6

Health Outcomes Associated with Seafood Consumption

Seafood provides an array of nutrients that are associated with beneficial health effects for those who are or may become pregnant or are lactating, and for infants, children, and adolescents. In particular, the long-chain omega-3 polyunsaturated fatty acids (n-3 LCPUFAs), as well as docosahexaenoic acid (DHA) and eicosapentaenoic acid, both synthesized from alpha-linolenic acid, have been shown to have a range of health benefits, including neurocognitive development.

As discussed earlier in this report, the average amount of seafood consumed by pregnant women is below recommended levels, and only about one in five U.S. women of childbearing age achieves an intake of seafood that is recommended for optimal maternal and child health. Furthermore, few children and adolescents in the United States and Canada consume the two or more servings of seafood per week recommended by the Dietary Guidelines for Americans 2020–2025 (DGA). The DGA, however, acknowledges that robust research on specific health benefits associated with seafood consumption by children is limited, and further evidence is needed to clarify the contribution of seafood consumption to children’s health (USDA and HHS, 2020).

This chapter considers seafood consumption patterns, dietary intake and nutrient composition of seafood, and exposure to contaminants associated with seafood consumption, and reviews the evidence on those relationships with child health outcomes. The committee determined that biomarkers of response to seafood consumption were outside the scope of its task.

SEAFOOD CONSUMPTION AND HEALTH OUTCOMES IN CHILDREN AND ADOLESCENTS

The discussion of evidence on seafood consumption and health outcomes in children and adolescents begins with results from systematic reviews commissioned by the committee (Appendix C). Additional evidence on outcomes not included in the commissioned systematic reviews was identified through a review of existing systematic reviews (Appendix D), as well as evidence submitted by the sponsor. Lastly, the committee obtained additional evidence through a narrative review of the published literature and technical reports. Examples of reports that the committee considered include the following:

• “Fish, Shellfish and Children’s Health: An Assessment of Benefits, Risks and Sustainability” from the American Academy of Pediatrics Council on Environmental Health Committee on Nutrition (Bernstein et al., 2019)

• The U.S. Food and Drug Administration’s (FDA’s) Quantitative Assessment of the Net Effects on Fetal Neurodevelopment from Eating Commercial Fish (FDA, 2014)
• The Report of the Joint FAO/WHO Expert Consultation on the Risks and Benefits of Fish Consumption (FAO/WHO, 2011)
• The Summary and Conclusions from the second Joint FAO/WHO Expert Consultation on Risks and Benefits of Fish Consumption (FAO/WHO, 2023)
• The Norwegian Scientific Committee for Food and Environment (VKM) report Benefit and Risk Assessment of Fish in the Norwegian Diet (Andersen et al., 2022).

Updated Reviews

The committee commissioned the Texas A&M Agriculture, Food, and Nutrition Evidence Center to update three existing systematic reviews previously published by the U.S. Department of Agriculture’s Nutrition Evidence Systematic Review team for the 2020 Dietary Guidelines Advisory Committee (DGAC). One review examined relationships between seafood consumption during pregnancy and lactation and neurocognitive development in the child, and another examined seafood consumption during childhood and adolescence and neurocognitive development. A third review examined seafood consumption during childhood and adolescence and cardiovascular disease. An analytic framework was developed to specify the interventions of interest (seafood consumption) during pregnancy and lactation and during childhood and adolescence and outcomes related to neurocognitive development or cardiovascular disease in children and adolescents (see Appendix D).

Summary of Key Systematic Reviews on Relationship Between Seafood Consumption During Pregnancy and Lactation and Neurocognitive Development in the Child

Two systematic reviews of seafood intake and neurocognitive outcomes conducted for the DGAC examined evidence on developmental outcomes, including:

• cognitive development,
• language and communication development,
• movement and physical development,
• social-emotional and behavioral development,
• attention deficit disorder (ADD)/attention deficit hyperactivity disorder (ADHD),
• autism spectrum disorder (ASD),
• academic performance, and
• anxiety and depression (Snetselaar et al., 2020a,b).

Snetselaar et al. (2020a) examined associations of seafood intake during pregnancy and lactation and a range of developmental outcomes. This systematic review found modest evidence that seafood intake during pregnancy is positively associated with measures of cognitive development in young children. The review also found limited evidence that seafood intake during pregnancy was associated with improved measures of language development in the child. There was insufficient evidence, however, to draw conclusions about other outcomes, and there were no studies of seafood intake during lactation that met their inclusion criteria.

Snetselaar et al. (2020b) described the results of the systematic review used by the 2020 DGAC to assess associations of seafood consumption with neurocognitive outcomes in children and adolescents. Based on 13 included studies, the review concluded that seafood intake during childhood and adolescence had either a beneficial or nonsignificant relationship across developmental domains, particularly in cognitive development, language and communication development, and movement and physical development. No conclusions were made regarding the relationships between seafood intake and developmental domains because of the inadequate number of studies, inconsistency in results, risk of bias in classification of exposures, and heterogeneity of outcome assessments.

Considerable variability was found across the instruments used to capture cognitive function; some authors used caregiver reporting and others used other tools to assess seafood consumption with outcomes.

Similarly, no conclusions were made regarding the relationships between seafood intake and academic performance, ADD/ADHD, anxiety and depression, or ASD, owing to an inadequate number of studies. It is important to note that neither review identified evidence of harm from maternal or child fish consumption in relation to child neurodevelopmental outcomes.

In the systematic review update commissioned by the committee, 11 additional studies were identified that met the review criteria. (Appendix C). Only one of the studies, a randomized controlled trial (RCT) conducted in a Norwegian population (Kvestad et al. 2021), found a significant beneficial effect of fish consumption during pregnancy and developmental outcomes in the child as measured by maternal reporting on developmental screening questionnaires. The study randomized 133 pregnant women (at 19 weeks of gestation or less) to receive 200 g cod fillet twice weekly (intervention) or to maintain their habitual diet (control) for 16 weeks. No difference was found between infants in the intervention and control groups on the total Ages and Stages Questionnaire, 2nd edition (ASQ-2), scores at 3, 6, and 11 months. A difference was found, however, on the ASQ: Social-Emotional scores in favor of the intervention group (P = 0.020). Thus, while there was no evidence for an effect of increased cod intake during pregnancy on overall child development in the first year of life, there was a positive effect on socioemotional development.

