Food Sources of EPA and DHA in the Diets of American Children, NHANES 2003-2010 BAOJ Nutrition

Objective Dietary eicosapentaenoicacid (EPA) and


Introduction
The human body has the ability to elongate and desaturate alphalinolenic acid (ALA) to the long chain omega-3 polyunsaturated fatty acids (n-3 PUFAs) eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) [1][2][3]. However, Western diets are low in n-3 PUFAs, specifically ALA found in plant oils and EPA and DHA found in oily fish [4][5][6]. In the United States (US), the mean intake of n-3 PUFAs is 0.7% of total energy consumed [7], and intake in adults is very low with a median intake of EPA and DHA only contributing 0.05% of total dietary energy [7,8]. Even individuals with low fish consumption have greater n-3 PUFA status than those individuals who do not consume any fish [9].
Western diets low in n-3 PUFAs and high in n-6 PUFAs contribute to poor brain development and function [4,10]. It is uncertain if the rate of DHA synthesis in the human body is sufficient to support optimal brain and retinal development, and these PUFA swould ideally be provided in the diet [11]. DHA is the major n-3 PUFA esterified in glycerophospholipids through the action of acyl-CoA synthases and acyl-CoA:lysophospholipid acyltransferases [12], that forms the structural matrix of brain grey matter and retinal membranes [13,14]. In addition, dietary intake of n-6 PUFAs play a role since n-6 PUFAs interact and compete with n-3 PUFAs in the fatty acid metabolic pathway [15][16][17][18][19][20][21]. In the National Health and Nutrition Examination Survey (NHANES) 1999-2006, a food frequency questionnaire (FFQ) specifically designed to assess habitual fish consumption over a 30day period was used. Tran et al. [22] found that the three most frequently consumed fish are tuna, salmon, and breaded fish. According to NHANES 2003NHANES -2008 dietary intake data, the food groups that contributed the greatest amount of DHA to the diets of American children were fish and shellfish (46.3%), poultry and poultry dishes (24.5%), eggs and egg dishes (19.6%), and pasta, rice and other grain dishes (2.2%) [23]. In children of all ages, racial groups, and ethnic groups (Non-Hispanic white or black and Mexican American) examined, the type of fish children most frequently consumed was white fish, followed by shellfish, and then oily fish. Young children in the US had the lowest reported EPA (6 mg/d) and DHA (20 mg/d) intake compared to children in Australia, Belgium, Canada, and China (mean ranges of 17-60 mg/d EPA and 23-96 mg/d DHA) [23,24]. At the same time, US children had a greater mean intake of total n-6 PUFAs (8.6 g/d) compared to Australia, Canada, and China (6.20-7.60 g/d, 7.40-7.74 g/d, and 2.14-2.32 g/d, respectively) [23].
Multiple factors influence the n-3 PUFA content of the natural dietary sources of EPA and DHA. Even within each food type, variations in fatty acid content are based on factors, such as location, season, water temperature, age, sex, and diet [25]. Other natural sources of EPA and DHA are fatty animal tissue (with fatty fish being the best dietary source) [4], products of select marine sources (e.g. fish oil, krill oil, and algal oil) and other meats (e.g. poultry, eggs, beef, and pork) [26][27][28]. In addition to the natural food sources of n-3 PUFAs, numerous alternative sources are available in the US food market. Food production companies incorporate n-3 PUFAs into breads and pastas, milk, eggs, processed meats, salad dressings, margarine, mayonnaise, peanut butter, pizza, nutrition bars, cereal, yogurt and juices [27][28][29][30][31]. Incorporating n-3 PUFAs into commonly consumed foods (that do not naturally contain n-3 PUFAs) provides a cost-effective, sustainable venue to increase n-3 PUFA consumption, specifically EPA and DHA [32,33]. Although the ALA, EPA, and/or DHA used in the fortification of foods is primarily from plant (ALA) or marine sources (EPA and DHA), the bioavailability and usefulness of these forms is currently understudied.
The 2010 Dietary Guidelines for Americans recommend that Americans consume seafood at least twice a week with "an intake of 8 or more ounces per week (less for young children)" that will provide a mean of 250 mg per day of EPA and DHA [34]. Furthermore, the Joint Food and Agriculture Organization (FAO) and World Health Organization (WHO) Expert Consultation on Fats and Fatty Acids in Human Nutrition provided recommendations for adequate intake levels of EPA and DHA for children and adults [35]. Although national and international public health organizations, professional organizations, and expert committees have made dietary recommendations for fish consumption or EPA and DHA intake [8,[36][37][38], to date, the Institute of Medicine has not established dietary reference intakes (DRIs) for EPA and DHA [39]. To develop effective public health nutrition policy, more specific EPA and DHA target intake levels (i.e. DRIs for EPA and DHA) must be established. There is a critical need for public health guidance on EPA and DHA intakes because substantial evidence indicates that these dietary fatty acids have many health benefits [39] and play important roles in heart health [40], brain [11,[41][42][43][44][45][46][47] and eye development [48][49][50][51].
The overall goal of this project was to describe the reported consumption levels and food sources of fish and shellfish as well as EPA and DHA reported intake in the diets of American children, thereby advancing public health policy.

