Each cooperative agreement program set recruitment goals and target population(s) as part of their funding proposal. Each program aimed to assess exposure to toxic substances in residents within their jurisdiction conducting population-based biomonitoring of susceptible population(s). All programs elected to target adults who had lived in a selected Great Lakes AOC for at least one year and were medically able to donate the required blood sample. For the proposed sample sizes, we assessed the ability to detect analyte levels above the most recent geometric mean (GM) reported for the adult U.S. population (NCEH, 2015) assuming 80% power and a one-sided test of significance at p = 0.05. Standard deviations for the U.S. population GMs were assumed the same for the BGLP-I program target populations. For core contaminants, sample sizes were sufficient to detect analyte levels 2-fold or higher than U.S. population levels. Recruitment goals and criteria are summarized in Supplemental Table 1.

Questionnaire interviews were conducted to collect information on respondent characteristics and possible sources of exposure to the environmental contaminants of interest. ATSDR and the state programs collaboratively developed survey questionnaires through a process of adopting questions from established surveys, reviewing questions/response options, and revising items according to group consensus. Demographic information such as race and ethnicity adhered to federal standards as required by the Office of Management and Budget (OMB). The questionnaires included a set of core sections devoted to demographic information, residential history, housing characteristics, job history, smoking history, dietary intake, and recreational water activities in contaminated areas. Much of the interview was devoted to questions about consumption of locally caught fish. Participants were asked how many years they have eaten fish of any type caught from local waters, to identify the location and species of locally caught fish eaten, and then asked to recall the number of times each species was eaten in the past 12 months. Similar questions were asked for wildlife, including birds and other animals that were consumed in the past 12 months. Visual aids were used to assist in referral to specific waterways, fish species, wild game, and fish portion size. The interview also collected information on the frequency of commercially purchased fish eaten in the past 12 months for specific species or species groupings. In addition to core questionnaire items, the biomonitoring interview for each state program was tailored to fit local concerns and cultural practices, and designed to assist in the interpretation of contaminant levels in the target population. For example, traditionally harvested wild plants such as locally cultivated wild rice were included in the FDL Community study interview; and, questions on fish consumption were framed in the context of traditional methods of catch and by season. Another example is inclusion of culturally relevant questions for the Burmese population in Buffalo, NY. For the Burmese, consumption of homemade or commercially prepared fish paste and the use of thanaka, a cosmetic made from ground bark, were of interest. State programs identified an advisory board or Advice Council (for FDL study) of local stakeholders and subject matter experts at the onset of program planning.

For all BGLP-I biomonitoring clinics, study participants provided written informed consent, responded to a face-to-face standardized questionnaire interview, provided blood and urine samples, and had their height, weight, and blood pressure measured. Phlebotomists screened each subject to determine if they were medically able to give blood. For Michigan and New York State, eligibility to provide the blood sample was based on the Red Cross criteria — i.e., blood pressure greater than 80 systolic/50 diastolic and less than 180 systolic/100 diastolic, no blood clotting disorders, and weight of at least 110 lbs. (American Red Cross). Laboratory standard operating procedures determined the aggregate minimum blood volume specified by each program. At the end of the clinic visit, participants received remuneration for their time and effort. Area-specific fish advisory brochures and “eat safe fish” guidelines were distributed to participants at the end of the clinic visit and/or in participant results mailings (MDHHS, 2018a; MDHHS, 2018b; NYSDOH, 2017).

All state programs were required to measure a core set of pollutants including lead, mercury, PCBs, and several organochlorine pesticides (OCPs) that included: hexachlorobenzene, mirex, DDT, and dichlorodiphenyldichloroethylene (DDE). Programs were also encouraged to measure additional pollutants of regional or local concern such as dioxins, furans, speciated mercury, toxaphene, PBDEs, and PFAS. While each state program has its own designated laboratories to conduct measurements, ATSDR implemented specific steps to ensure consistency across laboratories. In accordance with ATSDR program requirements, the designated laboratories provided documentation of Clinical Laboratory Improvement Act of 1988 (CLIA) certification, participated in appropriate external proficiency testing programs, provided proficiency testing results, and provided standard operation procedure for all analytical methods including detailed quality control and quality assurance procedures.

