Predictors of organochlorine pesticide plasma concentrations among reproductive-aged Black women

Abstract

Background:

Organochlorine pesticides (OCPs) are lipophilic persistent organic pollutants associated with adverse health outcomes. Black women have higher body burdens compared to other U.S. populations and the research on their predictors is limited.

Methods:

Using baseline data from a prospective cohort study of Black women aged 23–35 years from the Detroit, Michigan metropolitan area (enrolled 2010–2012), we examined predictors of plasma concentrations of the following OCPs: dichlorodiphenyltrichloroethane (p,p’-DDE), hexachlorobenzene (HCB), oxychlordane, and trans-nonachlor. At enrollment, we collected non-fasting blood samples from 742 participants. We also collected demographic, behavioral, dietary, occupational, and medical history data via self-administered questionnaires, telephone interviews, and in-person clinic visits. We fit linear regression models to calculate percent (%) differences across categories of each predictor and 95% confidence intervals (CIs).

Results:

In models adjusted for all other predictors, a 5-year increase in age was associated with 24% higher oxychlordane (95% CI: 12%, 38%) and 26% higher trans-nonachlor (95% CI: 12%, 42%). Heavy alcohol use was associated with 7–9% higher plasma concentrations of p,p’-DDE, oxychlordane, and trans-nonachlor. Current smoking was associated with 10–19% higher plasma concentrations of all four OCPs. Compared with never having been breastfed during infancy, having been breastfed for >3 months was associated with 15% higher concentrations of p,p’-DDE (95% CI: 6%, 25%), 14% higher oxychlordane (95% CI=5%, 24%), and 15% higher trans-nonachlor (95% CI=5%, 27%). Consumption of ≥5 glasses of tap or bottled water vs. <2 glasses/day of tap or bottled water was associated with 8–15% higher plasma concentrations of all four OCPs, and highest for trans-nonachlor (95% CI: 6%, 26%). No other dietary predictors were appreciably associated with plasma OCP concentrations. Obesity, parity, higher birth order, and longer lactation durations were inversely associated with plasma OCP concentrations.

Conclusions:

In Black U.S. women of reproductive age, exposure to OCPs earlier in life when these persistent pesticides were at higher concentrations in the environment appears to still contribute to current blood concentrations. In addition, residual contamination of tobacco, alcohol, and drinking water may be important current sources of exposure.

1. Introduction

Organochlorine pesticides (OCPs) are persistent organic pollutants associated with adverse neurodevelopmental outcomes (1), breast cancer (2, 3), metabolic dysfunction (4), and hormone disruption (5, 6). OCPs tend to accumulate in adipose tissue (7–12), especially in reproductive-aged women (13, 14). OCP concentrations in adipose tissue positively correlate with serum concentrations (15–17) although adipose concentrations may better reflect cumulative internal body burdens compared to serum (18, 19). The insecticide dichlorodiphenyltrichloroethane (p,p’-DDT) is one of the most well-known OCP, having been used extensively on crops and for vector control in the United States (U.S.) from 1945 to 1972, leading to widespread contamination of the environment (20, 21). In the U.S., racial differences in serum pesticide concentrations have been observed as far back as the 1960s (22), and concentrations are higher among Black women than White women (23). Nevertheless, there has been limited research on predictors of exposure to OCPs among reproductive-aged non-pregnant Black women.

With half-lives in humans of up to 7 years (24), many OCPs have been targeted by the Stockholm Convention on Persistent Organic Pollutants for elimination or restriction and labeled as probable carcinogens. Due to their environmental persistence, ability to bioaccumulate in humans, and adverse impacts on health, the U.S. banned the use of DDT and other OCPs for agricultural use starting in the 1970s. Uses for some extended into the late 1980’s, with Hexachlorobenzene as a fungicide until 1984 and chlordane-related pesticides, such as oxychlordane and trans-nonachlor, to control termites until 1988. Despite bans, these chemicals are still detectable in most of the U.S. population (14), the food chain (25, 26) and throughout the ecosystem (27–30). Once ingested and absorbed, OCP blood concentrations are influenced by factors related to lipid turnover and storage, including weight loss, body fat distribution, and lactation (15, 31, 32). A history of childbearing and lactation are proposed clearance routes inversely associated with OCP concentrations in blood and adipose tissue (33–44). Associations between OCPs and body mass index (BMI) are mixed (22, 39–41, 45, 46). For example, serum concentrations of p,p’-DDE have been positively (39, 46, 47) and inversely (22) associated with BMI, potentially due to the complicated nature of OCP distribution and metabolism across the body’s various adipose tissues (48).

