Biomonitoring of toxic metals, organochlorine pesticides, and polybrominated biphenyl 153 in Michigan urban anglers

Abstract

The 32-mile Detroit River and surrounding tributaries have been designated as a Great Lakes Area of Concern due to pollution from decades of municipal and industrial discharges, sewer overflows and urban development. The Agency for Toxic Substances and Disease Registry and the Michigan Department of Health and Human Services conducted a biomonitoring study to assess exposures to persistent toxic substances in Detroit urban shoreline anglers who may be at high exposure risk due to consumption of locally caught fish. Using a modified venue-based sampling approach, 287 adult shoreline anglers along the Detroit River were recruited and participated in the program. Study participants provided blood and urine specimens and completed a questionnaire interview. In this report, we examine percentile estimates for blood lead, blood manganese, urine arsenic, urine mercury, urine cadmium, organochlorine pesticides in serum (mirex, hexachlorobenzene, chlordane), and serum polybrominated biphenyl 153 (PBB 153) concentrations among study participants. Multiple linear regression was used to identify predictors of contaminant concentrations. The Detroit urban anglers’ blood lead concentrations were 2 times higher than the general adult U.S. population (median (95% CI): 2.9 μg/dL (1.8–2.3) vs. 0.94 μg/dL (0.90–0.98)). PBB 153 levels were 1.8 times higher than the general adult U.S. population at the 95th percentile (95th percentile, 95% CI: 62.7 ng/g of lipid, 53.2–75.2 vs. 34.6 ng/g of lipid, 12.8–66.8). Percentile estimates of the other study pollutants were similar to background levels found in the general U.S. population. Eating more locally caught fish was not associated with increased body burdens for any of the contaminants examined in this report. Higher blood lead was associated with increased age, male sex, current smoking, residing in a home built before 1960, an annual income less than $25,000, and a work history of lead paint removal. Evidence of PBB exposure in our study cohort likely reflects the continued effect of a widespread contamination of livestock feed in 1973 among Michigan’s lower peninsula population. These study results help determine if the pollutants examined warrant further consideration in subsequent population-based biomonitoring of frequent consumers of fish from the Detroit River and surrounding waterways. The biomonitoring data from this study also served to inform public health officials regarding the potential need for environmental public health actions to reduce harmful exposures.

1. Introduction

Pollution caused by persistent toxic substances in the Great Lakes and connecting waterways has long been a focus of environmental health concern. For more than a century, the Great Lakes and surrounding areas have been polluted by industrial and municipal waste. Since the 1970s, the U.S. Environmental Protection Agency (EPA) and Environment Canada have coordinated efforts to eliminate the discharge of persistent toxic substances to the Great Lakes and to restore the chemical, physical, and biological integrity of these waters (EPA, archive; GLWQA, 2012; EPA, 2018). Although releases of toxic pollutants have been reduced significantly over the last 40 years, so-called legacy contaminants (banned decades ago) including polychlorinated biphenyls (PCBs), and dichlorodiphenyldichloroethylene (DDE) persist in contaminated sediments and the Great Lakes ecosystem (EPA, 2018). Mercury and other pollutants continue to enter the Great Lakes from nearby and global sources through air deposition (EPA, 2017). These pollutants in the Great Lakes and connecting waterways can bio-accumulate in aquatic ecosystems and have the potential for adverse health effects in humans who are at the top of the food chain (GLWQB, 1985). Biomonitoring programs for legacy organochlorine pesticides, mercury and PCBs in fish tissue have been conducted since the 1970s as consumption of fish is a prominent route of exposure (McGoldrick and Murphy, 2016).

In 2010, the EPA, in collaboration with other federal agencies, launched the Great Lakes Restoration Initiative (GLRI) to provide resources to accelerate remediation efforts in the Great Lakes basin and to prevent environmental exposure related health risks (EPA, 2017). As part of the GLRI, the Agency for Toxic Substances and Disease Registry (ATSDR) established the Biomonitoring of Great Lakes Populations (BGLP) program to assess human exposure to toxic chemicals among susceptible populations (ATSDR, 2018). The BGLP program consists of several cross-sectional studies carried out collaboratively with state health departments in Michigan, Minnesota, New York, and Wisconsin. The state programs targeted diverse, susceptible adult populations living in designated areas of contamination across the Great Lakes basin and measured over 70 legacy contaminants and pollutants of local and/or emerging concern (Wattigney et al., 2019a).

The BGLP program conducted by the Michigan Department of Human Health Services (MDHHS) focused on urban, shoreline anglers in Detroit. During the 19th century, Detroit grew into a thriving hub of commerce and industry with multiple manufacturing firms taking advantage of the transportation resources afforded by the river and a parallel rail line. Detroit’s numerous and large industrial pollution sources such as coal-fired power plants, steel facilities, petroleum refineries, among others contribute to higher levels of toxic metals in the air. Moreover, Detroit has many low-income and minority neighborhoods and schools in proximity to large industrial facilities which has been shown to contribute to health disparities attributable to air pollutants for the Detroit urban area (Martenies et al., 2017). The Detroit River and surrounding tributaries have been designated as a Great Lakes Area of Concern (AOC) due to pollution from decades of municipal and industrial discharges and urban development (EPA, 2020). The BGLP Detroit urban angler program measured legacy contaminants (e.g., PCBs, dioxins, furans, pesticides, mercury, and lead) that were found in Michigan’s surface water resources, including the Great Lakes. In addition, the program measured contaminants that are of local concern, including manganese (prevalent in air and soil in southeast Michigan) (MDCH, 2009); arsenic and cadmium (prevalent soil and water contaminants); and 2,2′,4,4′,5,5′-hexabromo-biphenyl (PBB-153), the congener of highest concentration in commercial polybrominated biphenyls (PBBs). A report on the toxicants most closely related to fish consumption, i.e., total blood mercury, PCBs, DDE, and dioxin-like total toxic equivalency (TEQ) concentrations, among the Detroit urban angler cohort has been published elsewhere (Wattigney et al., 2019b).

