 is characterized by participants with HEIGHT > 44.16 (inches), HEIGHT > 47.16 (inches), WPERMSUM > 1.04334, SCHOOL < 5.5, WEIGHT
< 55.78 (pounds) and WEIGHT < 52.115 (pounds). The average level of WMETHSUM for these participants was 0.3524 nmoles/g Creatinine
(Log(WMETHSUM) = -1.0430).
Node 24 is characterized by participants with HEIGHT > 44.16 (inches), HEIGHT < 47.16 (inches), WHERMD4 < 1.5, and WCHLPYRF < 0.123506.
The average level of WMETHSUM for these participants was 0.0182 nmoles/g Creatinine (Log(WMETHSUM) = -4.0040) .
Figure G.3.10.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with Limited Questions and House Dust Measurements for 130 Yuma Study Participants:
Summary Statistics for Nodes in CART Tree (Figure G.3.10.a)
G-64
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WMETHSUM
j/hnchemo<6.5
wopbal<0.C
w herti
r
sprayf
wpermsun <1 0.082
swchlpyr' 1.19023
me<4vdustsum
~l I
d < 1.5 swdustsm
<4.84787
-3.0240-2.5140
-2.4850-1.3 55 0-3.6 220-2.0 63 0
-3.7820
chldtm
00298583 wchlpy rf<
<35.0927
0.751 59
height« 44.1 6
7 < 1.5
wchlpyrf<
-0.1 795
12
wdiaznon<
wopbal<0.C
0.60826
.0895515
ncatw
height<
46.595
00266113
w he rti
kd<4wchlpyrf<
wdustbal<
n e< 4
D. 450858
1.5 -2.7470
-1.2130 I I
5.1 41 61 8 -1.6450-2.3520
-3.0880
-2.5 400-1 .421 0-4.0500-3.01 20
57
-0.2036-3.1 61 0
18
-0.7838-1.9030
52
Figure G.3.11.a CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All Questions and House and School Dust Measurements for 130 Yuma Study
Participants: CART Tree
G-65
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
* denotes terminal node
1) ROOT 130 237.4000 -2.2580
2) WHNCHEMO<6.5 57 100.6000 -2.6210
4) SWCHLPYRCl.19023 48 76.1800 -2.4040
8) WPERMSUM<10.082 37 36.0000 -2.6580
16) WOPBALCO.000298583 23 11.9400 -2.4100
32) SPRAYFLDC1.5 12 2.3950 -2.8120
64) WHERTIME<4 7 0.9452 -3.0240 *
65) WHERTIME>4 5 0.6906 -2.5140 *
33) SPRAYFLD>1.5 11 5.4950 -1.9720
66) WDUSTSUM<4.84787 6 0.9541 -2.4850
67) WDUSTSUM>4.84787 5 1.0580 -1.3550
17) WOPBAL>0.000298583 14 20.3200 -3.0650
34) SWDUSTSM<35.0927 9 9.0480 -3.6220 *
35) SWDUSTSM>35.0927 5 3.4580 -2.0630 *
9) WPERMSUM>10.082 11 29.7300 -1.5480
18) WCHLPYRF<0.75159 6 4.0920 -0.2036 *
19) WCHLPYRF>0.75159 5 1.7830 -3.1610 *
5) SWCHLPYRXL.19023 9 10.0000 -3.7820 *
3) WHNCHEMO>6.5 73 123.4000 -1.9740
6) HEIGHT<4 4.16 24 46.8300 -1.2700
12) CHLDTM7<1.5 9 3.5000 -0.1795 *
13) CHLDTM7>1.5 15 26.2000 -1.9250
26) WCHLPYRFC0.60826 10 13.5800 -1.3430
52) WDIAZNON<0.0895515 5 6.0950 -0.7838
53) WDIAZNON>0.0895515 5 4.3540 -1.9030
27) WCHLPYRF>0.60826 5 2.4690 -3.0880 *
7) HEIGHT>4 4.16 49 58.8900 -2.3190
14) HEIGHT<4 6.595 23 29.1300 -2.7890
28) WOPBALCO.000266113 10 9.9560 -1.9810
56) NCATWRKD<4 5 4.0880 -2.5400 *
57) NCATWRKD>4 5 2.7380 -1.4210 *
29) WOPBAL>0.000266113 13 7.6150 -3.4110
58) WCHLPYRFC0.141618 5 1.5730 -4.0500 *
59) WCHLPYRF>0.141618 8 2.7310 -3.0120 *
15) HEIGHT>4 6.595 26 20.1500 -1.9020
30) WHERTIME<4 20 12.4800 -1.6490
60) WDUSTBALC0.450858 8 3.2770 -1.2130 *
61) WDUSTBAL>0.450858 12 6.6620 -1.9400
122) SEXC1.5 7 2.0010 -1.6450 *
123) SEX>1.5 5 3.2010 -2.3520 *
31) WHERTIME>4 6 2.1060 -2.7470 *
G-66
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
The four nodes with the highest average levels of Log(WMETHSUM) are 12, 18, 52, and 57. Node 5 has the lowest average level of Log(WMETHSUM). The
nodes are numbered in bold on Figure G.3.11.a and the final split characteristics are bolded in the above tree description.
Node 12 is characterized by participants with WHNCHEMO > 6.5, HEIGHT < 44.16 (inches), and CHLDTM7 < 1.5. The average level of WMETHSUM
for these participants was 0.8357 nmoles/g Creatinine (Log(WMETHSUM) = -0.1795) .
Node 18 is characterized by participants with WHNCHEMO < 6.5, SWCHLPYR < 1.19023, WPERMSUM > 10.082, and WCHLPYRF < 0.75159. The average
level of WMETHSUM for these participants was 0.8158 nmoles/g Creatinine (Log(WMETHSUM) = -0.2036) .
Node 52 is characterized by participants with WHNCHEMO > 6.5, HEIGHT > 44.16 (inches), CHLDTM7 > 1.5, WCHLPYRF < 0.60826, and WDIAZNON <
0.895515. The average level of WMETHSUM for these participants was 0.4567 nmoles/g Creatinine (Log(WMETHSUM) = -0.7838) .
Node 57 is characterized by participants with WHNCHEMO > 6.5, HEIGHT > 44.16 (inches), HEIGHT < 46.595 (inches), WOPBAL < 0.000266113, and
NCATWRKD > 4. The average level of WMETHSUM for these participants was 0.2415 nmoles/g Creatinine (Log(WMETHSUM) = -1.4210) .
