Demographic Profiles, Mercury, Selenium, and ...

J Community Health DOI 10.1007/s10900-015-0082-5

ORIGINAL PAPER

Demographic Profiles, Mercury, Selenium, and Omega-3 Fatty Acids in Avid Seafood Consumers on Long Island, NY Rebecca Monastero1

•

Roxanne Karimi2 • Susan Silbernagel3 • Jaymie Meliker4

Ó Springer Science+Business Media New York 2015

Abstract Seafood consumption is known to confer nutritional benefits and risks from contaminant exposure. Avid seafood consumers are neither well-characterized with regard to their demographic profile nor their underlying risk–benefit profile. Contaminants [e.g., mercury (Hg)] and nutrients [e.g., selenium (Se), omega-3 fatty acids] are prevalent in some seafood. Participants (N = 285) recruited on Long Island, NY, completed food frequency and health questionnaires and received blood draws analyzed for Hg, omega-3s, and Se. Participants were categorized based on frequency and type of seafood consumption. Logistic regression analyses evaluated relationships between seafood consumption and demographics, and were age- and sex-adjusted. t tests assessed relationships between seafood consumption patterns and biomarkers Hg, omega-3s, and Se. Consumption of both tuna and salmon was

& Rebecca Monastero [email protected] Roxanne Karimi [email protected] Susan Silbernagel [email protected] Jaymie Meliker [email protected] 1

Undergraduate Studies, Stony Brook University, Stony Brook, NY 11794, USA

2

School of Marine and Atmospheric Sciences, Stony Brook University, Stony Brook, NY 11794, USA

3

Division of Allergy and Infectious Disease, Department of Medicine, University of Washington, Seattle, WA 98195, USA

4

Department of Preventive Medicine and Program in Public Health, Stony Brook University, 071 Health Sciences Center, Stony Brook, NY 11794, USA

associated with older age: those aged 55–75 and over 75 years old were more likely than participants aged 18–34 to eat tuna and salmon (OR 2.27; 95 % CI 1.05, 4.89 and OR 3.67; 95 % CI 1.20, 11.20, respectively). Males were less likely than females to eat fish other than tuna or salmon (OR 0.58; 95 % CI 0.34, 0.97). Caucasians were more likely to consume tuna (OR 0.31; 95 % CI 0.10, 0.96) or salmon and tuna (OR 0.34; 95 % CI 0.12, 0.91), while non-Caucasians were more likely to consume other fish types (OR 2.73; 95 % CI 1.45, 5.12). Total blood Hg was associated with weekly consumption of any type of fish (p = 0.01) and with salmon and tuna consumption (p = 0.01). Salmon was associated with plasma omega-3s (p = 0.01). Se was not associated with fish intake categories. Risk communicators can use these findings to influence seafood preferences of different demographic groups. Keywords Seafood consumption
Mercury
Omega-3 fatty acids
Demographic
Selenium Abbreviations Hg Mercury omega-3s Omega-3 fatty acids RfD Reference dose USDA U.S. Department of Agriculture USEPA U.S. Environmental Protection Agency FFQ Food Frequency Questionnaire OR Odds ratio CI Confidence interval

Introduction Controversy exists regarding dietary seafood recommendations as a result of the conflicting risk of organic mercury (Hg) exposure and benefits thought largely to be due to

123

J Community Health

abundant omega-3 fatty acids as well as selenium [1–4]. Consumption of different amounts and types of seafood confers different risks and benefits, and there is a need for a more thorough understanding of seafood intake within and among different populations in order to identify those at high risk [1]; however, the demographic profiles of individuals with different seafood consumption habits are not well understood. There are many examples of studies designed to evaluate relationships between demographic characteristics and non-seafood dietary preferences [5–7]. These studies characterize consumers based on their food choices and aid in improving risk communication and health interventions, e.g., to improve vegetable and fruit intake for a particular population. Fewer studies focus specifically on correlations between demographic characteristics and seafood consumption, particularly correlations between demographic and consumption of specific fish types. A 1994 study examined the correlations between demographics and general seafood consumption in a Floridian population; however, it did not correlate demographics with consumption of particular seafood types [8]. Similarly, Sechena et al. [9] studied the differences in general seafood consumption among ten ethnic groups of second-generation Asian and Pacific Islanders, though the study did not include correlations of consumption of individual fish types among the demographic groups. A recent study examined effects of demographics and perception of seafood health and safety in relation to consumption, finding that less educated individuals, females, and high-end consumers placed more importance in explicit seafood labeling, which can be useful for communication of seafood nutrition and safety information [10]. A 2012 study observed the effect of demographic variables on fish consumption among men and women, and demonstrated the importance of controlling for demographics when predicting disease outcome [11]. Similarly, a recent study of the Portuguese population demonstrated variability of seafood consumption profiles based upon demographics, finding seafood consumption differences between men and women as well as among consumers from different geographic areas [12]. While informative, such studies of seafood consumer preferences and demographics would be improved by considering detailed characterization of seafood species that are eaten, as well as biomarkers for helping characterize the risk– benefit profile of seafood types. Biomarkers of Hg, Se, and omega-3s help to capture how demographic characteristics relate to the risk–benefit profile of each seafood species. Throughout the United States, total blood Hg concentrations based on seafood consumption patterns vary demographically and regionally, with the Northeast population having among the highest concentrations in the US [13]. In a 2013 study of the New

