We analyzed data on 7,108 women from 1999 to 2006: 1,260 pregnant women and 5,848 non-pregnant women who had completed interviews, exams, and valid pregnancy data. There were statistically significant differences in the distribution of pregnant and non-pregnant women by age group (p≤0.0001), with over 91% of pregnant women younger than 36 years of age, compared to approximately 53% of non-pregnant women in those age groups (Table 1). There were also significant differences by race/ethnicity (p≤0.0001), with a greater percentage of pregnant women compared to non-pregnant women being Mexican American, approximately 16 and 9%, respectively (Table 1). There were no significant differences by pregnancy status in the prevalence of fish, shellfish, or seafood (fish and/or shellfish) consumption over the previous 30 days (all p>0.05). Fish consumption was more commonly reported than shellfish consumption, for both pregnant women (63.0 and 46.7%, respectively) and non-pregnant women (68.4 and 51.2%, respectively). Based on data from the Day 1 dietary recall, daily mean grams of seafood intake, among women who reported seafood consumption in the last 24 h, was similar for both pregnant and non-pregnant women (110.4 and 117.6 g, respectively) (p=0.69).

Any seafood consumption in the 30 days prior to the MEC examination was significantly associated with increasing age for both pregnant and non-pregnant women (Table 2). Pregnant women 36–49 years of age were five and a half times more likely to have consumed fish or shellfish in the last 30 days (95% CI: 1.55, 19.52) compared to pregnant women aged 16–25 years; the corresponding association among non-pregnant women was less than two-fold (Odds Ratio [OR]: 1.98, 95% CI: 1.68, 2.33). Among non-pregnant women, non-Hispanic black race/ethnicity was significantly associated with increased odds of seafood consumption, compared to non-Hispanic white race/ethnicity (OR: 1.75, 95% CI: 1.33, 2.31). In addition, education greater than high school (OR: 1.80; 95% CI: 1.49, 2.18, reference less than high school graduate) was also associated with seafood consumption among non-pregnant women. Although similar patterns were seen among pregnant women for education, these associations were not statistically significant. There were no significant associations with poverty income ratio.

For both pregnant and non-pregnant women, those who were older, had higher education, and higher poverty income ratio were more likely to have reported consuming three or more servings of seafood in the past 30 days. Also for both pregnant and non-pregnant women, Non-Hispanic whites (53.3 and 55.8%, respectively) and Non-Hispanic blacks (53.0 and 63.6%, respectively) were more likely to report consuming three or more servings of seafood in the last 30 days compared to Mexican–Americans (38.9 and 48.2%, respectively) (Table 3). Non-pregnant women in the ‘other/multiracial’ category were also more likely to report higher seafood consumption (59.6%) compared to Mexican–Americans; sample size did not permit an estimate for this racial/ethnic group among pregnant women (Table 3).

Women who reported seafood consumption were then asked to report on how many occasions during the past 30 days they consumed specific types of fish and shellfish. Tuna, salmon, and catfish were the most commonly consumed fish, and shrimp and crab were the most commonly consumed shellfish among both pregnant and non-pregnant women (Table 4).

We examine trends in fish and shellfish consumption from 1999 to 2006 and found that trends appear stable over time for both pregnant and non-pregnant women (p _trend= 0.41 and 0.68, respectively) (Fig. 1).

We observed no significant differences in the prevalence, amount, or type of consumption of fish and shellfish, separately or combined, between pregnant and non-pregnant women. However, we observed that among both groups of women substantial proportions are not consuming any seafood, which may be particularly important during pregnancy. NHANES data do not allow us to assess potential motivations for lack of seafood intake, but a contributing factor may be concern about methylmercury contamination, particularly for pregnant women. A recent qualitative study of fish consumption during pregnancy found that women reported many barriers including not remembering which fish types were better to eat during pregnancy, advice to avoid fish, as well as perceiving fish to be costly (20). Some of motivations for fish consumption in pregnant women included having a portable list of fish types and having their obstetricians advised them to eat fish (20).

