BMJ Open Is Committed to Open Peer Review. As Part of This Commitment We Make the Peer Review History of Every Article We Publish Publicly Available

BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected] http://bmjopen.bmj.com/ on September 24, 2021 by guest. Protected copyright. BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from

Omega-3 Fatty Acid Serum Levels Across Life Stages in the United States: A Cross-sectional Analysis of NHANES 2011- 2012 ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2020-043301

Article Type: Original research

Date Submitted by the 04-Aug-2020 Author:

Complete List of Authors: Murphy, Rachel; BC Cancer Research Centre; The University of British Columbia, School of Population and Public Health Devarshi, Prasad; Pharmavite LLC Ekimura, Shauna; Pharmavite LLC Marshall, Keri; Pharmavite LLC Hazels Mitmesser, Susan; Pharmavite LLC, Science and Technology

PUBLIC HEALTH, NUTRITION & DIETETICS, Coronary heart disease < Keywords: CARDIOLOGY, EPIDEMIOLOGY

http://bmjopen.bmj.com/ on September 24, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 39 BMJ Open

1 2 3

4 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 5 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the terms applicable for US Federal Government officers or employees acting as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author Forunless you peer are acting as review an employee on behalf only of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 24, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Omega-3 Fatty Acid Serum Levels Across Life Stages in the United States: An Analysis of 4 5 6 NHANES 2011-2012 7 8 9 10 11 Rachel A Murphy1,2, Prasad P Devarshi3, Shauna Ekimura3, Keri Marshall3, Susan Hazels 12 13 14 Mitmesser3 15 16 For peer review only 17 18 19 1 th 20 Cancer Control Research, BC Cancer, 2-107, 675 W 10 Ave, Vancouver, BC, V5Z 1L3 21 22 2School of Population and Public Health, University of British Columbia, 2206 East Mall, 23 24 25 Vancouver, BC, V6T 1Z3, Canada 26 27 3 28 Science & Technology, Pharmavite LLC, West Hills, CA 91304, USA 29 30 31

32 http://bmjopen.bmj.com/ 33 Corresponding Author: 34 35 Rachel A Murphy, PhD 36 37 38 167-2206 East Mall 39

40 on September 24, 2021 by guest. Protected copyright. 41 University of British Columbia 42 43 Vancouver, BC 44 45 46 V6T 1Z3 47 48 49 Email: [email protected] 50 51 Phone: 604-822-1397 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 Abstract 12 13 14 Objective: To determine reference ranges of circulating levels of eicosapentaenoic acid 15 16 For peer review only 17 (EPA), docosahexaenoic acid (DHA) and the sum total of omega-3s 18 19 (EPA+docosapentaenoic acid; DPA+DHA) in a nationally representative population of 20 21 22 Americans. To provide context, serum levels of omega-3, a biomarker of omega-3 23 24 25 status, were compared to levels associated with consuming the recommended amount 26 27 of EPA and DHA by the Dietary Guidelines for Americans (DGA) and the Omega-3 Index 28 29 30 (EPA+DHA). 31

32 http://bmjopen.bmj.com/ 33 Design: Cross-sectional population-based study 34 35 Setting: The National Health and Nutrition Examination Survey (NHANES) 2011-2012 36 37 38 cycle 39

40 on September 24, 2021 by guest. Protected copyright. 41 Participants: Participants who had fatty acids measured in serum: 945 children, age 3- 42 43 19 years, and 1,316 adults, age 20 and older. 44 45 46 Main measure: Serum EPA, DPA, DHA and omega-3 fatty acids expressed as % of total 47 48 49 fatty acids. 50 51 Results: Among children, mean (SE) serum levels of EPA, DHA and omega-3s were 52 53 54 0.28% (0.01), 1.07% (0.02) and 1.75% (0.03). Among adults, mean (SE) of EPA, DHA and 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 omega-3s were, 0.61% (0.02), 1.38% (0.05), and 2.43% (0.08), all of which were 4 5 6 significantly higher than corresponding serum fatty acid levels in children (P<0.001). 7 8 9 Despite recommendations for higher intake, pregnant and/or breastfeeding women had 10 11 mean (SE) EPA, DHA and omega-3 levels of 0.34% (0.07), 1.52% (0.08) and 2.18% (0.15) 12 13 14 which were comparable to women of childbearing age; p=0.17, p=0.10 and p=0.73. 15 16 For peer review only 17 Over 95% of children and 68% of adults had omega-3 levels below those associated 18 19 with the DGA recommendation. Approximately 89% of adults had an Omega-3 Index in 20 21 22 the high cardiovascular risk category. 23 24 25 Conclusions: Continued emphasis on regular consumption of omega-3s is needed 26 27 given the high prevalence of suboptimal levels among Americans, particularly 28 29 30 considering the importance of omega-3s in cardiovascular health and brain 31

32 http://bmjopen.bmj.com/ 33 development. 34 35 Article Summary 36 37 38 Strengths and limitations of this study 39

40 on September 24, 2021 by guest. Protected copyright. 41  This study uses data representative of the US population, age 3 and older to provide 42 43 reference information on circulating levels of omega-3 fatty acids, a serum 44 45 46 biomarker of omega-3 nutritional status 47 48 49  Strengths of the study include the study design which allows inference to the 50 51 national-level, inclusion of multiple life stages (children, adolescents, adults, seniors 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 and pregnant or breastfeeding women) and serum biomarker measurement of 4 5 6 omega-3 fatty acid intake 7 8 9  Limitations of the study include the lack of information on long-term omega-3 fatty 10 11 acid intake, and possible shifts in omega-3 status since the time of measurement in 12 13 14 2011-2012 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

40 on September 24, 2021 by guest. Protected copyright. 41 42 43 Background 44 45 46 47 Consumption of seafood rich in long-chain omega-3 polyunsaturated fatty acids 48 49 50 (omega-3 PUFAs), from seafood and dietary supplements, is recommended as part of a 51 52 53 healthy diet (1,2). A large body of evidence has shown that dietary patterns including 54 55 high consumption of the omega-3 PUFAs eicosapentaenoic acid (EPA) and 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 docosahexaenoic acid (DHA), both from diet and supplementation, are associated with 4 5 6 decreased risk of cardiovascular disease (3), type 2 diabetes (4,5), lower overall mortality 7 8 9 (6), and lower blood pressure (7,8). Strong evidence also supports a positive relationship 10 11 between maternal dietary intake of EPA and DHA and neurodevelopment among infants 12 13 14 (9,10). Further research also supports the role of EPA and DHA in neuropsychological 15 16 For peer review only 17 health, such that EPA and DHA supplementation is considered as a part of integrative 18 19 therapies for depressive disorder by the American Psychiatric Association (11). 20 21 22 23 The most recent Dietary Guidelines for Americans (DGA 2010-2015 and 2015- 24 25 26 2020) recommends that individuals 2 and older consume 8 ounces per week of a variety 27 28 29 of seafood to provide approximately 250 mg of EPA and DHA per day (2). Consuming 30 31 this amount is associated with reduced cardiac death in those with and without pre-

32 http://bmjopen.bmj.com/ 33 34 existing cardiovascular disease (2). Similarly, the American Heart Association has 35 36 37 recommended consuming 1-2 seafood meals per week to reduce the risk of congestive 38 39 heart failure, coronary heart disease, ischemic stroke, and sudden cardiac death (12).

40 on September 24, 2021 by guest. Protected copyright. 41 42 Women who are pregnant or breastfeeding are encouraged to consume up to 12 43 44 45 ounces per week to obtain enough omega-3 in their diet to support infant health, 46 47 ensuring they consume at least 300 mg/d EPA+DHA, of which ≥200 mg/d should be 48 49 50 DHA specifically (2,13). 51 52 53 Despite the health benefits of omega-3s, data from the National Health and 54 55 Nutrition Examination Survey (NHANES) clearly shows that dietary intake of omega-3s in 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Americans is low and consistently well below the DGA recommendation across all age 4 5 6 groups, gender, and race/ethnicity (2,14–16). In 2015, the DGA Committee reviewed 7 8 9 data from What We Eat in America/NHANES and found 90% of Americans were below 10 11 the recommended intake of 250 mg DHA+EPA (2). Furthermore, an analysis of omega-3 12 13 14 dietary intake from NHANES 2003-2014 reported that younger individuals and women 15 16 For peer review only 17 may be particularly at risk: the mean EPA+DHA intake for children 1-5 years was just 18 19 23.1 mg per 1000kcal, which is below the DGA recommendation for EPA+DHA (17). A 20 21 22 further study of NHANES 2001-2014 reported that more than 95% of women of 23 24 25 childbearing-age did not meet the DGA recommendations for EPA+DHA (14). 26 27 Collectively, the dietary intake data from NHANES, as well as other populations 28 29 30 (18,19) provide compelling evidence of gaps in omega-3 intake. However, self-reported 31

32 http://bmjopen.bmj.com/ 33 dietary intake has also been criticized by some (20), due to biases of self-reporting 34 35 intake (21). Objective examination of markers of dietary intake (i.e. circulating omega-3s) 36 37 38 may therefore provide information on omega-3 status and evidence supporting studies 39

40 on September 24, 2021 by guest. Protected copyright. 41 of dietary intake. The DGA does not provide guidance on circulating levels of omega-3s 42 43 that correspond to recommended dietary intake of omega-3s, but comparisons can be 44 45 46 made to studies which have measured both intake and circulating levels (22) and have 47 48 49 established circulating omega-3 ranges associated with poor health outcomes (23,24). 50 51 Therefore, the aim of this study was to provide contemporary reference ranges for 52 53 54 omega-3 fatty acids in the US across all life stages using the most recent serum fatty 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 acid data (2011-2012) available in NHANES. An additional objective was to determine 4 5 6 the proportion of the US population with serum omega-3 fatty acid concentrations 7 8 9 below DGA recommendations and below those associated with cardioprotection. 10 11 12 13 14 Materials and Methods 15 16 For peer review only 17 This study uses the NHANES 2011-2012 public-use data files. NHANES is 18 19 designed to assess the nutritional status and health of children and adults in the US, 20 21 22 using a complex multistage probability sampling design that is representative of the 23 24 25 national civilian population in the US (25). Fatty acids were measured in serum with the 26 27 goal of obtaining US reference ranges for most circulating fatty acids. All participants, 28 29 30 age 3 to 11 who attended mobile examinations and a subsample of fasting participants, 31

32 http://bmjopen.bmj.com/ 33 age 12 and older who were examined in the morning session were eligible. Fatty acids 34 35 were measured using modified methods of Lagerstedt et al. (26). Briefly, total fatty acids 36 37 38 were hexane-extracted along with an internal standard solution for fatty acid recovery. 39

