PRELIMINARY 12-MONTH REPORT
of the
LEAD PAINT ABATEMENT AND
REPAIR AND MAINTENANCE STUDY IN BALTIMORE
Conducted by
KENNEDY KRIEGER RESEARCH INSTITUTE
for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Ben Lim, EPA Task Manager
Technical Programs Branch
Chemical Management Division
Office of Pollution Prevention and Toxics
July 21, 1995
-------
PRELIMINARY 12-MONTH REPORT
of the
LEAD PAINT ABATEMENT AND
REPAIR AND MAINTENANCE STUDY IN BALTIMORE
Conducted by
KENNEDY KRIEGER RESEARCH INSTITUTE
for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Ben Lim, EPA Task Manager
Technical Programs Branch
Chemical Management Division
Office of Pollution Prevention and Toxics
PREPARED BY:
Mark R. Farfel, Sc.D., Project Director
Charles Rohde, Ph.D., Project Statistician
Peter S. J. Lees, Ph.D., QC Officer
Brian Rooney, Data Analyst
Desmond I. Bannon, Laboratory Chief
Luban Alvi, Data Manager
July 21, 1995
-------
EXECUTIVE SUMMARY
This preliminary report is based on data from 60 (80%)
intervention houses and 29 controls that have completed one year of
follow-up in the Lead-Based Paint Abatement and Repair &
Maintenance (R&M) Study. The research aim is to characterize and
compare the short (2-6 mo.) and longer-term (12-24 mo.)
effectiveness of three levels of lower cost R&M interventions
designed to reduce children's exposure to lead in paint and dust,
known sources of high lead exposure in U.S. children. The study
population consisted of non-Hispanic black households (mostly
renters) residing in Baltimore rowhouses with reported low monthly
housing payments (mean $324). At the outset, mean ages of study
children ranged from 25 to 33 months across groups, and the GM PbB
levels in the three R&M groups and previously abated (PA) controls
ranged from 10 to 14.5 ng/dL. The GM PbB in children in the modern
urban (MU, post-1979) control houses was 4
The report presents longitudinal data from five study groups
across five campaigns. A number of types of graphical displays are
included to facilitate understanding of the changes found in dust
lead and children's blood lead. The reader who has a preferred type
of data display or limited time is directed to specific tables and
figures as follows: data tables- see Tables 5-8; box plots with
median traces- see Figures 1-13, 17-19; bargraphs - see Figures 14-
16, 20-23; for plots- see Figures 24-28 (blood data only). The
main findings are listed below:
All three levels of R&M intervention were associated with
sustained reductions in dust lead (PbD) loadings, lead
concentrations (PbD-C) , and dust loadings. The degree of
reduction in PbD and PbD-C was positively associated with the
level of the R&M intervention. Statistically significant
differences were found between R&M groups after intervention
and at two, six, and twelve months.
Over time, the MU control group had significantly lower dust
lead levels than the other four study groups. Dust lead levels
in the PA control group were generally intermediate between
the middle (R&M II) and high (R&M III) level of intervention.
Children with pre-intervention PbBs^20 jug/dL had statistically
significant reductions in PbB to levels <20 jug/dL at twelve
months post-intervention. Children with pre-intervention PbBs
<20 /ig/dL did not have statistically significant PbB changes
within or between R&M groups. An increasingly downward, but
not statistically significant, trend in PbB was found in
children in the high level intervention group (R&M III) .
Children in the MU control group had significantly lower PbBs
over time as compared with each of the other four groups.
Using all five study groups in longitudinal data analysis,
children's PbB was significantly related to dust lead.
-------
TABLE OF CONTENTS
1.0 INTRODUCTION 1
2.0 STUDY DESIGN AND SAMPLE COLLECTION PROCEDURES 3
2.1 Selection Criteria for Houses and Children 5
2.2 Repair & Maintenance Interventions 6
2.3 Recruitment and Enrollment 7
2.4 Sample Collection Procedures 8
3.0 LABORATORY ANALYSIS PROCEDURES 10
4.0 DATA PROCESSING AND STATISTICAL ANALYSIS PROCEDURES . . 11
4.1 Data Processing 11
4.2 Data Summary 12
4.3 Quality Control 13
4.4 Statistical Analysis 14
5.1 Side-by-Side Boxplots with Median Traces 23
5.2 Descriptive Statistics - Twelve Month Campaign . . 25
5.3 Longitudinal Data Analysis 29
6.0 DISCUSSION 34
7.0 REFERENCES 47
TABLES 1-10
FIGURES 1-28
APPENDIX A: Descriptive Statistics
APPENDIX B: Longitudinal Data Analysis: Dust Models
APPENDIX C: Longitudinal Data Analysis: Blood Lead Models
-------
List of Tables
Table 1:
Table 2:
Table 3:
Table 4:
Data collection plan
Types and numbers of 12-month campaign samples
collected and analyzed for lead
Types and numbers of 12-month campaign samples
collected by study group
Frequency of family moves and reoccupancies between the
initial and the twelve month campaigns among the 60 R&M
houses included in this preliminary report and the
control houses.
Overall dust lead loadings for houses completing the
12 -month campaign
Overall dust lead concentrations for houses completing
the 12 -month campaign
Overall dust loadings for houses completing the 12-
month campaign
Descriptive Statistics for Blood Lead Concentrations
for Children with Initial PbB <20 //g/dL
Twelve month campaign correlations of In (lead loading),
In (lead concentration) and In (dust loading) with
In (blood Atg/dL) for the youngest child per household
Table 5:
Table 6:
Table 7:
Table 8:
Table 9:
Table 10: Twelve month campaign correlations of In (lead loading),
In (lead concentration) and In (dust loading) with
In (blood /ug/dL) for all children
ii
-------
List of Figures
Figure l:
Figure 2:
Figure 3:
Figure 4:
Figure 5:
Figure 6:
Figure 7:
Figure 8:
Figure 9:
Figure 10:
Figure 11:
Figure 12:
Figure 13:
Figure 14:
Figure 15:
Figure 16:
Figure 17:
Figure 18:
Boxplot: Dust Lead Loadings across Campaigns for
Floor Surfaces
Boxplot: Dust Lead Loadings across Campaigns for
Window Sill Surfaces
Boxplot: Dust Lead Loadings across Campaigns for
Window Well Surfaces
Boxplot: Dust Lead Loadings across Campaigns for
Interior Entryway Surfaces
Boxplot: Dust Lead Concentrations across Campaigns
for Floor Surfaces
Boxplot: Dust Lead Concentrations across Campaigns
for Window Sill Surfaces
Boxplot: Dust Lead Concentrations across Campaigns
for Window Well Surfaces
Boxplot: Dust Lead Concentrations across Campaigns
for Interior Entryway Surfaces
Boxplot: Dust Loadings across Campaigns for Floor
Surfaces
Boxplot: Dust Loadings across Campaigns for Window
Sill Surfaces
Boxplot: Dust Loadings across Campaigns for Window
Well Surfaces
Boxplot: Dust Loadings across Campaigns for
Interior Entryway Surfaces
Boxplot: Blood Lead Concentrations for Children
with Initial Blood Pb <20 ug/dL
Bar Graph: Dust Lead Loadings over Time
Bar Graph: Dust Lead Concentrations over Time
Bar Graph: Dust Loadings over Time
Boxplot: Dust Lead Loadings across Groups at the
12 Month Campaign
Boxplot: Dust Lead Concentrations across Groups at
the 12 Month Campaign
111
-------
List of Figures (cont.)
Figure 19:
Figure 20:
Figure 21:
Figure 22:
Figure 23:
Figure 24:
Figure 25:
Figure 26:
Figure 27:
Figure 28:
Boxplot: Dust Loadings across Groups at the 12
Month Campaign
Bar Graph: Overall Lead Levels and Dust Loadings
by Group
Bar Graph: Overall Dust Lead Loadings by Group
Bar Graph: Overall Dust Lead Concentrations by
Group
Bar Graph: Overall Dust Loadings by Group
Plots of Children's Blood Lead Levels across Tine
- R&M I
Plots of Children's Blood Lead Levels across Tine
- R&M II
Plots of Children's Blood Lead Levels across Tine
- R&M III
Plots of Children's Blood Lead Levels across Tine
- Modern Urban
Plots of Children's Blood Lead Levels across Tine
- Previously Abated
IV
-------
Preliminary RtM Study Report-
Partial 12 Month Data 7/95
1.0 INTRODUCTION
This preliminary report is based on data from eighty percent
of the intervention houses that have completed the twelve month
campaign of the Lead-Based Paint Abatement and Repair & Maintenance
(R&M) Study which is underway at the Kennedy Krieger Research
Institute (KKRI) in Baltimore. The R&M study is being conducted
according to the Quality Assurance Project Plan approved by EPA in
November 1992.1
Included in this report are twelve month data from twenty
houses in each of the three R&M intervention groups (Levels I-III,
n=60 houses), 15 (100%) modern urban (MU) control houses, and 14
(93%) previously and comprehensively abated (PA) control houses.
The twelve month data from these 89 houses were collected between
March 1994 and May 1995. The level of participant cooperation with
home and clinic visits has been high. Consequently, we have
sufficient preliminary data to begin to address the following study
objectives:
Describe changes in lead levels in settled house dust for R&M
Levels I-III across five campaigns, i.e., pre- and immediate
post-intervention, and two months, six months, and twelve
months post-intervention.
Describe changes in environmental variables between baseline
and the twelve month campaign for the modern urban and
previously abated control houses.
Fit the study's statistical model for longitudinal data
analysis to the dust lead and blood lead data.
An earlier report provided descriptive statistics on baseline
environmental, biological, and demographic data for the five study
groups (total of 107 houses).: At baseline, the study population
consisted of non-Hispanic black households with reported low-to-
moderate monthly rents and mortgages (mean $324) residing in
Baltimore City row houses. Most (81%) households were renters.
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
The median household size was 4 persons, and the number of study
children per household ranged from 1 to 4 children. Mean baseline
ages of children ranged from 25 to 33 months across study groups.
Accounting for multiple children per household (clustering),
geometric mean baseline blood lead concentrations (PbB) were 5, 13,
10, 14.5, and 13 ng/dL for the modern urban, previously abated, and
R&M I-III groups, respectively.
At baseline, the three R&M groups were similar with respect to
demographic characteristics and lead levels in blood and
environmental samples. However, lead levels in settled house dust
and children's blood at baseline tended to be highest for the R&M
Level III houses (vacant at baseline), lowest for the R&M Level I
houses (occupied at baseline), and intermediate for the R&M Level
II houses (mix of vacant and occupied houses at baseline).2 Some
of the owners of the vacant R&M III and II houses accepted study
families with children with blood lead elevations >20 /xg/dL as per
their policy. Consequently, unlike R&M I households, R&M Levels II
and III had some children with initial baseline blood lead
concentrations >20 /xg/dL. These children were analyzed separately
in this report.
Purpose of R&M Study
The R&M study is designed to characterize and compare the
short (2 to 6 month) and longer-term (12 to 24 month) efficacy of
comprehensive lead-paint abatement and less costly and potentially
more cost-effective Repair and Maintenance (R&M) interventions for
reducing lead in settled house dust and children's blood. The
investigators plan to extend the follow-up to five years post-
intervention. This research is important because residential
paints and dusts have been identified as major sources of lead
exposure for U.S. children, particularly those living in houses
with deteriorated paint and high dust lead levels.3 Furthermore,
there is a dearth of research literature on the outcomes of
interventions for children with blood lead concentrations in our
-------
Preliminary R*M Study Report-
Partial 12 Month Data 7/95
primary study range of 10 to 20 jtg/dL.
The R&M approach may provide a practical means of reducing
exposure for future generations of children who will continue to
occupy housing containing lead-based paint. However, systematic
studies of the R&M approach are needed in order to provide a sound
scientific basis for prevention policies. EPA funding is
anticipated for two years of follow-up testing in each study house
to periodically measure lead in settled dust, soil, water, and
children's blood. Pre- and immediately post-R&M intervention data
provide baseline measurements for the study of short and longer-
term changes in lead levels in children's blood and settled house
dust.
The overall research goal of the R&M study is to contribute to
the scientific basis for a standard of care for lead-painted houses
via the conduct of a longitudinal intervention study. Specific
study objectives are as follows:
1. Measure and compare the short and longer-term changes of lead
in settled house dust and children's blood associated with R&M
Levels l-lll and previous comprehensive abatement.
2. Characterize the nature of the relationship between lead in
children's blood and settled house dust.
3. Evaluate dust collection methodologies for the determination
of lead in residential dusts, including wipe and cyclone
methods. This objective was addressed in past reports.4"6
2.0 STUDY DESIGN AMD SAMPLE COLLECTION PROCEDURES
This prospective study has two main components and five groups
of study houses (Table 1) . The first component is designed to
obtain serial measurements of lead in venous blood of children 6
months through 4 years of age at baseline, settled house dust,
soil, and drinking water in three groups of 25 R&M houses (total of
75 houses), each receiving one of three levels of R&M intervention
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
(Levels I-III) (Table 1). The study questionnaire, designed to
obtain information on demographics and covariates which could
influence lead exposure in the home (e.g., hobbies and child
behavior) are done at six month intervals beginning at enrollment.
Occupied study houses (target n=37) were randomly assigned to
receive either R&M Level I or R&M Level II interventions. Houses
vacant at the time of intervention (target n=38) were randomly
assigned to receive R&M Level II or Level III interventions. Since
R&M Level II interventions were done in both occupied and vacant
houses, the randomization scheme was designed to ensure equal
numbers of houses (n=25) at each R&M level. More frequent sampling
campaigns were planned for R&M houses during the first year to
allow for the estimation of the rate of reaccumulation of lead in
dust and an early assessment of the need for further cleanup or
repairs.
The second component of the study design is to obtain serial
measurements of lead in house dust, soil, drinking water, and
venous blood of children (aged 6 months through 4 years at
baseline), in two groups of control houses (Table 1). The study
questionnaire is administered at six month intervals in both
control groups.
One control group is a group of 16 scattered site row houses
in older housing neighborhoods which received a comprehensive type
of lead-paint abatement performed by pilot abatement projects in
Baltimore between May 1988 and February 1991. The two years of
planned follow-up will provide an opportunity to measure the
efficacy of comprehensive abatement practices at 4-to-6 years post-
abatement. Pre- and immediate post-abatement dust lead data are
available for these previously abated houses from past studies.
Modern urban houses built after 1979, and presumably free of
lead-based paint, constitute the second control group (n=16).
These houses are located in modern urban subdivisions in which all
houses are of this type.
-------
Preliminary RsM Study Report-
Partial 12 Month Data 7/95
2.1 Selection Criteria for Houses and Children
The following selection criteria were applied to all groups:
1. House size: approximately 800-1200 square feet.
2. House Condition: Structurally sound without pre-existing
conditions that could impede or adversely affect the R&M
treatments and the safety of the workers and field staff
(e.g. roof leaks or unsafe floor structures).
3. Utilities (heat, electric, and water) were available to
facilitate intervention and field sampling.
4. Occupants: Household included at least one child 6 months
through 4 years of age who was neither mentally retarded
nor physically handicapped with restricted movement;
family had no definite and immediate plans to move.
5. House was not excessively furnished. This criterion
enabled dust collection in all houses and intervention
and cleanup efforts in occupied R&M houses.
The following selection criteria were applied to R&M candidate
houses only:
6. House contained lead-based paint (defined in Maryland as
*0.7 mg Pb/cm2 or *0.5 percent lead by weight as
determined by wet chemical analysis) on at least one
surface in a minimum of two rooms or, in the absence of
testing, year of construction pre-1941.