In contrast, two studies identified negative effects of fish (cod) consumption, one on gross motor development (Varsi et al., 2022) and one using the Bayley Scales of Infant Development cognitive composite scores (Markus et al., 2021). The remaining studies reported null findings and none of the studies found evidence of harm.

Summary of a Key Systematic Review on the Relationship Between Seafood Consumption During Childhood and Cardiovascular Disease

In the third review conducted for the 2020 DGAC, Snetselaar et al. (2020c) examined the relationship of seafood consumption during childhood and cardiovascular disease. The health outcomes included in this review were blood pressure, total cholesterol, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, HDL cholesterol, serum triglycerides, and overall cardiovascular disease. One RCT and one prospective cohort study reported no significant relationship between tuna or fish intake and blood pressure in children ages 10–12 years. One RCT reported lower total cholesterol with higher tuna consumption in girls, but not boys; two RCTs and one prospective cohort study reported lower triglyceride levels associated with higher fish consumption or no significant relationship. No studies on cardiovascular disease met the inclusion criteria for the systematic review. The authors could not derive any conclusions on these outcomes because of insufficient evidence, and a grade was not assignable.

Supplemental Reviews of Systematic Reviews

The committee’s supplemental review (Appendix D) identified additional systematic reviews on associations of seafood consumption during pregnancy and lactation, and childhood, that included outcomes not covered by the commissioned systematic review update. The evidence from these reviews is summarized in the following sections.

Neurodevelopmental Outcomes

Hibbeln et al. (2019) conducted two systematic reviews on neurodevelopmental outcomes. The first review addressed the relationship between maternal seafood consumption during pregnancy and lactation and neurocognitive development of the infant. This review identified 29 studies representing 24 unique cohorts. Of the 29 studies, 24 reported beneficial outcomes associated with maternal seafood consumption and neurocognition on some or all the tests administered to children. Based on their analysis, the authors reported moderate and consistent evidence for an association of consumption of a broad range of amounts and types of commercially available seafood during pregnancy with improved neurocognitive development in the offspring as compared to not consuming seafood.

These studies suggested benefits to neurocognitive development occurred across the lowest (approximately 4 oz/week) to the highest (greater than 12 oz/week, including up to more than 100 oz/week) levels of intake. The authors further concluded that the evidence did not meet the criteria for “strong evidence” owing to a paucity of RCTs that may not be ethical or feasible to conduct during pregnancy.

The second review addressed the relationship between seafood consumption during childhood and adolescence (up to 18 years of age) and neurocognitive development. This review yielded 15 articles, comprising 25,031 children (six RCTs, four prospective cohorts, and nine case–control studies published between 2001 and 2019). They found moderate and consistent evidence indicating that seafood consumption of more than 4 oz/week and likely greater than 12 oz/week during childhood has beneficial associations with neurocognitive outcomes. No study in the review reported net adverse outcomes from seafood consumption in children; therefore, the authors concluded that it was unlikely that mercury exposure from seafood consumption by children was associated with substantive neurocognitive harms (Hibbeln et al., 2019).

Avella-Garcia and Julvez (2014) conducted a systematic review of primarily prospective cohort studies on seafood intake during pregnancy and childhood with a range of neurodevelopmental, behavioral, cognitive, and general developmental outcomes. Of the 16 publications identified in the review, 8 evaluated seafood exposure prenatally, 5 evaluated postnatal exposure, and 3 studies evaluated both pre- and postnatal exposure. All but two studies were prospective cohorts and most used food frequency questionnaires (FFQs) to assess seafood consumption.

Most of the studies found that seafood intake during pregnancy and postnatal periods appeared, overall, to be beneficial in connection with a wide range of neurodevelopmental outcomes, including neurological and cognitive functions, social competence, hyperactivity, and school performance, although potential risks such as inattention, hyperactivity, or impulsivity were identified at higher levels of intake. In general, prenatal exposure to seafood as part of the maternal diet was found to be associated with benefits in developmental outcomes in the child, including neurological, behavioral, and cognitive functioning. Seafood consumption was associated with improved neurological functions and lower disorder prevalence; better performances in cognitive functions related to verbal, memory, and visual-performance abilities; and improved behavioral outcomes such as hyperactivity, social competence, and school grades.

Some studies (de Groot et al., 2012; Freire et al., 2010; Mendez et al., 2009) observed an inverted U-shape pattern in the association between seafood intake and the neurodevelopment of children, which may be related to higher pollutant exposure in the heavy seafood consumers. Freire et al. (2010) suggested that consuming oily fish as a source of DHA concentrations counterbalanced some neurotoxic effects. Avella-Garcia and Julvez (2014) concluded from the review that public guidance should acknowledge the risk of contaminant exposure with large amounts of seafood, especially species with high mercury concentrations, while also noting the benefits of moderate intake.

Asthma and Allergic Outcomes

Yang et al. (2013) conducted a systematic review of studies on fish intake and risk of child asthma. A meta-analysis of three studies suggested that fish consumption in infants was associated with decreased risk of asthma during childhood. Two studies reported that higher levels of n-3 LCPUFAs in expressed breast milk were associated with lower incidence of asthma in their offspring. Maternal intake of fish during pregnancy was not reported to be related to development of asthma in their offspring.

A 2011 systematic review explored the associations between seafood intake and atopic or allergic outcomes in infants and children (Kremmyda et al., 2011). Positive associations were reported for maternal fish intake and atopic or allergic outcomes in their offspring, based on four cohort studies and one case–control study. When looking at associations between these outcomes and seafood intake by the infant or child, however, the results reported from five cohort, six cross-sectional, and three case-control studies were inconsistent. The authors concluded that more research is needed in this area before developing a conclusion.