Node 5 is characterized by participants with WHNCHEMO < 6.5, and SWCHLPYR > 1.19023. The average level of WMETHSUM for these participants
was 0.0228 nmoles/g Creatinine (Log(WMETHSUM) = -3.7820).
Figure G.3.11.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with All Questions and House and School Dust Measurements for 130 Yuma Study
Participants: Summary Statistics for Nodes in CART Tree (Figure G.3.11.a)
G-67
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Log WMETHSUM
heig ht<44.16
chldtml <1.5
wherm 12<1.5
-0.4536
4
wpermsurr
-3.3790
schoc
wherm
<3.78236
l<2.5 swdiazno<
wherti
ie<4
-1.8280
I I
-2.3140-2.9940
swdstbal
0.3975^9.
height* 46.03
i
height- 47.16
J4<1.5
4.01061
dadwo
7930-3.4770
-4.4730
swdiazno<
25
wchlpy rf"
k< 1.5
wpermsurr
<1.04334
schoc
0.197005
0.20494 swopbal
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Legend:
node#), split characteristic, n, std deviation, mean
* denotes terminal node
1) ROOT 130 237.4000 -2.2580
2) HEIGHT<4 4.16 43 101.7000 -1.8360
4) CHLDTMK1 .5 7 15.8700 -0.4536 *
5) CHLDTM1>1.5 36 69.8800 -2.1050
10) WHERMD2C1.5 5 14.2400 -3.3790 *
11) WHERMD2>1.5 31 46.2100 -1.8990
22) WPERMSUM<3.78236 18 15.5900 -2.4440
44) SCHOOL<2.5 5 6.7420 -1.8280 *
45) SCHOOL>2.5 13 6.2240 -2.6800
90) WHERTIME<4 6 3.2790 -2.3140 *
91) WHERTIME>4 7 1.4510 -2.9940 *
23) WPERMSUM>3.78236 13 17.9000 -1.1450
46) SWDIAZNOCO.397569 6 4.5090 -1.9740 *
47) SWDIAZNO>0.397569 7 5.7350 -0.4347 *
3) HEIGHT>4 4.16 87 124.2000 -2.4660
6) HEIGHT<47.16 52 61.4000 -2.8390
12) WHERMD4<1.5 16 11.1700 -3.5320
24) SWDSTBAL<4.01061 11 2.9830 -3.1040
48) HEIGHT<4 6.03 6 0.8057 -2.7930 *
49) HEIGHT>4 6.03 5 0.9004 -3.4770 *
25) SWDSTBAL>4.01061 5 1.7430 -4.4730 *
13) WHERMD4>1.5 36 39.1300 -2.5310
26) DADWORKC1.5 30 26.7600 -2.7440
52) SWDIAZNOCO.197005 10 3.6930 -3.3190
104) WCHLPYRFCO.20494 5 0.5163 -3.8550 *
105) WCHLPYRF>0.20494 5 0.3018 -2.7830 *
53) SWDIAZNO>0.197005 20 18.1000 -2.4560
106) SWOPBALCO.0769847 13 9.5600 -2.0990
212) WPERMSUM<3.97659 7 2.5990 -2.5670
213) WPERMSUM>3.97659 6 3.6320 -1.5520
107) SWOPBAL>0.0769847 7 3.7850 -3.1210 *
27) DADWORK>l.5 6 4.2220 -1.4670 *
7) HEIGHT>47.16 35 44.8800 -1.9130
14) WPERMSUMCl.04334 8 6.1140 -2.8630 *
15) WPERMSUM>1.04334 27 29.4000 -1.6310
30) SCHOOL<5.5 22 19.4500 -1.8310
60) WEIGHT<55.7 8 10 3.3010 -1.3620
120) WEIGHT<52.115 5 1.3320 -1.0430 *
121) WEIGHT>52.115 5 0.9485 -1.6820 *
61) WEIGHT>55.7 8 12 12.1200 -2.2220
122) WHERTIMEC1.5 7 2.9680 -1.5690 *
G-69
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
123) WHERTIME>1.5 5 1.9990 -3.1350 *
31) SCHOOL>5.5 5 5.1990 -0.7509 *
The four nodes with the highest average levels of Log(WMETHSUM) are 47, 4, 31, and 120. Node 25 has the lowest average level of Log(WMETHSUM). The
nodes are numbered in bold on Figure G.3.12.a and the final split characteristics are bolded in the above tree description.
Node 47 is characterized by participants with HEIGHT < 44.16 (inches), CHLDTM1 > 1.5, WHERMD2 > 1.5, WPERMSUM > 3.78236, and SWDIAZNO >
0.397569. The average level of WMETHSUM for these participants was 0.6475 nmoles/g Creatinine (Log(WMETHSUM) = -0.4347).
Node 4 is characterized by participants with HEIGHT < 44.16 (inches), and CHLDTM1 < 1.5. The average level of WMETHSUM for these
participants was 0.6353 nmoles/g Creatinine (Log(WMETHSUM) = -0.4536).
Node 31 is characterized by participants with HEIGHT > 44.16 (inches), HEIGHT > 47.16 (inches), WPERMSUM > 1.04334, and SCHOOL > 5.5. The
average level of WMETHSUM for these participants was 0.4719 nmoles/g Creatinine (Log(WMETHSUM) = -0.7509) .
Node 120 is characterized by participants with HEIGHT > 44.16 (inches), HEIGHT > 47.16 (inches), WPERMSUM > 1.04334, and SCHOOL < 5.5,
WEIGHT < 55.78 (pounds) and WEIGHT < 52.115 (pounds). The average level of WMETHSUM for these participants was 0.3524 nmoles/g Creatinine
(Log(WMETHSUM) = -1.0430).
Node 25 is characterized by participants with HEIGHT > 44.16 (inches), HEIGHT < 47.16 (inches), WHERMD4 < 1.5, and SWDSTBAL > 4.01061. The
average level of WMETHSUM for these participants was 0.0114 nmoles/g Creatinine (Log(WMETHSUM) = -4.4730) .