7. House dust lead levels prior to intervention exceeded
Maryland's interim post-abatement clearance levels
(<200 jiig/ft2, <500 Atg/ft2 and <800 Mg/ft2, respectively,
for floors, window sills, and window wells) at a minimum
of three locations.
8. House had 12 or fewer windows needing R&M work. This was
to facilitate the R&M work with limited resources.
Additional criterion applied to previously abated houses:
9. At least two pairs each of pre- and immediately post-
abatement dust-wipe lead measurements from the same
floor, window sill, and window well surfaces were
available from previously collected data. These data
provide baseline dust lead levels.
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
2.2 Repair & Maintenance Interventions
The Repair & Maintenance interventions under study were all
financed by the Maryland Department of Housing and Community
Development (DHCD) via a special loan program available to property
owners. The initial R&M intervention costs were capped by DHCD as
follows: Level I: $1,650; Level II: $3,500; Level III: $6,000-
$7,000. All R&M work was performed by workers trained in lead
paint abatement work according to Maryland regulations. These
interventions, described in detail elsewhere,1 are briefly
described below.
R&M Level I interventions included wet scraping of
deteriorating lead-based paint on interior surfaces, limited
repainting of scraped surfaces, installation of an entryway mat,
wet cleaning and vacuuming with a high efficiency particulate air
(HEPA) vacuum to the extent possible in an occupied unit, education
of occupants and owners, and stabilization of exterior painted
surfaces to the extent possible given the budget cap. R&M I
interventions were done exclusively in occupied units and generally
took a full day to complete.
Two key elements added to R&M Level II interventions were
floor treatments to make floors smooth and easily cleanable and in-
place window and door treatments to reduce abrasion of lead painted
surfaces. Half of the R&M II interventions were performed in
vacant units. In the occupied units, floor treatments were limited
to rooms and areas where the work was feasible. R&M II work was
generally completed within two days. Precautions were taken in the
occupied R&M I and II houses to protect children and furnishings
from lead exposure during the work phase. These measures included
having children out of the house until all work and cleanup was
completed and the use of containment measures such as plastic
sheeting to protect belongings.
For R&M Level III, the highest level of intervention,
-------
Preliminary RiM Study Report-
Partial 12 Month Data 7/95
additional work included window replacement as the primary window
treatment, encapsulation of exterior window and door trim with
aluminum coil stock, and more durable floor and stairway treatments
(e.g. coverings). All R&M households received cleaning kits for
their own wet cleaning efforts which included a bucket, sponge mop,
sponges, a replacement sponge mop head, TSP cleaning agent, and the
EPA brochure entitled "Lead Poisoning and Your Children."
2.3 Recruitment and Enrollment
The enrollment process entailed a three step process of pre-
enrollment, formal enrollment, and ongoing pre-enrollment and
enrollment activity as described below.
Extensive home visiting activity (1100 home visits to over 650
modern urban, previously abated, and older occupied dwellings) was
performed by field staff as part of pre-enrollment field activities
during the spring and summer of 1992. Over 90 percent of
households identified as potentially eligible for the study
indicated an interest in participating. This early pre-enrollment
activity yielded 100 interested and eligible households for formal
enrollment.
Formal enrollment refers to the obtaining of signed informed
consent statements for study participation from parents or legal
guardians for both environmental and biological sampling. Separate
consent statements were obtained for each child enrolled in the
study using forms approved by the Joint Committee on Clinical
Investigation of the Johns Hopkins Medical Institutions.
Between the time of formal enrollment and the initiation of
the initial data collection campaign in January of 1993, some
enrolled households were lost to the study, primarily due to
children aging and the moving of families to other dwellings. In
some cases, the losses necessitated ongoing pre-enrollment activity
to identify an increased pool of potential study participants. No
evidence was found for selection bias when excluded R&M-candidate
houses were compared to the R&M study houses.2
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
2.4 Sample Collection Procedures
Venous blood was collected from children at the KKRI Lead
Clinic by a pediatric phlebotomist into 3 mL Vacutainersฎ with EDTA
added as an anticoagulant.
Trained field teams collected all environmental samples,
including field QC samples (blanks and duplicates). Settled house
dust was collected using a modified high volume cyclone sampler
originally developed for EPA for the evaluation of pesticide
residues in house dust.7 The modified device, referred to as the
R&M cyclone, is described in detail and characterized elsewhere.4
The device consists of a cast aluminum cyclone attached to a hand
held Dirt Devilฎ vacuum as the air mover for the system. A 100 mL
Teflonฎ microwave digestion liner was used as the sample collection
container to eliminate a sample transfer step in the laboratory,
thereby reducing the risk of sample loss.
The sampling plan for settled dust included the collection of
three composite floor samples per house - one across rooms with
windows on the first story, one across rooms with windows on the
second story and one from first and second story rooms without
windows. Two randomly selected 1-ft2 (929-cm2) perimeter floor
locations were sampled in each room designated for inclusion in a
composite sample. Composite window sill and window well samples
were collected separately from all first and second story windows
available for sampling at each campaign, respectively. Settled
dust was collected as individual samples from horizontal and
accessible portions of air ducts, interior and exterior entryways,
and the main item of upholstered furnishing.
Soil core samples were collected as separate composites of the
top 0.5 inch (1.3 cm) of soil from 3 randomly selected locations at
the drip line. Cores were collected into polystyrene liners using
a 6-inch (15.2 cm) stainless steel recovery probe.
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
Drinking water samples were collected as 2-hour fixed-time
stagnation samples from the kitchen faucet rather than first draw
samples to avoid disrupting families early in the morning. The
procedure was to run the cold water for at least two minutes to
flush the pipes and then to collect the first flush of water after
a 2-hour interval. A list of field sample types is provided below.
Settled-dust fR&M cyclone)
Perimeter floor composite - 1st and 2nd story
rooms with windows and rooms without windows
Window sill composite - 1st and 2nd story
Window well composite - 1st and 2nd story
Air duct/upholstery
Interior entryway
Exterior entryway
Soil core Drip-line composite
Drinking water Kitchen faucet
Field QC
Blanks (one per house) and duplicates
(every tenth house) for all sample types.
Information on the study children and their households was
collected using a structured interview questionnaire. As planned,
families were informed by letter of the results of all dust lead
and blood lead tests. Dust test results were provided on a
qualitative basis with recommendations for priorities for
housekeeping. Families with water and soil lead concentrations
that exceeded EPA guidance levels were provided with additional
recommendations for avoiding lead exposure (e.g. EPA's guidance to
run the water to flush the pipes prior to use whenever the water
has been in contact with the plumbing for two hours or more) .
Additionally, separate letters were sent to the parents/guardians
with the results of the blood lead tests so that they could be
shared with the child's primary care provider. All blood lead
results were reported to the Maryland Childhood Lead Registry as
required by Maryland law.
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
3.0 LABORATORY ANALYSIS PROCEDURES
Interior and exterior settled dust, exterior soil, water, and
venous blood samples were analyzed at the Kennedy Krieger
Institute's Trace Metal Laboratory using established analytical
methods. Microwave digestion was used for dust, soil, and water
samples (modified SW 846 Methods 3015 and 3051). Analysis of dust
and soil digestates was performed using Inductively Coupled Plasma-
Atomic Emission Spectrometry (ICP) (SW 846 Method 6010) and/or
Graphite Furnace Atomic Absorption Spectrometry (GFAA) (SW 846
Method 7421). Analysis of drinking water was by GFAA (SW 846
Method 7421). Blood was analyzed by GFAA using matrix matched
standards.3 The table below summarizes the procedures.
Summary of Laboratory Procedures
Sample Type
Dust
Soil
Drinking Water
Blood
Pre- Preparation
Summary
Post-field drying
and gravimetrics
Sample drying and
homogenization
none
Stabilized in EDTA
Preparation Summary
Digest using 1:1 HN03: H20
with microwave heating
Digest using 1:1 HNO3: H20
with microwave heating
Digest using 1:1 HNO3: H20
with microwave heating
Addition of matrix
modifier/triton X-100
solution
Analysis
Summary
ICP/GFAA'
GFAA
GFAA
GFAA
Samples with lead concentrations below the limit of quantitation
of the ICP instrument were analyzed by GFAA.
10
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
4.0 DATA PROCESSING AND STATISTICAL ANALYSIS PROCEDURES
4.1 Data Processing
Sources of data include the field collection forms, the
questionnaire, and the laboratory data packages. Raw data of all
types were transferred to the Data Manager who uploaded the data to
a VAXStation 3100 computer and prepared the data for later
analysis. Below is a summary of the data processing steps employed
for the three sources of data.
The field data set consists of all data recorded on the field
collection forms for settled dust, soil, and drinking water
samples as well as room and window inventory data. A data
entry firm keypunched and verified (double entry) data from
the field forms into ASCII data files. These raw data files
were transferred to the Data Management team for management,
storage, and later analysis. Field data forms were checked
for completeness and accuracy by both the Outreach Coordinator
and Data Manager prior to data entry. Data were re-verified
by laboratory staff after data entry.
Laboratory data were electronically stored for each laboratory
instrument as follows: Mettler Balance: gravimetric data
(tared and loaded weights for cyclone-dust and soil samples)
ICE: lead measurements for cyclone-dust samples. Graphite
Furnace-AAS; lead content of drinking water, soil, blood, and
low lead concentration cyclone dust samples. Electronically
stored laboratory data from the Mettler, ICP, and GFAA
instruments were moved to Paradox (v.4.0) by laboratory staff
for tracking of samples. The data were re-verified by
laboratory staff after data entry. Paradox data were then
converted to ASCII files by the Data Management team for
uploading to the VAXStation. A SAS program read in the
laboratory data for environmental and blood samples and
created SAS datasets for data analysis.
Questionnaire data forms were keypunched and verified (double
entry) by the data entry firm into ASCII data files. These
raw data files were verified in-house and transferred to the
Data Manager. A SAS program read in the raw data and created
SAS datasets for data analysis.
11
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
4.2 Data Summary
This preliminary report includes data through the twelve month
campaign for 60 R&M houses and 29 control houses as follows:
R&M Level I: 20 (80%) of 25 planned houses
R&M Level II: 20 (80%) of 25 planned houses
R&M Level III: 20 (80%) of 25 planned houses
Modern Urban: 15 (100%) of 15 planned houses
Previously Abated: 14 (93%) of 15 planned houses.
Environmental dust data from four surface types (perimeter
floor, window sill, window well, and interior entryway) included
in each of the first five data collection campaigns (pre-R&M, post-
R&M, two months, six months, and twelve months post-R&M) are
included in this report as well as data collected less frequently
(i.e., airduct dust, upholstery dust, soil, and water). Tables 2
and 3 display the types and numbers of 12-month campaign samples
planned, collected, and analyzed for lead by study group for the 89
houses included in this report.
Some of the original study families moved or voluntarily
withdrew from the study between the initial and twelve month data
collection campaigns. Table 4 reports the frequency of family and
child moves, and reoccupancies by new study families, by study
group. Approximately 21 percent (13/60) of the original families
in the 60 R&M houses moved prior to the twelve month campaign. By
the twelve month campaign, all of these study families were
replaced by the next family that moved into the house. Despite our
success in gaining the participation of these new families, they
had fewer eligible children than the original families. By the
twelve month campaign, the study also gained 15 children who were
newborns that became of age (ฃ6 months) for blood lead testing.
One R&M II house was vacant at the time of the twelve month
sampling. None of the houses included in this report are known to
have had any major renovations or repairs during the first year of
follow-up. One R&M I house had its front and back doors replaced
12
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
due to break-ins that damaged the original doors, and in another
house the wallpaper was removed by the occupants from the first
floor rooms by a steam process.
Children in most of the occupied R&M I and II houses
experienced an extended time interval of 6 to 10 months between the
initial PbB and the completion of the R&M work. This interval was
due to the time taken by collaborating property owners to provide
a completed R&M loan application and the time required for loan
processing, loan approval, and loan closing. The actual R&M work
contributed relatively little to the actual time intervals between
the initial and subsequent campaigns. An effort is currently
underway, with the permission of study families, to obtain any
additional PbB data that might be available from these children's
medical records for the period between the initial study PbB and
the start of the R&M work that could serve as new baseline study
values closer to the times of the interventions. This effort was
not needed for children who moved into the vacant R&M II and III
houses after the R&M work was completed.
4.3 Quality Control
At the time of this report, the performance audit which
consists of a review of the laboratory QC charts has been an
ongoing activity of the QC Officer. QC samples include instrument
QC samples (initial and continuing calibration verification
samples, initial and continuing calibration blanks, standard
reference materials, spikes and spike duplicates), method blanks,
and field QC samples (field blanks, field duplicates). The data
quality objectives related to laboratory and field performance
continue to be met. The QC Officer has also completed the system
audit of the laboratory and field procedures. No significant
deficiencies have been noted. Since the data audit is ongoing and
the twelve month campaign has not been completed, this report is
termed preliminary and partial in nature.
13
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
The ongoing data review has recently revealed occasional
inconsistencies over time in the measurement of various window
components in some of the study houses. The measurement of window
components can be a problem due to the multiple window types
encountered; selection of the appropriate dimensions to
characterize the component requires considerable judgement by the
field teams. Shortly after field work commenced, the protocol for
window measurement was greatly expanded and was training conducted
to reduce potential measurement problems. Some problems remain
however. At this time, the principal investigator and the QC
Officer are in the process of personally inspecting and remeasuring
all windows identified with discrepancies in measurements in any
campaign. As a result of these remeasurements, it is expected that
some lead and dust loadings and statistical analyses based upon
these measures in the six- and twelve-month preliminary reports
will be revised slightly.
4.4 Statistical Analysis
In order to compare the same houses over time, we restricted
the longitudinal data analysis to the 60 R&M houses and 29 control
houses for which data were available from the twelve month campaign
as of May 1995.
For data analysis purposes, lead values less than the
instrument detection limit (IDL) were coded as the IDL/v/2.9 For
lead values less than the limit of quantitation (LOQ) but greater
than the IDL, the observed value was used in the data analysis.
One child in a previously abated house had a blood lead
increase to a concentration of 53 ng/dL at the 12 month campaign
and was provided with chelation therapy. This child is an outlier
in this study and was excluded from the statistical data analysis.
The mother reported that the affected child has more frequent hand-
to-mouth activity than the siblings who did not experience similar
blood lead increases. Inspection of the house revealed exterior
14
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
surfaces with deteriorated lead paint which had been stabilized,
but not removed, as part of the original abatement work 4 to 5
years ago. Also, lead in paint and dust was identified in the
basement which had not been treated at the time of abatement since
the basement was not a finished living area. The property owner
plans to make additional repairs.
Descriptive statistics
SASฎ PROC UNIVARIATE indicated that the environmental and
blood lead distributions were skewed. As expected, use of the log
transformation reduced the amount of skewness and produced
histograms and boxplots which were approximately normal (see
boxplot Figures 1 to 13) . Descriptive statistics for blood and
environmental variables were produced after transforming the data
using the natural logarithm (In).
SASฎ PROC UNIVARIATE was used to produce descriptive
statistics for all sample types with one observation per house.
