United States
Environmental Protection
Agency
Office of Research and
Development
Washington, DC 20460
EPA/600/3-90/
January 1990
'003
&EPA
Nonpccupationa!
Pesticide Exposure Study
(NOPES)
Final Report
-------
-------
EPA/600/3-90/003
January 1990
Nonoccupationa! Pesticide Exposure
Study (MOPES)
Final Report
Atmospheric Research and Exposure Assessment Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
-------
Notice
The information in this document has been funded wholly or in part by the U.S.
Environmental Protection Agency under Contract No. 68-02-4544. It has been submitted
to the Agency's peer and administrative review, and it has been approved for
publication as an EPA document. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
-------
Foreword
The Office of Research and Development (ORD), in 1979. first evaluated a new and
innovative research approach for assessing total human exposure to a variety of toxic
chemicals. Since that time, the Total Exposure Assessment Methodology (TEAM)
approach has been employed in several monitoring studies and has subsequently
become an integral component of the monitoring, Total Human Exposure Research
Program. The TEAM approach applies probabilistic population sampling techniques
indoor and outdoor microenvironmental personal exposure monitoring, and human
activity pattern data for multiple routes of exposure to support total human exposure
assessment. The Nonoccupational Pesticide Exposure Study (NOPES) carries this
process one step further by estimating potential human health effects associated with
nonoccupational exposures to pesticides in the study areas and associated monitoring
seasons.
The Atmospheric Research and Exposure Assessment Laboratory located at Research
Triangle Park, North Carolina, is committed to performing goal-oriented, high-quality
ORD studies to characterize air pollutant sources, sinks, transport, and transformations;
assess and predict exposure of humans and ecosystems to environmental pollutants'
and develop monitoring systems and other technologies to determine the status and
trends in pollutant concentrations and the condition of the nation's ecosystems.
Gary J. Foley, Ph.D.
Director
Atmospheric Research and Exposure Assessment Laboratory
Research Triangle Park, North Carolina
in
-------
Abstract
The Nonoccupational Pesticide Exposure Study was the first attempt to develop a
methodology for measuring the potential exposure of specified populations to common
pesticides. In this study, as in other studies utilizing the Total Exposure Assessment
Methodology (TEAM), the exposures were related to actual use patterns. A selected list
of 32 household pesticides were evaluated in two different cities during this study.
Air samples were collected over a 24-hour period in indoor, outdoor and personal
microenvironments. In addition, limited water and dermal contact samples were
collected for selected homes. The study households were selected from stratified
random population samples in two urbanized areas. The samples were collected over
several seasons in areas contrasting a relatively high and low use of pesticides. The
sampling strategy included within-home duplicate, triplicate and replicate samples, as
well as single- season and multi-season sampled homes. This comprehensive sampling
design permitted estimation of short-term and seasonal temporal differences as well as
interpersonal comparisons. Dietary recall, activity pattern, and pesticide use data were
collected through survey questionnaires.
The report discusses the results of the study with an emphasis on the various routes of
exposure (air, water, dermal, and indirectly, food) and their relative contribution to total
human exposure. The effectiveness of the exposure stratification, potential health
effects, consumer awareness highlights, and exploratory analyses of activity patterns
and pesticide use are also included.
IV
-------
Contents
Page
Foreword ...
Abstract '.'.'.'.'.'.'.'.'.'.'.'-.'.'.'-.'.'.' ' ' " ' ' ' •'"
Figures IV
Tables '.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'. V!
Acknowledgments ':'.'.'.'. V"
*"*"*""*•••••••••"•*•••«.... IX
1 Introduction
2 Conclusions
O
3 Future Research Recommendations g
4 Study Design
• Selection of Study Analytes 11
Target Population Definition '.'.'.'.'. 11
Sampling Design 14
Data Collection Procedures '.'.'.'.'.'.'.'.''''. 16
Response Rates ' ' ' ig
Respondent Characteristics •-..'.'.'.'.'.'.'.'.-'.'.'.'. 19
Laboratory Operations • • • •
5 Results and Discussion oq
Air Exposure '.'.'.'.'.'.'.' 23
Water Exposure • • • • .
Dermal Exposure '.'.'.'.'.'.'. ' 38
Dietary Exposure ..'.'.'.'.'.'.'.'.':'.'.'.'/. 42
Relative Contributions of Exposure Pathways 42
Air Exposure and Questionnaire Data Relationships ....'.'.'.'.'.'.'.'.'. • • • •
Effectiveness of Exposure Stratification 43
Exploratory Analyses . . AA
Potential Health Effects '.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'. " '• 53
Consumer Awareness Rn
Data Quality '.'.'.•'.'.'.'.'.'.'.'.'.'.'.',[ QQ
References „
D/
Appendices
A. NOPES Survey Instruments 71
B. Summary Statistics for All Analytes '.'.'. 1J
C. Weighted Percentiles for All Analytes '.'. 163
D. Glossary of Statistical and NOPES Terms '.'. 177
-------
Figures
Number
Page
1 NOPES Study Area in Jacksonville, Florida 15
2 NOPES Study Area in Springfield/Chicopee, Massachusetts 16
3 Chlordane Weighted Cumulative Frequency Distribution for Personal Air
Concentrations 28
4 Chlorpyrifos Weighted Cumulative Frequency Distribution for Personal Air
Concentrations , 28
5 Heptachlor Weighted Cumulative Frequency Distribution for Personal Air
Concentrations • - 28
6 ortho-Phenylphenol Weighted Cumulative Frequency Distribution for
Personal Air Concentrations 28
7 Propoxur Weighted Cumulative Frequency Distribution for Personal Air
Concentrations . 29
8 Chlordane Mean Concentrations ......... 30
9 Chlorpyrifos Mean Concentrations -. . 30
10 Heptachlor Mean Concentrations -.-••- 31
11 ortho-Phenylphenol Mean Concentrations . 31
12 Propoxur Mean Concentrations 32
13 Seasonal Relative Mean Indoor Air Concentrations in Jacksonville as
Percents of Summer Mean Concentrations . . . 34
14 Seasonal Relative Mean Indoor Air Concentrations jn Springfield/Chicopee as
Percents of Spring Mean Concentrations 34
15 Seasonal Relative Mean Outdoor Air Concentrations in Jacksonville as Percents
of Summer Mean Concentrations . 35
16 Seasonal Relative Mean Outdoor Air Concentrations in Springfield/Chicopee
as Percents of Spring Mean Concentrations 35
17 Seasonal Relative Mean Personal Air Concentrations in Jacksonville as Percents
of Summer Mean Concentrations 36
18 Seasonal Relative Mean Personal Air Concentrations in Springfield/Chicopee as
Percents of Spring Mean Concentrations 36
19 Relative Jacksonville and Springfield'Chicopee Mean Indoor Air
Concentrations , . 38
20 Relative Jacksonville and Springfield/Chicopee IvUsan Outdoor Air
Concentrations . . 39
21 Relative Jacksonville and Springfield/Chicopee Mean Personal Air
Concentrations 40
VI
-------
Tables
Number
Page
1 Overview of the NOPES Project 3
2 Estimated Percent of Population with Detectable Levels of Target
, Compounds in Personal Air, Jacksonville, FL - All Three Seasons ....... 7
3 Estimated Percent of Population with Detectable Levels of Target
Compounds in Personal Air, Springfield/Chicopee; MA - Both Seasons ... 8
4 Target Compounds in NOPES , 12
5 Planned Number of Respondents in the NOPES Third-Stage
Sample Design .,'..• 17
6 Response Rates 1 g
7 Third-Stage Respondent and Household Characteristics ; 20
8 Analytical Methods for NOPES Target Compounds '.'.'.'.'.'.'.'.'.'. 21
9 Number of Samples Collected and Analyzed 22
10 Estimated Percent of Jacksonville Population with Detectable
Levels in Air 24
11 Estimated Percent of Springfield/Chicopee Population with Detectable
Levels in Air 25
12 Weighted Arithmetic Mean Concentrations in Jacksonville Air 26
13 Weighted Arithmetic Mean Concentrations in Springfield/Chicopee
A'r --.: 27
14 Indoor, Outdoor, and Personal Air Concentration Correlations 32
15 Seasonal Variation in Number of Detected Analytes in Air . . . 33
16 Seasonal Comparisons 33
17 Local Weather During NOPES Data Collection Periods ........ . . . . . . . . 37
18 Study Area Comparisons 37
19 Replicate Relative Percent Differences \\\\ 40
20 Duplicate, Replicate, and Seasonal Indoor Air Concentration
Differences 41
21 Relative Air and Dietary Exposure Estimates . [ 43
22 Overall Effectiveness of the Exposure Stratification Model 45
23 Ranks of Exposure Category Mean Indoor Air Concentrations of
Commonly Detected Analytes 45
24 Indoor Air Concentration vs. Age of Housing Unit 47
25 Indoor Air Concentration vs. Type of Housing Unit . . 49
26 Indoor Air Fixed-Site Sampler Location Comparison ; 50
27 Indoor Air Concentrations vs. Presence in Household Pesticide Inventory .. 51
28 Indoor Air Termiticide Concentrations vs. Reported Termiticide Use ...... 53
29 Indoor Air Concentrations vs. Indoor Household Insecticide Use 54
30 Indoor Air Concentrations vs. Pesticide Use on Pets 55
VII
-------
Tables (continued)
Number Page
31 a Weighted Estimate of Annual Average Daily Concentrations,
Cancer Risk and Hazard Index for Jacksonville Air (Pesticides Other than
Cyclodiene Termiticides)
31 b Weighted Estimate of Annual Average Daily Concentrations, Cancer
Risk and Hazard Index for Jacksonville Air (Cyclodiene Termiticides) . .
32a Weighted Estimate of Annual Average Daily Concentrations, Cancer
Risk and Hazard Index for Springfield/Chicopee Air (Pesticides Other than
Cyclodiene Termiticides)
32b Weighted Estimate of Annual Average Daily Concentrations, Cancer
Risk and Hazard Index, for Springfield/Chicopee Air (Cyclodiene
Termiticides)
33 Matrix Spike Percent Recoveries
34 Duplicate Relative Percent Differences ,
35 Ranges of Estimated Limits of Detection for GCECD Target
Compounds by Site and Season ,
36 Ranges of Estimated Limits of Detection for GC
-------
Acknowledgments
The success of a project as large and complex as NOPES is a product of the work and
input of many people. The authors wish to thank all those who contributed to the
design, implementation, analysis, and reporting of this study. •
Overall project management was very capably handled by Mr. Andrew E. Bond. Dr.
David T. Mage and Dr. Wayne R. Ott developed the basic study design. Dr. Robert G.
Lewis, Dr. Lance A. Wallace, and Mr. Gerald G, Akland provided expert technical advice
throughout the course of the study. Project quality assurance activities were reviewed
by Mr. Howard Crist and Dr. C. R. Rhodes. Regional EPA assistance was provided by
Mr. Robert Kalayjian and Mr. Kent C. Williams.
Cooperative state and local .officials in Florida and Massachusetts helped ensure the
success of the NOPES data collection effort. We would particularly like to thank Mr.
Don Bayley and Dr. Pat Cowdrey of the Jacksonville Department of Bioenvironmental
Services, Mr. Jeffery Carlson and Mr. Paul Ricco of the Massachusetts Department of
Food and Agriculture, Dr. John Cipolla, Springfield Commissioner of Health, Ms. Colleen
Lasorsa of the Springfield Department of Public Affairs, and Mr. Randy White of the
Springfield Department of Public Health .
We would like to thank Mr. Ellis L. Gundersqn of the U.S. Food and Drug Administration
for providing dietary pesticide concentration data from the Market Basket Survey data
base.
The manuscript benefited from the attention of the following reviewers,who provided
many helpful suggestions:
Gerald Akland Bill Mitchell
David Camann
Michael Firestone
Thomas Hartlage
Merrill Jackson
Victor Kimm
Robert Lewis
John Moore
C. J. Nelson
Dale Pahl
Lance Wallace
Roy Whitmore
Anne Worth
Drs. Roy Whitmore, Robert Mason, and David Shanklin of Research Triangle
Institute (RTI) provided valuable advice on study design, conduct, and analysis. Other
RTI staff who made substantial contributions to this project were G.A. Rush, S.M.
Jones, J.G. Milne, S.L. Branson, D.W. Jackson, and R.C. Boytos. The various
iterations of this manuscript were expertly prepared by P.P. Parker and B.K. Porter.
Southwest Research Institute's (SwRI) activities on NOPES were directed by Mr. David
Camann. Other key staff at SwRI included H.J. Harding, J.P. Hsu, H.J. Schattenberg,
H.G. Wheeler, P. Kuhrt, M. Garza, J.A. Lawless, and D.E. Johnson.
IX
-------
Acknowledgements (continued)
Project management and quality assurance in the early phases of NOPES were
provided by Environmental Monitoring and Services, Inc. (EMSI), under the direction of
Drs. George Colovos and Miriam Lev-On. L. Levan, J.C. Delwiche, and C.C. Lin
contributed to the EMSI effort.
Finally, the authors would like to thank the study respondents, who by giving their time
to answer questions and wear monitors, provided new insight on the uses and
concentrations of pesticides in residential environments.
-------
Section 1
Introduction
In 1984, Congress appropriated FY85 monies to the
U.S. Environmental Protection Agency (EPA) to
assess the level of pesticide exposure experienced by
the general population. Occupational exposure of
specific groups of pesticide users, such as farm
workers and pest control operators, had been
examined and characterized by previous studies
(Wolfe, 1976; Bristol et al., 1984). However, little was
known about the general distribution of
nonoccupational exposures to household pesticides.
To begin to overcome this lack of knowledge, the EPA
Office of Research and Development, in conjunction
with the Office of Pesticide Programs, conducted the
Nonoccupational Pesticide Exposure Study (MOPES).
NOPES was designed to provide initial estimates of
nonoccupational exposure levels and to address the
nature of the variability in exposure.
NOPES was based on the Total Exposure
Assessment Methodology (TEAM) approach to
exposure estimation. The Agency began: developing
the TEAM approach in 1979 for measuring human
exposure to various environmental contaminants (Ott,
1985; Ott et al., 1986). In a TEAM study, probability-
based survey sampling procedures are combined with
questionnaire data collection and modern personal
monitoring techniques to obtain statistically defensible
estimates of exposure levels in the general population.
Data on exposure levels, rates of use, and activity
patterns are then used to develop predictive models
for exposure. The initial application of this innovative
approach (Wallace, 1987) was in the estimation of
exposures to volatile organic compounds (VOCs). The
TEAM approach was also applied to estimating
population exposures to carbon monoxide (Akland et,
al., 1985). The success of these projects prompted
the decision to conduct NOPES as a TEAM study.
NOPES had both methodological and analytical
objectives. NOPES sought to apply the TEAM
approach to a class of chemicals not previously
addressed by TEAM. Therefore, the primary
methodological objective of NOPES was to develop
and refine the monitoring instrumentation, laboratory
procedures, and survey questionnaires needed for a
TEAM study of pesticides. The overall analytical
objective of NOPES was to estimate the levels of
nonoccupational exposure to selected household
pesticides through air, drinking water, food, and
dermal contact. Specific objectives were as follows:
• Estimate exposure levels for the populations of
two urban areas of the United States.
• Assess the relative importance of each exposure
pathway to the overall level of exposure.
• Characterize the components of variability in the
observed exposure levels.
• Investigate and, if possible, model the
relationships between exposure levels, rates of
use, activity patterns, and other factors that could
contribute to variation in exposure levels.
Work on the design phase of NOPES began in 1985.
Southwest Research Institute. (SwRI), of San Antonio,
Texas, developed the methodology for collecting air
samples and analyzing them for 32 selected
pesticides and pesticide degradation products (Hsu et
al., 1988). Emphasis was placed on both identifying
and quantitating the target compounds. Research
Triangle Institute (RTI) of Research Triangle Park,
North Carolina, developed the probability-based
sampling design and the questionnaires needed to
collect information about pesticide use and activity
patterns. The questionnaires and monitoring and
analysis procedures were tested in a pilot study
conducted in Jacksonville, Florida in August and
September 1985 (Lewis et al., 1988).
To permit assessment of regional arid seasonal
variations in exposure levels, the main NOPES data
collection was conducted in three phases:
• Phase I : Summer 1986 in Jacksonville, Florida.
• Phase II : Spring 1987 in Jacksonville, Florida,
and Springfield and Chicopee, Massachusetts.
• Phase III : Winter 1988 in Jacksonville, Florida,
and Springfield and Chicopee, Massachusetts.
The findings of EPA's National Urban Pesticide
Applicator Survey and earlier studies were used to
select two study areas. Jacksonville was selected as
representative of an area of the country with relatively
high pesticide use, and the Springfield region was
-------
selected to represent an area of low to moderate
pesticide use. In both study areas, some sample
members were asked to participate in all phases of
the study, whereas others were recruited only for a
single phase. Monitoring some people in more than
one phase permitted assessment of whether the
overall differences observed between phases were
due to true seasonal variation or due to random
sampling variation. Short-term temporal variation was
addressed by monitoring selected respondents twice
in the same phase.
Sample members were selected from respondents to
a screening questionnaire. The questionnaire collected
data which were used to stratify screening
respondents into three potential-exposure categories.
Members of the high-potential-exposure category were
sampled at a higher rate than medium or low category
members. Members of the high-potential-exposure
category were sampled at a higher rate in an attempt
to improve the characterization of the upper tails of
the estimated air exposure distributions. Because of
the unequal sampling rates, estimation procedures
that incorporated sampling weights (essentially
reciprocals of the probabilities of selection) were used
to produce design-unbiased population estimates.
The following activities were performed for each
sample member who agreed to participate in the
study:
• A study questionnaire was administered.
• A personal air sampler was given to the
participant to wear or keep in close proximity for
24 h.
• Two or more fixed-site air samplers were set up
and run for 24 h. At least one sampler was run in
the respondent's home, and at least one was run
outside the home.
• At the end of the 24-h monitoring period, an
activity log questionnaire was administered.
In some households, drinking water samples were
collected for analyses. Dermal exposure during
pesticide application events was estimated for a small
number of respondents by analyzing cotton gloves
worn during typical application events following the
regular monitoring period.
In all phases, RTI recruited the sample households,
administered the questionnaires, and statistically
analyzed the questionnaire and chemical data. SwRI
performed the environmental monitoring and
laboratory analyses. In Phases I and II, Environmental
Monitoring and Services. Inc. (EMSI), of Camariilo,
California, provided overall program management and
quality assurance. EPA assumed these functions in
Phase III. A series of interim reports provides detailed
information on the conduct and results of each phase
(Lev-On et al., 1987; Immerman et al., 1988a).
During the Jacksonville portion of Phase III, a dust
sampling and analysis study was conducted in
conjunction with the NOPES data collection. This
study, designed to test a method for measuring the
level of pesticides present in residential floor dust, is
described in detail in a separate report (Budd et al.,
1988).
Table 1 presents an overview of the NOPES project.
-------
Table 1. Overview of the NOPES Project
1. Pilot Study, August 23 - August 29, 1985
Nine purposively selected people in Jacksonville. Florida, were recruited to test the procedures and
instrumentation developed for NOPES. Draft versions of the screening questionnaire, study questionnaire,
and activity log were administered to the participants by RTI. Air, water, and glove samples were collected
and analyzed by SwRI. The pilot study demonstrated that the TEAM approach proposed for NOPES was
feasible and indicated where revisions were needed in the questionnaires, sample collection protocols, and
analytical laboratory procedures (Lewis et a(., 1988)
2. NOPES Phase I, August 21 -September 18, 1986
Sixty-five people in Jacksonville, Florida, participated in the summer season of NOPES data collection.
Monitoring samples were analyzed for 30 selected pesticides and pesticide degradation products. RTI was
responsible for sample design and selection, questionnaire administration, and statistical analysis of the data.
SwRI collected and analyzed the monitoring samples. Overall project management and quality assurance
was provided by EMSI in this phase and Phase II (Lev- On et al., 1987).
3. NOPES Phase II, March 20 - April 13, 1987 (Jacksonville) and
May 29 - June 17, 1987 (Springfield/Chicopee)
Seventy-two people in Jacksonville and forty-nine people in Springfield and Chicopee, Massachusetts,
participated in the spring season of data collection. Nineteen of the Jacksonville respondents had also
participated in Phase I. Three compounds (4,4'-DDT, 4,4'-DDE, and 4,4'-ODD) were added to the set of
thirty studied in Phase I (Immerman et al., 1988).
4. NOPES Phase III, January 30 - February 17, 1988 (Jacksonville) and
March 11 - March 28, 1988 (Springfield/Chicopee)
Seventy-one people in Jacksonville and fifty-two people in Springfield/Chicopee participated in the winter
data collection. Sixteen of the Jacksonville respondents and fifteen of the Springfield/Chicopee respondents
had participated in the earlier phases (Immerman et al., 1988a).
5. Special Study - High-Volume Surface Sampling, February 1 - February 6, 1988
Nine of the Phase III respondents in Jacksonville also participated in this study, which was conducted to
(1) test the ability of a high-volume surface sampler (HVSS) recently developed by Envirometrics to work
effectively under field conditons, and (2) permit preliminary assessment of the levels of pesticides present in
residential floor dust. The dust samples collected by the HVSS were analyzed by SwRI by the same
protocols used for the main study samples (Budd et al., 1988).
-------
-------
Section 2
Conclusions
NOPES achieved both its methodological and
analytical objectives, and has provided a wealth of
new information about the magnitude of and variation
in nonoccupational pesticide exposure. The major
findings and conclusions of NOPES are described
below in relation to the study objectives specified in
the previous section.
Objective: Apply the TEAM approach to pesticides.
Conclusions: The Total Human Exposure Assessment
Methodology (TEAM) applies survey sampling
techniques, indoor and outdoor
microenvironmental monitoring, personal exposure
monitoring, and human activity pattern data to
assess total human exposures. NOPES applied
the TEAM concept to estimation of
nonoccupational exposures to pesticides. NOPES
investigated the air, water, food, and dermal
routes of exposure for probabilistically selected
study participants.
NOPES applied all the TEAM monitoring
procedures for the air route of exposure. Because
routine sampling of public water supplies prior to
NOPES did not identify any of the target
•compounds, a minimal water sampling effort was
implemented. Rather than incur the expense of
directly collecting and analyzing food samples,
dietary intake data were collected to indirectly
estimate food exposures. Special gloves were
developed and pilot tested in the NOPES for
monitoring dermal exposures during pesticide
application events.
NOPES demonstrated that the TEAM approach
could be successfully applied to estimate
nonoccupational pesticide exposures via
inhalation. The air sampling instrumentation and
analytical procedures proved capable of reliably
characterizing personal, indoor, and outdoor air
concentrations o,f the majority of the study
analytes. Because the study was based on a
probability-based sampling design, the NOPES
data can be used to make valid statistical
inferences about the distribution of exposures
experienced by the populations of the two study
areas. In addition, NOPES was the first study to
provide information on air concentration
relationships for many of the study analytes.
Objective: Estimate population pesticide exposure
levels.
Conclusions: NOPES yielded quantitative estimates of
air exposure concentrations, and qualitative
assessments of water, dietary, and acute dermal
exposure levels. All of the compounds studied in
NOPES were detected at least once in the
NOPES air samples. Some compounds were
detected in the majority of households studied
Tables 2 and 3 summarize the estimated
prevalence and mean concentration of the
NOPES target compounds in personal air in
Jacksonville and Springfield/Chicopee
respectively. The reported mean air
concentrations may underestimate the true mean
concentrations because of (a) lack of adjustment
for incomplete recovery from the sampling matrix,
and (b) the inclusion in the computations of zeros'
for samples in which analytes were not detected.
Substituting zeros for nondetections primarily
affects the analytes with a relatively high limit of
detection, such as dichlorvos, because the
measured amount of each detected analyte was
recorded and used in the computations.
Nearly all the pesticides studied in NOPES have
been used in residential settings. Given the
sensitivity of the air monitoring techniques (on the
order of nanograms per cubic meter of air), the
detection of the analytes in many NOPES air
samples is, therefore, not surprising.
Objective: Assess the relative importance of the
exposure pathways.
Conclusions: The NOPES data only support qualitative
evaluation of the relative importance of the
exposure pathways studied. For 14 of the 25
analytes for which dietary exposure estimates
could be calculated, food appears to be the major
contributor to total exposure, whereas air appears
to be the dominant contributor for six of the other
eleven compounds.
-------
On the basis of the limited number of water
samples analyzed in MOPES, exposure to the
study compounds from water ingestion appears to
be negligible in the two study areas.
NOPES evaluation of the dermal contribution to
total pesticide exposure was based on a small
number of pesticide application events. The data
tentatively suggest that the dermal pathway may
be a significant contributor to total exposure for
some pesticides.
A preliminary examination of pesticide
concentrations in surface dust indicated that dust
may be a significant contributor to total exposure
for some pesticides, especially for infants and
toddlers (Budd et al., 1988).
Objective: Characterize the components of
variability in exposure levels.
Conclusions: Estimates of spatial and temporal
variation in air exposures were developed from the
NOPES data. For the majority of study analytes,
indoor air concentrations were substantially higher
than outdoor concentrations, often by more than
an order of magnitude. Personal air
concentrations were usually very similar to indoor
concentrations, reflecting the high proportion of
time typically spent indoors at home by
respondents.
Personal and indoor air concentrations of many
compounds were 2 to 30 times higher in
Jacksonville than in Springfield/Chicopee. In
winter, outdoor air concentrations of most
detected analytes were higher in Jacksonville than
in Springfield/Chicopee, whereas in spring, no
consistent pattern of differences between the two
study areas in outdoor concentrations was
observed.
Patterns of seasonal variation in indoor, personal,
and outdoor air concentrations were observed for
many study compounds. The patterns were
compound specific and complex, and may reflect
interactions among pesticide use, household
ventilation, temperature, and other factors.
Air concentrations of some analytes varied
substantially over a period of several days,
perhaps in response to the same factors that
contributed to seasonal variation. This short-term
variation was generally greater than the estimated
measurement error variation and less than, but .
more comparable . to, the observed seasonal
variation.
Objective: Examine relationships between exposure
levels and questionnaire data.
Conclusions: A simple, potential-air-exposure
categorization was developed from screening
questions and used to stratify the sample. The
three categories were effective as a general
classification device. They consistently differed in
measures that summarized air exposure across all
analytes. However, the categorization had only
limited effectiveness as a predictive tool for air
concentrations of specific analytes.
Exploratory analyses indicated that more
predictive questionnaire-based models and
categorizations may be possible for particular
analytes. Termiticide concentrations were related
to reported termiticide treatment history, type of
housing unit, and age of housing unit. Age of
housing unit was also related to concentrations of
older pesticides that are now banned or much
less frequently used. Weaker relationships were
observed between mean concentrations of some
commonly detected analytes and presence in
household inventories or reported indoor
insecticide use.
-------
Table 2. Estimated Percent of Population with Detectable Levels of target
Compounds in Personal Air, Jacksonville, FL - All Three Seasons
category and Compound Range of %
Detectable
Commonly Found Compounds
Chlorpyrifos
Propoxur
Diazinon
ortho-Phenyiphenol
Chlordane
Often Found
Heptachlor
gamma-BHC (Ijndane)
Dieldrin
Aldrin
Dichlorvos
alpha-BHC
Bendiocarb
Malathion
Hexachlorobenzene
Occasionally Found
Chlorothalonil
Heptachlor epoxide
2,4-D (butoxyethyl or methyl ester)
4,4'-DDE
4,4'-DDT
Methoxychlor
Dacthal
Rarely or Never Found
cis-Permethrin
trans-Permethrin
Folpet
Carbaryl
Resmethrin
Atrazine
Captan
Ronnel
- Oxychlordane
Dicofol
4,4'-DDD
83 -
88 -
,79 -
71
50 -
41
32 -
22 -
20 -
11
19 -
14 -
11 -
6 -
<1
2 -
0 -
5 -
6 -
1
0 -
1 -
1
1
0 -
0 -
0 -
0 -
0 -
0
0
0
97
94
87
90
93
90
70
70
37
35
27
26
21
45
19
15
15
12
9
12
8
3
3
2
2
2
2
2
2
Range of Mean
Concentrations (ng/m3)
118
141
89
40
191
64 -
7
5
7
21
0.7
3
9
0.4
<0 1
0.1
NDa • -
0.5
0.4
0.1
ND
0.1
0.1
0.4
ND
ND
ND
ND
ND
ND
ND
ND
280
316
322
80
212
134
22
10
39
148
0.9
51
17
0.9
3
0.6
3
0.8
0.5
0.6
0.6
1-
0.5
0.8
28
0.4
0.3
0.1
0.1
aND = Not detected.
7
-------
Table 3. Estimated Percent of Population with Detectable Levels of Target
Compounds in Personal Air, Springfield/Chicopee, MA - Both
Seasons
Category and Compound
Commonly Found Compounds
ortho-Phenylphenol
Chlordane
Heptachlor
Often Found
Chlorpyrifos
Propoxur
4,4'-DDE
Dacthal
4,4'-DDT
Dieldrin
Diazinon
gamma-BHC (lindane)
Occasionally Found
Chlorothalonil
Dicofol
Aldrin
Rarely or Never Found
Malathion
Dichlorvos
Bendiocarb
Folpet
Ronnel
Captan
Carbaryl
alpha-BHC
Hexachlorobenzene
. Heptachlor epoxide
Oxychlordane
2,4-D (butoxyethyl ester)
Methoxychlor
cis-Permethrin
trans-Permethrin
4,4'-DDD
Atrazine
Resmethrin
Range of %
Detectable
82
50
50
30
32
19
5
12
12
10
8
2
0
0
0
1
1
<1
<1
0
0
0
0
-
-
-
-
-
-
-
-
.
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0
0
0
0
0
0
0
0
0
86
87
66
40
38
23
26
19
18
17
10
12
12
15
4
2
2
2
2
2
2
2
1
Range of Mean
Concentrations (ng/m3)
27
36
5
6
11
0.5
0.3
0.7
0.7
1
0.7
0.1
NDa
ND
ND
2
0.2
<0.1
<0.1
ND
ND
ND
ND
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
.
-
-
-
-
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
43
253
35
7
16
5
3
0.9
0.8
10
5
0.8
7
0.2
0.5
4
0.3
0.7
0.1
0.1
0.1
<0.1
<0.1
Not detected.
-------
Section 3
Future Research Recommendations
Evaluation of NOPES results, in addition to providing
important insights about the nature and magnitude of
nonoccupational pesticide exposure, suggests a
number of possible avenues for further research.
Specific recommendations are:
1. Develop guidance for conducting exposure
monitoring studies and associated methodologies
for assessing human non-dietary exposure to
pesticides in residential settings. These follow-up
studies will be designed to permit a more
comprehensive analysis of the health risks
associated with exposure to pesticides from
different routes.
2. Conduct prospective studies to estimate pesticide
concentrations in household dust in order to
explore the relationship between pesticide use
and exposure, and the relative importance of the
dust pathway to total human exposure, especially
for infants and toddlers.
3. Refine the dermal exposure sampling and
analytical methods required for quantifying dermal
exposures and the estimation of acute and
chronic pesticide exposures. These studies will
attempt to estimate transfer coefficients between
surface applications and the dermal and inhalation
routes of exposure.
4. Improve the PUF sampling technique to reduce
variability in matrix spike recoveries, evaluate
analytical methodology for new compounds of
interest, and prepare quality assurance standards
on PUF media.
5. Conduct similar NOPES studies following revision
of the population survey instruments. These
revisions would incorporate improvements to the
original survey design, develop more appropriate
stratification variables, and permit the
development of a survey data base with a larger
regional or national application. The survey
instruments -would incorporate more detailed
activity pattern information and pesticide use
applications. The data would be combined with
limited monitoring data and used to validate a
proposed human exposure model specifically
designed to estimate exposures to several of the
NOPES pesticides.
-------
-------
Section 4
Study Design
The NOPES project was designed to test whether the
TEAM approach could be adapted to develop
estimates of exposures to selected household
pesticides and pesticide degradation products from
air, drinking water, food, and dermal contact in
stratified random samples of the populations of two
urban areas in the United States. The design also
permitted examination of several components of the
variation in pesticide air exposures, including regional
differences, seasonal changes, short-term temporal
variations, and interpersonal differences in patterns of
use.
Selection of Study Analytes
Attempting to monitor the levels of all registered
pesticides in the air and water of a household would
be methodologically difficult and prohibitively
expensive. Therefore, a manageable subset of target
pesticides had to be defined for this study. EPA's
Office of Pesticide Programs recommended a
prioritized list of 24 pesticides and pesticide
degradation products for the study. These compounds
were selected because of current regulatory interests
and the potential for occurrence of the compounds in
household environments. Eight additional compounds
were also suggested on the basis of previous EPA
studies. Of the original 24 pesticides, four —
glyphosphate (RoundupR), acephate (OrtheneR),
paradichlorobenzene, and pentachlorophenol - were
subsequently removed from further consideration
because they were difficult to measure by the
protocols appropriate for the remaining pesticides.
Four other compounds -- hexachlorobenzene, 4,4'-
DDT, 4,4'-DDE, and 4,4'-DDD - were subsequently
added at the request of the World Health Organization
(WHO). Table 4 presents the final list of 32 target
compounds for NOPES and describes their primary
residential uses in the two study areas.
Target Population Definition
One of the first decisions to be made in the design
phase of NOPES was where to conduct the study.
Taking into account the NOPES objectives, the
following set of preferred characteristics for the two
study areas was developed.
• Each study area was to include both urban and
suburban areas and have a variety of housing types
and ages. These characteristics would help to
ensure that a wide variety of pesticide use patterns
was included in the study. Sites with large rural
areas were excluded to avoid the potential for
agricultural or forestry pesticide "contamination" of
the NOPES measurements.
• The population in each study area had to be large
enough to allow screening at a relatively low rate
and still yield a sufficient number of high-potential-
exposure households for monitoring. Using areas
with relatively large populations would keep the
perceived burden on the community low and would
also increase the likelihood that a range of
pesticide use patterns would be encountered.
• Each study area was to be similar in some respects
to a relatively large number of other urban/suburban
locations to permit limited, non-statistical
extrapolation of the NOPES findings.
• One study area had to be in a region of relatively
high pesticide use, and the other area was to be in
a region of lower, but not negligible, use. This
would permit regional comparisons of usage levels
versus analyte concentration levels. Data from
EPA's National Urban Pesticide Applicator Survey
were used to characterize relative regional
pesticide use.
Application of the set of preferred characteristics
resulted in the definition of northern Florida and New
England as the primary candidate regions for the high-
use and low-use study areas, respectively. Following
discussions with regional, state, and local officials,
Jacksonville, Florida, and Springfield, Massachusetts,
were chosen to be the NOPES study areas. The
Springfield area was broadened to include the
neighboring town of Chicopee to increase the variety
of housing types and. ensure a .sufficient population
size for screening.
Following the decision to conduct the NOPES
sampling in the Jacksonville and Springfield/Chicopee
areas, the study areas for the survey were defined
11
-------
Table 4. NOPES Target Compounds
Target
Compound
Common
Formulations3
Primary Residential
Uses in the Two Study Areas
DISINFECTANT:
ortho-Phenylphenol
FUNGICIDES:
Caplan
A, RL Active ingredient in many LysolR brand disinfectant products.
D, F, WP Widely used by consumers and pest control operators (PCOs) for
control of diseases on trees, shrubs, fruits, and vegetables. Usually
applied as wettable powder to plant surfaces.
Chlorothalonil
Fdpet
Hexachlorobenzene
HERBICIDES:
Alrazine
2,4-D (methyl and butoxyethyl esters)
Dacthal
INSECTICIDES:
Aldnn
alpha-BHC
Bendiocarb
Carbaryl
F, WP Widely used by PCOs to control lawn, tree, and greenhouse plant
diseases. Primarily applied as a flowable spray on lawns.
D, F, WP Used by consumers, rarely by PCOs, for control of leaf diseases of
vegetables, flowers, fruits, and roses. Applied as a dust to leaf
surfaces, not used in large volumes.
WP, D Currently used primarily as a seed protectant, primarily for wheat.
Extremely rare in residential settings. Also found as a contaminant in
pesticidal and non-pesticidal products.
F, G, WP Jacksonville: Commonly applied to suburban settings in granular
form combined with granular lawn fertilizer (weed and feed products).
Occasionally applied to suburban lawns in liquid formulation.
• SpringfieldiChicopee: Pre- and post-emergent selective herbicide,
primarily used on corn. Rare in residential settings.
G, EC Probably the most commonly used lawn herbicide for control of
dandelions and broadleaf weeds. Primary consumer applications are
in granular lawn fertilizers. PCOs commonly apply as spray,
sometimes in combination with other active ingredients. The methyl
ester was tested for in Phase I of NOPES; later phases tested for the
butoxyethyi ester.
G. WP Jacksonville: Not commonly used in suburban settings
SpringfieldiChicopee: Widely used as a pre-emergent on lawns to
prevent germination of crabgrass, annual weeds, and some broadleaf
weeds. Primarily applied in granular form.
D, G, EC, WP Formerly used as a termiticide, applied as. a soil treatment. Not as
commonly used as chlordane; probably comprised less than 10% of
termiticide use. Now withdrawn from use in the U.S.
Banned for use in U.S. Still entering the environment as a conversion
product of gamma-BHC.
D, WP Very widely used by PCOs for indoor control of ants, cockroaches,
pantry and clothing pests, fleas, and termites (wood surface
applications only). Probably most commonly used in multiunit
dwellings subject to cockroach infestations. Applied as a dust or
wettable powder.
B, D, F, Probably the most commonly used insecticide for broad-spectrum
G, WP chewing insect control on fruit, vegetables, flowers, trees, shrubs,
flea collars and lawns in residential settings. Applied to leaf surfaces primarily in
wettable powder and dust forms. Widely used by consumers and
PCOs. Also used for household pests and in flea dusts and flea
collars.
(continued)
12
-------
Table 4. (Continued
Target
Compound
Common
Formulations3
Primary Residential
Uses in the Two Study Areas
JSECTICIDES (continued):
Chlordane
Chlorpyrifos
EC, oil
A, EC, G
4,4'-DDD
4,4'-DDE
4,4'-DDT
Diazinon
Dichlorvos
(DDVP)
Dicofol
Dieldrin
gamma-BHC
Heptachlor
Heptachlor epoxide
Malathion
Methoxychlor
Oxychlordane
cis-Permethrin
trans-Permethrin
D, WP
A, D, EC, G
impregnated resin
strips, EC, A
WP, EC, D, A
EC
EC, WP
EC, WP, G, D
WP, EC, D, B, G, A
EC, WP, G, D, A
A, EC
The most widely used termiticide prior to being withdrawn in April
1988 (September 1985 in Massachusetts). Comprised approximately
80% of the termiticide market. Applied primarily by PCOs as liquid
poured or injected into soil around building foundations.