In epidemiologic studies examining predictors of OCP concentrations, most of which were in pregnant women, determinants of OCPs vary across individual pesticides and their metabolites. The most consistent predictor of elevated OCP concentrations in blood and adipose tissue is older age (39, 41, 42, 44–46, 49, 50). Although ingestion of contaminated food sources is a putative route of exposure (8, 25, 26), associations with self-reported diet are inconsistent (13, 39–41, 43, 44, 49, 51–53), with some studies showing evidence of increased OCP concentrations with higher meat and large fish consumption (44, 51, 52, 54–56). Trace residues of OCPs are still detectable in tobacco products and in soils used to produce wine, and beer (57–62). Therefore, exposure to cigarettes and alcohol may be important sources of human exposure. Given racial disparities in exposure to environmental chemicals across the life course (63), sociodemographic and lifestyle factors are also potentially important predictors of circulating concentrations of environmental chemicals, including OCPs (40, 46).

We used cross-sectional data from a prospective cohort study of 742 Non-Hispanic Black women in the Detroit, Michigan metropolitan area to identify predictors of exposure to OCPs. We focused on demographic, behavioral, dietary, occupational, and medical history factors associated with pesticide body burdens in previous literature.

2. Subject and Methods

2.1. Study Population

The Study of Environment, Lifestyle, and Fibroids (SELF) is a prospective cohort study of 1,693 reproductive-aged Black women recruited from the Detroit, Michigan metropolitan area during 2010–2012, as described in detail elsewhere (64). Eligibility criteria included age (23–35 years) and self-identification as African American, Black, or partly African American. Exclusion criteria included prior diagnoses of uterine leiomyomata (fibroids), cancer, or autoimmune disease. Participants who were pregnant at recruitment had enrollment delayed until three months postpartum. For this sub-study of selected endocrine disrupting chemicals and leiomyomata, we used a case-cohort study design in which we randomly selected a subcohort of 639 participants from the 1,693 fibroid-free SELF participants at baseline. We then supplemented the random subcohort of 639 with and the remaining incident cases of leiomyomata through 60-months of follow-up (this included 103 cases that occurred outside of the randomly selected subcohort). This sampling strategy resulted in a total of 742 participants for the present analysis, 639 selected at baseline (183 who went on to become a case) plus the 103 supplemented cases. Demographic, behavioral, dietary, occupational, and medical history data were collected via self-administered questionnaires, telephone interviews, and in-person clinic visits. Women also completed a self-administered semi-quantitative food frequency questionnaire (65). The Institutional Review Boards of the Henry Ford Health System, National Institute of Environmental Health Sciences (NIEHS), and Boston University Medical Campus approved the study. The involvement of the Centers for Disease Control and Prevention (CDC) did not constitute engagement in human subjects’ research.

2.2. Measurement of Organochlorine Pesticides in Plasma

At baseline, enrolled participants provided a non-fasting blood sample. Collected samples were shipped to the NIEHS repository for storage at −80 degrees Celsius, and then shipped on dry ice in three batches to the CDC for analysis. We used gas chromatography/isotope dilution high-resolution mass spectrometry (66) to measure plasma concentrations of nine OCPs: 2,2-bis(4-chlorophenyl)-1,1-dichloroethene (p,p’-DDE), 2-(4-chlorophenyl)-2-(2-chlorophenyl)-1,1,1-trichloroethane (o,p’-DDT), 2,2-bis(4-chlorophenyl-1,1,1-trichloroethane (p,p’-DDT), hexachlorobenzene (HCB), β-hexachlorocyclohexane (β-HCH), γ-hexachlorocyclohexane (lindane; γ -HCH), oxychlordane, trans-nonachlor, and mirex. The choice of OCPs was informed by their prevalence of exposure in humans (14) and evidence of their effects on reproductive hormones and processes that could influence risk of uterine leiomyomata (11, 67–71). Quality control (QC) samples, reagent blanks, and blind duplicates were distributed across batches and coefficient of variation estimates for the above mentioned analytes ranged from 3.3–13.9% (72). We determined total plasma lipids using an enzymatic summation method (73) and reported all POP concentrations on a lipid-adjusted basis.