This article presents blood lead, blood manganese, urine total arsenic, urine total mercury, urine cadmium, organochlorine pesticides in serum (mirex, hexachlorobenzene (HCB), chlordane), and serum PBB-153 concentrations for the Detroit urban anglers and an assessment of potential exposure sources. These metals, metalloids and pesticides were historically identified as critical pollutants present in the soil, water, sediment, or aquatic biota of the Great Lakes Basin for priority action (GLWQB, 1985). Although generally considered to be nontoxic, organic forms of arsenic can be present in seafood and excreted in urine within 48 h of ingestion (ATSDR, 2011a). Exposure to organochlorine pesticides, known to be toxic, can occur in the general population through consuming foods such as dairy products, fish or wildlife from contaminated areas (ATSDR, 2002). Detroit’s low-income urban population has a high risk of lead exposure due to industrial emissions and subsequent atmospheric deposition of lead (Sherman et al., 2015) and a high prevalence of homes built before 1960 having heavily leaded paint (DHD, 2016; MDHHS, 2016). Urinary mercury, presented in this report, reflects inorganic Hg exposures, not methylmercury. Methylmercury bioaccumulates in fish and is commonly measured in people by total blood mercury (CDC, 2017a). Body burden levels of PBB, a flame retardant used in the U.S. until the 1970s (Fries, 1985), were of interest due to the 1973 contamination of dairy cattle and other livestock feed in Michigan. This incident led to widespread PBB exposure among people in Michigan who consumed contaminated livestock, poultry, eggs, and dairy products (Wolff et al., 1982; Anderson, 1989). PBBs do not degrade either chemically or metabolically, and exposure to these compounds are still a public health concern due to their persistence.

2. Methods

All study activities were approved by the federal Office of Management and Budget (Control Number 0923–0044) and the MDHHS Institutional Review Board. The study methods, including sampling strategy and data collection, are described in previous publications (Wattigney et al., 2019a, 2019b). In summary, a modified venue-based sampling method was used to establish a sampling frame for shoreline anglers. Venue-based sampling is a strategy for targeting hard-to-reach populations whose members tend to assemble at discrete locations (Mackellar et al., 1996). Staff initially conducted an enumeration phase to estimate the number of adult shoreline anglers at identified fishing venues on various days of the week and times of the day. Stratified random sampling was then used to establish a venue-day-time calendar for the next phase of data collection. At each scheduled shoreline fishing venue-day-time, trained interviewers administered a standardized questionnaire to anglers to determine eligibility and willingness to participate in the biomonitoring study. Eligible anglers included anglers aged 18 years and older who ate at least two meals of locally caught fish per month. MDHHS program staff conducted follow-up phone calls with randomly selected eligible venue-based respondents to confirm eligibility and schedule clinic appointments. Study clinics were held from June through November 2013 at church facilities and community centers near the shoreline fishing locations to collect blood and urine samples; measure height, weight, and blood pressure; and administer a questionnaire. Participants provided written informed consent prior to data collection. Standardized questionnaire interviews were used to gather information on age, sex, race/ethnicity, education level, family income, years living in the Detroit area, year of construction for current housing, drinking water source, job history, hobbies, smoking history, breastfeeding history for females, recreational water activities in local waterways, and eating patterns in the past 12 months for locally caught fish, wildlife, and home-raised or home-grown food.

2.1. Laboratory analysis

The MDHHS Analytical Chemistry Section performed laboratory analysis of all analytes presented in this report. Lead and manganese in whole blood samples were measured using a dynamic reaction cell inductively coupled plasma mass spectrometry (DRC-ICP-MS). Lead was analyzed in standard mode, while manganese was analyzed in reaction cell mode utilizing ammonia as a cell gas to remove interferences (CDC, 2012a). Serum PBB-153 and organochlorine pesticides (mirex, hexachlorobenzene, heptachlor epoxide, trans-nonachlor, oxychlordane, lindane) were measured by gas chromatography with electron capture detection (GC/ECD). Positive identification of these analytes and possible interfering substances were performed by gas chromatography mass spectrometry, when possible.

Urinary total arsenic and cadmium levels were determined by DRCICP-MS (CDC, 2012b). Arsenic was analyzed in reaction cell mode utilizing oxygen as a cell gas to move arsenic away from interference to a new mass as arsenic oxide. Cadmium was analyzed in standard mode with an interference correction for tin. Total mercury in urine was measured using inductively coupled plasma mass spectrometry (ICP-MS) (CDC, 2012c). Urine creatinine is an indicator of urine dilution and was measured by Sparrow Laboratories using the Olympus AU2700, 640 E and 400 Creatinine procedure, which is a kinetic modification of the Jaffe procedure in which creatinine reacts with picric acid at alkaline pH and forms a yellow-orange complex. Laboratory methods for analyte measurements are described in more detail elsewhere (Wattigney et al., 2019a). The MDHHS laboratory was Clinical Laboratory Improvement Amendments (CLIA) certified, participated in appropriate external proficiency testing programs, and provided documented proficiency testing results and standard operating procedures for all analytical methods including detailed quality control and quality assurance procedures. A summary of analyte measurements, limits of detection (LOD) and the percent of measures below the LOD are presented in Supplementary Table 1.