Used both indoors and outdoors. Used outdoors by consumers and
PCOs for control of turf insects, ticks, chiggers, and ants. Applied
primarily in granular form outdoors. Used in aerosol form by
consumers and EC formulations, by PCOs for household insect
control. Widely used as a termiticide since ban on chlordane. Also
used in flea collars, shampoos, and sprays.
Environmental conversion product of DDT.
Environmental conversion product of DDT.
Very widely used from 1940s until early 1970s for control of
household, garden, ornamental, and public health insect pests. Also
used as a tracking powder for rodents and for control of public health
pests living on rodents. Banned from use in U.S.
Widely used outdoor soil insecticide for control of turf and garden soil
insects. Applied by consumers and PCOs in granular form. Indoors
applied in aerosol form by consumers for control of household insects
(ants, cockroaches).
Primarily used in "no-pest strips" by consumers to kill flying
household insects. Available to consumers only in concentrations of
1 % or less. .
Most common miticide in residential.settings for control of mites on
shrubs, fruits, vegetables, flowers, and houseplants. Applied by •
consumers and PCOs.
Formerly used by PCOs for subsoil control of termites' and on tree
bark to prevent borer infestations. Comprised less than 10% of
termiticides used. Withdrawn from use in U.S.
Primarily used to kill eggs of boring insects on tree bark. Applied as
liquid spray by consumers and PCOs. Used for head lice, but only by
prescription from a physician.
Until 1988, used alone or in combination with chlordane as a
subterranean termiticide. Comprised less than 5% of the termiticide
market. Now withdrawn from use in U.S.
Metabolite and/or environmental conversion product of heptachlor:
Widely used for control of insects on plant surfaces, especially on
trees, shrubs, fruit, vegetables, flowers, and houseplants. Primarily
used by consumers, less used by PCOs. Used in mosquito control
programs (discontinued in Florida in 1986).
Used for control of leaf-eating insects on trees, shrubs, fruit trees,
flowers, and vegetables. Primarily applied as a liquid spray. Formerly
used indoors to control pantry insect pests. Widely used for outdoor
control of mosquitos and flies.
Metabolite and/or environmental conversion product of chlordane.
Aerosols widely used by PCOs for control of household pests,
commonly in multiunit dwellings. Also used by consumers in aerosol
form for household insect control. Recently came into use by PCOs
as a termiticide, applied by liquid injection into soil.
(continued)
13
-------
Table 4. Continued
Target
Compound
Common
Formulations3
Primary Residential
Uses in the Two Study Areas
INSECTICIDES (continued):
Propoxur
Ronnel
Rosmothrin
EC, B, WP, A, Widely used for indoor pest control particularly cockroaches and
fogger, 'roach tape flies. Used by consumers and PCOs. Less commonly used in
granular applications for turf insect control.
A 'Use discontinued in U.S. Formerly used for indoor pest control,
especially fleas.
EC, D, WP, A Commonly used by PCOs for control of indoor household pests,
especially cockroaches, ants, and spiders. Less commonly used for
outdoor insect control on trees and shrubs. Applied by consumers in
liquid formulations to plant surfaces. Used in mosquito control
programs.
"Formulation codes: A = Aerosol, B = Bait, D = Dust, EC = Emulsifiable concentrate, F = Flowable. G = Granular. RL = Ready-to-use liquid,
WP » Wettable powder
more precisely in terms of standard geographic area
units used by the U.S. Bureau of ..the Census for the
1980 Decennial Census. Study area definition .was
governed by logistic considerations and the desire to
examine only urban and suburban areas. In
Jacksonville, the study area was restricted to the 10
centrally located Census County Divisions of Duval
County. (Governmentally, Jacksonville City and Duval
County are the same entities). The 1980 boundaries
of Springfield and Chicopee defined the northern
study area. The Jacksonville and Springfield/Chicopee
study areas are shown as shaded areas in Figures 1
and 2, respectively.
In each study area, the target population (i.e., the
population about which statistical inferences were to
be made) for NOPES consisted of individuals at least
16 years of age who satisfied the following criteria:
(1) primary place of residence was in the study area
when the household screening was conducted,
(2) not institutionalized or living in group quarters or
on a military reservation,
(3) not employed in a position in which the primary
activity involved the use or handling of pesticides,
nor residing in a household with one or more
members employed in such a position, and
(4) present in the study area at the time of personal
exposure and indoor/outdoor monitoring.
The age restriction was placed on the target
population because of the physical requirements and
level of responsibility imposed by the personal
exposure monitoring equipment. Individuals
occupationally exposed or residing with someone who
was occupationally exposed were excluded because
of potential problems in discriminating between
occupational and nonoccupational exposures.
The NOPES sampling weights were used to estimate
the size of the target population in each study area.
The target population in Jacksonville was estimated to
be approximately 290,000 people, residing in 150.000
housing units. The Springfield/Chicopee target
population was estimated to be approximately 135,000
people, residing in 73,000 housing units.
Sampling Design
Within the study areas of NOPES, participants were
selected at random using standard area household
survey sampling techniques. The NOPES sampling
design can be generally described as a three-stage
design. Probability sampling was used at all stages of
selection to ensure that the sample was statistically
representative and to allow valid statistical inferences
to be made from the data.
In the first stage of the sampling design, a stratified
sample of relatively small Census-defined geographic
areas (blocks or groups of blocks) was randomly
selected in each study area before beginning the first
season's data collection. By using 1980 Decennial
Census information, the first-stage sampling frame for
each study area was stratified by socioeconomic
status and by proportion of single-family housing units.
Socioeconomic status was selected as a stratification
variable because nonoccupational exposure to
pesticides was believed to be related to factors such
as type and quality of residence, composition of diet,
employment of professional pest control .services,
presence or absence of adequate air
conditioning/heating or other ventilation systems, and
other characteristics that may be correlated with
socioeconomic status. Use of the proportion of single
family housing units as a second dimension of the
stratification helped ensure that a variety of housing
types (e.g., single-family homes, apartment buildings,
mobile homes) were included in the sample.
14
-------
U.S. Census County Divisions
01 - Arlington
02 - Baldwin
03 - Cedar Hills
04 - Dewey Park - Venetia
05 - Dinsmore
06 - Eastport
07 - Jacksonville
08 - Jacksonville Beach
09 - Lake Forest-Riverview
10 - Lake Shore
11 - Mandarin-Loretto
12 - Marietta
13 - Pottsburg Creek
14 - San Jose
15 - Southside Estates
16 - Wesconnett
NASSAU CO
Scale of Miles
012345
Figure 1. NOPES study area in Jacksonville, Florida.
Within each sampled first-stage unit, all housing units
were identified and listed by field enumerators. At the
beginning of each season's data collection, a second-
stage random sample of housing units was selected,
and the sample households were screened to
ascertain characteristics of their dwellings and their
residents. The screening data were used to stratify
the sample households into three categories based on
the potential for exposure to pesticides from indoor
air. Characteristics used to define the categories
included:
• use of pesticides on indoor plants,
• use of insecticides (e.g., flea and tick powders,
dips, shampoos, collars) on household pets,
• treatment of the housing unit with termiticides, and
ST, JOHNS
• use of insecticides to control household insect
pests.
The stratification permitted respondents from high-
potential-exposure households to be included in' th'e
third-stage sample in higher proportion than they
occurred in the target population. The sample-
composition goal for each season's third-stage sample
of people was to have 50% "high-exposure"
respondents, 30% "medium-exposure" respondents,
and 20% "low-exposure" respondents.
Attempts were made to contact all persons selected in
the third-stage sample and ask them to participate in
the monitoring and interview portion, of the study.
Third-stage sample members were randomly selected,
and no more than one person was selected from any
household.
15
-------
Figure 2. NOPES study area in Springfield/Chicopee,
Massachusetts.
Conducting NOPES in phases at different times of
year permitted seasonal variation in air exposure
levels to be estimated. The NOPES study objectives
specified that the "within-home" or "within-individual"
(i.e., between seasons within the same home or for
the same individual) component of the overall
seasonal variation was also to be assessed. The third-
stage sample for each season and study city therefore
consisted of two components: a "single-season"
subsample that would be asked to participate in only
one phase of NOPES, and a "multiseason"
subsample consisting of people recruited to participate
in all phases conducted in the study city. Data from
the multiseason respondents were used to estimate
the between season component of variation for
persons and homes.
Prior to the first season's data collection in a study
city, third-stage sample members were randomly
designated as either single-season" or multiseason
subjects. Multiseason members who participated were
recontacted in each subsequent phase and asked to
participate again. New single-season subsamples
were selepted in each phase from households
screened in that phase. Table 5 summarizes the
planned number of respondents in each subsample in
the third-stage sample design.
Households of third-stage sample members were also
randomly designated as standard, duplicate, replicate,
or water households. These classifications determined
some of the sampling procedures to be used in the
households, as follows:
• Standard: air samples collected from one indoor
fixed-site, one outdoor fixed-site, and one personal
air sampling system.
• Duplicate: same as standard, with one additional
indoor fixed-site and one additional outdoor fixed-
site system operating concurrently and in close
proximity to the other fixed-site systems. Some of
these households were subsequently redefined in
the field as triplicate households, and additional
fixed-site systems were set up indoors and
outdoors. Duplicate and triplicate samples were
collected to help assess measurement error.
• Replicate: same as standard, with an additional
complete set of air samples collected at least three
days after the initial set. One set was to be
collected on a weekday and the other on a
weekend. Data from replicate sampling allowed
short-term temporal variation in air exposure levels
to be examined.
*»
• Water: same as standard, with a tap water sample
collected at the end of the monitoring period.
In each season and study area, five duplicate, three
triplicate, 10 replicate, and six water samples were to
be collected.
More detailed descriptions of the sampling design and
selection procedures used in each phase are
presented in the NOPES Interim Reports.
Data Collection Procedures
Field interviewers visited all housing units selected in
the second-stage sample. An attempt was made to
administer a screening questionnaire in each occupied
housing unit. Any responsible adult household
member was eligible to respond.
The screening questionnaire collected a variety of
information on potential exposure sources, such as
whether the home had .been treated with termiticides
or other insecticides. This information was
subsequently used to classify households into the
high-, medium-, and low-potential-exposure categories
that were used for third-stage stratification. The
screening questionnaire also recorded the name, age,
sex, and occupation of everyone 16 years of age and
older in, the household. This information was used to
16
-------
Table 5. Planned Number of Respondents in the NOPES Third-Stage Sample Design
Multiseason
respondents
Single-season
respondents
Total
Phase 1
Summer
(Aug-Sep 86)
30
40
70
Phase II
Spring
(Mar-Jim 87)
Jacksonville, FL
19
51
70
Phase III
Winter
(Jan-Mar 88)
15
55
7fl
Total
Number of
Different
Respondents
30
146
1 -7C
Springfield/Chicopee, MA
176
Multiseason
respondents
Single-season
respondents
Total
22
28
50
15
35
50
22
63
85-
select and identify specific people for the third-stage
sample.
All people selected into the third-stage sample were
contacted and asked to participate in the study. A
cash incentive for participation was offered, following
the standard practice used in earlier TEAM studies.
Upon agreeing to participate, the sample member was
administered a study questionnaire by a field
interviewer. The study questionnaire collected
demographic data, verified and updated eligibility and
exposure category inputs obtained during screening,
profiled occupational pesticide use, and inventoried
the pesticides at the respondent's residence.
Respondents were prompted to try to recall all
pesticides stored indoors or outdoors at the
.residence. The study questionnaire also included a
dietary intake record, which recorded all food items
consumed by the respondent the previous day. The
dietary information was collected to permit
development of estimates of individual dietary
pesticide exposures.
While the interview was being conducted, a monitoring
technician set up equipment to collect air samples.
Each air sample was collected by using a small,
portable, constant-flow air pump to draw air through a
clean polyurethane foam (PUF) plug. The pump
operated at a flow rate of 3.8 standard liters per
minute (SLPM) for the 24-h sampling period. A
portable calibrator was used to check the flow rate of
each pump prior to deployment. Indoor and outdoor
air samplers were located about 1.5 m above the floor
or ground in an area of high family use and were
plugged via a charger/converter into standard 110-V
AC outlets. The personal samplers operated for much
of the monitoring period on an internal battery supply,
but could be plugged in when the respondent was-
sedentary. The personal air sampler was housed in a
case with a shoulder strap, and participants were
instructed to keep the sampler in close proximity
throughout the monitoring period.
The interviewer and monitoring technician returned to
the household approximately 24 h after sample
collection began. The technician collected and
processed the PUF plugs, completed all necessary
documentation, and removed the sampling equipment.
If the household was designated as a "water
household," the technician also , collected a 1 -L
sample from the primary source of drinking water.
Meanwhile, the interviewer completed a 24-h activity
log for the participant.
The activity log recorded any pesticide exposures or
activities participated in by the respondent that could
have affected pesticide levels in the sampled air. Both
direct and indirecf sources of exposure were
examined. The amount of time spent by the
respondent in several general locations was profiled,
as was the ventilation pattern in the respondent's
home during the monitoring period.
After the activity log was completed, each respondent
was asked if he or she intended to perform a pesticide
application within several days and if the NOPES
sampling team could come back to monitor the
application event. The study design specified that six
events were to be monitored in each city each
season. Fewer than six were monitored in some
seasons because not enough cooperative
respondents planned to perform any application
events during the data collection period! '
Respondents who agreed to participate in the dermal
sampling were asked to wear a pair of precleaned
cotton gloves during the preparation, application, and
cleanup. These gloves were to be worn under the
17
-------
participant's regular work gloves, if work gloves were
normally worn during similar applications. The study
gloves were then collected and returned to the
laboratory for analysis. The glove data allowed an
estimate to be made of the amount of dermal
exposure experienced during the application event.
The individual's air exposure was monitored with a
personal air sampler during the application.
Copies of the Screening -Questionnaire, Study
Questionnaire, and Activity Log are provided in
Appendix A.
Response Rates
Overall second-stage sample sizes were 1,501
housing units in Jacksonville and 2,472 housing units
in Springfield/Chicopee. Screening information was
obtained from 1,005 Jacksonville households and
1,774 Springfield households. Second-stage response
rates, computed as the number of respondents
divided by the number of eligible sample members,
were relatively low for in-person household screening,
ranging from 66% for the Jacksonville spring season
to 84% for the Springfield;Chicopee winter season
(Table 6). Second-stage nonresponse was due more
to inability to contact household members during the
time period allotted for screening (56% of
nonresponding eligible sample members) than to
refusals (32% of nonresponding eligible sample
members).
Third-stage response rates varied by study area,
season, and whether sample members were single-
season or multiseason subjects. Nonresponse in the
third stage was primarily due to refusals to participate
(73% of nonresponding eligible sample members).
The two most commonly cited reasons for refusing to
participate were the amount of time required and the
perceived burden associated- with keeping the
personal sampler nearby.
The overall response rates presented in Table 6 were
computed by multiplying the second-stage response
rate by the third-stage response rate for first-time
sample members (i.e., multiseason sample members
were not included in the overall response rate
calculations after their first season). The NOPES
overall response rates were comparable to the 44%
response rate experienced in the New Jersey
segment of the TEAM-VOC study (Wallace, 1987).
Although these response rates are low relative to
those experienced in traditional area-household
surveys, they are typical of the rates experienced in
personal monitoring studies. Low personal-monitoring
response rates are believed to be primarily due to the
respondent burden imposed by the monitoring
systems and procedures.
In any sample survey, low response rates are
undesirable because they introduce the potential for
bias in the estimates computed from the survey data.
The extent to which the nonresponse actually
produces bias depends on the degree to which
respondents and nonrespondents differ in the
parameters being estimated. Although the size of the
difference can never be precisely quantified (due to
the lack of data for nonrespondents), a rough idea of
its magnitude can often be postulated by taking into
account the subject of the survey questions and the
characteristics of the population being surveyed.
Bias can be expected to be low if the following
statements are true/
1) The survey does not deal with a sensitive subject,
such as income level, sexual preference or habits,
or political or religious opinions.
2) The distributions of respondents' characteristics,
such as age, sex, race, and soeioeconomic status,
are similar to those of the population. (The
population distributions must be obtained from an
independent source.)
3) The parameters of interest are not functions of the
same factors that cause nonresponse (e.g.,
unusual work schedules that' make contact with an
interviewer unlikely).
Examining the NOPES results in light of the above
considerations leads to the conclusion that bias
related to the response rate was probably relatively
low. Neither the screening questionnaire nor the
monitoring phase of data collection dealt with subjects
typically considered sensitive.
The sex and race distributions of respondents,
discussed in the next section, were slightly different
from those observed in the 1980 Census. However,
the distributions were not different enough to
dramatically impact the pesticide concentration
estimates, even if, for example, all the
nonrespondents in one group (e.g., males) had higher
personal air pesticide concentrations than
nonrespondents in the other group (e.g., females).
Response rates were relatively similar in the different
geographic areas sampled within each study area.
Detailed data on pesticide use habits for different
segments of the population are not available.
Therefore, any scenario in which respondents and
sample members who refused to participate or who
could not be contacted differed regarding their use of
pesticides would be based only on speculation. The
available data suggest that although the low response
rate may have caused some bias in the sample
estimates, the magnitude of the bias is relatively
small.
18
-------
Table 6. Response Rates
Jacksonville
Springfield/Chicopee
Summer
'86
Spring
'87
Winter
'88
Total
Second Stage
Sample size
' Eligible
Respondents
Response rate
Third Stage
First-time sample:
Selected
Eligible
Respondents
Response rate
Overall Response Rate3
Followup sample:
Selected ,
Eligible
Respondents
Response rate
Total:
Selected
Eligible
Respondents
401
363
267
74%
125
120
65
54%
40%
550
510
336
66%
79
73
53
73%
48%
29
29
19
66%
550
499
402
81%
95
90
55
61%
49%
19
19
16
84%
1501
1372
1005
73%
299
283
173
61%
45%
48
48
35
73%
Spring
'87
1422
1361
956
70%
92
89
49
55%
39%
125
120
65
108
102
72
114
109
71
347
331
208
92
89
49
Winter
'88
1050
978
818
84%
73
72
37
51%
43%
20
20
15
75%
93
92
52
Total
2472
2339
1774
76%
165
161
86
53%
40%
20
20
15
75%
1185
1181
1101
"Overall response rate = (second-stage response rate) • (third-stage response rate) for first time members of the sample.
Respondent Characteristics
Selected characteristics of third-stage respondents
and their homes are presented in Table 7. In both
study areas, female respondents outnumbered male
respondents. The differential'was greater than
expected from the general population distribution
(48% male and 52% female in the 1980 Census for
persons 18 years of age and older in the two study
areas) and reflects slightly higher response rates
among female sample members in both the second-
and third-stage samples. Seventy-two percent of the
Jacksonville respondents and 86% of the
Springfield/Chicopee respondents were non-Hispanic
whites. The sample race/ethnicity distribution was very
similar in the 1980 Census population distribution in
Jacksonville, whereas in Springfield/Chicopee, whites
were slightly overrepresented among respondents
relative to,the Census-distribution. Approximately 70%
of the participants were employed.
The two study areas displayed some differences in
housing unit characteristics. Attached dwellings were
more common in Springfield/Chicopee, and mobile
homes occurred more frequently in Jacksonville,
although in both areas unattached, single-family units
were the predominant housing type in the sample.
The Spnngfield-'Chicopee housing units were on
average, 11 years older than the Jacksonville units
The average age for the Springfield/Chicopee housing
units was 42 years old, while the Jacksonville housing
units averaged 31 years of age. The oldest
Springfield/Chicopee sample housing unit was built in
1770, and the oldest home in the Jacksonville sample
was built in 1895.
In both areas, approximately half the responding
households said that their homes had been treated
with termiticides. The accuracy of this information is
unclear, because in some cases opposite answers to
the termiticide use questions were obtained in the
screening and the study questionnaires. A substantial
number of respondents indicated that they did not
know if their home had been treated for termites.
The average number of pesticide products listed in
the study inventory was comparable for the two sites'
4.2 pesticides per household for Jacksonville and 53
pesticides per household for Springfield/Chicopee
The maximum number of pesticide products listed in a
home was 23 for Jacksonville and 18 for
Springfield/Chicopee, which are again comparable
figures, given that the Jacksonville sample contains
about twice as many homes as the
Springfield/Chicopee sample. Some homes in each
sample did not have any inventoried pesticide
products in the home at the time of the study.
Laboratory Operations
Analysis of the PUF plug, water, and glove samples
followed protocols developed by SwRI for the MOPES
target compounds (Hsu et al., 1988). An additional
19
-------
Table 7. Third-Stage Respondent and Household Characteristics
Number of Respondents
Characteristics
Jacksonville
Springfield/Chicopee
Sex
Male
Female
Race/Ethnicity
White, non-Hispanic
Nonwhite or Hispanic
Age
16-25
26-45
46-60
Over 60
Employed
Yes
No
Occupational Exposure
Yes
No
Type of Housing Unit
Unattached Single Family
Attached Single Family
Multiunit (apartment)
Mobile home
Age of Housing Unit
Less than 6 years old
6-15
16-25
26-35
More than 35
Any Termiticide Treatment of Housing Unit
Yes
No
Don't know
85(41%)
123 (59%)
150(72%)
58 (28%)
39 (19%)
91 (44%)
41 (20%)
37 (18%)
143(69%)
65(31%)
8 (6%)
135 (94%)
153 (74%)
6 (3%)
29 (14%)
20 (10%)
17 (8%)
17 (8%)
58 (28%)
52 (25%)
64 (31%)
104(50%)
65(31%)
39 (19%)
42 (42%)
59 (58%)
87 (86%)
14(14%)
15(15%)
49 (49%)
20 (20%)
17(17%)
74 (73%)
27 (27%)
5 (7%)
69(93%)
71 (70%)
12 (12%)
16(16%)
2 (2%)
2 (2%)
12 (12%)
16 (16%)
23 (23%)
48 (48%)
46 (46%)
30 (30%)
25 (25%)
analytical protocol was used to determine chlordane
and heptachlor concentrations (ASTM, 1989).
After collection, PUF plugs and glove samples were
kept on dry ice until extracted by Soxhlet extraction.
Water samples were kept at 4°C until extracted and
analyzed according to EPA Method 608 (Method 608,
1984). Extractions were almost always completed
within seven days after collection.
The extract for each sample was concentrated and
divided into two aliquots, one for gas
chromatography/electron capture detection (GC/ECD)
analysis and the other for gas chromatography/mass
spectrometry/multiple ion detection (GC/MS/MID)
analysis. For the chlorinated target compounds,
GC/ECD was used for quantitation, and GC/MS/MID
served as a confirmation analysis (Table 8). For each
GC/ECD extract, a primary analysis was performed on
a megabore column, and a secondary analysis was
performed on a column with a dissimilar liquid phase.
The nonchlorinated target compounds were quantified
by using GC/MS/MID. More detailed descriptions of
the analytical instruments and conditions are
presented in the NOPES Interim Reports.
Analytical quality control steps were included
throughout the analysis activities. Stringent calibration
criteria were nearly always met for 31 of the analytes
on the column used for quantitation. Less precise
quantitation of dicofol was permitted because of its
poor chromatographic behavior. More than 98% of the
analyses were performed within 30 days after
extraction. Duplicate matrix-spiked samples were
prepared and analyzed with every batch of samples
from the field (20 to 30 samples per batch). The
matrix-spike solution included diazinon, propoxur,
alpha-BHC, heptachlor, chlorpyrif os,
hexachlorobenzene, and dieldrin. These compounds
spanned the chromatographic range and were
representative of the different chemical classes of the
analytes. Each sample was analyzed for a spiked
surrogate compound, octachloronaphthalene (OCN),
to monitor the integrity of the entire analytical system.
20
-------
Table 8. Analytical Methods for NOPES Target Compounds
Analyte
Dichlorvos (DDVP)
alpha-BHC
Hexachlorobenzene
gamma-BHC (LindaneR)
Chlorothalonil (BravoR)
Heptachlor
Ronnel
Chlorpyrifos (DursbanR)
Aldrin
Dacthal
Heptachlor epoxide
Oxychlordane
Captan
Folpet
2,4-D butoxyethyl esterc
Dieldrin
Methoxychlor
Dicofol
cis-Permethrin
trans-Permethrin
Chlordane (technical)
4,4'-DDTd
4,4'-DDDd
4,4'-DD£d
ortho-Phenylphenol
Propoxur (BaygonR)
Bendiocarb (FicamR)
Atrazine
Diazinon
Carbaryl (SevinR)
Malathion
Resmethrin
r-\i idiyuv^dl
GC/ECD
X
X
X
X
X
X
X
X
X
X
X \
X
X
X
X
X
x'
X
X
X
X
X .
X
X
IVICUIUU"
GC/MS ,
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Air Sample
QL Goalb
(ng/m3)
360
7
6
g
7 .
13
13
11
9
g
7
1 1
55
36
180
15
18
180
73
73
150.
11
11
11
36
18
45
45
55
45
45
j91 '
a GC/ECD = gas chromatography/electron capture detection, GC/MS = cms
chromatography/mass spectrometry. Compounds analyzed by both GC/ECD and GC/MS
were normally quantified by using the GC/ECD result uwmo
" The quantitation limit (QL) goal, established by the Quality Assurance Project Plan (Lev-
c £"pt ', £7a) waf defined as aPProximately five times the expected detection limit.
In Phase I, the methyl ester, rather than the butoxyethyl ester, was measured The QL
goal for the methyl ester was 110 ng/rr)3.
d Not measured in Phase I and not analyzed by GC/MS in Phase II.
To assess overall accuracy, performance audit
samples provided by EPA and EMSI were analyzed.
Field and laboratory blank samples were analyzed to
check for contamination.
The analytical protocol was refined and altered when
necessary over the three phases. On the basis of the
Phase I experience, the quantitation limit for ortho-
Phenylphenol was raised, and the methyl ester of 2 4-
D was replaced as a NOPES target compound by the
more commonly used butoxyethyl ester. Peak
identification on the GC/ECD chromatographs was
initially performed manually but was semiautomated in
Phase II. Heptachlor was quantitated from the
secondary GC/ECD column in Phase III to avoid a
frequently interfering peak on the primary column
chromatogram. Chlordane was quantitated without
inclusion of the heptachlor peak.
Experience brought refinements to the interpretation
of the chromatograms and quantitative data, and
efforts were made to review earlier data in light of the
revised interpretations. The analytical data for all
Phase I and Phase II samples were reexamined to
resolve coelution problems in a manner consistent
with the Phase III procedures. For heptachlor and
Chlordane, Phase I and Phase II values were reviewed
using the Phase III quantitation protocol, and the
earlier concentration estimates were revised as
21
-------
needed to establish concordance between the
chlordane and heptachlor results.
A problem developed in Phase I when BoileezerR
boiling chips were substituted for the standard Teflon^
boiling granules used during sample extraction and
concentration. The problem was recognized and
rectified, but only after the loss of accurate
quantitative data for specific compounds in 50 Phase I
samples (24 primary air, seven replicate air, seven
duplicate or triplicate air, four glove air, four sample
glove, one water, two air blank, one glove blank, and
one water blank sample). The inaccurate data were
excluded from all analyses presented in the next
chapter.
A total of 1,281 air, water, and glove samples were
collected, of which 1,277 were analyzed (Table 9).
This represented 97% of the samples originally
specified in the study design.
Table 9. Number of Samples Collected and Analyzed
Number of Analyzed Samples-
Sample Type
Jacksonville
Springfield/Chicopee
Personal air
Indoor air
Outdoor air
Water
Glove
QA/QO
Total
248
232
229
17
15
92
833
128
120
118
12
g
57
444
^Includes field blanks, duplicates, and triplicates.
22
-------
Section 5
Results and Discussion
Analyses of the NOPES data presented in this section
take into account both the NOPES sampling design
and the properties and limitations of the chemical
measurement process. The sampling design
influenced the way in which parameter estimates and
their standard errors were computed, and it provided a
framework for making inferences from the data. Both
sampling variance and measurement .error were
incorporated in the parameter estimates.
Measurement error reflected the limits in the resolving
power and precision of gas chromatography and mass
spectroscopy, as well as the variations inherent in
sample collection and laboratory activities. Factors
that contributed to measurement error included the
following:
• Less than 100% recovery of analytes from the
sampling devices (e.g., PDF cartridges, gloves).
• Abbreviated sampling times for some air samples
due to equipment malfunction.
• Failure by some respondents to keep the personal
sampler in close proximity throughout the
monitoring period.
The potential influence of such factors needs to be
recognized when interpreting the NOPES results.
When a survey includes respondents sampled at
different rates from the target population, weighted
analyses are needed to produce statistically unbiased
estimates from the survey data. The NOPES third-
stage samples were based on substantially different
selection rates for the three potential-exposure strata.
The third-stage samples were designed to be 50%
from the high-exposure stratum, 30% from the
medium stratum, and 20% from the low stratum. In
Jacksonville only 16% of the second-stage screened
households were in the high-exposure category, and
in Springfield/Chicopee, only 4% were in the high
category. Overall, high-exposure stratum members
were four times more likely than low-stratum members
to be selected in Jacksonville, and 28 times more
likely in Springfield/Chicopee. Sampling weights that
reflect the sampling design and are adjusted for
nonresponse were therefore computed for all NOPES
respondents and used for all estimates of population
parameters. Unweighted analyses were performed
when either very small sample sizes (e.g., the water
samples) or nonrandom selection procedures (e.g.,
the dermal sampling) made statistical inferences to
the target population inappropriate.
For each NOPES sample, every analyte was
categorized as either detected or not detected. The
actual detection limits are discussed in the Data
Quality Section. The measured amount of each
detected analyte was recorded and used to compute
concentration statistics (e.g.vmeans, standard errors,
medians, and percentiles). Tine computations included
zeros for samples in which the analyte was not
detected. Because the analyte may have been
present at less than detectable levels in some
samples, the use of zeros in the computations results
in statistics that may underestimate the actual
concentrations, especially for analytes with relatively
high detection limits.
Results of matrix spike quality control analyses
indicate that the percent recovery of compounds
under NOPES analytical conditions was often less
than 100%. When the concentration statistics were
computed, data from the field samples were not
adjusted to compensate for the percent recovery. In
addition, analyte concentrations were not adjusted for
the infrequent low level of contamination found in the
field blanks.
Air Exposure
The estimated percents of the target population with
detectable analyte air levels are profiled by sample
type (i.e., indoor, outdoor, and personal) and season
in Tables 10 and 11 for Jacksonville and
Springfield/Chicopee, respectively. More detailed
statistics, including the associated standard 'errors, are
presented in Appendix B (Tables B-1 and B-2).
Seasonal differences in the percent detectable may
be partly due to seasonal differences in the limits of
detection, which were generally lower for most
compounds quantitated from GC/ECD analyses (see
Table 8) for the Jacksonville summer season, as
discussed in the Data Quality Section.
All target compounds except 4,4'-DDD were detected
at least once in Jacksonville air samples. 4,4'-DDD
was detected in Springfield/Chicopee, but eight other
23
-------
Table 10. Estimated Percent of Jacksonville Population with Detectable Levels in Air
Indoor Outdoor
Analyte
Dichtorvos
alpha-BHC
Hexachlorobenzene
gamma-BHC
Chtocothatonil
Heptachlor
Ronnel
Chtorpyrifos
AWrin
Dacthal
Heptachlor epoxide
Oxychtordane
Cap tan
Folpet
2,4-D ester0
Dioldrin
Methoxychtor
Dicofol
cis-Permethrin
trans-Permethrin
Chtordane
4.4'-DDT
4,4'-DDD
4,4'-DDE
ortho-Phenylphenol
Propoxur
Bondiocarb
Atrazine
Diazinon
Carbaryl
Malathion
Resmethrin
Summer
33
25
50
34
9
58
3
100
21
1
16
2
4
2
8
79
7
* 0
3
1
61
-
-
-
85
98
23
0
83
17
27
1
Spring
14
23
6
47
13
71
0
88
19
0
3
0
5
1
0
37
1
5
3
3
54
14
0
6
84
93
20
0
83
1
32
0
Winter
10
22
7
68
20
92
0
96
31
9
5
1
<1
3
10
62
3
0
2
2
94
9
0
3
79
95
20
0
83
0
17
0
Summer
0
3
10
14
4
21
1
95
7
0
30
0
0
7
2
39
0
0
0
0
23
-
-
-
10
27
0
0
39
2
3
0
Spring
0
0
0
12
9
22
0
32
0
0
1
0
0
2
0
1
0
0
0
0
12
0
0
0
1
7
0
0
9
0
0
0
Winter
3
2
0
24
9
47
0
30
5
0
0
0
0
0
3
19
2
0
0
0
73
0
0
0
2
21
0
0
11
0
4
0
Summer
35
26
45
34
7
41
2
97
37
5
15
0
0
2
6
70
12
0
1
3
53
-
-
- ,
90
92
14
2
79
2
15
2
Personal
Spring
11
19
6
32
<1
68
0
83
20
0
3
0
2
2
0
22
1
0
3
1
50
9
0
5
73
94
21
0
83
2
21
0
Winter
16
27
6
70
19
90
1
97
36
8
2
0
1
2
15
51
3
0
2
2
93
6 '
0
12
71
88
26
0
87
0
11
0
"Methyl ester in summer; butoxyethyl in spring and winter.
compounds were not. Five analytes - chlordane,
chlorpyrifos, heptachlor, ortho-Phenylphenol, and
propoxur - were relatively common in indoor and
persona! air samples in both areas.
Tables 12 and 13 present the estimated arithmetic
mean air concentrations for the Jacksonville and
Springfield/Chicopee populations by sample type and
season. Standard errors of the means, as well as
maximum and weighted median concentrations, are
given in Appendix B (Tables B-3 and B-4).
Rgures 3 through 7 display the cumulative weighted
frequency distributions for personal air concentrations
of the five most prevalent analytes. The ordinates (Y-
axes) in these figures are in log scales to
accommodate the skewed distributions of
concentrations observed for the analytes. The
abscissas (X-axes) are in normal probability scales.
Data that exactly fit a log normal probability
distribution would lie on a straight diagonal line when
plotted on this combination of scales.
Appendix C presents the 25th, 50th, 75th, 90th, 95th
and 99th weighted percentiles for all analytes.
Indoor, Outdoor, and Personal Comparisons
In both study cities, "more analytes were detected in
indoor and personal air than in outdoor air. In
Jacksonville, 30 analytes were detected in indoor air,
and 29 were detected in personal air, whereas only 20
were detected in outdoor air. Corresponding counts
for Springfield/Chicopee are 24 analytes detected in
indoor air, 23 in personal air, and 11 in outdoor air.
Among analytes detected in all three sample types,
the estimated percent of the population with
detectable indoor and personalair levels was often
substantially higher than the estimated percent with
detectable outdoor levels.
24
-------
Table 11. Estimated Percent of Springfield/Chicopee Population with Detectable Levels in
Air
Indoor Outdoor " Personal
Analyte
Dichlorvos
alpha-BHC
Hexachlorobenzene
gamma-BHC
Chlorothalonil
Heptachlor
Ronnel
Chlorpyrifos
Aldrin
Dacthal
Heptachlor epoxide
Oxychlordane
Captan
Folpet
2,4-D butoxyethyl ester
Dieldrin
Methoxychlor
Dicofol
cis-Permethrin
trans-Permethrin
Chlordane
4,4'-DDT
4,4'-DDD
4,4'-DDE
ortho-Phenylphenol
Propoxur
Bendiocarb
Atrazine
Diazinon
Carbaryl
Malathion
Resmethrin
Spring
2
2
0
10
<1
50
2
29
0
21
0
0
<1
2
2
12
0
o
0
0
50
<1
0
13
90
49
2
0
16
2
2
0
Winter
1
0
4
21
2
70
<1
30
12
5
0
0
1
0
0
34
0
0
0
0
83
8
1
20
72
38
1
0
10
0
0
0
Spring
0
0
0
. 0
12
8
0
52
0
17
0
0
0
2
0
0
0
0
0
0
8
0
0
0
7
4
0
0
12
0
5
0
Winter
0
0
0
0
1
2
0
<1
0
0
0
0
0
0
0
0
0
0
0
0
16
1
0
0
0
1
o
0
8
0
0
0
Spring
1
2
0
10
12
50
2
30
0
24
0 .
0
2
2
0
12
0
12
0
0
50
12
0
23
82
' 32
2
0
17
2
4
0
Winter
2
0
1 .
8
2
65
<1
40
12
5
0
0
0
<1
0
18
0
0
0
0
87
19
0
19
86
38
1
0
10
0
0
0
The concentration data for indoor, personal, and
outdoor air show a similar relationship. Mean outdoor
concentrations were almost always lower than mean
indoor and personal concentrations of detected
compounds. Figures 8 through 12 display the
differences for the five most prevalent compounds. In
all, there are 157 sets of indoor-outdoor-personal air
mean concentrations, counting each season and
study area separately. In 122 of these sets there was
at least one mean greater than zero. The outdoor
concentration was higher than the indoor
concentration in only five sets, and higher than the
personal concentration in only six sets. In all eleven
sets, the mean concentrations were near or below the
QL goal. In most cases in which analytes were
detected indoors and outdoors, indoor concentrations
were 5 to 100 times higher than outdoor
concentrations.
These findings are similar to those reported for VOCs
in the initial TEAM study (Wallace, 1987). They
reaffirm the conclusion of the VOC TEAM study and
other studies (Lewis and Lee, 1976; Lewis and
MacLeod, 1982) that indoor and outdoor air
environments differ considerably in terms of toxic
substance levels, with greater levels indoors for many
compounds.
Mean personal air and indoor air levels were similar
for most analytes. The strength of the association
between indoor and personal air levels can be
measured by correlation analysis. Because of the
highly skewed distribution of concentration values,
nonparametric Spearman rank-order correlations were
computed. Correlations were computed for each
analyte detected in at least one sample in both the
indoor and personal air environments.
The correlation analyses, summarized in Table 14,
were performed for each study area and season, and
indicate a strong association between indoor and
personal air levels for the majority of detected
25
-------
Table 12. Weighted Arithmetic Mean Concentrations in Jacksonville Aira (ng/m3)
Analyte
Dichtorvos
alpha-BHC
Hexachtorobenzene
gamma-BHC
Chtocolhalonil
Heptachtor
Ronrtel
Chtorpyrifos
AMrin
Dacthal
Heptachtor epoxtde
Oxychkxdane
Captan
Fotpet
2,4-D esterb
Dieldrin
Methoxychlor
Dicofot
cis-Permelhrin
trans-Permethrin
Chtordane
4,4'-DDT
4,4'-DDD
4,4'-DDE'
ortho-Phenylphenol
Propoxur
Bondiocarb
Atrazine
Diazinon
Carbaryl
Malathion
Resmethrin
Summer
134.5
1.2
1.3
20.2
5.3
163.4
0.2
366.6
31.3
0.2
0.5
5.2
1.9
0.5
1.8
14.7
0.2
0
0.5
0.4
324.0
-
.