2.3. Potential Correlates

We focused on correlates associated with persistently occurring pollutants in previous literature, including: age (years, continuous); body mass index (BMI) calculated from height and weight as measured by technicians at baseline (kg/m², continuous); alcohol use in the past year (none, “moderate”: 1–5 drinks/day, or ≥4 drinks once per month or less, or “heavy”: ≥6 drinks/day, or ≥4 drinks twice per month or more); smoking history (never, former, or current); birth order (first, second, third or higher); parity (nulliparous or parous); cumulative duration of lactation (months, continuous, with nulliparous women coded as zero months); education (≤high school or General Education Diploma (GED), some college/Associate’s degree/technical degree, or ≥Bachelor’s degree); annual household income in U.S. dollars (<$20,000, $20,000-$50,000, or >$50,000); and cumulative residence in urban area (years, continuous). Dietary predictors of interest included frequency of consumption of the following food items: fish (grams/day, continuous), meat (grams/day, continuous), poultry (grams/day, continuous), eggs (grams/day, continuous), dairy (servings/day, continuous), vegetables (servings/day, continuous), fruit (cups/day, continuous), and tap or bottled water (glasses/day).

Measures of participants’ in utero exposures and early life experiences included: mother’s residence on a farm while pregnant with participant (yes/no), mother working on a farm while pregnant with participant (yes/no), or whether the participant lived on or visited a farm for ≥1 month before 18 years of age (yes/no). The majority of participants had the help of their mothers for questions on mother’s exposures and early life factors (85%), including whether they were breastfed as an infant (never, <3 months, or ≥3 months), both have been found to be reported with high reliability (74, 75). Measures of participant’s occupational exposure to pesticides included the following: “working on a farm or orchard” for ≥1 month (yes/no), “working with lawn care” for ≥1 month (yes/no), or “working with pesticides” ≥1 month (yes/no). We also assessed exposure to pesticides and bug repellant in the past 24 hours (yes/no).

2.4. Statistical Analysis

The median limit of detection (LOD) varied for each individual pesticide and ranged from 1.1 to 1.4 ng/g lipid (Table 2). Detection frequencies ranged from 0.3% (γ-HCH) to 100% (p,p’-DDE) and all analyses were restricted to chemicals with detection frequencies ≥60%, restricting all analyses to p,p’-DDE (100% detection), HCB (74.5% detection), oxychlordane (88.5% detection), and trans-nonachlor (91.4% detection). Values below the LOD were set to the LOD divided by the square root of two (76). We compared arithmetic means of each pesticide in the SELF cohort to the arithmetic mean of pooled serum samples representative of Non-Hispanic Black women aged 20–39 years from the National Health and Nutrition Examination Survey (NHANES) during 2009–2010 (14) (Table 2).

Table 2:

Characteristics of 742 participants in the prospective cohort Study of Environment, Lifestyle, and Fibroids (SELF) at baseline (2010–2012)