Since multiple observations were available per house for settled
dust from window sills, and window wells, floors in rooms with
windows, as well as for children's blood, additional analysis was
performed using SASฎ PROC MIXED with house as a random effect to
address the issue of clustering (i.e. multiple observations per
house). Geometric mean (GM) values, standard errors, and 95%
confidence intervals were obtained using the following PROC MIXED
models fitted separately for each study group (R&M Levels l-ni,
modern urban, previously abated), surface type (floors in rooms
with windows, window sill, window well), and matrix (dust, blood):
In(PbD) = 30 + ln(E) (Eq.l)
In(PbD-C) = 3o + ln(E) (Eq.2)
ln(DL) = 3o + ln(E) (Eq.3)
In(PbB) = 3o + ln(E) (Eq.4)
where, PbD = dust lead loading (/zg/ft2); PbD-C = dust lead
15
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
concentration (jig/g); DL = dust loading (mg/ft2) ; PbB = blood lead
concentration (ng/dL); 60 =ln(a); a = a constant; ln(E) is normally
distributed.
Side-by-side boxplots
Side-by-side boxplot figures with median traces are presented
in this report as a means of displaying lead levels across
campaigns within and across study groups. For this preliminary
report, the boxplots across time for the environmental data are
limited to the 89 houses with data available at the twelve month
campaign. Boxplots were generated using S-Plusฎ software10 (see
section 5.1 for a description of boxplot components). The
descriptive statistics presented in this report include "extreme
values" that are indicated by the symbol '*' in the boxplot
displays.
Statistical method for analysis of longitudinal data
Statistical methods for the analysis of longitudinal data have
developed rapidly over the last decade. These methods, which are
natural extensions of multiple regression and analysis of variance,
are extremely flexible. Current longitudinal methods allow for the
inclusion of random and fixed effects, longitudinal (time
dependent) covariates and constant covariates, as well as discrete
and continuous covariates, all in a multiple regression context.
In this study, for example, we have the following types of
covariates:
(1) type of house - fixed effect, discrete
(2) house - random effect, discrete
(3) dust lead - fixed time dependent continuous covariate
(4) child - random effect, discrete
(5) R&M Level - fixed effect, discrete
(6) time - fixed time dependent continuous covariate
(7) age of child - fixed time dependent covariate
(8) season - fixed discrete covariate
16
-------
Preliminary RซM Study Report-
Partial 12 Month Data 7/95
The response variable modeled was dust lead reading or blood
lead concentration (transformed using logarithms). These response
variables as well as their associated time dependent covariates
will be observed at times described in Table 1.
For the dust lead measurements let Y1 denote the vector of
responses over time for the ith house i.e. x is an nix vector of
the form Y, = (yiifyl2,... ,yllu)T where yx] is the response for the ith
house at time t: and "T" stands for the transpose operation. Then
the general form of the model is:
Y1 = XX0 + Z.b, + e1
where Xt is an /i xp matrix of covariate values for the fixed
effects, 3 is a pxl vector of parameters for the fixed effects, Zi
is an nAxq matrix of covariate values for the random effects, bx is
a qxl vector of random effect parameters and e,. is an r\xl vector
representing random error. We have N such models, one for each
house.
Estimates of the parameters in the overall model are obtained
using the methods outlined in published papers.11"17 The essential
feature of these methods is the use of weighted least squares with
a "working" estimate of the covariance matrix followed by iteration
with an updated estimate of the covariance matrix until
convergence. The estimate of the variance-covariance matrix of the
fixed effects is robust, in the sense that it is consistent,
regardless of the form of the "working" estimate of the covariance
matrix. The model for blood lead readings will be similar with the
above model specified for each child.
Our primary interest in this study is in the parameters of the
model which represent the effect of R&M interventions on dust lead
and blood lead. The fact that this model allows estimation of
these parameters in the presence of heterogeneity between houses
17
-------
Preliminary RSM Study Reporc-
Parcial 12 Month Data 7/95
and temporal correlation, and produces variance estimates which are
robust is extremely important. Should it be necessary we can, in
this framework, also consider response variables which are
discrete.
The general nature of the model makes it ideal for a study of
this type where there is the potential for unbalance. Since the
model is house or child specific, depending on whether dust lead or
blood lead is being modelled, we do not require that the number of
observations through time be equal. Thus, should a child move or
otherwise be eliminated from the study the house data can be
analyzed while the data for that child can be included up to the
point of departure. Should another child be entered into the study
at that house his or her blood lead readings can be included in the
blood lead analysis for the remainder of the study, thus providing
partial information for that child. The common residence of the
children is included in the house covariate which allows for
correlation structure between these observations.
Age-related effects in the analysis of blood lead responses
need to take into account the fact that blood lead is not linearly
related to age since it tends to increase between six months and
two years and decrease slowly among children over two years of age.
This is done by the use of linear and quadratic terms for age in
the model. The presence of several children in a house, which
introduces another source of correlation, (i.e. between children in
the same house) is accounted for by using house as a random effect
which introduces the required correlation.
SAS PROC MIXED AND GEE18 software were used for longitudinal
data analysis.
Specifications of longitudinal models for dust
In the analysis of the preliminary twelve month dust data, we
have fit the statistical model proposed in the QAPjP to the data.
The results of the compositing self study indicated that an overall
measure of lead exposure could be considered with little loss of
18
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
information. This was true for both dust lead concentrations and
lead loadings. These results suggest that the readings from
multiple sample sites in a house might be combined to produce an
overall measure to use as a covariate in the model relating
environmental lead levels to blood lead levels. Consequently, we
have explored the use of factor analysis as a method for combining
individual sample results. The results indicate that:
One factor, the first, accounts for (79% to 83%) of the
variability of environmental dust lead when all five groups
are analyzed together. The findings were similar when the
three R&M groups were analyzed separately.
The second factor measures the difference between the floor
lead readings and window sill and window well lead readings
and accounts for (12% to 14%) of the variability when all five
groups are analyzed together. The findings again were similar
when the three R&M groups were analyzed separately.
Thus far, the percentages of the variability of the dust readings
accounted for by the factor loadings have remained stable over
study groups and campaigns (see below).
Five Study Groups:
Oust
Measure
Pb Loading
Pb Cone.
Dust Loading
Initial
Campaign
factorl factor2
.81 .14
.82 .12
.61 .22
Six Month
Campaign
factorl factor2
.83 .12
.81 .14
.67 .18
Twelve Month
Campaign
factorl factor2
.83 .12
.79 .14
.69 .17
19
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
R&M Groups Only:
Dust
Measure
Pb Loading
Pb Cone.
Dust Loading
Initial
Campaign
factorl factor2
.80 .15
.73 .19
.69 .17
Six Month
Campaign
factorl factor2
.79 .15
.74 .18
.70 .16
Twelve Month
Campaign
factorl factor2
.79 .15
.75 .17
.70 .16
Given the stability of the factors over time, we feel comfortable
using them as the variable to measure environmental lead levels.
The first factor was used as the dependent variable in the
longitudinal data analysis of dust changes.
The following models were fit to the dust data:
ln(factorl)13kl = 00 + jS^season^ + 02*levellk
+ @3*campaign1 (level ฑk)
+ b^house, + eljkl
In(factorl)
+ ]32*levellk
+ 03*levelk (campaign^)
+ bl*housel +
where,
I refers to house, j to season, k to R&M Level, 1 to campaign,
level(campaign)= campaign nested within R&M level to compare dust
levels over time within R&M groups, and campaign(level)= R&M level
nested within campaign to compare R&M groups at each campaign over
20
-------
Preliminary RซM Study Report-
Partial 12 Month Data 7/95
time. Following standard practice regression coefficients
corresponding to "fixed effects" are denoted by Greek letters while
regression coefficients corresponding to "random effects" are
denoted by ordinary letters (e.g. b).
SAS PROC MIXED has been used for the analysis thus far since
a log transformation of lead values is indicated as previously
mentioned.
Specifications of longitudinal models for blood lead
In order to address the study objectives with regard to blood
lead changes, we fit two main types of models to the data. The
first model, hereinafter referred to as the exposure model, was
used to characterize the relationship between blood lead and dust
lead. in this model, the two dust lead factors were included as
dependent variables along with demographic and behavioral
variables. The second model, hereinafter referred to as the
comparison model, was used to investigate blood lead levels across
groups and within groups over time. The two models are as follows:
Exposure Model
ln(PbB)lklm = 0o + P!*factorliklm + ]32* factor 2 lklm
+ P3*ageiklm + P,*age2,klm + j35*seasonlklm
+ (36*genderlk,. + J37*mouthing,klm
+ 08*campaign1 + f3g* factor I*campaign1
+ jB10*factor2*campaign1
+ b^house, + bm(I)*childm(I) + eiklm
The initial campaign blood and dust lead values for children who
moved into the vacant R&M II and III houses after intervention were
excluded from the exposure model. Their initial PbB values at the
time of move-in reflect body burdens associated with exposures in
their past living environments and not their new home environments.
21
-------
Comparison Models
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
ln(PbB)lklm =
+ 02*age2lklm
+ J34*genderiklm + j35*mouthing,klm
+ J36*groupk + J37*campalgn1
+ j3g*groupk (campaignL)
+ b^house, + bm(I)*childmm + e
(33*seasonLklm
''iklm
ln(PbB),klm =
+ j33*seasoniklm
04*genderlklm + (35*mouthing,klm
06*groupk + /37*campaign1
jBg*campaign1(groupk)
b2*childm(I) + eiklm
where,
I refers to house, k to group, 1 to campaign, m to child within
house, group(campaign)= campaign nested within study group to
compare blood levels over time within groups, and campaign(group)=
group nested within campaign to compare groups at each campaign
over time. Following standard practice regression coefficients
corresponding to "fixed effects" are denoted by Greek letters while
regression coefficients corresponding to "random effects" are
denoted by ordinary letters (e.g. b). The comparison models were
fit using all blood lead data and then separately for children with
PbB levels <20/Ltg/dL and 220/jg/dL.
22
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
5.0 RESULTS
5.1 Side-by-Side Boxplots with Median Traces
In order to graphically display changes over time in
environmental and blood data this report provides a series of
boxplot displays with median trace lines connecting the median
values across time. These descriptive displays do not take into
account season or any other potential covariates.
In a boxplot display, 50 percent of the data is contained in
the box shown in the figure; the bottom of the box is the lower
quartile and the top of the box is the third quartile, the
horizontal line inside the box represents the sample median. The
vertical lines extending from the box represent the expected lower
and upper range of the data based on the variability of the central
portion of the data. The fences are 1.5 interquartile ranges from
the upper and lower edges of the box. Extreme values are indicated
by an asterisk.19 The widths of the boxes in any given side-by-side
boxplot are proportional to the number of observations.
Dust Data Boxplots
Figures 1-12 are boxplot displays presented by study group for
each of the main surface types (floors in rooms with windows,
window sills, window wells, interior entryways) showing the
distributions of dust lead loadings, dust lead concentrations, and
dust loadings, respectively, across campaigns. These displays
allow for comparisons both within and across groups over time.
The median traces for lead loadings across most surface types
in R&M Level I-III houses show a pattern of reduced levels at post-
intervention, most pronounced for R&M III, followed by increases at
two months to levels that remained below pre-intervention levels.
At six and twelve months, the median lead loadings tended to remain
relatively stable; R&M I and II houses had unchanged or moderately
increased levels across surface types and R&M III houses had
unchanged or moderately decreased levels (Figures 1-4). Deviations
from this pattern include: (a) interior entryways in R&M I houses
23
-------
Preliminary RsM Study Report-
Parcial 12 Month Data 7/95
which did not have reduced levels at post-intervention and floors
in R&M I houses which did not show a pattern of rebound at two
months and (b) floors and entryways in R&M II houses which had
decreased levels at post-intervention but no increases at two
months (Figures 1 and 4).
For dust lead concentrations, the median traces reveal a
downward trend at post-intervention and at two months across sample
types that is most pronounced in R&M III and R&M II houses compared
with R&M I houses. At six and twelve months, the lead
concentrations tend to remain relatively stable in R&M I houses and
show unchanged or moderately decreased levels across surface types
in R&M Level II and III houses (Figures 5-8).
The median traces for dust loadings in the three R&M groups
show patterns similar to those described above for the lead
loadings (Figures 9-12).
In the modern urban and previously abated control houses the
median traces for dust show a pattern of relatively stable lead
loadings, lead concentrations, and dust loadings over time with a
slight downward trend at six and twelve months for lead loadings
and dust loadings (Figures 1-12).
Blood Data Boxplots with Median Traces
Figure 13 provides boxplot displays of the unadjusted blood
lead concentrations (jug/dL) over time by study group for the subset
of children with initial PbB <20 /ng/dL. The outlier in the PA
group at twelve months with a PbB of 53 /ug/dL is excluded from
Figure 13 and the statistical data analysis for reasons explained
in section 4.4. The median traces for all five study groups
indicate modest or little change over time.
24
-------
Preliminary RSM Study Report-
Partial 12 Month Data T/95
5.2 Descriptive Statistics - Twelve Month Campaign
Settled Dust
Descriptive statistics for settled dust at the twelve month
campaign are displayed as follows:
Bar graphs of GM dust lead loadings (PbD, Mg/ft2), dust lead
concentrations (PbD-C, Mg/g), and dust loadings (DL, mg/ft2)
by study group and surface type (Figures 14 to 16).
Boxplot displays of PbD, PbD-C and DL distributions by study
group and surface type (Figures 17-19).
Descriptive statistics (geo. mean, n, min, max, sd) for the
twelve month PbD, PbD-C, DL respectively, for the five study
groups and seven surface types (interior entryways, floors in
rooms with and without windows, window sills, window wells,
upholstery and air ducts) are presented in Appendix A, Tables
a-c) . Sample sizes for upholstery and air duct data are
limited for reasons provided in Table 2. The geometric means
and 95% confidence intervals for floors in rooms with windows,
window sills, and window wells are the values produced by an
analysis that takes into account clustering (section 4.4).
Geometric mean (GM) lead loadings across all groups and
surface types at the twelve month campaign were <603 Mg/ft2, except
for air ducts in all groups (range of GMs 856 jug/ft2 (MU) to 18073
Mg/ft2 (R&M I)) and window wells in R&M I (19412 ng/fi ), R&M II
(1,761 /ig/ft2), and previously abated houses (1172 ng/ft2) (Appendix
A, Table b). For R&M Levels I-lll, respectively, the GM PbD values
were 112, 79, and 36 fig/ft2 for floors in rooms with windows; 603,
201, and 23 fig/ft2 for window sills and 19412, 1761, and 268 Mg/ft2
for window wells.
GM dust lead concentrations across all groups and surface
types at twelve months were <4000 Mg/g (ppm), equivalent to <0.40%,
except for window sills (7754 Mg/g) and window wells (22963 Mg/g)
in R&M I houses (Appendix A, Table a).
At the twelve month campaign, GM dust loadings across surface
types were <340 mg/ft2 except for window wells in MU, PA, and R&M
I and R&M II groups (range 456 to 841 mg/ft2) and air ducts (range
25
-------
Preliminary FUM Study Report-
Partial 12 Month Data 7/95
8474 to 15184 mg/ft2)(Appendix A Table c).
Modern urban houses continued to have the lowest lead loadings
at the twelve month campaign. GM values across surface types were
all <217 Mg/ft2 except for air ducts. For floors, windows sills,
and window wells at twelve months, R&M I houses had significantly
higher GM PbD values (112, 603, and 19412 jug/ft2, respectively)
compared to R&M III houses (36, 23, and 268 /ig/ft2, respectively);
GM values in R&M II houses were intermediate (79, 201, and 1761
Mg/ft2, respectively).
At twelve months, modern urban houses continued to have the
lowest GM PbD-C levels across surface types (<440 /ug/g) . The GM
PbD-C values for interior entryways and interior floors across the
other four study groups (R&M I-III and PA) were not statistically
different from each other (Appendix A, Table a) . For windows sills
and window wells, R&M I houses had significantly higher GM PbD-C
values (7754 and 22963 M9/9/ respectively) compared to R&M III
houses (796 and 1130 ng/g, respectively), with R&M II houses
intermediate (2941 and 3862 pq/q, respectively).