-
96.0
528.5
85.7
0
420.7
68.1
20.8
0.1
Indoor
Spring
86.2
1.2
0.4
13.4
2.2
154.9
0
205.4
6.8
0
0.8
0
2.2
0.7
0
8.3
0.3
11.0
1.9
1.1
245.5
1.0
0
0.6
70.4
222.3
5.5
0
109.2
0.4
14.9
0
Winter
24.5
1.1
0.3
6.0 :
6.7
72.2
0
120.3
6.9
0.3
0.8
6.5
0.1
0.6
2.5
7.2
0.2
0
1.3
0.8
220.3
0.5
0
0.2
59.0
162.5
3.4
0
85.7
0
20.4
0
Summer
0
0.0
0.2
1.3
0.2
30.2
0.1
16.7
0.2
0
0.7
0
0
0.3
0.0
0.7
0
0
0
0
38.4
-
-
-
1.2
10.2
0
0
12.6
0.2
0.3
0
Outdoor
Spring
0
0
0
0.5
0.3
10.7
0
3.5
0
0
0.1
0
0
0.4
0
0.0
0
0
0
0
9.5
0
0
0
0.0
0.8
0
0
1.1
0
0
0
Winter
3.2
0.0
0
0.6
0.6
2.8
0
2.5
0.1
0
0
0
0
0
0.8
0.8
0.1
0
0
0
27.3
0
0
0
0.1
2.5
0
0
13.8
0
0.2
0
Summer
147.6
0.9
0.9 -
22.1
0.5
129.1
0.1
280.4
19.9
0.6
0.6
0
0
0.4
0.7
10.1
0.3
0
0.1
0.1
212.0
-
-
79.7
315.6
51.4
0.3
321.6
28.3
9.2
0.4
Personal
Spring
40.2
0.8
0.4
7.0
0.0
133.7
0
182.8
38.5
0
0.5
0
0.1
0.4
0
5.4
0.1
0
1.3
0.3
190.7
0.5
0
0.5
55.6
141.1
4.4
0
112.7
0.8
10.1
0
Winter
21.4
0.7
0.4
8.5
2.5
64.2
0.0
118.2
6.9
0.2
0.1
0
0.1
0.8
3.5
4.8
0.6
0
0.8
0.5
194.8
0.4
0
0.8
39.7
142.8
3.5
0
89.0
0
16.8
0
"A weighted mean of "0" means no detectable levels were observed. A weighted mean of "0.0" means that the weighted mean was
less than 0.05.
I ester In summer, butoxyethyl ester in spring and winter. ,
approaches yields several insights into seasonal
variation.
Relatively minor variations occurred across seasons in
the number of analytes detected (Table 15). In
Jacksonville, the most analytes were found in
summer, followed by winter and then spring. In
Springfield/Chicopee, more analytes were detected in
spring than winter. The level of seasonal variation in
number of detected analytes was relatively small
compared to variation between sample types.
More information on seasonal variation is provided by
looking at patterns for each analyte. Inspection of
Tables 10 though 13 reveals that the analytes varied
considerably in their seasonal patterns. To summarize
these patterns across analytes, within each study area
each analyte was classified by the season that it
occurred with the greatest frequency or at the highest
analytes. As expected, the correlations between
personal and outdoor air are much weaker.
The general correspondence between indoor and
personal air concentrations is not surprising given the
amount of time spent indoors at home by
respondents. NOPES respondents spent an average
of 17 h per day indoors at home, a figure similar to
other survey-based estimates (Letz et al., 1984).
Seasonal Variation
Seasonal variation can be examined in terms of the
number of target compounds detected in each
season, the detection frequencies of particular
analytes across seasons, and the average
concentrations of particular analytes across seasons.
Analyzing the NOPES data by each of these
26
-------
Table 13. Weighted Arithmetic Mean Concentrations in Springfield/Chicopee Air*" (ng/m3)
Indoor Outdoor Personal
Analyte
Spring
Winter
Spring
Winter
Spring
Winter
Dichlorvos
alpha-BHC
Hexachlorobenzene
gamma-BHC
Chlorothalonil
Heptachlor
Ronnel
Chlorpyrifos
Aldrin
Dacthal
Heptachlor epoxide
Oxychlordane
Captan
Folpet
2,4-D butoxyethyl ester
Dieldrin
Methoxychlor
Dicofol
cis-Permethrin
trans-Permethrin
Chlordane
4,4'-DDT
4,4' -ODD
4,4'-DDE
ortho-Phenylphenol
Propoxur
Bendiocarb
Atrazine
Diazinon
Carbaryl
Malathion
Resmethrin
4.3
0.2
0 •
0.5
0.1
31.3
0.2
• 9.8
0
1.6
0
0
0.1
0.7
2.1
1.0 .
0
0
0
0
199.3.
0.0
0
0.9
44.5
26.7
0.2
0
48.4
0.3
5.0
0
1.5
0
0.1
9.5
0.1
3.6
0.0
5.1
0.3
0.3
0
0
0.0
0
0
4.2
0
0
0
0
34.8
0.5
0.0
0.6
22.8
17.0
0.4
0
2.5
0
0
0
0
0
0
0
0.4
0.3
0
13.9
0
0.9
0
0
0
0.5
0
0
0
0
0
0
3.1
0
0
0
1.6
0.8
0
0
8.2
0
0.8
0
0
0
0
0
0.8
0.1
0 .
0.0
0
0
0
0
0
0
o
0
0
0
0
0
2.0
0.2
0
0
0
0.1
0
0
9.2
0
0
0
3.7
0.0
0
0.7
0.8
34.7
0.1
7.5
0
2.6
0
0
0.1
0.7
0
0.8
0
7.0
0
0
252.9
0.9
0
4.9
43.4
16.2
0.3
0
10.1
0.1
0.5
0
2.1
0
0.0
5.4
0.1
4.6
0.0
5.9,
0.2
0.3
0
0
0
0.0
0
0.7
0
0
0
0
35.9
0.7
0
0.5
27.3
11.3
0.2
0
1 .4
0
0
0
aA weighted mean of "0" means no detectable levels were observed. A weighted mean of "0.0" means
that the weighted mean was less than 0.05.
mean concentration. The number of analytes in each
category was then compared (Table 16).
In Jacksonville, summer was the season in which the
greatest number of analytes had their highest
detection frequencies and highest mean
concentrations. In terms of the estimated percent of
the population with detectable levels, winter had the
next largest number of analytes, followed by spring.
Highest mean concentration levels occurred in the
summer for most analytes. Spring had the next largest
number of analytes for indoor air, whereas winter had
the next largest number for outdoor and personal air.
In Springfield/Chicopee, more analytes had their
highest mean concentrations in spring than winter. In
indoor and personal air, spring and winter differed not
at all or minimally in the number of analytes with
highest detection frequencies. The difference between
seasons was much more pronounced
considering mean concentrations.
when
These results imply that analytes that occurred at low
levels were not as consistent as common analytes in
their pattern of seasonal differences. The inconsistent
pattern of variation was probably due in part to
measurement error and in part to statistical sampling
variation associated with the small sample sizes.
Some of the inconsistent seasonal variation may also
reflect analytical protocol and reporting refinements
that occurred over the phases of NOPES. Larger
sample sizes and/or more refined analytical
techniques are needed to accurately assess the
seasonal variation for analytes found only at low
levels.
Analytes that occurred at higher concentrations
exhibited more consistent seasonal patterns. Figures
27
-------
5,000
3.000
Legend
JAX Summer
JAX Spring
JAX Winter
SP/CH Spring
Winter
5,000
3,000
25% 50% 75% 90% 95% 99%
JAX 72,500 145.000 217,500 261,000 275,500 287,100
SP/CH 33,750 67,500 101,250 121,500 128,250 133,650
Percent of Population Below Concentration Shown
Figures. Chlordane weighted cumulative frequency
distribution for personal air concentrations.
Legend
JAX Summer
JAX Spring
Winter
SP/CH Spring
SP/CH Winter
25% 50% 75% 90% 95% 99%
JAX 72,500 145,000 217,500 261,000 275,500 287,100
SP/CH 33,750 67,500 101,250 121,500 128,250 133,650
Percent of Population Below Concentration Shown
Figures. Heptachlor weighted cumulative frequency
distribution for personal air concentrations
5,000
3,000
Legend
—•-JAX Summer
-•-JAX Spring
-*-JAX Winter
-"-SP/CH Spring
-A-SP/CH Winter
25% 50% 75% 90% 95% 99%
JAX 72,500 145,000 217,500 261,000 275,500 287,100
SP/CH 33,750 67,500 101,250 121,500 128,250 133,650
Percent of Population Below Concentration Shown
Figure 4. Chlorpyrifos weighted cumulative frequency
distribution for personal air concentrations.
13 through 18 display the patterns for mean
concentrations of the five most prevalent analytes. For
most frequent and common analytes in Jacksonville,
summer season levels were highest, followed by
spring and then winter; however, winter levels were
5,000
3,000
1,000
O)
c
300
§
O
c
O
O
Legend
—*-JAX Summer
—•-JAX Spring
-A-JAX Winter
-•-SP/CH Spring
Winter
100
30
10
3
1
25% 50% 75% 90% 95% 99%
JAX 72,500 145,000 217,500 261,000 275,500 287,100
SP/CH 33,750 67,500 101,250 121,500 128,250 133,650
Percent of Population Below Concentration Shown
Figure 6. ortho-Phenylphenol weighted cumulative frequency
distribution for personal air concentrations.
higher than spring levels in outdoor and personal air
for some analytes. Spring levels were higher than
winter levels in Springfield/Chicopee for the majority of
frequent and common analytes.
28
-------
5,000
3,000
1,000 r
CO
e 300
O)
I
O
O
100 r
Legend
—»-JAX Summer
—»-JAX Spring
-*-JAX Winter
-•-SP/CH Spring
Winter
1
25% 50% 75% 90% 95% 99%
JAX 72,500 145,000 217,500 261,000 275,500 287,100
SP/CH 33,750 67,500 101,250 121,500 128,250 133,650
Percent of Population Below Concentration Shown
Figure?. Propoxur weighted cumulative frequency
distribution for personal air concentrations
Interpretation of the observed variations must account
for weather conditions during the data collection
periods. Table 17 summarizes the National Climatic
Data Center's Local Climatological Data for
Jacksonville and for Hartford, CT (the nearest
reporting station to Springfield/Chicopee) during the
sampling periods. The temperature data corroborate
Jacksonville respondents' comments that the spring
and winter sampling periods were colder than usual.
Whereas the Jacksonville spring sampling period was
locally described as being typical of early spring
conditions, the Springfield/Chicopee spring sampling
period was felt by some respondents to be
representative of late spring or even early summer.
Thus, the spring sampling periods were in some
sense not comparable in the two study areas. Despite
the somewhat below normal temperatures of the
Jacksonville winter sampling period, the weather was
still relatively mild compared to the
Springfield/Chicopee winter sampling period.
Weather may indirectly affect the air concentrations of
some analytes by influencing patterns of pesticide
use, heating, cooling, ventilation, and
peoples'activities. Temperature and humidity may
affect concentrations more directly by causing
changes in volatility or stability. The potential
complexity of the relationship between weather and air
concentrations, coupled with the limited number of
NOPES sampling periods, prevents development of
rigorous models of seasonal variation from the
NOPES data. For many analytes, the data are
sufficient only to permit rough approximation of the
annual levels of air exposure. Further work, building
on the NOPES findings, is needed to better
understand the seasonal variations in analyte levels in
air.
Study Area Comparisons
As expected, the two study areas showed marked
differences in the air levels of many target
compounds. The differences in personal air were
usually similar to those in indoor air but unlike those in
outdoor air. As was true in the analysis of seasonal
variation, alternative summarizations of the data yield
different conclusions and insights on regional
variation.
The total number of analytes detected were similar in
the two areas in the spring. In winter the number was
substantially lower in Springfield/Chicopee than in
Jacksonville (Table 15).
For each analyte, detection frequencies were
compared for Jacksonville and Springfield/Chicopee,
and the analyte was categorized as being higher in
Jacksonville, higher in Springfield/Chicopee, or
undetected in both. The number of analytes in each
category is tabulated by season and sample type in
Table 18. Results of a similar categorization based on
mean air concentrations are also presented in the
table. Figures 19 through 21 display the relative
differences in mean concentrations between the study
areas for the five most prevalent analytes.
These data clearly show that in indoor and personal
air the majority of analytes had higher concentrations
and occurred at greater frequencies in Jacksonville
than in Springfield/Chicopee. Among analytes
detected in both areas, Jacksonville mean
concentrations were often 2 to 30 times greater than
Springfield/Chicopee concentrations. Four analytes -
dacthal, folpet, 4,4'-DDT, and 4,4'-DDE - ran
counter to the others and usually had higher
concentrations in Springfield/Chicopee.
The relationship between outdoor air concentrations in
the study areas depended on the season. In winter
more analytes occurred' at higher levels in
Jacksonville than in Springfield/Chicopee, whereas in
spring neither area consistently prevailed. The "early
spring" conditions in Jacksonville and "late spring"
conditions in Springfield/Chicopee may partially
account for the higher Springfield spring
concentrations for some analytes (e.g., chlorpyrifos -
see Figure 20). Given the mild conditions in
Massachusetts, Springfield/Chicopee residents may
have been more likely to be outdoors using pesticides
than their Jacksonville counterparts, who were kept
inside by the relatively cool, wet weather. The harsher
winter conditions in Springfield/Chicopee undoubtedly
contributed to the more consistently low
Springfield/Chicopee winter outdoor air levels.
29
-------
Mean Concentration (ng/nr)
340
Outdoor Air hvil Personal Air
Summer
Spring
Winter
Spring
Winter
Jacksonville
Figure 8. Chlordane mean concentrations for indoor, outdoor, and personal air.
Springfield/Chicopee
380
Mean Concentration (ng/m )
Summer
Spring
Winter
Spring
Winter
Jacksonville
Figure 9. Chlorpyrifos mean concentrations for indoor, outdoor, and personal air.
Springfield/Chicopee
30
-------
Mean Concentration (ng/m3)
Outdoor Air EWl Personal Air
Summer
Spring
Winter
Spring
Winter
Jacksonville Springfield/Chicopee
Figure 10. Heptachlor mean concentrations for indoor, outdoor, and personal air.
100
Mean Concentration (ng/m3)
Summer
Spring
Winter
Spring
Winter
Jacksonville v Springfield/Chicopee
Figure 11. ortho-Phenylphenol mean concentrations for indoor, outdoor, and personal air.
31
-------
Mean Concentration (ng/m )
Summer Spring Winter Spring
Figure 12. Propoxur mean concentrations for indoor, outdoor, and personal air.
Winter
Table 14. Indoor, Outdoor, and Personal Air Concentration Correlations3
Number of Analytes with Spearman rank-order correlations of
Indoor vs Outdoor Air
Jacksonville
Summer
Spring
Winter
SpringfieW/Chicopee
Spring
Winter
Indoor vs Personal Air
Jacksonville
Summer
Spring
Winter
SpringfieloVChicopee
Spring
Winter
Personal vs Outdoor Air
Jacksonville
Summer
Spring
Winter
Springfield/Chicopee
Spring
Winter
<0.2
6
4
5
2
2
2
1
0
2
1
8
4
4
5
3
0.2-0.35
1
4
2
2
3
1
1 .
0
0
1
2
4
4
0
4
0.35-0.5
5
2
5
2
2
5
0
1
2
2
4
2
5
1
0
0.5-0.75
6
1
3
4
0
8
8
15
10
8
3
1
2
4
0
"0.75-1
0
0
0
0
0
10
14
9
6
5
1
0
0
0
0
'Correlations computed only for analytes detected in at least one sample from each air environment (e.g., indoor
and outdoor) in the given study area and season.
32
-------
Table 15. Seasonal Variation in Number of Detected
Analytes in Air
Number of Detected Analytes
Jacksonville3
Summer
Spring
Winter
Springfield/Chicopee
Spring
Winter
Indoor
Air
27
23
24
21
19
Outdoor
. Air
18
11
15
10
7
Personal
Air
26
po
24
21
18
34,4'-DDT, 4,4'-DDD, and 4,4'-DDE were excluded from the
calculations for Jacksonville because they were not included in
the summer season analyses.
Table 16. Seasonal Comparisons
Number of Analytes3
Seasonal Rankings of
Estimates
Estimated percent
with detectable levels
Jacksonville13
Summer > Spring, Winter
Spring > Summer, Winter
Winter > Summer, Spring
Springfield/Chicopee
Spring > Winter
Winter > Spring
Mean concentration
Jacksonville15
Summer > Spring, Winter
Spring > Summer, Winter
Winter > Summer, Spring
Springfield/Chicopee
Spring > Winter "
Winter > Spring
Indoor
Air
15
5
8
12
12
19
7
3
1.6
7
Outdoor
Air
12
o
7
g
2
12
1
7
g
3
Personal
Air
12
11
1 *3
I O
10
17
3'
7
Of\
tU
3
"Number of ana^ytes for which the seasonal ranking given in the row
termer3 T ^ SXamP'e' in the firSt row of the table,^n
terms of the estimated percent of the Jacksonville population with
detectable,ndoor air levels, 15 analytes had a higher pecenTn
aTr Tfln^" SPh'nH9 ?hf WLnter- '" b°th outdoor air and personal
air, 12 analytes had their highest "percent detectable" in the
summer.
"for'"?^' 4'4nD^D< 4'4'"DDE Wefe excluded from the Calculations
for Jacksonville because they were not included in the summer
season analyses. ~
Short-Term Temporal Variation
Short-term temporal variations in analyte levels were
examined in each season by selecting up to 10
respondents for replicate air samples collected at least
three days apart. The relatively small sample size
coupled with the low frequency of detection of many
analytes prevents precise quantitation of the levels of
short-term variation. However, the data do permit
assessment of the general magnitude of the variation
, and allow rough quantitative estimates for the
prevalent analytes.
For each replicate pair, the percent relative difference
between the replicates was computed for each analyte
detected in both samples. Table 19 summarizes the
replicate pair differences and also presents the
number of pairs in which an analyte was detected in
only one sample or in neither sample.
The data indicate that a substantial amount of
variation existed between some replicates The
variation was more pronounced in Jacksonville than
Springfield/Chicopee. |n a small number of cases ^
pairs differed by more than a factor of 10.
To assess the magnitude of short-term variability
relative to measurement error and seasonal variations
absolute differences between pairs of indoor air
measurements were computed for the five prevalent
analytes. Zeros were included in the calculations for
samples in which the analytes were not detected The
mean absolute differences in replicate indoor air
concentrations were computed for each study area
and season and compared to the mean absolute
differences between duplicate indoor air readings
(I able 20). Small sample sizes occasionally led to
considerable differences in the mean air
concentrations of the set of duplicate pair data and
the set of replicate pair data. Therefore, the mean
concentrations are presented in Table 20 to allow the
reader to assess the relative magnitude of the mean
absolute difference given the associated mean
concentration. The mean absolute differences
between seasons in multiseason respondent indoor air
concentrations were also computed and are presented
in Table 20.
The magnitude of the differences between estimated
measurement error variability (duplicates), estimated
short-term variability (replicates), and seasonal
variability (multiseason respondents) varied
considerably both within and between analytes
Because of the small sample size devoted to this
aspect of the study and the magnitude of the
variability observed, only qualitative conclusions are
supported regarding the relative magnitudes of these
components of variation in the two study areas.
Measurement error variability is generally less than
short-term variability, which itself is usually less than
seasonal variability. Moreover, short-term and
seasonal variability are generally more comparable
than short-term and measurement error variability
The fact that the short-term and seasonal variations
were generally comparable in magnitude suggests that
the factors contributing to short-term variations may
also be major components of seasonal variations
33
-------
Summer
Spring
Winter
Relative Mean Concentration (Percent)
100
Chlordane
Chlorpyrifos
Heptachlor ortho-Phenylphenol' Propoxur
Figure 13. Seasonal variation in relative mean indoor air concentrations in Jacksonville as percents of summer
mean concentrations.
I Spring
Winter
Relative Mean Concentration (Percent)
100
Chlordane
Chlorpyrifos
Heptachlor ortho-Phenylphenol Propoxur
Figure 14. Seasonal variation in relative mean indoor air concentrations in Springfield/Chicopee as percents of
spring mean concentrations.
34
-------
HH Summer
Relative Mean Concentration (Percent)
100
! Spring
I Winter
Chlordane Chlorpyrifos Heptachlor' ortho-Phenylphenol Propoxur
Figure 15. Seasonal variation in relative mean outdoor air concentrations in Jacksonville as percents of summer
mean concentrations.
Spring
Winter
Relative Mean Concentration (Percent)
100
Chlordane
Chlorpyrifos
Heptachlor ortho-Phenylphenol Propoxur
Figure 16. Seasonal variation in relative mean outdoor air concentrations in Springfield/Chicopee as oercents of
spring mean concentrations. 3 v fjeroenu, OT
35
-------
Summer
Spring
Winter
Relative Mean Concentration (Percent)
110
Chlordane
Chlorpyrifos
Heptachlor ortho-Phenylphenol
Propoxur
Rgure 17. Seasonal variation in relative mean personal air concentrations in Jacksonville as percents of summer
mean concentrations.
Spring
Winter
Relative Mean Concentration (Percent)
100
Chlordane Chlorpyrifos Heptachlor ortho-Phenylphenol Propoxur
Figure 18. Seasonal variation in relative mean personal air concentrations in Springfield/Chicopee as percents of
spring mean concentrations.
36
-------
Table 17. Local Weather During NOPES Data Collection Periodsa
Jacksonville
Springfield/Chicopee
Temperature (°F)
Minimum
Maximum
Avg Daily Minimum
Avg Daily Maximum
Avg Daily Average
Avg Departure from Normal
Precipitation
Days with Precip.
Total (in.)
Avg Percent of Total Possible
Sunshine
Summer
(8/21-9/18/86)
67
96
71
88
80
0.1
15
5.1
Spring
(3/20-4/13/87)
34
86
50
74
62
-2.6
9
3.1
Winter
(2/1-2/17/88)
25
82
40
65
52
-1.8
6
' 1.2
Spring
(6/1-6/17/87)
43
92
57
79
69
1-1.
7
2.5
Winter
(3/11-3/28/88)
12
74
29
48
39
1.0
9
1.4
~ 71
Climatologica. Data provided by the Nationa. Climatic Data Center. Springfield/Chicopee values are based on data for
Table 18. Study Area Comparisons
Number of Analytesa
Order of Study Area Estimates
Indoor Outdoor
Air Air
Personal
Air
Estimated percent
with detectable levels
Spring
Jacksonville > Springfield 21
Springfield > Jacksonville 7
Winter
Jacksonville > Springfield 24
Springfield > Jacksonville 4
Mean concentration
Spring
Jacksonville > Springfield 24
Springfield > Jacksonville 4
Winter
Jacksonville > Springfield 22
Springfield > Jacksonville 6
15
1
14
2
18
20
3
19
8
24
2
mTho H ana yteS f°r whlch the studv area ranki"9 given in the
row heading was true. For example, in the first row of the table in
spring the estimated percent of the population with detectable
£ r fJ^VelS was hi9her in Jacksonville than in
Springfield/Ch.copee for 21 analytes. In outdoor air, 7 analytes
qnrinnfi^f- PerCent detectable" "> Jacksonville than in
Springfield/Chicopee, and in personal air 18 analytes were higher
in J3CKsonvnl0,
Water Exposure
Water sampling was by design only a small
component of NOPES. Routine sampling of public
^DctT'lf3 by Jacksonvi"e and Springfield prior to
NOPES had not identified any contamination by the
target compounds, and water samples collected and
analyzed during the NOPES pilot study also did not
contain detectable levels of any analytes. Therefore a
minimal sampling effort was believed to be sufficient
for estimating water exposure to the tarqet
compounds. a
In all, 29 tap water samples were analyzed in NOPES
- 17 from Jacksonville and 12 from
Springfield/Chicopee. Six Jacksonville samples were
Q°m /I^u6 wells or water supplies; all
Spnngfield/Chicopee samples were from the public
water supply. Most of the samples contained no
detectable levels of any of the analytes. The only
analytes detected were as follows: V
Jacksonville
gamma-BHC -
6 ng/L in a summer sample from a
home served by a private water
company
58 ng/L in a spring sample from a
home served by a private water
company
327 ng/L in a winter sample from
a home served by the public
water supply
Springfield/Chicopee
ortho-Phenylphenol -110 ng/L and 36 ng/L in
two spring samples
Diazinon
Dichlorvos
- 30 ng/L in a spring sample.
Propoxur
Among households in the water subsample no
correlation was observed between indoor air
concentrations and water concentrations.
ah , SizSS Prevent estimation of
weighted population exposure estimates from these
data. However, the lack of detectable levels for
most analytes and the relatively low levels
occasionally detected for others suggest that
37
-------
Jacksonville
Springfield/Chicopee
Relative Mean Concentration (Percent)
100
Spring Winter Spring Winter Spring Winter Spring Winter Spring Winter
Chlordane
Chlorpyrifos
Heptachlor ortho-Phenylphenol Propoxur
Rguro19. Mean indoor air concentrations in Springfield/Chicopee as percents of Jacksonville mean
concentrations.
exposure to the NOPES target compounds from water
is minimal in the two study areas.
Dermal Exposure
The dermal exposure component of NOPES was
primarily a pilot study of a method for quantifying
dermal exposure levels during acute exposure events.
Chronic dermal exposure was not addressed. The
number of events monitored was small, and events
were not randomly selected, so estimated population
exposure levels cannot be developed. However,
analysis of the glove data does permit assessment of
the method, and provides an initial impression of the
relative importance of acute dermal exposure.
The monitored application events included spraying
and shampooing pets to eliminate fleas, spraying
insecticides inside and outside residences, spraying
herbicides, spraying disinfectants, and spreading
granular insecticides. Many applications involved
ready-to-use aerosols; in others, pesticides were
applied by hand, handsprayer, or mechanical
spreader. Precautionary measures, such as wearing
protective clothing or work gloves (over the sample
gloves), were rarely taken during the applications.
Twenty-two events were monitored, eight of which
involved products containing one NOPES target
compound, and four of which involved products
containing two target compounds. The compounds
applied were chlorpyrifos, diazinon, malathion,
carbaryl, dicofol, dichlorvos, resmethrin, and
methoxychlor.
In all events involving the application of one or more
target compounds, the compounds were measured on
the sample gloves, usually at high concentrations. In
the majority of these events, detectable levels of the
applied target compound were also measured in the
personal air samples collected during the events. The
lack of detected air levels in five cases may have
been due to the high limits of detection inherent in
short-duration sampling. This is especially likely for
the cases involving dicofol and dichlorvos, which had
relatively high detection limits compared to other
analytes.
When assessing the suitability of the cotton gloves
worn during application events as sampling devices,
the fact that the applied target compounds were
always found in the gloves is a desirable
characteristic. However, the data reveal that the
gloves collected more than just the applied
compounds. In 18 of the 22 monitored events,
analytes other than those being applied were also
detected in the gloves. These other analytes were
usually, but not always, at low concentrations. A
reasonable explanation for some of these findings is
that the "unexpected" analytes were present as
residues on the application equipment (especially
handsprayers and spreaders), in the application area,
38
-------
or on the respondent from a previous event. The
unexpected analytes were in some cases present in
the household pesticide inventory of the participant.
Before accurate estimates of acute dermal exposure
levels can be obtained, the measurement method
must be refined. Questions that need to be addressed
include the following:
• What were the sources of the unexpected
analytes?
• For a particular type of application event, such as
bathing a dog with a flea shampoo, how much do
concentration, measurements vary between
applications?
• Do the gloves overestimate certain types of dermal
exposure, such as that due to liquid contact,
because of their absorptive and/or adsorptive
nature? .«
• Are gloves adequate for assessing total acute
dermal exposure during application events,
especially during warm weather when people may
be working in shirt sleeves and shorts?
• What is the distribution of application events over
time and across the population?
Although the NOPES data cannot support estimation
of population dermal exposure levels, they can be
used to gain insight on the general magnitude of the
exposure experienced during application events. Lewis
(1988) used some of the NOPES data to model the
dose associated with a particular summer season
application and then compared it to the mean
estimated daily personal air exposure levels. The
event modelled (a 5 minute outdoor application of
granular chlorpyrifos by hand) had the highest glove
concentration observed during Phase I and so in
some sense represented a worst-case scenario. His
findings for the particular case examined indicated that
the dermal dose (assuming a 1% dermal absorption
factor) from the event was 40 times greater than the
daily air exposure. A lack of information on the
number of times similar events were performed over
the course of a year prevents computation of annual
exposure levels. Nonetheless, the single day
comparison indicates that dermal exposure is
potentially a significant contributor to overall exposure
levels.
§•( Jacksonville Springfield/Chicopee
Relatively/lean Concentration (Percent)
400
300
200
100
n
Spring Winter Spring Winter Spring Winter Spring Winter
Chlordane
Chlorpyrifos Heptachlor Propoxur
Figure 20. ^ea^outdoor air concentrations in Springfield/Chicopee as percents of Jacksonville mean
39
-------
Jacksonville
Springfield/Chicopee
Relative Mean Concentration (Percent)
Spring Winter Spring Winter Spring Winter Spring
Spring Winter
Chlordane Chlorpyrifos Heptachlor ortho-Phenylphenol Propoxur
Figure 21. Mean personal air concentrations in Springfield/Chicopee as percents of Jacksonville mean concentrations.
Table 19. Replicate Relative Percent Differences3
Number and Percent of Replicate Pairs
Relative Percent Difference15
Jacksonville
Indoor air
Summer
Spring
Winter
Outdoor air
Summer
Spring
Winter
Daren rial
r 01 oUl 101
Summer
Spring
Winter
Springfield/Chicopee
Indoor air
Spring
Winter
Outdoor air
Spring
Winter
Personal
Spring
Winter
<67%
43 (21%)
29 ( 8%)
46 (16%)
11 (6%)
1 ( 0%)
4(1%)
'"
28 (15%)
36(11%)
49(17%)
23 ( 7%)
27 ( 8%)
6 ( 2%)
0 ( 0%)
14 ( 4%)
28 ( 9%)
67-164%
12(6%)
24 ( 7%)
19 (6%)
3 ( 2%)
3(1%)
5 ( 2%)
18 (10%)
19 ( 6%)
9 ( 3%)
4 ( 1%)
6 ( 2%)
3(1%)
0 ( 0%)
6 ( 2%)
6 ( 2%)
>164%
2 (1%)
3 (1 %)
1 (0%)
0 (0%)
1 (0%)
0 (0%)
1 (0%)
3(1%)
6 (2%)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
1 (0%)
1 (0%)
Only Detected in
One Sample
28 (13%)
25 ( 8%)
24 ( 8%)
20 (12%)
21 ( 7%)
19 (6%)
20 (11%)
21 ( 6%)
24 ( 8%)
20 ( 6%)
9 ( 3%)
16 ( 5%)
2(1%)
13 (4%)
17(5%)
Not Detected in
Either Sample
116 (58%)
249 (76%)
207 (70%)
137 (80%)
304 (92%)
269 (91%)
121 (64%)
251 (76%)
209 (70%)
283 (86%)
287 (87%)
305 (92%)
294 (99%)
296 (90%)
277 (84%)
Total"
201 (100%)
330 (100%)
297 (100%)
171 (100%)
330 (1 00%)
297 (100%)
188 (100%)
330 (100%)
297 (100%)
330 (100%)
329 (100%)
' 330(100%)
296 (100%)
330(100%)
329 (100%)
a Relative percent difference, computed for pairs with detected values for both samples, calculated as: 100 " (primary value -
replicate valuel / (mean of the two values).
b Relative differences of less than 67% indicate that the paired values differed by a factor of 2 or less; relative differences
greater than 164% indicate the pair differed by an order of magnitude or more.
Total for up to 10 households and 33 analytes (including pentachlorophenol, for which all values were non-detect).
40
-------
Table 20. Duplicate, Replicate, and Seasonal Indoor Air Concentration Differences, (ng/m3)
Duplicates . Replicates
Multiseason Respondents
Chlordane
Jacksonville
Summer
Spring
Winter
Springfield
Spring
Winter
Chlorpyrifos
Jacksonville
Summer
Spring
Winter
Springfield
Spring
Winter
Heptachlor
Jacksonville
Summer
Spring
Winter
Springfield
Spring
Winter
ortho-Phenylphenol
Jacksonville
Summer
Spring
Winter
Springfield
Spring
Winter
Propoxur
Jacksonville
Summer
Spring
.Winter.
Springfield
Spring
Winter
Mean
Conc.a
55
505
145
51
54
247
268
187
63
18
13
142
43
5
7
81
101
51
107
54
142
378
92
48
10
Mean
Abs.
Diff.b
2
40
60
38
12
38
8
17
16
1
3
14
3
4
<1
2?
33
6
39
12
28
13
10
36
4
No. Of
Pairs
6
10
9
8
7
6
10
9
8
7
6
10
9
8
7
4
10
9
8
7
4
10
9
8
7
Mean
Conc.a
271
249
129
64
140
362
162
152
34
5
157
114
64
20
26
91
96
82
26
46
289
168
51
64
17
Mean
Abs.
Diff.b
98
55
22
43
32
169
101
198
14
2
41
75
22
11
3
46
145
87
22
23
138
137
30
18
12
No. of
Pairs
8
10
9
10
10
8
10
9
10
10
8
10
9
10
10
5
10
9
10
10
5
10
9
10
10
Mean
Cone.
Over
Seasons0
369
242
32
259
122
13
218
124
10
75
80
34
529
197
52
Mean
Abs. Diff.
Between
Seasonsd
343
1.14
29
276
114
11
223
108
15
72
117
38
629
184
77
No. Of
Pairs
19
16
15
19
16
15
19
16
. 15
17
16
15
17
16
15
aUnweighted mean of all matched pair data.
bllnweighted mean of the absolute differences between matched pairs.
Unweighted mean of data for two seasons from multiseason respondents. Values on the rows labelled'Spring' are means for combined
summer and spring data; rows labelled 'Winter' are for combined spring and winter data.
dValues on rows labelled 'Spring' are the unweighted mean absolute differences between summer and spring concentrations; values on rows
labelled 'Winter' are for mean absolute differences between spring and winter concentrations.
41
-------
The approach used in Lewis (1988) was applied to all
16 target compound applications monitored in
NOPES. The following assumptions were used to
compute dermal and daily air exposure doses for the
applied compounds:
• 20 m3 of air per day respired by a 70 kg adult
• dermal absorption factors of 0.01 for granular and
dust applications and 0.1 for liquid and spray
applications.
The estimated dermal doses, computed by multiplying
the glove concentration by the appropriate absorption
factor, ranged from 0.02 pg to 16,000 pg. Daily air
exposure doses were calculated as the mean personal
air concentration estimates (ng/m3) from Tables 12
and 13 multiplied by 20 m3 per day of respired air. In
only three of the 16 cases was the dermal dose less
than the estimated daily air dose. The dermal dose
was more than an order of magnitude greater than the
daily air dose in more than half the cases.
These results confirm the earlier conclusion that the
acute dermal component may for some analytes
contribute substantially to total exposure. More
accurate exposure measurements and data on the
frequency of applications are needed to better quantify
the actual contribution. A complete understanding of
dermal exposure also requires identification of all
chronic dermal exposure pathways, such as contact
with dust or residues on surfaces, and measurement
of exposures via those pathways.
Dietary Exposure
NOPES was not designed to directly measure dietary
exposure to the target compounds. Instead, dietary
exposures were estimated using residue concentration
information developed by the U.S. Food and Drug
Administration as part of their ongoing Total Diet
Study (TDS) program together with dietary intake data
from the survey participants. The uncertainties
associated with the estimated food exposures are
much greater than those for air exposures. Air
exposures were measured directly, but food
exposures were estimated indirectly. The TDS is not
designed to be statistically representative of all
commodities in commerce. The estimation procedure
does not account for the effects of food preparation in
the home (e.g., washing and cooking). Therefore, the
only conclusions that are supported by the study
regarding food exposures are qualitative comparisons
of the relative magnitudes of the food and air routes of
exposure, as reported in the next section.
Relative Contributions of Exposure
Pathways
One of the primary objectives of NOPES was to
assess the relative contributions of the four pathways
~ air, water, food, and dermal contact ~ to overall
exposure to the target compounds. The NOPES
findings, although not sufficient to permit precise
quantification of the relative contributions, do provide
some insight on the magnitude of exposure
attributable to each pathway.
For all target compounds, exposure from drinking
water appeared to be minimal in both study areas.
This conclusion is consistent with the NOPES pilot
study results and with the sampling performed in
ongoing municipal water quality testing programs.
The dermal exposure pathway could not be accurately
characterized, making conclusions about its relative
contribution to exposure tentative at best. The primary
conclusion is that acute dermal exposures that occur
during application events may contribute substantially
to total exposure for some analytes (Lewis, 1988).
Collection of house dust using a high-volume surface
sampler was pilot tested in NOPES in the Jacksonville
winter season. House dust may be a source of
exposure to pesticides via dermal contact, ingestion,
and inhalation of suspended particulates, especially
for infants and toddlers. The results of the pilot study
(Budd et al., 1988) suggest that further study is
warranted but are insufficient to support conclusions
about the relative 'importance of house dust at this
time.
Qualitative comparisons of the relative exposure
contributions of air and food were possible for some
of the target compounds. The relative air and food
contributions were computed for daily exposures.
Mean daily exposure from inhalation was estimated by
multiplying the mean personal air concentration
estimates (ng/m3) for each season (Tables 12 and 13)
by 20 m3 air respired per day. These daily air
exposure estimates were then compared to the daily
dietary estimates. Only qualitative comparisons were
supported by the data. The 25 analytes for which
dietary estimates were available were partitioned into
five categories on the basis of their estimated relative
exposure levels'in air and food as shown in Table 21.
For some of the analytes, the relationship between air
and dietary exposures seems to reflect the general
uses of the analytes. Prior to being withdrawn in 1988,
chlordane and heptachlor were used primarily as
42
-------
Table 21. Relative Air and Dietary Exposure Estimates
Relative Exposure Category Analytes
1. Mean air exposure always much higher
than estimated dietary exposure
2. Variable, but mean air exposure often
higher than estimated dietary exposure
3. Generally present in air, but mean air
exposure lower than estimated dietary
exposure
4. Mean air exposure usually much lower
'than estimated dietary exposure
5. Both exposures estmated to be very low
Chlordane
Heptachlor
Aldrin
Chlorpyrifos
Diazinon
gamma-BHC
4,4'-DDT
Dacthal
Dieldrin
alpha-BHC
Captan
Carbaryl
4,4'-DDE
Dicofol
Heptachlor epoxide
Hexachlorobenzene
Malathion
Methoxychlor
cis-Permethrin
trans-Permethrin
Ronnel
Oxychlordane
4,4'-DDD
Chlorothalonil
Folpet
termiticides and so would be expected to be present
in air at much higher levels than in food. Alternatively,
some of the pesticides in the second, third, and fourth
categories of Table 21 are used in agriculture or food
processing and distribution settings and so are
logically found in food as well as air. For analytes in
these categories, volatility, the amount of
nonagricultural use, and other factors determine the
relative exposure levels from air and food.