2.2. Data analysis

Statistical analyses were performed with SAS® software version 9.4 (SAS/STAT software,). Descriptive statistics are presented to describe characteristics and questionnaire responses for study participants. Percentiles and distribution-free 95% confidence intervals (CI) were calculated for each analyte using SAS PROC UNIVARIATE with CIQUANTDF option. Percentile estimates at the 50th and 95th percentiles are compared to National Health and Nutrition Examination Survey (NHANES) survey results reported in the Fourth National Report on Human Exposure to Environmental Chemicals (CDC, 2019). The NHANES, conducted by the National Center for Health Statistics, collects data to assess the health and nutrition status of the civilian, non-institutionalized U.S. population. Laboratory analyses of biological samples from NHANES participants in 2-year cycles provides an ongoing assessment and overview of the exposure of the U.S. population to environmental chemicals (Calafat, 2012). For assessing the metals and metalloids, we compared percentiles to those reported for the NHANES 2013–2014 survey (CDC, 2019). For the organochlorine pesticides and PBB-153, we compared percentiles to those reported for NHANES 2003–2004 (CDC, 2009), which is the most recent data available. Following NHANES guidelines, our study chemical concentrations less than the LOD were assigned a value equal to the LOD divided by the square root of 2 for the percentile estimates. Furthermore, the lipophilic organochlorine pesticides and PBB-153 are presented as per gram of total lipid. For chemicals measured in urine, levels are presented as per gram of creatinine (i.e., creatinine corrected) to adjust for urine dilution. We also calculated percentiles using NHANES public use data files for non-Hispanic adult black males only as a more restricted population closer to our study demographics.

For the four biomarkers detected in 60% or more of participants (blood lead, blood manganese, urine cadmium, and serum PBB-153), regression analysis was used to identify potential exposure sources. Random values from a uniform distribution between zero and the LOD were substituted for concentrations less than the LOD (Rocque and Winker, 2004). The regression analysis for serum PBB153 and urinary cadmium included the unadjusted analyte concentration with either serum total lipids or urinary creatinine, respectively, added as a separate independent variable. This approach has been shown to generated more precise effect estimates to control for the between-subject biological variations, while the statistical significance of other variables in the model are independent of the effects of lipid or creatinine concentration (Barr et al., 2005; Schisterman et al., 2005). In a preliminary set of analyses, we used multivariate analyses of variance and regression models to identify predictors of each biomarker. Potential predictors included age, sex, race/ethnicity (non-Hispanic black, non-Hispanic white, other), years residing in the Detroit area, BMI kg/m² category (<25 - normal, 25 to <30 - overweight, and ≥30 - obese), educational attainment (high school graduate or more, less than high school graduate), use of fish oil supplements (yes/no), use of herbal supplements (yes/no), recreational activities in local waterways in past year (yes/no), current smoker (yes/no), history of selected occupations, history of selected hobbies, locally caught fish consumption, and local wildlife consumption. For locally caught fish, species and location-specific data were combined into a single variable to represent the number of meals in the past year. Variables that were statistically significant in the preliminary analysis (p < .10) after adjusting for age were included in multiple regression models to evaluate their independent association with each biomarker. Biomarker concentrations and the number of locally caught fish meals in the past year were natural log (ln)-transformed to help normalize extreme values. Tests for normality and regression model diagnostics did not indicate a need for non-linear analysis. The multiple imputation (MI) procedure in SAS® was used to impute 5 iterations of binary values for the 32 missing values for annual family income (less than $25,000, $25,000 or more). The Markov Chain Monte Carlo algorithm produced imputed values that were set to 0 or 1 based on the distribution of non-missing values. Regression models that identified annual family income as a statistically significant predictor were reexamined excluding these imputed values to ensure similar variable parameters. The three missing values for race/ethnicity were coded as non-Hispanic black which was the predominant race/ethnicity of study participants. For HCB, mirex, trans-nonachlor and oxychlordane, the geometric mean number of locally caught fish meals in the past year was compared between urban anglers with measures < LOD and urban anglers with detectable levels of each chemical using analysis of variance unadjusted and adjusted for age.

3. Results

Initially, 2660 anglers were interviewed at shoreline fishing venues. Among the interviewed anglers, 1399 were potentially eligible, of those 582 declined further participation, resulting in 817 eligible participants to be recontacted. The recruitment results at each stage of the modified venue-based sampling are presented in detail elsewhere (Wattigney et al., 2019a). The program successfully recruited 287 (35.1%) anglers who were mostly males (80.8%), non-Hispanic black (78.7%), and aged 40 years or older (81.7%) (Table 1). The majority of participants had been living in the Detroit area for 21 years or longer (86.3%), lived in a home built before 1960 (65.1%), were low-income with 73.7% earning less than $50,000 per year (48.4% less than $25,000) (Table 1).

Table 1.

Selected characteristics of Detroit urban angler study participants.

Age group, race/ethnicity, by sex
	Males (n = 232)			Females (n = 55)
	Non-Hispanic black	White	Other	Non-Hispanic black	White	Other
	Number (%)			Number (%)
Age
18–39 years	27 (11.6)	a	a	11 (20.0)	a	a
40–59 years	113 (48.7)	10 (4.3)	17 (7.3)	18 (32.7)	a	a
60–79 years	43 (18.5)	a	a	14 (25.5)	0	a
Total	183 (78.9)	21 (9.1)	26 (11.2)	43 (78.2)	a	a
Characteristic		Number (%)		Characteristic		Number (%)
Current smoker		144 (50.2)		Home built before 1960 (55 don’t know)		149 (65.1)
BMI				Years residing in Detroit area
<25 (normal)		49 (17.2)		1–10 years		a
25–29 (overweight)		98 (34.4)		11–20 years		23 (8.1)
≥30 (obese)		138 (48.4)		21 years or more		246 (86.3)
Education				Annual family income (32 missing)
Less than high school		52 (18.3)		Less than 25 K		111 (43.5)
High school graduate		101 (35.4)		$25 K to less than 50 K		77 (30.2)
Some college or degree		82 (28.8)		$50 K or more		67 (26.3)
College degree		50 (17.5)

^aRepresents fewer than 10 individuals and suppressed to protect privacy.