Table 1 presents an overview of chemical analytes measured and biological matrices. For the MDHHS shoreline anglers study, the MDHHS Analytical Chemistry Section performed laboratory analysis of all analytes except dioxins, furans, co-planar PCBs, total cholesterol and triglycerides, which were measured by the Centers for Disease Control and Prevention (CDC) National Center for Environmental Health laboratories. Additionally, creatinine was measured at Sparrow Laboratories in Lansing, Michigan. For the FDL Community study, the MDH Public Health Laboratory measured mercury, lead, cadmium, PFAS, bisphenol A, triclosan, and 1-hydroxypyrene in biological specimens; selenium in urine was measured by the Colorado Public Health Laboratory; and PCBs, OCPs, and toxaphene were measured by the MDHHS Analytical Chemistry Section. Saturated and unsaturated fatty acids in blood plasma, additional biomarkers of fish consumption, were measured by the MDH Public Health Laboratory. Total cholesterol, hemoglobin A1C, creatinine and cotinine were also measured. Hemoglobin A1C and total cholesterol were measured as a public health service to participants at the request of the FDL Biomonitoring Study Advice Council; the results were not kept by MDH or ATSDR. For the New York State program, the chemicals measured and laboratory analysis methods were the same for both the licensed anglers and Burmese populations. In addition to the PCBs, OCP and metals measured by each state program, the NYSDOH program measured predominant PBDE congeners and PFAS. PBDEs are emerging chemicals of concern in the Great Lakes basin and Lake Ontario (Klecka et al., 2010). Lake Ontario, which contains several NYS's AOCs, has some of the highest levels of PBDEs reported in the US Great Lakes (Venier et al., 2014). PFAS measured in sediment samples from the Niagara River over the period 1980–2002 showed an increasing trend (EPA, 2007). All analytical tests were conducted at the Wadsworth Center, NYSDOH.

WSU interview teams made 439 VDT visits at 19 fishing venues along the Detroit River over 35 days, and 306 VDT visits at 11 fishing venues along the Saginaw River/Bay over 21 days. Of the 2660 Detroit River anglers interviewed, 1261 were not eligible, 582 were eligible but refused further participation, and 817 were eligible and willing to participate. Of the 843 anglers on the Saginaw River/Bay, 113 were eligible and willing to participate, 104 were eligible but refused further participation, and 626 were ineligible. The MDHHS aimed to collect biomonitoring study data from 200 anglers from each area for 400 study participants. While the target number of participants was met for the Detroit River area (n = 287), only 38 anglers from the Saginaw River/Bay area participated. Recruitment was limited to one fishing season, and severe spring flooding during the VDT screening phase that year limited the number of participants screened in the Saginaw River/Bay area. Furthermore, the majority of anglers in the Saginaw River/Bay area who were screened reported eating less than the required minimum of two meals per month of locally caught fish. Study results will only be presented for the Detroit River area participants. The recruitment process and outcome for the Detroit urban angler study is illustrated in Fig. 2.

The Detroit River angler participants were primarily male (81%), 40–59 years old (58%), African American of non-Hispanic origin (80%), and lived in the Detroit area for ≥ 21 years (86%) (Table 2). Non-response bias assessment based on venue-based screening questionnaire data was conducted for the Detroit River angler cohort at each of the two sampling phases. During the VDT enumeration phase, the 818 individuals who were eligible and willing to continue in the study were compared to the 582 individuals who were eligible but refused further participation. Eligible anglers willing to participate were significantly more likely to have been eating fish from the Detroit River for “more than 10 years,” 70% versus 62%. Demographic characteristics were similar between these two groups. In the second sampling phase non-response bias assessment, characteristics for the 287 anglers who completed the full biomonitoring study were compared to the 510 individuals who did not participate. Demographic differences were not noted between the two groups, but participants were more likely to be aware that some fish have harmful chemicals in them (91% versus 81%).

The FDL HSD service area covers a large land base, including multiple towns and cities within its boundaries (Supplemental Figure 2). The sampling frame for the FDL Community Biomonitoring Study started with a pool of 4732 potentially eligible participants included on the FDL HSD Client List. A random sample of 1421 people were selected from this pool and invited to participate in the study. Subsequently, nine people were identified as deceased and 69 had invalid contact information that couldn't be successfully updated resulting in a total of 1343 invitees. During recruitment, a total of 60 individuals were determined to be ineligible, and 278 invitees were either unable or unwilling to schedule a biomonitoring clinic appointment (refusals). An additional 514 couldn't be contacted by either mail or phone. Data collection occurred from January 2013 through October 2013 at two FDL clinic sites: The Min No Aya Win Human Services Center on the FDL reservation and the Center for American Indian Resources in Duluth. The target number of participants was 500; however, the FDL HSD clinic contract expired just short of this goal. Study staff successfully enrolled 491 participants (Fig. 3).