Figure G.3.12.b CART Analysis of LWMETHSM [LOG(WMETHSUM)] with Limited Questions and House and School Dust Measurements for 130 Yuma Study
Participants: Summary Statistics for Nodes in CART Tree (Figure G.3.12.a)
G-70
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Tables G.3.4 and G.3.5 show the questions and dust measurements selected by CART as classifiers of exposure measurement levels under the six
scenarios for LWETHSUM and LWMETHSM (Table G.3.2), respectively. For ease of presentation, these classifiers will be termed predictors,
although these analyses are not performed with the intent of offering traditional predictive tools as in regression analysis. Instead CART is used as a
tool to understand the factors and the interactions of the factors that may affect the exposure levels found in the Yuma Study participants.
The LTD questions (X in column LTD Q) were considered more likely to be predictors of exposure level. The shaded rows represent questions or
dust measurements that were selected as predictors or differentiators of exposure level in the scenarios a majority of the time (> 50%). Some
questions and the dust measurements were not used in all six scenarios. Questions marked as LTD were included in six scenarios; the other questions
were included in only three scenarios. House dust measurements were included in four scenarios; school dust measurements were included in only
two scenarios. Thus, the majority (>50%) criterion is based on the number of scenarios in which the question or measurement was included as a
potential predictor in a CART analysis.
The CART analyses were performed as preliminary indicators of questions and measurements that may be useful in classifying a child's level of
exposure to pesticides (section G.2.4.1). No fine-tuning of the CART trees was performed to create the "best" models under each scenario. Thus the
results in the following tables should be considered as identifying the predictors "more likely" to be the primary classifiers or surrogate predictors for
the primary classifiers.
Table G.3.4 Results of Classifying Yuma Study Children's Measurements of LWETHSUM [Log(WETHSUM)]a for Six Scenarios of Predictors
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
% Scenarios
with
Predictor
Selected
Number of
Scenarios
with
Predictor
Included
Questions
SEX
Child's gender
X
*
*
33
6
HEIGHT
Child's height (inches)
X
*
*
*
50
6
WEIGHT
Child's weight (lbs)
X
*
*
*
*
*
*
100
6
G-71
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
% Scenarios
with
Predictor
Selected
Number of
Scenarios
with
Predictor
Included
SCHOOL
Child's school
X
*
*
*
*
*
83
6
GRADE
Child's grade
X
*
*
*
*
67
6
ETHNIC
Child's ethnic and racial background
0
3
CHEMINHS
Pesticides used inside home last month?
0
3
WHOCHEMI
Who applied pesticides inside the house?
*
33
3
LIVINGRM
Living room treated with pesticides?
*
33
3
FAMILYRM
Family room treated with pesticides?
*
4c
67
3
DININGRM
Dining room treated with pesticides?
0
3
KITCHEN
Was kitchen treated with pesticides?
0
3
BATHROOM
Bathroom treated with pesticides?
0
3
BEDROOM
Bedroom treated with pesticides?
0
3
CHILDBED
Child's bedroom treated with pesticides?
0
3
BASEMENT
Basement treated with pesticides?
0
3
NRMSPRYD
Number of rooms sprayed last month
0
3
OTHERRM
Other rooms treated with pesticides?
*
*
67
3
OFTCHEMI
How often is home treaedt for pests?
0
3
CHEMOUTH
Pesticides used outside home last month?
0
3
WHOCHEMO
Who applied pesticides outside house?
0
3
G-72
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
% Scenarios
with
Predictor
Selected
Number of
Scenarios
with
Predictor
Included
FARFIELD
Distance between home and agricultural
field
*
*
67
3
CLOSEAPP
Distance between home and nearest
application of pesticides
*
33
3
WHEEL
Distance between home and field - rotary
wheel
0
3
HOWCHEMO
How pesticides were applied to fields
0
3
WHNCHEMO
Last time field treated with pesticides?
*
*
67
3
VEGGIES
How often child eats local fresh fruit/veg?
0
3
WASHVEGI
How often wash local fresh fruit/veg before
eating?
0
3
HOURAWAY
Number hours/wk child not at home
X
*
*
*
*
*
83
6
CHLDTM1
Child spends time in another home?
X
0
6
CHLDTM2
Child spends time at day care center?
X
0
6
CHLDTM3
Child spends time at school?
X
*
*
*
*
*
*
100
6
CHLDTM4
Child spends time at sport event?
X
0
6
CHLDTM5
Child spends time playing in field?
X
0
6
CHLDTM6
Child spends time playing in irrigation
water?
X
0
6
CHLDTM7
Child spends time playing outside?
X
0
6
WHERTIME
Room where child spends most awake time
X
4c
*
*
•k
67
6
G-73
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
% Scenarios
with
Predictor
Selected
Number of
Scenarios
with
Predictor
Included
SPRAYFLD
Child in yard when fields sprayed or
dusted?
*
33
3
WATERSR1
Drinking water source - public/commercial
X
0
6
WATERSR2
Drinking water source - private well
X
0
6
WATERSR3
Drinking water source - cistern
X
0
6
DADWORK
Is the father currently employed?
X
*
4c
4c
4c
67
6
NCATWRKD
Father's occupation - categories
*
*
67
3
DADPEST
Are pesticides used where father works?
0
3
DADCON2
Father's occupation - location and
pesticide use
*
33
3
MOMWORK
Mother now employed (not as housewife)?
X
0
6
NCATWRKM
Mother's occupation - categories
0
3
MOMPEST
Are pesticides used where mother works?
0
3
MOMCON2
Mother's occupation - location and
pesticide use
*
33
3
ADLTPEST
Non-parent in home works where pesticides
used?
0
3
CHILDFLD
Child worked in fields last month?
*
33
3
WHENFILD
Last time child was in work field
0
3
WHERMD1
Family med care at private medical clinic
X
*
*
33
6
G-74
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
% Scenarios
with
Predictor
Selected
Number of
Scenarios
with
Predictor
Included
WHERMD2
Family med care at health dept clinic
X
*
17
6
WHERMD3
Family med care at other med clinic
X
0
6
WHERMD4
Family med care in Mexico
X
0
6
WHERMD5
No access to medical care
X
0
6
WHERMD6
Family med care at other place
X
0
6
WHERMD7
Family med care - do not know
X
0
6
POISON
Anyone treated for pesticide poison?
X
0
6
HOWCHILD
Child's health in general
X
*
*
33
6
LICE
Child treated for head lice past six months?
0
3
INSURED
Is child covered by medical insurance?