Characteristic	Mean (SD) or N	%
Mean age in years (SD)	28.6 (3.5)	-
Age, years
23–25	179	24.1
26–28	182	24.5
29–31	198	26.7
32–35	183	24.7
Mean BMI, kg/m2 (SD)	33.7 (9.5)	-
BMI, kg/m2
<25	139	18.7
25–29.9	157	21.2
30–34.9	147	19.8
35–39.9	124	16.7
≥40	175	23.6
Education
≤High school diploma/GED	156	21.0
Some college/Associate’s/Technical	385	51.9
≥Bachelor’s degree	201	27.1
Annual household income, U.S. dollar
<20,000	340	45.8
20,000–50,000	280	37.7
>50,000	122	16.4
Urban residence in years
<5	70	9.4
5–9	40	5.4
10–19	155	20.9
≥20	477	64.3
Cigarette smoking history
Never	544	73.3
Past	59	8.0
Current	139	18.7
Alcohol consumption, past yeara
None	210	28.3
Moderate	383	51.6
Heavy	149	20.1
Birth order
First	308	41.5
Second	210	28.3
Third or higher	224	30.2
Breastfed as an infant
Never breastfed	521	70.2
<3 months	110	14.8
≥3 months	111	15.0
Parity
Nulliparous	281	37.9
1 birth	201	27.1
2 births	127	17.1
≥3 births	133	17.9
Duration of lactation among parous women (n=461)
None	159	34.5
1–3 months	113	24.5
4–6 months	70	15.2
7–12 months	63	13.7
>12 months	56	12.2
Mean dietary intake (SD)
Fish (grams/week)	27.8 (32.6)	-
Meat (grams/week)	53.5 (63.7)	-
Poultry (grams/week) (SD)	35.6 (36.5)	-
Egg (grams/week) (SD)	14.9 (14.4)	-
Dairy (servings/day) (SD)	1.0 (0.8)	-
Vegetable (servings/day) (SD)	3.0 (2.3)	-
Fruit (cups/day) (SD)	1.4 (1.2)	-
Water intake (glasses of bottled or tap/day)
≤2	215	29.0
3–4	298	40.2
≥5	229	30.9
Occupation (≥1-month):
Farm	4	0.5
Lawn care	13	1.8
Pesticides	5	0.7
Participant’s mother:
Lived on a farm for ≥1-month during pregnancy	4	0.5
Worked on a farm for ≥1-month during pregnancy	6	0.8
Before the age of 18, lived on or visited a farm for ≥1-month	29	3.9
Used pesticides in past 24 hours	11	1.5
Used bug repellant in past 24 hours	6	0.8

Abbreviations: BMI= Body mass index, GED= General education diploma, SD= standard deviation

^aAlcohol use in past year categories: none, “moderate”: 1–5 drinks/day, or ≥4 drinks once per month or less, or “heavy”: ≥6 drinks/day, or ≥4 drinks twice per month or more

We used Spearman correlation coefficients (rho) to evaluate correlations between individual OCPs. We fit linear regression models using log-transformed OCP concentrations as the dependent variable and each potential predictor as the independent variable, first unadjusted and then adjusted for all other predictors, as mentioned above. The percent (%) difference was calculated by exponentiating the beta coefficient comparing one category of the predictor to the reference category of that predictor, subtracting one, and then multiplying by 100 (i.e., (e^beta-1)*100). We used a Markov chain Monte Carlo method to impute missing data, which generated five imputation datasets that statistically combined estimates across the datasets (77). The percent missing for variables ranged from 0% (e.g., age) to 5% (breastfed as an infant). We imputed plasma OCP concentrations for <5% of participants: n=29 for HCB, n=4 for oxychlordane, and n=1 for trans-nonachlor.

To evaluate the extent to which inclusion of incident leiomyomata cases that were not part of the random subcohort influenced our results, we performed an additional sensitivity analysis restricted to the random subcohort of 639 women selected at baseline. Further, given the propensity for deposition of OCPs in adipose tissue, we performed analyses stratified by BMI (<30 vs. ≥30 kg/m²). Our hypothesis was that observed associations with serum OCP concentrations would be stronger in women with a BMI ≥30, assuming that they have higher cumulative body burdens due to the lipohilic nature of these chemicals.