For the remaining seven analytes, a relative
assessment of food and air exposures was not
possible because of the lack of dietary exposure
estimates. However, three of these analytes --
dichlorvos, ortho-phenylphenol, and propoxur -- occur
at relatively high levels in personal air in Jacksonville
and have a variety of household uses, leading to the
expectation that the air exposures were higher than
the dietary exposures. For the other four analytes, the
data are insufficient to assess the relative air and
dietary contributions. Estimated mean air
concentrations of bendiocarb were high in the summer
but low in all other seasons, whereas the mean air
concentrations of atrazine, 2,4-D, and resmethrin were
always low. - .
Air Exposure and Questionnaire Data
Relationships
A primary objective of exposure assessment research
is the development of validated, predictive models of
exposure. One approach to modelling is to identify
and quantify relationships between data collected in
questionnaires and exposure levels for the
corresponding survey respondents. The ability to
estimate exposure levels from questionnaire data is
desirable in terms of both cost and respondent
burden. Collection and laboratory analysis of samples
is costly compared to questionnaire data .collection.
Moreover, monitoring imposes a relatively high burden
of time and responsibility on study respondents.
Questionnaire-based modelling has been explored for
a number of compounds (see, for example, Akland et
al., 1985; Wallace, 1987; Ryan et al., 1988).
NOPES was designed to provide a data base from
which to develop and test air exposure models for the
garget compounds. Exploratory analysis of this data
base and the models it may support has begun and is
summarized in this section. Ultimately, the NOPES
data may be used to construct quantitative
multivariate air exposure models for some of the
NOPES analytes.
Effectiveness of Exposure Stratification
A preliminary, simplistic questionnaire-based model
was an integral part of the NOPES sampling design.
Potential indoor air exposure strata were defined using
screening questionnaire data and were used to control
the distribution of the third-stage sample. The
algorithm used to define strata was based on
responses to four questions:
• Was there any use of pesticides on indoor plants?
• Was there any use of pesticides on household
pets?
* Was the housing unit treated with termiticides?
• Were household insecticides applied within the
housing unit during the past year?
A housing unit was assigned to the high-exposure
stratum if the answer to at least three of the four
questions was "Yes", to the medium-exposure
stratum if any two answers were "Yes", and to the
low-exposure stratum otherwise.
The effectiveness of the stratification model can be
assessed in several ways. One way is to compute the
number of analytes detected in indoor air or personal
air for each respondent and to then compare the
mean number detected in each of the three strata.
Alternatively, for each analyte in each study area and
season, the mean concentration or the percent of the
population with detectable levels can be estimated for
each stratum. The strata can then be ranked or
otherwise compared, and the results can be
summarized across analytes. Both of these
approaches were used and are summarized in Table
22. (All analyses are based on the final exposure
43
-------
categorizations of respondents. Because of concerns
about the accuracy of the termiticide treatment
information obtained during screening, the termite
questions were asked again in the study
questionnaire. In some cases, the study questionnaire
answers prompted a revision of a respondent's
exposure category.)
The results presented in Table 22 indicate that the
stratification model was effective in a general way for
indoor air. The high-exposure category summary
measure was usually higher than the medium-
exposure measure, which in turn was usually higher
than the low-exposure measure. The model's
effectiveness was more limited for personal air, which
was expected given the definition of the model. The
indoor air findings imply that this type of model can be
useful when only broad relative categorizations are
needed.
However, examination of the stratum-specific statistics
for particular analytes reveals that the stratification
model was not generally effective as a predictive tool
for relative exposure levels for individual compounds.
Mean indoor air analyte concentrations rarely differed
significantly between strata, and the relative ranking of
high-, medium-, and low-exposure stratum means was
often inconsistent across seasons (Table 23). The
stratum-specific "percent detectable" estimates
displayed a similar lack of differentiation and
consistency.
This result was not surprising because the strata were
only intended to be predictive in a general sense, and
not for individual analytes. The NOPES air samples
were analyzed for compounds with a wide variety of
chemical properties and use characteristics. Any
single questionnaire- based index that tries to address
all of the compounds does so at the expense of
adequate prediction for the individual compounds.
Development of predictive models using the NOPES
data should therefore focus on individual analytes or
classes of analytes with similar properties (e.g.,
termiticides).
Stratification based on categories related to potential
exposure levels was somewhat effective in NOPES,
and would be a desirable feature in the design of any
subsequent surveys. Analysis of the NOPES data (see
below) indicates that the exposure categorization
criteria could be altered to make the stratification
more effective. Recommended changes include the
following:
• Delete the "indoor plant pesticide use" component
of the definition because of the infrequent incidence
of pesticide use on indoor plants.
• If the target pesticides are similar to those in
NOPES, the "use of pet pesticides" component of
the definition could be dropped because the
majority of current pet pesticide products do not
contain any of the NOPES analytes.
• Any survey studying termiticides or pesticides that
are no longer in use should incorporate age and
type of housing unit in the definition of the
exposure strata.
The NOPES experience also indicates .that differences
in sampling rates across strata should be much
smaller than those used in this study. The limited
effectiveness of the stratification was not great
enough to justify the vastly unequal sampling --rates,
which resulted in considerable unequal weighting and
variance inflation for overall population estimates.
Exploratory Analyses
The exploratory work done to date has examined the
relationships between specific analytes and a variety
of questionnaire items. Given the relatively small
sample sizes, many analytes were detected too
infrequently to permit statistical assessment of their
relationships to questionnaire data. Therefore, the
work has focused primarily on the common and
prevalent compounds. Even for these analytes,
standard errors for many estimates are large because
of small sample sizes, unequal weighting effects, and
the inherent variability of the measurement data. Most
observed differences were consequently not
statistically significant. However, the goal of the
exploratory work is to identify those differences that
are suggestively large and consistent with known
causal processes. Such differences can then be
explored further with more sophisticated analytical
techniques.
Two types of analyses have been performed: (1)
analyses based on general characteristics of
respondents' housing units, and (2) analyses focused
on potential uses of specific analytes. In analyses of
the first type, the reporting error for the questionnaire
data is believed to have been relatively slight,
because the concepts involved (age of housing unit,
type of housing unit, and location of, .the indoor fixed-
site sampler) are familiar and straightforward.
Reporting error may have been more of,a problem in
the second type of analysis because of
misunderstanding or lack of knowledge on the part of
some respondents.
Characteristics of Housing Units
Age of Housing Unit. A correspondence might be
observed between age of housing unit and indoor air
concentration of an analyte for several reasons:
44
-------
Table 22. Overall Effectiveness of the Exposure Stratification Model
Potential Indoor Air Exposure Category
High
Mean Rank Over All Analytes
(1 = highest value, 3 = lowest value)
Indoor Air
Percent Detect
Jacksonville
Springfield
Mean Concentration
Jacksonville
Springfield
Personal Air
Percent Detect
1.6
1.9
1.6
1.8
Medium
2.0
2.0
2.2
2.0
Low
Mean Number of Analytes Detected for a Respondent
(standard errors in parentheses)
Indoor Air
Jacksonville
Summer
Spring
Winter
Springfield
Spring
Winter
Personal Air
Jacksonville
Summer
Spring
Winter
Springfield
Spring
Winter
9.0(1.5)
8.0 (0.4)
9.3 (0.5)
4.0 (0.4)
4.1 (0.1)
7.7(1.2)
7.1 (0.6)
8.0 (0.3)
3.7 (0.4)
4.5 (0.1)
8.0 (0.5)
7.3 (0.5)
8.4 (0.8)
3.9 (0.7)
3.7 (0.3)
8.4 (0.5)
6.3 (0.5)
8.9 (0.9)
3.0 (1.2)
3.0 (0.5)
8.7 (0.5)
6.7 (0.8)
7.8 (0.6)
3.4 (0.6)
4.3(4.1)
7.8(1.1)
5.7 (1.3)
7.9 (0.4)
4.1 (0.7)
4.5 (0.6)
2.4
2.1
2.2
2.2
jacKsonvme
Springfield
Mean Concentration
Jacksonville
Springfield
1.9
1.8
2.0
1.8
2.0
2.2
2.0
2.1
2.1
2.0
2.1
2.0
• Some housing units were built after use of certain
analytes was discontinued.
• As units age, the need to use certain analytes may
increase or decrease.
• Units might "accumulate" an analyte over time
because of repeated applications.
• Residents of older units might typically differ from
residents of newer units in their use of pesticides
(for reasons not directly related to the age of the
unit).
Preliminary analyses suggest that the action of some
of these factors can be observed in the NOPES data.
For each analyte, mean indoor air concentrations were
computed for three housing unit age categories - 20
years and less, 21 to 40 years, and 41 years and
older. Some analytes were detected too infrequently
to provide useful comparisons. Others, including some
of the more common analytes, such as chlorpyrifos,
ortho-phenylphenol, and propoxur, showed no
evidence of substantial, consistent differences among
the categories. However, such differences were
observed for 10 analytes, and these are summarized
in Table 24.
Three of the analytes in the table have not been
registered for use in the United States since the
1970s (DDT and DDE since 1971, and alpha-BHC
since 1978), and the use of two others - aldrin and
dieldrin - has been drastically reduced. Concentrations
of DDT, DDE, alpha-BHC, and dieldrin were
consistently lower in housing units 20 years old or
less, presumably reflecting the lack of recent use. The
picture for aldrin is not as clear. The fact that older
pesticides such as DDT were detected in NOPES
samples is not surprising given the known persistence
45
-------
Table 23. Ranks of Exposure Category Mean Indoor Air Concentrations of Commonly Detected Analytes
(H = High exposure category, M = Medium exposure category, L = Low exposure category)
Jacksonville
Springfield/Chicopee
Analyte
gamma-BHC
Heplachlor
Chlorpyrifos
D'ioldrin
Chlordano
ortho-Phenylphenol
Propoxur
Diazinon
Summer Rank3
1 2 3
H M L
MHL
LH M
MHL
H M L
H ML
LH M
H M L
Sprinq Rank3
1 2 3
H M L
MLH
HML
LH M
H LM
H LM
LM H
HML
Winter Rank3
1 2 3
H LM
LHM
HLM
HLM
LHM
HLM
MLH
*LM H
Sprinq Rank3
123
LH = M
LHM
HML
LH = M
LHM
MLH
HML
MHL
Winter Rank3
123
LHM
HML
HML
L M H
HML
HLM
HLM
LHM
»1 = highest exposure category mean concentration, 3 = lowest exposure category mean concentration, equal signs
represent ties.
of these compounds. The dust sampling pilot study
results suggest that dust may be a significant
reservoir for older pesticides (Budd et al., 1988).
The two most frequently found termiticides in NOPES
- chlordane and heptachlor -- displayed a similar
pattern of generally lower levels in newer homes. Both
of these pesticides have been commonly used until
recently, and have long residual lifetimes. The
observed pattern may therefore be due to higher rates
of application in earlier decades, or accumulation of
the analytes. However, the pattern could also be an
artifact of a correlation of housing unit age with some
other factor, such as housing unit type (discussed
below).
The other three analytes in the table -- bendiocarb,
gamma-BHC, and hexachlorobenzene - were all
detected relatively frequently in Jacksonville, and
show some evidence of consistent patterns over the
three Jacksonville sampling seasons. Gamma-BHC
and hexachlorobenzene concentrations were lowest in
the newer homes, and bendiocarb concentrations
were highest in homes of intermediate age. These
patterns were not observed in Springfield/Chicopee,
perhaps simply because the analytes occurred there
with such low frequency.
Type of Housing Unit. A goal of the NOPES sampling
design was to ensure that various housing types were
represented in the sample. Monitored housing units
included unattached single-family dwellings, attached
single-family dwellings (e.g., duplexes, townhouses),
multiunit buildings (e.g., apartments), and mobile
homes (the latter primarily in Jacksonville). To
examine the possibility that indoor air concentrations
were related to type of housing unit, analyte mean
concentrations were computed for each type and
compared. The comparisons for some analytes are
summarized in Table 25.
For most analytes there was little evidence of a
relationship between indoor air concentrations and
housing type. However, consistent patterns were
observed for the analytes that have been primarily
used as termiticides - aldrin, chlordane, dieldrin, and
heptachlor. All of these compounds had higher levels
in unattached and attached single units than in
apartments, and were at much lower levels in mobile
homes (except for the one Springfield/Chicopee
mobile home). The most plausible explanation for the
lower multiunit and -mobile home levels is that
termiticide treatments either were not performed or
were needed less often because of the type of
construction materials used in these units.
Alternatively, because multiunit buildings and mobile
homes were younger on average than single-family
units, the observed differences between housing
types may reflect an age effect. Further analysis is
needed to separate the confounding effects of age
and type of housing unit.
Alpha-BHC, DDT, and DDE have all been out of use
since the 1970s (although alpha-BHC still enters the
system as an environmental conversion product of
gamma-BHC). All were absent from mobile homes,
which is to be expected given the limited lifespan of
mobile homes. These three analytes did not display a
uniform pattern of concentrations in single-unit versus
multiunit dwelling's. Although alpha-BHC displayed a
tendency toward lower multiunit levels, no such
tendency was .apparent for DDT or DDE.
46
-------
Table 24. Indoor Air Concentration vs. Age of Housing Unit
Mean concentration (ng/m3)
Analyte
Aldrin
alpha-BHC
Bendiocarb
Chlordane
4,4'-DDE
4,4'-DDT
Dieldrin
gamma-BHC
Age of
Housing Unit -
(years)
<20 Mean'
s.e.
21-40 Mean
s.e.
>41 Mean
s.e
<20 Mean
s.e.
21-40 Mean
s.e.
>41 Mean
s.e.
< 20 Mean
s.e.
21-40 Mean
s.e.
>41 Mean
s.e
<20 Mean
s.e.
21-40 Mean
s.e.
>41 Mean
s.e.
<20 Mean
s.e.
21-40 Mean
s.e.
>41 Mean'
s.e.
<20 Mean
s.e.
21 -40 Mean
s.e.
>41 Mean
s.e
< 20 Mean
s.e.
21-40 Mean
s.e.
>41 Mean
s.e
<20 Mean
s.e.
21-40 Mean
s.e.
>41 Mean
s.e
Summer
4.4
(4.9)
37.1
(25.5)
70.2
(76.4)
0.6
(0.6)
1.1
(0.6)
3.1
(2.4)
16.7
(6.6)
139.8
(111.3)
10.4
(13.0)
162.2
(75.2)
402.8
(131.8)
383.9
(88.2)
-
,
-
-
-
-
5.6
(2.8)
16.0
(1-9)
29.7
(10.9)
6.6
(2.8)
16.0
(1.9)
29.7
(10.9)
Jacksonville
Spring
19.6
(18.7)
4.6
(1-3)
1.6
(0.8)
0.2
(0.3)
1.1
(0.1)
1.8
(1.0)
5.8
, (3-3)
8.3
(5.3)
0.7
(0.7)
71.9
(47.5)
215.3
(90.1)
432.7
(86.7)
0
_
0.5
(0.6)
1.0
(0.3)
0.1
(0.1)
1.0
(0.7)
1.7
(0.6)
1.7 •
(1.7)
6.7
(2.1)
15.2
(5.1)
4.7
(1.7) :
6.7
(2.1)
15.2
(5..1)
Springfield/Chicopee
Winter
3.4
(2.8)
11.4
(2.7)
2.9
(0.4)
0.0
(0.0)
1.1
(0.7)
2.2
(1.8)
1.9
(1.2)
6.1
(2.1)
0.3
(0,3)
53.2
(19.7)
128.3
(13.3)
533.6
(93.9)
0
0.1
(0.1)
0.5
(0.2)
0
_
0.3
(0.1)
1.4
(0.7,
3.1
(1.8)
6.8
(2.1)
11.7
(4.0)
1.7
(1.8)
6.8
(2.1)
11.7)
(4.0)
Spring
0
0
0
0
0
0.4
(0.5)
0.9
(0.9)
0
0
18.3
(3.8)
428.6
(311.8)
41.3
(15.0)
0
2.0
(1.2)
0
-
0
_
0.0
(0.0)
0
-
0
0
2.7
(2.6)
0
0
2.7
(2.6)
Winter
0
0.8
(0.5)
0.0
(0.0)
o
0
0
o
1 1
(1.1)
0.0
(0.0)
30.6
(12.3)
43.9
(6.9)
29.8
(11-8)
0
1.2
(0.8)
0.4
(0.2)
0
0
1.3
(0.3)
0.5
(0.2)
0.2
(0^2)
9.1
(5.8)
40.9
(40.2)
0.8
• (0.5)
0.1
(0.1)
(continued)
47
-------
Table 24. Continued
Mean concentration (ng/m3)
Analyte
Heptachlor
Hexacntoro-
benzone
Sample size
Age of
Housing Unit -
(years)
<20 Mean
s.e.
21-40 Mean
s.e.
a41 Mean
s.e.
<20 Mean
s.e.
21-40 Mean
s.e.
a4l Mean
s.e.
£20
21-40
a4i
Summer
46.4
(23.0)
237.0
(99.2)
141.4
(40.4)
0.8
(0.2)
1.2
(0.6)
2.7
(1-5)
19*
32a
11"
Jacksonville
Spring
18.4
(15.0)
171.6
(106.2)
229.5
(77.2)
0
-
0.8
(0.4)
0
-
13
37
21
Springfield/Chicopee
Winter
13.8
(2.9)
38.3
(21.8)
184.6
(34.4)
0.0
(0.0)
0.1
(0.1)
0.8
(0.4)
15
37
19
Spring
3.7
(0.8)
64.6
(46.8)
8.9
(5.1)
0
-
0
-
0
-
13
16
19
Winter
1.5
(0.2)
5.1
(1.2)
3.7
(1.8)
0.3
(0.2)
0
-
0.1
(0.2)
9
22
20
»The sample sizes for bendiocarb were 14 "<20," 26 "21-40," and 9 ">4i."
Sampler Location. The possibility that the indoor air
concentration measurements were related to the room
in which the fixed-site sampler was set up was
examined by computing analyte mean concentrations
for each room type. The only consistent differences
observed were for bendiocarb, diazinon, and
malathion, which all had lower mean concentrations in
the kitchen, and propoxur, which had higher kitchen
concentrations in Jacksonville and lower kitchen
concentrations in Springfield/Chicopee (Table 26).
Potential Uses of Specific Analytes
Pesticide Inventory. A possible surrogate measure for
actual use of an analyte in a home is whether or not
pesticides containing the analyte are present in or
around the home. This measure might in turn be
expected to be correlated with analyte air
concentrations. The MOPES pesticide inventory data
were used to explore this possibility.
Active ingredients were identified for all pesticides
reported with valid EPA registration numbers in the
household pesticide inventories. For each MOPES
analyte, respondents were categorized by whether or
not the analyte was present in any of the pesticides in
the household inventory. Mean indoor air
concentrations were then computed for the two
categories. The results for all detected analytes
frequently present in the inventories --are shown in
Table 27.
Although none of the analytes had a completely
consistent pattern of concentration differences, ortho-
phanylphenol, chlordane (in Springfield/Chicopee), and
dichlorvos (in Jacksonville) showed some evidence of
having higher air concentrations when present in the
household inventories. Carbaryl also displayed this
pattern, but was rarely detected, despite being
relatively common in the inventories. The low volatility
of carbaryl probably accounts for its infrequent
detection. The relationship between mean
concentration and presence in inventories was
variable for chlorpyrifos, diazinon, malathion, and
propoxur, perhaps because of seasonal differences in
use. For folpet, the observed relationship was the
inverse of that expected; that is, indoor air means
were consistently lower in households with folpet in
their inventories. However, not much significance
should be attributed to this finding given the rare
occurrence of folpet in the inventories.
4
Termiticide Applications. All respondents were asked
in both the screening questionnaire and the study
questionnaire about the history of termiticide
applications in their home. Their answers were used
to classify the monitored homes as having or not
having been treated with termiticides. Units in which
the respondents did not know the termiticide history
were included in the untreated category. Mean indoor
air concentrations of analytes often used as
termiticides were then computed for the two
categories (Table 28).
Chlordane and heptachlor always displayed the
expected pattern of higher concentrations in treated
homes. Heptachlor epoxide, a degradation product of
heptachlor, was detected infrequently and was not
consistently higher in treated units.
In Jacksonville, aldrin and, to a lesser extent, dieldrin
concentrations were higher in treated homes.
However, in Springfield/ Chicopee, where both
analytes were detected infrequently, their concen-
trations tended to be higher in homes in the untreated
48
-------
Table 25. Indoor Air Concentration vs. Type of Housing Unit
Mean concentration (ng/m3)
Analyte
Aldrin
alpha-BHC
Chlordane
4,4'-DDE
4,4'-DDT
Dieldrin
Type of Housing
Unattached
single unit
Attached single
unit (e.g., duplex)
Multiunit
(e.g., apartment)
Mobile home
Unattached
single unit
Attached single
unit (e.g., duplex)
Multiunit
(e.g., apartment)
Mobile home
Unattached
single unit
Attached single
unit (e.g., duplex)
Multiunit
(e.g., apartment)
Mobile home
Unattached
single unit
Attached single
unit (e.g., duplex)
Multiunit
(e.g., apartment)
Mobile home
Unattached
single unit
Attached single
unit (e.g., duplex)
Multiunit
(e.g., apartment)
Mobile home
Unattached
single unit
Attached, single
unit (e.g., duplex)
Multiunit
(e.g., apartment)
Mobile home
Unit
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Summer
53.6-
(25.6)
0
_
0
_
0
-
1.9
(0.7)
0.6
(0.2) .
0.1
(0.1)
0
-
444.9
(117.6)
405.2
(128.5)
176.3
(56.5)
2.9
(2:4)
_
-
_
-,
-
_
20.0
(4.2)
13.0
(9.5)
8.2
(3.4)
2.2
(1.4)
Jacksonville
Spring
9.3
(3.7)
0
0
- =
2.4
0-7)
1.4
(0.4)
0
1.0
(0.6)
0
- -
308.8
(46.7)
779.0
(58.9)
56.5
(47.3)
0.
-
0.4
(0.1)
0
1.6
(1.5)
0
-
1.1 :
(0.7)
0.3
(0.5)
1.4
(1-3)
0
-
10.4
(3.0)
30.3
(2-3)
1.1
(1-0)
0
-
Springfield/Chicopee
Winter
8.4
(1-3)
0
6.0
6.4
.0
-
1.6
(0.9)
2.5
0.2
(0.1)
0
-
295.3
(100.2)
417.0
77.2
(11.0)
14.8
(9.8)
0.3
(0.2)
0
0
0 '
-
0.9
(0.5)
0
0
0
-
8.8
(2.0)
18.0
4.3
(1.7)
0
-
Spring
0
0
•• o
0
--
0.2
(0^2)
Q
o
0
293.2
(250.2)
24.8
(2.1)
5.5
(6.9)
41.0
- •
1.3
(0.9)
0
o
0
0.0
(0.0)
0
o
0
- -
1.5
1.7
0
o
0
Winter
Oe
.3
(0.4)
On
.u
(0.0)
•
0 ,
-
0
36.8
(4.7)
40.1
(18.4)
22.4
(2.6)
206.0
'
(0.6)
0"
1
0.0
(0.0)
0
0.7
(0.6)
0.6
(0.9)
0
R n
. o.u
(4.8)
1 7
(1-2)
2.2
0
(continued)
category. Possible explanations for the unexpected
Springfield/Chicopee finding include sampling error,
misclassification of the housing units, or use of aldrin
and dieldrin for purposes other than termite control.
Mean chlorpyrifos concentrations were usually similar
in the treated and untreated categories. This was
probably due to the wide variety of uses of
chlorpyrifos and its relatively limited use as a
termiticide until recently.
Use of Other Household Insecticides. To examine the
relationships between air concentrations and reported
treatment for insects other than termites, homes were
49
-------
Table 25. Continued
Mean concentration (ng/m3)
Analyte
Heptachlor
Sample size
Type of Housing
Unattached
single unit
Attached single
unit (e.g., duplex)
Multiunit
(e.g., apartment)
Mobile home
Unattached single unit
Attached single unit
Multiunit
Mobile home
Unit
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Summer
217.2
(76.7)
236.2
(93.1)
99.7
(78.7)
1.2
(1.0)
40
3
11
8
Jacksonville
. Spring
200.8
(69.8)
470.6 .
(18.0)
6.8
(3.1)
0.9
(0.8)
57
2
6
7
Springfield/Chicopee
Winter
102.8
(35.5)
156.0
-
9.1
(3.0)
1.3
(0.9)
53
1
12
5
Spring
45.9
(34.8)
10.7
(2.5)
0.5
(0.7)
3.9
-
35
6
7
1
Winter
4.1
(0.6)
4.8
(2.5)
1.2
(0.3)
35.0
-
36
5
9
1
Table 26. Indoor Air Fixed-Site Sampler Location Comparison
Jacksonville
Springfield/Chicopee
Analyte
Sample Location
Summer
Spring
Winter
Spring
Winter
Bendiocarb
Diazinon
Malathion
Propoxur
Sample size
Kitchen
Family room
Other
Kitchen
Family room
Other
Kitchen
Family room
Other
Kitchen
Family room
Other
Kitchen
Family room
Other
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
15.3
(14.0)
122.0
(89.5)
0
-
376.7
(221.2)
519.7
(89.0)
35.0
(16.9)
2.9
(2.0)
29.9
(20.3)
0
-
1541.7
(1075.1)
335.9
(42.5)
227.5
(16.3)
15a
41 a
48
5.8
(4.0)
7.0
(3.5)
0
-
51.2
(8.5)
138.7
(33.8)
72.4
(40.3)
0
-
11.2
(5.6)
3.7
(2.8)
455.6
(296.2)
188.2
(68.6)
146.8
(108.6)
12
42
13
0
-
3.8
(1.2)
3.2
(3.4)
31.0
(14.2)
90.7
(21.6)
159.9
(36.6)
0
-
23.5
(15.4)
0
-
257.2
(69.8)
154.6
(63.8)
66.2
(17.7)
6
61
3
0
- -
0.6
(0.8)
0
-
55.1
(77.2)
118.0
(141.2)
8.2
(6.3)
0
-
17.4
(21.1)
0
-
15.1
(9.3)
21.8
(21.1)
0
-
12
16
2
0
-
0.5
(0.5)
0
-
0.7
(0.4)
3.3
(1.8)
0
-
0
-
0
-
0
.
7.4
(4.0)
21.3
(6.5)
10.0
(4.7)
10
32
5
aThe sample sizes for bendiocarb, malathion, and propoxur were 9 "Kitchen," 35 "Family room," and 3 "Other."
categorized on the basis of responses to screening
questions on the subject. Units in which respondents
indicated that insecticides were applied at least once
a year were classified as treated; all others were
assigned to the untreated category. Table 29 presents
the mean indoor air concentrations in the two
categories for the comrnonly used or detected
insecticides.
50
-------
Table 27. Indoor Air Concentrations vs. Presence in Household Pesticide Inventory
Mean concentration (ng/m3)
Analyte
Carbaryl
Chlordane
Chlorpyrifos
Diazinon
Dichlorvos
Folpet
Jacksonville
Present in Inventory
Yes Mean
s.e.
n
No Mean
s.e.
n
Yes Mean
s.e.
n
No Mean
s.e.
n
Yes Mean
s.e.
n
No Mean
s.e
n
Yes Mean
s.e.
n
No Mean
s.e.
n
Yes Mean
s.e.
n
No Mean
s.e.
n
Yes Mean
s.e.
n
No Mean
s.e.
n
Summer
330.0
(346.3)
g
4.3
(2.6)
40
NRb
438.1
(60.3)
20
338.3
(88.2)
42
1034.3
(594.6)
8
370.5
(97.2)
54
200.8
(85.4)
20
80.9
(58.6)
29
0
4
0.5
(0.5)
45
Spring
NDa ,
9-
78.0
0
1
246.6
(46.0)
68
212.4
(55.6)
18
210.5
(55.3)
51
76.0
(36.5)
10
115.7
(26.6)
59
191.3
(112.9)
25
52.2
(42.3)
44
0
1
0.7
(0.4)
68
Winter
ND
324.2
(48.0)
3
217.4
(88.8)
68
118.0
(22.1)
22
121.3
(26.3)
49
175.9
(102.1)
18
61.5
14.4)
53
46.4
(18.1)
32
8.3
(8.6)
39
0
7
0.6
- (0.5)
64
Springfield/Chicopee
Spring
0.9
(1.1)
17
0
32
683.5
(506.8)
5
30.9
(11.6)
44
35.6
(25.4)
•6
9.1
5.8
43
90.4
47.6
13
40.9
38.4
36
0
(4.9)
14
4.7
(4.9)
35
0
7
0.8
(0.7)
42
Winter
ND
51.1
(7.1)
8
31.3
(6.0)
43
7 fl
/ .0
(6.5)
7
4.4
(0-7)
. 44
8
P Q
^.9
(1.9)
43
n Q
\Jr3
(2.0)
16
1.7
(2-0)
35
ND
(continued)
Only chlorpyrifos was routinely at higher
concentrations in treated homes than in untreated
homes. Chlordane, diazinon, gamma-BHC, and
carbaryl (when detected) were higher in treated units
in Jacksonville, but the pattern did not hold up in
Springfield/Chicopee. In contrast, bendiocarb and
propoxur exhibited the expected pattern (i.e., treated
higher than untreated) in Springfield/Chicopee, but not
in Jacksonville. Dichlorvos and malathion displayed no
evidence of a consistent pattern of variation.
Use of Pesticides on Pets. The household pesticide
inventory data and the dermal exposure component of
NOPES made it clear that some of the pesticides
more commonly used by respondents were pet
pesticides. The screening questionnaire responses
were used to categorize households according to
whether or not they reported any use of pet
pesticides. Mean indoor air concentrations were
calculated for seven analytes that are used in pet
products (Table 30).
51
-------
Table 27. Continued
Mean concentration (ng/m3)
Jacksonville
Springfield/Chicopee
Anatyte
Malathion
ortho-Phanyl-
phcnol
Present in Inventory
Yes Mean
s.e.
n
No Mean
s.e.
n
Yes Mean
s.e.
n
No Mean
s.e.
n
Summer
2.6
(2.2)
7
23.7
(16.5)
42
84.8
(29.7)
9
99.3
(24.2)
40
Spring
1*3.9
(8.6)
8
15.8
(10.8)
61
178,7
(47.4)
17
43.1
(9.9)
52
Winter
66.9
(57.3)
12
15.3
(13.0)
59
135.3
(50.8)
24
24.4
(7.8)
47
Spring
24.0
(11.7)
12
0
37
67.6
(24.0)
18
32.7
(7.1)
31
Winter
ND
50.1
(15.2)
12
16.3 '
(8.1)
39
Propoxur
Yes Mean
s.e.
n
No Mean
s.e.
n
854.8
(434.0)
22
215.0
(26.9)
27
191.3
(99.6)
26
229.7
(138.9)
43
135.4
(45.8)
32
182.4
(83.8)
39
25.5
(5.8)
20
27.1
(17.7)
29
17.1
(8.3)
21
17.0
(8.7)
30
BND «
Not detected in indoor air, or estimated to be detectable in less than 1% of the population.
Not reported in any household pesticide inventories.
Classifying households according to reported pet
pesticide use did not result in consistent patterns of
variation in mean air concentrations for any of the
analytes. One possible explanation for this result is
that the analytes also had non-pet uses that
overshadowed the effect of the pet applications. An
alternative or contributing explanation could be that
many of the pet pesticide products did not contain any
of the NOPES target compounds. This possibility is
borne out by the household inventory data. If the
NOPES analytes were not very common in the pet
pesticide products, the predictive power of the general
screening questions on pet pesticide use has to be
limited.
Summary. In summary, the exploratory analyses
indicated the following:
• Termiticide air concentrations were related to
reported termiticide treatment history, type of
housing unit, and age of housing unit.
• Indoor air concentrations of older pesticides that
are now banned or much less frequently used were
related to the age of the housing unit.
• For some pesticides, presence of the pesticide in
the household pesticide inventory was associated
with higher indoor air concentrations.
• Very general information on indoor insecticide use
was related to indoor air concentrations for a few
pesticides.
Future Analyses. The exploratory analyses performed
to date indicate that further investigation of the
relationships between air concentrations and
questionnaire data is warranted. The next generation
of analyses should include:
• Review of reported pesticide use during or
immediately preceding the monitoring period to
assess the impact on air concentrations.
• Further analysis of termiticide and household
insecticide concentrations using the detailed
questionnaire information on frequency of
application and type of applicator (e.g., professional
service or householder).
• Assessment of the degree to which the
assumptions underlying standard .statistical
52
-------
Table 28. Indoor Air Termiticide Concentrations vs. Reported Termiticide Use
Mean concentration (ng/m3)
Analyte
Aldrin
Chlordane
Chlorpyrifos
Dieldrin
Heptachlor
Heptachlor epoxide
Sample sizes
Termiticide Use
Yes
Noa
Yes
Noa
Yes
Noa
Yes
Noa
Yes
Noa
Yes
Noa
Yes
Noa
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Summer
81.5
(42.5)
0.6
(0.4)
473.1
(87.6)
232.6
(114.2)
343.1
: (63.8)
381.0
(70.8)
22.5
(4.0)
9.9
(2.3)
216.3
(48.2)
130.9
(73.9)
0.6
(0.4)
0.5
(0.3)
31
31
Jacksonville
Spring
11.7
(4.3)
0.5
(0.4)
309.7
(22.4)
162.3
(118.0)
190.9
(38.1)
224.1
53.7
8.0
(2.0)
8.6
(4.2)
184.8
(75.3)
113.9
(82.4)
1.5
(1.1)
0
-
50
22
Springfield/Chicopee
Winter
8.9
(1.8)
5.8
(0.7)
248.8
(52.3)
204.5
(112.3)
117.0
(29.8)
122.2
(22.4)
10.0
(2.2)
5.6
(1.8)
83.8
(23.4)
65.8
(37.6)
0.4
(0.3)
1.0
(0.5)
31
40
Spring
0
0
470.5
(329.9)
20.2
(10.5)
19.2
(11.7)
3.6
(3.2)
0
1.7
(1.8)
75.1
(44.2)
2.3
(0.8)
0
0
33
16
Winter
0.1
0.0
0.4
(0.3)
54.6
(7.1)
26.0
(5.9)
4.0
(1-9) '
5.5
(2.3)
1.8
(0.7)
5.2
(4.1)
8.0
(1.9)
1.7
(0.6)
0
o
21
30
alncludes households that did not provide information on termiticide use.
inference procedures are met by untransformed
and transformed concentration data.
• Multivariate analyses to evaluate the explanatory
power of more than one variable at a time, and to
look for confounding effects or collinearities.
After these analyses are completed, quantitative
regression models may be developed from the
NOPES data for some analytes. The outlook is
especially promising for the termiticides and some of
the discontinued compounds, and it is hopeful for
most of the other commonly detected analytes.
Potential Health Effects
The NOPES health effects evaluation concentrates on
the air exposure route because air was the major
focus of the NOPES sampling effort. The limited
NOPES water sampling indicated that exposure from
water ingestion in the two study areas appeared to be
low for the pesticides studied. The dermal exposure
data collected in NOPES were insufficient to support
any comprehensive conclusions. The air data from the
personal air samplers were used to estimate
exposure, since these were considered most
representative (if the air data from the general indoor
air measurements had been used, very similar results
would have been obtained).
It is not surprising that so many pesticides were
detected in the indoor residential environment
especially considering that the study was targeted
toward commonly used household pesticides and
employed sensitive analytical techniques (as low as
10-10 gm/m3). The presence of these compounds is
not necessarily synonymous with the advent of health
effects. The following evaluation will describe the
health risk implications of the NOPES air monitorinq
data.
Estimation Procedures
The health risk estimates were derived using average
personal air concentrations. The seasonal daily mean
air concentrations in Tables 12 and 13 were averaged
using a seasonal weighting that provides an
approximation of the annual average of daily mean
53
-------
Table 29. Indoor Air Concentrations vs. Indoor Household Insecticide Use
Mean concentration (ng/m3)
Analyte
Bcndiocarb
Carbaryl
Chtordane
Chlorpyrifos
Oiazinon
Dichlotvos
gamma-BHC
Malalhion
Propoxur
Sample sizes
Household —
Insecticides Used
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Summer
72.8
(52.0)
0
-
58.1
(54.6)
0
-
340.6
(93.9)
21.4
(16.5)
382.1
(66.4)
85.0
(54.7)
441.4
(131.3)
42.7
(10.5)
111.8
(55.2)
455.3
(350.7)
21.3
(9.7)
0
18.3
(12.7)
0
-
309.3
(35.9)
3616.7
(2770.7)
60a
2a
Jacksonville
Spring
5.4
(2.4)
5.9
(6.1)
0.6
(0.4)
0
-
269.0
(64.1)
162.7
(64.5)
208.4
(43.8)
194.8
(97.3)
133.1
(21.4)
24.6
(10.7)
110.6
(68.2)
0
-
15.8
(7.3)
4.8
(2.0)
5.6
(1.7)
47.8
(29.9)
189.0
(46.9)
339.8
(303.3)
64
8
Springfieid/Chicopee
Winter
3.2
(1.0)
4.3
(3.0)
0
-
0
-
243.5
(99.3)
33.3
(9.3)
125.3
(16.7)
80.1
(40.3)
94.7
(19.0)
13.8
(3.9)
27.6
(8.7)
0
-
6.6
(2.4)
1.0
(0.7)
22.9
(14.7)
0
-
157.9
(62.4)
200.0
(118.6)
66
5
Spring
0.6
(0.5)
0
-
0
-
0.4
(0.4)
40.5
(13.5)
271.3
(250.2)
25.8
(11.0)
2.5
(2.3)
154.5
(108.0)
0.4
(0.4)
13.9
(14.6)
0
-
0
0.7
(0.7)
15.9
(16.6)
0
-
39.1
(7.9)
21.1
(17.8)
38
11
Winter
2.0
(2.2)
0
'
0
~
0
-
33.2
(12.4)
33.8
(6.8)
15.5
(6.8)
2.1
(0.8)
2.2
(0.9)
2.6
(2.0)
1.1
(1.0)
1.6
(2.0)
0.2
(0.1)
12.2
(15.1)
0
- -
0
"
21.6
(1.1.0)
15.7
(6.7)
33
18
"The sample sizes for bendiocarb, carbaryl, dichlorvos, malathion, and propoxur were 47 "Yes" and 2 "No."
concentrations. The annual average daily air
concentration (Ca) was estimated for Jacksonville as
Ca = (Summer + 2 * Spring + Winter) / 4
and for Springfieid/Chicopee as .