Study groups were most similar to each other in terms of dust
loadings. However, dust loadings tended to be highest in R&M I
houses, lowest in R&M III houses and intermediate in R&M II houses.
For windows sills and window wells, R&M I houses had significantly
higher GM dust loading values (78 and 841 mg/ft2, respectively)
compared to R&M III houses (29 and 237 mg/ft2, respectively), with
R&M II houses intermediate (68 and 456 mg/ft2, respectively)
(Appendix A Table c).
Dust Lead Correlations between Surface Types
Statistically significant correlations (p<.05) were found between
the dust lead levels from most of the surfaces types at the twelve
month campaign. Tables d-f in Appendix A display the correlation
matrices for dust load loadings, lead concentrations, and dust
loadings, respectively. The highest correlation coefficients were
for window sills and window wells (.73 for PbD and .67 for PbD-C).
26
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
Overall Summary Measures of Dust Data
Summary measures of the dust data for each house were
calculated based on the weighted average of the readings across the
four surface types common to all campaigns, i.e., floors,
entryways, window sills, and window wells. The weighting factor
for lead loadings and dust loadings was the surface area sampled.
For lead concentrations the weighting factor was the sample mass.
Figure 20 displays the overall geometric mean PbD, PBD-C and
DL values at the twelve month campaign by study group. Figures 21-
23 display the GM values for the three overall dust measurements by
study group for each of the five data collection campaigns through
the twelve month campaign. Descriptive statistics (GM, range) for
the overall PbD, PbD-C, and DL readings respectively by campaign
are displayed in Tables 5-7. Based on these summary measures,
overall GM lead loadings were approximately 35 times higher in R&M
I houses as compared with R&M III houses. This difference is due
to order of magnitude higher lead concentrations and several fold
higher dust loadings in R&M I houses relative to R&M III houses.
Drip-Line Soil
Drip-line soil samples were not collected at the twelve month
campaign. Therefore, this report provides preliminary six month
data on lead concentrations of drip-line soil (PbS) by study group
(Appendix A Table g) . These data are limited due to the fact that
most study houses have no drip-line soil. The PbS levels in modern
urban houses remained similar to the initial levels (six month
GM=73 nq/g, range 34 to 229 Mg/g versus initial GM=63 ng/g, range
29 to 154 Mg/g) Across previously abated and R&M houses
individual PbS values ranged from 182 to 7845 /xg/g at six months
compared to the range of 233 to 15968 Mg/g observed at pre-
intervention/baseline.
Drinking Water
Drinking water samples also were not collected at the twelve
month campaign. Water lead concentrations (PbW) at six months were
27
-------
Preliminary RSM Study Report-
Partial 12 Month Daca 7/95
unchanged from their geometric mean baseline levels of ฃ4 /ig/L
(ppb) across groups. The range of values also remained the same -
less than the instrumental limit of detection (
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
clustering) (max n=79).
5.3 Longitudinal Data Analysis
Dust Lead
The longitudinal models described in section 4.4 were fit to
the data for lead loadings, lead concentrations, and dust loadings
from the three R&M groups and all five study groups. Fitting the
models to the data from the three R&M groups allowed for
investigation of changes across all five campaigns (initial, post-
intervention, and two, six, and twelve month) , two of which (post-
intervention and two month) were not common to the control groups.
The three campaigns common to all study groups were included in the
five group models.
In these models the dependent variable was factorl obtained
from the factor analysis, a factor that accounted for most of the
variability of environmental dust lead. For each dust variable
(PbD, PbD-C, DL), the model was fit to the data with campaign
nested within study groups in order to examine trends over time
within study groups, and with study group nested within campaign in
order to compare study groups at each campaign.
The main findings of the longitudinal analysis of the dust
data after controlling for season are listed below. In the
presence of other covariates (campaign, group), season was a
significant fixed effect in the lead loading and dust loading
models but not in the lead concentration models.
The PROC MIXED output is displayed in Appendix B. The
printouts have the following codes for campaign and group: 00=pre-
intervention, PI=post-intervention, 02=two months, 06=six months
and 12=twelve month, IN=initial campaign (5 group model), 1=R&M I,
2=R&M II, 3=R&M III). Interpretation of the estimates obtained by
SAS PROC MIXED obey the usual rules of interpretation of regression
coefficients, i.e., the coefficient of a covariate is the expected
change in the response variable associated with a unit change in
29
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
the covariate in the presence of the other covariates. When the
covariate is a dummy variable, a unit change in the covariate
corresponds to the expected difference between the response at the
level of the covariate compared to the omitted level. If the
effect is a nested effect, e.g., level nested within campaign the
coefficient represents the comparison of that level versus the
omitted level for a fixed campaign and this allows the
determination of differences between levels of intervention.
Dust Lead loadings:
R&M group comparisons across time
Despite higher baseline levels, lead loadings were
significantly lower (p<.01) in R&M III and R&M II houses as
compared with R&M I houses at immediately post-intervention,
two months (R&M III only), six months, and twelve months. At
each campaign, PbD was lowest in R&M III, highest in R&M I,
and intermediate in R&M II.
Changes over time within R&M groups
For all three R&M groups, post-intervention lead loadings were
significantly lower than their pre-intervention levels. For
R&M II and III houses, PbD levels at two months, six months,
and twelve months were significantly higher than their
corresponding post-intervention levels. In R&M I houses PbD
levels over time were not significantly different from their
immediate post-intervention levels.
Lead concentrations:
R&M group comparisons
Pre-intervention dust lead concentrations (PbD-C) were not
significantly different across the three R&M groups. PbD-C
levels were significantly lower (generally p<.01) in R&M II
and III as compared with R&M I at immediately post-
intervention, two months, six months and twelve months. At
each campaign PbD-C was lowest in R&M III, highest in R&M I,
and intermediate in R&M II.
30
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
Changes over time within R&M groups
For R&M II and III (but not R&M I), lead concentrations were
significantly reduced (p<.01) at immediately post-
intervention. For both R&M I and II houses, PbD-C levels at
two months, six months and twelve months were significantly
lower than their corresponding levels at immediately post-
intervention. In R&M III, only the twelve month level was
significantly lower (p<.01) than the post-intervention level.
Dust loadings:
R&M oroup comparisons
At pre-intervention, dust loadings (DL) were significantly
higher in R&M II and III houses as compared with R&M I houses.
At each point in time post-intervention the estimates for R&M
II and III were negative relative to R&M I; however, none of
differences reached statistical significance. R&M II houses
had intermediate DL levels at each follow-up campaign.
Changes over time within R&M groups
In all three R&M groups, dust loadings were significantly
reduced (p<.01) at immediately post-intervention. None of the
follow-up levels were significantly different from their post-
intervention levels although the levels tended to be higher at
follow-up (positive coefficients).
Control houses:
Dust lead loadings in the modern urban (MU) houses were
significantly lower (p<.01) than those in the three R&M groups
and the previously abated (PA) houses at baseline, six month
and twelve months. PbD levels tended to decrease over time in
MU and PA houses; however, none of the changes were
statistically significant.
Dust lead concentrations at baseline, six month and twelve
months were significantly lower in MU houses as compared with
each of the other four groups. At six and twelve months, lead
concentrations in MU and PA houses were not significantly
different from their baseline levels.
Dust loadings at baseline were significantly lower in modern
urban houses as compared with each of the R&M groups but not
the previously abated group. At follow-up, none of the study
groups had dust loadings that were statistically different
from those in the modern urban study houses.
31
-------
Preliminary RSM Study Report-
Parcial 12 Month Data 7/95
Blood Lead
The exposure and comparison models described in section 4.4
were fit to the blood lead data for children in the three R&M study
groups and for children in all five study groups. Fitting the
models to data from the three R&M groups allowed for investigation
of changes across all four PbB testing campaigns (initial, two
month, six month and twelve month). The three campaigns (initial,
six month, twelve month) common to all five study groups were
included in the five group models.
The main findings of the comparison models for investigating
PbB changes within and between groups are listed below. The PROC
MIXED output is displayed in Appendix C with the following codes:
(campaigns)0=pre-intervention/initial, 2=two month, 6=six month and
12=twelve month, L1=R&M Level I, L2=R&M II, L3=R&M III.)
R&M group comparisons
For children with initial PbB <20 jiig/dL, no statistically
significant blood lead differences were found between the
three R&M groups at baseline and two, six and twelve months
post-intervention controlling for age and season. R&M I
children tended to have lower PbB levels at each campaign,
including baseline, relative to R&M III children.
Changes over time within R&M groups
For children with initial PbB <20 M9/dL in each of the three
R&M groups, no statistically significant changes in PbB were
found at the two, six and twelve month campaigns controlling
for age and season. However, a consistently downward but
nonsignificant trend in blood lead was found in children in
R&M III houses. (The coefficients at two, six, and twelve
months were negative and increased in magnitude over time.)
32
-------
Preliminary R4M Study Report-
Parcial 12 Month Data 7/95
Control Houses; Chances within groups and group comparisons
For children with initial PbB <20 jxg/dL in the previously
abated control houses, no statistically significant PbB
changes were found at the six and twelve month campaigns
controlling for age and season.
Controlling for age and season, children in modern urban
houses had PbB levels that were lower and statistically
different from those of children in each of the other four
study groups at the initial, six and twelve month campaigns.
Modern urban children had a small but statistically
significant increase in blood lead concentration over baseline
at the six month campaign only.
PbB changes in children with baseline PbB s?20 ug/dL
For the small numbers of children with initial PbB 120 /ig/dL,
a downward trend in blood lead concentration was found at the
two, six and twelve month campaigns. These changes were
statistically significant (p<.05) at twelve months in R&M II,
R&M III and PA groups, controlling for age and season.
Exposure Model
The main findings of the exposure models for investigating the
relationship between blood lead and dust lead are listed below (see
Appendix C for the PROC MIXED output):
Age, age2, and season (summer vs nonsummer) were significant
contributors to the 3 and 5 group models; gender and hand-to-
mouth activity (high vs low) were not. Blood lead tended to be
higher in males and significantly higher in males in the 5
group model with dust loadings only.
Controlling for age and other covariates (campaign, group,
dust factors) included in the various blood lead models, the
seasonal change in children's PbB was estimated to be +1.3
/xg/dL in summer relative to winter.
Using all five study groupsr dust lead loadings and
concentrations (factor1 and factor2) and dust loadings
(factor2 only), were significantly related to children's PbB
after adjusting for age, season, campaign and the inclusion of
random effects for houses and multiple children per house.
33
-------
Preliminary RsM Study Report-
Partial 12 Month Data 7/95
Using the three R&M groups. dust lead loadings and
concentrations (factor1 and factor2) and dust loadings
(factorl), were not significantly related to children's PbB
after adjusting for age, season, campaign and the inclusion of
random effects for houses and multiple children per house.
Factor2 was a significant factor in the PbB model with dust
loading factors.
The interactions of factorl and factor2 with campaign were not
statistically significant for lead concentration factors and
dust loading factors. For lead loading, the interaction of
factor2 and campaign was only marginally significant. For
these reason, the exposure models were rerun without these
interaction terms.
6.0 DISCUSSION
To date, this study has met its enrollment goals and has
completed nearly all of the planned data collection across the
first four sampling campaigns. Laboratory performance and data
quality objectives continue to be met. The study also continues to
benefit from a high level of family cooperation with both blood
lead and environmental sampling well into the twelve month
campaign. Through the twelve month campaign approximately 20
percent of the original families in the 60 R&M houses included in
this report moved. In every case, we successfully enrolled the
next family to occupy the house, assuring that at minimum we can
continue to do the environmental sampling. Most new families also
had eligible children who were enrolled in the blood lead testing
component of the study.
Dust Lead
All three levels of R&M intervention under investigation were
found to be associated with overall reductions in interior dust
lead loadings, lead concentrations, and dust loadings that were
sustained for twelve months below pre-treatment levels. Moreover,
the degree of reduction in lead exposure was positively associated
with the level of the R&M intervention, and statistically
significant differences were found between R&M groups over time.
34
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
After intervention and at two, six, and twelve months, lead
loadings and lead concentrations were lowest in R&M III houses,
intermediate in R&M II houses, and highest in R&M I houses. A
similar but nonstatistically significant pattern was observed for
dust loadings over time post-intervention (Figures 21-23).
At the outset, R&M II and III houses had significantly higher
lead loadings and dust loadings as compared with R&M I houses
despite random assignment of houses to intervention groups as
explained in section 2.0. Since lead concentrations at the outset
were similar, the R&M group differences in lead loadings were due
to higher dust loadings which in turn may be attributed to the fact
that half of the R&M II and all of the R&M III houses were vacant
at the outset. Lead dust has been shown to reaccumulate in vacant
houses.20
Reaccumulation of lead-containing dust in R&M houses was
greatest during the first two months post-intervention as compared
to the relatively stable period from two months to twelve months
(Figures 1-8 and 21-23). This early reaccumulation in R&M II and
III houses may be due to a combination of reaccumulation of lead in
dust after intervention and prior to occupancy as well as the
potential importation of lead containing dust into the house during
move-in by study families.
At the twelve month campaign, overall weighted averages of the
dust lead loadings (based on floor, window sill, window well, and
interior entryway samples) were 26-fold higher in R&M I houses as
compared with R&M III houses, and 5-fold higher in R&M I houses as
compared with R&M II houses (Table 5). These substantial
differences in lead exposure are attributable mainly to differences
in lead concentrations and secondarily to differences in dust
loadings (Figure 20) . This was also true with regard to the
observed 2.7-fold difference in overall PbD levels between R&M III
and the modern urban houses at twelve months (Figure 20).
35
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
Differences in overall dust lead concentrations between R&M
groups following intervention were expected since R&M III
interventions addressed lead paint sources to a greater degree
(e.g. by window replacement) than either R&M I or R&M II
interventions. Despite this, it is interesting to note that dust
lead concentrations in R&M I and II houses at two months, six
months, and twelve months were significantly lower than their
corresponding levels immediately post-intervention. This was also
the case with R&M III houses at the twelve month campaign. These
findings may be due to a reduced rate of input of lead into dust
from high lead-content paint sources during the one-year period of
follow-up. Given a reduced rate of high lead input, housekeeping
by families might then further remove some of the remaining
reservoir of highly concentrated lead dust. Such dust may have
accumulated in R&M houses during the last fifty years or more since
the time they were built. A rise in the lead concentration (Mg/g)
of settled house dust in future campaigns might signal the
presence of new lead paint hazards and the need for further
remediation activities.
The patterns observed in dust loadings and concentrations
across R&M groups (highest in R&M I, lowest in R&M III) may be due
to the degree to which smooth and easily cleanable surfaces were
provided as part of the interventions. The provision of smooth and
easily cleanable surfaces has been shown to be an important element
of effective residential lead paint abatement.21 In this study,
surface conditions would have influenced the effectiveness of the
post-R&M cleanup by contractors as well as the housekeeping of the
resident families.
One year following intervention, R&M III was the only R&M
group to have dust lead loadings on a geometric mean basis that
were less than the current EPA guidance levels of 100 /^g/ft2
(floors) , 500 M9/ft2 (window sills), and 800 nq/ft2 (window wells).