Methods
Dietary intake and socio-economic data including sex, race/ ethnicity, and poverty income ratio (PIR) ( Table 1) from children (n=13,441) ages 2-18 years in the NHANES 2003-2010 were extracted. PIR is the ratio of family income to the appropriate poverty threshold [52]. Ratios below 1.00 indicate that the family income is below the poverty threshold whereas a ratio of 1.00 or greater indicates income above the poverty threshold. Households with a PIR <1.3 are eligible for the United States Department of Agriculture (USDA) Supplemental Nutrition Assistance Program (SNAP); households with a PIR ≤1.85 PIR are eligible for participation in the USDA Women, Infants, and Children (WIC) Program; households with a PIR of 1.86-3.49 are defined as medium income; households with a PIR of 3.50-5.00 are defined as high income with all values >5.00 truncated to five [53].NHANES data are publicly available, de-identified data, thus, this study was "exempt" by the Institutional Review Board for Human Research.

Nutritional Variables
To code the nutrient and food-level data intakes, the USDA's Food and Nutrient Database for Dietary Studies (FNDDS), 5.0 (2012) [54] was used. The USDA food coding system was used to group foods into foodgroups with shared characteristics, such as "fish and shellfish", "meat, poultry, and fish with non-meat items", or "frozen and shelf-stable plate meals, soups, and gravies". Dietary intake data was provided by 24-hour recalls, which were used to rankorder (highest to lowest) the foods reportedly consumed by 2-18 year old children by EPA and DHA density (mg/g of food) and to calculate average EPA and DHA intake amounts. The nutrients EPA and DHA were analyzed as combined variable (EPA+DHA). Since the goal of this study was to examine the nutrients critical for brain and eye development, EPA and DHA content of food and their consumption levels were combined to account for the bioconversion from EPA to DHA in the human body. Consumption of EPA and DHA was examined to determine tertiles of the total diet EPA and DHA density (mg/100g of food consumed) to identify and describe three levels of EPA and DHA consumers. Food-level data was analyzed using three approaches: a) the density of EPA and DHA in the foods (mg/100g of food), b) mean intake of EPA and DHA from foods (mg/d and number of children reportedly eating the food), and c) food list of all foods containing at least some EPA and DHA (more than zero mg/100g of food) and the number of children eating the food. To assess the effect of fish and shellfish intake on EPA and DHA intake, the responses to the FFQ assessing habitual fish intake were used to discern those children who ate fish or shellfish in the past 30 days from those who were classified as "not fish or shellfish consumers".

Statistical Analysis
All analyses were corrected for survey design and weighted (using the standard one-day dietary weight) to maintain the nationally representative character of the data and conducted in SAS V9.3 (SAS Institute Inc., Cary, NC, USA). Proc Survey Freq in SAS was used to estimate the percentage of the population in each race/ethnic, PIR, and sex category and to estimate the proportion of children who ate fish or shellfish in the past 30 days (FFQ responses). To determine the foods with the highest EPA and DHA density, all reported food items were identified and rank-ordered descending order of EPA and DHA density (mg/100g of food). Total EPA and DHA intake was calculated by multiplying EPA and DHA density with amount of food eaten. The mean intake of each food item and each food group was calculated. Also, the number of children reporting each food item with EPA or DHA was ascertained. Consumption data are reported as mean ± standard error or as percentages.