123

York population, it was determined that seafood consumption was a major source of organic Hg exposure for 99 % of individuals in the State Heavy Metals registry showing a range of total blood Hg from 5 to 760 lg/L) [14]. Another study found that almost 25 % of New York City adults and almost 50 % of Asian residents of New York City had total blood Hg levels at or above the USEPA RfD [15]. Regionally focused studies such as these suggest that a considerable proportion of the nation’s Hg-related exposures are observed in the northeastern U.S. among different demographic groups. These studies also demonstrate that organic Hg exposure is typically dietary, and is most often due to high seafood consumption [14, 15]. Additional studies reporting seafood consumption as the primary source of organic Hg exposure, particularly focusing on demographics, include: a 2013 study in Duval County, Florida [16], which found that women of Asian/ Pacific Islander origin were at greatest risk for organic Hg exposure due to seafood consumption; and a 2012 study in Brooklyn, New York [17], which found that umbilical cord total blood Hg in a Caribbean immigrant population was predicted by the mother’s foreign birth and her seafood consumption. However, neither study detailed relationships between organic Hg levels and individual fish species or demographics and individual fish species. Seafood consumption is also known to be one of the primary sources of omega-3 fatty acids and Se [1, 18]. Omega-3 fatty acids bring benefits against cardiovascular disease [19], and have a clear association with specific types of seafood [20]. Se is an essential nutrient that is a known antioxidant and provides protection against cardiovascular disease and thyroid disease [18]. Average serum concentrations of Se in populations vary geographically within the US, though all regions have levels well above the RfD [21]. Se concentrations also vary within individual fish types, with the highest levels found in tuna and shrimp [22]. Few studies have reported relationships between blood Se and consumption of individual fish types. However, studies have found correlations between Se levels within the US and demographics: older consumers, males, and African American consumers have higher levels of Se in comparison to younger consumers, females, and non-Hispanic Caucasian or Hispanic consumers [23]. This correlation suggests the importance of including demographics in a study of seafood consumption to best evaluate groups at risk for lower Se levels, as these demographic differences in Se levels may be due to varying seafood intakes among demographic groups. The primary objectives of this study were to: (1) investigate demographic characteristics in relation to specific seafood species in a cohort of avid seafood consumers on Long Island, NY, USA; and (2) use biomarkers of total blood Hg, Se, and omega-3s to help characterize

J Community Health

the risk–benefit profile of different types of seafood eaters. These objectives are intended to facilitate subsequent risk communication efforts by identifying specific demographic groups at risk of elevated total blood Hg exposure and reduced omega-3 and Se intake.

themselves or by someone they know, along with questions about potential inorganic Hg exposures due to lifestyle factors (e.g., dental amalgams, skin lightening creams, or occupational exposures). Blood Hg, Se, and Omega-3 Collection and Analysis

Materials and Methods Participant Recruitment Process The study was approved by Stony Brook University’s Committee on Research Involving Human Research Subjects (CORIHS) Institutional Review Board (IRB # 2010-1179). We recruited 290 adult, avid seafood consumers from the general Long Island population in 2011–2012. In-person recruitment efforts occurred at seafood-serving establishments, libraries, gyms, university bulletin boards, fishing piers, and seafood markets; flyers and advertisements targeting avid seafood consumers were also posted at these locations. Interested participants were informed that they were being recruited for a study to investigate the benefits and risks of seafood consumption. An article about the study ran in Newsday, the main newspaper on Long Island; three advertisements also ran in Newsday. Interested participants filled out an online survey regarding frequency and type of seafood consumption, which was used to estimate their total blood Hg concentrations using total seafood Hg concentrations reported in the Seafood Hg Database presented in Karimi et al. [24]. Individual eligibility was established through estimated total blood Hg levels above the USEPA RfD of 0.1 lg kg-1 day-1 corresponding to a total blood Hg concentration of 5.8 lg L-1 to ensure power for subsequent analyses [25]. Of the 996 individuals who completed the screening questionnaire, 746 individuals were eligible. Of those who were eligible, 290 enrolled in the study.

Participants also scheduled an appointment at the Clinical Research Core at Stony Brook University Medical Center for blood to be drawn. Blood samples were collected from 285 participants; there were no blood samples for the remaining five participants due to complications with the blood draw. Sample digestion, quality control, and quality assurance measures were taken for Hg and Se digestion and analysis as reported elsewhere [27, 28]. In brief, whole blood specimens for total serum Hg and Se analysis were collected in trace element collection tubes (BD medical), stored at 4 °C, and subsequently sent to be analyzed using ICP-MS (Thermo X-Series II) at RTI International’s Trace Inorganics Laboratory (Research Triangle Park, NC). Hg and Se concentrations for 2 sample blanks and 4 method blanks per batch were negligible, confirming no background contamination. Detection limits ranged from 0.10 to 0.70 lg Hg L blood-1 and 2 to 8 lg Se L blood-1. Samples that were below the detection limit (n = 2 for Hg) were assigned a value of one-half the detection limit for that batch. We collected fasting blood specimens for fatty acid analysis in 3 mL vacutainer tubes with K2EDTA (BD Medical) [27, 28]. Blood was centrifuged within 45 min of collection, and an aliquot of plasma was removed and stored at -80 °C. Plasma specimens were sent to Lipid Analytical Lab, (University of Guelph, Canada), for analysis. We analyzed fatty acid concentrations in the phospholipid fraction as described previously [27], which is correlated with fatty acid intake estimated by a FFQ over 1 year [29]. Data and Statistical Analysis