X
0
6
House Dust Measurement Sums
WCHDNSUM
Weighted sum of alpha-chlordane and
gamma-chlordane
0
4
WCHLPYRF
Weighted chlorpyrifos
*
*
50
4
WDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and
4,4'DDT
*
25
4
WDIAZNON
Weighted diazinon
*
*
50
4
WDUSTBAL
Weighted sum of dust analytes except OP
pesticides
*
*
50
4
WDUSTSUM
Weighted sum of all dust analytes
*
*
*
*
100
4
G-75
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
% Scenarios
with
Predictor
Selected
Number of
Scenarios
with
Predictor
Included
WOPBAL
Weighted sum of OP pesticides except
chlorpyrifos, diazinon, permethrins, and o-
phenyiphenoi
0
4
WOPSUM
Weighted sum of OP pesticides
4c
4c
4c
4c
100
4
WPERMSUM
Weighted sum of cis-permethrin and trans-
permethrin
4c
4c
4c
4c
100
4
School Dust Measurement Sums
SWCHDNSM
Weighted sum of alpha-chlordane and
gamma-chlordane
0
2
SWCHLPYR
Weighted chlorpyrifos
0
2
SWDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and
4,4'DDT
4c
50
2
SWDIAZNO
Weighted diazinon
0
2
SWDSTBAL
Weighted sum of dust analytes except OP
pesticides
4c
4c
100
2
SWDUSTSM
Weighted sum of all dust
0
2
SWOPBAL
Weighted sum of OP pesticides except
chlorpyrifos, diazinon, permethrins, and o-
phenylphenol
4c
50
2
SWOPSUM
Weighted sum of OP pesticides
4c
4c
100
2
SWPERMSM
Weighted sum of cis-permethrin and trans-
permethrin
0
2
Number of predictors selected in scenario
15
11
14
14
18
13
G-76
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
a log (molar-weighted sum of ethylated DAPs adjusted for creatinine) - see Appendix F.
b See Table G.2.1 for extended descriptions of questions, and Appendix F for descriptions of dust sums.
c Questions considered more likely to be predictors.
* Question or measurement was selected as predictor in CART analysis for this scenario.
G-77
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table G.3.5 Results of Classifying Yuma Study Children's Measurements of LWMETHSM [Log(WMETHSUM)]a for Six Scenarios of Predictors
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
%
Scenarios
Predictor
Selected
Number of
Scenarios
Predictor
Included
Questions
SEX
Child's gender
X
*
*
33
6
HEIGHT
Child's height (inches)
X
*
*
*
*
*
*
100
6
WEIGHT
Child's weight (lbs)
X
*
*
*
*
67
6
SCHOOL
Child's school
X
*
*
*
*
*
83
6
GRADE
Child's grade
X
0
6
ETHNIC
Child's ethnic and racial background
0
3
CHEMINHS
Pesticides used inside home last month?
0
3
WHOCHEMI
Who applied pesticides inside the house?
*
33
3
LIVINGRM
Living room treated with pesticides?
0
3
FAMILYRM
Family room treated with pesticides?
0
3
DININGRM
Dining room treated with pesticides?
0
3
KITCHEN
Kitchen treated with pesticides?
0
3
BATHROOM
Bathroom treated with pesticides?
0
3
BEDROOM
Bedroom treated with pesticides?
0
3
CHILDBED
Child's bedroom treated with pesticides?
0
3
G-78
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
%
Scenarios
Predictor
Selected
Number of
Scenarios
Predictor
Included
BASEMENT
Basement treated with pesticides?
0
3
NRMSPRYD
Number of rooms sprayed last month
0
3
OTHERRM
Other rooms treated with pesticides?
0
3
OFTCHEMI
How often is home treated for pests?
*
33
3
CHEMOUTH
Pesticides used outside home last month?
0
3
WHOCHEMO
Who applied pesticides outside house?
*
33
3
FARFIELD
Distance between home and agricultural
field
*
33
3
CLOSEAPP
Distance between home and nearest
application of pesticides
0
3
WHEEL
Distance between home and field - rotary
wheel
*
*
67
3
HOWCHEMO
How pesticides were applied to fields
0
3
WHNCHEMO
Last time field treated with pesticides?
*
100
3
VEGGIES
How often child eats local fresh fruit/veg?
0
3
WASHVEGI
How often wash local fresh fruit/veg before
eating?
0
3
HOURAWAY
Number hours/wk child not at home
X
*
17
6
CHLDTM1
Child spends time in another home?
X
*
*
*
50
6
CHLDTM2
Child spends time at day care center?
X
0
6
G-79
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
%
Scenarios
Predictor
Selected
Number of
Scenarios
Predictor
Included
CHLDTM3
Child spends time at school?
X
0
6
CHLDTM4
Child spends time at sport event?
X
0
6
CHLDTM5
Child spends time playing in field?
X
0
6
CHLDTM6
Child spends time playing in irrigation
water?
X
0
6
CHLDTM7
Child spends time playing outside?
X
4c
4c
4c
50
6
WHERTIME
Room where child spends most awake time
X
4c
*
4c
ilc
100
6
SPRAYFLD
Child in yard when fields sprayed or
dusted?
4c
33
3
WATERSR1
Drinking water source - public/commercial
X
0
6
WATERSR2
Drinking water source - private well
X
0
6
WATERSR3
Drinking water source - cistern
X
0
6
DADWORK
Is the father currently employed?
X
4c
4c
4c
50
6
NCATWRKD
Father's occupation - categories
4c
4c
67
3
DADPEST
Are pesticides used where father works?
0
3
DADCON2
Father's occupation - location and pesticide
use
4c
33
3
MOMWORK
Mother now employed (not as housewife)?
X
4c
17
6
G-80
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
%
Scenarios
Predictor
Selected
Number of
Scenarios
Predictor
Included
NCATWRKM
Mother's occupation - categories
*
33
3
MOMPEST
Are pesticides used where mother works?
0
3
MOMCON2
Mother's occupation - location and
pesticide use
0
3
ADLTPEST
Non-parent in home works where pesticides
used?
0
3
CHILDFLD
Child worked in fields last month?
0
3
WHENFILD
Last time child was in work field
0
3
WHERMD1
Family med care at private medical clinic
X
*
17
6
WHERMD2
Family med care at health dept clinic
X
*
*
*
50
6
WHERMD3
Family med care at other med clinic
X
0
6
WHERMD4
Family med care in Mexico
X
*
*
*
50
6
WHERMD5
No access to medical care
X
0
6
WHERMD6
Family med care at other place
X
0
6
WHERMD7
Family med care - do not know
X
0
6
POISON
Anyone treated for pesticide poison?