40 on September 24, 2021 by guest. Protected copyright. 41 The extract was derivatized to form pentafluorobenzyl esters and injected onto a 42 43 capillary gas chromatograph column. A total of 30 dietary fatty acids were quantitated. 44 45 46 The lower limit of detection for EPA, DHA and DPA were 0.79µmol/L, 1.84µmol/L and 47 48 49 0.24µmol/L. Fatty acids were expressed as a % of total fatty acids (respective fatty 50 51 acid/sum of 30 fatty acids measured) in this analysis to facilitate comparison with other 52 53 54 studies, which predominately report fatty acids as a relative percent (27–29). As relative 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 fatty acids require a common denominator of total fatty acids, individuals who were 4 5 6 missing data for any fatty acid were excluded, which resulted in an overall sample of 7 8 9 2,261 participants: 945 children, age 3-19 and 1,316 adults, age 20 and older. 10 11 Mean EPA, DHA and the sum of omega-3s (EPA+DPA+DHA), were calculated for 12 13 14 children and adults as well as by key life stages: early childhood (3-5 years), middle 15 16 For peer review only 17 childhood (6-11 years), adolescents (12-19 years), adults (20-55), seniors (>55) and 18 19 women who were pregnant and/or breastfeeding. Pregnancy status was determined 20 21 22 using the variable pregnancy status at exam (RIDEXPREG) from the demographics data 23 24 25 and the variable pregnancy test (URXPREG) from the laboratory data. Breastfeeding 26 27 status was determined from the reproductive health data (RHQ200). Women of 28 29 30 childbearing age was defined as those 15-44 years of age, which corresponded to the 31

32 http://bmjopen.bmj.com/ 33 age range of women who reported being pregnant and/or breastfeeding. Each life stage 34 35 was also analyzed by gender and race/ethnicity. 36 37 38 Supplements containing fish oil, cod liver oil, salmon oil, krill oil and DHA were 39

40 on September 24, 2021 by guest. Protected copyright. 41 referred to in this analysis as ‘omega-3 supplements’ and were identified from 42 43 participants report of dietary supplement use in the past 30 days. 44 45 46 Serum omega-3 fatty acids reflect short-term (weeks to months) intake of dietary 47 48 49 fat (30,31). To provide additional public health context to findings, comparisons were 50 51 made to fatty acid concentrations (omega-3s=2.49%), previously identified as 52 53 54 approximately equivalent to DGA dietary recommendations of 250 mg per day of EPA 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 and DHA (22,32). Comparisons were also made to serum fatty acid levels associated with 4 5 6 cardioprotection. Virtanen et al. (33) reported lower risk of atrial fibrillation; a common 7 8 9 cardiac arrhythmia among participants with serum DHA ≥2.85%. Reduced risk of 10 11 mortality, sudden cardiac arrest and other adverse cardiac events have been consistently 12 13 14 associated with higher red blood cell (RBC) levels of EPA+DHA, also known as the 15 16 For peer review only 17 Omega-3 Index (34). To compare serum NHANES data with the Omega-3 Index, an 18 19 equation relating the RBC-based metric with plasma EPA + DHA levels was applied to 20 21 22 calculate risk categories for cardiovascular disease (23,35). This equation was chosen 23 24 25 based on previous evidence that the fatty acid composition of serum and plasma are 26 27 similar (36). Application of the equation resulted in the following Omega-3 Index 28 29 30 categories: low risk >6.23%, intermediate risk 3.11-6.23%, or high risk <3.11%. 31

32 http://bmjopen.bmj.com/ 33 The following covariates were used to characterize the sample population: age, 34 35 gender, race/ethnicity (Non-Hispanic White, Hispanic/Mexican-American, Non-Hispanic 36 37 38 Black, Asian and other/unknown), and for adults: poverty income ratio (PIR; low: 0-1.85, 39

40 on September 24, 2021 by guest. Protected copyright. 41 medium >1.85-3.50 and high: >3.50), education (high 42 43 school), current smoking status (yes/no) and BMI. 44 45 46 Sample sizes in the text and tables are unweighted. However, estimates for 47 48 49 means, proportions, standard errors (SE) and 95% confidence intervals (CI) were 50 51 weighted using sampling weights for the fatty acid subsample. Variance estimate was 52 53 54 determined using the Taylor Series Linearization approach. Differences between groups 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 were determined by linear regression at p<0.05. Statistical analyses were performed 4 5 6 using survey procedures in STATA software version 14.2 (StataCorp, College Station, TX, 7 8 9 USA). 10 11 12 13 14 Results 15 16 For peer review only 17 Demographic characteristics 18 19 A total of 2,261 participants were included (Table 1), 47.6% were male, 52.5% 20 21 22 were female; 41.8% were 3-19 years of age; 58.2% were 20 years of age and older; 16.5% 23 24 25 were Mexican American or Hispanic, 65.2% were non-Hispanic White, 12.1% were non- 26 27 Hispanic black and 4.55% were Asian. With respect to education, 5.36% had education 28 29 30 below high school, 11.2% had a high school education, 63.7% had some college or 31

32 http://bmjopen.bmj.com/ 33 higher education, and 19.7% had education in between (more than high school). Among 34 35 adults, 20.2% were current smokers; 9.90% had low PIR, 32.7% had medium PIR and 36 37 38 66.3% had high PIR. The mean (SE) BMI among adults was 28.7 (0.34) kg/m2. Omega-3 39

40 on September 24, 2021 by guest. Protected copyright. 41 supplement use was low, with only 7.3% (N=166) in the overall population reporting use 42 43 in the prior 30 days. 44 45 46 Mean concentrations of omega-3 fatty acids 47 48 49 Mean serum concentrations of EPA, DHA and omega-3s by age group are shown 50 51 in Table 2. Among children 3-19 years, mean (SE) EPA, DHA and omega-3s were 0.28% 52 53 54 (0.01), 1.07% (0.02), and 1.75% (0.03), respectively. Among adults 20 and older, mean 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 (SE) EPA, DHA and the sum of omega-3s were 0.61% (0.02), 1.38% (0.05) and 2.43% 4 5 6 (0.08), respectively. Concentrations of EPA, DHA and the sum of omega-3s were all 7 8 9 significantly lower (p<0.001) in children compared to adults. Mean (SE) concentrations 10 11 of EPA, DHA and the sum of omega-3s in early childhood were 0.26% (0.01), 1.01% 12 13 14 (0.03), 1.65% (0.04); middle childhood: 0.29% (0.02), 1.09% (0.03), 1.78% (0.06); and 15 16 For peer review only 17 adolescents: 0.29% (0.02), 1.08% (0.03) and 1.76% (0.04). Compared to adolescents, the 18 19 sum of omega-3s were lower in early childhood (p=0.03). Similarly, EPA and DHA 20 21 22 tended to be lower in early childhood (p=0.06 and p=0.07). Mean (SE) concentrations of 23 24 25 EPA, DHA and the sum of omega-3s in adults 20-55 years old were 0.51% (0.02), 1.29% 26 27 (0.06), 2.23% (0.08), which were significantly lower than in seniors (all p<0.001): 0.79% 28 29 30 (0.06), 1.56% (0.06), 2.83% (1.00). The prevalence of omega-3 supplement use ranged 31

32 http://bmjopen.bmj.com/ 33 from low use N=3 (1.21%) in adolescents to high use N=84 (16.9%) in seniors. 34 35 Mean serum concentrations of EPA, DHA and the sum of omega-3s by life stage 36 37 38 and gender are shown in Supplementary Table 1 and by race/ethnicity among children 39

40 on September 24, 2021 by guest. Protected copyright. 41 and adults in Supplementary Table 2. Among adults 20 and older, males had 42 43 significantly lower serum levels of EPA (p=0.01), DHA (p<0.001) and the sum of omega- 44 45 46 3s (p<0.001) compared to females. When further examined by life stage, men 47 48 49 consistently had lower levels beginning with adolescent males having significantly lower 50 51 DHA levels than females (p=0.03). Additionally, males age 20-55 and males >55 had 52 53 54 significantly lower DHA and sum omega-3 fatty acids than females 20-55 (p=0.001 and 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 p=0.01) and females >55 (p=0.03 and p=0.04). Race/ethnicity differences between 4 5 6 omega-3s were similar among children and adults: relative to non-Hispanic White 7 8 9 individuals, EPA was significantly lower in Mexican American/Hispanic individuals, DHA 10 11 was significantly lower in non-Hispanic Black individuals, and EPA, DHA and the sum of 12 13 14 omega-3s were all significantly higher among Asian individuals. Mean concentrations of 15 16 For peer review only 17 EPA, DHA and omega-3s are also shown in Supplementary Table 3 by Institute of 18 19 Medicine (IOM) age categories to facilitate comparison with dietary recommendations. 20 21 22 Reference concentrations of all 27 fatty acids additionally measured in the NHANES 23 24 25 2011-2012 cycle are shown in Supplementary Table 4. 26 27 Mean (SE) serum concentrations of EPA, DHA and the sum of omega-3s by 28 29 30 pregnancy/breastfeeding status are shown in Table 3. There was no statistical difference 31

32 http://bmjopen.bmj.com/ 33 in the concentrations of EPA among pregnant/breastfeeding women compared to 34 35 among women of childbearing age (p=0.17). DHA and the sum of omega-3s among 36 37 38 pregnant/breastfeeding women which were not statistically different than those in 39

40 on September 24, 2021 by guest. Protected copyright. 41 women of childbearing age. Omega-3 supplement use was reported by two (9.52%) 42 43 pregnant/breastfeeding women compared to 21 (6.03%) women of childbearing age. 44 45 46 The prevalence of individuals with concentrations below those approximately 47 48 49 equivalent to the DGA recommendations is shown in Table 4 by gender and life stage. 50 51 Overall, 95.5% of children and 68.3% of adults age 20 years and older had omega-3 52 53 54 concentrations below 2.49%, the omega-3 level recommended by the DGA. The 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 prevalence of those below the DGA recommendations ranged from 97.5% of females in 4 5 6 early childhood to 52.3% of female seniors. Early childhood (age 3-5 years), middle 7 8 9 childhood (age 6-11 years and adolescents (age 12-19 years), had the largest 10 11 prevalence of low omega-3 levels, with 97.4%, 94.8% and 95.5% of the population in 12 13 14 these age groups below omega-3 levels (<2.49%) recommended by the DGA, 15 16 For peer review only 17 respectively. Figure 1 depicts the prevalence of participants with serum DHA and 18 19 Omega-3 Index in previously established risk categories (23,33). The vast majority of 20 21 22 adults had DHA concentration levels below that associated with lower risk of atrial 23 24 25 fibrillation (2.85%): 97% of all adults, 98% of males and 96.2% of females. Among all 26 27 adults, 88.7% of the population had an Omega-3 Index in the high cardiovascular risk 28 29 30 category, 10.4% of the adults were in the intermediate risk category and just 1.0% of the 31

32 http://bmjopen.bmj.com/ 33 adults were in the low risk category. The prevalence of those with an Omega-3 Index in 34 35 the low cardiovascular risk category ranged from a low of 0.34% of adult males to a high 36 37 38 of 3.08% among senior females (Supplementary Table 5). 39

