EXPOSURE INVESTIGATION REPORT
SOLUTIA INCORPORATED/MONSANTO COMPANY
ANNISTON, CALHOUN COUNTY, ALABAMA
The Monsanto company produced polychlorinated biphenyls (PCBs) at a plant in
Anniston, Alabama, from 1935 to the 1970s. Hazardous wastes, including PCB still
bottoms, were disposed in two unlined landfill areas located adjacent to the
production facility. Investigations conducted by Monsanto (now Solutia, Inc.)
under a Consent Order with the Alabama Department of Environmental Management
(1985) documented the presence of PCB contamination in sediment samples from
off-site drainage ditches and in soil samples from private residences east and
north of the facility. These findings led to the remediation of off-site contaminated
areas and property buyouts for some homeowners.
Recent investigations have detected elevated blood levels of PCBs in some residents
of the community surrounding the Solutia facility and other neighborhoods in
Anniston . The source and exposure
pathways by which residents have been exposed to PCBs have not been defined.
Furthermore, it is uncertain whether significant exposures are still occurring.
To address these questions, the Agency for Toxic Substances and Disease Registry
(ATSDR) conducted this Exposure Investigation (EI).
Prior to conducting the EI, staff from ATSDR and the Alabama Department of
Public Health (ADPH) met with community representatives to explain the EI and
solicit their input. In March 2000, ATSDR met with families who lived within
a 1/2-mile radius of the site and invited them to participate in the EI. In
order to be eligible for the investigation, at least one family member had to
be a child between 1 and 7 years old. ATSDR staff and representatives of the
community group, CAP, went door-to-door in the designated neighborhoods to invite
eligible families to participate. A total of 18 families fully participated
in the EI. Environmental samples were collected from these 18 homes, and biological
samples were collected from 78 residents of these homes. In addition, environmental
samples, only, were obtained from one home; and biological samples, only, were
obtained from two people who lived in the designated area.
ATSDR staff gave the participants an appointment to come to the Calhoun County
Health Department where blood samples were collected. A few of the participants
were unable to come to the collection center. ATSDR staff traveled to the homes
of these individuals to collect the blood samples.
A licensed phlebotomist collected a 7-ml blood sample from each participant
using a Vacutainer tube with no anticoagulant. After collection, the blood samples
were allowed to clot for 2 hours at room temperature. The tubes were then placed
on ice until they were delivered to the laboratory for analysis. Blood collection
supplies and laboratory analyses were provided by the National Center for Environmental
Health laboratory at the Centers for Disease Control in Atlanta, Georgia.
Blood serum samples were analyzed for PCB congeners using gas chromatography/mass
spectroscopy (GC/MS). Results were reported as concentrations of individual
PCB congeners per unit volume of blood serum, and also as lipid-based concentrations.
Individual congeners were added together to yield total PCB concentrations.
ATSDR staff used a metal trowel to collect a composite surface soil sample
(0-3 inches in depth) from an area in the yard that the parent or child identified
as a frequent play area. In addition, ATSDR staff used a Nilfisk HEPA vacuum
cleaner to collect an indoor house dust sample from a room frequented by family
members; this was typically the living room of the house. The environmental
samples were shipped by overnight mail to the Midwest Research Institute in
Kansas City, Missouri, for analyses of PCB congeners and total PCBs. The house
dust samples were strained through a wire mesh screen to remove food particles,
fibers, and other debris prior to analyses. Soil and house dust samples were
analyzed for PCB congeners using gas chromatography/mass spectroscopy (GC/MS).
Prior to testing, each adult and a parent or legal guardian of each minor participant
was required to sign an informed consent/assent form. A separate informed consent
form for environmental testing was also obtained for each house prior to testing.
Descriptive statistics were used to characterize the blood serum and environmental
media PCB data. To assess correlations between biological and environmental
data, Spearman rank correlation coefficients were calculated. The correlation
analyses were conducted using the lipid-based concentrations of blood serum
To interpret the complex PCB congener profiles in the EI participants, ATSDR
employed principal components analyses (PCA). The goal of PCA is to linearly
transform possibly correlated variables (PCB congeners) into a smaller number
of uncorrelated variables called principal components (PC). The PC scores are
then plotted to graphically reveal patterns in a complex data set that are not
readily seen by visual inspection.
Principal components analyses (PCA) and cluster analyses were carried out using
SAS V.8 statistical software . Blood
PCB data (lipid-based) were normalized by expressing the concentration of the
individual PCB congeners in a sample as a percentage of the combined sum of
all PCB congeners. The factor loadings and scores were rotated according to
the Varimax rotation to assist in the interpretation of the factors .
The cluster analysis was performed using squared Euclidean distances with an
average linkage .
The concentration of PCBs was determined in blood serum samples from 37 children
(16 years old or less) and 43 adults. In adults, the blood PCB concentrations
ranged from non-detected to 210 parts per billion (ppb). The mean concentration
in adults was 14.2 ppb, and the median concentration was 2.2 ppb. In children,
the blood PCB levels ranged from non-detected to 4.6 ppb. The mean concentration
in children was 0.37 ppb, and the median concentration was non-detected.
Among the adults, five people had a blood PCB concentration in excess of 20
ppb; the PCB levels in these five people were: 22, 54, 93, 97, and 210 ppb.
These five high values skewed the arithmetic mean of the adult population. If
the five individuals with elevated PCB levels are not included in the mean,
the average PCB concentration in the rest of the population of 38 adults was
Blood serum concentrations of PCBs were also calculated as a blood lipid concentrations.
For adults, the mean blood PCB concentration was 2,537 ng/g lipid , and the
median concentration was 392 ng/g lipid. Previous studies have shown that blood
concentrations of PCBs can be influenced by the intake of dietary fat .
Therefore, blood PCB concentrations are best expressed as a blood lipid concentration
so comparisons can be made between individuals and with other studies. However,
to date, few published studies have reported PCBs as blood lipid concentrations.
In this EI, blood PCB concentrations are reported as µg/L (ppb) and as ng/g lipid.
The concentration of PCBs detected in composite surface soil samples from 19
homes ranged from non-detected to 11.7 ppm. The mean concentration of PCBs in
soil was 1.4 ppm, and the median concentration was 0.60 ppm.
The concentration of PCBs detected in house dust samples from 18 homes ranged
from non-detected to 10.3 ppm. The mean concentration of PCBs in house dust
was 0.81 ppm, and the median concentration was 0.11 ppm.
Indoor surface loading concentrations of PCBs in house dust ranged from non-detected to 2,960 ng/ft2. The mean surface loading concentration was 192 ng/ft2, and the median concentration was 6.6 ng/ft2.