Ca = (3 * Spring + Winter) / 4.
The annual average may be underestimated for
Springfieid/Chicopee because this site was not
monitored in the summer season, which generally had
the highest concentrations in Jacksonville.
Where pesticides were never found above the
detection limit, an upper bound risk was calculated
using the maximum detection limit encountered in this
study. In none of these cases did the cancer risks
exceed 3 x 10'6 or the hazard index exceed 1 x 10-1.
This demonstrates that even if these pesticides were
present at just below the detection limit, they would
54
-------
Table 30. Indoor Air Concentrations vs. Pesticide Use on Pets
Mean concentration (ng/m3)
Analyte
Carbaryl
Chlorpyrifos
Diazinon
Dichlorvos
Malathion
Methoxychlor
Propoxur
Sample sizes
Pesticides Used
on Pets
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
No
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Mean
s.e.
Summer
6.0
(5.6)
83.8
(85.9)
421.6
(78.3)
335.8
(89.6)
' 663.2
(290.2)
284.7
(32.6)
125.3
(70.6)
140.1
(78.9)
45.8
(30.8)
2.1
(0.7)
0.4
(0.2)
0.1
(0.1)
267.3
(45.8)
686.0
(345.7)
33a
29a
Jacksonville
Spring
1.0
(0.7)
0
-
283.1
(39.1)
146.2
(49.2)
92.0
(28.0)
122.2
(31.9)
86.9
(62.3)
85.7
(69.3)
4.0
(2.0)
23.3
(16.6)
0.7
(0.6)
0
-
168.5
(72.5)
263.2
(115.5)
51
21
Springfield/Chicopee
Winter
0
_
0
-
128.1
(29.1)
115.3
(23.7)
80.1
(27.1)
89.4
(34.1)
49.8
(25.4)
8.2
(5.4)
18.5
(16.5)
21.7
(19.2)
0.1
(0.1)
0.2
(0.2)
136.3
(44.0)
179.4
(93.7)
40
31
Spring
0.9
(1.0)
0
8.0
(3.2)
10.6
(8.1)
42.8
(33.6)
51.0
(47.2)
13.8
(15.1)
0
0
_
7.2
(7.2)
0
0
-
29.6
(8.2)
25.4
(18.9)
32
17
Winter
0
0
5.1
(3.3)
5.1
(1.6)
0.4
(0.2)
4.5
(4.1)
2.9
(2.3)
0.0
(0.0)
0
0
0
0
14.2
(4.7)
19.8
(7.0)
36
15
aThe sample sizes for propoxur, dichlorvos, and malathion were 27 "yes" and 22 "No."
have associated risks in a range that the Agency
generally considers negligible.
The administered dose, as calculated using the
"lifetime average daily exposure" or LADE (mg/kg-
day) from inhalation was estimated as:
LADE = Ca * IR / BW
where
Ca = air concentration (ng/m3) '* (1 x 10-6
mg/ng),
IR = inhalation rate (20 m3/day), and
BW = body weight (70 kg).
The individual excess lifetime cancer risk or ELCR
was calculated as:
ELCR = LADE*q1*
where q1* is the potency-slope factor (mg/kg-day)-i
for the analyte. The noncancer risk was calculated as
a hazard index defined as the ratio of the LADE
divided by the Reference Dose or RfD (mg/kg-day).
Discussion of Results
Tables 31 a, 31 b, 32a and 32b summarize the risk
estimates due to inhalation of pesticide vapors in
nonoccupational settings in Jacksonville and
Springfield/Chicopee. Table 31 a presents the risks for
pesticides other than the cyclodiene termiticides in
Jacksonville and Table 31 b presents the risks for the
55
-------
cyclodiene termiticides (chlordane, heptachlor, aldrin,
and dieldrin) in Jacksonville. Similarly, Table 32a
presents the risks for the pesticides other than the
cyclodiene termiticides in Springfield/Chicopee and
Table 32b presents the risks for the cyclodiene
termiticides in Springfield/Chicopee. The cyclodiene
termiticides were separated from the other pesticides
due to the fact that their registrations have been
cancelled, suspended or withdrawn (Velsicol
voluntarily withdrew chlordane and heptachlor in an
August 1987 agreement with the Agency).
These estimates were derived assuming an inhalation
rate of 20 m3/day and 70 years exposure at the-
average concentrations. They should be interpreted as
general indications of risk levels, not precise values
because of major sources of uncertainties. To better
understand these risk estimates, one should consider
the key assumptions and associated uncertainties.
Where available, the risk estimates were based on the
cancer potencies and reference doses presented in
the Integrated Risk Information System (IRIS). All
values in IRIS have been rigorously reviewed and
officially accepted by the Agency. Unfortunately 13 of
the 32 pesticides studied are not currently included in
IRIS. For 12 of these 13 pesticides, • the cancer
potencies and reference doses were based on
assessments from EPA's Office of Pesticide Programs.
(OPP). The OPP values were derived from their
review of the open literature and proprietary data from
pesticide manufacturers. No risk estimate could be
made for oxychlordane because toxicity values were
not available from IRIS or OPP.
The cancer risk estimates indicate that the four
pesticides which present the highest risks in both
areas were the cyclodiene termiticides (chlordane,
heptachlor, aldrin, and dieldrin), which have either
been cancelled, suspended or voluntarily withdrawn.
None of the other pesticides were estimated to
present cancer risks exceeding 2 x 10-6 in either area.
The Agency generally considers risks less than about
10-6 as negligible.
The estimated hazard indices for noncancer risks
were less than one for all pesticides in both areas.
However, several approached one including
chlordane, aldrin and diazinon. The RfD is a peer-
reviewed estimate of the time-weighted average daily
lifetime exposure that is likely to occur, without
appreciable risk of deleterious effects. The Agency
generally considers hazard indices less than one as
low.
Discussion of Uncertainty
The key assumption in this assessment is that the
concentration averages represent the true average of
levels to which a person is exposed over a 70 year
lifetime. The seasonal weighting scheme results in an
annual average that at least partially accounts for
seasonal changes in pesticide use. However,
averages of 24-hour samples collected during two or
three seasons in one year is an uncertain basis for
deriving an average representative of a 70 year
period. Such short term surveys cannot account for all
changes that occur over time. For example, the
introduction of new pesticides or registration changes
could affect current residential practices and resulting
exposure levels. The "highest risks were associated
with four cyclodiene termiticides (chlordane,
heptachlor, aldrin and dieldrin) that have been
cancelled, suspended or withdrawn. Although these
chemicals are highly persistent, slow degradation or
dilution (due to physical processes such as leaching
or diffusion) will occur Over 70 years resulting in some
reduction in exposures. No reliable data could be
found on the degradation rates that may occur for
these termiticides when applied in and around
foundations. The possible reductions in risks
corresponding to a range of half-life assumptions are
shown below (assuming degradation does occur and
proceeds according to first order kinetics):
Half-Life (years)
2
4
10
20
30
50
Factor by Which 70 Year
Risk is Reduced
25
12
4.8
2.7
2.0
1.6
Moves to other homes (within the target areas) or
differences in personal pesticide use habits introduce
variability. This variability was accounted for by
sampling a large number of homes in each city ~ 173
in Jacksonville and 86 in Springfield/Chicopee.
Because the NOPES surveys were limited to
Jacksonville, Florida and Springfield/Chicopee,
Massachusetts, the results cannot be directly
extrapolated to other areas. Pesticide use and home
ventilation systems vary across the country and are
likely to lead to different exposure levels. Considering
the widespread use of pesticides and the frequency
with which they were detected, it does suggest some
exposure will occur in other areas.
Inhalation rate's vary with body size and activity level.
The assumed 20 m3/day is a widely accepted average
for the adult population and is probably not an
important source of uncertainty.
56
-------
•
Table 31 a. Weighted Estimate of Annual Average Daily Concentrations, Cancer Risk and Hazard Index for Jacksonville Air
(Pesticides other than Cycfodiene Termiticides)
Analyte
Dichlorvos
alpha-BHC
Hexachlorobenzene
gamma-BHC
Chlorothalonil
Ronnel*
Chloropyrifos
Dacthal
Captan
Folpet
2,4-D esterd
Methoxychlor
Dicofol'
cis-Permethrin
trans-Permethrin
4,4'-DDTe
4,4'-DDD'
4,4'-DDEe
ortho-Phenylphenol
Propoxur
Bendiocarb
Atrazine
Diazinon
Carbaryl
Malathion :
Resmethrin
Annual Avg.
Daily Concen.
(ng/cu.m.)
62.4
0.8
0.5
11.2
0.8
<2.5
191.1
0.2
0.1
0.5
1.1
0.3
<33
0.9
0.3
. 0.5
<3.1
0.6
57.7
185.2
15.9,
0.1
159
7.5
' 11.6
0.1
Slope
Factor
(kg-day/mg)
6.3a
0.011&
0.0023b
0.00353
0.019"
0.34b
0.022b
0.022b
0.34a
0.34t>
0.34b
0.0016b
0.0079b
0.22b
Excess
Lifetime
Cancer Risk
2E-06
3E-09
7E-1 1
5E-10
6E-09
<3E-06f
6E-09
2E-09
5E-08
<3E-07'
6E-08
3E-08
4E-07
6E-09
Reference
Dose
(mg/kg-day)
0.0008°
0.0008C
0.0003a
0.015°
0.015°
0.0033
0.5a
0.1 3=
0.1°
0.01a
,0.05°
0.001 °
0.0053
0.05°
O.OOOS3
0.004a
0.005°
0.0053
0.00009°
0.13,
0.02a
0.033
Hazard
Index
2E-02
2E-04
1E-02
2E-05
<5E-05f
2E-02
1E-07
2E-07
1E-06
2E-06
2E-06
<1E-02f
5E-05
2E-06
3E-04
1E-02
9E-04
6E-06
5E-01
2E-05
2E-04
1E-06
aSource: Integrated Risk Information System (IRIS) '
bSource: Memorandum from Reto Engler to Health Effects Division Branch Chiefs and Selected OPP Division Directors US EPA
October 27, 1989. ' . . ' '
°Source: Reference Dose Tracking Report, Health Effects Division, Office of Pesticides, US EPA, October 12 1989
dMethyl ester in summer, butoxyethyl ester in' spring and winter.
^Concentration calculated as (3 * spring + winter)/4 because this analyte was not measured in the summer season •
'Not found above detection limit. Concentration listed is the maximum detection limit encountered in this study. The correspondina
' risk estimate represents an upper limit. ' ; y
The chemical toxicity assumptions also play a major
role in risk estimation. These assumptions are
chemical-specific and normally derived from animal
studies which are outside the purview of this study.
However, they also introduce substantial uncertainty
and must be considered in interpreting the risks. The
major sources of uncertainty concern the validity of
extrapolating animal data to humans and extrapolating
high dose animal experiments to low dose
relationships.
The dose response relationships for cancer risks are
expressed as slope factor values which are estimated
as 95th percentile confidence limits using the
linearized multistage model. As such, they are
conservative estimates of the chemical's hazard or
potential to cause cancer. Risks estimated by
combining these slope factors with exposure
estimates are commonly referred to as upper bound
risks. It should be recognized, however, that the
exposure estimates used in this assessment are
believed to represent average conditions. Accordingly
the risk estimates resulting from a combination of an
upper bound slope factor and average exposure
estimates cannot be characterized as an upper bound
risk nor an average. Some individuals may be
exposed to concentrations that are higher than the
mean throughout their lifetime and have greater risk
than those presented in Tables 31 and 32.
57
-------
Tablo 31b. Weighted Estimate of Annual Average Daily Concentrations, Cancer Risk and Hazard Index for Jacksonville Air
(Cyclodiene Termiticldes)
Analyte
Heptachlor
AWrin
DtekJrin
Chlordane
Heptachlor Epoxido0
Oxychtordaneo.'
Annual Avg.
Daily Concen.
(ng/cu.m.)
115.2
26
6.4
197.1
0.4
<2
Slope
Factor
(kg-day/mg)
4.53
173
163
1.33
9.1a
Excess
Lifetime
Cancer Risk
2E-04C
1 E-04<=
3E-05C
7E-05<=
1E-06<=
Excess
Lifetime
Cancer Risk
6E-06"
5E-06d
lE-06d
3E-06d
4E-081
Reference
Dose
(mg/kg-day)
0.0005b
0.00003b
0.00005b
0.000063
0.00001 b
Hazard
Index
7E-02<=
3E-01C
4E-02<=
1E + OOC
1E-02C
Hazard
Index
3E-03d
1E-02d
lE-03d
4E-02d
5E-04d
*>Source: Integrated Risk Information System (IRIS)
^Source: Reference Dose Tracking Report, Health Effects Division, Office of Pesticides, US EPA, October 12,1989.
cTh9 risk estimates presented in this table assume that the concentrations remain constant over 70 years. Since all have been cancelled or
withdrawn, some reduction in risk will occur due to degradation. Although these degradation rates are not known, possible reductions based
on halflile assumptions are presented in the text
<* These risk estimates were computed assuming that the pesticide degrades with a 2 year half life. As explained in the text, no reliable
degradation data are available and these estimates are included as an example of the possible reductions in risk due to degradation.
*Theso pesticides are included in this table because they are breakdown products of the cyclodiene termiticides.
I Not found above detection limit Concentration listed is the maximum detection limit encountered in this study.
Another source of uncertainty concerns the
application of toxicity values derived from animal
experiments using oral administration and applying
them to human inhalation scenarios. For the
carcinogens, this uncertainty has been reduced by
using cancer slope factors that have been either
adjusted for application to inhalation exposures or
deemed applicable without adjustment. The reference
doses used for the noncarcinogens are all based on
ingestion and were simply assumed to apply equally
to inhalation. This assumption can be a significant
source of uncertainty for these compounds, due to
differences in the absorption between the routes and
possibility of direct effects at the point of entry.
The risk estimates account for the air pathway only.
Additional exposure will occur as a result of ingestion
and dermal contact. Contamination of food and water
results from residues from agricultural practices or
contact with pesticides used in the home. Hand to
mouth activity can also cause ingestion exposure at
homes, especially among young children. These
contributions could not be quantified.
Finally, although the risks were presented as average
values, it should be understood that they will vary
significantly across the population. The measured
concentrations varied substantially; lifestyle, personal
pesticide use, etc. will all contribute to the variability.
Much more data are needed to estimate exposure
levels other than the mean, especially data near the
tails of the distribution.
Summary
In summary, this assessment provides a reasonable
indication of the possible risks due to inhalation of
pesticide vapors in nonoccupational settings, but has
some important limitations and uncertainties. The
most important limitations are the consideration of
only the air pathway and evaluation of the risks only
under average exposure conditions. The major
sources of uncertainty are the assumption that the
estimated air concentrations represent true averages
for lifetime exposures and the validity of the toxicity
standards. Bearing these points in mind, the
assessment showed that the noncancer risks were
generally low, and the cancer risks were in a range
the Agency generally considers negligible with the
possible exception of heptachlor and aldrin in
Jacksonville. The estimated risks for these two
compounds were on the order of 10'6 to 10-4,
depending on degradation. As noted above, the
registration of both of these compounds has been
cancelled, suspended, or withdrawn; and, although
they are very persistent, some degradation will occur
over time.
An earlier assessment of the risks posed by the
cyclodiene termiticides concluded that each
individual's chances of developing symptoms are low
and because of the large numbers of people exposed
to the cyclodienes the aggregate risk is a real one for
the U. S. population as a whole (US EPA, 1988,
Termiticides - Consumer Information, OPA-87-014).
This document also describes techniques
58
-------
Table 32a. Weighted Estimate of Annual Average Daily Concentrations, Cancer Risk and HazaYd Index for
Spnngfield/Chicopee Air (Pesticides other than Cyclodiene Termiticides)
Analyte
Dichlorvos
alpha-BHC
Hexachlorobenzene*
gamma-BHC
Chlorothalonil
Ronnel
Chloropyrifos
Dacthal
Captan
Folpet
2,4-D esterd.'
Methoxychlor'
Dicofol
cis-Permethrin'
trans-Permethrin1
4,4'-DDTe
4,4'-DDD'
4,4'-DDEe
ortho-Phenylphenol
Propoxur
Bendiocarb
Atrazine*
Diazinon
Carbaryl
Malathion
Resmethrin'
Annual Avg.
Daily Concen.
(ng/cu.m.)
3.3
0.2
<2.2
1.9
0.6
0.1
7.1
2
0.1
0.5
<30
<7.8
5.3
<53
<38
0.9
<5.3
3.8
39.4
15
0.3
<45
7.9
0.1
0.4
<25
Slope
Factor
(kg-day/mg)
6.3a
0.011"
0.0023"
0.0035a
0.019"
0.34"
0.022"
0.022"
0.34a
0.34"
0.34"
0.0016"
0.0079"
0.22"
Excess
Lifetime
Cancer Risk
4E-07
2E-09
7E-11
5E-10
<2E-07*
<3E-07'
<2E-07«
9E-08
<5E-07'
4E-07
2E-08
3E-08
<3E-06f
Reference
Dose
(mg/kg-day)
0.0008°
0.0008°
0.00033
0.015°
0.015°
0.003a
0.53
0.13°
0.1°
0.01 a
0.05°
0.001°
0.005a
0.05°
0.00053
0.0043
0.005°
0.005a
0.00009=
0.1a
0.02a
0.033
Hazard
Index
1E-03
<8E-04' •
2E-03
<1E-05
2E-06
7E-04
"
1 E-06
2E-07
1E-06
<9E-04f
<5E-05'
2E-03
<3E-03f
<2E-04'
5E-04
1E-03
2E-05
<3E-03f
3E-02
3E-07
6E-06
<2E-04f
aSource: Integrated Risk Information System (IRIS)
"Source: Memorandum from Reto Engler to Health Effects Division Branch Chiefs and Selected OPP Division Directors, US EPA,
October 27, 1989. '
°Source: Reference Dose Tracking Report, Health Effects Division, Office of Pesticides, US EPA, October 12, 1989
aMetnyl ester in summer, butoxyethyl ester in spring and winter.
^Concentration calculated as (3 * spring + winter)/4 because this analyte was not measured in the summer season.
'Not found above detection limit. The concentration shown is the highest detection limit encountered in this study and
corresponding risk is an upper bound.
homeowners can use to improve indoor air quality
such as increasing the air exchange rate, sealing
treated areas and installing outside air supplies to
appliances.
Follow-up studies are recommended to determine a
more comprehensive analysis of the risks. Research
is planned to develop guidance for conducting
exposure monitoring studies and associated
methodology for assessing human non-dietary
exposure to pesticides in a residential setting.
Reports to Participants
An individualized report of the NOPES findings will be
provided to each respondent who completed the
monitoring phase of the study. The intent of the report
is to inform participants of their measured analyte
concentrations and to discuss the significance of
themeasurements. In addition to presenting the
concentrations for the particular respondent and
housing unit, each report will present 'summary
statistics for each study area, so that the respondent
can assess his or her concentrations relative to those
59
-------
Table 32b.Weighted Estimate of Annual Average Daily Concentrations, Cancer Risk and Hazard Index for Springfield/Chicopee
Air (Cyclodiene Termiticides)
Annual Avg. Slope Excess Excess Reference
Daily Concen. Factor Lifetime Lifetime Dose Hazard Hazard
Analyte (ng/cu.m.) (kg-day/mg) Cancer Risk Cancer Risk (mg/kg-day) Index Index
Hoptachlor
Aldnn
Dicldnn
Chlordarte
Heplachlor Epoxtde6-'
Oxychtordane'.'
27.2
0.1
0.8
198.7
<3.3
<3.3
4.53 4E-05C,
173 5E-07=
16a 4E-06<=
1.3a 7E-05C
9.13 <1E-06<:
1E-06d
2E-08d
1E-07d
3E-06d
<4E-08d
0.0005"
0.00003b
0.00005b
0.000063
0.00001 b
2E-02=
1 E-03<=
5E-03C
1E + OQC
<1E-02<=
6E-04d
4E-05d
2E-04d
4E-02d
<5E-04d
aSource: Integrated Risk Information System (IRIS)
^Source: Reference Dose Tracking Report, Health Effects Division, Office of Pesticides, US EPA, October 12,1989.
cTho risk estimates presented in this table assume that the concentrations remain constant over 70 years. Since all have been cancelled,
suspended or withdrawn, some reduction in risk will occur due to degradation. Although these degradation rates are not known, possible
reductions based on halflife assumptions are presented in the text.
dTheso risk estimates were computed assuming that the pesticide degrades with a 2 year half life. As explained in the text, no reliable
degradation data are available and these estimates are included as an example of the possible reductions in risk due to degradation.
"Tlwse pesticides are included in this table because they are breakdown products of the cyclodiene termiticides.
'Not found above detection limit. Concentration listed is the maximum detection limit encountered in this study.
of the study area population. The reports will then
discuss the potential health implications of the
findings, and describe how participants can reduce
their exposure through proper use, storage, and
disposal of the target pesticides.
Development of the specific format and content of the
reports will follow standard EPA review procedures.
The participant reports will be prepared and distributed
soon after the release of this report.
Consumer Awareness
Although NOPES was not designed to provide an in-
depth look at consumer awareness about pesticides
and their safe use, the study yielded some anecdotal
information on the subject. Respondents' comments,
interviewers' observations, and questionnaire data
provide insight on how pesticides are used in
nonoccupational settings, and they indicate areas in
which exposure could be reduced by alternative
practices.
In general, respondents seemed to be using
appropriate pesticides given their pest problems,
although a few instances of questionable use were
observed. Label directions on mixing and applying the
pesticide were usually not read just before the
application, but were generally followed.
Of more concern from an exposure standpoint was
the lack of precautions taken by some respondents to
limit their exposure during or after pesticide
applications. Few respondents wore gloves, other than
those provided for the dermal sampling. Many did not
wash their hands or change clothes after an
application. Previous work (Lewis, 1988) suggests that
acute dermal exposure could be reduced through the
use of these precautions.
Air and chronic dermal exposure might be reduced by
decreasing the amount of pesticides stored in and
around the home. The pesticide inventory data
indicate that some respondents kept large inventories
of pesticides, some of which were rarely used. DDT
was found in a few homes, despite having been
banned for use by the general public for years. A few
respondents asked about how they could safely
dispose of unused pesticides, and indicated that they
had previously been unsuccessful at identifying safe
disposal methods. Making safe disposal methods
widely available and encouraging the sale of
pesticides in small amounts for home use could lead
to a desirable decrease in household pesticide
inventories.
Data Quality
Throughout NOPES, quajity assurance and quality
control activities were an integral part of data
collection and laboratory procedures. These activities
provided an ongoing review of field and laboratory
practices, and they permit assessment of the quality
of the NOPES data.
System and Performance Audits
System audits designed to review the overall
measurement process and evaluate its ability to yield
accurate data were performed several times over the
course of the study. EPA and EMSI conducted
external system audits, and SwRI performed internal
audits in Phases II and III. The early audits identified
several areas in both the field procedures and
laboratory protocols where corrective actions were
needed. For example, a recommendation was made in
the audit following Phase I to label each PDF
cartridge, in addition to its container, with a unique
identifier so as to improve sample tracking and reduce
the risk of sample misidentification. Such actions were
taken prior to the subsequent rounds of sample
collection and analysis. Few problems were noted in
the later audits..
6.0
-------
Field performance audits were conducted in each
study area each season to check the flow rates of
theair sampling pumps. In the majority of cases, the
difference between the audit standard and the pump
flow rate was less than 5%. Those few pumps with
flow rates that differed by more than 10% from the
audit standard were all checked and either
recalibrated or taken out of service.
Analytical Data Quality
A number of steps were taken to assess and quantify
analytical precision and accuracy. Laboratory and field
blanks were analyzed to check for contamination. An
octachloronaphthalene (OCN) spike was added to
each sample to evaluate the recovery efficiency of the
analytical system. Matrix spikes were run with each
extraction batch of samples to assess the accuracy of
the laboratory measurement process. Duplicate'
samples were collected and analyzed so that the
precision of the measurement process could be
quantified. To assess the laboratory component of
measurement error, some samples and standards
were analyzed by laboratories other than SwRI. The
findings of each of these activities are summarized
below.
Blanks. Laboratory: solvent blanks and laboratory
water blanks were analyzed in each extraction batch.
Only two instances of contamination were found, and
both involved very low levels of single analytes.
Air, water, and glove field blanks were also collected
and analyzed each season. The PUF cartridge used
as an air field blank was taken to the home at the end
of the sampling period, opened, and assembled as if
for use. No contamination was found in 28 of the 31
air field blanks. The sources of the low level of a
single contaminant in the other three blanks were
apparent. The propoxur and ortho-phenylphenol found
in two blanks represented about five percent of the
amount sampled in the indoor air. An air field blank
contained a low level of methoxychlor because its
storage jar broke during shipment, and the shipment
included a methoxychlor-laden glove sample. All five
water field blanks were clean. All the glove field
blanks, which were opened in the vicinity of the
pesticide application, contained low levels of one or
two contaminants. Because the field blanks were
contaminated only infrequently and at low levels that
were often attributable to a known causative factor, no
adjustment of the data for contamination or
background levels was performed.
OCN Recovery Efficiency. The OCN mean recoveries
for the matrix spike samples ranged from 86% for
Springfield/Chicopee in Phase II to 97% for
Jacksonville in Phase I, with coefficients of variation
(CVs) of 11% to 18%. Recoveries from gloves tended
to be lower than air and water sample recoveries.
Overall, 94% of Jacksonville samples (777/829)'and
93% of Springfield/Chicopee samples (413/444) had
OCN recoveries within the 75% to 125% advisory
limits specified in the NOPES Quality Assurance
Project Plan.
Matrix Spike Recovery. The mean, range, and
standard deviation of recoveries over extraction
batches for the matrix spikes are summarized ih Table
33. Coefficients of variation for the spike recoveries
ranged from 15% to 35% in the summer, 6% to 39%
in the spring, and 9% to 27% in the winter. Because
of coelution problems, the gamma-BHC and
chlorothalonil components of the spike mixture were
replaced with alpha-BHC and hexachlorobenzene
during the summer season analysis.
The matrix spike data indicate that the mean
recoveries were good for most analytes, but the range
was larger than desirable. Propoxur was the only
analyte with consistently low recoveries. Analyte
concentrations were not adjusted for recovery
efficiency because of the variations in matrix spike
recoveries were so dramatic.
Elevated and variable recoveries for heptachlor
prompted a change to using the DB-5 column for
quantification of this analyte in Phase III. Low matrix
spike recoveries for two summer season extraction
batches alerted SwRI to the problem caused by
switching to BoileezerR boiling chips, which resulted in
rectification of the problem before subsequent batches
were extracted. The inaccurate data resulting from
this problem were excluded from all statistical
analyses presented in this report.
Duplicates. Table 34 summarizes the percent relative
differences (defined in Table 34) observed for the
duplicate samples. This is an effective method of
expressing the pairwise deviation between duplicate
measurements when collectively summarizing many
different constituents where the deviations are a
function of level. However, caution must be used
when dealing with individual compounds on a
seasonal basis at different sites for both outdoor and
indoor locations. A close examination of the data
revealed that for most constituents the variability
between duplicates was not clearly a function of
concentration levels across all seasons and even
within a season.
Small sample sizes and low detection frequency of
many analytes prohibited exact quantification of
precision by using duplicates. The tabled results do,
however, indicate that the paired values were often
similar. Differences are believed mainly to be due to
field and laboratory measurement error, although the
possibility of some contribution from microspatial
variation cannot be discounted. Comparison of Table
34 and Table 19 confirms the earlier conclusion (see
61
-------
Table 33. Matrix Spike Percent Recoveries
Jacksonville
Analytea
Summer
Spring
Winter
Springfield/Chicopee
Spring
Chlorothalonilb
mean 62 - -
s.d. 19
range 38-89 -
n 7 - -
Chlorpyrifosb
mean 87 93 83 92
s.d. 15 6-11 7
range 69-113 81-105 50-108 82-m
n 8 24 32 21
Diazinon0
mean 70 .-. 75 79 73
s.d. 19 11 11 9
range • 31-92 52-88 48-104 60-96
n 8 24 32 21
Dieldrinb
mean 89 85 99 97
S.d. 24 13 16 15
range 46-124 75-101 70-138 85-155
n 8 24 32 21
gamma-BHCb
mean 108 -
s.d. 33
range 73-163 - - -
n 7 - - -
Heptachlorb
mean 107 117 83 103
S.d. 24 23 12 20
range 69-133 70-118 45-107 82-126
n 8 24 32 21
Hexachlorobenzeneb
mean 96 95 73 86
S.d. 13 5 10 7
range 86-111 91-109 44-91 73-107
n 3 24 32 21
Winter
alpha-BHCb
mean
s.d.
range
n
80
5
71 -87
24
80
16
33-113
32
79
7
73-98
21
78
9
65-96
22
88
8
76-105
22
70
7
56-84
22
96
10
84-116
22
88
9
74-102
22
73
7
60-86
22
Propoxur0
mean
s.d.
range
n
52
20
18-80
8
53
21
8-76
24
67
11
44-88
32
56
16
, 31-84
21
66
18
36-108
22
°Blank PUF plugs and gloves and split water samples were spiked with a solution containing the analytes
listed, which were selected to be representative of the different structural classes covered by the GC/ECD
and GC/MS analyses and to cover the chromatographic range.
bAnalyzed by GC/ECD.
<=Analyzed by GC/MS.
62
-------
Table 34. Duplicate Relative Percent Differences3
Number (and Percent) of Duplicate Pairs
Relative Percent Difference13
Jacksonville
Indoor air
Summer
Spring
Winter
Outdoor air '
Summer
Spring
Winter
Springfield/Chicopee
Indoor air
Spring
Winter
Outdoor air
Spring
Winter
<41%
26 (17%)
61 (19%)
64 (21 %)
8 ( 6%)
9 ( 3%)
9 ( 3%)
17 ( 6%)
17 (7%)
10 (4%)
1 ( 0%)
41 -67%
1 (1%)
7 (2%)
2 (1%)
0 (0%)
4(1%)
2 (1%)
1 (0%)
1 (1%)
4 (2%)
0 (0%)
>67%
2 (1%)
10 (3%)
2(1%)
1 (1%)
1 (0%)
2 (1%)
5 (2%)
1(1%)
0 (0%)
0 (0%)
Only Detected in
One Sample
10 (7%)
3(1%)
8 (3%)
9 (7%)
7 (2%)
8 (2%)
8 (3%)
11 (5%)
4 (2%)
2 (1%)
Not Detected in
Either Sample
115 (74%)
249 (75%)
221 (74%)
119(86%)
309 (94%)
276 (93%)
233 (89%)
200 (86%)
246 (92%)
228 (99%)
Totalc
154 (100%)
330 (100%)
297 (100%)
137 (100%)
330 (100%)
297 (100%)
264 (100%)
230 (100%)
264 (100%)
231 (100%)
"Relative percent difference, computed for pairs with detected values for both samples.calculated as 100" (primary value - duplicate
value|/(mean of the two values).
Relative differences of 40% or less indicate that the paired values differed by a factor of 1.5 or less whereas relative differences greater
than 67% indicate that paired values differ by a factor of two or more.
°Total for up to 10 households and 33 analytes (including pentachlorophenol, for which all values were non-detect)
pg. 33) that the variability represented by the duplicate
pairs (measurement error) was less than the short-
term temporal variation addressed by the replicate
pairs.
Laboratory Comparisons. Several types of samples
were independently analyzed by SwRI and EMSI.
Triplicate air samples collected by SwRI in Phase II
were analyzed by EMSI, and the data was compared
with the corresponding primary and duplicate sample
data developed by SwRI. Split extracts prepared by
SwRI were analyzed by both laboratories. Both
laboratories also analyzed a standard reference
material provided by EPA, as well as two sets of blind
spike samples, one prepared by EMSI and the other
prepared by another EPA contractor.
The results of the laboratory comparisons indicate that
both laboratories generally achieved the desired
accuracy and precision limits defined for NOPES. In
most cases, differences between the laboratories
were relatively minor compared to other sources of
variability. More substantial interlaboratory differences
were evident for heptachlor and, to a lesser extent,
propoxur. This may reflect the analytical difficulties
associated with these analytes.
Detection Limits. Detection limits for NOPES target
compounds were estimated for each sampling season.
The actual limits of detection varied between analytical
batches within sampling seasons, being higher in
batches in which the instrument gave less response to
the standard. Limits of detection were also higher for
"dirty" samples than for "clean" samples. Moreover,
the procedures used to calculate detection limits were
different for GC/ECD and GC/MS compounds
because of differences in these two analytical
techniques. Therefore, ranges of estimated limits of
detection are presented in Table 35 for NOPES target
compounds quantitated using the GC/ECD technique
and in Table 36 for those quantitated using GC/MS.
•Inspection of Table 35 reveals that the detection limits
for many GC/ECD target compounds were lower in
Season 1 (Summer, Jacksonville) than in the other
seasons. This occurred because of a change from a
labor-intensive, manual method of interpreting the
GC/ECD chromatographs in Season 1 to a more
automated procedure for the second two seasons.
Using the more labor-intensive procedures in Season
1, the analytical chemists could detect lower levels of
occurrence for many compounds than was possible in
the other two seasons, especially in clean samples.
Although the detection limits vary across seasons, the
chemists consistently attempted to ensure that levels
of the analytes exceeding the QA goals established in
the QA Project Plan (see Table 8) were accurately
quantitated.
63
-------
Table 35. Ranges of Estimated Limits of Detections for GC/ECD Target Compounds by Site and Season
(ng/nv»)
Analyte
Dichlorvos
a!pha-BHC
Hexachlorobenzene
gamma-BHC
Chlorothalonil
Heptachlor
Ronnel
Chlorpyrifos
Aldrin
Dacthal
Heptachlor epoxide
Oxychlordane
Captan
Folpet
2,4-D butoxyethyl esterb
Dioldrin
Methoxychlor
Dicofol
cis-Permethrin
trans-Permethrin
Chlordane
4.4'-DDT
4,4'-DDD
4.4'-DDE
Summer
Jacksonville
1.5 -2.2
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
1.6-2.5
0.5 - 3.1
0.5 - 0.8
0.5
1.0-2.7
18
2.2 - 3.6
2.0 - 4.2
20 - 40
Jacksonville
39 -49
1.6-2.0
1.1 - 1.3
1.7- 1.8
1.2- 1.4
1.7 -2.4
2.2 - 2.5
2.4 - 2.7
1.7 -2.0
1.7 - 1.8
1.8 -2.0
1.7-2.0
2.5 - 3.5
3.6 - 5.1
15 - 17
1.7-2.0
4.2 - 5.1
18 -33
29 -38
19 -22
35 -50
2.2 - 2.5
2.7 - 3.1
2.0 - 2.9
Spring
Springfield
56-79
2.7 - 3.0
1.9 -2.2
2.9 - 3.1
2.3 - 2.5
2.8 - 3.1
4.1 - 4.4
4.1 - 4.5
2.9 - 3.2
2.7 - 3.1
3.1 - 3.3
3.0 - 3.3
5.4 - 7.8
7.5 - 1 1
24-30
3.0 - 3.3
7.2 - 7.8
46-74
43 - 53
30-38
24 - 33
3.8 - 4.1
4.7 - 5.3
3.3 - 3.6
Winter
Jacksonville
31 - 35
1.0-1.3
1.0 - 1.1
1.2 - 1.4
1.3 - 1.9
1 .2 - 1 .4
1.7 -2.2
2.5 - 3.1
1.2 - 1.5
1.8 -2.1
1.5 - 1.9
1.4-1.7
6.1 - 14
4.7 - 10
11 - 14
1.3 - 1.6
3.6-5.1
10 -25
20 - 23
14 - 16
4 - 13
1.9 -2.9
1.8 -2.7
1 .4 - 1 .7
Springfield
40 -45
1.3-1.5 "
1 .2 - 1 .4
1.5-1.7
1 .3 - 1 .4
1.4-1.6
2.2 - 2.4 ,
3.3 - 3.5
1 .7 - 1 .9
2.4 - 2.6
2.3 - 2.5
1.8-2.1
4.5-13
11-37
14-16
1.9 -2.1
4.5. - 5.0
9 - 10
28-31
19-21
5-11
2.1 - 2.3
2.3 - 2.6
1.8-2.0
"Lowest value normally detectable. .
bAnalyte was the methyl ester for the Jacksonville summer season.
Table 36. Ranges of Estimated Limits
(ng/irfl)
Analyte
ortho-Phenylphenol
Propoxur
Bendiocarb
Atrazine
Diazinon
Carbaryl
Malathton
Resmethrin
Summer
Jacksonville
5- 15
3- 9
13-38
12-42
11 -22
9-28
11 -48
12-28
of Detection3
for GC/MS Target Compounds by
Spring
Jacksonville Springfield
13
8
22
32
48
25
60
48
12
7
20
45
22
25
25
16
Site and Season:
Winter
Jacksonville
7 -22
4 - 16
7-38
14 - 40
16-45
8 -42
11-42
10-35
Springfield
5-20
4-12
9-30
1 1 - 45
13-42
1 1 - 32
10 -45
8 - 25
aConservative estimate -- lower values detectable in clean samples.
64
-------
The detection limits shown in Table 35 for GC/ECD
target compounds were calculated in such a manner
that they estimate the minimum possible detection
limits across sample batches. Because of differences
in the analytic methods, a slightly different estimation
procedure was necessary for the GC/MS compounds.
The detection limits shown in Table 36 for the GC/MS
compounds are more conservative and estimate the
analyte level that could be consistently detected
across batches. Lower levels could often be detected
in clean samples. Details of the methods used to
estimate the detection limits are provided in Hsu et al
(1988).
Questionnaire Data Quality
Questionnaire data from any survey is affected to
some degree by nonsampling error. Below are listed
some of the sources that can contribute to
inaccuracies in the data:
1. Respondents may not understand a question and
therefore are unable to answer. This is particularly
true if the question contains technical terms or
addresses a complex subject.
2. Respondents can misinterpret a question, and
then inadvertently provide an incorrect response.
3. Even if a respondent correctly understands a
question, he or she may not know the answer, or
may provide an inaccurate response.
4. Respondents may refuse to answer a question.
This is often because the question deals with a
sensitive subject, or because of the time needed
to provide an answer.
5. Interviewers can make mistakes when reading the
questions or when recording the responses.
6. Errors can be introduced when transferring the
questionnaire data from hard copy to machine-
readable format. Illegible responses, mistakes in
editing or coding, and keying errors may all result
in inaccuracies in the questionnaire data file used
for analysis.