Table 2 presents a summary of self-reported work history, hobbies, and local wildlife and fish consumption. The two most frequent jobs reported were “worked for an automobile manufacturer” (44.9%) and “worked in a foundry, a smelter, a welding facility or steel mill” (36.1%), with 23.6% of the participants reporting both. A smaller percentage of urban anglers reported having performed other jobs in categories that pose a risk of occupational chemical exposure, such as performing maintenance work in a type of heavy industry (19.4%), applying pesticides (16.8%), or working for a trash or recycling company (15%). Specific hobbies with potential for chemical exposure included gardening or farming (54.9%), painting or glazing (37.3%), woodworking (20.7%), and assembling electronics (15.7%). Study participants consumed a median of 64 meals in the past year of locally caught fish from the Detroit River, with an interquartile range of 26–155 meals. Eating wild game such as deer, raccoon, rabbit, squirrel, or porcupine from the Detroit River AOC in the last 12 months was reported by only 11.5% of participants.

Table 2.

Work history of selected jobs, selected hobbies, and local wildlife and fish consumption among Detroit urban anglers study participants (n = 287).

Response (number excluded)a	Percentage, Yes
As part of a job, have you ever.…
applied pesticides that kill insects, fungus, or weeds? (1)	16.8%
worked for a trash or recycling company? (4)	15%
worked in a foundry, a smelter, a welding facility or steel mill? (2)	36.1%
removed lead paint? (14)	11.7%
worked with commercial electrical equipment such as transformers, or capacitors? (3)	10.2%
been a maintenance worker in any type of heavy industry? (4)	19.4%
worked for a battery manufacturing or recycling company? (4)	4.2%
worked for a chemical manufacturing company? (2)	9.8%
worked for an automobile manufacturing company? (2)	44.9%
In the past 12 months, hobbies and activities done by either yourself or someone inside your home.
Dyeing material (1)	4.9%
Electronic assembly (1)	15.7%
Gardening or farming (1)	54.9%
Glass crafting (3)	<4%b
Leather crafting (1)	<4%b
Metal working (1)	8.7%
Painting or glazing (3)	37.3%
Printmaking (2)	4.6%
Woodworking (2)	20.7%
Local wildlife and fish consumption
Ate locally caught wild game or birds in the past 12 months	11.5%
	Median (25th and 75th percentile)
Detroit River caught fish, meals in past 12 months (7)	64 (26, 155)

^aA response of “don’t know” or refusal and missing values.

^bPercentages that represent fewer than 10 individuals are suppressed to protect privacy.

Percentile concentrations of toxic metals, HCB, and PBB-153 for study participants and the U.S. population are shown in Table 3. Blood lead and manganese were detected in 99.3% of the study participants. Compared to the NHANES 2013–2014 survey sample results, blood lead levels of the urban anglers are 2 times higher at both the 50th and 95th percentiles (median, 95% CI: 2.0 μg/dL, 1.8–2.3 vs. 0.94 μg/dL, 0.90–0.98), and blood manganese levels are similar. The median blood lead level of the urban anglers was comparable to the NHANES 90th percentile (2.26 μg/dL, 95% CI: 2.1–2.5). The detection rates in the urban angler cohort for urinary cadmium, mercury, and arsenic were 62.3%, 51.4%, and 11.0%, respectively. Median and 95th percentile creatinine corrected urinary cadmium levels were similar or lower than the NHANES 2013–2014 levels, both in the entire cohort and by smoking status. The 95th percentiles for creatinine corrected urinary mercury and arsenic were lower in the urban angler cohort compared to the NHANES 2013–2014 sample. PBB-153 was detected in 63.9% of angler participants. The median PBB-153 in this study was 15.4 ng/g of lipid (95% CI: 13.5–18.5), which was 6 times greater than in NHANES 2003–2004 (2.50 ng/g of lipid, 2.20–2.80); the 95th percentile (62.7 ng/g of lipid, 53.2–75.2) was 1.8 times higher than the NHANES 2003–2004 results (34.6 ng/g of lipid, 12.8–66.8) (Table 3). The median serum PBB-153 level of the urban anglers was comparable to the NHANES 90th percentile (13.6 ng/g of lipid, 95% CI: 7.20–34.6). HCB was the most prominent organochlorine pesticide measured, detected in 44.0% of participants. The other organochlorine pesticides were predominantly non-detectable: mirex (89% <LOD), trans-nonachlor (83% <LOD), oxychlordane (89% <LOD), heptachlor epoxide (100% <LOD), and lindane (100% <LOD). For HCB, mirex, trans-nonachlor and oxychlordane, the geometric mean number of locally caught fish meals in the past year was not statistically different between urban anglers with measures < LOD and urban anglers with detectable levels of each chemical (data not shown).

Table 3.

Concentrations of toxic metals, hexachlorobenzene (HCB), and hexabromo-biphenyl (PBB-153) for Detroit urban angler study participants and the U.S. population.