X
0
6
HOWCHILD
Child's health in general
X
*
*
33
6
LICE
Child treated for head lice past six months?
0
3
G-81
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
%
Scenarios
Predictor
Selected
Number of
Scenarios
Predictor
Included
INSURED
Is child covered by medical insurance?
X
0
6
House Dust Measurement Sums
WCHDNSUM
Weighted sum of alpha-chlordane and
gamma-chlordane
0
4
WCHLPYRF
Weighted chlorpyrifos
ilc
4c
100
4
WDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and
4,4'DDT
0
4
WDIAZNON
Weighted diazinon
4c
4c
4c
75
4
WDUSTBAL
Weighted sum of dust analytes except OP
pesticides
4c
4c
50
4
WDUSTSUM
Weighted sum of all dust analytes
4c
25
4
WOPBAL
Weighted sum of OP pesticides except
chlorpyrifos, diazinon, permethrins, and o-
phenylphenol
4c
4c
50
4
WOPSUM
Weighted sum of OP pesticides
0
4
WPERMSUM
Weighted sum of cis-permethrin and trans-
permethrin
4c
4c
4c
4c
100
4
School Dust Measurement Sums
SWCHDNSM
Weighted sum of alpha-chlordane and
gamma-chlordane
0
2
SWCHLPYR
Weighted chlorpyrifos
4c
50
2
SWDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and
4,4'DDT
0
2
G-82
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor
Brief Description11
LTD
Qc
Scenario 1
All
Questions
Scenario 2
Ltd
Questions
Scenario 3
All
Questions
and House
Dust
Scenario 4
Ltd
Questions
and House
Dust
Scenario 5
All
Questions,
House Dust
and School
Dust
Scenario 6
Ltd
Questions,
House Dust
and School
Dust
%
Scenarios
Predictor
Selected
Number of
Scenarios
Predictor
Included
SWDIAZNO
Weighted diazonon
*
50
2
SWDSTBAL
Weighted sum of dust analytes except OP
pesticides
*
50
2
SWDUSTSM
Weighted sum of all dust analytes
*
50
2
SWOPBAL
Weighted sum of OP pesticides except
chlorpyrifos, diazinon, permethrins, and o-
phenylphenol
*
50
2
SWOPSUM
Weighted sum of OP pesticides
0
2
SWPERMSM
Weighted sum of cis-permethrin and trans-
permethrin
0
2
14
10
14
12
15
13
a log (molar-weighted sum of methylated DAPs, adjusted for creatinine) - see Appendix F.
b See Table G.2.1 for extended descriptions of questions, and Appendix F for descriptions of dust sums.
c Questions considered more likely to be predictors.
* Question or measurement was selected as predictor in CART analysis for this scenario.
G-83
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
In the CART analysis results (Table G.3.4), the questionnaire variable CHLDTM3 (Child spends time at school) was the first-level
predictor for LWETHSUM (Log(WETHSUM)) in all six scenarios and no further splits of the node with the eight "No" responses
occurred. Since all children whose responses were included in the CART analyses were in kindergarten or first grade, neither the
intent of the response nor the reason for the segmentation based on that question were clear. Other CHLDTM questions asked about
the child's extracurricular activities, thus, CHLDTM3 was hypothesized to indicate that a child may have spent "additional" time at
school. The subset of children with a "No" response to CHLDTM3 had the highest measure of ethylated DAPs. This situation may
reflect additional exposure from the home environment because the children were not spending more time at school. Table G.3.6
shows the predictors selected with and without CHLDTM3 as a potential predictor in the CART analyses.
Table G.3.6 Results of Classifying Yuma Study Children's Measurements of LWETHSUM [Log(WETHSUM)]a for Six Scenarios of Predictors
Including and Excluding CHLDTM3 (Questions Sorted Alphabetically)
Scenario 1
Scenario 2
Scenario 3
Scenario 4
Scenario 5
Scenario 6
Predictor
Description11
LTD
Q
All
Questions
Ltd
Questions
All
Questions
and House
Dust
Ltd
Questions
and House
Dust
All
Questions,
House
Dust and
School
Dust
Ltd
Questions,
House
Dust and
School
Dust
% Scenarios
Predictor
Selected0
Number of
Scenarios
with
Predictor
Included
W
WOd
W
WO
W
wo
W
wo
W
wo
W
wo
W
WO
Questions
ADLTPEST
Non-parent in home works where
pesticides used?
0
0
3
BASEMENT
Basement treated with pesticides?
0
0
3
BATHROOM
Bathroom treated with pesticides?
0
0
3
BEDROOM
Bedroom treated with pesticides?
0
0
3
CHEMINHS
Pesticides used inside home last month?
0
0
3
CHEMOUTH
Pesticides used outside home last
month?
A
0
33
3
CHILDBED
Child's bedroom treated with pesticides?
0
0
3
CHILDFLD
Child worked in fields last month?
*
33
0
3
CHLDTM1
Child spends time in another home?
X
0
0
6
G-84
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Scenario 1
Scenario 2
Scenario 3
Scenario 4
Scenario 5
Scenario 6
Predictor
Description11
LTD
Q
All
Questions
Ltd
Questions
All
Questions
and House
Dust
Ltd
Questions
and House
Dust
All
Questions,
House
Dust and
School
Dust
Ltd
Questions,
House
Dust and
School
Dust
% Scenarios
Predictor
Selected0
Number of
Scenarios
with
Predictor
Included
W
WOd
W
WO
W
WO
W
WO
W
WO
W
WO
W
WO
CHLDTM2
Child spends time at day care center?
X
0
0
6
CHLDTM3
Child spends time at school?
X
*
*
*
*
*
*
100
0
6
CHLDTM4
Child spends time at sport event?
X
0
0
6
CHLDTM5
Child spends time playing in field?
X
0
0
6
CHLDTM6
Child spends time playing-irrigation
water?
X
0
0
6
CHLDTM7
Child spends time playing outside?
X
A
A
A
0
50
6
CLOSEAPP
Distance between home and nearest
application of pesticides
*
33
0
3
DADCON2
Father's occupation - location and
pesticide use
A
*
33
33
3
DADPEST
Are pesticides used where father works?
A
0
33
3
DADWORK
Is the father currently employed?
X
*
*
A
*
*
67
17
6
DININGRM
Dining room treated with pesticides?