All of these sources were recognized during ^the
development phase of MOPES as potential
contributors to nonsampling error, and steps were
taken to minimize their impact. Questionnaire
wordings were tested in the pilot study and revised
when necessary to improve respondent
understanding. Prompts were used in the
questionnaires and by the interviewers to promote
complete and accurate response. Some questions,
such as those on termiticide use and age of home,
were asked in both the screening questionnaire and
the study questionnaire so that responses could be
compared. Accurate questionnaire administration and
response recording were stressed during interviewer
training. Completed questionnaires went through a
field-edit before being sent to RTI, where they were
manually edited again for legibility, completeness,
andlogical consistency prior to keying. Questionnaire
data were keyed twice to minimize the incidence of
keying errors.
Despite the above actions, some nonsampling error
was unavoidable, and was present in the
questionnaire data file used in the analyses. Most of
the error is believed to be relatively minor. The area in
which the nonsampling error is of more concern is the
reported history of termiticide use in the monitored
housing units.
The primary sources of problems in the termiticide
use data were respondent misinterpretation and lack
of knowledge. The potential for misinterpretation was
recognized during the pilot study, when some
respondents indicated that they did not differentiate
between inspection visits by pest control professionals
and actual treatment of their homes, with termiticides.
As a result, the questionnaire wording was changed to
at least partially alleviate this problem. In addition
interviewers were instructed on how to clarify
questions for respondents that expressed confusion
about, what was being asked. Nonetheless, the
occasional dissimilarity between screening and study
questionnaire responses may have been caused in
part by misinterpretation by either the screening
respondent or the monitored individual.
Lack of knowledge,about a housing unit's termiticide
history is a more difficult problem to overcome
Because termiticides are applied infrequently, people
who have lived in a unit for only a few years or less
often will not know if the unit has been treated. This is
especially true in rental units. Overcoming this lack of
knowledge would be a resource-intensive activity
requiring the identification of and contact with previous
owners or landlords, and was not attempted for this
study. , v
The exploratory analyses of termiticide air
concentrations versus reported termiticide use have
yielded promising results. However, nonsampling error
may limit the degree of precision that can be
ultimately expected from models that predict
termiticide air concentrations from questionnaire data.
Comparisons to Other Studies
This study was the first to examine the
nonoccupational indoor air concentrations for many of
the target compounds. However, some of the NOPES
analytes, including chlordane, heptachlor, aldrin
dieldrin, chlorpyrifos, diazinon, propoxur, dichlorvos'
malathion, and ronnel, have been addressed by other
studies. Lewis (1988) provides an overview of the
indoor air concentrations observed in a variety of
studies.
Comparison to data produced by other studies
provides an independent assessment of the quality of
the NOPES data, although caution must be exercised
65
-------
when comparing the findings of a general-purpose,
probability-based study like NOPES to results from
special-purpose or non-probability studies. None of
the other studies examined were designed to produce
probability sampling estimates of the mean
concentrations experienced by a specific population.
Comparisons of the ranges of concentrations
observed in the studies are not as greatly affected by
differences in the study designs as other statistics.
Therefore, Table 37 summarizes the ^maximum
concentrations observed in studies reported in the
literature.
Termiticides, especially chlordane and heptachlor, are
the most widely studied of the NOPES analytes.
Several studies (Wright and Leidy, 1982; EPA, 1983)
focused on concentration profiles over time in homes
treated as part of the study. Sample sizes in these
studies were relatively small. Sample sizes in two
other studies (Olds, 1987; Lillie and Barnes, 1987)
were much larger, but only military housing units were
examined. Variation in housing unit age and types as
limited in the two military studies, which were
prompted by earlier reports of health problems related
to high pesticide concentrations in some military
dwellings.
Given the focus of studies other than NOPES, the
maximum concentrations observed in such studies
might be expected, a priori, to be higher than those in
a general-purpose survey, such as NOPES. For
chlordane and aldrin, this expected outcome was
observed in some, but not all, studies. NOPES
maxima were similar to those observed in other
studies for heptachlor and dieldrin.
Chlorpyrifos and diazinon have been the subjects of
several studies, including military housing studies and
temporal profile studies. The a priori expectation,
given the study populations, was for higher maxima in
these studies than in NOPES, which was generally
observed.
Comparisons for the remaining analytes were limited
to single studies. Some of these involved a limited
number of homes, while others (propoxur) were based
on observations from single rooms or dwellings.
NOPES maxima for these analytes were within an
order of magnitude of the maxima observed in the
other studies.
In summary, the NOPES findings are similar to those
of earlier studies. NOPES confirmed earlier
observations that indoor air pesticide concentrations
are commonly substantially higher than outdoor air
concentrations (Lewis and Lee, 1976; Lewis and
MacLeod, 1982). The ranges of indoor air
concentrations observed in NOPES were usually
comparable to those measured in other studies.
Table 37. Comparison of Maximum Indoor Air Concentrations
Analyte
Chlordane
Heptachlor
Aldrin
Dieldrin
t
Chlorpyrifos
Diazinon
Propoxur
Dichlorvos
Malathion
Ronnel
Study
Lillie (1981)
Livingston and Jones (1981)
Wright and Leidy (1982)
Lewis and MacLeod (1982)
EPA (1983)
Leidy etal. (1985)
Olds (1987)
Qazi (1987)
Lillie and Barnes (1987)
NOPES
Wright and Leidy (1982)
EPA (1983)
Jurinski (1984)
Leidy etal. (1985)
NOPES
EPA (1983)
Olds (1987)
Jacquith etal. (1987)
NOPES
EPA (1983)
NOPES
Lewis and MacLeod (1982)
EPA (1983)
Leidy and Wright (1987)
Bush etal. (1987)
Olds (1987)
NOPES
Leidy etal. (1982)
Lewis and Macleod (1982)
Leidy etal. (1984)
Olds (1987)
NOPES
Jackson and Lewis (1981)
NOPES
Lewis and MacLeod (1982)
NOPES
Lewis and MacLeod (1982)
NOPES
Lewis and MacLeod (1982)
NOPES
Maximum
Concentration
(ug/m3)
37.8
264
5.8
5.5
3.6
9.9
130
52
>5
4.4
1.8
0.6
14.8
2.0
2.4
7
1.6
5
1.8
0.17
0.18
7.0
37
8.5
4.5
11.9 ,
4.4-
38
2.0
149
34.6
13.7
0.79
7.9
28
2.9
1.0
1.9
10
0.0016
66
-------
References
Akland, G.G., Hartwell, T.D., Johnson, T.R., and
Whitmore, R.W., 1985, Measuring Human
Exposure to Carbon Monoxide in Washington,
D.C., and Denver, Colorado, during the Winter of
1982-1983. Environ. Sci. Technol. 19:911-918.
ASTM, 1989, Standard Method for Chlordane and
Heptachlor Residues in Indoor Air. ASTM
Philadelphia, PA.
Baker, S.R. and Wilkinson, C.F. (Editors). The Effects
of Pesticides on Human Health. Proceedings of a
Workshop, May 9-11, 1988, Keystone Co., Task
Force of Environmental Cancer and Heart and
Lung Disease, Volume XVIII, Advances in Modern
Environmental Toxicology, Princeton Scientific
Publishing Co., Inc. Princeton, NJ 438 pp.
Bristol, D.W., MacLead, K.E., and Lewis, R.G., 1984,
Direct and Indirect Chemical Methods for
Exposure Assessment. In Determination and
Assessment of Pesticide Exposure, M. Siewierski
(ed.), Elsevier, New York, pp. 79-119.
Budd, W.T., Roberts, J.W., and Ruby, M.J., 1988,
Field Evaluation of a High Volume Surface
Sampler for Pesticides in Floor Dust. EPA/600/3-
90/030. U.S. Environmental Protection Agency,
Washington, D.C. 52 pp.
Bush, P.B., Taylor, J.W., McMahon, C.K., and Neary,
D.G., 1987, Residues of Lindane and Chlorpyrifos
in Firewood and Woodsmoke. J. Entomol. Sci. 22:
131-139.
EPA, 1983, Analysis of Risks and Benefits of Seven
Chemicals Used in Subterranean Termite Control.
EPA-540/9-83-005. U.S. Environmental Protection
Agency, Washington, DC. 66 pp.
Hsu, J. P., Wheeler, H.G., Camann, D.E.,
Schattenberg, H.J., Lewis, R.G., and Bond, A.E.,
1988, Analytical Methods for Detection of
Nonoccupational Exposure to Pesticides. J.
Chromatogr. Sci. 26:181- 189.
Immerman, F.W., Rush, G.A., Jones, S.M., Camann,
D.E., Harding, H.J., Hsu, J.P., Lev-on, M.,
Delwiche, J.C., Levan, L, and Lin, C.C., 1988,
Nonoccupational Pesticide Exposure Study
(NOPES) Phase II Interim Report, Vols. 1 and 2
EPA/600/0-90/003. U.S. Environmental Protection
Agency, Washington, D.C. 649 pp.
Immerman, F.W., Rush, G.A., Jones, S.M., Camann
D.E., Harding, H.J., Schattenberg, H.J., Hsu, J.P.,
1988a, Nonoccupational Pesticide Exposure Study
(NOPES) Phase III Interim Report, Vols. 1 and 2
EPA/600/0-90/004. U.S. Environmental Protection
Agency, Washington, D.C. 652 pp.
Jackson, M.D. and Lewis, R.G., 1981, Insecticide
Concentrations in Air After Application of Pest
Control Strips. Bull. Environm. Contam Toxicol
27: 122-125.
Jacquith, D.A., McDavit, W.M., and Reinart, J.C.,
1987, Monitoring of Air Levels of Termiticides in
Homes in the United States. Proc 4th Internatl.
Conf. Indoor Air Quality and Climate, Berlin, West
Germany, pp. 230-234.
Jurinski, N.B., 1984, The Evaluation of Chlordane and
Heptachlor Vapor Concentrations Within Buildings
Treated for Insect Pest Control. Proc. 3rd
Internatl. Conf. Indoor Air Quality and Climate,
Stockholm, Sweden, pp. 51-56.
Leidy, R.B., Wright, C.G., and Dupree, Jr., H.E.,
1982, Concentration and Movement of Diazinon in
Air. J. Environ. Sci. Health, B17: 311-319.
Leidy, R.B., Wright, C.G., MacLeod, K.E., and
Dupree, Jr., H.E., 1984, Concentration and
Movement of Diazinon in Air. II. Vertical
Distribution of Rooms. J. Environ. Sci. Health
B19: 747-757. '
Leidy, R.B., Wright, C.G., Dupree, Jr. H.E., and
Sheets, T.J., 1985, Subterranean Termite Control:
Chlordane Residues in Soil Surrounding and Air
Within Houses. In R.C. Honeycutt. G. Zweig, and
N.N. Ragsdale, eds., Dermal Exposure Related to
Pesticide Use, ACS Symposium Series 273,
67
-------
American Chemical Society, Washington, DC. pp.
265-276.
Leidy, R.B. and Wright, C.G., 1987, Airborne
Residues of a Termiticide Formulation of
Chlorpyrifos in Homes. 193rd American Chemical
Society National Meeting, Denver, CO. Paper No.
AGRO 2.
Letz, R., Ryan, P.B., and Spengler, J.D., 1984,
Estimated Distributions of Personal Exposure to
Respirable Particles. Environ. Mon. Assess.
4:351-359.
Lev-On, M., Delwiche, J. C., Lin, C.C., Immerman, F.
W., Rush, G.A., Jones, S. M., Camann, D. E.,
Harding, H.J., and Hsu, J.P., 1987,
Nonoccupational Pesticide Exposure Study
(NOPES) Phase I: Jacksonville, Florida, Summer,
1986 Interim Report with addendum. EPA/600/0-
90/001. U.S. Environmental Protection Agency,
Washington, DC. 309 pp.
Lewis, R.G. and Lee, Jr., R.E., 1976, Air Pollution
from Pesticides: Sources, Occurrences and
Dispersion. In Air Pollution from Pesticides and
Agricultural Processes, R.E. Lee, Jr. (ed.), CRC
Press, Boca Raton, FL, pp. 5-55.
Lewis, R.G. and MacLeod, K.E., 1982, A Portable
Sampler for Pesticides and Semi-volatile Industrial
Organic Chemicals. Anal. Chem. 54:310-315.
Lewis, R.G., Bond. A.E., Johnson, D.E., and Hsu,
J.P., 1988, Measurement of Atmospheric
Concentrations of Common Household Pesticides:
A Pilot Study. Environ. Mon. Assess. 10: 59-73.
Lillie, T. H., 1981, Chlordane in Air Force Family
Housing: A Study of Houses Treated After
Construction. OEHL 81-45, USAF Occupational
and Environmental Health Laboratory, Brooks
AFB, Texas.
Lillie, T.H. and Barnes, E.S., 1987, Airborne
Termiticide Levels in Houses on United States Air
Force Installations. Proc. 4th Internatl. Conf.
Indoor Air Quality and Climate, Berlin, West
Germany, pp. 200-204.
Livingston, J.M. and Jones, C.R., 1981, Living Area
Contamination by Chlordane Used for Termite
Control. Bull. Environm. Contam. Toxic. 27: 406 -
411.
Method 608, 1984, Organochlorine Pesticides and
PCBs. Federal Register 49:209, 89-104.
Olds, K.L., 1987, Indoor Airborne Concentrations of
Termiticides in Department of the Army Family
Housing. Proc. 4th Internatl. Conf. Indoor Air
Quality and Climate, Berlin, West Germany, pp.
205-209.
Ott, W.R., 1985, Total Human Exposure. Environ. Sci.
Technol. 19: 880-886.
Ott, W.R., Wallace, L., Mage, D., Akland, G., Lewis,
R., Sauls, H., Rodes, C., Kleffman, D., Kurada,
D., and Morehouse, K., 1986, The Environmental
Protection Agency's Research Program on Total
Human Exposure. Environ. Int. 12:475-494..
Qazi, A.H., 1987, Termiticides in the Indoor
Environment. Proc. 4th Internatl. Conf. Indoor Air
Quality and Climate, Berlin, West Germany, pp.
248-251. .
Ryan, P.B., Soczek, M.L., Treitman, R.D., Spengler,
.J.D., and Billick, I.H., 1988, The Boston
Residential NOa Characterization Study - II.
Survey Methodology and Population
Concentration Estimates. Atmos. Environ. 22,
2115-2125.
U.S. Bureau of the Census, 1983a, County and City
Data Book, 1983. U.S. Government Printing
Office, Washington, DC, pp. 690 and 720.
U.S. Bureau of the Census, 1983b, Statistical Abstract
of the United States: 1984. U.S. Government
Printing Office, Washington, DC, p. 128.
Wallace, L.A., 1987, The Total Exposure Assessment
Methodology (TEAM) Study: Summary and
Analysis: Volume 1. EPA-600/6-87-002. U.S.
Environmental Protection Agency, Washington,
DC, 192pp.
Wolfe, H.R., ,1976, Field Exposure to Airborne
Pesticides. In Air Pollution from Pesticides and
Agricultural Processes, R;E. Lee, Jr. (ed.), CRC
Press, Boca Raton, FL, Chapter 5.
Wright, C.G.'and Leidy, R.S., 1982, Chlordane and
Heptaehlor in the Ambient Air of Houses Treated
for Termites. Bull. Environm. Contam. Toxicol. 28:
617-623.
68
-------
Appendix A
NOPES Survey Instruments
69
-------
-------
HOUSEHOLD SCREENING QUESTIONNAIRE
NOPES STUDY/RTI PROJECT 3620
O.M.B. No. 2080-0022
Expires 9/30/88
7/86
FOR OFFICIAL USE ONL
A. HOUSEHOLD IDENTIFICATION
Segment No. I I I I
SHU No. | | | |
STREET ADDRESS
TELEPHONE NUMBER
Obtained:
City
State Zip Code
i I M - JIT
NO PHONE 00
REFUSED 01
B. RECORD OF SCREENING CALLS
DAY OF WEEK
DATE
TIME
RESULTS
CODE Fl ID No.
am/pm
am/pm
am/pm
am/pm
am/pm
am/pm
FINAL SCREENING RESULT
D. INFORMANT ID
FS USE ONLY
SAMPLE DROP: (CIRCLE ONE)
Vacant 01
Not an HU 02 SECTION
Temporary/vacation home 03 D
SCREENING NOT COMPLETED:
Refusal Q4
No one at home (after
repeated visits) 05
No eligible respondent HOLD FOR
(after repeated visits) . . 06
Language barrier 07
Other (SPECIFY) 08
SCREENING COMPLETED ... 09
NOTES:
NAME:
ADDRESS:
CITY
STATE
ZIP
RELATIONSHIP/TITLE:
TELEPHONE NUMBER:
NO PHONE 00
REFUSED 01
FS
Approval:
Verified?
Yes .... 01
No 02
Date of
Verification
COMMENTS:
71
-------
A. HOUSEHOLD ROSTER
Hello, I'm (NAME) from the Research Triangle Institute. (THEN, AFTER IDENTIFYING THE HEAD OF
HOUSEHOLD, SPOUSE OF HEAD, OR OTHER RESPONSIBLE ADULT, SAY). We are conducting a
research study for the U.S. Environmental Protection Agency. Here is a letter that explains the study.
(HAND LETTER AND ALLOW TIME FOR READING.)
1.
2.
First, how many people 16 years of age or older (including friends or roomers) live in this
household? How many younger than 16?
a.
HOUSEHOLD MEMBERS
16 OR OLDER
b.
HOUSEHOLD MEMBERS
UNDER 16
ENTER NAME IN FIRST COLUMN OF THE
What is the name of the head of this household?
ROSTER: TOP OF PAGE 3.
What are the names of all other persons 16 years of age or older who live here? Let's list them
in order of age, beginning with the oldest. ENTER NAME(S) AND RELATIONSHIP TO THE -
HEAD OF HOUSEHOLD IN ROSTER.
NOTE- IF THERE ARE MORE THAN FOUR HOUSEHOLD MEMBERS 16 OR OLDER, USE
ANOTHER SCREENING QUESTIONNAIRE TO COMPLETE THE ROSTER AND
APPROPRIATE QUESTIONS ABOUT EACH OF THE INDIVIDUALS. (INSERT THE SECOND
SCREENER INSIDE THE FIRST SCREENER UPON COMPLETION.) IF THERE ARE MORE
THAN 10 HOUSEHOLD MEMBERS, CONSIDER THE UNIT TO BE GROUP QUARTERS
RATHER THAN A-HOUSEHOLD, STOP INTERVIEW AND EXCUSE YOURSELF.
CHECKPOINT:
DOES NUMBER OF NAMES LISTED IN ROSTER EQUAL NUMBER OF HOUSEHOLD
MEMBERS RECORDED IN Q. 1a?
YES (GO TO QUESTION 5)
NO (RECONCILE DISCREPANCY WITH RESPONDENT AND
CORRECT Q. 1a OR ROSTER AS NECESSARY.)
72
-------
Name of Person Interviewed:
B. PERSONAL DATA (FOR EACH PERSON LISTED BELOW, ASK Q's 4-12)
'•SSET FIRSTNAME
LAST NAME
RELATIONSHIP TO HEAD OF HH
PERSON NUMBER
5. Is (PERSON) male or
female? CIRCLE CODE
6. In what range is (PERSON'S)
age?,
1. 16-25
2. 26-45
3. 46-60
4. Over 60
(RECORD PROPER CODE)
7. Is (PERSON) presently
employed in any capacity?
IF "NO", SKIP TO Q. 10
8. Is (PERSON) employed in any
of the jobs listed on this card?
(HAND RESPONDENT CARD
A) RECORD PROPER CODE
9. In (PERSON'S) current job,
does he/she use or handle
any insecticides,
fungicides or_herbicides
such as weed killers,
wood preservatives or
, nsect/pest killers?
10. Which of the following
best describes (PERSON'S)
status? CIRCLE ONE
•
Other (Specify)
1 . Is (PERSON) involved in any of
the following activities at any
time of year?
CIRCLE ALL THAT APPLY
IF "NONE", SKIP TOO. 13 [
12. In that activity does (PERSON)
use any insecticides,
fungicides or herbicides such
as weed killers, flower/ plant
sprays, or insect/pest killers?
•
01
M F
D Yes
D No
D Yes
D No
EH Don't know
| SKIP TO Q. 1 1 |
01 Housewife
02 Student
03 Unemployed
04 Retired
05 Disabled
01 Outdoor/Gardening
lawnwork
02 Indoor flower or
plant care
03 None of these
08 Don't know
D Yes
D No
D Don't know
02
M F
n Yes
D No
D Yes
D No
Q Don't know
SKIP TO Q. 1 1
01 Housewife
02' Student
03 Unemployed
04 Retired
05 Disabled
01 Outdoor/Gardening
lawnwork
02 Indoor flower or
plant care
03 None of these
08 Don't know
D Yes
fj No
Q Don't know
03
M F
D Yes
D No ,
D Yes
D No
Q Don't know
| SKIP TO Q. 1 1
01 Housewife
02 Student
03 Unemployed
04 Retired
05 Disabled
01 Outdoor/Gardening
lawnwork
02 Indoor flower or
plant care • •
03 None of these
08 Don't know
n Yes
D No
C] Don't know*.
04
M F
D Yes
D No
D, Yes
D No
C] Don't know
SKIP TO Q. 1 1
01 Housewife
02 Student
03 Unemployed
04 Retired
05 Disabled.
01 Outdoor/Gardening
lawnwork
02 Indoor flower or
plant care
03 None of these
08 Don't know
D Yes
D No .
C] Don't know
73
-------
C. HOUSEHOLD DATA
The next few questions are about this household in general.
13. Does this household have any cats or dogs?
[T| Yes
|~2~] No (SKIP TO Q. 18)
14. How many cats and/or dogs does this household have? ,
a. Cats
b. Dogs
15. Are any of the following used on your cats and/or dogs to control fleas or ticks at any time of year?
(CHECK ALL THAT APPLY)
a.
Flea/tick shampoos or dips
b. Flea powders
c. Flea collars
>• D
None of these
Don't know
SKIP TOO. 18
16. Are the treatments, shampoos, or sprays usually done inside or outside your home, or are they done
at a veterinarian's or professional pet groomer's?
I 1 I Inside
2 Outside
8
Veterinarian or professional pet groomer
Don't know
17. At this time of year, how often (generally) are the treatments, shampoos, or sprays performed?
At least once a month
Less than once a month
8 I Don't know
74
-------
18. How old is this house/building?_
19. What type of foundation does this house/building have?
Slab
Crawl space
Combination crawl space/basement
Full basement
Oher (Specify).
8 Don't know
20. Since this house/building was built, have pesticides or chemicals ever been used to control a termite
problem in it?
'Yes
No (SKIP TO Q. 22)
Don't know (SKIP TO Q. 22)
21. When was the last time pesticides or chemicals were used to control termites in this house/building?
Less than 1 year ago >
More than 1 year ago
Don't know
22. Excluding termite treatments, are insecticides ever applied in this home/apartment for roaches ants
silverfish, fleas, or other household insect pests?
Yes
8
No (SKIP TO Q. 25)
Don't know (SKIP TO Q. 25)
75
-------
23. How often is this home/apartment usually treated for these pests?
11 | Every month or more often
12 I Every 2-4 months
Every 5-11 months •
Every year
Less frequently than every year
Don't know
8
24. Who usually treats this home/apartment for these pests? (READ CATEGORIES-CHECK ONE BOX)
I T | A professional service
[£J Someone in the household (PERSON NUMBER(S) FROM ROSTER _
13 [ Both professional service and one or more household members
|4| Other (SPECIFY) •_..
Don't know .
8
25. Does this house (or apartment/mobile home) have air conditioning?
Central air conditioning
Window unit(s)
3 No air conditioning (SKIP TO Q.27)
76
-------
26. Hdw often, if ever, are the windows or doors opened and left open for several hours at this time of
year?
Less than once a week
More than once a week
Don't know
27. What is your primary source of drinking water?
City Water
Private company
Private well
Bottled water
28. What is your home telephone number, starting with your area code?
c
Iheck if no home phone.
29. Is this phone number unlisted?
1
2
Yes
No
77
-------
D. RECORD BY OBSERVATION: (if unable to accurately record by observation, BE SURE TO ASK THE
RESPONDENT.)
30. Type of structure in which the Housing Unit is located
Unattached single unit
Attached single unit (e.g., duplex, row house)
Multi-unit building (e.g., apartment building)
Mobile home
Other (SPECIFY) -
* •
0
31. Outdoor area around the structure
0
2
IT
[4"
Private yard area with lawn, trees, and/or shrubs
Private yard without lawn, trees, or shrubs
Common area with lawn, trees and/or shrubs
Common area without lawn, trees, or shrubs
Other (SPECIFY)
(END OF QUESTIONNAIRE. THANK RESPONDENT FOR HIS/HER TIME AND COOPERATION.)
78
-------
Response Card A
79
-------
-------
CARD A
OCCUPATIONS
1. Pest control operator (PCO)/ professional pesticide applicator
2. Construction worker
3. Employee at a facility where items such as furniture or garments are fumigated
4. Landscaper or nursery worker
5. Employee at a golf course
6. Maintenance worker such as building janitor or groundskeeper
7. Food processing plant employee
8. Veterinarian, veterinary assistant or worker at a zoo
9. Agricultural worker
10. Employee at a facility that manufactures, formulates, or distributes pesticides
11. Chemist or chemical laboratory technician
12. None of the above occupations
81
-------
-------
Study Questionnaire
83
-------
STUDY QUESTIONNAIRE
NON-OCCUPATIONAL PESTICIDE EXPOSURE SURVEY
OMB No. 2080-0022
Expires: 9/30/88
Rrst, I would like to ask some general questions about you.
1. Sex (by observation):
1
Male
Female
2. Race (by observation):
Q] Black
|2| White
American Indian/Alaskan Native
Asian/Pacific Islander
Other (specify)
3. What was your age in years on your last birthday?
Years
IF AGED 15 YEARS OR LESS, ASK:
Previously, when I was asking questions about people in your household, I was told that you were
older than 15. Just to make sure I have this right, are you presently years old?
IF AGED 15 YEARS OR LESS, SELECT ALTERNATE RESPONDENT WITHIN THE HOUSEHOLD, IF
POSSIBLE NOTE IDENTITY OF THE ALTERNATE ON THE COVER SHEET. IF NO ALTERNATE
RESPONDENT IS AVAILABLE, THANK RESPONDENT FOR HIS/HER COOPERATION AND END
INTERVIEW.
Next, I would like to ask some questions about your occupation.
4. Are you presently employed in any capacity?
Yes (CONTINUE)
No (SKIP TO QUESTION 10)
5. Do any of the jobs on this list describe your current occupation or occupations?
HAND RESPONDENT SHOW CARD A
RECORD PRIMARY OCCUPATION CODE
RECORD SECONDARY OCCUPATION CODE, IF ANY
IF OCCUPATION CODE = 1 (e.g., PCO), THANK RESPONDENT AND TERMINATE INTERVIEW.
84
-------
I am interested in finding out whether you use or handle any pesticides in your current job.
A pesticide is a chemical used to destroy, prevent, control or repel pests. By pests I mean such thinqs as
insects, spiders, fleas, fungus, mildew, and weeds. 9
HAND RESPONDENT SHOW CARD B AND READ EXAMPLES
Weed killers ; ,
Wood preservatives
Lawn sprays
Fruit and vegetable sprays or dusts
Rose sprays
Insect killers or repellants
Mold inhibitors
Flea or tick treatments *-••••
6. Do you ever use or handle any of these types of pesticides in your job?
1
Yes (CONTINUE)
No (SKIP TO QUESTION 11)
7. Does your primary activity at work involve using or handling pesticides?
YES (THANK RESPONDENT AND TERMINATE INTERVIEW)
NO (CONTINUE)
85
-------
8.
I would like to find out some more details about the pesticides you use or handle in your job.
Could you please tell me the brand names of the pesticides you use or handle?
1
Yes (CONTINUE)
No/Don't Know
I—I (SKIP TO QUESTION 9)
ENTER PESTICIDES IN COLUMN 1 AND FOR EACH PESTICIDE ASK:
What is this pesticide used for? ENTER IN COLUMN 2.
How often do you use or handle this pesticide? ENTER CODE IN COLUMN 3.
PESTICIDE NAME
Column 1
DESCRIPTION OF USE
Column 2
DAILY 01
WEEKLY 02
MONTHLY 03
LESS FREQUENTLY THAN
MONTHLY -04
Column 3
AFTER COMPLETING COLUMNS 1, 2, AND 3 FOR ALL PESTICIDES MENTIONED, SKIP TO
QUESTION 11.
86
-------
9. Please describe in what ways you use or handle these pesticides in your job. (For example, use them
for weed control or to control ants.)
SKIP TO QUESTION 11
10. Which of the following best describes your status?
3 Unemployed
1
Housewife
Student
Retired
Disabled
The next set of questions are about this household in general.
11. Do you have any of the following at this home? (CHECK ALL THAT APPLY.)
Lawn or yard (By observation)
Ornamental shrubs and/or fruit or flowering trees
Vegetable garden
Flower garden/other outdoor plants
Indoor plants
Detached greenhouse
Attached greenhouse
Detached garage, shed or other buildings
Attached garage or storage room
Household pets
12. When was this house/building built?
10
13. Since this house/building was built, have pesticides or chemicals ever been used to control a termite
problem in it?
Yes (CONTINUE)
No
Don't Know
SKIP TO DIETARY INTAKE RECORD
87
-------
14. When was the last time pesticides or chemicals were used to treat this house/building for termites?
[ 1 I Less than 1 year ago
1-5 years ago
|3 I More than 5 years ago
|_jj Don't know
15. At that time, who treated this home for termites? (READ CATEGORIES)
A professional service
Yourself
Someone else in the household
Other (specify)
GO TO DIETARY INTAKE RECORD.
88
-------
OMB No. 2080 -0022
Expires: 9/30/88
DIETARY RECALL INTERVIEW
DATE COMPLETED:
MONTH
DAY
YEAR
ID LABEL:
Day of Week
Monday ..
Tuesday .
Wednesday
Thursday .
Friday . . .
Saturday .
Sunday . .
01
02
03
04
05
06
07
OPENING
"Now I need to know everything you ate or drank yesterday from midnight of . "/ ' ' to
midnight of _. Please try to remember everything you ate or drank during the niqht
and day no matter how much or how little you had. Include food or drink you had at home or away from
home. As you tell me what you had, I will ask you how the food or drink was prepared. For example if
you ate eggs, I will need to know if they were scrambled, fried, poached or hard cooked. I will also need
to know the amount you ate or drank. I will help you use these models to describe how larqe the
portions were that you had. I will also ask if you added anything to the foods or drinks. Now starting
with midnight of , what was the first thing you ate or drank?"
89
-------
DIETARY INTAKE LISTING
Line Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Name of Food or Drink
Amount Consumed
For Office Use Only
90
-------
Line Number
26
27
28
29
30
31
32
33
... 34
35
36 ,
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Name of Food or Drink
1
. .- - , . . ,
-
Amount Consumed
For Office Use Only
. '• -
91
-------
MEASUREMENT CONVERSIONS
DECIMAL CONVERISON
3 teaspoons = 1 tablespoon
2 tablespoons = 1 fluid ounce
4 tablespoons = 1/4 cup
51/3 tablespoons = 1/3 cup
16 tablespoons = 1 cup = 8 ounces = 1/2 pint
2 cups = 1 pint
2 pints = 1 quart
4 quarts = 1 gallon
1
1/2
1/3
1/4
1/8
1.00
0.50
0.33
0.25
0.13
NOTES:
92
-------
HOUSEHOLD PESTICIDE INVENTORY
Now I'd like to ask you about pesticides you have around the house, I am going to read a list of
some common househould pests and mention examples of products that may be used to control the pests
This list was designed to help you remember what products you have here and help identify items not
usually thought of by most people as pesticides. For each pest I mention, please tell me if you currently
have any products here that can be used to control that pest. It doesn't matter if you've ever used the
product.
Please take your time as you consider each pest problem, and think carefully if you have anv
products that could be used for the problem. y
23.* Do you currently have any products that can be used to control:
YES
NO
a. Ants, cockroaches, or other crawling insects; products
such as Raid and Black Flag bug sprays? 01 02
b. Flies, gnats, and other flying insects; products such as Raid
and Black Flag bug sprays? . . 01 02
c. Bees, hornets, or wasps; products such as Raid Wasp and
Hornet Killer or Ortho Hornet and Wasp Killer? 01 Q2
d. Spiders and mites; products such as Defend? 01 02
e. Fleas; products such as Holiday or Four-Gone foggers? ...... 01 02
--Do you have any products that can be used to: .,.-..-
f. Treat or prevent indoor plant insects or diseases;
products such as Ortho Indoor Plant Spray? 01 02
g. Treat or prevent termites; products such as Chlordane? ...... 01 .02
h. Preserve wood; products such as Capernol or Creosote? 01 02
--Do you have any products to control or prevent:
i. Outdoor plant insects or diseases; products such as Sevin,
Malathion, rose dust, or tomato dust? . 01 02
j. Weeds; products such as crabgrass killers, dandelion killers, and
chickweed killers? 01 02
k. Tree diseases; products such as orchard or fruit tree sprays? 01 02
I. Fleas and ticks on pets; products such as soaps, shampoos
dips.or powders? ..01 02
-Do you have any:
m. Outdoor foggers; products such as Raid Yard Guard and Ortho
Yard and Patio Insect Spray? . . 01 02
1MOTE: QUESTIONS 16-23 were originally included in the Dietary Intake Questionnaire but were
deleted following the pilot testing.
93
-------
n. Continuous use products, such as flea and tick collars, no-pest
strips, ant traps, or roach baits? 01 02
o. Lysol disinfectant spray? 01
02
-Do you have any pesticides that the Health Department or any other
government agency gave you? •
01
02
24. Do you have any other products here that might be considered pesticides that you haven't already
told me about?
Yes
No
1 (CONTINUED)
2 (SKIP TO QUESTION 26)
25. What are they? LIST BELOW
a.
b. :
c.
d.
26. INVENTORY SECTION
Now let's talk about the products you have here. I'll need to see the containers to copy down the
exact name and EPA registration number of each one. I would like to go to the places where you store
your pest products; but if you prefer, you can collect them and bring them to me.
IF RESPONDENT ELECTS TO BRING PESTICIDES TO YOU, SAY:
If you would like, you can use this bag or these gloves to carry the containers. As you go through
your storage areas, please check for any products you may have forgotten. Thanks.
94 -
-------
1- CM
CO
0)
CO
CO
o
o
c?
-^ "CD ^ m CD
3*11*
•*-* 4^ +•* +^ ._ $
CO CO CO CO £ o
_eo _cg _co _eg w c
null
CM eo •* in co
01
a
2
2
C/3
"8
E c . •§
c o ® ffl a>
.-S ^
£! c! ® CO CO CD
O += CO £ •& ¥
*i CO CO CO 3 ±:
i^ CQ CQ OOO
i- CM co •
to CD
CD CD
co co
3 3
CD
CD!?
2-^ 2 ri
~l£.g 3
^o
o
3 T3 CO
O 9 Q.
.5 CO
O O i- O.CL-= «°
f r « o ^ E o
-o u co
=.toS S
Illfl-ls
CD
.a
< d>£
a.
CM
95
-------
-------
Response Card B
97
-------
CARDS
Weed killers
Wood preservatives
Lawn sprays
Fruit and vegetable sprays or dusts
Rose sprays
Insect killers or repellants
Mold inhibitors
Flea or tick treatments
98
-------
24-Hour Activity Log
99
-------
OMB No. 2080-0022
Expires: 9/30/88
24-HOUR ACTIVITY LOG
The following questions are designed to find out about the activities you were involved in during the 24-hour
monitoring period you have just completed. For some questions, you will also be asked to'report on
activities during the 24 hours before the monitoring period. Please be certain to answer separately for both
time periods.
Please answer Yes or No for whether you were involved in each activity I read to you during 24 hours
before the start of the monitoring period and/or during the monitoring period.
24 HOUR'S
BEFORE START
OF MONITORING
MONITORING
PERIOD
1. a) Rrst, were you involved in
gardening/lawn plant care?
IF YES,
b) About how much time did you
spend in this activity?
YES
NO
HRS.
MINS.
YES
NO
HRS.
MINS.
2. a) Were you involved in pet
handling^brushing/bathing?
IF YES.
b) About how much time did you
spend in this activity?
YES
NO
HRS.
MINS.
YES
NO
HRS.
MINS.
IF NOT CURRENTLY EMPLOYED, SKIP TO QUESTION 4.
a)
b)
Were you involved in using or
handling pesticides, insecticides,
fungicides or herbicides in your
current job?
IF YES,
About how much time did you
spend in this activity?
YES
NO
HRS.
MINS.
YES
NO
HRS. MINS.
100
-------
4.
Now I would like to read a list of different sorts of pesticides you may have used or handled at work
home, or somewhere else. Please answer YES or NO to whether or not you have used or handled
any of these products in the last 48 hours. -'..,. '^NUIWU
Did you use or handle any products to control ....
a. Ants, cockroaches, or other crawling insects; products such as Raid and
Black Flag bug sprays? . . . .-. YES NO
b. Flies, gnats, and other flying insects; products such as Raid and Black
Flag bug sprays? _ YES NO
c. Bees, hornets, or wasps; products such as Raid Wasp and Hornet
Killer or Ortho Hornet and Wasp Killer? . .-,." YES NO
d. Spiders and mites; products such as Defend? * YES NO
Did you use or handle any products to ...
e. Treat or prevent indoor plant insects or diseases; products such as
Ortho Indoor Plant Spray? YES NO
f. Treat or prevent termites; products such as Chlordane? ......... . YES NO
g. Preserve wood; products such as Capernol and Creosote ......... YES NO
Did you use or handle any products to control or prevent
h. Outdoor plant insects or diseases; products such as Sevin, Malathion
rose dust, or tomato dust? YES NO
i. Weeds; products such as crabgrass killers, dandelion killers, and
chickweed killers YES NO
j. Tree diseases; products such as orchard or fruit tree sprays? YES NO
k. Fleas and ticks on pets; products such as soaps, shampoos dips
or powders? ' .YES NO
Did you use or handle any ...
I. Indoor foggers to control fleas; products such as Holiday or Four-Gone
f°99ers? . YES NO
m. Outdoor foggers; products such as Raid Yard Guard and Ortho Yard
and Patio Insect Spray? YES NO
n. Lysol disinfectant spray? _ YES NO
101
-------
5 NOW FOR EACH PESTICIDE YOU SAID YOU USED OR HANDLED IN THE LAST 48 HOURS, I
WOULD LIKE TO KNOW THE PRODUCT NAME, WHERE YOU USED IT, WHEN YOU USED IT,
METHOD OF APPLICATION AND ANY PRECAUTIONARY ACTIONS YOU TOOK WHILE USING IT.
— : 1 ; — • 2
a.
b.
Product Name
24 Hours
Before During
Monitoring Monitoring
Where did you use it?