		Detroit Urban Anglers				NHANESa
				Percentile (95% CI)			Percentile (95% CI)
Analyte, units	LODb	Sample sizec	% > LOD	Median	95th	LODc	Median	95th
Blood metals
Lead, μg/dL	0.5 μg/dL	273	99.3	2.00 (1.80–2.30)	6.10 (4.80–7.60)	0.07 μg/dL	0.94 (0.90–0.98)	3.03 (2.65–3.55)
Manganese, μg/L	4 μg/L	273	99.3	8.10 (7.70–8.50)	13.0 (12.0–15.0)	0.99 μg/L	9.22 (9.01–9.39)	16.2 (15.5–17.0)
Urinary metals, creatinine-adjusted, μg/g creatinine
Total Arsenic	6 μg/L	284	51.4	<LOD	28.1 (21.8–36.1)	0.26 μg/L	6.39 (5.66–7.29)	54.0 (47.9–66.1)
Total Mercury	0.4 μg/L	283	11.0	<LOD	0.90 (0.79–1.10)	0.13 μg/L	0.30 (0.28–0.33)	1.76 (1.50–1.88)
Cadmium, all	0.2 μg/L	284	62.3	0.22 (0.20–0.26)	0.70 (0.63–0.82)	0.036 μg/L	0.18 (0.16–0.20)	0.87 (0.75–1.0)
Non-smokers		142	59.9	<LOD	0.56 (0.17–1.12)		0.16 (0.15–0.18)	0.75 (0.67–0.84)
Smokers		142	64.8	0.29 (0.25–0.39)	0.71 (0.65–0.96)		0.26 (0.22–0.31)	1.16 (0.98–1.41)
Serum persistent organic pollutants, lipid-adjusted, ng/g of lipid
HCB	0.0625 ng/mL	275	44.0	<LOD	20.5 (17.7–24.3)	7.8 ng/g lipid	15.1 (14.5–15.9)	29.0 (25.6–33.6)
PBB-153	0.0625 ng/mL	274	63.9	15.4 (13.5–18.5)	62.7 (52.2–75.2)	0.8 ng/g lipid	2.50 (2.20–2.80)	34.6 (12.8–66.8)

^aNHANES 2013–2014 for metals and NHANES 2003–2004 (most recent available data) for the persistent organic pollutants (CDC, 2019).

^bLimits of detections (LODs) for urinary metals were based on un-adjusted concentration (ug/L) both in this study and in NHANES. For the persistent organic pollutants, LODs were based on un-adjusted concentrations (ng/mL) in this study, while NHANES LODs were based on lipid-adjusted concentration (ng/g lipid).

^cSample sizes do not equal the 287 study participants due to missing measurements and/or inadequate blood specimen.

A summary of the results for the multiple regression analyses to identify statistically significant predictors of blood lead, blood manganese, total urinary mercury, and serum PBB-153 is presented in Table 4. Blood lead levels increased with age and were higher in males, current smokers, people whose home was built before 1960, people with an annual income less than $25 K, and those with a work history of lead paint removal. The total R² for these factors was 0.28, with age, male sex and current smoking contributing to most of the explained variation. Being non-Hispanic black, male, and a current smoker were associated with lower blood manganese concentrations and together explained 10% of the variance in manganese levels. Creatinine adjusted urinary cadmium concentrations were lower in non-Hispanic black participants and those classified as overweight or obese (BMI ≥ 25). Creatinine adjusted urinary cadmium concentrations increased with age and were higher in current smokers, males and participants who worked with commercial electrical equipment, transformers, or capacitors. The total explained variation for urinary cadmium was 15.4%. Age was the most important predictor of lipid-adjusted serum PBB-153 concentrations, accounting for 19.3% of the variance. PBB-153 concentrations also increased with years living in Detroit and were higher in males, while PBB 153 concentrations were lower in participants who were overweight or obese.

Table 4.

Significant predictors of blood lead, blood manganese, urinary cadmium, and serum PBB-153 concentrations.

Biomarkera	Parameterb	Exp(β)	P-value	Partial R2
Blood lead, (μg/dL)	Age (years)	1.0166	<0.0001	0.1067
	Male	1.4390	<0.0001	0.0615
	Current smoker	1.3254	<0.0001	0.0585
	Home built before 1960	1.1821	0.02	0.02135
	Annual income < 25 K	1.1904	0.01	0.0167
	Work history of lead paint removal	1.2621	0.03	0.01261
			Total R2	0.2774
Blood manganese, (μg/L)	Non-Hispanic black	0.8664	0.0009	0.0377
	Male	0.8735	0.0035	0.0343
	Current smoker	0.9166	0.01	0.0294
			Total R2	0.1014
Urinary cadmium, (μg/L)	Age	1.0185	<0.0001	.0048
	Non-Hispanic black	0.6956	0.004	.0085
	Current smoker	1.3403	0.004	.0132
	BMI ≥ 25	0.7205	.02	.0060
	worked with commercial electrical equipment, transformers, or capacitors	1.344	.03	.0157
	Creatininea	2.4022	<0.0001	0.3955
			Total R2	0.1445
PBB-153, (ng/mL)	Age	1.0300	<0.0001	0.2025
	Years living in Detroit	1.0202	0.04	0.0545
	Male	1.2809	0.10	0.0176
	BMI ≥ 25	0.7193	0.04	0.0052
	Total lipidsa	3.2430	<0.0001	0.0542
			Total R2	0.3340

^aNatural log-transformed.

^bAge in years was modeled as a continuous variable and other parameters were modeled as dichotomous variables (1 vs 0 for otherwise).