Social-Emotional Outcomes

An RCT by Hysing et al. (2018) investigated associations between fish consumption and social/emotional development in Norway. Children ages 4–6 years were randomized to a fish lunch group (three lunches per week of fatty fish including herring and mackerel) or to a meat lunch group. No significant differences in social and/or emotional development were assessed using the Strengths and Difficulties Questionnaire (SDQ) for young children. In a second Norwegian RCT, teens (average age 14.6 years) were randomized to a high-fish lunch (three lunches of salmon, mackerel, or herring), a high-meat group, or a supplement group (fish oil supplement three times per week). Using the SDQ for older children, no significant differences were observed between groups (Skotheim, et al., 2017).

In a third RCT conducted in Denmark, children ages 9–10 years were randomized to a high-fish consumption group (compared with a high-poultry consumption group). Using the Behavior Rating Inventory of Executive Function, no significant differences were found in social and/or emotional developmental scores (Teisen et al., 2020).

Other Reports Related to Seafood Consumption and Child Health

The Food and Agriculture Organization of the United Nations (FAO) and World Health Organization (WHO) convened two Expert Consultations on the risks and benefits of fish consumption. The first, published in 2011, had the following three conclusions related to fish consumption by women and children:

1. When comparing the benefits of n-3 LCPUFAs with the risks of Hg among women of childbearing age, maternal fish consumption lowers the risk of suboptimal neurodevelopment in their offspring compared with the offspring of women not eating fish in most circumstances evaluated.
2. At levels of maternal exposure to dioxins (from fish and other dietary sources) that do not exceed the provisional tolerable monthly intake (PTMI) of 70 pg/kg body weight established by the Joint FAO/WHO Expert Committee on Food Additives (for polychlorinated dibenzodioxins [PCDDs], polychlorinated dibenzofurans [PCDFs], and coplanar PCBs), neurodevelopmental risk for the fetus is negligible. At levels of maternal exposure to dioxins (from fish and other dietary sources) that exceed the PTMI, neurodevelopmental risk for the fetus may no longer be negligible.
3. Among infants, young children, and adolescents, the available data are currently insufficient to derive a quantitative framework of the health risks and health benefits of eating fish. However, healthy dietary patterns that include fish consumption and are established early in life influence dietary habits and health during adult life (FAO/WHO, 2011, p. 33).

The most recent FAO/WHO Expert Consultation occurred in 2023, and while the full report has not yet been released, the published executive summary includes the following three conclusions relevant to this report:

1. Strong evidence exists for the benefits of total fish consumption during all life stages: pregnancy, childhood, and adulthood. For example, associations are found for maternal consumption during pregnancy with improved birth outcomes and for adult consumption with reduced risks for cardiovascular and neurological diseases. This evidence for health benefits of total fish consumption reflects the overall effects of nutrients and contaminants in fish on the studied outcomes, including nutrients and contaminants not specifically considered in the evidence review.
2. Benefits derived from general population studies and individual effects will vary depending on overall diet (e.g., selenium intake, exposure to other contaminants) and characteristics of consumers (e.g., n-3 [LC]PUFA status and individual susceptibility) and fish consumed (e.g., fish species and food preparation methods).
3. Risk–benefit assessments at regional, national, or subnational levels are needed to refine fish consumption recommendations considering local consumption habits, fish contamination levels, and nutrient content; nutritional status of the population of interest; cultural habits; and demographics (FAO/WHO, 2023, p. 3).

The VKM conducted a benefit and risk assessment of fish in the Norwegian diet (Andersen et al., 2022). This report consisted of a quantitative assessment of benefits and risks from fish consumption, a semiquantitative benefit

assessment of nutrients in fish, and a semi-quantitative risk assessment of contaminants in fish. No associations were categorized as “convincing,” but “probable” associations were found for a number of adult health outcomes, as well as for a benefit of fish consumption during pregnancy for preventing both preterm birth and low birth weight. The VKM found “limited, suggestive” evidence for a “protective association between maternal total fish consumption and child neurodevelopment” and also “limited, suggestive” evidence that “child fish consumption (total and fatty fish) benefits neurodevelopment” (Andersen et al., 2022, pp. 303, 320).

The committee notes that 62 percent of Norwegian women consume two or more meals of fish per week and mean intake was 238 grams (8 oz) per week. The overall conclusion of the Norwegian analysis was that low fish consumption is a potential health risk, and that optimal beneficial health effects of fish consumption are not obtained at current fish intake levels in Norway, which are substantially higher than those observed in the United States and Canada. The overall recommendation was for 300–450 g (about 10–16 oz) prepared fish per week for adults, of which at least 200 g should be fatty fish such as salmon, trout, mackerel, or herring. This recommendation is higher than the DGA, which recommends 8–12 oz of fish per week.

Summary on Evidence on Health Outcomes Associated with Consumption of Seafood in Pregnancy and Childhood

Taken together, the evidence reviewed by the committee indicates that higher fish consumption by women of childbearing age, including women who are pregnant and lactating, and by children, is not associated with any detectable adverse health outcomes, although one review found a neurocognitive benefit in 13 of 15 studies in the review. Moreover, there is evidence that greater fish consumption by women during pregnancy is likely associated with a number of health benefits, including improved birth outcomes. To date, available evidence for health benefits of child seafood consumption is overall not sufficient for conclusions to be drawn, but evidence does not suggest any harm from seafood consumption among children, in the studied populations, and potential benefits exist.

EXPOSURE TO TOXICANTS IN SEAFOOD AND CHILD GROWTH AND DEVELOPMENT

The Texas A&M Agriculture, Food, and Nutrition Evidence Center was commissioned to conduct a de novo systematic review to examine associations between seafood-related contaminants and health outcomes during pregnancy, lactation, childhood, and adolescence, and child growth and development.

An additional review of systematic reviews was carried out for mercury (Hg) exposure. An analytic framework was developed to specify the toxicant exposure to women during pregnancy or lactation and to children and adolescents, and the outcomes in infants, children, and adolescents (Appendix C).

De Novo Systematic Review

The de novo review began with a scoping review to determine the state of the current literature on specific exposure–outcome relationships. Some of these studies also reported associations of seafood consumption with the studied health outcomes; where such results are available, they are also summarized below.