The twelve month GM PbD levels in the R&M groups for floors, window
36
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
sills and window wells, respectively were 112, 603, and 19412
jug/ft2 for R&M I; 79, 201, and 1761 /ig/ft2 for R&M II; and 36, 23,
and 268 jttg/ft2 for R&M III. It should be noted that although the
R&M cyclone has been shown to produce higher estimates of dust lead
loadings compared to wipes across a range of surface types and
conditions, the cyclone tends to yield lower PbD estimates on
smooth surfaces with low lead loadings (<~100 /xg/ft2) as compared
with wipe sampling.6
Control groups
The modern urban and previously abated control houses were
characterized by the relative stability of lead loadings, lead
concentration and dust loadings over time (Tables 5-7) . The
downward trend found in dust loadings and lead loadings in these
houses may be attributable in some part to (a) families becoming
more aware of the importance of lead dust control as a result of
their study participation and (b) the fact that dust was removed
from their homes by the process of environmental sampling. This
applies to the intervention groups as well. Moreover, this study
was not designed to assess the specific effects of education on the
study outcomes across groups (e.g. study has no education only
control group) but rather to investigate the effectiveness of the
R&M interventions as a whole which included the provision of
information to all families as explained in section 2.0.
The modern urban study houses are all located in housing
subdivisions built after 1978 and were presumed to be free of lead-
based paint. This presumption is supported by the consistently low
overall interior dust lead concentrations (GM *300 jug/g (ppm),
equivalent to 0.03%) and low soil lead concentrations (GM*75 /xg/g).
At each campaign in which they were tested, the modern urban houses
had significantly lower lead loadings and concentrations as
compared with each of the other study groups. Dust loadings in the
modern urban houses were significantly lower than those in the
other study groups at the initial campaign only.
37
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
In the previously abated group, geometric mean lead loadings
and concentrations tended to be bracketed by the levels found in
R&M II and III houses at the six and twelve month campaigns. The
latter campaigns were conducted approximately four to five years
after the houses had been abated. As was illustrated by the case
in which a child's blood lead rose to 53 /ig/dL during follow-up,
the previously abated houses were comprehensively but not fully
abated of lead paint. Some interior (basement) surfaces that had
not been treated originally due to resource limitations and some
exterior surfaces that had been stabilized as part of the original
abatement were found to be in deteriorated condition and a likely
source of this child's exposure along with deteriorated exterior
paint observed on neighboring houses. This case also points to the
need for ongoing inspection and maintenance of houses that receive
any type of interim control or partial abatement intervention and
the need for long-term follow-up of study houses to assess long-
term effectiveness of the R&M interventions.
Soil and Drinking Water
Due to low concentrations, drinking water was not found to be
an important source of lead exposure in study children. The
geometric mean water lead concentrations across study groups were
all ฃ4 M9/L (ppb) at the initial and six month campaigns. Only a
small number of readings exceeded the EPA drinking water standard
of 15 jug/L.. The maximum reading was 44 Aig/L.
Soil lead data were limited due to the absence of drip-line
soil at most study houses except for the modern urban houses. Soil
lead concentrations ranged from 30 to 16,000 Mg/g across houses at
the initial and six month campaigns. GM levels ranged from 700 to
730 ng/g in R&M I and II houses at the six month campaign (Appendix
A, Table g). The low levels (GMs of 63 to 73 /xg/g) measured in the
drip-line soil next to the modern urban houses over time suggest
that replacement sod or soil might have been used at these houses
at, or since, the time of construction.
38
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
Blood lead
The majority of U.S. children with elevated blood lead levels
as defined by the U.S. CDC (*10 nq/dL), have levels in the range of
10 to 20 ng/dL-22 Little is known about PbB changes associated with
lead paint hazard remediation interventions in the homes of
children with this low-to-moderate level of lead toxicity. Most of
the children in the three R&M groups had baseline PbBs in the range
of 5 to 20 /jg/dL. They were recruited from households residing in,
or moving into, eligible study houses identified from lists
provided by collaborating property owners who manage low-income
rental property in minority neighborhoods in Baltimore City.
Children in these types of neighborhoods are at known high risk of
lead poisoning.
All study children are African Americans living in low-to-
moderate income housing. At the outset, their geometric mean (GM)
ages across the five study groups ranged from 25 months to 33
months. Their baseline GM PbB values across groups ranged from 10
to 14.5 /*g/dL. It can be noted here that in U.S. non-Hispanic black
children from low-income families living in central cities
(populations *1 million, 1988-1991), the GM PbB was recently
estimated at 9.7 /ng/dL.22
The R&M study's maximum baseline PbB was in a child who moved
into a vacant R&M Level III house with a PbB reading of 42 nq/dL.
As anticipated, the largest numbers of children with baseline PbBs
>20ng/dL were in the R&M II and III groups. This was due the fact
that one of the housing organizations had a policy of renting its
improved properties to families with lead poisoned children.
Children in the three R&M groups with PbB <20 /xg/dL at the
outset tended to remain in that range throughout the twelve month
period of follow-up. After controlling for age and season, an
increasingly downward trend in PbB level was noted across the two,
six, and twelve month campaigns in the R&M III group, the group
that also had the lowest house dust lead levels over time.
39
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
Furthermore, no statistically significant PbB changes were found in
any of the R&M groups in children with baseline PbB <20 /xg/dL. One
could hypothesize that the R&M interventions prevented an increase
in blood lead that study children might have experienced otherwise
in the absence of the R&M interventions. For ethical reasons, the
study design did not include a non-intervention control group
against which to test this hypothesis.
R&M children with pre-intervention PbBs *20 ng/dL did
experience statistically significant reductions in blood lead over
time to levels below 20 nq/dL at twelve months. Swindell et. al.
recently reported that blood lead levels also declined up to one
year post-abatement in children with baseline PbBs ^20 /ig/dL among
moderately lead poisoned children in central Massachusetts.23 Their
study was based on a retrospective record review and did not
include the collection of environmental lead data pre- and post-
abatement. Our finding of an absence of a blood lead rise in
children with initial PbBs <20 M9/dL differ from those of Swindell
et. al. who reported an overall rise in PbB from 16.8 ^g/dL to 19.3
jig/dL at up to one year post-abatement in children with pre-
abatement PbB <20 jitg/dL.23
A noteworthy finding in the R&M study is the absence of
statistically significant increases in children's PbB measured at
two months post-intervention, especially in the R&M I and II houses
that were occupied at the time of treatment. This finding,
together with the dust lead data, provides some evidence that the
precautions used in R&M houses were successful in preventing
significant short-term increases in children's PbB that have been
attributed to improper abatement practices in past studies.24'25
Precautions included having children out of the house while R&M
work was in progress, and the use of work practices to minimize,
contain, and cleanup lead dust.
40
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
Control groups
The findings for children with PbBs <20 /ig/dL and *20 jug/dL in
the previously abated group were similar to those described above
for children in the R&M groups. In contrast to this group and the
three R&M groups, the GM PbBs across time (range:3-4 jiig/dL) in
children living in the modern urban study houses were similar to
the national GM of 3.6 /xg/dL reported for U.S. children aged 12 to
60 months.22 The GM estimated for all U.S. non-Hispanic black
children in this age range was 5.6 /xg/dL.22 Unlike the other four
study groups, nearly all of the modern urban houses were occupied
by owner occupants. Also, as reported previously, the mean monthly
mortgage/rent payment was higher ($406) in this group as compared
with the other groups (range of means:$288 to $330).2
Additionally, all but one of the PbB readings in the modern
urban group across time were below the CDC's level of concern (10
/xg/dL) . One child had a PbB increase from 3 to 10 jug/dL at the six
month campaign which may have contributed to the significant PbB
rise found between the initial and six month campaigns in this
group. The PbB levels in children in the modern urban group were
significantly lower than those in the children in each of the other
four study groups at the initial, six and twelve month campaigns.
This finding is the only statistically significant difference in
PbB found in the R&M study to date among children who had an
initial PbB <20 /xg/dL.
Exposure Model
Using data from all five study groups, the exposure models
(longitudinal data analysis) indicated that the blood lead levels
of study children were significantly related to their house dust
lead loadings and lead concentrations after controlling for
significant fixed effects, i.e., age, (age2), and season. The
guadratic term indicates a nonlinear relationship between blood
lead and children's age. The absence of a statistically significant
relationship between dust lead and blood lead in the three R&M
41
-------
Preliminary RsM Study Report-
Partial 12 Month Data 7/95
group exposure model was likely due to the narrower range of dust
lead levels over time post-intervention as compared with pre-
intervention and the absence of the low-lead modern urban houses
from the analysis. Other studies, including the recent study in
Rochester,26 have found a relationship between lead in children's
blood and settled dust in their homes.
Based on the various longitudinal data analyses performed, the
seasonal change in children's blood lead levels was estimated to be
+1.3 /ig/dL in summer relative to winter. Others have reported
seasonal trends in children's blood lead levels for different years
and populations which vary in the estimated magnitude of the
seasonal difference (unpublished review, PL Reagan, July 1992) .
Considerations in interpretation of PbB findings
A number of factors that can mediate a child's blood lead
response to an intervention, including age, degree of hand-to-mouth
activity, total cumulative body lead burden, the timing of the PbB
measurements, and neighborhood housing characteristics need to be
considered in the interpretation of blood lead findings from this
or any other intervention study. For this study, these factors are
of particular relevance with regard to the following blood lead
findings:
The lack of statistically significant PbB changes within and
across R&M groups in children with baseline PbB <20 /zg/dL
despite significant differences in house dust lead levels
between R&M groups.
The consistently downward, but not statistically significant,
trend in PbB over time post-intervention in the R&M III group.
The significantly lower PbBs in modern urban children relative
to all other groups, in particular R&M III where dust lead
exposures were most similar to those in modern urban houses.
At enrollment, the mean ages of the children in the R&M and
previously abated groups were in the range of 25 to 33 months. On
average, these children were past the age range at which blood lead
42
-------
Preliminary RtM Study Report-
Partial 12 Month Data 7/95
levels tend to peak (i.e., 18-24 months of age).27 Given their
reported housing histories, it is likely that the R&M study
children spent most or all of their young lives prior to enrollment,
in low-income housing with potentially high lead dust and paint
exposures. For the modern urban group, the blood lead and housing
history data suggest that the children were at lower risk of high
lead exposure prior to enrollment compared to the other groups.
Most children in the modern urban group were residents of the same
house since birth.
Assuming that children in the R&M and previously abated houses
had chronically elevated PbBs prior to enrollment, additional time
beyond twelve months post-intervention may be needed to measure
significant PbB changes in R&M children for the reasons mentioned
briefly below. Lead in blood reflects a mixture of recent exposure
and body stores of lead. Most (approximately 70%) of the lead in
children is stored in bone.28 The half-life of lead in human adult
cortical bone is estimated to be 20 years.29 Skeletal lead can be
an ongoing internal source of lead measured in blood even after
external exposure and children's lead ingestion are reduced
following lead remediation interventions. Unfortunately, the bone
lead concentrations of study children are unknown and the kinetics
of lead mobilization from bone is not well understood in children.
This makes it difficult to estimate the magnitude and duration of
bone lead's contribution to children's blood lead measured in the
post-intervention phase of this study.
In an earlier study in Baltimore, children with GM PbBs of 63
Mg/dL prior to receiving inpatient chelation therapy were monitored
after discharge to "lead free" public housing and housing abated
according to local ordinances in effect prior to 1982. These
children had post-discharge GM PbBs at one month that changed very
little prior to 24 to 30 months post-discharge.30 This earlier
study highlights the need to continue the investigation of PbB
changes in the R&M Study at the eighteen and twenty-four month
43
-------
Preliminary RsM Study Report-
Partial 12 Month Data 7/95
campaigns.
At the twelve month campaign, most study children were 36 to
48 months of age, an age range in which the frequency of children
engaging in mouthing behavior is likely to be less compared to
younger children. Since hand-to-mouth activity is recognized as a
a major route of entry of lead into pre-school children,27 less
frequent hand-to-mouth activity over time might account in part for
the lack of statistically significant blood lead changes within and
between R&M groups in children with baseline PbB <20 jug/dL, despite
the differences in dust lead exposure between and within groups
over time. To date, hand-to-mouth activity has not been found to be
a significant covariate in the blood lead exposure models. This
may be due in part to the more or less truncated PbB distribution
and relatively small sample sizes.
The small number of children with initial PbB *20 /ig/dL at the
outset may have had higher blood lead levels due to more frequent
hand-to-mouth activity. Therefore, they may have had a relatively
greater contribution to their blood lead from their current
exposure versus bone lead compared to the children with PbB <20
/xg/dL. Thus, their PbB levels may have been more responsive to the
reduction in lead exposure associated with the R&M interventions as
compared to children with lower baseline PbB levels.
Neighborhood housing characteristics may have contributed to
the observed PbB differences between the modern urban group and the
other four study groups. By design, the modern urban study houses
were all located in housing subdivisions built after 1978 and
presumably free of lead-based paint. As mentioned above, this
presumption is supported by the very low dust and soil lead
concentrations found in the modern urban study houses. In addition
to having low exposures to lead in their own house dust and soil
(up to more than two orders of magnitude lower than those in the
other study groups) , children in the modern urban houses also have
fewer opportunities to be exposed to high lead sources in, or
44
-------
Preliminary RCM Study Reporc-
Partial 12 Month Data 7/95
originating from, neighboring houses by virtue of the fact that
they are located in subdivisions with similar low lead housing.
On the other hand, children in each of the other study groups,
including the previously abated group, share a common
characteristic of being located in neighborhoods with older lead-
painted rowhousing built before 1940 and in generally poor
condition. These neighborhoods, and the blocks in which the R&M
houses are located, often have abandoned and boarded housing as
well. Compared to the modern urban group, children in the R&M and
previously abated groups have had more opportunities for exposure
to high lead sources in paint and dust from surrounding houses.
Neighboring houses can be sources of lead particles and paint chips
which can be tracked or blown inside study houses. This was the
rationale for including walk-off mats at the main entryways in R&M
houses. Neighboring houses can also be sites where children are
directly exposed to lead during visits and outside play activities.
The study sample sizes of 20 to 26 children per group across
campaigns limits the degree to which small proportional changes
over time in PbB can be detected with high statistical power. The
group differences in PbB between the modern urban group and each of
the other four study groups were found to be statistically
significant (differences on the In scale of approximately 0.8 or
ratio of approximately 2:1), whereas the smaller log differences in
blood lead levels over time in the R&M III group were not large
enough to be statistically significant (alpha s.05) in this study.
In closing, it should be emphasized that by their nature, the
three R&M interventions under investigation are interim control or
partial abatement approaches to reducing lead hazards in housing.
As such they are not expected to be long-lasting or permanent. A
major study objective remains the determination of the longevity of
the R&M interventions. To date, dust lead loadings at specific
sites in individual study houses (particularly R&M I houses) have
reaccumulated to levels close to pre-intervention levels. However,
45
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
none of the R&M interventions in individual houses through the
twelve month campaign have failed on a wholesale basis, whereby all
or most of the interior dust lead samples have had lead levels at
or above pre-intervention levels. If wholesale failures do occur,
we will use contingency funds to make additional repairs and
reclean houses as needed.
Lastly, it is recognized that the findings reported herein are
preliminary in nature and that the results can change as more data
from the twelve month and subsequent campaigns become available.
We anticipate that data from the twelve month campaign in R&M
houses will be complete and available for inclusion in the eighteen
month preliminary report scheduled for February of 1996. Future
campaigns will allow us to investigate the longer-term changes in
lead in settled house dust and children's blood and move towards
the overall research goal of contributing to the scientific basis
for a standard of care for lead-painted houses. Also, it is hoped
that future campaigns will provide sufficient longitudinal data on
children born into study houses to allow for a separate analysis of
a subgroup which has the potential to increase our understanding of
the role of R&M intervention in the primary prevention of lead
poisoning in children.