In the regression analysis, sex was associated with blood lead and blood manganese and race/ethnicity was associated with blood manganese and urinary cadmium. Analysis restricted to only non-Hispanic Blacks and males was conducted to help understand if the combined analyses are influenced by the sparse data for other groups. When the analysis to identify predictors of blood lead was restricted to non-Hispanic blacks, the p-value for “home built before 1960” changed from p = .02 to p = .20; and, for blood manganese the p-value for current smoking changed from p = .03 to p = .12. No other changes were observed. We also examined median levels of these analytes by sex and race/ethnicity strata for indications of possible elevated levels within the demographic strata. These strata specific results are not reported due to the small sample sizes. Strata-specific analyses did not alter the results reported above. For example, the assessment of elevated blood lead holds for both males and females [males: n = 223 median = 2.20 vs NHANES median = 0.994; females: n = 50 median = 1.45 vs NHANES median = 0.746], and blood manganese concentrations were not elevated for non-Hispanic whites [non-Hispanic whites: n = 26 median = 9.3 vs NHANES median = 9.27; non-Hispanic blacks: n = 214 median = 7.80 vs NHANES median = 8.23]. Furthermore, blood lead concentrations for non-Hispanic black males in our study remained elevated when compared to NHANES 2013–2014 public use data restricted to adult non-Hispanic black males females [non-Hispanic black males: n = 176 median = 2.25 vs NHANES median = 1.15].

4. Discussion

The Detroit River, a 32-mile connecting channel linking Lake St. Clair and the upper Great Lakes to Lake Erie, drains approximately 700 square miles of land in Michigan and Ontario (EPA, 2020). Environment Canada and the U.S. EPA identified the Detroit River as one of several Great Lakes AOCs due to legacy pollutants such as lead, mercury and mercury compounds, organochlorine pesticides, PCBs, dioxins and furans (GLWQB, 1985). Studies have been conducted since the 1990s to assess these critical pollutants in Great Lakes basin residents who consume sport fish from the Great Lakes and surrounding waterways (Hovinga et al., 1993; Anderson et al., 1998; Falk et al., 1999; Dellinger, 2004; Fitzgerald et al., 2004; Bloom et al., 2005; Schantz et al., 2010; Christensen et al., 2016; Savadattia et al., 2019; Wattigney et al., 2019b). For most of the Great Lakes, significant decreasing trends from 1999 to 2014 for PCBs, DDTs, HCB, dieldrin, endrin, chlordane, oxychlordane, nonachlor, and mirex have been observed in tissue samples from top predator fish, indicating successful efforts to remove or sequester these persistent pollutants (Zhou et al., 2018). Our findings indicate lower levels of cadmium, arsenic and mercury in urine, blood lead and organochlorine pesticides than levels observed in studies of Great Lakes sport fish consumers conducted in the 1990s (Hovinga et al., 1993; Anderson et al., 1998). Despite the decreasing trends, concentrations of these legacy pollutants continue to be the predominant chemicals found in Great Lakes top predator fish, particularly PCBs, DDT and methylmercury which are the basis for limited fish consumption advisories (McGoldrick and Murphy, 2016).

As part of the BGLP program, concentrations of over 70 legacy pollutants were measured in biological specimens collected from Detroit urban angers in 2013. The current report presents results for toxic metals (other than blood mercury), organochlorine pesticides, and PBB-153 in the Detroit urban angler cohort. In this cohort only blood lead and PBB-153 were elevated relative to the general U.S. population. The toxicants most closely associated with fish consumption have been published elsewhere. In a previous publication, total blood mercury and PCBs were shown to be elevated among the Detroit urban angler study participants and were associated with higher locally caught fish consumption (Wattigney et al., 2019b).

4.1. Metals, arsenic

Arsenic, cadmium, lead, manganese, and mercury occur naturally, and trace amounts are found in air, water, rocks, and soil. These metallic elements are used in multiple industrial processes, including agricultural, medical, and technological applications (Tchounwou et al., 2012). Laboratory studies have demonstrated the high toxicity and carcinogenicity of arsenic, cadmium, lead, and mercury. These metals can induce multiple organ damage, even at lower levels of exposure, and are classified as either “known” or “probable” human carcinogens (Tchounwou et al., 2012). Our biomonitoring study included total arsenic measured in the urine which includes all species of inorganic and organic arsenic (ATSDR, 2011a). Inorganic arsenic compounds, such as copper chro-mated arsenate (CCA), are mainly used to preserve wood. Although no longer used in the U.S. for residential uses, CCA is still used in industrial applications. Organic arsenic compounds are used as pesticides, primarily on cotton fields and orchards. Fish, shellfish, and some other types of seafood can contain organic forms of arsenic which are absorbed and quickly excreted in the urine (WHO, 2001). Urinary arsenic was detected in only 11% of our urban angler participants, and the creatinine corrected 95th percentile was lower than in the general U.S. population.

4.2. Urinary mercury

Mercury in urine is typically inorganic, while mercury in blood is organic and likely to be a result of eating contaminated fish (CDC, 2017a). Urinary mercury levels can increase with more dental amalgams containing elemental mercury – i.e., silver fillings (Becker et al., 2003; Woods et al., 2007). Urine mercury data can assist in describing total mercury exposure in an urban environment. Urinary mercury was detected in 11% of our urban angler participants with a laboratory LOD of 0.4 μg/L, which is higher than the NHANES 2013–2014 50th percentile of 0.24 μg/L. Creatine corrected urinary mercury in our urban angler participants was lower than the general U.S. population at the 95th percentile. Contrary to urinary mercury which is presented in this report, a previous publication showed that the Detroit urban anglers had total blood mercury concentrations 3.2 times higher than that for the general adult U.S. population (median: 2.4 μg/L vs. 0.74 μg/L), and eating more locally caught fish was associated with higher total blood mercury concentrations (Wattigney et al., 2019b).