0
0
3
ETHNIC
Child's ethnic and racial background
A
A
0
67
3
FAMILYRM
Family room treated with pesticides?
*
*
67
0
3
FARFIELD
Distance between home and agricultural
field
*
A
*
67
33
3
GRADE
Child's grade
X
*
A
*
A
*
A
*
67
50
6
HEIGHT
Child's height (inches)
X
*
A
*
A
*
A
50
50
6
HOURAWAY
Number hours/wk child not at home
X
*
A
*
A
*
A
*
A
*
A
A
83
100
6
HOWCHEMO
How pesticides were applied to fields
0
0
3
G-85
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Scenario 1
Scenario 2
Scenario 3
Scenario 4
Scenario 5
Scenario 6
Predictor
Description11
LTD
Q
All
Questions
Ltd
Questions
All
Questions
and House
Dust
Ltd
Questions
and House
Dust
All
Questions,
House
Dust and
School
Dust
Ltd
Questions,
House
Dust and
School
Dust
% Scenarios
Predictor
Selected0
Number of
Scenarios
with
Predictor
Included
W
WOd
W
WO
W
WO
W
WO
W
WO
W
WO
W
WO
HOWCHILD
ChildAs health in general
X
*
*
33
0
6
INSURED
Is child covered by medical insurance?
X
0
0
6
KITCHEN
Was kitchen treated with pesticides?
0
0
3
LICE
Child treated for head lice past six
months?
0
0
3
LIVINGRM
Living room treated with pesticides?
A
*
33
33
3
MOMCON2
Mother's occupation - location and
pesticide use
*
33
0
3
MOMPEST
Are pesticides used where mother works?
0
0
3
MOMWORK
Mother now employed (not as
housewife)?
X
0
0
6
NCATWRKD
Father's occupation - categories
A
*
*
67
33
3
NCATWRKM
Mother's occupation - categories
0
0
3
NRMSPRYD
Number of rooms in house sprayed last
month
A
A
A
0
100
3
OFTCHEMI
How often is home treated for pests?
0
0
3
OTHERRM
Other rooms treated with pesticides?
*
*
67
0
3
POISON
Anyone treated for pesticide poison?
X
0
0
6
SCHOOL
Child's school
X
A
*
*
*
A
*
*
A
83
50
6
SEX
Child's gender
X
*
*
A
33
17
6
SPRAYFLD
Child in yard when fields sprayed or
dusted?
*
A
33
33
3
VEGGIES
How often child eats local fresh fruit/veg?
A
A
0
67
3
G-86
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Scenario 1
Scenario 2
Scenario 3
Scenario 4
Scenario 5
Scenario 6
Predictor
Description11
LTD
Q
All
Questions
Ltd
Questions
All
Questions
and House
Dust
Ltd
Questions
and House
Dust
All
Questions,
House
Dust and
School
Dust
Ltd
Questions,
House
Dust and
School
Dust
% Scenarios
Predictor
Selected0
Number of
Scenarios
with
Predictor
Included
W
WOd
W
WO
W
WO
W
WO
W
WO
W
WO
W
WO
WASHVEGI
How often wash local fresh fruit/veg
before eating?
0
0
3
WATERSR1
Drinking water source-public/commercial
X
0
0
6
WATERSR2
Drinking water source-private well
X
0
0
6
WATERSR3
Drinking water source-cistern
X
0
0
6
WEIGHT
Child's weight (lbs)
X
*
A
*
A
*
A
*
A
*
A
*
A
100
100
6
WHEEL
Distance between home and field- rotary
wheel
A
A
A
0
100
3
WHENFILD
Last time child was in work field
0
0
3
WHERMD1
Family med care at private medical clinic
X
*
A
*
33
17
6
WHERMD2
Family med care at health dept clinic
X
*
17
0
6
WHERMD3
Family med care at other med clinic
X
0
0
6
WHERMD4
Family med care in Mexico
X
0
0
6
WHERMD5
No access to medical care
X
0
0
6
WHERMD6
Family med care at other place
X
0
0
6
WHERMD7
Family med care-do not know
X
0
0
6
WHERTIME
Room where child spend most awake
time
X
*
A
*
A
*
*
67
33
6
WHNCHEMO
Last time field treated with pesticides?
*
A
A
*
A
67
100
3
WHOCHEMI
Who applied pesticides inside the house?
*
33
0
3
WHOCHEMO
Who applied pesticides outside house?
0
0
3
House Dust Measurement Sums''
G-87
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Scenario 1
Scenario 2
Scenario 3
Scenario 4
Scenario 5
Scenario 6
Predictor
Description11
LTD
Q
All
Questions
Ltd
Questions
All
Questions
and House
Dust
Ltd
Questions
and House
Dust
All
Questions,
House
Dust and
School
Dust
Ltd
Questions,
House
Dust and
School
Dust
% Scenarios
Predictor
Selected0
Number of
Scenarios
with
Predictor
Included
W
WOd
W
WO
W
WO
W
WO
W
WO
W
WO
W
wo
WCHDNSUM
Weighted sum of alpha-chlordane and
gamma-chlordane
A
0
25
4
WCHLPYRF
Weighted chlorpyrifos
*
A
*
A
A
50
75
4
WDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and
4,4'DDT
*
25
0
4
WDIAZNON
Weighted diazinon
*
*
50
0
4
WDUSTBAL
Weighted sum of dust analytes except
OP pesticides
A
*
A
A
*
A
50
100
4
WDUSTSUM
Weighted Sum of all dust analytes
*
A
*
A
*
A
*
A
100
100
4
WOPBAL
Weighted sum of OP pesticides except
chlorpyrifos, diazinon, permethrins, and
o-phenylphenol
0
0
4
WOPSUM
Weighted Sum of OP Pesticides
*
*
A
*
*
A
100
50
4
WPERMSUM
Weighted sum of cis-permethrin and
trans-permethrin
*
A
*
A
*
A
*
A
100
100
4
School Dust Measurement Sums''
SWCHDNSM
Weighted sum of alpha-chlordane and
gamma-chlordane
A
0
50
2
SWCHLPYR
Weighted chlorpyrifos
*
*
100
0
2
SWDDSUM
Weighted sum of 4,4'DDD, 4,4'DDE and
4,4'DDT
*
50
0
2
SWDIAZNO
Weighted diazinon
0
0
2
SWDSTBAL
Weighted sum of dust analytes except
OP pesticides
*
*
100
0
2
SWDUSTSM
Weighted sum of all dust analytes
A
0
50
2
G-88
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Scenario 1
Scenario 2
Scenario 3
Scenario 4
Scenario 5
Scenario 6
Predictor
Description11
LTD
Q
All
Questions
Ltd
Questions
All
Questions
and House
Dust
Ltd
Questions
and House
Dust
All
Questions,
House
Dust and
School
Dust
Ltd
Questions,
House
Dust and
School
Dust
% Scenarios
Predictor
Selected0
Number of
Scenarios
with
Predictor
Included
W
WOd
W
WO
W
WO
W
WO
W
WO
W
WO
W
wo
SWOPBAL
Weighted sum of OP pesticides except
chlorpyrifos, diazinon, permethrins, and
o-phenylphenol
*
A
A
50
100
2
SWOPSUM
Weighted sum of OP pesticides
*
A
*
A
100
100
2
SWPERMSM
Weighted sum of cis-permethrin and
trans-permethrin
A
0
50
2
a log (molar-weighted sum of ethylated DAPs, adjusted for creatinine) - See Appendix F for detailed descriptions of weighted measurement sums.