Results
This study included a nationally representative sample of children whose sociodemographic characteristics reflect the demographic profiles of the American population with approximately 50.7% of the children male, 60.2 non %Hispanic white, and 32.0% from a household with a PIR < 1.30 ( Table 1). The foods with the highest EPA and DHA density (mg/100g food) reportedly consumed by children ages 2-18 years old are reported in (Table2) and separated by age group (Supplemental Table 1). To directly compare the EPA and DHA density of food items consumed by age group, refer to Figures 1-4. Analysis showed that 19 of the top 20 foods with the highest EPA and DHA density were from the fish and shellfish food group, with the exception of cooked brains (consumed by one child); the greatest densities were found in sturgeon roe, baked/ broiled mackerel, and skinless, boneless, water-packed sardines. Analysis of the top 20 food source of the highest contributors to dietary EPA and DHA in 2-18 year olds (Table 3) showed that all of those foods were fish or shellfish and were eaten by only small numbers of children, i.e. only three children consumed sardines or cooked salmon, one child ate fried carp, but 16 children ate battered and fried fish (fish not specified) and 34 children consumed baked or broiled salmon. The data are presented by age group (2-5, 6-11, and 12-18 year olds in Supplemental Table 2).
Overall, the mean EPA and DHA intake was 48 ± 0.002 mg/d. This value varied by age group with 37 ± 0.002 mg/d, 45 ± 0.003 mg/d, and 58 ± 0.003 mg/d of EPA and DHA in 2-5, 6-11, and 12-18 years old, respectively [ Figure 5]. However, when children not consuming dietary EPA and DHA were excluded from the calculation, the mean daily intake of EPA and DHA increased to 59 ± 0.002 mg/d for 2-18 year old children, 44 ± 0.003 mg/d for 2-5 year old children, 53 ± 0.004 mg/d for 6-11 year old children, and 72 ± 0.004 mg/d for 12-18 year old children. Furthermore, when only those children who were identified as "fish and shellfish eaters" using the FFQ were included in the calculations, the mean EPA and DHA intake further increased to an average of 64 ± 0.003 mg/d for 2-18 year olds (62.0% consumed seafood) and 49 ± 0.004 mg/d for 2-5 year olds (63.4% consumed seafood), 61 ± 0.006 mg/d  When EPA and DHA from all sources were considered, the tertiles of total dietary EPA and DHA density in the foods were <5 mg/100g in the lowest tertile, 5-21 mg/100g in the medium tertile, and >21 mg/100g in the highest consumption tertile.The USDA food groups contributing most to the daily EPA and DHA intake (mg/d) were calculated and reported in (Table 4), it at least one of the food items in each food group was reportedly consumed and if at least 15 children reported eating food items from this food group. Fish and shellfish contributed the most EPA and DHA, followed by "meat, poultry, and fish with non-meat items", and "frozen and shelf-stable plate meals, soups, and gravies". Results  Considering EPA and DHA intake from all sources, the food items that provided the most EPA and DHA were "skinless, boneless,  Data are presented as mean intake of EPA and DHA intake in mg/d, rank-ordered by EPA and DHA intake with n = number of children reportedly consuming food water-packed sardines", "cooked salmon", and "floured/breaded fried carp". When examining all foods that were reportedly consumed by the children and that contained any EPA and DHA, non-fish items emerged as most consumed contributors to dietary EPA and DHA intake, i.e. ice cream, salty snacks, and eggs ranked at the top (Table 5). After expanding the list and examining the 50 foods that contributed the most EPA and DHA to children's diets, tuna salad was the only fish/shellfish-based food and ranked in the 38 th place. The tuna salad, commonly assumed to be a regularly consumed food, was eaten by only 127 children and provided on average 167.6 mg of EPA and DHA per day.