40 on September 24, 2021 by guest. Protected copyright. 41 42 43 Discussion 44 45 46 This study provides new information on reference levels of serum omega-3 (a biomarker 47 48 49 of omega-3 nutritional status) in a large, nationally representative sample of Americans 50 51 across all life stages, from early childhood into later life. The mean EPA and DHA serum levels 52 53 of 0.28% and 1.07% in children (3-19 years), and 0.61% and 1.38% in adults (20 years+), 54 55 56 reflects the persistently low consumption of seafood and omega-3s in general, in the US 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 population despite ongoing scientific research to support guidance from expert committees to 4 5 6 increase intake and for the general public to consume 8 ounces per week of a variety of 7 8 9 seafood, providing approximately 250 mg of EPA and DHA per day (EPA+DHA 10 11 concentrations of 2.49% are approximately equivalent to DGA dietary recommendations 12 13 14 of 250 mg per day of EPA and DHA) (1,2). Using estimates of omega-3 concentrations 15 16 For peer review only 17 approximately equivalent to the DGA dietary intake recommendations for EPA and DHA, 18 19 nearly all children in the study, and over 68% of adults had serum omega-3 levels below 20 21 22 those associated with US Dietary recommendations for omega-3 consumption. When 23 24 25 further examined by life stage and demographics, several potentially vulnerable 26 27 populations emerged that may warrant targeted promotion of EPA and DHA 28 29 30 consumption; specifically early childhood (ages 3-5 years), adult males, Mexican 31

32 http://bmjopen.bmj.com/ 33 American/Hispanic individuals, and non-Hispanic Black individuals tended to have 34 35 particularly low serum concentrations of omega-3 fatty acids. Pregnant and/or 36 37 38 breastfeeding women also appear to be a population of concern as omega-3 fatty acid 39

40 on September 24, 2021 by guest. Protected copyright. 41 concentrations were similar to women of comparable age despite recommendations for 42 43 higher EPA and DHA consumption to support infant development (2). 44 45 46 Adults, particularly seniors, had the highest levels of serum EPA, DHA and the 47 48 49 sum of omega-3s, although only around one-quarter of adults, and approximately half 50 51 of seniors had concentrations that met the DGA recommendations. However, these 52 53 54 levels are not adequate for cardioprotection; less than 2% of seniors had an Omega-3 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Index in the low cardiovascular disease risk category. The notably higher serum 4 5 6 concentrations of omega-3s in seniors >55 relative to adults 20-55 years mirrors 7 8 9 findings from previous studies of omega-3 dietary intake (15,17) that report higher 10 11 intake and also greater use of omega-3 supplementation among seniors. Currently, 12 13 14 there is no established Adequate Intake (AI) for EPA or DHA; the AI for alpha-linolenic 15 16 For peer review only 17 acid for males and females is the same for those age 14 and older (37). However, 18 19 epidemiologic and physiologic evidence suggests that older adults have an increased 20 21 22 omega-3 requirement due to the aging body, including brain atrophy and cognitive 23 24 25 aging (38,39), mobility, bone health, and sarcopenia (40–42). Despite omega-3 26 27 concentrations being highest in seniors, the implications with respect to adequacy to 28 29 30 support the normal physiological processes underlying healthy aging are unclear. 31

32 http://bmjopen.bmj.com/ 33 Our findings suggest similar DHA and total omega-3 fatty acids between women 34 35 who were pregnant and/or breastfeeding and women of reproductive age. Both groups 36 37 38 have potential implications for maternal health as well as for fetal and infant 39

40 on September 24, 2021 by guest. Protected copyright. 41 development. Omega-3s, specifically DHA, are important for fetal and infant neuron and 42 43 retinal membrane development in late pregnancy and early life, essentially the first 1000 44 45 46 days of life (9,43,44). Omega-3 intake requirements in pregnancy have not been 47 48 49 established, however recommendations from a number of international organizations 50 51 have been put forth (14) and public health messaging on the importance of omega-3s 52 53 54 intake in maternal health has been promoted in recent years (13). DHA demands during 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 pregnancy are higher due to diversion to the fetus, especially in the last trimester as the 4 5 6 nervous system is rapidly developing, as well as expanded maternal cell mass and 7 8 9 general requirements of the placenta (45). Pregnant women can become low in DHA if 10 11 dietary intakes do not compensate for increased demands. Public health messaging has 12 13 14 encouraged women to increase omega-3 intake, especially DHA through a variety of 15 16 For peer review only 17 sources (low-mercury fish, and DHA found in supplements and fortified foods) (10). 18 19 However, the prevalence of omega-3 supplementation in pregnant/breastfeeding 20 21 22 women was just 9.5%. In addition, it was found that serum EPA tended to be lower 23 24 25 among pregnant/breastfeeding women, although not statistically significant, possibly 26 27 reflecting the small sample size. This may have relevance for maternal health as 28 29 30 epidemiologic evidence suggests that omega-3s play a role in prevention and/or 31

32 http://bmjopen.bmj.com/ 33 treatment of neuropsychiatric diseases (46). In recent years, a number of systematic 34 35 reviews and meta-analyses have concluded that omega-3 supplementation with a 36 37 38 higher EPA to DHA ratio, with at least 1 gram of EPA, in conjunction with anti-depressive 39

40 on September 24, 2021 by guest. Protected copyright. 41 therapy, may be more beneficial than either monotherapy alone (47–49). Associations 42 43 with postpartum depression specifically are unclear, but warrant further investigation 44 45 46 considering the prevalence of postpartum depression (50). The estimates for pregnant 47 48 49 and breastfeeding women, met analytic and reporting criteria for NHANES (25), but 50 51 nonetheless, the sample size was small and caution is warranted for interpretation of the 52 53 54 findings. The findings are however consistent with a study of EPA and DHA intake in 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 NHANES which have pooled the samples across several cycles and found EPA and DHA 4 5 6 intake in pregnant women did not differ from non-pregnant women in terms of being 7 8 9 generally inadequate (17). 10 11 A strength of this study is the measurement of omega-3s in serum rather than 12 13 14 estimates from dietary intake which minimizes error from self-reporting. However, 15 16 For peer review only 17 serum fatty acids represent short term dietary consumption (30,31) and our results may 18 19 therefore not reflect longer term status. Omega-3 Index risk categories were determined 20 21 22 from an equation relating concentrations in RBCs to plasma which may introduce bias in 23 24 25 our estimates. The large population allowed us to examine reference concentrations for 26 27 multiple life stages and demographic categories which was weighted to be nationally 28 29 30 representative of the US, although sample sizes for some subgroups e.g. pregnant and 31

32 http://bmjopen.bmj.com/ 33 breastfeeding women were limited as described above. The present study provides the 34 35 most contemporaneous data in a US nationally representative population, but 8 to 9 36 37 38 years has elapsed between the MEC examination and the present study since the 39

40 on September 24, 2021 by guest. Protected copyright. 41 NHANES serum fatty acid data only became publicly available in 2020. It is therefore 42 43 possible that concentrations reported herein differ from those currently in the US. Data 44 45 46 on dietary intake of EPA and DHA in pregnant women and women of childbearing age 47 48 49 in NHANES 2001-2014 suggest slight increases in intake across 2-year survey periods 50 51 (14). Unfortunately, it is not possible to assess trends in circulating omega-3 fatty acids 52 53 54 as NHANES infrequently collects circulating fatty acid data. Fatty acids were last 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 measured in plasma collected in the 2003-2004 NHANES cycle which captured a 4 5 6 different number of fatty acids (24 versus 30 in 2011-2012), and was limited to 7 8 9 individuals age 20 and older (32). The absence of standardized guidelines or definitions 10 11 of adequate or optimal serum omega-3 concentrations to support overall health also 12 13 14 precluded our ability to provide inferences from these results. The cross-sectional nature 15 16 For peer review only 17 of NHANES can be limiting in its ability to study possible causal relationships and their 18 19 resultant physiological function. 20 21 22 Conclusions 23 24 In conclusion, this study provides reference ranges on serum omega-3 fatty acids, a 25 26 27 biomarker of omega-3 status, across all life stages in a nationally representative population, 28 29 30 information that until now, was critically lacking. The findings demonstrate overall low serum 31

32 concentrations of omega-3 fatty acids including both EPA and DHA and the sum of omega-3 http://bmjopen.bmj.com/ 33 34 fatty acids across all life stages. This data is supported by findings from previous dietary intake 35 36 analyses (14–16). Owing to the role of omega-3s in brain development, there is particular need 37 38 39 for public health strategies aimed at increasing omega-3 intake in children and women of

40 on September 24, 2021 by guest. Protected copyright. 41 childbearing age including those who are pregnant, breastfeeding or looking to become pregnant. 42 43 Based on life stages, gender and demographic factors, we demonstrate that children, 44 males, 45 46 Mexican American/Hispanic and non-Hispanic Black individuals have lower EPA and/or 47 48 49 DHA serum concentrations than adults, females, and non-Hispanic White individuals, 50 51 respectively. This further demonstrates the need for public health awareness and action. 52 53 54 Omega-3 fatty acids play a critically important role in many aspects of human health across the 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 lifespan. Healthcare practitioners need to ensure their patients are consuming enough omega-3 4 5 6 fatty acids in their daily diet, and if they are not, they need to consider the role of 7 8 supplementation for their patients to ensure nutrient gaps are being met. Furthermore, there is 9 10 increased need for guidance and education for healthcare practitioners to ensure they are 11 12 prepared to make recommendations needed that include specific ways to increase EPA and DHA 13 14 15 intake for the entire US population, with particular attention to vulnerable populations. 16 For peer review only 17 18 19 20 Figure 1 legend: 21 22 Parallels to omega-3 levels were drawn from Sun et al. (22) to approximate the prevalence of 23 24 those with omega-3 levels equivalent to dietary intake recommendations (EPA+DHA) by the 25 26 27 DGA, the lower range of the tertiles of serum DHA associated with lower risk of atrial 28 29 fibrillation by Virtanen et al. (33), and cardiovascular risk categories defined by the Omega-3 30 31

32 Index (23,35). http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

40 on September 24, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 4 5 6 7 8 References 9 10 1 1. Fish and Omega-3 Fatty Acids. American Heart Association. Cited 2020 Mar 25. Available 11 2 from: https://www.heart.org/en/healthy-living/healthy-eating/eat-smart/fats/fish-and- 12 3 omega-3-fatty-acids 13 14 15 4 2. Dietary Guidelines Advisory Committee. Scientific Report of the 2015 Dietary Guidelines 16 5 Advisory Committee,For Advisory peer Report review to the Secretary ofonly Health and Human Services and 17 6 the Secretary of Agriculture. Washington, DC: US Department of Agriculture, Agricultural 18 7 Research Service; 2015. Cited 2016 Nov 28. Available from: 19 8 https://health.gov/dietaryguidelines/2015-scientific-report/pdfs/scientific-report-of-the- 20 9 2015-dietary-guidelines-advisory-committee.pdf 21 22 23 10 3. Bowen KJ, Harris WS, Kris-Etherton PM. Omega-3 Fatty Acids and Cardiovascular 24 11 Disease: Are There Benefits? Curr Treat Options Cardiovasc Med. 2016 Nov;18(11):69. 25 26 12 4. Djoussé L, Gaziano JM, Buring JE, Lee I-M. Dietary omega-3 fatty acids and fish 27 13 consumption and risk of type 2 diabetes. Am J Clin Nutr. 2011 Jan;93(1):143–50. 28 29 14 5. Kaushik M, Mozaffarian D, Spiegelman D, Manson JE, Willett WC, Hu FB. Long-chain 30 15 omega-3 fatty acids, fish intake, and the risk of type 2 diabetes mellitus. Am J Clin Nutr. 31 16 2009 Sep;90(3):613–20.