Development and Distribution of Questionnaires Participants filled out: (1) a semi-quantitative Block Food Frequency Questionnaire (FFQ) which collected information about frequency, quantity, and types of food consumed [26]; (2) a semi-quantitative supplement to the FFQ consisting of 71 questions which collected information on frequency and daily quantity of seafood consumption; and (3) a health questionnaire adapted from the CDC Behavioral Risk Factor Surveillance System which was designed to obtain information regarding health endpoints and demographics. The questionnaire also included information on whether or not participants ever ate fish caught by

We generated variables for each of six demographic factors obtained from the health questionnaire, which are shown in Table 1. We calculated percentages of the study’s convenience samples by each of the six demographic groups to evaluate similarities between study participants and Suffolk County residents (Table 1). Demographic data regarding the Suffolk County Population was collected from the 2007–2011 census data [30–32]. For coding seafood consumption, we used a cutoff frequency of at least once per week (corresponding to a designated frequency value of 5) for each fish type to

123

J Community Health

enhance the sample size in the different demographic categories. Five discrete groupings of consumers were generated, membership of which was exclusive; i.e., each study participant was assigned to a specific, single group with no overlap. We selected the following seafood consumer variables: predominantly tuna weekly or greater, predominantly salmon weekly or greater, predominantly tuna and salmon weekly or greater, other seafood consumed weekly or greater, and a group where seafood was not eaten at least once per week. We generated a binary ‘‘any fish’’ variable as well, indicating whether or not a participant consumed any type of fish once per week or greater. We divided the total blood Hg variable a priori into four categories based on previously established cut-offs: \5.8 (reference), 5.8–10, 10–15, and [15 lg/L. We used tertiles for categorizing Se (\265 (reference), 265–346, [346 ng/ mL) and omega-3s (\5.29 (reference), 5.29–6.65,[6.65 % fatty acids) [25]. To investigate the association between demographic factors and fish consumption patterns, we performed logistic regression analyses, with a binary outcome of either belonging or not belonging to one of the previously described fish consumption groups (predominantly tuna, predominantly salmon, predominantly tuna and salmon, predominantly other fish, no predominant fish species consumption, and weekly consumption of any fish species). Age and gender are adjusted for in every model, along with one other demographic factor. To investigate the relationship between biomarker levels and fish consumption patterns, we performed a series of t-tests comparing membership in one of the six fish consumption groups with levels of each of three blood biomarkers: Hg, omega-3s, and Se. All analyses were run using IBM SPSS Statistics (V. 21).

Caucasians, though the study had approximately 7 % more Asian participants and the county population had approximately 12 % more Hispanics. Frequencies of demographic and biomarker variables among different categories of seafood consumers are shown in Table 2, and regression analyses predictive of membership in the different seafood consumer categories are reported in Table 3. In a logistic regression model including age and gender, older age predicted weekly consumption of both tuna and salmon. The odds of consuming both tuna and salmon increased twofold in the 55–74-year-old age group, and increased threefold to fourfold in the over 75-year-old age group. Males were nearly half as likely as females to consume ‘‘other’’ (nonsalmon/tuna) fish types weekly. Caucasians were about three times more likely to consume tuna and tuna/salmon, and nearly three times less likely to consume other types of non-salmon/tuna fish than non-Caucasians. Similar to the trends observed in Table 2, in t test comparisons between biomarker levels and fish consumption patterns, blood Hg levels of consumers who ate any type of fish per week were higher than those of participants who consumed fish less than weekly (p = 0.01). Similarly, consumers who ate tuna and salmon every week had higher Hg levels compared to those who did not consume these fish types weekly (p = 0.01). Blood omega-3 levels were associated with weekly salmon consumption (p = 0.01), and borderline associated with weekly consumption of any fish type (p = 0.06) and weekly consumption of both salmon and tuna (p = 0.10). Blood Se showed a borderline association with weekly consumption of any seafood type (p = 0.06), driven by weekly consumption of fish other than tuna or salmon (p = 0.09). No other groupings of seafood consumers were associated with these biomarkers.