Scoping Review

To determine the status of the evidence to support a de novo systematic review, a scoping review was conducted that included polychlorinated biphenyls (PCBs), dioxins, polybrominated diphenyl ethers, per- and polyfluoroalkyl substances (PFAS), arsenic, cadmium, lead, selenium, iron, magnesium, zinc, and dichlorodiphenyltrichloroethane (DDT) as exposures through fish consumption. Outcome measures included:

• cognition,
• language,
• motor skills,
• academic performance,
• neurological disorders (ASD, ADHD, seizures, tremors, and gait abnormalities),

• measures of growth and body composition,
• failure to thrive,
• malnutrition and protein deficiency,
• blood pressure, and
• allergy and immune response.

The scoping review was used to identify exposure–outcome pairs with at least three studies to justify a de novo systematic review. Of the 12 exposure categories only two exposure–outcome pairs had three or more articles. These were lead (Pb) exposure in pregnant and lactating women and child developmental domains (four studies) and PCB exposure in pregnant and lactating women and child growth-related outcomes (three studies).

Characteristics of Studies for Lead Exposure Through Seafood Consumption

Table 6-1 describes the characteristics of the identified studies for Pb exposure in pregnant and lactating women and child developmental domains.

Summary of Key Studies for Lead Exposure Through Seafood Consumption

Jeong et al. (2017) assessed associations of heavy metal concentrations in pregnant women with the concentrations in their offspring as part of the Mothers’ and Children’s Environmental Health study of 1,751 pregnant women in Korea. Metal concentrations included Pb, total Hg, and cadmium (Cd) in cord blood during early and

late pregnancy, and in children who were age 2, 3, and 5 years. Concentrations of Pb were lowest in cord blood, highest in children 2 years of age, and thereafter decreased with child age. Overall, the study found strong correlations between the Pb and total Hg levels during late pregnancy with the levels in cord blood. Correlations between the Pb and Hg levels in women during late pregnancy and cord blood with the levels in their children were weak.

Rothenberg et al. (2016) looked for associations of maternal consumption of seafood and Pb levels in rural China and found no significant correlation. All three studies (Jeong et al., 2017; Rothenberg et al., 2016, 2021) examined relationships between maternal Pb level and child development outcomes. Jeong et al. (2017) found no significant association, while Rothenberg et al. (2016) found a significant negative association between maternal Pb level and psychomotor index in the child at 12 months of age. Both Rothenberg studies found negative but nonsignificant associations between maternal Pb level and psychomotor index at 36 months of age. Taken together, the evidence from the three studies showed inconsistency in associations between Pb exposure through seafood consumption and health outcomes. As shown in Table 6-1, all the studies identified had a high risk of bias, which the committee considered when assessing the totality of the evidence on health outcomes related to exposure to metals.

Characteristics of Studies for PCB Exposure Through Seafood Consumption

Table 6-2 shows characteristics of the studies for PCB exposure during pregnancy and child development outcomes.

Summary of Key Studies for Exposure to PCBs

Using data from the Danish National Birth Cohort, Halldorsson et al. (2008) examined the associations between fatty fish intake and plasma PCB levels during pregnancy and the association of maternal PCB levels with fetal growth. The investigators obtained dietary intake data from a food frequency questionnaire and PCB concentrations from blood samples during pregnancy and at delivery. The study further found a positive correlation between total maternal lipid concentration and plasma PCB level, adjusting for plasma lipid concentrations. The study concluded that maternal exposure to PCBs through regular consumption of fatty fish was positively associated with PCB levels and inversely associated with both birth weight and placental weight.

In a study examining associations of maternal seafood consumption during pregnancy and small-for-gestational age (SGA) as a birth outcome, Mendez et al. (2009) obtained data from a population-based longitudinal study (Infancia y Medio Ambiente) in Spain with women enrolled in the first trimester of pregnancy. A food frequency questionnaire was used to obtain data on seafood consumption. Birth weight and lengths of infants and length of gestation were gathered at birth. The study found that higher maternal intake of crustaceans and tuna during pregnancy was associated with significantly greater risk of SGA, whereas there was no association among women who consumed lean fish or other shellfish. Associations of risk of SGA for infants of women who consumed fatty fish were not significant. The results of the study suggest that high maternal consumption of crustaceans and tuna may be associated with increased risk of SGA, although the causal chain that links fish consumption to SGA is unclear.

Miyashita et al. (2015) assessed the effects of in utero exposure to PCBs and methylmercury (MeHg) on birth size in the Japanese population. Levels of PCBs and n-3 LCPUFAs in maternal blood and the total Hg in hair were measured during pregnancy and at delivery. An FFQ, administered at delivery, was used to estimate maternal fish and shellfish consumption. This study found no association of PCB concentration and birth size. The study did find that risk of SGA decreased with increasing MeHg concentration when adjusted for n-3 LCPUFA intake.

Wohlfahrt-Veje et al. (2014) examined associations of dioxin and PCB exposure with insulin-like growth factor 1 (IGF1) levels and growth in early childhood. The investigation used data from the longitudinal Copenhagen Mother Child Cohort of Growth and Reproduction study of 2,098 mother–infant pairs recruited from 1997 to 2001. The study found associations of levels of exposure to environmental dioxin-like chemicals and accelerated infant height and weight gain as well as increased IGF1 concentrations at 3 months of age. Associations between dioxins and weight-for-length at birth were negative but not significant. The investigators concluded that early exposure to dioxins and dioxin-like compounds was associated with accelerated early childhood growth in height and weight, which is further associated with greater risk of adult obesity and associated chronic disease.

Summary of Findings on Exposure to PCBs

Evidence from the four studies reviewed showed a significant positive association between seafood consumption during pregnancy and maternal PCB level. This finding was consistent irrespective of the PCB exposure indicator (i.e., maternal whole blood, plasma, serum, or human milk). Mendez et al. (2009) found a significant positive association between prenatal consumption of crustaceans and odds of an SGA infant. Other associations between seafood consumption by type and birth weight or SGA were nonsignificant. Miyashita et al. (2015), however, found no significant associations between seafood intake during pregnancy and growth outcomes. Among the three studies that examined an association of maternal PCB level and child growth outcomes, the findings were inconsistent (Halldorsson et al., 2008; Miyashita et al., 2015; Wohlfahrt-Veje et al., 2014).