3. Results

The mean age of the analytic sample was 28.6 years (standard deviation=3.5) and 60.1% had a BMI ≥30 kg/m² (Table 3). Nearly 75% of participants never smoked, and 20.1% reported heavy alcohol consumption in the past year. Approximately 62% of participants were parous, and among them 65% had breastfed at least one infant. Twenty-nine percent reported being breastfed as an infant, and 41.5% were first-born children. Few participants had potential occupational exposure to OCPs or exposure to farms or farm work (<5%). Mean total daily intakes of fish, meat, poultry, and eggs were within recommended levels from the U.S. Department of Agriculture, while intakes of dairy, fruit, and vegetable were generally below recommended levels (78). Baseline plasma OCP concentrations (ng/g lipid) in SELF were comparable to or lower than non-Hispanic Black women of similar age from NHANES in 2009–2010 (Table 3). Correlations across individual OCPs were generally moderate to high, with correlations ranging from 0.51 to 0.86 (Table 4). Chlordane and trans-nonachlor were highly correlated (rho=0.86).

Table 3:

Spearman correlations for log-transformed plasma organochlorine pesticide concentrations (ng/g lipids) among 742 Study of Environment, Lifestyle, and Fibroids (SELF) participants.

Organochlorine pesticides	p,p’-DDE	Hexachlorobenzene	Oxychlordane	Trans-nonachlor
p,p’-DDE	1.0	0.58	0.70	0.67
Hexachlorobenzene		1.0	0.58	0.51
Oxychlordane			1.0	0.86
Trans-nonachlor				1.0

Abbreviations: p,p’-DDE= 2,2-Bis(4-chlorophenyl)-1,1-dichloroethene

In models mutually adjusted for all other predictors, a 5-year increase in age was associated with a 8.3% increase in p,p’-DDE (95% CI: −2.7%, 20.7%), 24.3% increase in oxychlordane (95% CI: 12.4%, 37.5%), and a 26.1% increase in trans-nonachlor (95% CI: 12.0%, 41.9%) (Figure 1a). A 5-kg/m² increase in BMI was associated with a 8.9% decrease in p,p’-DDE (95% CI: −10.3%, −7.4%), a 5.3% decrease in HCB (95% CI: −6.3%, −4.3%), a 6.8% decrease in oxychlordane (95% CI: −8.2%, −5,4%), and a 5.5% decrease in trans-nonachlor (95% CI=−7.2%, −3.8%) (Figure 1b). Compared with no alcohol use, moderate and heavy alcohol use were weakly associated with higher concentrations of plasma OCPs, with percent increases ranging from 3.7% to 8.0% (Figures 1c and 1d). Plasma OCP concentrations for never and past smokers did not differ appreciably (Figure 1e). Current smokers had higher plasma concentrations than never smokers across all pesticides, with the strongest association for trans-nonachlor (19% difference, 95% CI: 9.0%, 30.4%) (Figure 1f). We provide the percent differences and 95% confidence intervals for all predictors in the Supplemental Table.

Figure 1: Adjusted percent differences and 95% confidence intervals in organochlorine pesticides by age, body mass index (BMI), alcohol use in the past year, and cigarette smoking status.

Abbreviations: p,p’-DDE= 2,2-Bis(4-chlorophenyl)-1,1-dichloroethene; HCB=Hexachlorobenzene.

All x-axis are percent differences and 95% confidence intervals. All models use the log-transformed plasma concentrations as the dependent variable and include age, education, household income, occupation (farm or orchard, lawn care, or pesticides), cigarette smoking status, alcohol use, body mass index (BMI), birth order, breastfeeding history as an infant, parity, lactation history, early exposure to pesticides (maternal farm residence or work during pregnancy, and lived on or worked on a farm before the age of 18), fish intake, meat intake, poultry intake, dairy intake, egg intake, fruit intake, vegetable intake, water intake, and total caloric intake.