Home, Indoor
Home, Outdoor |~~
: n
i n
24 Hours
Before
Monitoring
D
During
Monitoring
D
Work, Indoor
Work, Outdoor
Elsewhere, Indoor
Elsewhere, Outdoor
n
n
n
n
n
n
n
What method of application did you use? (CHECK ALL THAT APPLY)
Handsprayer | |
Pressurized/Hose Sprayer | |
Brush or Cloth | I
I—I
Lawn or Garden Spreader
Hand Duster or Shaker Container
Aerosol Can
Other
What precautionary actions did you take while using it? (CHECK ALL THAT APPLY)
n n
n
n
n
n
n
n
n
Wore Protection Clothing (e.g.
Gloves, Apron, Boots, or Mask)
Held Breath
Covered or Removed Food
and/or Furniture
Washed Hands or Showered
Changed Clothes
None
D
n
n
n
n
n
n
n
102
-------
a. Product Name
24 Hours
Before
b. Where did you use it? Monitoring
Home, Indoor 1 — 1
Home, Outdoor | |
Work, Indoor 1 1
Work, Outdoor [ |
Elsewhere, Indoor | |
Elsewhere, Outdoor
c. What method of application did you use? (Cl
Handsprayer i
24 Hours
During Before During
Monitoring Monitoring Monitoring
4ECK
n n
n n
a
n
ALL THAT APPLY)
a
a
a
a
Pressurized/Hose Sprayer
Brush or Cloth
Lawn or Garden Spreader
Hand Duster or Shaker Container
Aerosol Can
Other
n
n
What precautionary actions did you take while using it? (CHECK ALL THAT APPLY)
Wore Protection Clothing (e.g.
Gloves, Apron, Boots, or Mask)
Held Breath
Covered or Removed Food
and/or Furniture
Washed Hands or Showered
Changed Clothes
None
n
n
103
-------
a. Product Name
b. Where did you use it?
Home, Indoor
Home, Outdoor
Work, Indoor
Work, Outdoor
Elsewhere, Indoor
Elsewhere, Outdoor
24 Hours 24 Hours
Before During Before
Monitoring Monitoring Monitoring
n :
c
n
: n
During
Monitoring
n
n
n n n n
n n n n
n n n n
What method of application did you use? (CHECK ALL THAT APPLY)
Handsprayer
Pressurized/Hose Sprayer
Brush or Cloth
Lawn or Garden Spreader
Hand Duster or Shaker Container
Aerosol Can
Other
n
n
n
n
n
n
What precautionary actions did you take while using it? (CHECK ALL THAT APPLY)
n n
Wore Protection Clothing (e.g.
Gloves, Apron, Boots, or Mask)
Held Breath
Covered or Removed Food
and/or Furniture
Washed Hands or Showered
Changed Clothes
None
n
n
n
104
-------
6. a)
To your knowledge, did anyone else in your household apply any pesticides in or around your
home during or in the 24 hours before the monitoring period?
YES
NO
D
(SKIP TO QUESTION 8)
7. For each such application, please describe the product used, when it was applied, the general location
of the application, and how the product was applied.
1 2
a. Product
24 Hours 24 Hours
. "... Before During Before During
b. Where was it used? Monitoring Monitoring Monitoring . Monitoring
Home, Indoor
Home, Outdoor I I
Work, Indoor I I
Work, Outdoor I I
Elsewhere, Indoor I I
Elsewhere, Outdoor
D
D D ..
n
* — *
n n
n n
What method of application was used? (CHECK ALL THAT APPLY)
Handsprayer
Pressurized/Hose Sprayer
Brush or Cloth
Lawn or Garden Spreader
Hand Duster or Shaker Container
D
n
Aerosol Can
Other
n
n
n
n
105
-------
a.
Product
b. Where
Home,
Home,
Work,
24 Hours 24 Hours
Before During Before During
was it used? Monitoring Monitoring Monitoring Monitoring
Indoor
Outdoor f [
Indoor
Work, Outdoor | |
Elsewhere, Indoor [ |
Elsewhere, Outdoor |
D Q D
D D
D D D
II 1 J
D D D
| |
D
D
What method 'of application was used? (CHECK ALL THAT APPLY)
Handsprayer
Pressurized/Hose Sprayer
Brush or Cloth
Lawn or Garden Spreader
Hand Duster or Shaker Container
Aerosol Can
Other
D
D
D
106
-------
a.
Product
b. Where
Home,
Home,
Work,
Work,
24 Hours 24 Hours
Before During Before During
was it used? Monitoring Monitoring Monitoring Monitoring
Indoor 1 1
Outdoor J~[
Indoor 1 1
Outdoor
Elsewhere, Indoor | |
Elsewhere, Outdoor
n n
ii ii
D
n n n
c.
What method of application was used? (CHECK ALL THAT APPLY)
n
n
Handsprayer
Pressurized/Hose Sprayer
Brush or Cloth
Lawn or Garden Spreader
Hand Duster or Shaker Container
Aerosol Can
Other
n
n
n
n
8. Were any of the following products in use in or around your home or job during the monitoring period?
CHECK YES OR NO FOR EACH. ,
HOME
WORKSITE
a) flea/tick collars
b) no-pest strips
c) ant or roach baits
YES
n
n
NO
D
C
n
YES
|
n
n
NO
D
1
107
-------
9. During the 24-hour monitoring period, in your home, how much time was spent with ......
ENTER HOURS AND MINUTES .,
•HRS. MINS. NONE
a) central heat or air conditioning in
operation?
b) window air conditioner in operation?
windows or doors open?
c)
d)
windows and doors shut, no air conditioner
in operation
10. About how much time during the past 24-hour monitoring period did you spend .......
ENTER HOURS AND MINUTES
HRS. MINS. NONE
indoors at home
a)
b)
c)
d)
indoors at work
indoors at other locations
outdoors
11. During the time you spent indoors at home during the monitoring period, how much time
was spent with
ENTER HOURS AND MINUTES
HRS. MINS. NONE
a) any heating or air conditioning in operation?
b) windows or doors open?
c) windows and doors shut, no heating or
air conditioning in operation?
12. During the past 24-hour monitoring period, did you spend any time in areas where fruit or vegetable
crops are currently being grown? If so, how much time did you spend there?
YES
NO
HRS.
MINS.
13. At any time during the montioring period, were you in an area that was bneing sprayed for mosquitoes
or other insect pests? If so, how long were you in the area?
D
YES
NO
HRS.
MINS.
108
-------
14. Please indicate any OTHER things that you did or that happened to you which brought you into
contact with pesticides, insecticides, fungicides, or herbicides during the 24-hour monitioring period,
and the length of time involved.
15.
Do you plan to make any application of pesticides in the next several days?
I I YES -+ When? | | II -» GO TO Q. 16
NO ->
16.
THANK RESPONDENT FOR THEIR COOPERATION AND END THE
INTERVIEW.
Would you be willing to allow us to observe you during that application and wear a pair of cotton
gloves that we will provide to you to measure what gets on your hands?
YES
MAKE THE APPOINTMENT
Month:
Time:
Day:
a.m.
p.m.
NO
THANK RESPONDENT AND END INTERVIEW.
17. Would you like to receive a copy of the results of your monitoring sample analyses?
YES
NO
THANK YOU FOR YOUR COOPERATION.
109
-------
-------
Appendix B
Summary Statistics for All Analytes
111
-------
O CM O CM
v£> CO
CM CM
CM SO
tv CM O
o r^i os r/» o
H CO sO
!_
a
C
in co
O^ CM
Ul CO
CM Ul
CM
00 O O
•8
O F»
'tn-t
CM CO
co
in co CM
- in
in
tn
•8
112
-------
i-
a
rH Os
O CM
•*• O
M i-H
Os Os
Os CM
s± rH' SO i-l
in o
SO rH
03 10 H Ifl
•H CO
M W
O O
O so
M si-
O
IT)
I i.
O
I
-H
8
D)
Os CM
M sO
CM Ul
Csj CSJ
CO
tfi
CO
sj- si-
CO CM r-H CM
OS
sO CO
CO sd-
o
sO
si- CO
cl
f- O CsJ O
O SO O so
Kl I
Ifl
O ifl O Ifl
si- co so CO
M so
' sO O sO rH
i co r-.
co
CM
Os
si-
co o
X o x w
•H
ra
c
g
ra
H--
O
m
H-1
I
2
O
O
C£
O
31
o
45
o
01
•p
a)
a
_Q
(I
i
113
-------
CO
g
a:
a tz
UJ M
c
a: ui
2d
Vt 0
in
a :«:
Z 0
M TJ
ui z a
t^ i_| *j
«t c
S W -H
M -J-H
1- UJ C
w > o
UI UJ U
0 ~
UI UJ
w r-
ui u
S. ui
1-
• UI
rH 0
1
ca g
UI U-
s
H
•
•H
m
i
UI
t_
o
0.
•
I_
o
o
2
i
I,
8
•g
1-1
i.
a '
•*••
c
• H
Z
g
• rl
.£-
a.
w
i.
3
(/I
L.
0)
"c
2
O)
£
• H
i.
W
{.
i
|=
c/>
fc
01
•H
C
s
i?
• rt
t-
Q-
t.
o
§
3
V)
r-. co
ON sO
CO
r-l W
CO NO
o -d-
rH ON
\*
•
rM vO
r>- IN~
*
ON <)•
CM
O r-l
(H FN.
CM
CM -a-
CM .»
ON
CO rH
o ui
IN.
M CM
CO UI
UI
0)
X VI
M CO
O NO rH
rH IN.
ON Ifl
r-J NO rH
Ifl IN.
CJ i*.
ON CNf rfl
UI NO
ON NO
CM IN. O
in IN-
rH »d"
* .
COM CM
O IH
ON IN. O
IN. NO
CO •*
IN^ %± H
CM rH
ON UI CM
CM IN.
IN. CM
rH U| CM
-------
a
co
CM O CO O
Is- M CsJ tA
M Is. O. !>.
10 CO IT)
> «0 M CO i-l
Is- so
W
f*- tH M rt
tH
f-.
CO M CO
CO
OL
O
CSL
a. uj
. r»
K\ r-f 1^. ,_)
CM Ifl N. Ifl C\J
M vO IX
Is- M MM
* rH ui rt o
00 00
o o o o CM
O is.
N CM CM CM
a*
M r-- is. is.
*& CM M CM tH
CM U) CO UI
CO CJ iH CM CM
CO l-l - fs.
O O 00
00 O O , .X w 5
Q) 4)
M ro
s
<
a:
it
o:
o
—i
x
o
0)
a
D:
a
•8
•H
' O
Q)
•*•*
I
115
-------
en
C
• H
S_
a.
V)
rH CO Os CO
CO tn rH M
o o o o
00 00 rH
o rs-
sO CM v± CM
OS , sO
rH O Os O
CM CM • r** CM
CM CM h» CM"
CO tf\ CM tn
ui •* tn
co so
(/)
o:
o
o
c: i
«x .
z <
** •> ~
1— UJ C
« > 0
uj uj o
O C3
O O
O O
O C3
O O
O O
O O
O
O O
O
O O
o
o o
o o
^ vO
o o
M tv
o o
o
1-1
r~. rs-
UI UJ
I 1
X tf>
C3
-------
a
co
o o
o o
o o
o o
o o
o o
o o
o
o o
o
o o
o
CO s£>
o d
o o
o o
CM sO
sO M
!•«. CO
o o
o
l-t
o
U1
CO
C£
O
ct
Oi OL
UJ M
O
C£ UJ
< -J
a -i
2 HI
O O
o o
o o
o
o
r-.
o o
o d
o o
o
rH
to
a ii
z o
< <
-5
8Z
1- H
<
•z. co
h- UJ
CO >
UJ UJ
1
0
UJUJ
3u CO
O) <
Hi ^
UJ O
S UJ
•D
01
1
o
1_
o
o
"a
1
E1
• H
a.
CO
J_
OJ
V)
I_
Cflg
UJ U-
CO
H
a
co
o o
o o
o o
d o
o o
o
o o
o
o o
o o
o o
o o
o o
o
I-l
o o
o
CO sO CO sO
O O Os o
00
o o
O !•-.
CO r-i
o o
o
I-l
CO i-l CO iH
OJ
.V tit
OJ 0) ~
x ui N: o —
§
X
u
I
-H
OJ
•
<
a.
u
%
.. J
J2
ra
117
-------
CM O CO O
CM CM h- CM 1-1
ON r-
in r- in f»
|H O CO O r-l
CM CM r«- CM
ON
Is- CM
N± M
O O
O O
UI ON
NO NO
in
CO
O O
O
UI
O
Ifl
ce
o: c£
ui M
D
fy iti
< -J
O _)
t>i
CO
o ^
W Tl
Ul 2 g
< M C
s: co -H
I-l -1 -p
K 111 C
CO > O
Ul Ul O
Q ~
Ul Ul
3= CO
CD <
I-l 1-
Ul CJ
Su;
. UJ
H a
ig
UJ U.
CO
i.
o
•8
i.
o
I
M
D)
C
• H
i.
a
co
H1
• H
£.
a
CO
o o
o o
CO M CO
o o
o
co m CM m
l-i |-J CO rH
* CM
ui
M CM
M in
O O
O O
CO NO
N.
•d-
o o
o
CO
ON
I-l O ON O
CM NO
ON
I N± ON N±
id wo CM
CM CM FN* CM ON
ON N^
IN. IN. W r^-
ON CM O CM rH
O O O O
00 00 CM
O IN-
Nt CM \0 CM
co in i-5 in ON
0) 0) —
N: o x o —
(D
-H
H g
UJ -H
Q. Q)
-j a
o
45
o
01
-H
-8
4!
J3
4?
o
•8
•g
118
-------
CO rH
CO Kl
Kl CM NO CM rH
ON f*
ra
c
o
(/)
01
D.
i.
a
co r-
-------
tt)
a.
g
iri
£.
a
o o
o o
o o
o o
o o
o o
o o
o
o o
o
o o
o o
I O CO O
CM CM t»- CM rH
CO rH CM rH
r- rH rH
(T- f-.
..O
Ul
v± vO M> vO
CO i
O"
tn
so
i
Oi Oi
UJ M
O
n* UJ
O —1
Z rH
^1
in
§S~
V) T3
UJ Z C9
S If) -H
KH -I-H
h- UJ C
M > O
UJ 111 O
o o
o o
o o
o o
o o
o o
o o
o
o o
o
o o
o
o o
o o
o o
o o
o o
o
o o
o
o o
o
o
•JD
UJ Ul
I- -J
M I-
rH Q
I
co a
o
UJ U.
CO
I.
o
-8
o o
• i>
00
U) rH
00
O O
o o
• •
00
O O
o
CM O CO O
CM CM r- CM
O CO O CO
o r«. o CM
0s r*«
CO 10 CM Ul
CM O N. O CM
0)
N: w
fl) — .
w —
.V M .X
1 2
ra
c
£
o
u
o
ai
•**
I
UJ
a.
•i
0)
i-H
_Q
0)
"O
120
-------
i.
a
CM Q CO O
CM CM rv CM rH
O^ Iv
03 CO CM CO
O O O^ o
CO •*
CM O
Ifl Kl
CM M
O ••£>
tfl
tn iv
ifl
U] Kl
rv KI
Kl rH IV i-(
•d- ^
2.-
O
c:
a: a:
at M
O
C/)
UJ C
> o
uj o
a
UJ UJ
1- _J
X 03
CD <
1-1 I-
LU U
3 UJ
I—
• UJ
|H O
I
ed cz
O
UJ U.
9
i.
O
I
O)
C
o o
o o
o o
o o
o o
o
o o
o
o o
o
CM O CO O
CM CM rv CM rH
O CO O CO
Kl O IV o CM
CM •
fv
Kl
CM
IV I
CO
in a-
CM
IV
O
•-D
O CM O CM
Ifl
K> Ifl ^O Ifl CM
ui -d1 rv
rv CM M CM
CM
vO
0) 0) m
N- VI .NO £
0) a) 2
X n X in —
>
r-H
ra
«t
i-
LU
cc
-Q
45
HI
a
uj m
Z -H
< O
O Q)
o; -t^
o 0)
-i o
u
.a
ro
I
121
-------
s.
en
c.
• H
a
in
i-l O ON O
\O *jj" iT/l ^f
mo u) o
CO Vf 'iH ^J"
o o
o o
o o
o o
o o
o
o o
o
T-l
D: OL
Ui H
o"*
o: uj
-------
rH O O- O
CM IT) r-. 10 rH
r-i co Is-
M UI M rH
O"* f*»
h- C\l
O O O O
MM IN M i-l
IN «M is.
o-. in o in o
CO . rH UI
C£
O
o:
UJ M
< -i
o —i
a ^
z o
< <
r- M 3
< C
S CO -M
M —I -P
feSfS
UI UI O
—I —
a
UJ UI
i.
o
|
3
O
o o
o o
o o
o o
o o
o
o o
o
M CO
CvJ r-<
O O
t- CO
I CO.
O
M h-
UJ CJ
3:u;
• uj
rH Q
I
CO QZ
O
s_
a
v>
3
W
M ui r- in
M C\l t~- CJ rH
0s t*>
Is-
Ifl
O -f rH
Cvl r-.
in •* in
M vO
CO
rH CJ C^ CM
S" rH •* i-! O.
rH tf
X
o
a
i
-Q
(I
01
a
UI
X
a.
a.
o
g.
45
a
01
U
01
H-"
I
123
-------
Ol
i.
a
CO
l CM
•* CO
CO CM
0^ f- i-i r-
CM N" • t*-
O > O
ui ui u
i ^
o
Ul Ul
1— _J
X CO
CD
CO
o
CO
m in
1^. vO
X
-------
0)
O.
H1
a.
V)
o o o o
o o o o
o
rH
O O O O
00 00
o
CM CM f>- CM O
Os U)
M Os Os Os
CO H N.
u) o mo
CO rH l>
s* m so m
Os M O M K\
N. CM sO
i.
O
-8
W -rt
O)
C
• H
a
CO
00 00
o o o o
o
o o o o
O O O O CM
o r-.
o o o o
o o o o
o
CM M CO M
CO •
CO
CO
CO iH CO O
sO vO
.£S
CB <
M I-
UJ O
• UJ
1-1 a
i
ca a
o
UJ U.
J9
i.
O
I
s.
a
CO
i.
0)
£
O O O O
o o o o
o
o o o o
00 O O CM
O N.
o o o o
o o
o
i in -* in
CM
CO
in in m
CM
CO
Kl CM
CO
a o ~
x w x w —
0)
•f
>\
3
(V
0)
a
(II
1-1
-i
I
•8
125
-------
O)
E
•rt
i.
a
10
o o
O O
o
I O
O rH O rH
CM CM CO CM
co in
CO
ui in o
co ifl
to
g
a:
o: eg
UJ M
<
a
en ui
< —i
C/l O
CO
o *:
2; u
I- M 3
gco-5
i.
o
o
T3
I- UJ C
O
Ul UJ U
_1 —
a
UJ UJ
i.
a
c/i
£.
O
o o
o o
o o
o o
l-i
o
0)
126
-------
o o
o o
o o
o
O)
c
o o
o o
o o
o
s
3
CO
M
CM i-H
CO
££
O
o:
Q; Q;
LU W
O
a: LU
O -1
Z HI
CO O
CO
a ii
z o
CO
LU ~Z.
<
r: co
fr— LU
CO >
LU LU
a
LU LU
H- -1
X CO
CD <
LU O
s: LU
1-
• LU
iH O
1
CO K.
O
LU U-
CO
t-
•o
01
c
• H
c
o
o
~
t-
•§
-H
3
o
t-
o
-8
c
H
!.
.^_i
E
• H
O)
c
S-
Q.
'-
01
1
CO
s.
0)
c
tlH
2
Dl
C
• H
i.
co
L.
O O
o o
o o
o o
o o
o o
o o
o
o o
o
o o
o
o
vO
O O O O
o o o o
o
o o o o
o o o o
o
1-1
3
CO
r^ co
o o
CO
o
C 01
I/) E
ra 3
01 co
CO
01
J!
TJ
0)
X M .X M —
i. O
3 C
I/I
ra 01
01 i-
LU
s:
CO
LU
O
, 01
-f
. 01
"1
s
ID
01
•+- 0)
•H -C
D * h-
* *
127
-------
13
g
M
t_
C
• |H
s:
g
CM SO
CM i-l
ui NO ui NO
CO I
ON
CM
Ul
co
•a-
o o
O 0
o o
o
I-H
CM i-H CO rH
ON
CO
•d-
O O
O O
00
ON i-l
O O
O
1<
eg ui
Ul Ul
o§>
ce o
z u
£S5
CO -I
Ul
M i
Ul I
• H
g
• H
ft
O O
o o
o o
o o
o o
o
o o
o
o
Ul
ON
N*
o o
o o
o o
o
I-l
o o
o
o o
o o
o o
o o
o o
o
o o
o
rH
o
Ul
Ul 10
si
r- Ul
• r-
X
fl) -N
CM rH CO r-l - ON
ON >d-
NO X M
O O
O O
O O
O
0) —
I
•8
•F
O
01
H^
O
a
0)
rH
-8
•H
O
4?
0)
•u
ca
i
a:
a.
•8
1
g
u
<
X
.fl
43
0)
a
128
-------
_
0)
0.
D)
c
CJ M
CO ON
M CJ
O O
O NO
CM O
h* cj
co °
CO
•a-
co o
ON
CO M
CO rH
CM
Ifl
§ "*
Q; uj
UJ UJ
Q.
Q O
OL U
< M
a x
I- a
CO -J
UJ
a HI
Z u.
< CD _
CO M 13
uj o; o
I- a. 3
w o
UJ W O
_J _
a uj
UJ >
I- uj
o
|
3
O
o o
o o
o o
o o
o o
o
o o
o
o
Lfl
. co
CO
CO
cu o
I Ul
CO I-
UJ
uj a
ca o:
< o
I- u.
i.
O
o
T3
C
CO CO CM 03
CM fH 1^ r-i Ul
D)
C
ON O
r-t
O O
ON rH
CM IT) CM
O O ON O ON
ON ^>
01
i-H
O "S
X -»^
ca o
I Q)
u
01
.a
O 01
a: -p
o a)
•8
-*^
o
01
-H
I
129
-------
(0
§
(A
i.
o
a.
DI
in ON
in co
ON IN-
tf i-H
in ON
* cd
o I
Ifl
O 0s
CM rH
O ON
CO C£
o: n
o <
OS
a: ui
UJ Ul
o.
a o
o: u
a x
z o
r- a
CO -J
UJ
Z U.
«*§-
CO I-HTJ
< CO C
I-H Z -f
r- H C
CO O
UJ CO O
_J ~~
a ui
ui >
I— ui
<.
o
•8
i a
i en
K1 CO
CM rH
M ON
CO M
CO
O O
O O
O O
O O
O O
O
O O
O
O
Ifl
CM O
I UI
CO r-
UJ
Ul Q
CO IX
< O
r- U.
0)
• H
a
co
ON rH O rH rH
NO rH K> rH W
CM O CO O
O ^1* ON *d* ON
0) 0) 2
X 0) X W —
O ON O ON
CM rH CO rH ON
X O X W —
en
c
JD
(B
01
a
1
I
a
a
0)
rH
J3
•*••
o
•8
130
-------
O)
Ico1
O rH
O ON
ON W
CM r-1
O
NO
O Ul
IN, M
CM
Ul
CO
•a-
if]
O O
O O
- 10
O5
CO
O O
O
t-i
CM
Ul
1- UJ
X —1
CD
H 111
i-
o
o
-a
•H
fH i-l
o o
ON rH
o
It)
o o
o o
o o
o o
o o
o
o o
o
t-t
LU
ui a
ca a:
-------
Dl
C
• H
i.
CL
in
*'
in i
in in \i-in
CM r-l t^ i-l
o o
o o
o o
o
o o
o
ce
1<
D: ui
ui ui
a o
o: u
§«e
fc3
Ul
a M
li= I
2S"5
H- a.
S "1
co o
ui to o
_i —
Q Ul
Ul >
I- Ul
H Ul
Ul -J
^ CO
D)
• rt
a
CO
O O
o o
r-. o
1-4 r-<
o o
o
M O
CO rt
O
in
o o
o o
O C3
O O
O O
O.
o o
o
o
Ul
i i
m i
i tz
; o
D)
c
• H
i.
a
to
Ifl
so CO <* CO
0)
N: w
in
O O
o o
o o
o
rH
1X1
o o o o
o o o o
o
0) 0) ^
X O X M —
ui
a
ra
-H
O
Q)
•H
01
a
0)
>-i
St
o
a.
ui 0)
i—i
3 I
I— g)
a. a
Ul
3
I
-F
01
132
-------
o
I
g.
• ri
0.
10
O O
o o
o o
o o
o o
o
o o
o
CM
ui
CO
-a-
o o
o o
o
CM i
o o
o
O Cf
co .i-J
Oi UJ
ai uj
a.
o o
tt O
< H
a x
z u
< \
I- a
M -i
UJ
a u
to H TJ
uj ex a
1-0-3
< co c
£ -H
M 2 -H
I- W C
CO O
UJ CO U
O UJ
o
-§
•+••
3
o
c
• H
Z
D.
CO
o o
o o
o o
o o
o o
o
o o
o
o o
o o
o o
o o
o o
o
o o
o
z: _i
CD
M UI
UJ —I
S 03
• I-
OJ O
UI
uj a
OQ <£
1 O
Ol
c
o o
o o
o o
o o
o o
o
o o
o o
•* in
MM r>- M
o o o^ o
.X W
X
Ql ->
UJ
2
a
J3
(I
•A
-Q
45
u
I
0)
•o
I-
a.
0)
a
••8
133
-------
c
• l-t
2:
O)
c
S.
Q.
O
CM
r- M
o o
cd i-i
CO
Ul
<0
-a-
o o
, o o
o o
o o
o o
o
CM
Ul
CO
•a-
CC. M
s*
ex ui
ui ui
,s
z u
1- Ul
CO
M Ul
Ul —1
i.
O
o
5
5
a
• M
f
c1
• ri
C.
Q.
to
o o
00
o o
o
CO O
O
Ul
o o
o o
o o.
o o
o o
o
o o
o
CM O
I Ul
CO I-
Ul
ui a
o
|
M
s?
1 -H
I J-
i a
CO
(30
0 O
o
CM
O O
O
O 0^ '
co i-5
o o
o o
O
CM !
0)
X O
o o
o
CO i
cc>
01
a
•s
ID
•H
u
01
-K
134
-------
HI
-H
'CO
Cd
Ifl
o o
o o
o o
o
i-l
i-
0)
Q.
Q.
CO
CM M
i-l i-l
CO
CO
o o
o o
O O
o
10 £*
O <
D£ UJ
UJ LU
Q.
Q O
OL CJ
< 1-1
O X
•z. o
< X
h- Q
CO -I
LU
i.
O
•8
D)
a
CO
o o
o o
o o
o o
o o
o
o o
o
o o
o o
o o
o o
o o
o
1-1
o o
o
o
in
CD
M LU
LU _l
S Oi
CM O
I LJJ
ca o:
< o
i.
a
CO
CM
i-l
Cvl CM
•$• CO
CO Cx)
CO i-l
O O
O O
O O
O O
.V ui
O O
o
o o
o
1-1
a:
a
LU
M
a
01
a
45
o
a
I
O
>•
X
O
X
\-
UJ
(I
-P
o
: 0)
-f
0)
Q
i
•+-»
o
0)
-H
01
I
135
-------
O O
O O
«O
tO
o o
o o
o o
o
O O 00
o o
o
01
cr ui
UJ Ul
g
I- o
co -i
HI
HTJ
iu c: Q)
H n. 3
i.
o
•8
M O
UJ W O
o
M Ul
Ul — 1
E1
a
V)
oo
oo
O O
d o
o
o
o o
o
O p
d d
o o
o
1-H
o o
o
O
Ul
N U
I Ul
CO t-
Ul O
§g
I- u,
o o
o d
o o
d d
o o
d d
o o
o
o o
d d
o o
o
O O
o
flj 0) ^
X W X O —
flj
X 0
ro
-H
_i o
S 4!
8 &
M
J3
(D
o
HI
.
a
LU
D.
I
10
H
CJ
JD
(U
01
a
•8
j
01
•a
136
-------
i.
HI
CL
_
0)
-H
E1
II-I
i.
D.
V)
O O
O O
O O
o d
o o
o
o o
o
r-l
O O1* O '
CO "° ™
f- Ul
o- r-.
o
Ifl
co o:
OL M
o; LJJ
UJ UJ
Q.
O O
.
o
•8
CO M T!
LU CC 0)
t- O. 3
I- UJ
•x. -i
UJ
O O
o o
o o
d o
o o
o
o o
o
o
ID
co co CM co
in ui * in
l-l .CO
CM O
I UJ
00 I-
UJ
UJ
m o;
< o
i.
o
I
l-l
Dl
c
• H
I
.0 O
o o
o o
o o
0)
X w
00 iH
O Ul
o o
o
CO
xD CO "iH
rH 10
* i
o •
QJ ^1
N: ci —
•s
01
a
45
u
3
I
HI
O
_a
ro
•D
137
-------
CO O
co <^
O O
CM m
CM O
CO
CO
CO
O O
O O
CO
•a-
O O
O
O O
O .
CM
U1
a: uj
III 1U
a.
O O
o: u
gs
II
ui
O M
{2S"g
§S)|
PS?
W O
ui en o
o ui ~
ui >
o
|
o
.s
cl
o o
o o
CO O
CO rH
o o
o
o
m
o o
o o
o o
o o
o o
o
o o
o
l-t
O
M 10
UJ -I
CM O
UJ
uj a
o
I
M
cl
•* o
co ui
CM CM CO CM
o o 0*0
X
i-l
IT)
I rH I-l
UI
o o
o o
o o
o
ffl Q) ^
« x —
03
C
0)
rH
J3
o o
a ai
i +•
0)
•;)• o
•8
a
a
o
0)
a
•8
J
138
-------
ra
c
o
u
L.
0)
0.
i a
I V)
oo m CM m
w in i-! in
i-l i-l CO i-l
- Ul
CM CM
CM 1-1
N- in
tor,'
M LCI
r-I CM
r-l sO
CO
O iH O
CO ^> CO
o: iu
LU LU
a.
a o
ce o
< w
a x
z o
«t N
1- a
CO -i
iu
a H
2 LU
< CD
Z "-•
CO H T3
LU Of OJ
H a 3
< to c
z: -H
1-1 z -H
l- M c
co o
LU CO 0
_J *~
o ui
tu >
I- LU
X _l
CB
HI LU
LU -J
•z. ca
-------
M M>
CO -O
CM
in
CO
i-H -
tt\
Nl CO
CO
r-i o
O CT-
CM rH
co o^
cd o
O 0^
CO rH
co
-a-
CO CH
C£. H
S"
a. uj
ui ui
a o
a. u
§*e
I- Q
wsl
CO H 73
UJ CS O
gSsg
o
o
-a
H H C
CO O
Ul CO O
.-J ^~
a ui
ui >
t_ ui
co
H ui
ui —i
SJ
c-j u
I Ul
CO h-
ui
ui a
—i
ca oz
D)
C
a
co
r^ o
CO rH
O
-------
o o
o o
o p
o o
o o
o o
o
o o
o
I-i
CO CM CM CM
Os sO O so
sO r-.
so in
CO
CM
Ul
CO
si-
as uj
111 111
111
O HI
Z U.
W M "
uj a^ aj
h- D. 3
gw.S
M Z -f
i.
o
•8
-H
3
O
O O O O
00
o
O
U1
Q.
OT
O O
O O
O O
O N.
CO so
CM O
CM sO
o !•-
CM so
CO O
ft. sO
O
U|
UJ W O
—I —
a tii
-
O
HI UJ
CM U
I UJ
OJ H-
UJ
uj a
_j
cfl a:
< o
I.
OJ
-H
C '
• H
3C
*
O O
o o
o o
o o
O O rH
q in
o o
o
OS
s*
ffl 0) ~
-V w N- w £
o U) a- 10 rH
r-l CO U|
sO If) H\ Ifl Os
i-i (O •it
0) 0) ~
X M • X HI —
I
<
cc
o
rH
_Q
0)
a
0)
•-H
i
Z -H
1 j
N d)
< a
141
-------
Dl
C
o o
O O
o a*
CM r-i
O O
O
o o
co r-i
o o
a o
O O
o
*$• •&
OJ
IX)
to a
a M
o <
Of
(V nt
UJ UJ
Q.
0 O
§x
u
H O
CO ->
UJ t-
Q H O
Z U. O
t UI
s
H UJ
UJ -J
i.
a
c
>H
s
H1
• H
1_
o o
o o
p o
o o
O .0
o
o o
o
o
U1
o o
o o
Kl !*•
ui H
O O
o
CsJ O
I UJ
CO I-
UI
uj a
_j
co o:
PS
1
O)
c
o o
o o
o o
o
O C3
O O
o o
o
CO CO CM CO
i-l iH CO r-5
CO CO CM CO
rH r-i CO H
OJ
N: o
X —
0)
N: o
E
<
CD
a:
•8
01
a
OJ
rH
ra
-H
g
-H
I
O
H
H
<
-8
0)
a
01
i-l
I
142
-------
ID
C
o
Ifl
i.
01
Q.
O)
c
a
co
o p
00
O O
O O
O O
o
o o
o
to a:
tx. u
o <
OL
a. iu
UJ UJ
a.
a o
OL U
< H
,° X
z u
0,0
o o
o o
o
O
Ul
uj o: o)
I- a D
i co c
2: •!-!
.
o
-§
-H
3
O
01
• H
i.
a
co
o o
o o
o o
o
UJ CO O
.1 ^
a uj
UJ >
h- LU
X —I
cs
l-l UJ
UJ -I
s: 00
<
CJ u
a:
o
!_
O
|
o o
C) O
o o-
O
-------
r-4
19
C
O
UI
i.
0
Q.
i.
a
• rt
x
B)
C
• rt
i.
Ci.
w
£.
a
g
V)
CO
M
CM
in
rH
O
•a-
o
•*
t-.
o-
rH
f»- a
<*• o
o o
i-t 0
CM O
CM
CM O
tfl O
*O o
tn
o
0
^
rH
O> rH
^
O
O
vO
M
CO rH
t*~
O
O
O
v£>
•3- 0
CM U)
f2
d
CO
o
CO
CO
o
CM
d
S
d
CO
CM
o
o
o
o
o
o
o
0
o
d
o
o
<0
rH
**"
O
O
CM
J"H
•^
a
o
CO
CM
vO
CT>
M
d
tfl
0
*
o
CM O
O O
CM O
O O
M 0
CM O
d d
o
CO
0^
rH
O
0
CM
rH
rH
r*.
O
O
CO
U
MM 0
r- r- O
ft . 01 ~
K 0 i.
< Z Q)
S. °"S
co a: „
M O
Q . 0 0
CM *d* O O
vO rH O O
*° *° ^ O O
rH O O CM
M rH
f~. CO O O
rH CM 0 0
rH O O CO
CM
- N- o o
M rH O r»-
o o o t-
CM
•"O f*- O O
CM M O O
r4ddrH
CM
sO
ca
i
UI
N
Z
1U
I
i ui.-O X "
• 0) ID C
; rt irt
uil) X •-•
144
-------
I
CM O O O
in co c?v o
CO tft r-t vO
CM r-l
*""
CM M O O
o cr. o o
vj* vO O O
m M o o
CM r-l 0 CO
CO
r-t
r-
M M O O
OOOO
CO K) O O
rH vj- O O
vj- U1 O IH
vO M i-l W
vO r-l
>-
Ul 0 O 0
v£> f{\ Q- Q
M M 00
O O O C1*
0 O O M
OOOO
o o o a
vO r-t O O
O CM O O
MO O CM
CM r-t ,-t
o o
O •a-
o
in
i-f •* O O
i-» vO O O
co vo a o
oooo
O O O M
rt
C/>
U
M C/l
<5 i.
MM O
(- i- O
<< -a
r- CC ~H
C/5 r- 3
"Z. ^ o
a: '_) i-
< Z 01
s o •+» —
s u o -a
HOC
Q < .rl .rH
IU JQ -I-1
r- UJ 3 C
X —I O O
U _J \ 0
^|5
! 0
•^
01
— I.Gi
? ""
O
O
•g
^_{
s.
01
-H
c
i?
g
5.
*
i.
D
£_
GJ
•<-*
C
w
?
• H
t.
a
w
i.
a
5
Cfl
O v£> O O
vO CM O Kl
O O O vO
O
r«.
C?v r-t O O
•vf i-l O O
O O O Qv.
CJ
r-l vO 0 0
, IA CM 0 0
r-l O O CO
CM
0
CO O 0 0
Ov NT I— 1 O
1C) CM CM Lli
T—I
f»*
O CO O O
•* 10 O O
M U! O O
r-l K\
in CM
rH iv.
vO f^. O O
r-l U> O O.
O r-l O [^
O O O ID
CM
f.
CM lf\ O O
CM r-l O O
O O O M
rt
W) Ifl-O O
r*" M o o
vO N^ O tf\
CM
in rH
r-l 0 0 0
CM CO 0 0
CM 0 0 t4
10
CM
h"
Ifl Ul 0 0
CM CM O O
LO vj- O vi-
vO
CM Cv
*
Ifl vO O O
fx. C^ O O
CM O O v*
CM
0
CM vj- O O
f- co a o
O H\ O O
rt CM
CM CM
frt M O O
C4 vO O O
O O O f~
M CM CM
vO O
v£>
r-l CO O O
C%J ^J* >^. O'
CM r-l CO vj-
vO r-l
Is.
M CM 0 O
ov ^ o o
M O t>- O
ui iet r-i r-
r-l M CM
CM fs.
IT. rH O O
M r-l 0 0
M O O O
vO vO r-l O
r-l . vO CM
r-l vO
OOOO
OOOO
OOOO
o
a o o o
oooo
0 CD 0 CJ
CM
f.
vo in o o
O O O CM
O O O vO_
. o-
0 00 0
o o a o
ca cJ a a
r-l
h-
oooo
oooo
oooo
CM
IS.
vO, cvl O O
r-l r-l 0 0
oooo
CM
CM
o
-r>
>.
CO
I
(n u •
I
o
—i
X
u
as as
x ^
c ..S-.5
ra iu TJ x
c -..§.
§UJ TJ
• O
2:
-------
r-l
2
o
I/}
0)
D.
(.
0)
c
3C
g
• •-1
to
£.
Si
2
CO
in MJ o o
i-< O O O
co r*- M> o
rH CM m CM
H rl£
in o^ o o
f~ * o o
CM m m o
co KI co in
r-l i-l r-l
CO O O O
Kl i-l 0 0
O O O O
co in VD ^
CM i-l !*• Kl
CM vO
' M> CO O O
co in o o
in i-i o co
^° /-i
r--
^ o o o
u\ r-l O O
CO CO O O
0>, o O O'
rH rH CM
O Kl
I-l M>
Kl O O O
CM rH O CM
O O O Kl
^
^
O O O O
o o o o
o o o o
J_l
^
Kl I1- O O
•xO vd* O O
o o o t-
rt
•
Ml vO O O
O O O H
O O O Kl
i-i in o o
in M o o
do o r^
H M 0 0
vO CM O O
O O O CM
8
ti §
I/JO J-
t~4 1— 1 O
SS -8
r- CC -K
"'S S
>• Ul ->
a: o £.