As noted above for Pb exposure, the evidence from these studies also showed inconsistency in associations between PCB exposure through seafood consumption and health outcomes. Table 6-2 shows high risk of bias across all studies except Mendez, which was rated as “some concerns.” The committee considered this inconsistency when assessing the totality of the evidence on health outcomes related to exposure to PCBs.

Review of Systematic Reviews on Maternal Exposure to Mercury

The Texas A&M Agriculture, Food, and Nutrition Evidence Center carried out a review of existing systematic reviews on maternal or childhood exposure to Hg and child growth to determine the state of current evidence on development outcomes not included in the de novo systematic review (see Appendix C). Eleven existing systematic reviews were identified that met the committee’s criteria. No existing systematic reviews were identified for

TABLE 6-3 Systematic Reviews by Prioritized Outcomes and the Overall Quality of Evidence

SOURCE: Texas A&M Agriculture, Food, and Nutrition Evidence Center, 2023.

academic performance or failure to thrive. Table 6-3 shows the studies identified in the initial scoping review by the committee’s prioritized outcomes for exposure to Hg, and the overall quality of the evidence. Among these studies, three (Dack et al., 2022; Eola Tapia et al., 2023; Saavedra et al., 2022) were of moderate to high quality and covered the range of developmental domains included in the search.

Summary of Key Studies on Exposure to Mercury

Dack et al. (2022) conducted a systematic review of evidence for associations of prenatal Hg exposure and early neurodevelopment in children ages 3 days to 59 months. The investigators examined patterns of results by Hg biomarker (umbilical cord blood, whole blood, erythrocytes, and hair), the timing of measurement, and child age. Outcome groupings included cognition and language; attention, executive function, and memory; motor function; communication and language development; and composite measures of neurodevelopmental functioning. The analysis found that, at the levels of Hg recorded in studies reviewed, the evidence for an association of prenatal Hg exposure and neurodevelopmental functioning was weak.

The systematic review by Saavedra et al. (2022) examined the effects of maternal exposure to Hg during pregnancy on health outcomes in utero, and among newborns and children up to age 8 years. The review included studies from Europe, Asia, Africa, North America, and South America, and all were observational cohort studies. Levels of Hg were measured in cord blood, maternal venous blood, maternal hair, and umbilical cord tissue. Health outcomes included both cognitive and physical development. The age at assessment of health effects following in utero exposure ranged from birth to 7 years. The analysis suggested that maternal dietary exposure to Hg has a significant effect on the neurological and physical development of children, although the investigators noted wide heterogeneity among studies.

Ealo Tapia et al. (2023) conducted a systematic review to evaluate the scientific evidence on the effects of Hg exposure during the prenatal and postnatal periods and associations with neurodevelopmental disorders. Of 31 studies reviewed, 18 focused on the effects of Hg exposure on neurodevelopment and 13 on behavioral disorders. Of the 18 studies that assessed effects on neurodevelopment, 4 studies assessed the effect of pre- and postnatal exposure to Hg, 8 assessed prenatal exposure, and 6 assessed postnatal period exposure. The reported effects included association of Hg exposure with impaired cognitive and language functions and development of ASD and ADHD. The primary finding of the review indicates that the current scientific evidence on the effects of prenatal and postnatal Hg exposure and its incidence on neurodevelopmental and neurobehavioral disorders is limited.

Evidence from the Peer-Reviewed Published Literature and Technical Reports

Health Outcomes Related to Exposure to Metal and Metalloid Compounds Through Seafood

In 2014, FDA published an assessment of the overall net effects on the developing nervous system of the fetus from the consumption of commercial fish during pregnancy (FDA, 2014). The approach used was a quantitative risk assessment with the added component of calculating dose–response relationships for both adverse and beneficial effects. The assessment assumed that the beneficial and adverse effects act independently of one another but occur at the same time and included estimates of both population-level effects and individual effects.

The modeling component of the assessment included the net effects of consuming fish during pregnancy on three neurodevelopmental endpoints: (1) full IQ at age 9 years, (2) early age verbal development up to 18 months, and (3) later age verbal development through age 9 years. The model first estimated as a current baseline effect that on a population basis, the average neurodevelopment benefits were approximately 0.7 of an IQ point (95% confidence interval [CI] of 0.39–1.37 IQ points) in the United States from maternal consumption of fish during pregnancy. Because no IQ tests are sensitive to such a small difference, it is debatable whether a 0.7-point IQ difference is significant at the level of the individual. However, the assessment then estimated that optimum amounts of each of 47 individual species and market types sold in the United States could result in full-scale IQ gains in the vicinity of three IQ points depending on species and market type.

Optimum amounts always involved multiple IQ points although fish with lower levels of MeHg tended to produce greater gains than fish with higher levels. The assessment estimated somewhat greater gains for verbal IQ when optimum amounts are consumed and somewhat lesser gains for early age verbal development, again involving multiple IQ points. It further estimated that consumption beyond optimum amounts would result in gradual declines in the sizes of the gains, although the amounts necessary for the gains to dissipate completely or be replaced by net adverse effects were typically beyond what most people eat or could eat.

Collectively, these estimates are consistent with the current state of the evidence: higher fish consumption is associated with lower risk of IQ loss, and there is no evidence of increased risk of net adverse effects in the amounts that have been studied to date. At a population level, however, loss of a single IQ point may be meaningful. Furthermore, economic analyses have estimated each additional IQ point, at the population level, may result in 1.3 to 2.2 percent greater lifetime earnings (Grosse and Zhou, 2021).

The 2014 FDA report summarized the available literature on prenatal exposure to Hg from maternal fish consumption (primarily during pregnancy) and fetal neurodevelopment in children. It also examined postnatal exposure from the child’s consumption of fish or breast milk that was consumed after the mother consumed fish. Whether the net effects of consuming fish were beneficial or harmful appears to depend on the amounts and types of fish consumed during pregnancy. In some studies, methylmercury (MeHg) likely reduced the size of beneficial effects and possibly caused the net effects to become adverse under some circumstances.