Discussion Results In comparing the study population with the demographic characteristics of the county, a few factors were different in this population of avid seafood eaters (Table 1). The county had a larger percentage of individuals aged 35–54, while the study had a larger percentage aged 55–74. Proportionally more women were in the study population than in the county. Levels of education attained also differed, as the study sample consisted of approximately 40 % more college-educated participants than the Suffolk population. The characteristics ‘‘yearly household income’’ and ‘‘born in the US’’ were similar in percentage frequencies between Suffolk County and the study’s sample population. Ethnicity was somewhat similar between both groups, with approximately 75 % of both groups consisting of

123

Given high levels of both contaminants and nutrients in seafood, identifying characteristics of different types of seafood consumers is important for risk communication efforts. Earlier work in this arena by McKelvey et al. [15, 33, 34], Tsuchiya et al. [35, 36], and Geer et al. [17] shed light on relationships between demographic groups and total blood Hg levels, or Hg levels and seafood consumption, and thus conveyed the importance of populationbased biomonitoring, particularly among seafood-consuming populations. However, these studies did not have sufficient power to relate demographics to consumption of specific seafood species. Our study uncovered several demographic characteristics associated with different patterns of seafood intake which we will discuss below. In addition, one of the study’s main strengths is our blood

J Community Health Table 1 Demographics of study participants in comparison with suffolk county residents

Demographics

Long Island study of seafood consumption

Suffolk county

N

%

N

Female

171

59.0

755,674

50.8

Male

119

41.0

731,503

49.2

15–34

84

29.2

361,179

30.2

35–54

76

26.4

461,098

38.6

55–74

108

37.5

280,168

23.4

20

6.9

92,848

7.8

African American

10

3.5

103,638

7.0

Asian American Hispanic/Latino

29 13

10.0 4.5

50,752 237,061

3.4 15.9

White/Caucasian

228

78.6

1,074,429

72.3

10

3.5

21,297

1.4

%

Gender

Age (years)a

[75 Ethnicity

Other/[1 ethnicity Born in United States Yes

246

84.8

1,254,570

84.4

No

44

15.2

232,607

15.6

Yearly household incomeb \$25,000 $25,000–$200,000

48

17.1

55,988

11.3

213

76.1

391,870

78.9

19

6.8

48,909

9.8 4.6

[$200,000

Education level among those Cage 25c

a

Below Grade 9

0

0.0

45,466

Grades 9–11

2

0.8

59,863

6.0

High School Graduate

15

6.3

302,697

30.3

College 1–3 years

48

20.2

268,172

26.8

College C 4 years

173

72.7

323,578

32.4

Study survey data begins age 18. Census data begins age 15

b

Three common income brackets among datasets were established for comparison; the five income brackets established within the study survey were used in subsequent analyses (\$25,000 (reference); $25,000–70,000; $70,000–110,000; $110,000–200,000; [$200,000) c Census data begins education at age 25; survey data were accordingly stratified

biomarker measures of total Hg, Se, and omega-3s, allowing us to shed light on the demographics of specific groups of seafood consumers with higher concentrations of blood Hg, Se, and omega-3s [12, 34]. Participants over the age of 75 were 3–4 times more likely to consume both tuna and salmon weekly than participants aged 18–34, and participants aged 55–74 were two times more likely to do so. Caucasians were three times more likely to consume salmon/tuna weekly, as well as tuna weekly, compared with non-Caucasians. These correlations are of particular interest due to additional correlations between tuna/salmon consumption and both Hg and omega-3s. Future studies should assess why older Caucasian consumers prefer these types of fish, and how to steer them toward lower Hg, higher omega-3 fish types (e.g., salmon), and away from higher Hg fish (e.g., tuna).

Females were approximately twice as likely as males to consume seafood ‘‘other’’ than tuna or salmon at least weekly. Non-Caucasians were nearly three times as likely to eat seafood in this ‘‘other’’ category at least weekly. However, because the individual fish species in this ‘‘other’’ category are unspecified, it is difficult to communicate risk beyond conveying information about ‘‘other fish’’ consumed once per week or greater to the targeted consumers. Thus, future studies should consider designing for greater power to gain more information about consumers of specific types of ‘‘other’’ non-salmon/tuna fish. Weekly intake of fish, regardless of type, was significantly associated with levels of total Hg in the blood, and showed borderline association with omega-3s in the blood. Looking at specific types of fish, plasma omega-3 levels were associated with salmon intake at least weekly, in line

123

J Community Health Table 2 Frequencies of demographic and biomarker variables by seafood consumption categories

Total (N)

All participants (%)

Any fish C1x/week (%)

Only tuna C1x/week (%)

Only salmon C1x/week (%)

Salmon and tuna C 1x/ week (%)

Other fish types C 1x/ week (%)

No fish types C 1x/ week (%)

290

223

37

29

62

95

67

Gender Female Male

58.6 41.4

59.2 40.8

46.0 54.1

48.3 51.7

59.7 40.3

67.4 32.6

56.7 43.3

15–34

29.0

28.7

35.1

27.6

17.7

33.7

29.9

35–54

26.2

23.3

29.7

10.3

27.4

22.1

35.8

55–74

37.2

39.9

35.1

55.2

43.6

34.7

28.4

6.9

7.2

0.0

6.9

11.3

7.4

6.0

Age (years)