In models mutually adjusted for all other predictors, relative to never having been breastfed in infancy, having been breastfed for <3 months was not materially associated with plasma OCP concentrations (Figure 2a), while having been breastfed for ≥3 months was associated with 15.2% higher p,p’-DDE (95% CI: 6.0%, 25.2%), 13.8% higher oxychlordane (95% CI: 4.9%, 23.6%), and 15.1% higher trans-nonachlor (95% CI: 4.7%, 26.6%) concentrations (Figure 2b). Compared to nulliparous women, parous women had suggestively lower HCB, oxychlordane, and trans-nonachlor, although estimates were not statistically significant (Figure 2c). Among parous women, a 3-month increase in lactation was associated with 1–4% lower plasma concentrations of OCPs (Figure 2d). We observed 2–9% lower plasma concentrations of all OCPs comparing participants who were third born or higher to those who were first-born (Figure 2f). Greater plasma concentrations of all OCPs were observed comparing the highest (≥5 glasses/day) with the lowest water consumers (≤2 glasses/day), with percent increases ranging from 4–15% (Figure 3b). Farm work or residence were rare (<4%) and observed associations with plasma OCP concentrations were imprecise. For example, residence on a farm or farm work prior to the age of 18 was associated with 31.6% higher p,p’-DDE (95% CI: 12.9%, 53.0%) (Figure 3c).

Figure 2: Adjusted percent differences and 95% confidence intervals in organochlorine pesticides by history of breastfeeding in infancy, parity, lactation history, and order of birth.

Abbreviations: p,p’-DDE= 2,2-Bis(4-chlorophenyl)-1,1-dichloroethene; HCB=Hexachlorobenzene.

All x-axis are percent differences and 95% confidence intervals. All models use the log-transformed plasma concentrations as the dependent variable and include age, education, household income, occupation (farm or orchard, lawn care, or pesticides), cigarette smoking status, alcohol use, body mass index (BMI), birth order, breastfeeding history as an infant, parity, lactation history, early exposure to pesticides (maternal farm residence or work during pregnancy, and lived on or worked on a farm before the age of 18), fish intake, meat intake, poultry intake, dairy intake, egg intake, fruit intake, vegetable intake, water intake, and total caloric intake.

Figure 3: Adjusted percent differences and 95% confidence intervals in organochlorine pesticides by water intake (tap or bottled), and farm exposure in early life.

Abbreviations: p,p’-DDE= 2,2-Bis(4-chlorophenyl)-1,1-dichloroethene; HCB=Hexachlorobenzene.

All x-axis are percent differences and 95% confidence intervals. All models use the log-transformed plasma concentrations as the dependent variable and include age, education, household income, occupation (farm or orchard, lawn care, or pesticides), cigarette smoking status, alcohol use, body mass index (BMI), birth order, breastfeeding history as an infant, parity, lactation history, early exposure to pesticides (maternal farm residence or work during pregnancy, and lived on or worked on a farm before the age of 18), fish intake, meat intake, poultry intake, dairy intake, egg intake, fruit intake, vegetable intake, water intake, and total caloric intake.

Education, income, and all other dietary variables were not meaningfully associated with plasma OCP concentrations (Supplemental figures a–k). Mother’s exposure to farm work or farm residence was not associated with participant’s plasma OCP concentrations (Supplemental figures l and m). Exposure to pesticides, farm work, lawn care, and bug repellant were also rare (< 2%). Occupations were not associated with plasma OCP concentrations (Supplemental figures m–p), while pesticide exposure in the past 24 hours was associated with 35% higher plasma concentrations of p,p’-DDE (95% CI=6.7%, 70.7%) (Supplemental figure q). In sensitivity analyses restricted to the random subcohort selected at baseline, heavy alcohol use was no longer associated with elevated levels of HCB. All other associations were similar (data not shown). Predictors of p,p’-DDE were suggestively stronger among women with BMI ≥30 kg/m² (relative to BMI <30 kg/m²) for heavy alcohol use (8.3% higher vs. 3.7% higher) and having been breastfed for ≥3 months in infancy (19.9% higher vs. 10.6% higher).

4. Discussion

Four out of the nine pesticides selected for analysis were detectable in ≥60% of SELF participants, and p,p’-DDT and β-HCH were detectable in 50% and 36% of participants, respectively. Lindane (γ-HCH), mirex, and o,p’-DDT were detectable in <10% of SELF-participants. Chlordane-related OCPs (oxychlordane and trans-nonachlor) were highly correlated with one another. In models adjusted for all potential predictors, older age, heavy alcohol use, current cigarette smoking, having been breastfed ≥3 months in infancy, and greater water intake were associated with higher plasma OCP concentrations. BMI, parity, longer lactation, and higher birth order were inversely associated with plasma OCP concentrations. Education, income, maternal exposures, and other dietary factors were not meaningfully associated with elevated plasma concentrations.