From the report’s literature review, two studies from Poland examined exposure to MeHg from prenatal fish consumption and found an inverse association with the Bayley’s Scale at 1 year but not at age 2 or 3 years. Three studies on prenatal fish consumption from the Avon Longitudinal Study of Parents and Children (ALSPAC) cohort in the United Kingdom found fatty fish consumption associated with development of stereoscopic vision (the perception of shape and distance of an object with binocular vision) at 3.5 years. Additionally, consumption of fish was shown to have a positive association with the Macarthur Communicative Development Inventory at age 15 months and with the Denver Developmental Screening Test at 18 months of age. Peak benefits were seen with 1–3 servings/week, with a waning or a plateau in benefits at higher intakes. No association was found with elemental Hg.

Four studies in the ALSPAC review from the United States showed positive associations between maternal fish consumption and multiple neurocognitive tests (including IQ) in children from age 6 months through 8 years. Each additional weekly serving of fish consumed by the mother was associated with a 4-point gain in IQ in the child, and each 1.0 ppm of Hg detected in maternal hair was associated with a loss of 7.5 points on visual recognition memory at age 5.5 to 8.4 months. Overall, fish consumption during pregnancy was associated with improvements

in the Peabody Picture Vocabulary Test and the Wide Range Assessment of Visual Motor Abilities, and MeHg exposure was associated with reductions in those beneficial outcomes.

Among mothers who consumed two servings of fish per week, lower Hg exposure was associated with higher test scores than higher exposures—with both subgroups experiencing net benefits (compared with no fish consumption, with an overall plateau of effects at two servings/week). Maternal blood Hg levels were associated with reduced test scores and maternal fish consumption was associated with higher scores on the Bayley Scales at age 12, 24, and 36 months and with Wechsler Preschool and Primary Scale of Intelligence scores at age 48 months. In older children, higher maternal Hg and lower fish intake was associated with ADHD-related behaviors at age 8 years.

One study from Denmark did not directly measure exposure through maternal fish consumption, but assumed MeHg exposure was low in fish commonly consumed by the population. This study found that higher fish intake compared with no or lower intake during pregnancy resulted in better performance on developmental milestones including sitting unsupported at age 6 months and climbing stairs and drinking from a cup at age 18 months. Another study in Japan found beneficial outcomes with overall fish consumption, and adverse outcomes with higher Hg exposure in the “motor cluster” facet of the Neonatal Assessment Scale in infants at 3 days of age, but no evidence of a plateau associated with maternal fish consumption.

To determine the correlation of seafood intake and adverse cognitive outcomes associated with exposure to Hg through consumption of seafood, Golding et al. (2022) observed several groups of children across time in both a British and Seychelles cohort. From 1990 to 1992, the British Pre-Birth Cohort Study enrolled a geographic population of pregnant women and followed their pregnancies and their offspring throughout childhood, adolescence, and into adulthood. The results of the study found positive indicators of fish consumption among the children of this cohort.

Choi and Grandjean (2008) reviewed evidence for the effects of MeHg on developmental neurotoxicity and risk of heart disease. The review included two major prospective cohort studies from the Faroe Islands and the Seychelles. The primary sources of exposure were freshwater and marine fish. Biomarkers of exposure for MeHg included hair, cord blood, and cord tissue. The Faroe Island study associated prenatal exposure with decrements in attention, language, verbal memory, and motor speed at 7 and 14 years of follow-up. However, it did not find a protective effect of n-3 LCPUFA or selenium on neurocognitive outcomes. In the Seychelles study, no clear evidence was found for associations of prenatal exposure to MeHg and neurotoxicity, although decreased fine motor function was found among those with fetal exposure levels at or above 10 µg/g in maternal hair by age 9 years.

Murcia et al. (2018) conducted a prospective cohort study of pregnant women in four Spanish regions between 2003 and 2008 and followed their children up to 4–5 years. The participants were recruited from the Spanish cohort study Infancia y Medio Ambiente (Environment and Childhood). The study found inverse associations between urinary iodine during pregnancy and both cognitive and motor function at ages 4–5 years, but no association with seafood intake was detected.

In a prospective cohort study conducted with a subset of the ALSPAC cohort, researchers examined the association between fish consumption (assessed via FFQ) among 3-year-olds. Using the SDQ, the authors reported no associations between fish consumption and problems reported on the conduct subscale. Collectively there was no evidence to support an association between fish consumption and social and/or emotional development (Ajmal et al., 2022).

The joint and interactive effects of prenatal exposure to Pb, manganese (Mn), Se, and MeHg on executive function behaviors were evaluated in 1,009 mother–child pairs from the Project Viva cohort in the United States (Fruh et al., 2021). An FFQ was used to assess food and beverage intake in early pregnancy and included questions on fish consumption. Parent and teacher evaluations of social, emotional, and self-regulatory behaviors were also included in the analysis. The elements of interest were measured in blood samples from women collected at the beginning of the second trimester of pregnancy. Executive function behaviors and behavioral difficulties were inferred from the Behavior Rating Inventory of Executive Function and the SDQ. The study did not find evidence of interaction between Pb, Mn, Hg, and Se. Overall, the study did not find strong evidence of adverse effects from exposure to the elements on neurobehavioral outcomes, but the investigators did observe a trend of worsening behavior associated with increasing concentrations of the mixture of elements.

As discussed above, exposure to metals and metalloids in early childhood can have adverse effects on neurodevelopment (Snetselaar et al., 2020a,b). A comprehensive study by Rahbar et al. (2020) examined a panel of metals (Pb, As, Mn, Hg, Cd, and aluminum [Al]) against the generalized Weighted Quantile Sum (gWQS) regression models algorithm developed by Lee et al. (2019) to compare exposure to the metals with each of three glutathione S-transferase (GST) genes (GSTP1, GSTT1, and GSTM1) and their possible interactions in relation to ASD. The analysis included an FFQ to collect information about the type and frequency of seafood consumed weekly by children. Specifically, data were collected on the consumption of sardine, mackerel, tuna, salt fish, shellfish, and shrimp. The investigators then compared their findings against a multivariable Conditional Regression Model. They also evaluated the overall mixture association of metals with ASD based on additive positive and negative gWQS models. Overall, the study found inverse associations of the mixture gWQS index score of the six metals. There were also marginally significant interactions identified between GSTP1 and Pb, Hg, and Mn in a mixture with regard to ASD status, although the interaction was no longer significant in the adjusted model.