32 http://bmjopen.bmj.com/ 33 34 17 6. Wang DD, Li Y, Chiuve SE, Stampfer MJ, Manson JE, Rimm EB, et al. Association of 35 18 Specific Dietary Fats With Total and Cause-Specific Mortality. JAMA Intern Med. 2016 36 19 01;176(8):1134–45. 37 38 20 7. Liu JC, Conklin SM, Manuck SB, Yao JK, Muldoon MF. Long-Chain Omega-3 Fatty 39 21 Acids and Blood Pressure. Am J Hypertens. 2011 Oct 1;24(10):1121–6.

40 on September 24, 2021 by guest. Protected copyright. 41 22 8. Ueshima H, Stamler J, Elliott P, Chan Q, Brown IJ, Carnethon MR, et al. Food Omega-3 42 23 Fatty Acid Intake of Individuals (Total, Linolenic Acid, Long-Chain) and Their Blood 43 44 24 Pressure: INTERMAP Study. Hypertension. 2007 Aug;50(2):313–9. 45 46 25 9. Jensen CL. Effects of n−3 fatty acids during pregnancy and lactation. Am J Clin Nutr. 2006 47 26 Jun 1;83(6):1452S-1457S. 48 49 27 10. Greenberg JA, Bell SJ, Ausdal WV. Omega-3 Fatty Acid supplementation during 50 28 pregnancy. Rev Obstet Gynecol. 2008;1(4):162–9. 51 52 29 11. Gelenberg AJ, Freeman MP , Markowitz JC, Rosenbaum JF, Thase ME, Trivedi MH, et al. 53 54 30 Practice Guideline for the Treatment of Patients With Major Depressive Disorder, Third 55 31 Edition. page 51 and page 92. 2010. Available from: 56 32 http://psychiatryonline.org/pb/assets/raw/sitewi de/practice_guidelines/guidelines/mdd.pdf. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 33 12. Rimm EB, Appel LJ, Chiuve SE, Djoussé L, Engler MB, Kris-Etherton PM, et al. Seafood 4 34 Long-Chain n-3 Polyunsaturated Fatty Acids and Cardiovascular Disease: A Science 5 6 35 Advisory From the American Heart Association. Circulation. 2018 03;138(1):e35–47. 7 8 36 13. Joint FAO/WHO Expert Consultation on Fats and fatty acids in human nutrition: report of 9 37 an expert consultation. Rome, Italy: Food and Agriculture Organization, 2010. 10 11 38 14. Zhang Z, Fulgoni VL, Kris-Etherton PM, Mitmesser SH. Dietary Intakes of EPA and DHA 12 39 Omega-3 Fatty Acids among US Childbearing-Age and Pregnant Women: An Analysis of 13 40 NHANES 2001-2014. Nutrients. 2018 28;10(4). 14 15 41 15. Richter CK, Bowen KJ, Mozaffarian D, Kris-Etherton PM, Skulas-Ray AC. Total Long- 16 For peer review only 17 42 Chain n-3 Fatty Acid Intake and Food Sources in the United States Compared to 18 43 Recommended Intakes: NHANES 2003–2008. Lipids. 2017 Nov;52(11):917–27. 19 20 44 16. Papanikolaou Y, Brooks J, Reider C, Fulgoni VL. U.S. adults are not meeting 21 45 recommended levels for fish and omega-3 fatty acid intake: results of an analysis using 22 46 observational data from NHANES 2003–2008. Nutr J. 2014 Dec;13(1):31. 23 24 47 17. Thompson M, Hein N, Hanson C, Smith LM, Anderson-Berry A, Richter CK, et al. Omega- 25 26 48 3 Fatty Acid Intake by Age, Gender, and Pregnancy Status in the United States: National − 27 49 Health and Nutrition Examination Survey 2003 2014. Nutrients. 2019 Jan 15;11(1). 28 29 50 18. Villegas R, Takata Y, Murff H, Blot WJ. Fish, omega-3 long-chain fatty acids, and all- 30 51 cause mortality in a low-income US population: Results from the Southern Community 31 52 Cohort Study. Nutr Metab Cardiovasc Dis. 2015 Jul;25(7):651–8.

32 http://bmjopen.bmj.com/ 33 53 19. Berry JD, Prineas RJ, van Horn L, Passman R, Larson J, Goldberger J, et al. Dietary Fish 34 54 Intake and Incident Atrial Fibrillation (from the Women’s Health Initiative). Am J Cardiol. 35 36 55 2010 Mar;105(6):844–8. 37 38 56 20. Archer E. The NHANES dietary data are physiologically implausible and inadmissible as 39 57 scientific evidence. Am J Clin Nutr. 2017;106(3):951–2.

40 on September 24, 2021 by guest. Protected copyright. 41 58 21. Thompson FE, Kirkpatrick SI, Subar AF, Reedy J, Schap TE, Wilson MM, et al. The 42 59 National Cancer Institute’s Dietary Assessment Primer: A Resource for Diet Research. J 43 60 Acad Nutr Diet. 2015 Dec;115(12):1986–95. 44 45 61 22. Sun Q, Ma J, Campos H, Hankinson SE, Hu FB. Comparison between plasma and 46 47 62 erythrocyte fatty acid content as biomarkers of fatty acid intake in US women. Am J Clin 48 63 Nutr. 2007 Jul;86(1):74–81. 49 50 64 23. Harris WS. The omega-3 index as a risk factor for coronary heart disease. Am J Clin Nutr. 51 65 2008 Jun 1;87(6):1997S-2002S. 52 53 66 24. Virtanen JK, Laukkanen JA, Mursu J, Voutilainen S, Tuomainen T-P. Serum long-chain n- 54 67 3 polyunsaturated fatty acids, mercury, and risk of sudden cardiac death in men: a 55 56 68 prospective population-based study. PloS One. 2012;7(7):e41046. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 69 25. Johnson CL, Paulose-Ram R, Ogden CL, Carroll MD, Kruszon-Moran D, Dohrmann SM, 4 70 et al. National health and nutrition examination survey: analytic guidelines, 1999-2010. 5 6 71 Vital Health Stat 2. 2013 Sep;(161):1–24. 7 8 72 26. Lagerstedt SA, Hinrichs DR, Batt SM, Magera MJ, Rinaldo P, McConnell JP. Quantitative 9 73 Determination of Plasma C8–C26 Total Fatty Acids for the Biochemical Diagnosis of 10 74 Nutritional and Metabolic Disorders. Mol Genet Metab. 2001 May;73(1):38–45. 11 12 75 27. Sun Y, Koh HWL, Choi H, Koh W-P, Yuan J-M, Newman JW, et al. Plasma fatty acids, 13 76 oxylipins, and risk of myocardial infarction: the Singapore Chinese Health Study. J Lipid 14 77 Res. 2016;57(7):1300–7. 15 16 For peer review only 17 78 28. Wu JHY, Marklund M, Imamura F, Tintle N, Ardisson Korat AV, de Goede J, et al. 18 79 Omega-6 fatty acid biomarkers and incident type 2 diabetes: pooled analysis of individual- 19 80 level data for 39 740 adults from 20 prospective cohort studies. Lancet Diabetes 20 81 Endocrinol. 2017;5(12):965–74. 21 22 82 29. Imamura F, Fretts A, Marklund M, Ardisson Korat AV, Yang W-S, Lankinen M, et al. 23 83 Fatty acid biomarkers of dairy fat consumption and incidence of type 2 diabetes: A pooled 24 25 84 analysis of prospective cohort studies. PLoS Med. 2018;15(10):e1002670. 26 27 85 30. Holman RT. Control of polyunsaturated acids in tissue lipids. J Am Coll Nutr. 28 86 1986;5(2):183–211. 29 30 87 31. Kwon JS, Snook JT, Wardlaw GM, Hwang DH. Effects of diets high in saturated fatty 31 88 acids, canola oil, or safflower oil on platelet function, thromboxane B2 formation, and fatty

32 89 acid composition of platelet phospholipids. Am J Clin Nutr. 1991 Aug;54(2):351–8. http://bmjopen.bmj.com/ 33 34 90 32. Murphy RA, Yu EA, Ciappio ED, Mehta S, McBurney MI. Suboptimal Plasma Long Chain 35 36 91 n-3 Concentrations are Common among Adults in the United States, NHANES 2003-2004. 37 92 Nutrients. 2015 Dec 9;7(12):10282–9. 38 39 93 33. Virtanen JK, Mursu J, Voutilainen S, Tuomainen T-P. Serum Long-Chain n-3

40 94 Polyunsaturated Fatty Acids and Risk of Hospital Diagnosis of Atrial Fibrillation in Men. on September 24, 2021 by guest. Protected copyright. 41 95 Circulation. 2009 Dec 8;120(23):2315–21. 42 43 96 34. von Schacky C. Omega-3 Index and Cardiovascular Health. Nutrients. 2014 Feb 44 97 21;6(2):799–814. 45 46 47 98 35. Chaudhary R, Saadin K, Bliden KP, Harris WS, Dinh B, Sharma T, et al. Risk factors 48 99 associated with plasma omega-3 fatty acid levels in patients with suspected coronary artery 49 100 disease. Prostaglandins Leukot Essent Fatty Acids. 2016 Oct;113:40–5. 50 51 101 36. Armstrong JM, Metherel AH, Stark KD. Direct Microwave Transesterification of Fingertip 52 102 Prick Blood Samples for Fatty Acid Determinations. Lipids. 2008 Feb;43(2):187–96. 53 54 103 37. Panel on Macronutrients, Panel on the Definition of Dietary Fiber, Subcommittee on Upper 55 56 104 Reference Levels of Nutrients, Subcommittee on Interpretation and Uses of Dietary 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 105 Reference Intakes, Standing Committee on the Scientific Evaluation of Dietary Reference 4 106 Intakes, Food and Nutrition Board, et al. Dietary Reference Intakes for Energy, 5 6 107 Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. Washington, 7 108 D.C.: National Academies Press; 2005. Cited 2020 Apr 20. Available from: 8 109 https://www.nap.edu/catalog/10490. 9 10 110 38. Issa AM, Mojica WA, Morton SC, Traina S, Newberry SJ, Hilton LG, et al. The Efficacy of 11 111 Omega–3 Fatty Acids on Cognitive Function in Aging and Dementia: A Systematic 12 112 Review. Dement Geriatr Cogn Disord. 2006;21(2):88–96. 13 14 113 39. Denis I, Potier B, Heberden C, Vancassel S. Omega-3 polyunsaturated fatty acids and brain 15 16 114 aging: Curr OpinFor Clin Nutr peer Metab Care. review 2015 Mar;18(2):139–46. only 17 18 115 40. Reinders I, Murphy RA, Song X, Visser M, Cotch MF, Lang TF, et al. Polyunsaturated 19 116 fatty acids in relation to incident mobility disability and decline in gait speed; the Age, 20 117 Gene/Environment Susceptibility-Reykjavik Study. Eur J Clin Nutr. 2015 Apr;69(4):489– 21 118 93. 22 23 119 41. Harris TB, Song X, Reinders I, Lang TF, Garcia ME, Siggeirsdottir K, et al. Plasma 24 25 120 phospholipid fatty acids and fish-oil consumption in relation to osteoporotic fracture risk in 26 121 older adults: the Age, Gene/Environment Susceptibility Study. Am J Clin Nutr. 2015 27 122 May;101(5):947–55. 28 29 123 42. Dupont J, Dedeyne L, Dalle S, Koppo K, Gielen E. The role of omega-3 in the prevention 30 124 and treatment of sarcopenia. Aging Clin Exp Res. 2019 Jun;31(6):825–36. 31