4.3. Cadmium

Cadmium is extracted during the production of other metals and may be emitted into the air from zinc, lead, or copper smelters (EPA, 2000). Inhalation of cigarette smoke is a leading source of cadmium exposure, and smokers tend to have urine and blood cadmium levels twice that of nonsmokers (CDC, 2017b; CDC, 2019). Non-occupational exposure to cadmium for nonsmokers is mainly through food (CDC, 2017b). Cadmium measured in urine reflects cumulative exposure and the amount of cadmium in the kidney. Urinary cadmium concentrations among our urban angler study participants were higher in smokers, as expected, and comparable to the general U.S. population even when stratifying on current smoking status.

4.4. Manganese

Manganese is used in steel production to improve hardness and strength, and welders are particularly at risk of exposure (Ellingsen et al., 2008). Overexposure to manganese through occupational or environmental sources has been associated with neurological effects such as cognitive disorders and Parkinson’s disease-like symptoms (ATSDR, 2011b; Bowler et al., 2015). The urban anglers in our study had blood manganese concentrations similar to the general U.S. population. Among our study participants, being non-Hispanic black, male, and a current smoker were associated with lower blood manganese concentrations.

There is little definitive research on the effect of smoking on manganese status. A few previous studies in adults did not find a relationship between smoking and manganese levels (Kristiansen et al., 1997; Diaz et al., 2001), while a study looking at manganese levels during pregnancy found that smokers were more likely to have low manganese levels (Takser et al., 2004). A possible explanation for lower manganese levels in smokers is that smoking increases one’s metabolic rate, which in turns increases the body’s need for manganese, thus depleting this nutrient (Seelig and Rosanoff, 2003).

4.5. Lead

A common source of lead exposure is contaminated dust and soil in and around older homes that contain lead-based paint (ATSDR, 2018). Also, old pipes may contain lead and lead soldered connections which can contaminate drinking water. Because of health concerns, the U.S. banned lead as an additive in house paint in 1978 and in gasoline in 1996 (ATSDR, 2018). In adults, high lead exposure is associated with adverse reproductive, neurological, hematological, and immunological effects, kidney damage, and high blood pressure (ATSDR, 2018). Blood lead levels in U.S. adults decreased from a median of 1.7 μg/dL (95% CI: 1.6–1.8) in NHANES 1999–2000 to a median of 0.94 μg/dL (95% CI: 0.90–0.98) in NHANES 2013–2014 (CDC, 2019).

The urban anglers in our study, conducted in 2013, had median blood lead levels twice as high (2.0 μg/dL, 95% CI: 1.8–2.3) as the general U.S. population, but well below levels of clinical concern or levels typically seen in lead-exposed workers. At the time of the study data collection, an elevated blood lead level was defined as ≥ 10 μg/dL (CDC, 2018). More recently, a blood lead level ≥5 μg/dL became the threshold for elevated levels in adults (CDC, 2021). Very few urban angler participants (number not provided due to data privacy restrictions) had a blood lead level above 10 μg/dL. We found blood lead levels to be associated with increasing age, male sex, current cigarette smoking, living in a home built before 1960, an annual income less than $25,000, and a work history of lead paint removal.

The population in Detroit, Michigan in general has a history of a higher risk of lead exposure. Recent studies have shown that atmospheric deposition of heavy metals contributes significantly to mercury and lead pollution in the Detroit, Michigan urban and industrial areas (Sherman et al., 2015). Furthermore, 93% of housing in Detroit carries a high risk of lead poisoning, with homes built before 1960 having heavily leaded paint (DHD, 2016; MDHHS, 2016). In a 2014 data report on lead testing in Michigan, children tested for lead poisoning in Detroit were twice as likely to have elevated blood lead levels than children state-wide (MDHHS, 2016). Smoking cigarettes has also been associated with lead exposure in other studies. Analysis of NHANES data, 1988–1994 and 1999–2008, showed a positive linear relation between tobacco smoke exposure and blood lead levels in vulnerable populations and the U.S. population (Mannino et al., 2005; Richter et al., 2013). Geometric mean blood lead levels were 30% higher in adults with cotinine levels that indicated tobacco exposure than in adults with no detectable cotinine (Mannino et al., 2005). NHANES data also have shown that males have higher blood lead levels than females, and blood lead levels increase monotonically with age (Mannino et al., 2005; Richter et al., 2013). We did not find an association between blood lead levels and locally caught fish consumption. An earlier study found that mean blood lead was significantly higher in Great Lakes fish eaters, however, like our study, the primary predictors of lead were smoking and other non-incidental sources of exposure (Hovinga et al., 1993).

4.6. Organochlorine pesticides

Organochlorine pesticides are among the families of priority contaminants in the Great Lakes and basins that pose concern about human exposure. The general population may be exposed to organochlorine pesticides primarily through consuming fatty foods, such as dairy products and fish or wildlife from contaminated areas (ATSDR, 2002; CDC, 2017c). Consumers of contaminated Lake Ontario sportfish species during the years 1980–1990 were found to have statistically significant higher serum mirex than non-consumers (median for non-consumers = 7.9 ng/g serum lipids vs. 18.4 ng/g serum lipids for consumers, p < .0001), while serum HBC concentrations were similar for these two groups (Bloom et al., 2005). Studies found that HCB and mirex concentrations were higher in Lake Ontario fish consumers (Kearney et al., 1999; Kostyniak et al., 1999; Fitzgerald et al., 2001) despite a documented declining trend in these pollutant levels that began in the 1970s (Schmitt et al., 1999).