Among our study population of 23–35 year-old Black U.S. women, plasma concentrations of the selected OCPs were either comparable to or lower than those from a NHANES sample of Non-Hispanic Black U.S. women of a similar age range (14). With the exception of oxychlordane, pesticides that were not estimable by NHANES due to insufficient detection levels were also present at low frequencies of detection in our sample (Table 1). In contrast, oxychlordane was not estimable by NHANES yet it was highly detectable among SELF-participants (88.5%). Trans-nonachlor, another chlordane-related compound, was also commonly detected (91.4%). We may have expected to find higher chlordane residues among SELF-participants because participants resided in Michigan, where chlordane-related compounds have been detected in the nearby environment (79, 80). In contrast, NHANES participants were sampled from across the U.S. HCB and p,p’-DDE were also highly prevalent in our participants (74.5% and 100%, respectively) and observed in women globally (31, 81, 82).

Table 1:

Distribution of plasma organochlorine pesticides (ng/g lipid) in the Study of Environment, Lifestyle, and Fibroids (SELF) cohort (N=742) and NHANES (2009–2010)

	SELF Participants (2010–2012) 23–35 year old Black Women				NHANES 2009–2010 20–39 year old Non-Hispanic Black Women
Organochlorine Pesticide	Median LOD	% Detected (>LOD)	Median (25th, 75th percentiles)	Mean	Mean
p,p’-DDT	1.4	50.1	>60% below the LOD	-	3.0
o,p’-DDT	1.1	3.0	>60% below the LOD	-	NC
p,p’-DDE	1.4	100.0	51.2 (38.6, 69.3)	57.7	110.0
Hexachlorobenzene	1.4	74.5	5.9 (5.0, 7.2)	6.3	6.0
β-HCH	1.1	36.3	>60% below the LOD	-	NC
γ-HCH (Lindane)	1.1	0.3	>60% below the LOD	-	NC
Oxychlordane	1.1	88.5	2.1 (1.6, 2.8)	2.3	NC
Trans-nonachlor	1.2	91.4	2.4 (1.8, 3.4)	2.8	6.8
Mirex	1.1	8.1	>60% below the LOD	-	NC

Abbreviations: LOD=Limit of detection, NC= Not calculated “the proportion of results below LOD for that congener was too high to provide a valid result”, p,p’-DDT= 2,2-Bis(4-chlorophenyl-1,1,1-trichloroethane, HCH= Hexachlorocyclohexane

Higher plasma concentrations of p,p’-DDE and chlordanes, in particular, are consistent with their reported persistence and biochemistry. p,p’-DDE is the main bioaccumulative metabolite of DDT and, when compared with other DDT metabolites, is detectable at higher concentrations in human blood and adipose tissue. Positive associations between OCPs and age in our study of non-pregnant Black U.S. women of reproductive-age are consistent with previous studies (15, 32, 39, 41, 42, 44–46, 49, 50), albeit most were of predominantly White populations. Given that older participants were more likely to be exposed to OCPs earlier in life and continue to have higher plasma OCP concentrations compared with younger participants, these findings suggest that exposure to OCPs earlier in life may continue to contribute to current blood concentrations.

We observed inverse associations between increasing BMI and plasma OCP concentrations. Storage of OCPs in fat has been called “protective,” as it decreases the amount of bioavailable pesticides for biological activity (83). In line with this hypothesis, inverse associations between BMI and blood OCP concentrations are plausible and previously reported (22). A nine percent decrease in p,p’-DDE per 5-kg/m² increase in BMI in our study (Supplemental Table) is nearly identical to estimates reported by James and colleagues (22) using pre-pregnancy BMI and serum collected from U.S. pregnant women in the 1960s. Lactation mobilizes fat and is a major route of excretion of OCPs and other persistently occurring pollutants (39, 49, 84), and is inversely associated with p,p’-DDE and HCB concentrations in adipose tissue and positively associated with concentrations in milk (15, 31). Higher plasma concentrations of p,p’-DDE and chlordane-related pesticides among participants who themselves were exposed to prolonged breastfeeding in infancy are also consistent with this hypothesis, and may suggest that inherited body burdens from early life may persist into adulthood.