Fetuses are a high-risk group for methylmercury exposure because of the increased susceptibility of the developing brain to this exposure; however, studies examining the associations between methylmercury exposure and children's neurodevelopment have had inconsistent findings (26, 27). In some of the studies assessing the impact of methylmercury on children's health, detrimental associations between prenatal mercury exposure and the neurological development of the children have been observed (13–15), while others have reported no significant associations between methylmercury exposure and adverse outcomes in children (16, 17). Measures of neurodevelopment studied included decreased physical activity levels, loss of IQ points, and decreased performance on standardized tests, including those assessing memory, attention, language, and spatial cognition. Given that permanent damage to the developing brain can possibly occur with methylmercury exposure, the FDA and EPA have issued advisories on seafood consumption for pregnant women and women of childbearing age (18). Although swordfish, king mackerel, and shark are recommended by the FDA (18) and EPA (19) to be avoided during pregnancy, there were pregnant women in NHANES that reported consuming these types of fish, although the number was small.

Although the FDA advisory clearly states the types of fish that should be avoided, it does not provide a comprehensive list of fish that are considered safe to be consumed by pregnant women. For example, the advisory states ‘Eat up to 12 ounces (two average meals) a week of a variety of fish and shellfish that are lower in mercury’ and lists five of the most commonly eaten fish that are low in mercury (18). A qualitative study of fish consumption during pregnancy found that many pregnant women knew that fish might contain mercury and had received advice on limiting fish consumption but had not received advice on the types of fish that are safe to consume (20). Therefore, many of the women avoided fish consumption altogether due to advice to limit intake of certain fish intake combined with a lack of information about which fish types they should be eating.

The same qualitative study of fish consumption during pregnancy revealed that women were less likely to know that fish contains DHA or what function DHA served, than they were to know about mercury contamination in seafood. Most women surveyed had not received information about which type of fish contains more DHA or less mercury (20). The DGA 2010, from the US Department of Agriculture (USDA) and the US Department of Health and Human Services (HHS), recommends consumption of at least eight ounces (227 g) of variety of seafood per week for the general public, which corresponds to intake of an average of 250 mg/d of fatty acids, including DHA; and 8 to 12 (227 to 340 g) ounces of variety of seafood per week for pregnant and breastfeeding women (12). As we reported earlier, we found that daily means grams of seafood intake was similar for both pregnant and non-pregnant women (110.4 and 117.6 g, respectively), among women who reported seafood consumption in the last 24 h.

According to the USDA, the average intake of seafood in the US is low, around ~85–113 g/week. In order to achieve the recommended amounts of seafood intake, Americans would, on average, have to double their consumption of seafood (28). The DGA 2010 recommendations are based on fatty acid intake from seafood and other dietary sources and not from supplementation. Although DHA can be consumed via supplementation, data from at least one study suggests that omega 3 (N-3) intake from oily fish is better incorporated into plasma lipids than from supplemental forms (29). A recent study found that blood mercury concentrations were low and significantly below the reference dose set forth by the EPA in both pregnant and non-pregnant women in the US (30). Given that blood mercury concentrations are low in addition to that seafood consumption is lower than recommended in the US population, including among pregnant women, better communication of the benefits of seafood overall and in particular DHA is warranted. A study on fish consumption among childbearing age women and risk-benefit analysis on neurodevelopment of the fetus concluded that food interventions and advisories on fish consumption should focus more on promoting intake of fish species with a high DHA content and avoiding those with high methylmercury content (31). In addition to the benefits from DHA content in fish, fish in considered a healthy food and is rich in vitamins, minerals, and proteins (32).

Strengths of our study include its large sample size and detailed data on the types of fish consumed. In addition, the oversampling of pregnant women during the 1999–2006 NHANES gave us sufficient power to stratify our analysis and make comparisons between pregnant and non-pregnant women. Additionally, the large sample size as well as multiple years of data allowed us to examine trends in seafood consumption over 8 years. Furthermore, our findings of lower overall seafood consumption are generalizable to the population of the US given that NHANES is a nationally representative sample of the US non-institutionalized population. There are also several limitations of the NHANES data on fish consumption. Self-reported data on fish and shellfish consumption is subject to misreporting; therefore, women may underestimate or overestimate their fish and shellfish consumption. Although the NHANES questionnaire on fish and shellfish consumption is detailed, some of the fish that were mentioned in the FDA advisory were not specifically queried in the 30-day questionnaire, for example tilefish and king mackerel. Also, information on portion sizes as well as preparation is not collected in the 30-day questionnaire. The FDA advisory recommends consumption of canned tuna due to lower concentrations of mercury, but NHANES only asks about tuna and whether it is canned or fresh tuna is not known. We attempted to address some of these limitations by using data from the first 24-h dietary recall.

Our findings complement those of other studies on seafood consumption among pregnant women and women who may become pregnant. More detailed information about health benefits and risks as well as types of seafood that are safe for consumption should be provided to pregnant women.