32 125 43. Middleton P, Gomersall JC, Gould JF, Shepherd E, Olsen SF, Makrides M. Omega-3 fatty http://bmjopen.bmj.com/ 33 126 acid addition during pregnancy. Cochrane Database Syst Rev. 2018 15;11:CD003402. 34 35 36 127 44. Carlson SE, Gajewski BJ, Valentine CJ, Rogers LK, Weiner CP, DeFranco EA, et al. 37 128 Assessment of DHA on reducing early preterm birth: the ADORE randomized controlled 38 129 trial protocol. BMC Pregnancy Childbirth. 2017 13;17(1):62. 39

40 130 45. Rees A-M, Austin M-P, Owen C, Parker G. Omega-3 deficiency associated with perinatal on September 24, 2021 by guest. Protected copyright. 41 131 depression: case control study. Psychiatry Res. 2009 Apr 30;166(2–3):254–9. 42 43 132 46. Larrieu T, Layé S. Food for Mood: Relevance of Nutritional Omega-3 Fatty Acids for 44 133 Depression and Anxiety. Front Physiol. 2018;9:1047. 45 46 47 134 47. Liao Y, Xie B, Zhang H, He Q, Guo L, Subramaniapillai M, et al. Efficacy of omega-3 48 135 PUFAs in depression: A meta-analysis. Transl Psychiatry. 2019 05;9(1):190. 49 50 136 48. Hallahan B, Ryan T, Hibbeln JR, Murray IT, Glynn S, Ramsden CE, et al. Efficacy of 51 137 omega-3 highly unsaturated fatty acids in the treatment of depression. Br J Psychiatry J 52 138 Ment Sci. 2016;209(3):192–201. 53 54 139 49. Wojcicki JM, Heyman MB. Maternal omega-3 fatty acid supplementation and risk for 55 56 140 perinatal maternal depression. J Matern Fetal Neonatal Med. 2011 May;24(5):680–6. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 141 50. Browne JC, Scott KM, Silvers KM. Fish consumption in pregnancy and omega-3 status 4 142 after birth are not associated with postnatal depression. J Affect Disord. 2006 Feb;90(2– 5 6 143 3):131–9. 7 8 Footnotes 9 10 Twitter: @RaAMurphy 11 12 Contributors: RAM, KM, PPD, SE, and SHM were involved in study conception and design 13 14 15 and advised on interpretation of the data. RAM compiled the dataset and performed the statistical 16 For peer review only 17 analysis. RAM drafted the manuscript. KM, PPD, SE, and SHM review the manuscript, provided 18 19 amendment and approved the final version. RAM had full access to study data and takes 20 21 22 responsibility for its accuracy and the integrity of the analysis. 23 24 Funding: N/A 25 26 Competing interests: RAM carried out the work presented in the manuscript as a consultant for 27 28 Pharmavite. PPD, SE, KM and SHM are employees of Pharmavite, LLC, manufacturers and 29 30 31 suppliers of omega-3 nutritional lipids.

32 http://bmjopen.bmj.com/ 33 Patient and public involvement: Patients and/or the public were not involved in the design, or 34 35 conduct, or reporting, or dissemination plans of this research. 36 37 38 Patient consent for publication: Not required 39

40 Ethics approval: The NCHS Ethics Review Board protects the rights and welfare of NHANES on September 24, 2021 by guest. Protected copyright. 41 42 participants. The NHANES protocol complies with the U.S. Department of Health and Human 43 44 45 Services’ Policy for Protection of Human Research Subjects. NCHS IRB/ERC Protocol number: 46 47 2011-17. This study was exempt from Institutional Review Board at the University of British 48 49 Columbia related to the use of publicly available data for research and publication. 50 51 Provenance and peer review: Not commissioned; internally peer reviewed by Pharmavite 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Data availability statement: Data are available in a public, open access repository. The dataset 4 5 6 used for this study was generated from data publicly released by the National Health and 7 8 Nutrition Examination Survey (NHANES). 9 10 Word count: 3,525 11 12 13 14 144 15 16 145 For peer review only 17 18 146 19 20 21 147 22 23 148 24 25 149 26 27 28 150 29 30 151 31

32 152 http://bmjopen.bmj.com/ 33 34 35 153 36 37 154 38 39 155

40 on September 24, 2021 by guest. Protected copyright. 41 42 156 43 44 157 45 46 158 47 48 49 159 50 51 160 52 53 161 54 55 56 162 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 163 4 5 6 164 Table 1. Characteristics of study participants in the NHANES 2011-2012 fatty acid 7 165 sample 8 9 Characteristics N (%) 10 11 Gender Male 1098 (47.6) 12 Female 1163 (52.5) 13 Age group Children, 3-19yrs 945 (41.8) 14 15 Adults, 20yrs+ 1316 (58.2) 16 For peerEarly review childhood, 3- only 179 (2.40) 17 5yrs 18 19 Middle childhood, 6- 519 (6.52) 20 11yrs 21 Adolescents, 12-19yrs 247 (10.3) 22 23 Adults, 20-55yrs 818 (53.9) 24 Seniors, >55yrs 498 (26.9) 25 Race/ethnicity Mexican 605 (16.5) 26 27 American/Hispanic 28 Non-Hispanic White 729 (65.2) 29 Non-Hispanic Black 592 (12.1) 30 31 Asian 265 (4.55)

32 Other/unknown 70 (1.66) http://bmjopen.bmj.com/ 33 Education-adults High school 279 (19.7) 37 Some college or 734 (63.7) 38 39 above

40 Current smoking Yes 265 (20.2) on September 24, 2021 by guest. Protected copyright. 41 No 1,049 (79.8) 42 43 Poverty Income Ratio Low 0-1.85 20 (9.90) 44 Medium >1.85-3.50 423 (32.7) 45 High >3.50 712 (66.3) 46 BMI-adults Mean, kg/m2 28.7 (0.34) 47 48 Omega-3 supplement use, N Yes 166 (7.30) 49 (%) 50 166 51 Sample sizes are unweighted however, estimates for means, proportions and standard 52 167 errors are weighted. Poverty to income ratio is defined as total family income relative 53 168 to poverty guidelines. Omega-3 supplement use defined from self-reported 54 169 55 supplement use (fish oil, cod liver oil, salmon oil, krill oil and DHA) in prior 30 days. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 170 4 171 5 6 172 7 173 8 174 9 175 10 176 11 177 12 178 13 14 179 15 16 180 Table 2. Mean serumFor omega-3 peer fatty acidreview concentrations only (% of total fatty acids) by 17 181 age group in NHANES 2011-2012 18 19 Age group Variable N Mean (SE) 95%CI P 20 a 21 Children EPA 945 0.28 (0.01) 0.25-0.31 <0.001 22 3-19yrs DHA 945 1.07 (0.02) 1.03-1.10 <0.001 23 EPA+DPA+DHA 945 1.75 (0.03) 1.68-1.83 <0.001 24 25 Adults EPA 1,316 0.61 (0.02) 0.56-0.66 Ref 26 20yrs+ DHA 1,316 1.38 (0.05) 1.27-1.49 Ref 27 EPA+DPA+DHA 1,316 2.43 (0.08) 2.28-2.59 Ref 28 b 29 Early childhood EPA 179 0.26 (0.01) 0.24-0.27 0.06 30 3-5yrs DHA 179 1.01 (0.03) 0.96-1.07 0.07 31 EPA+DPA+DHA 179 1.65 (0.04) 1.58-1.73 0.03

32 http://bmjopen.bmj.com/ Middle 0.99 33 EPA 519 0.29 (0.02) 0.24-0.34 34 childhoodb 35 6-11yrs DHA 519 1.09 (0.03) 1.02-1.16 0.76 36 37 EPA+DPA+DHA 519 1.78 (0.06) 1.65-1.90 0.78 38 Adolescents EPA 247 0.29 (0.02) 0.25-0.32 Ref 39 12-19yrs DHA 247 1.08 (0.03) 1.02-1.13 Ref

40 on September 24, 2021 by guest. Protected copyright. 41 EPA+DPA+DHA 247 1.76 (0.04) 1.78-1.84 Ref 42 Adultsc EPA 818 0.51 (0.02) 0.47-0.55 Ref 43 20-55yrs DHA 818 1.29 (0.06) 1.18-1.41 Ref 44 45 EPA+DPA+DHA 818 2.23 (0.08) 2.08-2.39 Ref 46 Seniors EPA 498 0.79 (0.06) 0.67-0.92 <0.001 47 >55yrs DHA 498 1.56 (0.06) 1.44-1.68 <0.001 48 49 EPA+DPA+DHA 498 2.83 (1.00) 2.62-3.04 <0.001 50 182 Comparisons are a) children 3-19 years versus adults 20+years, b) early childhood and middle 51 183 childhood versus adolescents, and c) adults versus seniors 52 184 53 185 54 186 55 187 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 188 4 189 5 6 190 7 191 8 192 9 193 10 194 11 195 12 13 14 15 16 For peer review only 17 18 19 20 21 Table 3. Mean serum omega-3 fatty acid concentrations (% of total fatty acids) in 22 women who were pregnant/breastfeeding (N=21) and women of childbearing age 23 (N=319) 24 25 95% CI for Women of 95% CI for P Pregnant/ 26 Mean childbearing Mean 27 breastfeeding 28 age 29 Fatty acid Mean (SE) Mean (SE) 30 EPA 0.34 (0.07) 0.19-0.49 0.47 (0.03) 0.39-0.53 0.17 31

32 DHA 1.52 (0.08) 1.35-1.68 1.37 (0.07) 1.21-1.52 0.10 http://bmjopen.bmj.com/ 33 Sum Omega- 1.87-2.49 2.02-2.46 0.73 2.18 (0.15) 2.24 (0.10) 34 3 35 36 Sample sizes are unweighted, however, estimates for means and SE are weighted. 37 Women of childbearing age are those, age 15-44 years. 38 39