In the current Detroit urban angler cohort, we measured HCB, mirex, trans-nonachlor, oxychlordane, heptachlor epoxide, and lindane which are organochlorine pesticides historically of concern in the Great Lakes. Concentrations of these organochlorine pesticides in our urban angler participants had low detection rates and were lower than the general U. S. population at the 95th percentile. Our study did not find an association between HCB, mirex, trans-nonachlor and oxychlordane and local fish consumption, based on urban anglers with measures < LOD compared to urban anglers with detectable levels of each chemical. Our findings are consistent with fish monitoring data which indicate that these contaminants have decreased over the past several decades, particularly in Lake Erie, which is one of the most urban influenced Great Lakes due to the proximity of Toledo, Cleveland, and Detroit (Zhou et al., 2018). Furthermore, the organochlorine pesticides in our report have lower bioaccumulation potential than PCBs and DDT (Zhou et al., 2018).

4.7. PBB 153

PBB 153 is a flame retardant phased out of use in the U.S. in the 1970s following accidental PBB contamination of manufactured livestock feed, particularly in the 1973 historical widespread incident in Michigan (Kay, 1977; Wolff at al., 1982; Fries, 1985). Following the 1973 PBB incident, the Michigan Department of Health established a registry of over 6000 residents to study the long-term health effects of exposure to PBB (Landrigan et al., 1979). Studies of these registry participants have shown evidence of altered pubertal development (Doufas and Mastorakos, 2000), increased risk of miscarriage (Small et al., 2011), and thyroid dysfunction (Jacobson et al., 2017). PBB-153 was detected in 63.9% of our urban angler study participants and at levels significantly higher than the U.S. population, with the study median 6 times higher than that of NHANES 2003–2004. The elevated PBB-153 exposure in our study cohort indicates the continued effect of the 1973 incident among Michigan’s lower peninsula population. Age followed by years living in Detroit, Michigan (our study area) were important predictors of lipid-adjusted serum PBB-153 concentrations. Our study results are consistent with other studies of PBB-153. A biomonitoring study of contaminants in Great Lakes sport fish consumers found that Michigan residents had higher mean serum PBB-153 concentrations than participants from other Great Lakes states (Anderson et al., 1998). In general, concentrations of PBB 153 have been found to increase with age as with other chlorinated persistent organic pollutants that were phased out in the 1970s (Sjodin et al., 2008). Older people have a longer period of possible exposure, and at a time when PBBs were at higher levels in the environment and food chain, than younger people. PBB-153 concentrations among the Detroit urban anglers may reflect past and not ongoing exposure.

4.8. Limitations

We acknowledge limitations to our study. First, non-response bias could have been introduced based on the large number of anglers interviewed at fishing venues who were not successfully recruited (64.9%). Although a modified venue-based sampling strategy was employed, final recruitment included only shoreline anglers who expressed interested in participating in the biomonitoring clinic. As such, participants may not be statistically representative of the population of urban, shoreline anglers in Detroit. Compared to non-participants, anglers who completed the full study were older in age and more identified as non-Hispanic black (Wattigney et al., 2019a). Second, a cross-sectional study design was used to ascertain concentrations of select heavy metals and persistent pollutants among Detroit urban anglers. While this design is appropriate for estimating exposure to these chemicals at a given point in time, limitations exist in drawing inferences for chemicals that have short biological half-lives. Moreover, biological samples were collected in 2013 and may not reflect current exposure. In addition, associations measured in cross-sectional studies can also be biased due to lack of well-defined temporality. Third, data on fish consumption was collected by species and location and may be under- or over-estimated due to error in recall, potentially contributing to differential measurement error by biomarker level. Fourth, sparse data for females and race/ethnicity other than non-Hispanic blacks limits analysis stratified by sex or race/ethnicity. As such, meaningful sex or race/ethnicity differences may be masked in our results. Another limitation is that NHANES 2003–2004 data were used as the U.S. general population reference for the organochlorine pesticides for our urban angler study a decade later. A declining trend has been noted in these pollutants. NHANES 2003–2004 was the last survey cycle for which these analyte measurements were made on individual serum samples. Furthermore, the MDHHS laboratory LODs for the urban angler study measurements were often higher than the referent NHANES median leading to a high rate of ‘non-detects’ in our study. As such, we could only assess exposure for selected contaminants with sufficient detection rate. Moreover, study contaminant concentrations 1.8 or 2 times higher compared to NHANES data are not meaningful in the context of health effects. Lastly, while beyond the data collected for our study, air pollution data in the Detroit urban area as a source of exposure could allow comparisons to exposure from fish consumption and or exposure to local river water.

5. Conclusions

Blood lead levels were somewhat elevated, although not at levels of clinical concern, among the urban angler study participants. Blood lead was associated with lifestyle characteristics such as low-income, older homes and smoking, not local fish consumption. Residents of Detroit in general have been noted to have a greater risk of exposure to lead. PBB 153 levels were also higher in this study than in the U.S population. Evidence of PBB exposure in our study cohort reflects among others the continued effect of the widespread contamination of livestock feed in 1973 among Michigan’s lower peninsula population. Profiles of other study contaminants did not indicate that these priority pollutants are higher among the Detroit urban angler study participants than background levels found in the U.S. population.

Abbreviations:

ATSDR

Agency for Toxic Substances and Disease Registry

AOC

Area of Concern

BGLP

Biomonitoring of Great Lakes Population

BMI

Body Mass Index

DDE

Dichlorodiphenyldichloroethylene

DRC-ICP-MS

Dynamic reaction cell Inductively coupled plasma mass spectrometry

GC/ECD

Electron capture detection

GLRI

Great Lakes Restoration Initiative

HCB

Hexachlorobenzene

ICP-MS

Inductively coupled plasma mass spectrometry

LOD

Limits of detection

MDHHS

Michigan Department of Health and Human Services

NHANES

National Health and Nutrition Examination Survey

PBBs

Polybrominated biphenyls

PBB 153

Polybrominated biphenyl 153

PCBs

Polychlorinated biphenyls

TEQ

Total toxic equivalency

EPA

U.S. Environmental Protection Agency