b See Table G.2.1 for extended descriptions of questions, and Appendix F for descriptions of dust sums.
0 Percents in bold are questions or measurements for which the percent of the analyses without CHLDTM3 was > 50%. (See summary in Table G.3.7)
d W: CHLDTM3 was included as a potential predictor. WO: CHLTM3 was not included as a potential predictor. *, A indicate predictor was selected in W or WO scenarios,
respectively.
G-89
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
A summary of this comparison (Table G.3.6) is included in Table G.3.7. Additional
discussion of results from the CART analyses is included in the section 4.3.2.4.
Table G.3.7 compares the selected predictors from the CART analyses with and without
CHLDTM3 as a potential predictor. The predictors selected when CHLDTM3 was not
included in the CART analyses can be considered surrogate predictors for CHLDTM3 or the
situation that the children represented. Surrogate predictors, although not interchangeable,
offer options for differentiating exposure levels based on available responses. Examples of
surrogates can be seen in the questions selected. For example, the W case includes
FARFIELD, and the WO case includes WHEEL. Both are measures of proximity to
agricultural fields. Pesticide treatment in the house switches from specific rooms to number
of rooms sprayed. Although different dust measurement variables were selected for both
cases (with and without CHLDTM3), overall the analyses track the same pesticides. The
scenarios with CHLDTM3 seemed to include the more specific school parameters, such as
SCHOOL and GRADE, while the scenarios without CHLDTM3 picked up ETHNIC and
VEGGIES.
Table G.3.7 Predictors Selected for Classifying Yuma Study Children's Measurements of
LWETHSUM [Log(WETHSUM)]a for Six Scenarios of Predictors Including and
Excluding CHLDTM3
Predictor"
Description':
LTD
Q
% Scenarios
Predictor
Selected
Number of
Scenarios
with
Predictor
Included
Selection
Comparison
Code0
Wd
WOd
Questions
CHLDTM3
Child spends time at school?
X
100
0
6
A
DADWORK
Is the father currently employed?
X
67
17
6
A
ETHNIC
Child's ethnic and racial
background
0
67
3
C
FAMILYRM
Family room treated with
pesticides?
67
0
3
A
FARFIELD
Distance between home and
agricultural field
67
33
3
A
GRADE
Child's grade
X
67
50
6
A
HOURAWAY
Number hours/wk child not at home
X
83
100
6
B
LIVINGRM
Living room treated with pesticides?
67
33
3
A
NCATWRKD
Father's occupation—categories
67
33
3
A
NRMSPRYD
Number of rooms sprayed last
month
0
100
3
C
OTHERRM
Other rooms treated with
pesticides?
67
0
3
A
SCHOOL
Child's school
X
83
50
6
A
G-90
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Predictor"
Description0
LTD
Q
% Scenarios
Predictor
Selected
Number of
Scenarios
with
Predictor
Included
Selection
Comparison
Code0
Wd
wod
Questions
VEGGIES
How often child eats local fresh
fruit/veg?
0
67
3
C
WEIGHT
Child's weight (lbs)
X
100
100
6
B
WHEEL
Distance between home and field -
rotary wheel
0
100
3
C
WHERTIME
Family room treated with
pesticides?
X
67
33
6
A
WHNCHEMO
Last time field treated with
pesticides?
67
100
3
C
House Dust Measurement Sums
WCHLPYRF
Weighted chlorpyrifos
50
75
4
C
WDUSTBAL
Weighted sum of dust analytes
except OP pesticides
50
100
4
C
WDUSTSUM
Weighted sum of all dust analytes
100
100
4
B
WOPSUM
Weighted sum of OP pesticides
100
50
4
A
WPERMSUM
Weighted sum of cis-permethrin and
trans-permethrin
100
100
4
B
School Dust Measurement Sums
SWCHLPYR
Weighted chlorpyrifos
0
100
2
C
SWDSTBAL
Weighted sum of dust analytes
except OP pesticides
100
0
2
A
SWOPBAL
Weighted sum of OP pesticides
except chlorpyrifos, diazinon,
permethrins, and o-phenylphenol
50
100
2
C
SWOPSUM
Weighted sum of OP pesticides
100
100
2
B
a log (molar-weighted sum of ethylated DAPs, adjusted for creatinine) — See Appendix F for detailed
descriptions of weighted measurement sums.
b Predictors listed are only those that were selected for > 50% of the scenarios either with or without CHLDTM3.
c See Table G.2.1 for extended descriptions of questions, and Appendix F for descriptions of dust sums.
d W: CHLDTM3 was included as a potential predictor.
WO: CHLDTM3 was not included as a potential predictor.
e A: Selected with CHLDTM3, but not selected without CHLDTM3
B: Selected with and without CHLDTM3
C: Selected without CHLDTM3, but not selected with CHLDTM3.
G.3.4 Comparison of Questionnaire Responses for High- and Low-end Measurements
A non-statistical approach was implemented to identify any predictors that could differentiate
between the high and low exposure levels based on the DAP urinary metabolites. In the
previous analyses, CART and CDC (2002), the questionnaire responses, dust measurements,
and urine measurements for all of the principal participants were considered. Because the
G-91
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
range of the distribution of the urine and dust measurement values is limited, it seemed
reasonable to compare the information of participants from the extremes of the distribution.