46
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
7.0 REFERENCES
1. "Quality Assurance Project Plan for the Kennedy Krieger
Institute Lead Paint Abatement and Repair and Maintenance
Study in Baltimore" November 1992. (Battelle Subcontract No.
41950(2348)-2207; EPA Contract No. 68-DO-0126) Office of
Pollution Prevention and Toxics, Design and Development
Branch, Washington DC.
2. "Draft Final Report - Descriptive Statistics from the Initial
Sampling Campaign of the Lead Paint Abatement and Repair &
Maintenance Study in Baltimore," Submitted by Kennedy Krieger
Research Institute, December 6, 1994, (under EPA Contract No.
68-D4-0001) to the Office of Pollution Prevention and Toxics,
Design and Development Branch, Washington DC.
3. U.S. Agency for Toxic Substances and Disease Registry. "The
Nature and Extent of Lead Poisoning in United States Children:
A Report to Congress," Washington, DC 1988.
4. Farfel MR, Bannon D, Lees PSJ, Lim BS and Rohde CA. 1994.
Comparison of Two Cyclone-Based Collection Devices for the
Evaluation of Lead-Containing Residential Dusts. Journal of
Industrial and Occupational Hygiene. 9:212-217.
5. Farfel MR, Bannon D, Chisolm JJ Jr, Lees PSJ, Lim BS and Rohde
CA. 1994. Comparison of a Wipe and a Vacuum Collection Method
for the Determination of Lead in Residential Dusts.
Environmental Research, 65:291-301.
6. Farfel MR, Lees PSJ, Rohde CA, Lim BS and Bannon D. 1994.
Comparison of Wipe and Cyclone Methods for the Determination
of Lead in Residential Dusts. Applied Occupational &
Environmental Hygienet 9:1006-1012.
7. Research Triangle Institute and Engineering Plus. "Development
of a High Volume Small Surface Sampler for Pesticides and
Toxics in House Dust - Final Report" Submitted to US EPA
Exposure Assessment Research Division. Research Triangle Park,
NC (June 29, 1990) .
47
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
8. Bannon DI, Murashchik C, Zapf CR, Farfel MR and Chisolm, JJ
Jr. 1994. A Graphite Furnace AAS Method of Blood Lead
Measurement Using Matrix Matched Standards. Clinical
Chemistry. 40:1730-1734.
9. Hornung RW, Reed LD. 1990. Estimation of Average Concentration
in the Presence of Nondetectable Values. Applied Occupational
& Environmental Hygienef 5(1):46-51.
10. S-PLUSฎ Version 3.2 for Windows, March 1993, "Read Me First"
StatSci, A Division of Microsoft, Inc., Seattle, Washington.
11. Laird Nan, Ware James. 1982. Random-Effects Model for
Longitudinal Data. Biometrics 38:963-974.
12. Zeger SL, Liang K-Y (1986). Longitudinal Data Analysis for
Discrete and Continuous Outcomes. Biometrics 42:121-130.
13. Zeger SL, Liang K-Y, Albert PS. 1988. Models for Longitudinal
Data: A Generalized Estimating Equation Approach. Biometrics
44:1049-1060.
14. Waternaux C, Laird N, Ware J. 1989. Methods for the Analysis
of Longitudinal Data: Blood Lead Concentrations and Cognitive
Development. J Amer Stat Assoc 84:33-41.
15. Liang K-Y, Zeger SL. 1986. Longitudinal Data Analysis Using
Generalized Linear Models. Biometrics 73:13-22.
16. Moulton LH, Zeger SL. 1989. Analyzing Repeated Measures on
Generalized Linear Models via the Bootstrap. Biometrics. 45:
381-394.
17. Royall RM. 1986. Model Robust Inference using Maximum
Likelihood Estimators. Internatl Stat Rev 54:221-226.
18. GEE: A SAS Macro for Longitudinal Data Analysis. Technical
Report #674 Department of Biostatistics, The Johns Hopkins
University.
19. Tukey JW. 1977. Exploratory Data Analysis. Addison-Wesley,
Reading, Massachusetts.
48
-------
Preliminary RSM study Report-
Partial 12 Month Data 7/95
20. Sayre JW, Katzel MD. 1979. Household surface lead dust: its
accumulation in vacant homes. Environ. Health Perspect.
29:179-182.
21. Farfel MR, Chisolra JJ Jr, Rohde CA. 1994. The longer-term
effectiveness of residential lead paint abatement. Environ.
Research 66:217-221.
22. Brody DJ, Pirkle JL, Kramer RA, et al. 1994. Blood lead levels
in the US population: Phase 1 of the third national health and
nutrition examination survey (NHANES III, 1988 to 1991). JAMA
272:277-283.
23. Swindell SL, Charney E, Brown MJ, Delaney J. 1994. Home
Abatement and Blood Lead Changes in Children with Class III
Lead Poisoning. Clin. Pediatrics Sept.1994: 536-541.
24. Rey-Alvarez S and Menke-Hargrave T. 1987. Deleading dilemma:
pitfall in the management of childhood lead poisoning.
Pediatrics 79:214-217.
25. Farfel MR and Chisolm JJ. 1990. Health and environmental
outcomes of traditional and modified practices for abatement
of residential lead-based paint. Am. J. Public Health 80:1240-
1245.
26. Lanphear B. et. al. The Relationship of Lead-Contaminated
House Dust and Blood-Lead Level Among Urban Children. Final
Report to the National Center for Lead-Safe Housing, 1994.
27. U.S. Centers for Disease Control. Preventing Lead Poisoning in
Young Children; A Statement by the Centers for Disease Control
- October 1991. USDHHS/PHS/CDC Atlanta, Georgia.
28. Barry, PSI. 1981. Concentrations of lead in tissues of
children. Br. J. Industrial Med. 316:1037-1043.
29. Rabinowitz MB, Kopple JD, Whetherhill GW. 1976. Kinetic
analysis of lead mobilization in healthy humans. J. Clin.
Invest. 58:260.
30. Chisolm JJ, Mellits ED, Quaskey. 1985. The relationship
between the level of lead absorption in children and the age,
type, and condition of housing. Environ. Research 38:31-45.
49
-------
Preliminary RCM Study Report-
Partial 12 Month Data 7/95
Table l: Data collection plan*
Study
Group
R&M I
R&M II
R&M III
Previously
Abated
Modern
Urban
Pre-R&M/
Initial
B
D
S
w
Q
B
D
S
W
Q
B
D
S
-
Q
B
D
S
W
Q
B
D
S
W
Q
Post-
R&M
-
D
S
-
"
-
D
S
-
"
-
D
S
W
"
n/a
n/a
2
Month
B
D
-
-
B
D
-
-
"
B
D
-
-
-
-
-
"
-
-
-
6
Month
B
D
S
w
Q
B
D
S
W
Q
B
D
S
W
Q
B
D
S
W
Q
B
D
S
W
Q
12
Month
B
D
-
-
Q
B
D
-
-
Q
B
D
-
-
Q
B
D
-
Q
B
D
-
-
Q
18
Month
B
D
S
W
Q
B
D
S
W
Q
B
D
S
W
Q
B
D
S
W
Q
B
D
S
W
Q
24
Month
B
D
_
_
Q
B
D
_
_
Q
B
D
_
_
Q
B
D
-
Q
B
D
-
-
Q
B=Blood; D=Dust; S=Soil; W=Water; Q=Questionnaire
50
-------
Preliminary RsM Study Report-
Partial 12 Month Data 7/95
Table 2: Types and numbers of 12-month campaign samples collected
and analyzed for lead (not including field QC)
SAMPLE TYPE
SBTTLKD DOST:
Perimeter Floor
Composite-- room*
with window.
Ferinater Floor
Composite room
without window*
Window Sill Ccopoaita
Window Wall ConpoBite
Interior Bntryway
Exterior Bntryway
Air Duct/
Upholstery
Total Dust par
Dwelling
TOTAL DOST
Soil Cora -
drip line
Drinking Mater*
Venous Blood -
JUKI 6 TO 60 MDl.
TOTAL
Planned
per House
2a
1
2
2
1
0
I9
9
0
0
1 per
child
10+
Collected in
89 Study
Bouse*
182b
45
177
174
89
0
53
35
755
0
0
109
864
Chemically
Analyzed
for Lead
182
45
177
174
89
0
53
35
755
0
0
109
864
Not
Collected
in the 89
Bouses
0
44C
ld
4e
0
Of
lh
50
Of
Of
51
55
* One composite sample was obtained per story.
" Includes 4 samples collected from basements used as living spaces in 3 RSM I
houses and 1 Modern Urban house.
c 44 houses did not have rooms without windows.
d Sills on one story were inaccessible in 1 RSM I house.
Wells on one story were inaccessible in 2 R&M I houses and on both stories in 1
RSM I house, for a total of 4 samples not collected, primarily due to plastic
being sealed over the windows in winter.
' This sample type was not part of the 12-month campaign.
9 Upholstery sample was collected if air duct sample could not be obtained.
h Air duct & upholstery were inaccessible/not present in 1 RSM II house.
1 3 Modern Urban children had moved at the time of blood collection, and 2 Modern
Urban children were missed by the Primary Care Physician.
51
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
Table 3: Types and numbers of 12-month campaign samples collected
by study group (not including field QC)
SAMPLE TYPE
SETTLED DUST:
Perimeter Floor
Composite room*
with window*
Perimeter Floor
Compositerooms
without window*
Window Sill
Composite
Window Nell
Composite
Interior Batryway
Exterior Entryway
Air Duct
Upholstery
TOTAL DOST
Soil Cora
- drip lino
Drinking Mater
Venou* Blood -
agea 6 to 60 no*.
Collected
in IS
Modern
Urban
Bouses
31
4
30
30
15
0
11
4
125
0
0
14
Collected
in 14
Previously
Abated
Houses
28
6
28
28
14
0
5
9
118
0
0
24
Collected
in 20
Rat i
House*
43
15
39
36
20
0
8
12
173
0
0
21
Collected
in 20
IUM II
Houses
40
11
40
40
20
0
14
5
170
0
0
26
Collected
in 20
RCM III
Bouses
40
9
40
40
20
0
15
5
169
0
0
24
52
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
Table 4: Frequency of family moves and reoccupancies between the
initial and the twelve month campaigns among the 60 R&M
houses included in this preliminary report and the
control houses.
STUDY GROUP
R&M I
(20 houses)
R&M II
(20 houses)
R&M III
(20 houses)
Modern Urban
(15 houses )b
Previously Abated
(15 houses )b
Total
Moved
No.
Families
6
3
4
1
1
15
No.
Children3
12
5
6
3
1
27
Replaced
No.
Families
6
3
4
1
0
14
No.
Children
1
2
6
0
0
9
' Includes children who moved although family did not.
b One extra house was included for a total of 16 houses. Of these one household
withdrew from the study before the 6 month campaign leaving 15 houses.
53
-------
Preliminary RIM Study Report-
Partial 12 Month Data 7/95
Table 5: Overall dust lead loadings* (f/g/ft2) for houses completing the 12-month campaign
Study
Group
(n)
Modern
Urban
(15)
Previously
Abated
(14)
R&M
Level I
(20)
R&M
Level II
(20)b
R&M
Level III
Pre-
Intervention/
Initial Campaign
GM
min max
90
10 540
890
120 5420
16200
1440 70700
23800
3100 124000
39300
3260 127000
Post-
Intervention/
R&M Campaign
GM
min max
N/A
N/A
1490
280 13100
300
3 4910
90
7 3760
Two Month
Campaign
GM
min max
N/A
N/A
3420
560 11200
1440
160 13560
220
40 1140
Six Month
Campaign
GM
min max
70
6 750
470
50 2970
4020
950 19100
1230
40 14000
180
10 1610
Twelve Month
Campaign
GM
min max
60
10 260
630
100 5300
4140
590 22600
840
40 24900
160
30 3050
' Based on weighted averages of floor, entryway, window sill, and window well samples within houses.
Eighteen houses were tested in the two month campaign.
54
-------
Preliminary RtM Study Report-
Partial 12 Month Data 7/95
Table 6: Overall dust lead concentrations* C^g/g) for houses completing the 12-month campaign
Study
Group
(n)
Modern
Urban
(15)
Previously
Abated
(14)
R&M
Level I
(20)
R&M
Level II
(20)b
R&M
Level III
(20)b
Pre-
Intervention/
Initial Campaign
GM
min max
260
90 510
2360
430 16400
20500
3300 106000
13900
1240 65000
16800
1440 64800
Post-
Intervention/
R&M Campaign
GM
min max
N/A
N/A
8300
730 66500
6000
480 38400
2800
830 47100
Two Month
Campaign
GM
min max
N/A
N/A
13400
3800 34900
6200
700 53300
1500
500 6700
Six Month
Campaign
GM
min max
300
100 830
1900
180 18000
12700
2400 50900
4700
500 45500
1400
360 12600
Twelve Month
Campaign
GM
min max
300
150 1500
2800
500 15600
12900
2300 54400
3400
300 36900
1100
200 16300
Based on weighted averages of floor, entryway, window sill, and window well samples within houses
Eighteen houses were tested in the two month campaign.
55
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
Table 7: Overall dust loadings* (mg/ft2) for houses completing the 12-month campaign
Study
Group
(n)
Modern
Urban
(15)
Previously
Abated
(14)
R&M
Level I
(20)
R&M
Level II
(20)b
R&M
Level III
Pre-
Intervention/
Initial Campaign
6M
min max
330
50 1690
380
160 1340
790
300 3360
1710
350 7920
2340
970 6570
Post-
Intervention/
R&M Campaign
GM
min max
N/A
N/A
180
50 1530
50
10 260
30
10 710
Two Month
Campaign
GM
min max
N/A
N/A
260
100 490
230
100 1090
140
70 640
Six Month
Campaign
GM
min max
240
60 2950
250
100 830
320
140 1330
260
50 1230
130
30 300
Twelve Month
Campaign
GM
rain max
200
90 680
220
50 620
320
100 1340
250
40 930
140
40 320
Based on weighted averages of floor, entryway, window sill, and window well samples within houses
Eighteen houses tested included in the two month campaign.
56
-------
Preliminary RcM Study Report-
Partial 12 Month Data 7/95
Table 8:
Descriptive Statistics for Blood Lead Concentrations (PbB //g/dL) for
Children with Initial PbB < 20 //g/dL
STUDY GROUP
R&M Level I
R&M Level II
R&M Level III
Previously
Abated*
Modern Urban
GM
range
(n)
GM
range
(n)
GM
range
(n)
GM
range
(n)
GM
range
(n)
CAMPAIGN
INITIAL
8
2 to 16
(24)
10
3 to 19
(20)
11
2 to 19
(19)
12
6 to 18
(18)
3
1 to 6
(18)
2 MONTH
8
1 to 16
(22)
10
4 to 20
(15)
11
3 to 25
(16)
N/A
N/A
6 MONTH
9
2 to 17
(19)
12
5 to 21
(16)
10
3 to 23
(18)
13
5 to 32
(17)
4
2 to 10
(16)
12 MONTH
8
2 to 20
(15)
10
5 to 18
(15)
10
6 to 16
(14)
12
7 to 21
(16)
3
2 to 6
(14)
* Excludes child requiring chelation therapy during follow-up.
57
-------
Preliminary RซM Study Report-
Partial 12 Month Data 1/95
Table 9: Twelve month campaign correlations of In (lead loading). In (lead concentration),
and In (dust loading) with In (blood ^g/dL) for the youngest child per household.