Discussion
This study provides new data on fish and shellfish as well as EPA and DHA consumption of American children ages 2-18 years old. Results showed that the majority of children had at least some dietary EPA and DHA intake. Food sources of EPA and DHA included nutrient dense foods, such as fish and shellfish, but also high fat, sweet desserts and salty snacks (i.e. ice cream and chips). On average, only 35.9% of the children consumed shellfish and 49% consumed fish;with increasing age, more children ate shellfish Number of children reportedly consuming at least one food item from the food group. Food groups with less than 15 children reportedly consuming at least one food item from the USDA food group are not shown.
but less children had fish.
The foods with the greatest EPA and DHA density (mg/100g food) that were consumed by any 2-18 year old child were roe, mackerel, and sardines and other fish and shell fish, however these EPA and DHA-rich foods were not commonly consumed by the majority of children. This might be due to barriers such as lack of food access (i.e. ability to obtain or retrieve food from local stores), lack of availability (i.e. the quantity and quality of food that is provided by caretakers and caretakers' knowledge and ability to prepare fish) [55], and the social context in which the food is encountered such as role models not eating seafood and/or habitual consumption i.e. the children are not used to the odor or flavor of these food sources [56].
The weighted mean EPA and DHA in takes per day were 37 mg, 45 mg, and 58 mg for 2-5, 6-11, and 12-18 year olds, respectively -much below the intake recommendations by the Joint FAO and WHO Expert Consultation on Fats and Fatty Acids in Human Nutrition Adequate Intake levels of 100-150 mg for 2-4 year olds, 150-200 mg for 4-6 year olds, 200-250 mg for 6-10 year olds, and 250-2000 mg for people 10 years and older [35] as Figure 5 shows.
This finding is alarming since, especially in younger children, EPA and DHA are critical nutrients to promote healthy brain and eye development and function. Based on the food sources identified in this study, the dramatic under consumption of dietary EPA and DHA may be due to the low proportion of children consuming fish and shellfish, which were the most EPA-and DHA-dense foods consumed by the children. The proportion of children consuming these foods was 1: 34 children or < 0.3% of the sample population (N=13,441).
The majority of children were not consuming the foods highest in EPA and DHA and further analysis showed that consumption of foods contributing any EPA and DHA was too low to meet the intake recommendations, i.e. total daily contribution of ice cream was 3.6 mg (which was the single most consumed EPA and DHAcontaining food) and that of egg omelet or scrambled egg was 39.0 mg (highest contribution to daily EPA and DHA intake but only consumed by 537 children). Tuna salad, commonly served in childcare centers and schools, ranked only thirty-eighth place and contributed small amounts of EPA and DHA to children's diets.
Not surprisingly, fish and shellfish was the food group contributing the largest amount of EPA and DHA to 2-18 year olds diets. However, as the analysis of consumption amounts and frequency showed, very few children consumed these foods (0.3% of the population).
Thus, it appears that an important public health policy should be to increase the proportion of the pediatric population consuming these foods. Previous research to examine acceptability of fatty fish meals in 2-5 year old children has shown that young children accepted oily fish when it was incorporated into familiar foods such as macaroni-and-cheese and wraps [57], thereby significantly increasing EPA and DHA intakes.
The present study had several strengths and limitations. The strengths include, but are not limited to, the large sample size and representation of the US population, the use of validated instruments to estimate EPA and DHA intake, and the pooling of four NHANES survey years (2003-2004, 2005-2006, 2007-2008, and 2009-2010). One limitation of the analysis was that the dietary intake estimates were based on one single 24-hour dietary recall. The 24-hour dietary recall is one of the most commonly used dietary intake estimation methods, especially in large nutrition surveys and has been validated for adults and children in the US  [58-61] because 24-hour dietary recalls provide comparable data to other methods and with reduced effort and cost [62]. However, one day of intake does not provide a good estimate of usual intake especially of rarely consumed foods such as fish and shellfish, and children's diets might not be accurately captured in a single 24hour recall. However, some of this variation was accounted for by the use of the FFQ assessing fish and seafood consumption over a 30-day period.

Conclusions
This study provides important information on the food sources and the dietary intake of EPA and DHA, which are important essential fatty acids for brain and eye development and function, in 2-18 year old American children. Results indicate that fish and shellfish provide the highest EPA and DHA density, but are not commonly consumed foods. Accordingly, estimated average EPA and DHA intake amounts were dramatically below the current intake recommendation levels. Higher consumption levels of EPA and DHA would be beneficial and should be recommended, however, as data from this study show, fish and shellfish are not well accepted in children ages 2-18 years old. Future research should investigate the barriers (e.g. fish consumption advice, availability of and accessibility to oily fish, preparation ability by parents, parental eating habits, and cost) to elucidate why so many children are not consuming fish and shellfish. In addition, healthy alternative dietary sources of EPA and DHA may have to be explored.