Thus, approximately 10% of the respondents from the low end of a specific distribution and
approximately 10% of the respondents from the high end of the distribution were selected.
Twenty-one questions considered more likely predictors of a child's pesticide exposure level
were identified. The weighted sum of the responses for each respondent was created from 18
of the questions where the weight was added to the sum if the response indicated a potential
exposure to pesticides. Table G.3.5 shows the questions used in the exposure weighted sum,
and the amounts added to the sum based on the responses. The values of this weighted sum
and the responses to the 21 individual questions were compared between the high- and low-
end values of each measurement sum to determine if any patterns in the responses were
evident.
Table G.3.8 Questions and Weights Used to Create the Exposure Weighted Sum for Comparing High
and Low End Measurements
Q Name
Q Description''
Value
Value Label
Amount Added to
Sum"
adltpest
Nonparent in home works where pesticides
used?
1
Yes
1.0
2
No
childfld
Child worked in fields in last month?
1
Yes
1.0
2
No
3
Do not know
4
No Response
chldtm5
Child spends time playing in field?
1
Yes
1.0
2
No
chldtm6
Child spends time playing in irrigation
water?
1
Yes
1.0
2
No
chldtm7
Child spends time playing outside?
1
Yes
1.0
2
No
closeapp
Distance between home and nearest
application of pesticides
1
In your
yard/garden
1.0
2
In neighbor's
yard
0.5
3
Further away
G-92
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q Name
Q Description''
Value
Value Label
Amount Added to
Sum"
4
Not used near
home
5
Do not know
6
No Response
dadcon2
Father works outdoors + with pesticides
1
Works Inside,
no pesticides
assumed
2
Works Outside,
no pesticides
assumed
3
Works Inside,
pesticides
assumed
0.5
4
Works Outside,
pesticides
assumed
1.0
5
Dad doesn't
work
6
No job response
farfield
Distance between home and agricultural
field
1
250 feet or less
1.0
2
Over 250 feet
3
Do not know
4
No Response
howchemo
How pesticides were applied to fields
1
By airplane
1.0
2
Mechanized
spraying
0.5
3
Hand
application
4
Other (Specify)
5
Not used near
home
6
Do not know
7
No Response
lice
Child treated for head lice past six months?
1
Yes
1.0
2
No
3
No Response
momcon2
Mother works outdoors + with pesticides
G-93
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q Name
Q Description''
Value
Value Label
Amount Added to
Sum"
1
Works Inside,
no pesticides
assumed
2
Works Outside,
no pesticides
assumed
3
Works Inside,
pesticides
assumed
0.5
4
Works Outside,
pesticides
assumed
1.0
5
Mom doesn't
work
6
No job response
oftchemi
How often is home treat for pests?
1
About once a
week
1.0
2
About once a
month
0.5
3
Several times a
year
0.5
4
About once a
year
5
Infrequently
6
Never or not yet
7
Do not know
8
No Response
poison
Anyone treated for pesticide poison?
1
Yes
1.0
2
No
3
Do not know
4
No Response
sprayfld
Child in yard when fields sprayed or
dusted?
1
Yes
1.0
2
No
3
Do not know
4
No Response
veggies
How often child eats local fruit/veg?
-1
No Response
0
Do not know
1
Never
G-94
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Q Name
Q Description''
Value
Value Label
Amount Added to
Sum"
2
About once a
year
3
About once a
month
4
About once a
week
0.5
5
About once a
day
1.0
wheel
Distance between home and field-rotary
wheel
1
< 250 feet
1.0
2
> 250 and < 500
feet
3
> 500 feet
whenfild
Last time child was in work field
1
Today
1.0
2
Yesterday
1.0
3
> 2 days ago
0.5
4
A week ago
0.5
5
> a week ago
6
Do not know
7
Child not in field
8
No Response
whnchemo
Last time field treated with pesticides?
1
Today
1.0
2
Yesterday
1.0
3
> 2 days ago
0.5
4
A week ago
0.5
5
> a week ago
6
Other
7
Do not know
8
Not applicable
9
No Response
a See Table G.2.1 for extended descriptions of questions, and Appendix F for descriptions of dust sums.
b Blank indicates no amount was added to the exposure sum.
Table G.3.9 shows the measurement sums considered and any of the questions that showed a
difference (not statistically significant) in responses between the two ends of the
measurement distribution.
G-95
August 2005
-------
Results Between Questionnaire Responses and Children's Pesticide Exposure Measurements
Table G.3.9 Results from Non-statistical Comparison of Questionnaire Responses between High and
Low End Measurements
Measurement Sum3
Questions Indicating Differences between High and Low Groups'3
WETHSUM +
WMETHSUM
EXPOSURE SUM d, FARFIELDc, WHNCHMOd'e, WHEELd, DADCON2d,
MOMCON2c
WOPSUM
SCHOOL, HOWCHEMOde, FARFIELDd, CLOSEAPPd e, WHEELd,
CHLDTM7d, WHENFILDd, CHLDFLDd
WDUSTSUM
SCHOOL, NRMSRYD0, HOWCHEMOde, OFTCHEMIc, FARFIELDd,
WHNCHMOde, WHEELd, SPRAYFLDd, DADCON2c, MOMCON2d
a See Appendix F for description of sums.
b See Table G.2.1 for abbreviated description of question variables.
c Some difference (> 15%) in responses between participants at both ends of measurement distribution was
evident. Difference was in direction expected, that is, exposure to factor is associated with high-end
measurement value.
d Some difference (> 15%) in responses between participants at both ends of measurement distribution was
evident. Difference was not in direction expected based on current knowledge; that is, t exposure to factor is
associated with low-end measurement values.
e Some difference (> 15%) in responses between participants at both ends of measurement distribution was
evident. Difference is based on response (some exposure to factor) compared to non-response (Don't know,
No response).
The questions that point to some differentiation of the exposure levels are reasonable;
however, most of them show the difference to be in the direction opposite of what is
expected based on current knowledge (Table G.3.9). As noted for the results in CDC (2002),
relationships with the responses are considered one question at a time. This view may hide
interactions with other risk factors or it may point to other factors that have a related effect.
G-96
August 2005
-------
-------
SEPA
United States
Environmental Protection
Aqencv
Office of Research
and Development (8101R)
Washington, DC 20460
EPA/600/R-05/092
August 2 0 05
www.epa.gov
------- |