Pearson correlation coefficients / Prob > |R| under Ho: Rho=0 / Number of observations
DUST
VARIABLE
In (lead loading)
A/9/ft2
In (lead cone)
/*g/g
In (dust loading)
mg/ft2
Surface Type
r
P
n
r
P
n
r
P
n
Floor
.32
< .01
79
.45
< .01
79
.01
NS
79
Window
Sill
.22
.05
79
.22
.05
79
.13
NS
79
Window
Well
.22
.05
78
.29
.01
78
- .06
NS
78
Interior
Entryway
.15
NS
79
.31
< .01
79
- .003
NS
79
Floors - Rms
w/o Windows
.30
.05
41
.36
.02
41
.01
NS
41
Air
Duct
.31
.03
50
.37
< .01
50
.12
NS
50
Upholstery
.49
< .01
29
===^=^=
.61
< .01
29
.19
NS
29
NS = Non-significant p-value>.05
58
-------
Preliminary RSM Study Report-
Partial 12 Month Data 7/95
Table 10: Twelve month campaign correlations of In(lead loading), In(lead concentration)
and In (dust loading) with In (blood //g/dL) for all children.
Pearson correlation coefficients / Prob > |R| under Ho: Rho=0 / Number of observations
DUST
VARIABLE
In (lead loading)
MQ/ft2
In (lead cone)
M9/g
In (dust loading)
mg/ft2
r
P
n
i i
r
P
n
r
P
n
Surface Type
Floor
.34
< .01
109
.45
< .01
109
< .01
NS
109
====d
Window
Sill
.22
.02
109
.23
.02
109
.10
NS
109
Window
Well
.22
.02
107
.27
.01
107
- .04
NS
107
Interior
Entryway
.11
NS
109
.34
< .01
109
- .01
NS
109
Floors -Rms
w/o Windows
.36
.01
55
.37
< .01
55
.02
NS
55
Air
Duct
.39
< .01
62
.46
< .01
62
.14
NS
62
Upholstery
.30
.04
47
.50
< .01
47
.09
NS
47
NS = Non-significant p-value>.05
59
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 1 Dust Lead Loadings (ug/ftA2) across Campaigns for Floor Surfaces
RM-I
Q
.a
o,
Q
.0
Q-
o,
Previously-Abated
RM-II
RM-III
7 -
6 -
5
4
2
1 -
0 -
-1 -
r-r-i r^-i |
MUb g g
IN PI 2M 6M 12M
7
6
5
4
3
2
1 -
0
-1 -
^-fl h h n
ug u 2
IN PI 2M 6M 12M
7 -
6 -
5
4
3
2
1 -
0 -
-1
Vn JH Pi '
IN PI 2M 6M 12M
KENNEDY KREGER RES
Modern
7
6
5
4
3
2
1
0
-1 i
&
-&
O
H
H
C
H
m
IN
6M
12M
IN
6M
12M
data based on the 89 houses completing the twelve month campaiqn
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 2 Dust Lead Loadings (ug/ftA2) across Campaigns for Window Sill Surfaces
RM-I
RM-II
RM-III
6
.ฃ)
CL
0
0)
o
Q
_Q
CL
O
^j>
0
7
6 -
5
A -
*t
3 -
2 -
1 -
0 -
-1
T
_L -T-
i i i ' y T"1 j . ^
S^sb rh n -Fi
M U U r_p
1 1 *-^J
! ' '
^
IN PI 2M 6M 12M
Previously-Abated
7 -
6
5 -
4
3
2
1 -
0
-1
-r
Bn I3!
M g
"
7
6
5
4
3
2 ^
1
0
-1
IN PI 2M
6M
12M
5
4 -
3
2
1
0
Modern
IN
6M
12M
IN
-Q-
6M
7
6
5
4
3
2 -\
1
0
-1 -
m
m
o
rn
IN PI 2M
6M
12M
w
m
O
nvซ
CU.SJ
c
H
12M
data based on the 89 houses completing the twelve month campaign
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 3 Dust Lead Loadings (ug/ftA2) across Campaigns for Window Well Surfaces
o
o
7
6
5
4
3
2
1
0
-1
RM-I
Previously-Abated
a
RM-II
RM-III
7 -
6 -
5 J
og10(PbD)
M CO -ฃ>.
1 '
0 -
-1 -
PI
Bv - ^
NI r1-! [-1-] i i
a tj LJ
IP -
7 -
6 -
5 -
4 -
3 -
2 -
1 -
0
-1 -
-T-
ง T- T-
\T n h n
^ " !
7 -
6
5 -
4
3 -
2
1 -
0
-1
j=
s
\ . *
\* n n R
ps/^r M y
-^J
_
m
z:
z
ru
3
3?
JS
rn
o
IN PI 2M 6M 12M IN PI 2M 6M 12M IN PI 2M 6M 12M
^J
ni
m
Modern
5
4
3
2
1
-1
B-
-e
-B
o
m
IN
6M
12M
IN
6M
12M
data based on the 89 houses completing the twelve month campaign
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 4 Dust Lead Loadings (ug/ftA2) across Campaigns for Interior Entryway Surfaces
RM-I
RM-II
RM-III
O)
o
0
.a
CD
O
7
6
5
4
3
1
0
-1
7
6
5
4
2
1
0
-1
B-e-f
IN PI 2M 6M
12M
Previously-Abated
7
6
5
4
3
2 \
1
0
-1 -I
7
e ^
5
4
3
2
1
0 -
-1 -
7 -
6
5
4
3
2
1
0
-1 ^
IN PI 2M
6M
12M
Modern
9-
-B
IN PI 2M
6M
12M
K
rn
a
-<
rn
Q
m
po
20
m
c#
g
73
O
T
z
w
H
H
C
H
m
IN
6M
12M
IN
6M
12M
data based on the 89 houses completing the twelve month campaign
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 5 Dust Lead Concentrations (ug/g) across Campaigns for Floor Surfaces
RM-I
RM-II
RM-III
6
5
9 4
Q
{ฃ 3
0
g, 2 -
1 -
0 -
~ n ^
B ^~-fฐ~i fl r~i
^ .L 2 H hd
IN PI 2M 6M 12M
6 -
5 -
4 -
3 -
2 -
1 -
0 -
m
5n -f- T ;
U tl -pi n -"
" J^ : M ^P
IN PI 2M 6M 12M
6 -
5 -
4 J
3
2
1 -
0
^i-
n^T T :
: fj^CL 5 "^
: Sr M Q
IN PI 2M 6M 12M
^
m
z.
z
m
^
"v
ง
m
0
m
73
70
m
C0
5
5
Previously-Abated Modern O
X
6
5
9 4
Q
n
t 3 -
o
D) 2 -
1 -
0 -
*
i i *
r-i .-i-.
S
ซ_L_i * '
6
5 -
4 -
3
2 -
1 -
0 -
-r
PI n rn
~ rn F I
Z
Cfl
-^
H
c
H
m
IN
6M
12M
IN
6M
12M
data based on the 89 houses completing the twelve month campaign
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 6 Dust Lead Concentrations (ug/g) across Campaigns for Window Sill Surfaces
RM-I
RM-II
RM-III
o
Q
a
o
en
a
O
Q
.0
O
O)
0
6
5
4 -
3 -
2 -
1 -
0 -
6
5 -
4 -
3
2 -
1 -
0 -
F^ Fh~PI n n
_J_ LJ [J ^J LJ
-j-j j^ '-^
IN PI 2M 6M 12M
Previously-Abated
PI n 5
tpr LJ M
^L.
6
5
4
3
2 H
1
0 H
S
IN PI 2M
6M
12M
6
5
4
3
2 -I
1
o H
Modern
IN
6M
12M
IN
-B-
6M
-a
12M
6
5
4
3 -
2 -
1
0
IN PI 2M
6M
12M
m
m
o
m
Q
m
5
^0
O
X
c
m
data based on the 89 houses completing the twelve month campaign
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 7 Dust Lead Concentrations (ug/g) across Campaigns for Window Well Surfaces
RM-I
IN PI 2M
6M
12M
Previously-Abated
IN
6M
12M
RM-II
IN PI 2M
6M
12M
IN
Modern
0
Q
Q.
O)
0
6 -
5 -
4 -
3
2
1 -
0 -
i rh ri
n n - Hi
u u
L-^J
6 -
5
4 -
3
2
1 -
0 -
ฃ
^ j^
6M
12M
RM-III
6 -
5 -
9 4
Q
t 3 -
o
| 2 -
1 -
0 -
&Pl n rh n
U U u3~~ u
6 J
5
4
3
2
1 -
0 -
M-a-S T
1
6 -
5 -
4 -
3 -
2
1 -
0 -
N-D D S
I I i-J LJ
KENNEDY KRIEGI
IN PI 2M
6M
1?M
m
3)
o
c
m
data based on the 89 houses completing the twelve month campaign
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 8 Dust Lead Concentrations (ug/g) across Campaigns for Interior Entryway Surfaces
RM-I
RM-II
RM-III
6
5 -
4 -
-Q
cT
*a> 2 -
1 -
0 1
*
EHIH3 Q Q
*
*
6
5 -
4 -
3
2 -
1 -
0 -
CrQ^fu J5
. UJ I 1
' -" IH. , .
J^
6
5 -
4
3 -
2
1 -
0
.
ri T "
rn
z
m
D
^
2
m
o
IN PI 2M 6M 12M IN PI 2M 6M 12M IN PI 2M 6M 12M ^
PI
Previously-Abated Modern ^J
o
6
5
9 4 -
J3
n 3 -
^> 2 -
o
1 -
0 -
i-^-i
: Un~~ -M
- ' *
6
5
4 -
3
2
1 -
0 -
*
g^ ง ~^^~^
,
X
nra
z
H
H
m
IN
6M
12M
IN
6M
12M
data based on the 89 houses completing the twelve month campaign
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 9 Dust Loadings (mg/ftA2) across Campaigns for Floor Surfaces
RM-I
RM-II
RM-III
o
2
1 -
Q
o 0 -
D)
-1 -
-2 -
-3 -
3 -
2
1 '
Q
o 0 -
O)
-1 -
-2
-
Eh-T| P R 5
IN PI 2M 6M 12M
Previously-Abated
ง n a
'-p-J I I UJ
o
2 -
1 -
0 -
-1 -
-2 -
-3 -
3 -
2
1 -
0 -
-1 -
-2
_
^\T " 5
. b^ง &~~ S
1
IN PI 2M 6M 12M
Modern
*
J-
o
2
1 -
0
-1
-2
-3 -
" Vx& S ~ 13
,_^
IN PI 2M 6M 12M
<*^
rn
z
mm
2,
m
Q
m
Q
m
71
m
m
>
0
S
z
c
m
IN
6M
12M
IN
6M
12M
data based on the 89 houses completing the twelve month campaign
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 10 Dust Loadings (mg/ftA2) across Campaigns for Window Sill Surfaces
RM-I
RM-II
RM-III
3
2
^ 1 -
Q
o 0
en
-1 -
-2 -
-3 -
:\rr jru -R Pi
,-ij b^-f trJ i i
IN PI 2M 6M 12M
3
2
1 -
0
-1 -
-2 -
-3
^N R rn rh
V-T J&- td tp
IN PI 2M 6M 12M
3 -
2 -
1 -
0 -
-1 -
-2
-3
w n n n
Yl^L U V
IN PI 2M 6M 12M
KENNEDY KRSEGER RES
Previously-Abated
Modern
^
Q
o
o
3 -
2 -
1 -
0
-1
-2
-3 -
,
n n a
u U tj
1 *
3
2
1
0
-1 -
-2
-3
^~ -S i
* I
1 1
O
f.v.fl
m-fl
m
IN
6M
12M
IN
6M
12M
data based on the 89 houses completing the twelve month campaign
-------
Q
o
cr>
o
3
2
1
0
-1
2
-3 -
R&M Study: Preliminary 12 Month Report (7/95)
Figure 11 Dust Loadings (mg/ftA2) across Campaigns for Window Well Surfaces
RM-I
-9-
IN PI 2M
6M
12M
3 ^
2
1
0 i
-1
-2 -
-3 -
RM-II
-B-
IN PI 2M
6M
12M
3
2 ^
1
0
-1
-2
-3
RM-III
IN PI 2M
6M
12M
m
%.
z:
ni
m
Q
m
m
Q
o
^i
o
3
2
1
0
-1
-2
-3 -
Previously-Abated
3
2
1
0
-2 -
-3 -
IN
6M
12M
IN
Modern
o
H
m
6M
12M
data based on the 89 houses completing the twelve month campaign
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 12 Dust Loadings (mg/ftA2) across Campaigns for Interior Entryway Surfaces
RM-I
RM-II
RM-III
O)
o
O)
o
3 -
2
1 -
0 -
-1 -
-2 -
-3 -
3
2 -
1 -
0 -
-1 -
-2 -
-3 -
9
N
Ji H .a R
Lj-tr~" XT M
: -^ ^
3 -
2 -
1 -
0 -
-1 -
-2 -
-3 -
,_.
: p_
rk^ h fl Q
; fJ-'-V U
uL
3
2 -
1 -
0 -
-1 -
-2
-3
_
ri "^
Mv ^__ ^
B^i ~~^
ฑ^ ^
'
m
MBBi
m
o
Q
PI 2M 6M 12M IN PI 2M 6M 12M IN PI 2M 6M 12M HI
20
0)
ง
Previously-Abated Modern ^
O
X
-Q o
^_, ;
l_i_,
3 -
2 -
1 -
0 -
-1 -
-2 -
-3 -
^^-, *
: g
y~ ~&~ ~^
=,
H
m
IN
6M
12M
IN
6M
12M
data based on the 89 houses completing the twelve month campaign
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 13 Blood Lead Concentrations for Children with Initial Blood Pb < 20 ug/dL
RM-I
RM-II
RM-III
2.0
1.5-
CD
1 1.0-
O)
0
0.5-
o.o-
55-
iiL
*
*
*
IN 2M
-R- i
ฑ B
* 1
2.0-
1.5-
1.0-
0.5-
0.0-
6M 12M
Previously
2.0
1.5-
m
JD
| 1.0-
O)
o
0.5-
0.0-
^~
_L
E3-B-& e
.
IN 2M 6M 12M
2.0-
1.5
1.0-
0.5-
0.0
BB n n
^. V LJ LJ
ซ-!-
IN 2M 6M 12M
m
;z
ni
71
ni
Q
rn
m
w
m
-Abated
2.0
1.5
1.0-
0.5-
0.0-
Modern Urban
. .
S" ง S
>
ya
o
z
d
c
m
IN
6M
12M
IN
6M
12M
data based on the 89 houses completing the twelve month campaign
-------
Figure 14: Dust Lead Loadings
R & M Study - Partial 12 Month Campaign
100,000
Geo. Mean 0"g/ft2)
Ifit Eniry Floors w/windows Floors w/o windows Sills
Wells
Air Due!