C75 Ethnicity African American Asian American

3.8

3.1

5.4

0.0

1.6

6.3

6.0

10.0

8.5

5.4

6.9

1.6

14.7

14.9

Hispanic/Latino

4.5

4.9

0.0

6.9

3.2

5.3

3.0

White/Caucasian

79.0

80.3

89.2

86.2

91.9

67.4

74.6

[1 Ethnicity/other

2.8

3.1

0.0

0.0

1.6

6.3

1.5

Born in United States Yes

84.8

86.1

91.9

89.7

91.9

79.0

80.6

No Yearly income

15.2

13.9

8.1

10.3

8.1

21.1

19.4

\$25,000

16.6

15.7

16.2

6.9

14.5

19.0

19.4

$25,000–$70,000

29.7

28.7

35.1

31.0

24.2

28.4

28.4

$70,000–$110,000

24.1

24.2

13.5

13.8

30.7

27.6

23.9

$110,000–$200,000

20.7

21.5

24.3

31.0

19.4

19.0

17.9

6.6

6.3

10.8

13.8

6.5

2.1

7.5

Below grade 9

0.0

0.0

0.0

0.0

0.0

0.0

0.0

Grades 9–11

0.7

0.0

0.0

0.0

0.0

0.0

3.0

High School Grad

6.6

7.6

8.1

6.9

4.8

9.5

3.0

College 1–3 years

26.2

26.5

18.9

20.7

32.3

27.4

25.4

College C4 years

66.6

65.9

73.0

72.4

62.9

63.2

68.7

[$200,000 Education level

Blood Hg level (lg/L) \5.8

57.6

52.0

59.5

41.4

43.6

57.9

76.1

5.8–10

16.2

17.9

5.4

31.0

21.0

16.8

10.4

10–15 [15

11.7 14.1

13.5 16.1

16.2 18.9

10.3 17.2

14.5 19.4

12.6 12.6

6.0 7.5

Plasma Omega-3 level (% total fatty acids) \5.29

33.1

30.9

40.5

17.2

24.2

35.8

40.3

5.29–6.65

32.4

30.9

35.1

27.6

25.8

33.7

37.3

[6.65

32.8

35.9

24.3

55.2

46.8

27.4

22.4

\265

32.8

30.5

35.1

34.5

25.8

30.5

40.3

265–346

33.4

33.6

37.8

41.4

33.9

29.5

32.8

[346

31.7

33.2

24.3

20.7

35.5

35.8

25.4

Blood Se level (ng/mL)

with the reported literature [37, 38]. Similarly, weekly consumption of both salmon and tuna was associated with total blood Hg and borderline associated with omega-3

123

levels. Weekly consumption of tuna alone was not associated with Hg in this study, although the literature suggests this correlation [1, 3]. This is likely because we

J Community Health Table 3 Logistic regression models: ORs and CIs for predictors of seafood consumption groups Any fish type C1x/week OR (CI)

Only tuna C1x/week OR (CI)

Only salmon C1x/week OR (CI)

Salmon and tuna C1x/week OR (CI)

Other fish types C1x/week OR (CI)

No fish types C1x/week OR (CI)

Gendera Female

–

Male Age (years)b

–

0.94 (0.54,1.64)

18–34

–

35–54

1.79 (0.89,3.60)

–

–

1.72 (0.79,3.75)

–

–

0.96 (0.54,1.71) –

0.39 (0.10,1.53)

1.97 (0.86,4.52)

– 0.58* (0.34,0.97) – 0.59 (0.30,1.16)

– 1.06 (0.61,1.86) –

0.67 (0.33,1.32)

0.93 (0.39,2.22)

1.5 (0.76,3.05)

55–74

1.42 (0.70,2.86)

0.79 (0.34,1.80)

1.74 (0.70,4.30)

2.27* (1.05,4.89)

0.66 (0.36,1.19)

0.71 (0.35,1.43)

C75

1.21 (0.36,4.04)

0.0 (0.00,?)

1.10 (0.21,5.63)

3.67* (1.20,11.20)

0.81 (0.29,2.26)

0.83 (0.25,2.76)

Race/ethnicityc White/Caucasian

–

Other

–

0.75 (0.38,1.50)

0.31* (0.10,0.96)

–

–

0.60 (0.19,1.93)

0.34* (0.12,0.91)

– 2.73* (1.45,5.12)

– 1.33 (0.67,2.65)

Born in United Statesc Yes

–

No

–

0.68 (0.32,1.44)

0.37 (0.10,1.29)

–

–

0.65 (0.18,2.38)

0.55 (0.20,1.52)

– 1.76 (0.88,3.50)

– 1.48 (0.70,3.14)

Yearly incomec \$25,000

–

–

–

–

–

–

$25,000–$70,000

1.29 (0.56,2.99)

1.48 (0.50,4.34)

3.08 (0.62,15.29)

0.78 (0.30,2.01)

0.86 (0.40,1.86)

0.78 (0.34,1.80)

$70,000–$110,000 $110,000–$200,000

1.34 (0.55,3.26) 1.64 (0.64,4.21)

0.52 (0.14,1.92) 1.26 (0.39,4.06)

1.49 (0.25,8.85) 4.65 (0.91,23.83)

1.3 (0.51,3.35) 0.84 (0.31,2.31)

1.2 (0.55,2.71) 0.89 (0.38,2.07)

0.75 (0.31,1.82) 0.61 (0.24,1.58)

[$200,000

1.01 (0.28,3.68)

1.96 (0.42,9.12)

5.68 (0.83,38.81)

0.76 (0.19,3.03)

0.26 (0.05,1.32)

0.99 (0.27,3.58)