I§"S9
3s||
Ul Is -H
I- Ul"3 C
§-1 O O
_1 V 0
• CO
1 (J
01 <
Ul
gg
s-
o
1
H
t_
a
-H
c
*•
g
• rl
£.
c^
£_
|
C/>
£.
Q)
•f
• H
31
0)
• rl
£.
a
3
CO
>d* CM o o
CM O O Ml
CM
o
fi
in KI o o
>i- m o o
Kl O O •#•
CO
£
vO Kl CO O
MJ Kl CO vO
r-l O
CM O
CM CO O O
Kl Is- O Kl
d * <> KI
CM r-l M> *
rl 0 r-l
rH Is"
KI m o o
' W O O^ O
o * N- in
CM Kl CM
«d* t^-
o r~ o o
SO tf CM O
Kl rl i-l CM
CM vO
rl O O O
O O O Kl
O
r-.
o o o o
o o o o
o o o o
M
CO 0s O O
rl 0 0 CP-
O Q O Kl
O
o in o o
C^ CM O O
M> rH O M5
O
rH i-H
I*"
O O O O
CO O O O
vO K> O O
CM
Kl CM
*d" O O O
Kl Vi- O O
rH CO 0 0
Kl rH
-------
rH
S
o
w
i.
0)
o.
, 0)
"c
s
O)
c
i.
t.
Ci
S
w
o
o
0
o
o
0
0
0
o
o o
o o
o o
0 0
o o
0 0
o o
0 0
0 0
o
0
o
£
o
o
o
rH
o
o
o
M
vO
O f-
rH 0
0 0
I-. N.
O O
O 0
0 0
0 0
o"o
o o
o o
rH'
0 0
O rH
O M
rH
O O
0 0
0 0
o
ui
o
CO
o
o
0
*
" 'o
^f o
o o
CM O
CM 0
O O
tA O
«c!" O
O O
o
o
N.
rH
f-.
O
O
!>!
r-l
^
0
0
o
CM
O
UI
N.
tft
0
o
o
CM
O
rH O O
r^ o o
o o o
r-{
O O O
O O O
0 O 0
rH
N
N. O O
!*- o o
O O IT)
CM
o
w
cy)
o
§H '/)
s.
H W O
£1 -• •§
W 1- "3
z o
>- UJ —
(K C
*r
05 <
-5
UJ
co o
2 "•
r-
i.
O
O
c
HH
1.
0
H-*
c
.ri
"*
I
L.
Q)
I
5
«
•
£.
(U
"r5
S
g
• r*
L,
O
i
rrt
OOOO
OOOO
oooo
o
"*"
o o a o
o a o o
oooo
CM '
oooo
oooo
o o o o
o
^>
h/1 M O O
O O O IT)
O O O vO
1^]
oooo
oooo
oooo
CM
O O O CM
CD 0 (D UI
CM
VO
OOOO
OOOO
o a o o
0
tN-
oooo
oooo
o o o o
CM
oooo
oooo
oooo
f^
*
CO CO O O
oooo
0 0 0 rH
CM
r^
O CM O O
CM •& 0 0
JM rH 0 •*
CM CM
^
Lf> tft O O
CO CO O O
M^d*
i ry.
"*
oooo
0000
oo ci o
0
r-
vO O O'O
m KI o o
O O O rH
CM
CM
CM i-H O O
N^ M O O
O O O M
rH
*
rH CO O 0
vO -* O O
0 0 0 *
CM
rH
CM O O
SO M O O
* • O . O O Ut
vO
CM
N.
CO O^ O O
^ ^J" O O
CS C3 0 M
CM
Ov
>*
O ^ o O
CO * O O
o o o *
•3"
o
oooo :
oooo
oooo
fj.
CM CM O O
O O O CM
O O O r-l
. *
* rH O'O
UI CO 0 0 . , .
CM O O CO'
rH
^
OOOO
OOOO
OOOO
CM
N.
O O^ O O
CO CM 0 0
rH r-H O CO*
ON
-------
• rt
B1
•H
I"
CO r» o o
ts.t-.r-o
SO
S±
so <*• o o
M K> O O
i p-t o CM
CO
rH
r-
i-t rs. o o
I-I • Ul «
ill?
V> C£ 3
HOC
a < .n .n
U) .Of
fc UJ 3 C
x — i a o
Sw^i-
^|5
*? O
UJ
p
o
•§
H
t-
a
•*••
*rt
•*
w
c
Sf
£.
(U
3
cy)
L
•H
£
S
01
£
'2
a
V)
t.
E
I
CO Ul O O
p^ KI o o
o o o o
CM
o
*"
Kl K) 0 0
O O O so
O O O Ul
CM
CO If O O
so rH O O
O O O CM
O
sO
so in o o
rt CO 1- 0
|s- H Kl t^
*** rt
h*
SO Kl O O
CM Ul O O
cd CM o i-i
sO
CM
N
Ul i-I 0 O
so si- so O
•d- CM ui r--
5cM
K» K» O O
I-! i-I O sT
0 CD 0 t-
o
o o o o
o o o o
o o o o '
CM
N.
o a o o
o o o o
o o o o
SO
CO s* 0 0
rH H O O
O O O !•»
1-1
r—
CN t>- O O
CM CM O O
o o o in
in
CM
^
CM O O O
CM i-I 0 O
O CD 0 t*
I-I
CM
o o o o
o o o o
o o o o
o
h-
o o o o
o o o o
o a o o
CM
(s.
0000
o o o o
a o o o
s*
o o o o
O 0 0 0
o o o o
r-I
^
Os i-t O O
o in o 1-1
i-I CO
LC| CM
o o o o
o o o o
o o o o
Os
S*
o o o o
o o o o
o o o o
s
o o o o
o o o o
0 0 0 O
CM
o o o o
o o o o
o o a o
o
SO
s* st O O
M CM O O
H i-< 0 CM
sO
f-1
H i-I O O
os m o o
H' ° ° 8 CM
Kl s^l* O O
LC| Kl O O
O O O Kl
CM
H
S 2,0
Ul
W
a
£ . .rl .rt
(0 UIT3 X —
" • .0) _flj £
X
o
£ • .rt .rt
ra tu t3 X •-;
a • o) ro £
TS: vs:?: —
o
u.
o
u
ui 13 X ^
• a ra £
M
I
B
ra s
.rl .rl
148
-------
r-i
§
01
a
01
•V
C
• M
s
H1
• H
i.
w
O
£
1
O so O O
lit s± 0 0 .
0 d 0 tft
CM
s
CO 0 0 O
CM t-
is.
•
0
0000
o o o o
o o o o
o o o o
o o o o
o o o o
• • • •
OT
o
W M
H 2
W O £.
gp §
c P 2
wi ' 8
CZ O £.
*t iZ Q)
E 8 "S -S
3 E (D
M o: 3
HOC
O o o
Sri^,^
III ^ (™
3T 2 —
o
(2 <
~J
UJ
-1 Oi
ca o
< u.
H
i.
o
1
M
1.
_g
C
»r<
TO
• rt
t.
1.
o
E
§
0)
•f
1
C
a
w
L
1
w
0 O 0 O
0 0 O 0
0 CD 0 0
o
*^
0000
p O O O
o o o o
CM
O 0 0 0
o o o o
o o o d
0
so
0 -* 0 0
co r-. a o
O O O N.
M
N.
rs- tfi o o
O i-l O 0
r*t O O ^O
U)
CM
M OS o O
•* M 0 0
o d d .-5
M
CM
sO
Ifl r-1 O O
N} Os o O
is. ui o in
CM CM is.
•H O
fs»
f> tfl O O
Os i-l O sO
SO
CM (SJ
hi Is. O O
s* Ifl 0 0
CO r4 0 CO
M CM CM
so O
so
rt O 0 O
M r< o o
O I> i-l O
CM co o in
CM O rH
CM Is.
in co o o
U) IH O 0
ui d r-5 d
•* s* in co
CM M CM
st fs>
o o o o
. O s^t* O O
s* sO >j- O
CM CO CO CM
M 0 CM ,
M so
o o o o
o o o o
o o o o
o
^
o o o o
0000
o p d d
CM
• • • •
o
si- sO O O
W Nl O 0
do d r-5
I-l
is.
1-1*00
o in o o
1-5 0 0 M
r-l
CM
Is.
o
o o o o
o o o o
o o d d
o
o a o o
0000
o d d d
CM
Is.
• . •. .
a
o o o o
o o o o
o o o o
s
o o o o
O 0 O 0
d d d d
CM
Is.
* " * *
o .
o: c 5
y _ <5 s
Q. C ••rl.rl
I <5 Ul T3 X "
UJ
2
<
a
S 111 T3 'x -s
I
si-
*%
si-
i iu-a x "•
• ID ra c
w s s >-
o
§
si-
A
si-
149
-------
r
ir4
L.
Q.
Vt
IN.
O
•3"
O
•
co o
I"A O
O O
NO O
Kl O
O O
• •
0
0
in
1-1
rH
IN.
O
O
CM
IN.
•
0
"CO IN-
NO U\
ON rH
Ul M
W CM
- o o
0000
C-J tft O O
o o o tf
CM
r-.
IV) CO O O
CM m o o
S*3»
r- lil 3 C
S -I 3 0
1-1 O O
vO O O
CM i-l O 1/1
M 0 0 0
co m o o
I-l r-i O O
CM CM O O
8
o o o o
o o o o
o o o o
o o o o
0000
o o o o-
o o o o.
0 0 0 0
0
IN. CM O O
r-l rH O CO
O O O
CM O O
CO NO O O
Ul CM CM N±
CM NO O O
in o o o
CM ON CM O
NO Ul Ul IN-
rH M r-l
rH IN.
NO CM O O
tft O O O
M r-l O CO
NO
•§
H
Ul ON O O
in CN! o o
o o o m
I fN, O O
• CO O O
': 10 M CD
• rH Nl* N±
CO N* O O
CM CO O O
CM CM Ul O
CM ON NO M
CM O CM
CM IN.
NO CO O O
Nt NO O O
m CM o ON
CO
N* ON O O
O Kl O O
NO NO CM O
ON iH IN. >±
O ON
>-i -i-
NO ON O O
•• Nt^ ^— O O
CO IN- Kl O
CM O iH CM
Ul CM i-l ON ON
Nl NO O O
IN. IN. O O
S' ON CD C3
in o
rH <*
a
I
X ~
• o ra c
W S S *-
IU
o
x
S
§i
o
a.
§
a.
111 T3 X m
• o ra c
UJ
£0
150
-------
r-1
«
C
Perso
JL
0)
H-1
•r
Dl
L.
Q.
">
C
Q)
s
3
CO
o o o o
o o o o
o o o o
rH
IN,
O O O O
O O O O
0 0 0 0
i-l
Iv.
M * O O
r/1 M O O
O O O NO
rH
O
in
ON O O O
ON CM O O
CO ON IN. o
CO CM CM M
M i-l
m i-l 0 0
NO rH O O
CM tA NO O
i-i CM rO NO
I-l NO I-l
i-l IN.
£.
< 2: 01 m
2£ O Q) TU
~ § ° c
D -s: -rH -iH
IU _Q -J
r- UJ 3 C
X —I O O
CB -I \ C)
W M O) —
UJ > C
^i"
1 O
CO <
Uj^
03 O
< lu
r-
O
' i
M
01
~r>
'£.
gi
'£
M
1;
CO
0)
C
• H
?
• ri
£.
0)
s
3
V)
O O O O
0000
o o o o
o
rv.
o o o o
o o o o
0 0 C5 C>
CM
o o o o
0000
o o o o
o o o o
o o o o
0 0 C5 0
rH
O O O O
O O O O
o o o o
CM
IV.
O O O O
o o o o
0 0 0 0
ON
%t
Ul CM 0 O
fv. O O
rH O O in
CM
N Iv. o O
m NO o o
CM «1 O
CM CM Iv. rH
* rH CM
NO
O O O O
O O O O
O O O O
O
0000
O O O O
o o o o
CM
fv
M * O O
CM ivl o O
O O O i-l
rH
0000
O O O O
0 0 0 0
rH
tv.
tA O O O
•d- in o o
o o o Iv.
ON
CM
rH O O O
i-i W 0 0
CO NO O O
rH ON
ON M O O
rH rH 0 0
O O O M
rH
O
Iv.
s
O O O O
o o o o
0 0 0 C3
CM
rH <4> O O
W rH 0 0
O O O rH
CM
O NO O O
* -o o o
O CM O O
i> rH
rH ^
LCI ON O O
ON O O O
* O O O
rH rH <}•
CM CM
Iv.
•a- o o o
CO M O O
O & O O
CM rH ON
CO ON
rH H
§U1 TD X —
• 0) TO C
S CO S £ -
v. C • -ri >ri
o; ra ui T> x —
2 2»i££
o:
2
z
o
M
81UTJ X *-<
• 01 ra c
s w E: z —
151
-------
o o o o
0000
0000
o o o o
•d- o^ o o
tf M O O
L.
o
•8
•F
a
M O C
I < -ri
-------
H1
• r4
I
O v£> O O
I-! U) O O
CM i-i O Ul
O
1-1 N. a o
(^ r-s. o o
03
•$•
O O O O
a oo o
o o o o
Nl tfl O O
o o o in
CSI
in
M M O O
O O O Is,
o o o o
o o o o
CO
H 1-
W S
Pz
£S
w z
>• u ~
< tu 0)
S H +•
(/> HI
UJ O
Q D. -H
1U O jQ
fey 8
Sg th
UJ X C
I
3
O
H1
• rt
0000
o o o o
o o o o
o o o o
o o o o
O 00 O
o o o o
o o o. o
o o o o
o
in
o o o o
o o o o
o o o o
o o o o
•a-
o
m
Os
sfr
• UJ
•a- M
I U-
oa g
Of
o
Ifl O O O
•» M O O
w o o
* O O
•* o
o o o o
o o o o
SO U) O O
l-l rH O O
O O O CO
l-l
in
CJ O O O
r-l rH O N-
O O O O
0000
OS
>*
u
s
il
C * *H iri
as'ss 2
£
-------
CM O O O
v* O- O O
U) v
CM
Ul
CO M O O
OOOO
CM
Ifl
CO O O O
Ul CO fv O
r-t i-i O O
O O O vD
O O O Kl
L.
a
a.
E1
• H
*
i-l O O O
r- r>- o o
o o o CM
fi
N» r-l O O
CQ ON O O
m co o o
t*. r* O O
-d- CM O
Kl M
i-t rH O ON
O O O vO
CO
•3-
O O O O
o o o o
vo >a o o
r~ •* cs cj
o o a CM
ui
co m o o
oooo
oooo
oooo
o
Ifl
I
• rl
tl
oooo
oooo
O I-~ O O
<* CM O O
vo vO O O
CM i-t O CM
oooo
oooo
I
• H
r-l CO O O
ui o o o
O CM O O
O O O CM
C^ O O
CO M O
CM CM O O
O O O CO
ui
• rl
cl
oooo
U| Ul O O
O O O UI
OS Ul O O
O O 00
O O O Ul
M
f~ o o
* vi- o
>i- tn M
CM Ul
CM
ON
If
CO vO O O
H rH O CO
ffi
(0 O
£«££ £
0) Tl X "
154
-------
i^. in o o
05 M O O
Ul rH O O
CM
O Os O O
m r*- o o
<0
CM
UI
CM ffl O O
CM i-« O r-t
0000
O O O O
LA CM O O
CM rH O O
O >* O O
VO Os O O
CM |H O NV
rH
CM
lit
CO
-a-
o o o o
o o o o
o o o o
o o o o
y STATISTICS
CONCENTRATIOi
0£
vo o o
co in o o
o o o •o
CM
C O O O
0000
o o o o
o o o o
0000
o
m
o in o o
in rH O rH
O O
- o o
U) O CM
rH
m
CO i-! O O
CM CM O OS
O O O O
o o o o
;-a- o o
CM O O
o o o m
O fH O O
s0 Os O O
rH O O CM
tfi
m
O O O O
0000
o o o o
o o o o
UJ
a
0.
IU
I
O
_1
u
I 2
M
a
0) 13 X —
n
• ••-*
OTJ
X ~
i
ex
UI
0) T3 X
• o ro
o s: E
155
-------
o o o
00 =
3 •= 3 o
3 3 -3 O
i-l rt O O
o o o o-
o o o o
o o o o
Ul
o
Ul
i.
o
a.
I
o c o o
o o a o
.0 O O O
.3.000
o o o o
0000
o o o o
CO
s±
i
eg
wz
o o o o
o o o o
o
Ul
o o o o
0000
o o o o
o o o o
Ul
0000
o o o o
o
Ul
o o o o
o o o o
o o o o
o o o o
o o o o
SO MO O
sf «* o o
0000
o o o o
o>
•i-
.
CO
o o o o
o o o o
M M O O
o o o >d-
o o o o
o o o o
o o o o
o o o o
Ul
c
• M
S
cl
0000
o o o o
Ifl sS- O O
o o o o
o o o tsi
Cvl
CM t
r«-
O O
o o
l^ M O O
o
-------
in
i.
I
tri co o o
r- CM o in
o o o in
CM ON O O
co co o o
CM
ui
N?
o o o o
o o o o
o o o o
o o o o
o o o o
CM
Ul
CO
o o o o
o o o o
ON CO O O
ON NO O O
o ON
ut
CNl
Ul
CO
o o o o
0000
o o o o
o o o o
CM
ut
lu O ^_
I- U 3 C
X M O O
i.
o
•§
•H
o
o o o o
o o o o
o o o o
0000
o o o o
o o o o
o o o o
o o o o
o o o o
O-.
•if
o o o o
o o o o
o o o o
o o o o
o o o o
o o o o
o o o o
o o o o
o o o o
o
U)
• Ul
•* w
I U-
CQ Ul
lit M
—i o:
CO a.
< OT
g
g,
• H
I
Lfl
CM CM O O
O iH O CO
C2 CO
o o o o
o o o o
o o o o
o o o o
o o o o
o o o o
o o o o
o o o o
o o o o
i-l
Ifl
o o o o
o o o o
o o o o
o o o o
o o o o
o o o o
4!
>1
i-H
IS
5 c
S §
Ul
a
!£ £
0) 7}
_i e . .H .H
o fli o) ~a x <-•
u. o) • o) ra c
O S O EE -
w
i c -.2.
ui ra 0) ti
157
-------
a
•H
o o o o
o o o o
o o o o
CM
in
O O
O O
. in
CM
vO O O
- t/i o CM
CM
in
o o o o
o o o o
CM
LCI
£
Q.
0000
o o o o
r-t i-l O O
O CO O O
CM O C3 O
in n\ CM
CM CM CM
CM
O CO O O
O^ ON O in
O O O O
o o o o
CO
o o o o
o o o o
CM SO O O
o LO a o
o
Ifl
Kl CO O O
CM tH O O
o o o o
o o o o
o
m
I
ui g
o o o o
o o o o
o o o o
o o o o
tn H o in
«t M
ift
o o o o
in
i in
! W
in rH O O
O O O UI
in
O O O CO
o o o o
;§:
(5 E
X ~
x m
ra a)'
r^
I
s: o :
IX — -
!£ ^
ro at
0) • o ra c
£ M S S —
158
-------
S1
• H
I
o CM o a
in so o o
CO
si-
o o o
O O O
i^ m un
r-1 CM
sO
CO CO O O
tfl in O O
HI O O M
<(• r-( M CO
CM
Ul
s
CO !-« O O
CM CM O O
1-5 si- O M
i-i CO
i-i
- o —••
§* o
13-S9
3 EC!
CO Ul 3
Ul O C
Q Q. i»^ trf
Ul O
I- O
: M
i -P
i c
_ . . J O
O a: x o
u u oi~
ui s. c
so —
E1
o o a o
o o o o
o o o o
o o o o
o o o o
o
ifl
0000
o o o o
1T| CO O O
to en o o
I-l O O CM
U)
o
Ul
I1- -d- o o
o o o a
co 1-1 o O'
N vO O O
o
Ul
o o o o
0000
o o o o
o o o o
o o o o
• UI
* M
I U.
fft tJD
UI HI
_i o:
CD a
a
o
D)
c
0000
M> M O Ul
ui in o o
co vo o >d-
O O O CO
r-I
U)
CM
CM
O O
o o
i-l CO
CM
M CO O O
Ul 1^ O O
* r^ ui o
•O: CM so
Ul
0*00
<3 O O CD
o o o o
!>• CO O O
SO CM O U)
CM t-l O
m
rH
m
O O
o o
CO Os O O
i-l i-l O O
o o o o
Ul
3;
QL
I
si-
st
Ul
S
C 3
a s
fl)-p x —
ID E
C • -in -H
«J 0) T3 X
O • 0) IU
s: vizis
<
8
UI
ffi
159
-------
£.
a
O O O O
O O O O
Cvl
Ul
ON o o
so O O
r-l O O Ul
CM
Ul
0000
Q o o o
CM
Ul
O O O O
O O O O
O O O O
Ul
«
t.
.5
a
OT
O O O O
O O O O
O O O O
CO
sr
r*. co o o
o •* o o
M CM O O
r-l r-i O CM
CO CO O O
>± CM O O
O O O O
Cvl
0)
0000
o o o o
o
in
M CM O O
CO CO O O
o o o o
o o o o
o o o o
o
Ul
o o a o
o o o o
o o o o
o
Ul
-§
c.
C/)
o o o o
o o o o
C\I
. .
eo r- o
O O
O O
o o o o
o o o o
C\l CM O O
CO N1 O O
a
o o o o
o o o o
CO CM O O
>* <* o o
0000
o o o o
0000
o o o o
U)
ft
Ul
i.
o
|
§000
000
o o a o
ro i
•*
o o
C3 O
'ON ON O O
CM CM O O
NO O O O
ON O O O
CO N± O
•* tn
CO
B
2
(0 0) T3 X "•>
o • a ro c
(5 OT3
O • O
s: o s:
o;
<
c • '^
ro CDT
«J OT3
X —
160
-------
o o o o
0000
GJ
c
o
2
2
E>
o o o o
o o o o
o o o o
i.
o
-8
•H
o
E1
.
a
in
o o o o
0000
o o o o
o o o o
0000
o
U)
o o o o
o o o o
O)
c
o o o o
o o o o
V)
til
a
161
-------
-------
Appendix C
Weighted Percentiles for All Analytes
163
-------
s
.s
_ 01
CD C
1 I
1
CO
Q>
CD
o-
T3
&
!
o
Z
75
s
o
13
o o T- co
o o »- co cxj c\j oooooo oocomi-i- oooooo
t- t- c\j CD t- w in
OOOOOCM OOOOT-CO OOOOOO OOOOOO OOOOOO
ooojojor^ oooooo ooor-inco oooooo
'
OOOCVJCOCD OOOOOO OOOOOO OOOO)r-m OOOOOCO
t- T- T— CM T- C\J LO
10
of
oooooo oooooo oooooo oooooo oooooo
OOOCQCOO) OOOOOO OOOOCOO) OOOOOO
t~»-cn csj^-co co
O O r- CO ^f CO OOCNICDT-IO OOOOOCD OOOOOCO OOOOOO
CM i— ^± O t- OJCD^1
r^. T- co o)
ooooo^- OOOOCM^J- oooooo oooooo oooooo
OOOCXIGOCM O O O CO CO tO OOOOOO O O O O IO O OOOOOO
CO C\Jh-CO COCOO ^~
tooiooioo toomoiocn tooiooiog) in o LO o LD
CMior^cncno) cNjinr^oicncn c\imN-cno>O) ojio^-
in o in o LO 03
o
164
-------
15
c
o
£2
£
$> 8
I I
l
CL
CO 0
III
co co b
LLI
CL
O
o
o
•a
c
1
oooooo co o Tt co in o
eg r^ ^- ^r in o
T- CO O) CM
eg"
o^ooooo oococotom
eg co CD O) r^
oooooo mi-o^-coo
co i^ co co in in
r- in O) O
eg"
oooooo> oincoiooo
eg co o T- tj- o
CM in T- in
oooooo ooo o in ^
in in o>
OOOOOO Oi-I^cOOO
\f> CM CM O i-
CM CO !--_ O)
oooooo oinginoo
Tfr^ V— CO O^ CO
o o o o o T- o o o c\j ^- oo
*- r^ ^- cvj
OJ CO
OOOCOCOO O^-h-OOO
t-tO5 COOCvJWCVJ
»- CO O h- O
CO" T-" r-" CO"
to o m o LO o) moiooiDcn
c\jLnh*o5O3o> cvjinr^-OTOcn
000^1-00 cocDcotor^o
t — co co to ^f co r**- c\j
i- eg CM in
. r-
OOOOCOi- OOCOCOCOCD
r- - T- eg
o o o r-~ T— cn ^•cncnegin'—
COCO CMCDCDCOCncO
•- T- eg co cn
oooooo cninincocno
eg co eg eg CM co
CM co Tt m
CM"
ooooegin oOTtror^-*
oootocnco coo)^-como
*~ co eg r^ ^
-------
i
I
OT
O
«
X
CO *± CO ^
CO O O) *~ i- Tfr .C
i- CO *?
oooooco oooooo ooooj^tp
sa • -
oooocoo) oooooo o -
-------
8
. o
co
' CO
to P
«•§.
03
S
Q.
I
CO
UJ
Q.
O
Q.
co
2
O
o
S2
8
I
03
D-
oooooi-~ ocMini^Ttco
co CM r^
eg
OOOOOO OOOCM'
oooooo oeg^tT-Ttin
eg »- CM t^
^ o o co in in oooc»oco oooocoo:
*~ i- co co ^~ —
oocococo-^- oooooo oooooo
'CQCOCOi~G3 OOOOh^O
co m in *t co
i- in co
o o o o o
OO^fCMrtr^ OOCOCOOO OOOOO»-
T—T—o i— oo r^ 03 co T-
T- co o in
OOOOCMCO
OOOOOO OOOCMCOI-- OOOCM^CO OOOOOO OOOOOO
"^ CO O3 CO
OOOOOCM OOOCOOC3 O^ggOO 00000,;
*" CO CM CO
OOOOCOCO
^-TtcM
T-rtr- OOOT-r-CO
OOOOOO O O LO G) CO CM O O O CO IO O
oj co (Q i— *— en co T-
T~ ^ T— C\| OO CO
oooo^tco ooomoco ooootof^
T- »- w in o) CM
t- CO
OOOOOCO OOOJOOCO1O OO^l-OTOO
CM *— co r^> ^ ^-cocnr^co
^ CNJ CO T- 10
SgKgolS? SSKiSg aSK'858» cQSSggS? SgEggg!
03
O
03
O)
0>
•a
•
03
o
Q.
03
167
-------
CO
.s
1
c
o,
CD
~
If
CD
I
p> fe
1 I
CO 03
O
o
o
15
CL
I
CD
CL
OOOCOIOCVI OOOOOCD O O (O CO O> <- OOOOOO
cjo CM n- en CD oj
CO »- 10
CD
• i*-* ^~ to T~ co ^
• co 0) r^ in ^ .£
to [^ o) -e
oooooo oooooi^ ooooo-* oooooo ooomcnio
OOOCOT-CM oooooo r^-oi--r-oo oooooo
- -
o o o CD CM CD oooooo OCVJLOCOOO oooooo 't co-co oj 03 o
cut-in cococo-*co T- •* T- co co co
-— * --
OOOOOO OOOOOO OOOOOO OOOOOO OOOOOTIO
oor-ocno oooooo COCONCOCOO oooooo c\jjni-c\ioo
coo^ i-Ttcocg^r- ojcor^ocNcq
"CN T-OJCO i-inco_o_
ooo-co*
oooooo
en
co
o o o o o to oooooo o o o o i- r-« oooooo o o CD in CM to
oooooco
o in o in 01 inomomo) moioomoi moiooincn inomoujcn
iiot^cncno) coiot^cnmcn cumr-cncno) o) oomt^cncno)
I
fo
(0
J_
^—
s
^
"S
5
CD
£2
a.
&*
CD
.C
O-
£
0
§
CO
"2
o
^*
O
3
a
s
Q_
168
-------
Continued)
Sp
l£
co
05
Q.
3
sz
CO 03
LL1
Q.
O
Outdo
CD
o
1
CD
Q.
-CD
I
s.
oooooo oooooo ooooo
w
oooooo oooooooooooo ooooo^ oooooo
OOOOOO OOOOOO OOOOOO '0000(00 OOOOOO
CO C\J CO
oooooo oooooo oooooo
oooooo oooooo
oooooo oooooo oooooo oooooo oooooo
oooooo oooooo oooo
OO) OOOOOO OOOOOO
^
OOOOO
ooooocvj ooooor- o o o o to
OOOOOO OOOOOO OOOOOO OOOOO'
OOOOOO O O O O O CO O O O O O t- O O O O CO CO
CO ' T- T*
tp
a.
§
DC
2
Q
Q
•*
CJ
Q
Q
Q
V
•*"
169
-------
1
5
CD
=3
i s
a.
to
§
-------
NOPES Weighted Percentile - Springfield/Chicopee
(ng/m3)
Spring
Winter
Analyte
alpha-BHC
Aldrin
-
Atrazine
Bendiocarb
Captan
Carbaryl •
Chlordane
Percentile
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
Indoor
0
0
0
0
0
8
0
0
0
o -
0
0
0
0
0
0
0
0
0
0
0
0
0
10
0
0
0
0
0
0
0
0
0
0
0
16
0
29
70
544
1,700
1,700
Outdoor
0
0
0
0
" 0
0
0
0
0
0
0
0
0
0
0
0
0
a
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
35
43
. Personal
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
13
0
0
0
0
0
6
0
0
0
0
0
6
0
0
99
2,220
2,220
2,220
Indoor
0
0
0
0
0
0
0
0
0
2
2
2
0
0
0
0
0
0
0
0
0
0
0
38
0
0
0
0
0
0
0
0
0
0
0
0
14
25
45
62
76
206
Outdoor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
• 0
0
0
0
0
0
0
0
0
0
0
0
0
8
10
40
Personal
o
o
o
o
o
0
o
o
o
2
2
2
0
0
o
0
0
0
o
0
o
o
o
13
0
0
o
o
o
0
o
o
o
o
o
0
16
25
37
85
103
298
(continued)
171
-------
NOPES Weighted Percentile - Springfield/Chicopee (Continued)
(ng/m3)
Spring
Winter
Analyte
Chtorothatonil
Chlorpyrifos
cis-Permethrin
Dacthal
Otazinon
Dichtorvos
DiCOfo)
Percentile
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
Indoor
0
0
0
0
0
0
0
0
13
19
38
179
0
0
0
0
0
0
0
0
0
5
6
'32
0
0
0
20
45
1,810
0
0
0
0
0
241
0
0
0
0
0
0
Outdoor
0
0
0
3
3
3
0
3
9
12
34
523
0
0
0
0
0
0
0
0
0
2
4
26
0
0
0
7
14
391
0
0
0
0
0
0
0
0
0
0
0
0
Personal
0
0
0
7
7
7
0
0
15
17
25
92
0
0
0
0
0
0
0
0
2
13
13
13
0
0
0
19
26
318
0
0
0
0
0
241
0
0
0
59
59
59
Indoor
0
0
0
0
0
3
0
0
7
11
38
45
0
0
0
0
0
0
0
0
0
0
0
8
0
0
0
7
27
27
0
0
0
• 0
0
115
0
0
0
0
0
0
Outdoor
0
0
0
0
0
40
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
116
116
0
0
0
0
0
0
0
0
0
0
0
0
Personal
0
0
0
0
0
3
0
0
7
9
9
186
0
0
0
0
0
0
0
0
0
0
0
5
0
0
0
0
8
20
0
0
0
0
0
96
0
0
0
0
0
0
(continued)
172
-------
NOPES Weighted Percentile - Springfield/Chicopee (Continued)
(ng/m3)
Spring
Winter
Analyte
Dieldrin
Folpet
gamma-BHC
Heptachlor
Heptachlor epoxide
Hexachlorobenzene
Malathion
Percentile
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
Indoor
0
0
0
9
9
9
0
0
0
0
0
36
0
0
0
0
5
5
0
3
6
231
253
253
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
275
Outdoor
0
0
0
0
0
0
0
0
0
0
0
21
0
0
0
0
0
0
0
0
0
0
2
5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
11
24
Personal
0
0
0
7
7
7
0
0
0
0
0
32
0
0
0
7
7
7
0
0
6
313
313
313
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
20
Indoor
0
0
4
6
40
40
0
0
0
0
0
0
0
0
0
2
118
118
0
1
3
8
17
35
0
0
0
0
0
0
0
0
0
0
0
6
0
0
0
0
0
0
Outdoor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 .
0
0
0
0
0
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Personal
0
0
0
4 :
6
6
0
0
0
0
0
0
0
0
o
0
70
70
0
4
7
9
12
23
0
0
0
0
0
0
0
0
0
0
0
3
0
0
0
0
0
0
(continued)
173
-------
NOPES Weighted Percentile - Springfield/Chicopee (Continued)
(ng/m3)
Spring
Winter
Analyto
Mothoxychtor
ortho-Phenylphenol
Oxychtordano
Propoxur
Rosmolhrin
Ronnel
trans-Pormethrin
Percentile
25
50
75
90
95
gg
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
gg
25
50
75
90
95
99
25
50
75
go
95
gg
Indoor
0
0
0
0
0
0
15
25
79
126
126
181
0
0
0
0
0
0
0
0
33
111
111
121
0
0
0
0
0
0
0
0
0
0
0
9
0
0
0
0
0
0
Outdoor
0
0
0
0
0
0
0
0
0
0
10
52
0
0
0
0
0
0
0
0
0
0
0
33
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Personal
0
0
0
0
0
0
17
30
81
128
128
177
0
0
0
0
0
0
0
0
13
79
79
79
0
0
0
0
0
0
0
0
0
0
0
7
0
0
0
0
0
0
Indoor
0
0
0
0
0
0
0
11
15
87
96
96
0
0
0
0
0
0
0
0
20
53
53
129
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Outdoor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Personal
0
0
0
0
0
0
9
15
24
50
133
133
0
0
0
0
0
0
0
0
9
39
47
61
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
(continued)
174
-------
Analyte
NOPES Weighted Percentile - Springfield/Chicopee (Continued)
(ng/m3)
Spring
Percentile
Indoor
Outdoor
Personal
2,4-D (butoxyethyl
ester)
4,4'-DDD
4,4'-DDE
4,4'-DDT
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
25
50
75
90
95
99
0
0
0
0
0
104
0
0
0
0
0
0
0
0
0
6
6
8
0
0
0
0
0
0
0
0
0
0
0
0 .
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
o
0
0
0
0
0
0
0
0
0
0
0
0
38
38
38
0
0
0
8
8
8
Indoor
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
4
4
4
0
0
0
0
5
15
Winter
Outdoor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
19
Personal
0
0
0
0
0
0
0
0
0
0
0
. 0
0
0
0
2
4
4
0
0
0
2
6
6
175
-------
-------
AppendixD
Glossary of Statistical and NOPES Terms
area householding sampling - a standard survey
sampling method in which sample households or
people are chosen from sample areas selected at
a previous stage of sampling. The sample areas
are selected from a sampling frame that provides
complete geographic coverage of the area in
which the target population resides.
bias - the difference between the expected value of a
sample statistic and the corresponding population
parameter. The expected value of a statistic is the
average value of the statistic over all possible
samples.
census - a survey of all units in the target population.
detection limit - the minimum analyte concentration
that consistently produces responses above the
instrument background signal under typical
operating conditions. Defined in NOPES as three
to five times the instrument background signal.
duplicate air sample - an air sample collected for
essentially the same time and space as the
prirnary air sample. Duplicate samples were
collected both indoors and outdoors in a
subsample of households.
measurement error - error that occurs because the
measurement process, including environmental
sampling, laboratory analysis, sample
identification, questionnaire administration, and
data entry, yields an incorrect result for the
characteristic being measured.
multiseason respondents - sample members that
participated in more than one of the NOPES
phases. Prior to initial contact in a study area, a
subset of the sample was randomly selected to be
recruited as multiseason participants.
population parameter - a characteristic based on or
calculated from all units in the target population.
probability sample - a sample for which every unit
on the sampling frame has a known, positive
probability of being selected into the sample. The
terms "probability sampling" and "random
sampling" are some times used interchangeably.
quantitation limit - the minimum analyte
concentration that yields relatively precise
response values under typical operating
conditions. Defined in NOPES as approximately
five times the detection limit.
replicate air sample - an air sample collected from a
household or for an individual that had provided a
primary air sample three to ten days earlier.
Replicate samples were collected for a subset of
sample members. Indoor, outdoor, and personal
replicate air samples were collected.
sample (statistical) - a set of units selected from the
target population. , --
sampling design - the method used to select a
sample of units from the target population.
sampling error - error that occurs because
inferences are made from a sample rather than
from a census of the entire population.
sampling frame - a list from which a sample is
selected. An ideal sampling frame contains one
and only one entry for each member of the target
population. In practice, sampling frames usually
miss some members of the target population, and
. include some individuals who are not members of
the target population.
sampling variance (of a statistic) - the variance of
the sampling distribution of the statistic, which is
generated by the sampling design.
sampling weights - factors used to compute design-
unbiased population estimates from sample data.
For probability sampling designs, a unit's sampling
weight is the reciprocal of its probability of
selection. Adjustments of sampling weights are
often made to partially compensate for the
potential bias due tor- nonresponse. If the sampling
design results in unequal probabilities of selection
for sample members, sampling weights must be
used to compute unbiased population estimates.
177
-------
standard error (of a statistic) - the square root of
the sampling variance of the statistic.
statistic - a sample-based estimate of a population
parameter.
stratified sample - a sample selected from a
sampling frame which is partitioned into disjoint
subsets called strata, and composed of
subsamples selected independently from each
stratum.
target population - the set of units or elements for
which a sample survey is designed to provide
statisitical inferences. The target population is
sometimes simply referred to as the population or
universe of inferential interest.
triplicate air sample - an air sample collected for
essentially the same time and space as the
primary and duplicate air samples. Triplicate
samples were collected from a small subset of
sample households, and were collected both
indoors and outdoors.
OU.S. GOVERNMENT PRINTING OFFICE: 1990-718-159/20 W
178
-------
-------
TJ O
CD X
01 O
=ug
S S
CO
O
O
I
o
CO
1
o
o
a*
co
If you do n
detach, or
upper left
3- o o
0) O "*
g.3|
o 31 ^
X 5-5
3 0 -t
2 o to
r1 < o
=f. o
ho
o>
00
T3
m
g
-a
O
CO
1
------- |