32 http://bmjopen.bmj.com/ 33 34 196 Table 4. Prevalence, N (%) of EPA+DHA fatty acid concentrations <2.49%, which is 35 197 approximately equivalent to DGA dietary recommendations for EPA and DHA intake 36 198 by gender and age group in NHANES 2011-2012. 37 38 Children, Adults, Early Middle Adolescen Adults Seniors 39 3-19 yrs 20 yrs+ childhoo childhoo ts 40 on September 24, 2021 by guest. Protected copyright. 41 d d 42 <2.49% EPA + DHA 43 44 All 880 856 (68.3) 168 488 224 (95.5) 593 (76.0) 263 (52.9) 45 (95.5) (97.4) (94.8) 46 Male 429 445 (72.4) 85 (97.3) 239 105 (97.6) 313 (80.7) 132 (53.7) 47 48 (96.6) (94.7) 49 Femal 451 411 (64.6) 83 (97.5) 249 119 (93.6) 280 (71.4) 131 (52.3) 50 e (94.5) (94.9) 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

BMJ Open Page 32 of 39 A 100

1 90 2 80 3 4 70 5 97.1 98.0 96.2 6 60 7 50 8 9 40 Percent 10 For peer review only 30

11 http://bmjopen.bmj.com/ 12 20 13 14 10 15 0 16 17 All Males Females 18 DHA ≥2.85% DHA <2.85% 19 on September 24, 2021 by guest. Protected copyright. 20 B 21 100 22 90 23 24 80 25 26 70 27 60 28 88.7 92.1 85.6 29 50 30 40 31 Percent 32 30 33 34 20 35 10 12.9 36 10.4 7.58 37 0 38 All Males Females 39 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 40 Low risk Intermediate risk High risk 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from

Page 33 of 39 BMJ Open

1 2 3 4 Supplementary Table 1. Mean omega 3 fatty acids (% of total fatty acids) by gender and age group in NHANES 2011-2012 5 6 Males Females P 7 Age group Variable N Mean (SE) 95%CI N Mean (SE) 95% CI 8 Children, 3-19yrs EPA 457 0.30 (0.02) 0.26-0.33 488 0.27 (0.01) 0.25-0.30 0.45 9 10 DHA 457 1.05 (0.02) 1.00-1.10 488 1.10 (0.03) 1.04-1.16 0.18 11 sum omega 3 457 1.75 (0.04) 1.67-1.84 488 1.75 (0.04) 1.66-1.84 0.98 12 Adults, 20yrs+ EPAFor 64peer1 0.55 (0.03) review0.50-0.61 675 0.65only (0.03) 0.58-0.73 0.01 13 DHA 641 1.27 (0.05) 1.17-1.37 675 1.49 (0.07) 1.34-1.63 <0.001 14 sum omega 3 641 2.27 (0.07) 2.11-2.42 675 2.58 (0.09) 2.39-2.78 <0.001 15 Early childhood EPA 90 0.26 (0.01) 0.23-0.29 89 0.25 (0.01) 0.23-0.28 0.64 16 3-5yrs DHA 90 1.06 (0.04) 0.98-1.14 89 0.97 (0.04) 0.89-1.04 0.73

17 http://bmjopen.bmj.com/ 18 sum omega 3 90 1.71 (0.06) 1.59-1.83 89 1.60 (0.05) 1.50-1.70 0.15 19 Middle childhood EPA 255 0.31 (0.03) 0.25-0.36 264 0.27 (0.02) 0.23-0.31 0.02 20 6-11yrs DHA 255 1.10 (0.05) 1.00-1.21 264 1.08 (0.03) 1.01-1.15 0.63 21 sum omega 3 255 1.83 (0.08) 1.66-2.00 264 1.73 (0.05) 1.62-1.84 0.17 22 Adolescents EPA 112 0.30 (0.02) 0.25-0.34 135 0.28 (0.02) 0.25-0.31 0.40 23 12-19yrs DHA 112 1.01 (0.03) 0.95-1.07 135 1.14 (0.05) 1.04-1.23 0.03 24 sum omega 3 112 1.71 (0.05) 1.62-1.81 135 1.80 (0.06) 1.67-1.93 0.25

25 on September 24, 2021 by guest. Protected copyright. 26 Adults EPA 404 0.49 (0.02) 0.46-0.53 414 0.53 (0.03) 0.46-0.60 0.29 27 20-55yrs DHA 404 1.18 (0.05) 1.08-1.28 414 1.40 (0.07) 1.25-1.55 0.001 28 sum omega 3 404 2.11 (0.06) 1.99-2.24 414 2.35 (0.10) 2.13-2.57 0.01 29 Seniors EPA 237 0.68 (0.06) 0.56-0.80 261 0.88 (0.10) 0.67-1.09 0.11 30 >55yrs DHA 237 1.47 (0.05) 1.36-1.58 261 1.64 (0.08) 1.47-1.80 0.03 31 sum omega 3 237 2.62 (0.11) 2.39-2.85 261 3.00 (0.15) 2.68-3.32 0.04 32 33 p-values represent differences between males and females within a respective life stage. 34 35 36 37 38 39 40 41 42 43 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from

BMJ Open Page 34 of 39

1 2 3 4 Supplementary Table 2. Mean omega 3 fatty acids (% of total fatty acids) by race/ethnicity in NHANES 2011-2012. 5 6 Children, 3-19 yrs Adults, 20 yrs+ 7 Race/ethnicity Fatty acid N Mean (SE) 95%CI P N Mean (SE) 95%CI P 8 Mexican American/Hispanic EPA 312 0.23 (0.01) 0.21-0.26 0.01 293 0.44 (0.02) 0.40-0.47 <0.001 9 DHA 1.02 (0.03) 0.97-1.08 0.72 1.26 (0.04) 1.20-1.45 0.29 10 sum omega 3 1.62 (0.04) 1.54-1.71 0.08 2.10 (0.05) 1.99 -2.21 0.01 11 Non-Hispanic White EPA 202 0.30 (0.02) 0.26-0.34 Ref 527 0.63 (0.04) 0.55-0.70 Ref 12 ForDHA peer 1.04review (0.03) 0.98-1.10 onlyRef 1.32 (0.06) 1.20-1.45 Ref 13 14 sum omega 3 1.75 (0.05) 1.65-1.85 Ref 2.40 (0.09) 2.21-2.60 Ref 15 Non-Hispanic Black EPA 269 0.29 (0.02) 0.25-0.32 0.67 323 0.55 (0.03) 0.49-0.61 0.08 16 DHA 1.20 (0.03) 1.14-1.26 <0.001 1.55 (0.06) 1.43-1.68 0.01

17 sum omega 3 1.88 (0.05) 1.78-1.97 0.02 2.55 (0.09)http://bmjopen.bmj.com/ 2.36-2.73 0.19 18 Asian EPA 98 0.37 (0.02) 0.32-0.42 0.04 167 0.94 (0.07) 0.79-1.10 0.003 19 DHA 1.40 (0.05) 1.30-1.51 <0.001 2.13 (0.09) 1.95-2.30 <0.001 20 21 sum omega 3 2.18 (0.06) 2.06-2.30 <0.001 3.54 (0.15) 3.22-3.87 <0.001 22 Other/unknown EPA 41 0.27 (0.02) 0.22-0.32 0.29 29 0.60 (0.06) 0.47-0.73 0.75 23 DHA 1.03 (0.05) 0.95-1.11 0.81 1.59 (0.21) 1.15-2.04 0.22 24 sum omega 3 1.68 (0.06) 1.56-1.80 0.38 2.67 (0.26) 2.12-3.21 0.35

25 p-values represent differences between race/ethnicity within a respective life stage. For example, relative to on September 24, 2021 by guest. Protected copyright. non-Hispanic white 26 children, Asian children had higher EPA (P=0.04). 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 35 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 4 Supplementary Table 3. Mean omega 3 fatty acids (% of total fatty acids) by IOM age groups 5 in NHANES 2011-2012. 6 7 Age group Variable N Mean (SE) 95%CI 8 1-3yrs EPA 55 0.25 (0.02) 0.20-0.31 9 10 DHA 55 0.98 (0.04) 0.88-1.07 11 sum omega 3 55 1.60 (0.08) 1.44-1.76 12 4-8yrs EPA 382 0.27 (0.01) 0.25-0.28 13 DHA 382 1.04 (0.02) 1.00-1.08 14 sum omega 3 382 1.70 (0.03) 1.65-1.76 15 9-13yrs EPA 325 0.29 (0.03) 0.23-0.34 16 For peerDHA review325 only1.13 (0.04) 1.05-1.20 17 18 sum omega 3 325 1.82 (0.07) 1.67-1.96 19 14-18yrs EPA 145 0.29 (0.02) 0.26-0.33 20 DHA 145 1.05 (0.03) 0.99-1.11 21 sum omega 3 145 1.74 (0.04) 1.66-1.83 22 19-30yrs EPA 258 0.41 (0.03) 0.35-0.48 23 DHA 258 1.27 (0.08) 1.11-1.43 24 sum omega 3 258 2.09 (0.10) 1.88-2.30 25 26 31-50yrs EPA 425 0.52 (0.02) 0.48-0.56 27 DHA 425 1.26 (0.05) 1.15-1.36 28 sum omega 3 425 2.21 (0.07) 2.06-2.36 29 51-70yrs EPA 409 0.76 (0.06) 0.63-0.89 30 DHA 409 1.49 (0.06) 1.36-1.61 31 sum omega 3 409 2.71 (0.11) 2.49-2.94

32 http://bmjopen.bmj.com/ 33 >70yrs EPA 156 0.80 (0.09) 0.62-0.98 34 DHA 156 1.73 (0.11) 1.51-1.96 35 sum omega 3 156 3.02 (0.20) 2.61-3.43 36 37 38 39