Education levelc Not a college graduate College graduate

–

–

0.94 (0.51,1.71)

1.43 (0.64,3.19)

–

–

1.50 (0.62,3.62)

0.76 (0.41,1.40)

– 0.85 (0.50,1.44)

– 1.07 (0.58,1.96)

The asterisk indicates significance of the odds ratio at the 95 % confidence interval All CIs are at the 95 % level ‘‘–’’ reference variable a

Adjusted for age;

b

adjusted for gender;

c

adjusted for age and gender

grouped tuna steak consumers, canned white tuna consumers, and canned light tuna consumers together into one group to increase power due to limited sample size, and doing so may have masked relationships among demographics, total Hg, and the individual tuna types. This was a limitation of the study: although all three types of tuna have higher levels of Hg, each provides a different risk level in consumers due to differing average organic Hg concentrations; e.g., tuna steak is suggested to have higher levels of organic Hg when compared to both types of canned tuna [27]. As a result, future work with greater power to investigate individual tuna types, their respective Hg risk levels, and demographic profiles may prove useful for advising species-specific risk communication efforts. Similarly, although correlations between Se and particular fish types were not significant, the trends suggest that higher Se levels may be associated with frequent consumption of a combination of fish types. Consistent with this relationship between elevated Se and elevated seafood

consumption, approximately 90.5 % of our population of avid seafood consumers had Se levels above the mean US concentration of 136.7 ng/mL [39]. A study with a larger sample size may be needed to shed insights on relationships between Se and individual fish types. Aside from tuna, the study’s survey also collected information regarding other high-Hg fish identified in Karimi et al. [24], including swordfish, shark, and marlin. However, the number of participants who consumed these fish types was too small to have power in these analyses; thus, future work may seek to explore correlations between these higher-Hg fish and demographics with a larger sample size. Another fish type that may be important in future work is shrimp, as it is the most commonly purchased seafood in the United States, at an average of approximately 4.1 pounds per American in 2010 [37]. A grouping of shrimp consumers was considered in this study, but only a few individuals reported eating solely shrimp weekly. Focusing

123

J Community Health

on the demographics of shrimp consumers in a targeted study or larger dataset will expand upon the risk communication implications of this type of work. The ‘‘other’’ fish category in this study includes some higher-Hg fish (swordfish, shark, and marlin), shrimp, along with twenty-eight other fish types, aside from salmon and any of the tuna varieties. All of these seafood varieties were combined into a single category due to a very small number of participants who consumed them regularly, if at all. As a result of how inclusive the ‘‘other’’ category is, information gained from that category is limited, and therefore analysis of consumers of those individual seafood types must be reserved for future work. It is important to note that this study population is not representative of the surrounding area due to oversampling of avid seafood eaters, and thus it is expected that the population of the surrounding area might have different patterns of eating specific types of seafood, and would likely have lower levels of Hg, omega-3s, and Se than the sample population. Most participants were drawn from the Suffolk County area by virtue of the geographic extent of recruitment efforts. The Suffolk County population was partially reflected in the seafood consumption cohort, although there were differences in age, education, gender, and ethnicity. This may be partially inherent to the recruitment process: older and/or more educated members of the population may generally be more inclined to participate in studies [38], but also may reflect that these groups are more likely to be avid seafood consumers. The higher education level of the participants in our sample size may also have an effect in strategies for future risk communication in determining how best to target the most consumers: a majority (72 %) of the participants have a college degree or higher, and this might suggest risk communication efforts might be most effective when targeted particularly to that demographic group using strategies similar to the participant recruitment efforts completed within this study (Table 2). The underlying differences in demographic characteristics need to be considered in evaluating the generalizability of study results to other populations of avid seafood consumers.

Conclusion This study sheds light on relationships between demographics of seafood consumers, patterns of seafood intake, and biologic measures of total blood Hg, Se, and omega-3s. Several types of demographic characteristics were associated with different patterns of seafood intake, plasma omega-3s, and total blood Hg and Se, highlighting opportunities for risk communication among avid seafood consumers. As seafood consumption continues to rise [40, 41]

123

additional regional seafood consumption studies are needed to help understand risks and benefits and to guide risk communication selection efforts in the seafood-eating population. Acknowledgments This work was generously supported by the Gelfond Fund for Mercury Research and Outreach and the Simons Foundation; we also thank the participants for their time and interest in our study.