Reported associations between OCPs and tobacco smoking are mixed (40, 85). Elevated concentrations of serum HCB, β-HCH, and p,p’-DDE concentrations have been associated with smoking habits among pregnant U.S. women (40) and in Finnish samples (50% males) (86). Pesticides have also been detected in tobacco smoke (87) and tobacco leaves (88, 89). Our study suggests that trace residues in the soil may still be an important source of human exposure. However, in contrast, other studies report no associations with smoking habits (85). Pesticides have also been detected in vineyard soils and wines (58, 59), as well as in hops used in beer brewing (60, 61). Trace levels of OCPs in the soils of these agricultural products may also help to explain elevated plasma concentrations in heavy drinkers in our cohort. Ingestion of contaminated food sources is a putative route of OCP exposure (8, 25, 26). However, associations with self-reported diet are inconsistent across the literature (13, 39–41, 43, 44, 49, 51–53). In our study, water consumption was the most important dietary predictor. Although pesticides from runoff and residential uses can be detected in water sources (90) and public water contamination of chlordane has been reported in the U.S. (91), this finding should be interpreted with caution. We did not distinguish tap from bottled water consumption on our questionnaire, and we could not eliminate the possibility of unmeasured or residual confounding from factors associated with water consumption and potential exposure to OCPs (e.g. gardening). Further studies are needed to determine whether these findings are robust and what water sources, if any, may be implicated.

Identified differences in predictors of exposure across OCPs are likely due to differences in chemical properties related to absorption, distribution, metabolism, excretion, and toxicity (92). For example, the chlordane-related compounds (oxychlordane and trans-nonachlor) have similar chemical structures and effect estimates were similar for most predictors. Chlordanes and p,p’-DDE were also strongly positively correlated with one another and have long environmental half-lives in soil (5–15 years) (82). At a constant rate of intake, the concentration of DDT in adipose tissue is known to reach a steady-state value and remain relatively constant (83). Therefore, elevated plasma levels of p,p’-DDE and chlordane isomers may reflect both early-life exposures (e.g., having been breastfed in infancy) as well as prolonged exposure to residues in agricultural and food products (e.g., tobacco, alcohol, and water).

Our study has some limitations. First, the study analysis was cross-sectional in that predictors and plasma concentrations were assessed at baseline only. Reporting of some predictors, such as dietary intake in the last year, was retrospective. Dietary measurement error can lead to underestimated effect estimates and reduced precision (93). However, we do not believe that reporting of most correlates would be differential with respect to OCP plasma concentrations as women are not usually aware of their exposure. Second, we did not have information on water sources, making conclusions with water consumption tenuous and requiring additional research. Lastly, findings may not be generalizable to all U.S. women, which is especially the case for chlordane-related pesticides, where the historic sources of exposure in Michigan differed from those of other states (79, 80). Our study has a number of strengths. We collected data on a wide range of potential predictors and confounders, including sociodemographic, educational, behavioral, dietary, early-life, occupational, and medical history factors. These factors were mutually controlled for in all analyses. Furthermore, we used state-of-the-art detection methods for measuring OCP concentrations in plasma which are positively correlated with adipose concentrations (15, 16). Lastly, SELF fills an important gap in the environmental health literature by focusing on an understudied population.

5. Conclusions

Older age, heavy alcohol use, current cigarette smoking, having been breastfed ≥3 months in infancy, and greater water intake were associated with higher plasma OCP concentrations in our study. Our findings suggest that exposure to OCPs earlier in life may directly or indirectly contribute to current blood concentrations, and that agricultural and water sources may be sources of pesticide exposure in U.S. Black women of reproductive age.

Highlights.

• Older age was associated with higher plasma pesticide concentrations

• Contamination of tobacco, alcohol and water may be important sources of pesticides

• Diet was not associated with plasma pesticide concentrations