FDL Community Biomonitoring Study participants were primarily female (57%); 37% aged 18–39 years, 42% aged 40–59 years, and 21% aged 60 years or older; and 83% had lived in the study area for 21 years or longer (Table 2). To provide an overview of the representativeness of the study sample, the age/gender distribution of study participants was compared to the total pool of 4732 potentially eligible participants. The FDL Community Biomonitoring study sample tended to underrepresent the total population until the late 30's for women and late 40's for men, had more people of both genders after that until the late 70s, at which point men were again underrepresented.

The enrollment goal for the licensed angler target population in NYSDOH's Healthy Fishing Communities Project was 400 licensed anglers. A total of 409 licensed anglers participated in the biomonitoring study clinics that were held at local health departments from February through October 2013 (Fig. 4). Participants were mostly non-Hispanic whites (83%), males (86%), aged 40–59 years (48%), and had lived in the area for 21 years or longer (90%) (Table 2). Non-response bias was assessed among the 883 respondents to the initial screening survey. Characteristics for the 409 anglers who completed the full biomonitoring study were compared to the 474 individuals who did not participate. Participants were more likely to be 60 years of age or older (33%) compared to nonparticipants (23%). No gender differences were noted between these two groups.

To facilitate NYSDOH's study of Burmese immigrants, the program worked closely with the JRCHC, a faith-based organization dedicated to facilitating wellness and self-sufficiency in the lives of refugee and low-income community members in Buffalo, NY. This collaboration was critical for refining the questionnaire interviews, identifying seeds, and employing interpreters for the project. The enrollment goal of 200 Burmese immigrants was quickly met using RDS with 209 participants (Liu et al., 2018). Clinics were held at the JRCHC one evening per week from July through October 2013 and served as recruitment, screening and data collection venues. The Burmese participants included slightly more females (60%) than males and 55% were aged 18–39 years, and 52% had lived in Buffalo for at least 4 years with only 18% having lived elsewhere in the United States (Table 2).

From inception of the FDL Community Biomonitoring Study, the MDH-led initiative recognized FDL's sovereign status – a key element for conducting culturally competent projects with American Indian tribes. Staff from MDH and FDL developed a collaborative partnership through a series of educational and administrative exchanges to design and conduct the study. The majority of study activities (community outreach, participant recruiting, interviews, and specimen collection) took place at the Min No Aya Win Clinic by staff hired by the FDL Band from within the FDL Community. The FDL Biomonitoring Study Advice Council, comprised mainly of tribal members, reviewed all proposed plans, draft materials and procedures, and offered guidance on topics related to cultural appropriateness, community collaboration and involvement, results communication, and protections for participants. Recommendations from the FDL Biomonitoring Study Advice Council were incorporated into the study design wherever feasible.

The need for conducting a biomonitoring study with a potentially susceptible population living within the SLRAOC was the key factor that led MDH to initiate the partnership with the FDL Band. Study results established baseline measurements of exposure to environmental contaminants which has laid the groundwork permitting later study, if the tribe should choose, to access time trends and changes affecting the FDL Community. Furthermore, MDH collaborated with FDL to implement a broader array of culturally congruent public health activities related to Great Lakes contaminants of concern and safe fish consumption. The relationship between MDH and FDL has been characterized by mutual trust, respect, and commitment. As a result, this MDH-led tribal community biomonitoring study serves as an example of how state agencies can successfully collaborate with sovereign tribes.

The BGLP-I program has several limitations. First, the program aimed to assess contaminant exposure in target Great Lakes populations deemed to be at high risk of exposure and did not collect health outcome information. Hence, we cannot identify potential adverse health outcomes associated with chemical exposure. Second, the program consists of diverse target populations and certain elements of each state program's study designs were specific to the respective populations. As an applied public health program, it was not the intent to directly compare analyte levels across target populations nor to combine the information collected as part of the three separate programs. We also acknowledge that the study objectives could have been strengthened had it been feasible to include non-contaminated reference sites to compare the results for each target population as opposed to using the general U.S. population for comparisons. Third, the program bears the same limitations inherent to biomonitoring, such as inability to distinguish sources and route of exposure based purely on biomonitoring results. Nevertheless, questionnaire results are expected to provide information about exposure sources (e.g., eating locally caught fish) within the target communities, which allows the study of association between biomonitoring results and specific sources of interest, and guide public health actions by state health departments to protect these communities within their jurisdiction.

A few state-specific limitations are noted. The NYSDOH licensed angler cohort included sport anglers who tend to practice catch-and-release, thus reducing their chances for dietary exposure. In addition, severe spring flooding in Michigan's Saginaw River/Bay area during the study period limited fishing events. The study period included only one fishing season and recruitment in the Saginaw AOC resulted in only 38 participants, which is not a large enough sample size to support generalizable results.