References 1. Nesheim, M. C., & Yaktine, A. L. (Eds.). (2006). Institute of medicine, seafood choices: Balancing benefits and risks. Washington, DC: Institute of Medicine of the National Academies. 2. Rheinberger, C. M., & Hammitt, J. K. (2012). Risk trade-offs in fish consumption: A public health perspective. Environmental Science and Technology, 46(22), 12337–12346. doi:10.1021/ es302652m. 3. Mozaffarian, D., & Rimm, E. B. (2006). Fish intake, contaminants, and human health: Evaluating the risks and the benefits. Journal of the American Medical Association, 296(15), 1885–1899. 4. Goepp, J. G. (2006). Selenium: Important health benefits from an overlooked trace mineral. Life Extension, 12(11), 66–73. 5. Gustafson, A., Christian, J. W., Lewis, S., Moore, K., & Jilcott, S. (2013). Food venue choice, consumer food environment, but not food venue availability within daily travel patterns are associated with dietary intake among adults, Lexington Kentucky 2011. Nutrition Journal. doi:10.1186/1475-2891-12-17. 6. Dave, J. M., Evans, A. E., Pfeiffer, K. A., Watkins, K. W., & Saunders, R. P. (2010). Correlates of availability and accessibility of fruits and vegetables in homes of low-income Hispanic families. Health Education Research, 25(1), 97–108. doi:10.1093/ her/cyp044. 7. Nesbitt, A., Majowicz, S., Finley, R., Pollari, F., Pintar, K., Marshall, B., et al. (2008). Food consumption patterns in the Waterloo Region, Ontario, Canada: A cross-sectional telephone survey. BMC Public Health. doi:10.1186/1471-2458-8-370. 8. Denger, R., Adams, C., Moss, S., & Mack, S. (1994). Per capita fish and shellfish consumption in Florida. Florida Agricultural Market Research Center. Gainseville, FL: University of Florida. 9. Sechena, R., Shiquan, L., Lorenzana, R., Nakano, C., Polissar, N., & Fenske, R. (2003). Asian American and Pacific Islander seafood consumption—A community-based study in King County, Washington. Journal of Exposure Analysis and Environmental Epidemiology, 13(4), 256. doi:10.1038/sj.jea.7500274. 10. Wang, H. H., Zhang, X., Ortega, D., & Olynk Widmar, N. J. (2013). Information on food safety, consumer preference and behavior: The case of seafood in the US. Food Control, 33(1), 293–300. doi:10.1016/j.foodcont.2013.02.033. 11. Wennberg, M., Tornevi, A., Johansson, I., Hornell, A., Norberg, M., & Bergdahl, I. A. (2012). Diet and lifestyle factors associated with fish consumption in men and women: A study of whether gender differences can result in gender-specific confounding. Nutrition Journal. doi:10.1186/1475-2891-11-101. 12. Cardoso, C., Lourenc¸o, H., Costa, S., Gonc¸alves, S., & Nunes, M. L. (2013). Survey into the seafood consumption preferences and patterns in the Portuguese population. Gender and regional variability. Appetite, 64, 20–31. doi:10.1016/j.appet.2012.12.022. 13. Mahaffey, K. R., Clickner, R. P., & Jeffries, R. A. (2009). Adult women’s blood Hg concentrations vary regionally in the United States: Association with patterns of fish consumption (NHANES

J Community Health

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

1999–2004). Environmental Health Perspectives, 117(1), 47–53. doi:10.1289/ehp.11674. Fletcher, A. M., & Gelberg, K. H. (2013). An analysis of Hg exposures among the adult population in New York State. Journal of Community Health, 38(3), 529–537. doi:10.1007/ s10900-012-9646-9. McKelvey, W., Gwynn, R. C., Jeffery, N., Kass, D., Thorpe, L. E., Garg, R. K., et al. (2007). A biomonitoring study of lead, cadmium, and Hg in the blood of New York City adults. Environmental Health Perspectives, 115(10), 1435–1441. doi:10. 1289/ehp.10056. Traynor, S., Kearney, G., Olson, D., Hilliard, A., Palcic, J., & Pawlowicz, M. (2013). Fish consumption patterns and Hg exposure levels among women of childbearing age in Duval County, Florida. Journal of Environmental Health, 75(6), 8–15. Retrieved from http://apps.webofknowledge.com.libproxy.cc.stonybrook. edu/full_record.do?product=UA&search_mode=GeneralSearch& qid=38&SID=1DWD2CobATkPz3r5IC2&page=1&doc=1. Geer, L. A., Persad, M. D., Palmer, C. D., Steuerwald, A. J., Dalloul, M., Abulafia, O., & Parsons, P. J. (2012). Assessment of prenatal Hg exposure in a predominately Caribbean immigrant community in Brooklyn, NY. Journal of Environmental Monitoring, 14(3), 1035–1043. doi:10.1039/c2em10835f. Sunde, R. A. (2012). Selenium. In A. C. Ross, B. Caballero, R. J. Cousins, K. L. Tucker, & T. R. Ziegler (Eds.), Modern nutrition in health and disease (11th ed., pp. 225–237). Philadelphia, PA: Lippincott Williams & Wilkins. Wen, Y. T., Dai, J. H., & Gao, Q. (2014). Effects of omega-3 fatty acid on major cardiovascular events and mortality in patients with coronary heart disease: A meta-analysis of randomized controlled trials. Nutrition Metabolism and Cardiovascular Diseases, 24(5), 470–475. doi:10.1016/j.numecd.2013.12. 004. Bemrah, N., Sirot, V., LeBlanc, J. C., & Volatier, J. L. (2009). Fish and seafood consumption and omega-3 intake in French coastal populations: CALIPSO survey. Public Health Nutrition, 12(5), 599–608. doi:10.1017/S1368980008002681. Niskar, A., Paschal, D., Kieszak, S., Gunter, E., Pirkle, J., Rubin, C., et al. (2003). Serum Se levels in the US population: Third National Health And Nutrition Examination Survey, 1988–1994. Biological Trace Element Research, 91(1), 1–10. doi:10.1385/ BTER:91:1:1. U.S. Department of Agriculture, Agricultural Research Service. (2015). USDA National Nutrient Database for Standard Reference, Release 27. Nutrient Data Laboratory Homepage 2011. Laclaustra, M., Stranges, S., Navas-Acien, A., Ordovas, J. M., & Gullar, E. (2010). Serum Se and serum lipids in US adults: National Health and Nutrition Examination Survey (NHANES) 2003–2004. Atherosclerosis, 210, 643–648. doi:10.1016/j.athero sclerosis.2010.01.005. Karimi, R., Fitzgerald, T. P., & Fisher, N. S. (2012). A quantitative synthesis of Hg in commercial seafood and implications for exposure in the United States. Environ Health Perspectives, 120(11), 1512–1519. Agency for Toxic Substance and Disease Registry. (1999.) Mercury. CAS#7439-97-6 ToxFAQs. Retrieved from: \http:// www.atsdr.cdc.gov/toxprofiles/tp.asp?id=115&tid=24[. Block, G., Hartman, A. M., & Naughton, D. (1990). A reduced dietary questionnaire: development and validation. Epidemiology, 1, 58–64. Karimi, R., Fisher, N., & Meliker, J. (2014). Hg-nutrient signatures in seafood and the blood of avid seafood consumers.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