Kim et al. (2020) investigated heavy metal exposure from domestic consumption of fishery products and proposed guidelines for the safe intake of seafood products to reduce consumer exposure. In this study, fishery products, including flounder, Pacific cutlass fish, chub mackerel, Pacific cod, Alaska pollock, Pacific saury, bastard halibut, Korean rockfish, Japanese Spanish mackerel, Okhotsk atka mackerel, shrimp, crab, croaker, alfonsino, monkfish, Filipino venous, mussel, squid, long arm octopus, and webfoot octopus were collected from local fish markets in Korea. Exposure assessment was conducted for children (age 1–11 years), adolescents (age 12–18 years), and adults (age 19+ years) by combining the heavy metal contamination values with the fishery product consumption data and dividing by body weight. Consumption data were obtained from the Korea National Health and Nutrition Examination Survey from 2010 to 2015. At the 95th percentile of consumption, squid was the leading Cd contributor, followed by Filipino venues, and chub mackerel was the leading contributor of Hg, followed by Alaska pollock. No species-specific differences were found for Pb exposure.

Chronic Disease Risk Associated with Exposure to Toxicants by Consumption of Seafood

An FAO/WHO (2011) risk–benefit analysis of fish consumption included a review of the evidence on the risk of chronic disease from seafood consumption. This study considered the effect of both nutrients and toxicants in seafood on health outcomes. Based on a review of previous reports and reviews, the authors concluded that seafood consumption lowers the risk of mortality from coronary heart disease in the adult population and little or no evidence exists for other health benefits including ischemic stroke or cancer. This report, however, did not include studies of fetal or childhood exposure.

Summary of Evidence on Health Outcomes Associated with Exposure to Toxicants by Consumption of Seafood

Taken as a whole, the evidence reviewed by the committee indicates that higher fish consumption is associated with lower risk of adverse health outcomes or no association with health outcomes. The evidence for increased risk of adverse health outcomes associated with seafood consumption was insufficient to draw a conclusion.

COMMON MECHANISMS OF ACTION OF CONTAMINANTS COMMONLY FOUND IN SEAFOOD

The committee was asked to consider evidence on how various compositions of a seafood matrix (i.e., the relative proportions and concentrations of nutrients or contaminants present) may function as exposures in the human body that have various mechanisms of action and interplay, including interactions between toxic and nutritive components. While there are many potential contaminants in seafood, the committee identified two common contaminants commonly found in seafood, MeHg and PCBs, for further consideration of seafood matrices.

Mechanism of Mercury-Mediated Toxicity

The mechanism of Hg toxicity is a complex chain of events. It has been postulated that co-exposure to selenium in seafood can modulate the effects of MeHg. For example, Moniruzzaman et al. (2021) conducted a study in mice to evaluate the effect of Se, vitamin C, and vitamin E on MeHg-toxified olive flounder fish muscle powder. The study assessed tissue Hg bioaccumulation, antioxidant enzyme activities, and oxidative stress. Mice were fed diets with three levels of Hg (0, 50, or 500 μg per kg) and two levels of Se in combination with vitamin C and vitamin E (Se: 0, 2 mg per kg; vitamin C: 0, 400 mg per kg; vitamin E: 0, 200 mg per kg). The findings included that dietary supplementation of antioxidants such as Se, vitamin C, and vitamin E had no effect on Hg bioaccumulation in mice. The antioxidants did, however, have a partial effect on reducing serum lipid peroxidation and a significant effect on the cumulative survival rate in the mice fed Hg-intoxicated diets.

Spiller (2018) described the role of selenium as follows.

Reduction of Hg toxicity depends in part on the form of Hg and may be multifaceted and may include:

• facilitating demethylation of organic Hg to inorganic Hg;
• redistribution of Hg to less sensitive target organs;
• binding to inorganic Hg and forming an insoluble, stable, and inert Hg:Se complex;
• reduction of Hg absorption from the gastrointestinal tract;
• repletion of Se stores (reverse Se deficiency); and
• restoration of target selenoprotein activity and restoring the intracellular redox environment (Fernandes et al., 2020).

The key interactions of Hg with Se identified are the sequestering of the chemical form and the counteracting of oxidative stress.

James et al. (2022) showed that for long-term low-level MeHg exposure from consuming fish, Hg speciation in human brain tissue demonstrates that MeHg coordinated to an aliphatic thiolate, resembling the coordinated environment observed in marine fish. The investigators also observed the presence of less bioavailable mercuric selenide deposits using high-energy resolution fluorescence detected X-ray absorption spectroscopy.

The selenoenzyme thioredoxin reductase (TrxR) is a critical component of the thioredoxin system. For example, selenoenzymes restore antioxidants such as vitamin C, thioredoxin, and glutathione to their functionally active forms. Branco and Carvalho (2019), Nogara et al. (2019), and Spiller (2018) reviewed the evidence about the interaction between Hg compounds, the thioredoxin system, and its implications for toxicity development due to co-exposure to selenium and Hg compounds. The TrxRs are a group of Se-containing pyridine nucleotide-disulphide oxidoreductases, whereby the mechanism of toxicity of Hg compounds involves a complex chain of events.

First, inhibition of TrxR, glutathione, and the glutaredoxin system (glutathione peroxidase) can result in the proliferation of intracellular reactive oxygen species. Next, glutamate excitosis, calcium dyshomeostasis, mitochondrial injury/loss, lipid peroxidation, compromised protein repair, and apoptosis can occur. Supplemental Se has been shown to have a protective effect on TrxR from the toxicity of inorganic Hg, but not from MeHg. The inhibition of TrxR can reduce thioredoxin by alternative mechanisms, which involve glutathione and glutaredoxin. If this pathway is compromised, apoptosis can occur. MeHg is a more potent inhibitor of the thioredoxin system, partially explaining its increased neurotoxicity. Whether the amount of selenium in seafood produces a similar response has not been established.