Upholstery
Surface Type
MU PA R&M I R&M II R&M III
m
3
7;
3D
m
m
I
2
H
I
7/95
-------
Figure 15: Dust Lead Concentrations
R & M Study - Partial 12 Month Campaign
100,000
10,000
1,000
100
10
Geo. Mean (/ug/g)
Int Entry Floors w/windows Floors w/o windows Sills Wells Air Duct Upholstery
Surface Type
MU I PA IR&MI R&M II IR&MIII
I
m
5
m
c
MM
m
7/95
-------
Figure 16: Dust Loadings
R & M Study - Partial 12 Month Campaign
100,000
10,000
1,000
100
10
Geo. mean (mg/ft2)
Int Entry Floors w/windows Floors w/o windows Sills Wells Air Duct Upholstery
Surface Type
MU I PAH R&M I R&M II I R&M III
m
5
5
m
30
3D
m
I
mm
I
7/95
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 17 Dust Lead Loadings (ug/ftA2) across Groups at the 12 Month Campaign
Floor
Window Sill
Window Well
7
6
5 -
logld(PbD)
ro co .u
1 -
0 -
-1 -
B a ph D _
1 1 *~l
B
7 -
6
5 -
4
3
2
1 -
0
-1 -
*
Q P ^ ^
; ^
7
6
5
4
3
2
1 -
0
-1 -
. . T^
R . _
2 Q _
M H S p
y-
m
z
z
m
a
F
m
Q
RMI RMII RMIII PA MU RMI RMII RMIII PA MU RMI RMII RMIII PA MU ^
m
C/>
Interior Entryway
Floors w/o windows
7 -
6 -
5 -
4
3
2 -
1 -
0
-1 -
^T^ *
i B B B s
-;-
7 -
6 -
5 -
4 -
3 -
2
H
1 '
o -
-1 -
, -^
f=l rn ^ D
B a
. : . p
5
I
C
rn
RMI RMII RMIII PA MU
RMI RMII RMIII PA MU
data based on the 89 houses completing the twelve month campaign
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 18 Dust Lead Concentrations (ug/g) across Groups at the 12 Month Campaign
Floor
Window Sill
Window Well
9
Q
.0
f
o
Q
n
6
5
4 -
3 -
2 -
1 -
0 -
i i 5 i
ง
6
5
4
3
2 -
1 -
0
*
a 3
-0555
^^ r~l
6
5 -
4
3
2
1 -
0 -
T r~:~1
_^ rb Pi
d g Q
m
z
m
3
7;
^
Q
RMI RMII RMIII PA MU RMI RMII RMIII PA MU RMI RMII RMIII PA MU ITS
7ฎ
m
v*
Interior Entryway Floors w/o windows 20
6
5 -
4
3
2
1 -
0 -
.
r ^-> .-^-,
Q i ฃ Q ^
" Q
6 -
5 -
4
3
2
1 -
0 -
*
m
._
m
H
ซv|
c
H
m
RMI RMII RMIII PA MU
RMI RMII RMIII PA MU
data based on the 89 houses completing the twelve month campaiqn
-------
R&M Study: Preliminary 12 Month Report (7/95)
Figure 19 Dust Loadings (mg/ftA2) across Groups at the 12 Month Campaign
Floor
Window Sill
Window Well
3
2
IT 1 '
Q
o 0
CT)
-1
-2 -
-3 -
*
3 a a 9 |
3 -
2
1 -
0 ~
-1 -
-2 -
-3 -
a
RMI RMII RMIII PA MU RMI
Interior Entryway
3
2
1 .
Q
0 0
cn
-1
-2
-3
, ,
i , ^_,
.J-. : : : ,
ง 5 i P a
! *
>-i-'
3
2
1 -
0
-1
-2
-3 -
' '
^
ง ง B 5
l-i--
3 -
2
1 -
0
-1
-2
-3 -
*
- ? Q i I
5^
m
z
m
^
ffi
STJ
RMII RMIII PA MU RMI RMII RMIII PA MU 94jj
*>?#
m
CO
s
P"
Floors w/o windows Q
B g 5 I
Z
(ft
swvjj
RMI RMII RMIII PA MU
RMI RMII RMIII PA MU
data based on the 89 houses completinq the twelve month campaign
-------
Figure 20: Overall lead levels and dust loadings by group*
R & M Study - Partial 12-Month Campaign
Geo. Mean
100,000
10,000
1,000
100
m
5
5
m
33
X
m
Lead Concentration
Lead Loading
(/jg/ft" 2)
Dust Loading
(mg/ft~2)
Modern Urban 11 Previously Abated R&M I
R&M
R&M III
7/95
* Summary measures are based on weighted averages of
floors, entryways, window sills, and window wells within houses
-------
Figure 21: Overall dust lead loadings by group
R & M Study - Partial 12 Month Campaign
100,000
10,000
1,000
Geo. Mean (jug/ft2)
100
Initial
Post-IN
2 month
6 month
12 month
Modern Urban Previously Abated R&M I R&M II R&M III
7/95
* Summary measures are based on weighted averages of
floors, entryways, window sills, and window wells within houses
i
m
s
33
I
m
ao
33
m
c
FT
-------
Figure 22: Overall dust lead concentrations by group*
R & M Study - Partial 12 Month Campaign
Geo. Mean
100,000
10,000
1,000
100
Initial
Post-IN
2 month
6 month
12 month
Modern Urban Previously Abated R&M I R&M II R&M
7/95
* Summary measures are based on weighted averages of
floors, entryways, window sills, and window wells within houses
i
I
*
35
m
o
m
m
(ft
-------
Figure 23: Overall dust loadings by group
R & M Study - Partial 12 Month Campaign
10,000
1,000
100
Geo. Mean (mg/ft2)
Initial
Post-IN
2 month
6 month
12 month
Modern Urban Previously Abated R&M I R&M II R&M
7/95
* Summary measures are based on weighted averages of
floors, entryways, window sills, and window wells within houses
m
3
*
35
m
6
m
30
as
c
ft
-------
Figure 24:
40
R &M Study- Preliminary 12 Month Report (7/95)
Children's Blood Lead Levels Across Time - R &M I
for children with 12-month values
35-
30-
D>
g
8 20
I
^ 15-
CQ
Dec-1992
Jun-1993
Dec-1993 Jun-1994
Blood Collection Date
Dec-1994
Jun-1995
-------
Figure 25:
40
35-
30-
25-
8 20
<3
|
^ 15~
o
_o
CD
10-
5_
Dec-1992
R & M Study - Preliminary 12 Month R eport (7/95)
Children's Blood Lead Levels Across Time - R&M I
for children with 12-month values
Jun-1993
Dec-1993 Jun-1994
Blood Collection Date
m
1
s
i
s
m
o
z
C0
nl
Dec-1994
Jun-1995
-------
Figure 26:
R &M Study- Preliminary 12 Month Report (7/95)
Children's Blood Lead Levels Across Time - R&M II
for children with 12-month values
Dec-1992 Jun-1993 Dec-1993 Jun-1994
Blood Collection Date
The bloodleadvaluelhatappearsas40isactjallya value of43ug/dL.
Dec-1994
Jun-1995
-------
40
35_
30-
20-
8
1
0)
-1 15-
8
CD
10-
5_
Dec-1992
Figure 27: R &M Study- Preliminary 12 Month Report (7/95)
Children's Blood Lead Levels Across Time - Modern Urban
for children with 12-month values
T~
Jun-1993
T
T
m
m
m
o
m
50
o
^r^
|
m
Dec-1993 Jun-1994
Blood Collection Date
Dec-1994
Jun-1995
-------
40
35-
30-
125
o
8 20
8
_g
en
10-1
5-
Dec-1992
Figure 28: R &M Study- Preliminary 12 Month Report (7/95)
Children's Blood Lead Levels Across Time - Previously Abated
for children vvith 12-month values
Jun-1993
Dec-1993
m
Jun-1994
Blood Collection Date
"The child whh blood lead value of 53ug/dL is excluded from analysis.
Dec-1994
Jun-1995
-------
Preliminary R&M Study Report
Partial 12 Month Data 7/95
APPENDIX A
Descriptive Statistics
-------
Table di Pearson correlation coefficients between log transformed dust lead loadings (*/g/ft') at the twelve month campaign
Floors
Window Sills
Window Wells
Interior
Entryway
Floors - Rms
w/o Windows
Air Ducts
Window Sills
.49"
(89)
"
"
Window Wells
.49"
(88)
.73"
(88)
~
*
Interior
Entryway
.38"
(89)
.39"
(89)
.40"
(88)
Floors -
Rms w/o Windows
.46"
(45)
.27
(45)
.17
(44)
.29'
(45)
-
Air
Ducts
.16
(53)
.25
(53)
.26
(52)
.15
(53)
-.11
(28)
Upholstery
.28
(35)
.04
(35)
.07
(35)
.16
(35)
.17
(17)
p < .05; * p < .01
Air Duct or Upholstery Samples were collected, not both.
Page A-4
-------
Table ei
Pearson correlation coefficients between log transformed dust lead concentrations at the twelve month campaign
Floors
Window Sills
Window Hells
Interior
Entryway
Floors - Rms
w/o Windows
Air
Ducts
Window
Sills
.51"
(89)
-
~
Window Wells
.54"
(88)
.67"
(88)
~
-
~
Interior
Entryway
.54"
(89)
.47"
(89)
.51"
(88)
-
Floors -
Rms w/o
Windows
.55"
(45)
.46"
(45)
.30
(44)
.48"
(45)
-
-
Air
Ducts
.47"
(53)
.42"
(53)
.33
(52)
.34
(53)
.37
(28)
-
Upholstery
.30
(35)
.30
(35)
.31
(35)
. 38
(35)
.43
(17)
^ซ
p < .05; ** p < .01
Air Duct or Upholstery Samples were collected, not both.
Page A-5
-------
Table f: Pearson correlation coefficients between log transformed dust loadings (mg/ftj) at the twelve month campaign
Floors
Window Sills
Window Wells
Interior Entryway
Floors - Rms w/o
Windows
Air
Ducts
Window Sills
.08
(89)
-
Window
Wells
.14
(88)
.66"
(88)
"
-
Interior
Entryway
.08
(89)
.56"
(89)
29
(88)
"
-
Floors -
Rms w/o Windows
.27
(45)
.04
(45)
.18
(44)
- .01
(45)
-
-
Air
Ducts
- .10
(53)
.05
(53)
- 07
(52)
-.04
(53)
- .35
(28)
-
Upholstery
.04
(35)
- .14
(35)
- .23
(35)
- .12
(35)
- .15
(17)
_a
p < .05; ** p < .01
Air Duct or Upholstery Samples were collected, not both
Page A-6
-------
Table gi Descriptive statistics on soil lead concentrations O/g/g) from the preliminary six month campaign report
Study
Group
Modern Urban
Prev'ly Abated
RM-I
RM-II
13
3
10
9
35
Minimum
34
304
182
428
Maximum
229
7,845
4,530
2,608
Geometric
Mean
73
1,521
730
708
std(ln)
0.489
1.625
1.000
0.599
Table hi
Descriptive statistics on water lead concentrations /g/L) from the preliminary six month campaign report
Study
Group
Modern Urban
Prev'ly Abated
RM-I
RM-II
RM-III
15
14
17
19
17
82
Minimum
0
0
0
0
0
Maximum
40
32
11
18
30
Geometric
Mean
4
1
2
3
3
stddnl
1.316
1.448
0.964
1.047
1.161
Upper
95% CI
8
3
4
4
5
Page A-7
-------
Preliminary R&M study Report
Partial 12 Month Data 7/95
APPENDIX B
Longitudinal data analysis output for dust lead
-------
DUST LEAD LOADINGS WITHIN COMPLETE R_M HOUSES AS AN OUTCOME OF ENVIRONMENTAL FACTORS: 07/17/95
12-month campaign all data model: Tactorl = level + season + eampaign(level) + randomjiouse
The MIXED Procedure
Class Level Information
Class Levels Values
DID 60 310 316 343 347 348 353 355
357 358 359 366 372 373 381
386 403 406 409 418 419 302
312 335 338 345 350 354 365
370 374 376 377 384 385 391
397 407 412 416 420 306 309
317 320 323 325 326 328 332
336 383 389 390 402 405 411
413 415 417 404
LEVEL 3321
CAMPAIGN 5 00 02 06 12 PI
UST3.LIS (source) Page B-1
-------
DUST LEAD LOADINGS WITHIN COMPLETE R_H HOUSES AS AN OUTCOME OF ENVIRONMENTAL FACTORS: 07/17/95
12-month campaign all data model: Tactor1 = level + season + campaign(level) + random house
The NIXED Procedure
Covariance Parameter Estimates (REML)
Cov Parm
Ratio
Estimate
Std Error
Z Pr > !Zi
DID
Residual
0.52166498
1.00000000
0.09194687
0.17625655
0.02427637
0.01687193
3.79
10.45
0.0002
0.0000
The MIXED Procedure
Model Fitting Information for FACTOR1
Description
value
Observations 293.0000
Variance Estimate 0.1763
Standard Deviation Estimate 0.4198
REML Log Likelihood -215.239
Akaike's Information Criterion -217.239
Schwartz's Bayesian Criterion -220.855
-2 REML Log Likelihood 430.4771
Null Model LRT Chi -Square 44.3687
Null Model LRT OF 1.0000
Null Model LRT P-Value 0.0000
The MIXED Procedure
Solution for Fixed Effects
Parameter
INTERCEPT
SPRING
SUMMER
FALL
LEVEL 3
LEVEL 2
LEVEL 1
CAMPAIGN(LEVEL)
CAMPAIGN(LEVEL)
CAHPAIGN(LEVEL)
CAMPAIGN(LEVEL)
CAHPAIGN( LEVEL)
CAMPAIGN(LEVEL)
CAMPAIGN(LEVEL)
CAHPAIGN(LEVEL)
CAMPAIGN(LEVEL)
CAHPAIGN(LEVEL)
CAMPAIGN(LEVEL)
CAMPAIGN(LEVEL)
CAMPAIGN(LEVEL)
CAMPAIGN(LEVEL)
CAMPAIGN LEVEL)
00
02
06
12
PI
00
02
06
12
PI
00
02
06
12
PI
3
3
3
3
3
2
2
2
2
2
Estimate
0.21071261
-0.05365035
-0.15434789
-0.00393296
-1.57870478
-0.79301191
0.00000000
3.05571972
0.70179258
0.57870463
0.47188651
0.00000000
2.03467544
0.60115318
0.41829973
0.42521529
0.00000000
0.93998960
0.06323570
0.17211068
0.14439255
0.00000000
Std Error
0.
0.
0.
0.
0.
0.
0.
0.
0.
12117595
07966301
07064688
08553026
16707386
16524626
B
13688774
13823696
13526552
0.13338638
0.
0.
0.
0.
0.
0.
0.
0.
13359146
13780818
13331476
13336752
13534171
13992696
13372528
13503716
.
DDF
218
218
218
218
218
218
m
218
218
218
218
218
218
218
218
218
218
218
218
.
T
1.74
-0.67
-2.18
-0.05
-9.45
-4.80
<
22.32
5.08
4.28
3.54
15.23
4.36
3.14
3.19
6.95
0.45
1.29
1.07
.
Pr > !T!
0.0835
0.5014
0.0300
0.9634
0.0000
0.0000
.
0.0000
0.0000
0.0000
0.0005
0.0000
0.0000
0.0019
0.0016
0.0000
0.6518
0.1994
0.2861
.
Aloha
0.05
0.05
0.05
0.05
0.05
0.05
_
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
.
Lower
-0.0281
-0.2107
-0.2936
-0.1725
-1.9080
-1.1187
2.7859
0.4293
0.3121
0.2090
1.7714
0.3295
0.1555
0.1624
0.6732
-0.2125
-0.0914
-0.1218
m
Itooer
0.4495
0.1034
-0.0151
0.1646
-1.2494
-0.4673
3.3255
0.9742
0.8453
0.7348
2.2980
0.8728
0.6811
0.6881
.
1.2067
0.3390
0.4357
0.4105
Source
The MIXED Procedure
Tests of Fixed Effects
NDF DDF Type IMF Pr > F
SPRING
SUMMER
FALL
LEVEL
CAMPAIGN(LEVEL)
1
1
1
2
12
218
218
218
218
218
0.45
4.77
0.00
30.48
78.00
0.5014
0.0300
0.9634
0.0000
0.0000
JST3.LIS (source)
Page B-2