40 on September 24, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Supplementary Table 4. Fatty acids (% of total fatty acids) measured in NHANES 2011-2012 4 5 Children, 3-19 Adults, 20+ 6 Fatty acid Mean (SE) 95% CI Mean (SE) 95% CI 7 8 Capric acid, 10:0 0.02 (0.00) 0.02-0.02 0.03 (0.00) 0.03-0.04 9 Lauric acid, 12:0 0.10 (0.01) 0.09-0.11 0.15 (0.01) 0.13-0.17 10 11 Myristic acid, 14:0 1.10 (0.02) 1.07-1.13 1.09 (0.03) 1.03-1.15 12 Pentadecanoic acid, 15:0 0.21 (0.00) 0.20-0.22 0.21 (0.00) 0.20-0.22 13 Palmitic acid, 16:0 24.1 (0.07) 23.9-24.2 23.1 (0.13) 22.8-23.3 14 15 Margaric acid, 17:0 0.27 (0.00) 0.26-0.27 0.28 (0.00) 0.27-0.28 16 Stearic acid, 18:0 For peer 5.80review (0.02) 5.76 -only5.84 6.29 (0.06) 6.16-6.42 17 18 Arachidic acid, 20:0 0.20 (0.00) 0.20-0.21 0.22 (0.00) 0.21-0.22 19 Docosanoic acid, 22:0 0.58 (0.01) 0.56-0.60 0.61 (0.01) 0.59-0.62 20 Tricosanoic acid, 23:0 0.25 (0.00) 0.25-0.26 0.26 (0.00) 0.26-0.27 21 22 Lignoceric acid, 24:0 0.50 (0.01) 0.48-0.51 0.52 (0.01) 0.50-0.53 23 Myristoleic acid, 14:1n-5 0.07 (0.00) 0.06-0.07 0.07 (0.00) 0.06-0.07 24 25 Palmitoleic acid, 16:1n-7 2.03 (0.03) 1.97-2.10 1.56 (0.03) 1.49-1.62 26 cis-Vaccenic acid, 18:1n-7 1.24 (0.01) 1.21-1.26 1.10 (0.01) 1.08-1.13 27 28 Oleic acid, 18:1n-9 18.2 (0.09) 18.0-18.4 17.3 (0.13) 17.0-17.6 29 Eicosenoic acid, 20:1n-9 0.12 (0.00) 0.12-0.12 0.12 (0.00) 0.12-0.12 30 Nervonic acid, 24:1n-9 0.74 (0.01) 0.72-0.77 0.74 (0.01) 0.72-0.75 31

32 Linoleic acid, 18:2n-6 31.4 (0.12) 31.2-31.7 34.5 (0.28) 34.0-35.1 http://bmjopen.bmj.com/ 33 Alpha-linolenic acid, 18:3n-3 0.73 (0.01) 0.71-0.76 0.71 (0.01) 0.68-0.74 34 35 Gamma-linolenic acid, 18:3n-6 0.51 (0.01) 0.49-0.53 0.44 (0.02) 0.41-0.48 36 Stearidonic acid, 18:4n-3 0.03 (0.00) 0.03-0.03 0.02 (0.00) 0.02-0.02 37 Eicosadienoic acid, 20:2n-6 0.19 (0.00) 0.19-0.20 0.20 (0.00) 0.19-0.21 38 39 Homo-gamma-linolenic acid, 20:3n-6 1.35 (0.01) 1.33-1.37 1.36 (0.03) 1.31-1.42

40 Eicosatrienoic acid, 20:3n-9 0.07 (0.00) 0.06-0.07 0.06 (0.00) 0.05-0.06 on September 24, 2021 by guest. Protected copyright. 41 42 Arachidonic acid, 20:4n-6 7.35 (0.07) 7.20-7.50 6.87 (0.08) 6.71-7.04 43 Docosatetraenoic acid, 22:4n-6 0.22 (0.00) 0.21-0.23 0.26 (0.00) 0.25-0.27 44 45 Docosapentaenoic acid, 22:5n-6 0.17 (0.00) 0.16-0.17 0.22 (0.00) 0.21-0.23 46 Data shown are weighted means and 95% confidence intervals. Fatty acids shown represent 32 47 of the 35 fatty acids measured in NHANES. Together with EPA, DHA and DPA (shown in the 48 manuscript tables) were summed to generate total fatty acids to derive the relative concentration 49 (%) of fatty acids. 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 4 Supplementary Table 5. Prevalence, N (%) of omega 3 fatty acid concentrations below those 5 identified as having cardiovascular benefits by gender and age group in NHANES 2011-2012. 6 7 8 Adults, 20 yrs+ Adults Seniors 9 DHA ≥2.85%, lower risk of atrial fibrillation 10 All 61 (2.93) 27 (2.07) 34 (4.65) 11 Male 19 (1.98) 8 (1.39) 11 (3.30) 12 13 Female 42 (3.80) 19 (2.72) 23 (5.70) 14 Omega-3 Index >6.23%, Low risk 15 All 20 (1.0) 10 (0.57) 10 (1.86) 16 Male For peer7 (0.34) review4 (0.37) only 3 (0.29) 17 Female 13 (1.60) 6 (0.78) 7 (3.08) 18 Omega-3 Index 3.11-6.23%, Intermediate risk 19 20 All 138 (10.4) 61 (6.59) 77 (17.9) 21 Male 48 (7.58) 21 (4.61) 27 (14.3) 22 Female 90 (12.9) 40 (8.52) 50 (20.7) 23 Omega-3 Index <3.11%, High risk 24 All 1158 (88.7) 747 (92.8) 411 (80.2) 25 26 Male 586 (92.1) 379 (95.0) 207 (85.4) 27 Female 572 (85.6) 368 (90.7) 204 (76.2) 28 Parallels to omega 3 levels were drawn from Sun et al. (22) to approximate the prevalence of those 29 30 with omega 3 levels equivalent to omega 3 dietary intake recommendations by the DGA, the lower 31 range of the tertiles of serum DHA associated with lower risk of atrial fibrillation by Virtanen et

32 al. (33), and cardiovascular risk categories defined by the Omega-3 Index (23,35). http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

STROBE (Strengthening The Reporting of OBservational Studies in Epidemiology) Checklist BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 A checklist of items that should be included in reports of observational studies. You must report the page number in your manuscript 4 where you consider each of the items listed in this checklist. If you have not included this information, either revise your manuscript 5 accordingly before submitting or note N/A. 6 7 8 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and published 9 examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely available on the Web 10 sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at http://www.annals.org/, and Epidemiology 11 12 at http://www.epidem.com/). Information on the STROBE Initiative is available at www.strobe-statement.org. 13 14 Section and Item Item Reported on 15 Recommendation No. Page No. 16 Title and Abstract 1 For(a) Indicate peer the study’s design review with a commonly only used term in the title or the 17 18 abstract 19 20 (b) Provide in the abstract an informative and balanced summary of what was 21 done and what was found 22 23 Introduction 24 Background/Rationale 2 Explain the scientific background and rationale for the investigation being 25 26 reported 27 28 Objectives 3 State specific objectives, including any prespecified hypotheses 29 30 Methods 31 Study Design 4 Present key elements of study design early in the paper

32 http://bmjopen.bmj.com/ 33 Setting 5 Describe the setting, locations, and relevant dates, including periods of 34 35 recruitment, exposure, follow-up, and data collection 36 37 Participants 6 (a) Cohort study—Give the eligibility criteria, and the sources and methods of 38 selection of participants. Describe methods of follow-up 39 Case-control study—Give the eligibility criteria, and the sources and methods of 40 on September 24, 2021 by guest. Protected copyright. 41 case ascertainment and control selection. Give the rationale for the choice of 42 cases and controls 43 44 Cross-sectional study—Give the eligibility criteria, and the sources and methods of 45 selection of participants 46 47 (b) Cohort study—For matched studies, give matching criteria and number of 48 exposed and unexposed 49 50 Case-control study—For matched studies, give matching criteria and the number 51 52 of controls per case 53 54 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and 55 effect modifiers. Give diagnostic criteria, if applicable 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from Section and Item Item Reported on 1 Recommendation 2 No. Page No. 3 Data Sources/ 8* For each variable of interest, give sources of data and details of methods of 4 Measurement assessment (measurement). Describe comparability of assessment methods if 5 there is more than one group 6 7 Bias 9 Describe any efforts to address potential sources of bias 8 9 Study Size 10 Explain how the study size was arrived at 10 11 Quantitative Variables 11 Explain how quantitative variables were handled in the analyses. If applicable, 12 describe which groupings were chosen and why 13 14 Statistical Methods 12 (a) Describe all statistical methods, including those used to control for 15 16 Forconfounding peer review only 17 18 (b) Describe any methods used to examine subgroups and interactions 19 20 (c) Explain how missing data were addressed 21 22 (d) Cohort study—If applicable, explain how loss to follow-up was addressed 23 24 Case-control study—If applicable, explain how matching of cases and controls was 25 addressed 26 27 Cross-sectional study—If applicable, describe analytical methods taking account of 28 sampling strategy 29 30 (e) Describe any sensitivity analyses 31

32 Results http://bmjopen.bmj.com/ 33 34 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers potentially 35 eligible, examined for eligibility, confirmed eligible, included in the study, 36 37 completing follow-up, and analysed 38 39 (b) Give reasons for non-participation at each stage

40 on September 24, 2021 by guest. Protected copyright. 41 (c) Consider use of a flow diagram 42 43 Descriptive Data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) and 44 information on exposures and potential confounders 45 46 (b) Indicate number of participants with missing data for each variable of interest 47 48 (c) Cohort study—Summarise follow-up time (eg, average and total amount) 49 50 Outcome Data 15* Cohort study—Report numbers of outcome events or summary measures over 51 time 52 53 Case-control study—Report numbers in each exposure category, or summary 54 measures of exposure 55 56 Cross-sectional study—Report numbers of outcome events or summary measures 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from Section and Item Item Reported on 1 Recommendation 2 No. Page No. 3 Main Results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates 4 and their precision (eg, 95% confidence interval). Make clear which confounders 5 were adjusted for and why they were included 6 7 (b) Report category boundaries when continuous variables were categorized 8 9 (c) If relevant, consider translating estimates of relative risk into absolute risk for a 10 meaningful time period 11 12 Other Analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and 13 14 sensitivity analyses 15 16 Discussion For peer review only 17 18 Key Results 18 Summarise key results with reference to study objectives 19 20 Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or 21 imprecision. Discuss both direction and magnitude of any potential bias 22 23 Interpretation 20 Give a cautious overall interpretation of results considering objectives, limitations, 24 multiplicity of analyses, results from similar studies, and other relevant evidence 25 26 Generalisability 21 Discuss the generalisability (external validity) of the study results 27 28 Other Information 29 30 Funding 22 Give the source of funding and the role of the funders for the present study and, if 31 applicable, for the original study on which the present article is based

32 http://bmjopen.bmj.com/ 33 34 35 *Give information separately for cases and controls in case-control studies and, if applicable, for exposed and unexposed groups in 36 cohort and cross-sectional studies. 37

38 Once you have completed this checklist, please save a copy and upload it as part of your submission. DO NOT include this 39 checklist as part of the main manuscript document. It must be uploaded as a separate file. 40 on September 24, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from

Long Chain Omega-3 Fatty Acid Serum Concentrations Across Life Stages in the United States: An Analysis of NHANES 2011-2012 ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2020-043301.R1

Article Type: Original research

Date Submitted by the 18-Nov-2020 Author:

Primary Subject Nutrition and metabolism Heading:

Secondary Subject Heading: Epidemiology, Mental health

PUBLIC HEALTH, NUTRITION & DIETETICS, Coronary heart disease < http://bmjopen.bmj.com/ Keywords: CARDIOLOGY, EPIDEMIOLOGY

on September 24, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 38 BMJ Open