Science of the Total Environment, 496, 636–643. doi:10.1016/j. scitoenv.2014.04.049. Karimi, R., Silbernagel, S., Fisher, N. S., & Meliker, J. R. (2014). Elevated blood Hg at recommended seafood consumption rates in adult seafood consumers. International Journal of Hygiene and Environmental Health, 217, 758–764. doi:10.1016/j.ijheh.2014. 03.007. Hodge, A. M., Simpson, J. A., Gibson, R. A., Sinclair, A. J., Makrides, M., O’Dea, K., et al. (2007). Plasma phospholipid fatty acid composition as a biomarker of habitual dietary fat intake in an ethnically diverse cohort. Nutrition, Metabolism, & Cardiovascular Diseases, 17(6), 415–426. U.S. Census Bureau. (2007–2011). ACS demographic and housing estimates: 2007–2011 American Community Survey 5-year estimates. Retrieved 23 July 2013 from http://factfinder2.census. gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk. U.S. Census Bureau. (2007–2011). Selected economic characteristics: 2007–2011 American Community Survey 5-year estimates. Retrieved 27 August 2015 from http://factfinder.census. gov/faces/tableservices/jsf/pages/productview.xhtml?src=bcmc. U.S. Census Bureau. (2007–2011). Selected Social Characteristics in the United States: 2007–2011 American Community Survey 5-Year Estimates. McKelvey, W., Chang, M., Arnason, J., Jeffery, N., Kricheff, J., & Kass, D. (2010). Hg and polychlorinated biphenyls in Asian market fish: A response to results from mercury biomonitoring in New York City. Environmental Research, 110, 650–657. doi:10. 1016/j.envres.2010.07.002. McKelvey, W., Jeffery, N., Clark, N., Kass, D., & Parsons, P. (2011). Population-based inorganic Hg biomonitoring and the identification of skin care products as a source of exposure in New York City. Environmental Health Perspectives, 119(2), 203–209. doi:10.1289/ehp.1002396. Tsuchiya, A., Duff, R., Stern, A. H., White, J. W., Krogstad, F., Burbacher, T. M., et al. (2012). Single blood Hg samples can result in exposure misclassification: Temporal monitoring within the Japanese community (United States). Environmental Health: A Global Access Science Source. doi:10.1186/1476-069X-11-37. Tsuchiya, A., Hinners, T. A., Krogstad, F., White, J. W., Burbacher, T. M., Faustman, E. M., & Marien, K. (2009). Longitudinal Hg monitoring within the Japanese and Korean communities (United States): Implications for exposure determination and public health protection. Environmental Health Perspectives, 117(11), 1760–1766. doi:10.1289/ehp.0900801. NOAA. (2010). Seafood and Human Health. NOAA Fisheries. Retrieved July 3, 2014 from http://www.nmfs.noaa.gov/aqua culture/faqs/faq_seafood_health.html#8what. Strobel, C., Jahreis, G., & Kuhnt, K. (2012). Survey of n-3 and n-6 polyunsaturated fatty acids in fish and fish products. Lipids in Health and Disease. doi:10.1186/1476-511X-11-144. National Institute of Health: Office of Dietary Supplements. (2013, July 2). Selenium: Dietary supplement fact sheet. http://ods.od.nih.gov/factsheets/SeleniumRetrieved from: HealthProfessional/. Jensen, H. H. (2006). Changes in seafood consumer preference patterns and associated changes in risk exposure. Marine Pollution Bulletin, 53(10–12), 591–598. doi:10.1016/j.marpolbul.2006. 08.014. Delgado, C. L., Wada, N., Rosegrant, M. W., Meijer, S., & Ahmed, M. (2003). Fish to 2020: Supply and demand in changing global markets. Washington, D.C. & Penang, Malaysia: International Food Policy Research Institute and WorldFish Center.

123