The current evidence shows that the MeHg and Se interaction is complex (Choi et al., 2008). The interactions can result in decreased Hg or Se toxicity, or an increase in Se deficiency. The nature or the interactions depends on several factors such as the form of Hg or Se, the affected organ, and the dose. Other factors have been shown to counteract the effects of Hg and other heavy metals. The nuclear factor erythroid 2 (Nrf-2) regulates cellular resistance to oxidants by controlling the expression of an array of genes dependent on the antioxidant response element. Glutathione (GSH) prevents damage to cellular components caused by reactive oxygen species and heavy metals.

In an investigation into the effect of zinc (Zn) on inorganic Hg-induced cytotoxicity in cultured cells, Hossain et al. (2021) used cellular toxicity assays to show that pretreatment was partially effective in reversing inorganic

Hg-induced toxicity. The mechanism of action was proposed to be through inhibiting formation of reactive oxygen species, GSH increase, and the Nrf-2-mediated pathway. In this model, Zn pretreatment was partially effective in protecting cells from inorganic Hg-induced cytotoxicity and intrinsic apoptosis through its antioxidant properties.

Mechanisms of PCB Toxicity

Yazdi et al. (2021) investigated the effects of PCBs found in food on hormone-responsive and nonresponsive cell lines. In this study, all cell lines were treated with serial concentrations of PCBs for 48 hours and tested for cell viability. The most effective PCB concentration was then applied and levels of paraoxonase-1—an HDL-associated protein that hydrolyzes LDL cholesterol with potential atheroprotective effects—were evaluated. In addition, a molecular modeling technique was used to determine the binding mechanism and predict the binding energies of PCB compounds to the aryl hydrocarbon receptor (AhR). The results showed that the hormone-responsive cell lines tested showed different concentrations of PCBs after treatment. Furthermore, the molecular modeling showed that PCBs had steric interaction with the AhR. The investigators concluded that PCBs have a significant effect on hormone-responsive cells.

Health Effects of Chemical Mixtures in Seafood

The traditional risk assessment approach relies on testing single chemicals across dosage concentrations in animal models or measuring selected biomarkers of exposure for selected chemicals in the studied populations and evaluating their associations with health outcomes (IOM, 2007). Therefore, the traditional risk assessment approach is only useful in considering selection of the nutrients and contaminants that are present in seafood. Moreover, as the relative proportions and concentrations of nutrients or contaminants in seafood vary, it is difficult to assess the cumulative effects or risks of the combined “chemical” exposure from seafood consumption using traditional approaches. There have been many recent efforts to enhance contaminant mixture modeling in epidemiological research (Yu et al., 2022) and efforts to enhance in silico prediction models (Gao et al., 2023). While a research priority, formal methods to assess risk from joint exposure to chemical mixtures have yet to be implemented (Nikolopoulou et al., 2023).

As reported in Liew and Guo (2022), six varying concentrations of an endocrine-disrupting chemicals (EDCs) mixture were identified in a cohort of Swedish mother–child pairs. These EDC mixtures were tested in multiple experimental models, including human cerebral organoids, Xenopus laevis tadpoles, and zebrafish larvae. Gene networks that were altered by the EDC mixture were identified in human cerebral organoids, and validated thyroid, estrogen, and peroxisome proliferator-activated receptor endocrine pathways as major convergent targets in vivo. The investigators estimated that about half of the children in the cohort had been prenatally exposed to EDC mixture concentrations that could cause biological effects, such as language development (Liew and Guo, 2022).

Summary of Evidence on Mechanisms of Action and Contaminants Found in Seafood

Some experimental evidence supports that the toxicity of mercury and PCBs can be modified by other factors (i.e., in antioxidant response pathways). This literature is complex, and the committee was not able to identify supportive evidence in humans.

FINDINGS AND CONCLUSIONS

Findings

Outcomes Associated with Seafood Consumption

1. Many of the studies reviewed by the committee reported outcomes correlated with seafood consumption generally and without differentiation as to species. One commonly accepted assumption has been that

omega-3 long-chain fatty acids in seafood, particularly DHA, contribute benefits, possibly in combination with other nutrients.

1. The evidence reviewed indicates that some gains in neurodevelopment may be achieved during childhood and are apparent in the children of women who consume greater quantities of seafood during pregnancy compared to those who consume lower quantities or no seafood.
2. Seafood consumption by women during pregnancy may also have a protective effect against adverse neurocognitive outcomes in their children that is linked to the nutrients in seafood, particularly the n-3 long-chain polyunsaturated fatty acids that are essential to brain development. Associations of health outcomes with seafood intake differ between the general populations and recreational and subsistence fishers.

Conclusions

1. The results of the studies identified through the committee’s evidence reviews did not support a beneficial association of seafood consumption during childhood and adolescence and reduced risk of cardiovascular disease.
2. No evidence was identified to determine whether seafood consumption among children or adolescents is associated with benefits to reducing risk of other diseases such as immune disease.
3. The evidence reviewed on health outcomes associated with seafood consumption for women of childbearing age, children, and adolescents is not adequate to support an accurate assessment of the health benefits and risks associated with meeting the recommended intakes of seafood for this population group.

RESEARCH GAPS

• Additional research is needed to assess childhood health outcomes related to seafood consumption by children. This should include not only amounts and types consumed but also the age of introduction of seafood to infants and children.
• Additional research is needed to determine whether there are sensitive periods in child development during which seafood consumption or exposure to contaminants in seafood might have different effects on child health.
• Population studies that examine the effects of maternal and child seafood consumption on child health outcomes need to better characterize the seafood species (e.g., type of fish, source, and location) as well as nutrient composition and contaminant concentrations in the seafood consumed.
• Additional research is needed on the health effects of contaminant mixtures and varied exposure levels to determine applicability for these observations in seafood-consuming populations in the United States and Canada.
• Additional research is needed to assess how to effectively communicate seafood consumption recommendations to women of childbearing age, children, and adolescents.

REFERENCES

This page intentionally left blank.