1 2 3

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 24, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Long Chain Omega-3 Fatty Acid Serum Concentrations Across Life Stages in the United 4 5 6 States: An Analysis of NHANES 2011-2012 7 8 9 10 Rachel A Murphy1,2, Prasad P Devarshi3 Shauna Ekimura3, Keri Marshall3, Susan Hazels 11 , 12 13 Mitmesser3 14 15 16 For peer review only 17 1 th 18 Cancer Control Research, BC Cancer, 2-107, 675 W 10 Ave, Vancouver, BC, V5Z 1L3 19 20 2School of Population and Public Health, University of British Columbia, 2206 East Mall, 21 22 Vancouver, BC, V6T 1Z3, Canada 23 24 3 25 Science & Technology, Pharmavite LLC, West Hills, CA 91304, USA 26 27 28 29 Corresponding Author: 30 31

32 Rachel A Murphy, PhD http://bmjopen.bmj.com/ 33 34 167-2206 East Mall 35 36 University of British Columbia 37 38 39 Vancouver, BC

40 on September 24, 2021 by guest. Protected copyright. 41 V6T 1Z3 42 43 44 Email: [email protected] 45 46 Phone: 604-822-1397 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Abstract 4 5 6 Objective: To determine reference ranges of circulating long chain (LC) omega-3 fatty 7 8 acids: eicosapentaenoic acid (EPA), docosapentaenoic acid (DPA), and docosahexaenoic 9 10 acid (DHA) in a nationally representative population of Americans. To provide context, 11 12 13 serum concentrations of LC omega-3 were compared to concentrations associated with 14 15 consuming the recommended amount of EPA and DHA by the Dietary Guidelines for 16 For peer review only 17 Americans (DGA) and the Omega-3 Index (EPA+DHA). 18 19 20 Design: Cross-sectional population-based study 21 22 Setting: The National Health and Nutrition Examination Survey (NHANES) 2011-2012 cycle 23 24 Participants: Participants with fatty acids measured in serum: 945 children, age 3-19 years, 25 26 27 and 1,316 adults, age 20 and older. 28 29 Main measure: Serum EPA, DPA, DHA and sum of LC omega-3 fatty acids expressed as % of 30 31 total fatty acids.

32 http://bmjopen.bmj.com/ 33 34 Results: Among children, mean (SE) serum concentrations of EPA, DHA and omega-3s were 35 36 0.28% (0.01), 1.07% (0.02) and 1.75% (0.03). Among adults, mean (SE) of EPA, DHA and 37 38 omega-3s were 0.61% (0.02), 1.38% (0.05), and 2.43% (0.08), all of which were 39

40 on September 24, 2021 by guest. Protected copyright. 41 significantly higher than corresponding serum fatty acid concentrations in children 42 43 (P<0.001). Despite recommendations for higher intake, pregnant and/or breastfeeding 44 45 46 women had mean (SE) EPA, DHA and LC omega-3 concentrations of 0.34% (0.07), 1.52% 47 48 (0.08) and 2.18% (0.15) which were comparable to women of childbearing age; p=0.17, 49 50 p=0.10 and p=0.73. Over 95% of children and 68% of adults had LC omega-3 51 52 53 concentrations below those associated with the DGA recommendation. Approximately 89% 54 55 of adults had an Omega-3 Index in the high cardiovascular risk category. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Conclusions: Contemporary reference ranges for circulating LC omega-3s are critical for 4 5 6 setting public health recommendations. Our findings show the need for continued 7 8 emphasis on regular consumption of LC omega-3s among Americans, particularly 9 10 considering the importance of LC omega-3s in cardiovascular health, brain health and 11 12 13 development throughout life. 14 15 Article Summary 16 For peer review only 17 Strengths and limitations of this study 18 19 20  This study is one of the few to provide reference range data on circulating LC omega-3 21 22 fatty acids, a biomarker of LC omega-3 nutritional status using data that is 23 24 representative of the children (3 and older) and adults in the US population 25 26 27  Strengths of the study include the study design which allows inference to the general 28 29 population in the US, inclusion of multiple life stages (children, adolescents, adults, 30 31 seniors and pregnant or breastfeeding women) and serum biomarker measurement of 32 http://bmjopen.bmj.com/ 33 34 LC omega-3 fatty acid intake 35 36  Limitations of the study include the lack of information on long-term LC omega-3 fatty 37 38 39 acid intake, and possible shifts in diet and LC omega-3 status since the time of

40 on September 24, 2021 by guest. Protected copyright. 41 measurement in 2011-2012 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 4 5 6 Background 7 8 9 Consumption of seafood rich in long-chain (LC) omega-3 polyunsaturated fatty acids 10 11 12 (omega-3 PUFAs), from seafood and dietary supplements, is recommended as part of a 13 14 healthy diet (1,2). A large body of evidence has shown that dietary patterns including high 15 16 consumption of LCFor omega-3 peer PUFAs eicosapentaenoic review acid only (EPA) and docosahexaenoic acid 17 18 19 (DHA), both from diet and supplementation, are associated with decreased risk of 20 21 cardiovascular disease (3), type 2 diabetes (4,5), lower overall mortality (6), and lower 22 23 blood pressure (7,8). Strong evidence also supports a positive relationship between 24 25 26 maternal dietary intake of EPA and DHA and neurodevelopment among infants (9,10). 27 28 Further research also supports the role of EPA and DHA in neuropsychological health, such 29 30 that EPA and DHA supplementation is considered as a part of integrative therapies for 31

32 http://bmjopen.bmj.com/ 33 depressive disorder by the American Psychiatric Association (11). 34 35 36 The most recent Dietary Guidelines for Americans (DGA 2010-2015 and 2015- 37 38 2020) recommends that individuals 2 and older consume 8 ounces per week of a variety of 39

40 on September 24, 2021 by guest. Protected copyright. 41 seafood to provide approximately 250 mg of EPA and DHA per day (2). Consuming this 42 43 amount is associated with reduced cardiac death in those with and without pre-existing 44 45 46 cardiovascular disease (2). Similarly, the American Heart Association has recommended 47 48 consuming 1-2 seafood meals per week to reduce the risk of congestive heart failure, 49 50 coronary heart disease, ischemic stroke, and sudden cardiac death (12). Women who are 51 52 53 pregnant or breastfeeding are encouraged to consume up to 12 ounces per week to obtain 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 enough omega-3 in their diet to support infant health, ensuring they consume at least 300 4 5 6 mg/d EPA+DHA, of which ≥200 mg/d should be DHA specifically (2,13). 7 8 Despite the health benefits of omega-3s, data from the National Health and Nutrition 9 10 Examination Survey (NHANES) clearly shows that dietary intake of omega-3s in Americans 11 12 13 is low and consistently well below the DGA recommendation across all age groups, gender, 14 15 and race/ethnicity (2,14–16). In 2015, the DGA Committee reviewed data from What We 16 For peer review only 17 Eat in America/NHANES and found 90% of Americans were below the recommended 18 19 20 intake of 250 mg DHA+EPA (2). Furthermore, an analysis of LC omega-3 dietary intake 21 22 from NHANES 2003-2014 reported that younger individuals and women may be 23 24 particularly at risk: the mean EPA+DHA intake for children 1-5 years was just 23.1 mg per 25 26 27 1000kcal, which is below the DGA recommendation for EPA+DHA (17). A further study of 28 29 NHANES 2001-2014 reported that more than 95% of women of childbearing-age did not 30 31 meet the DGA recommendations for EPA+DHA (14).

32 http://bmjopen.bmj.com/ 33 34 Collectively, the dietary intake data from NHANES, as well as other populations 35 36 (18,19) provide compelling evidence of gaps in LC omega-3 intake. However, dietary intake 37 38 is limited by measurement error (20). Use of error-prone dietary assessment tools can lead 39

40 on September 24, 2021 by guest. Protected copyright. 41 to spurious findings, contributes to heterogeneity in evidence on diet and health and has 42 43 led some to be critical of self-reported dietary data being used to inform public health 44 45 46 recommendations (21). Biomarkers of dietary intake (i.e. circulating omega-3s) may 47 48 provide more accurate information on LC omega-3 status, and may also confirm findings 49 50 from studies of dietary intake. Accurate reference range data for fatty acids is critical for 51 52 53 identifying at risk populations and setting appropriate public health recommendations. 54 55 However, reference data on fatty acids in large, representative populations is scarce. The 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 last data of this kind in the US was from NHANES 2003-2004 in adults only. Therefore, the 4 5 6 aim of this study was to provide contemporary reference ranges for LC omega-3 fatty acids 7 8 in the US across all life stages using the most recent serum fatty acid data (2011-2012) 9 10 available in NHANES. An additional objective was to determine the proportion of the US 11 12 13 population with serum LC omega-3 fatty acid concentrations below concentrations 14 15 equivalent to DGA intake recommendations and below concentrations associated with 16 For peer review only 17 cardioprotection. 18 19 20 21 22 Materials and Methods 23 24 This study uses the NHANES 2011-2012 public-use data files. NHANES is designed 25 26 27 to assess the nutritional status and health of children and adults in the US, using a complex 28 29 multistage probability sampling design that is representative of the national civilian 30 31 population in the US (22). Fatty acids were measured in serum with the goal of obtaining

32 http://bmjopen.bmj.com/ 33 34 US reference ranges for most circulating fatty acids in a subsample of participants. The 35 36 subsample was selected to be a nationally representative sample. Participants were chosen 37 38 at random from all participants, age 3 to 11 who attended mobile examinations and fasting 39

40 on September 24, 2021 by guest. Protected copyright. 41 participants aged 12 and older who were examined in the morning session. Of the 5,643 42 43 participants in the subsample, 397 participants did not have a lab specimen. Fatty acids 44 45 46 were measured using modified methods of Lagerstedt et al. (23). Briefly, total fatty acids 47 48 were hexane-extracted along with an internal standard solution for fatty acid recovery. The 49 50 extract was derivatized to form pentafluorobenzyl esters and injected onto a capillary gas 51 52 53 chromatograph column. A total of 30 dietary fatty acids were quantitated. The lower limit 54 55 of detection for EPA, DHA and DPA were 0.79µmol/L, 1.84µmol/L and 0.24µmol/L. Fatty 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 acids were expressed as a % of total fatty acids (respective fatty acid/sum of 30 fatty acids 4 5 6 measured) in this analysis to facilitate comparison with other studies, which 7 8 predominately report fatty acids as a relative percent (24–26). Missing data (not including 9 10 those with values below the limit of detection) for a given fatty acid ranged from a low of 11 12 13 N= 1,222 (EPA) to 1,861 (eicosatrienoic acid). As relative fatty acids require a common 14 15 denominator of total fatty acids, individuals who were missing data for any fatty acid were 16 For peer review only 17 excluded (N=2,983) which resulted in an overall sample of 2,261 participants: 945 18 19 20 children, age 3-19 and 1,316 adults, age 20 and older. 21 22 Mean EPA, DHA and the sum of LC omega-3s (EPA+DPA+DHA), were calculated for 23 24 children and adults as well as by key life stages: early childhood (3-5 years), middle 25 26 27 childhood (6-11 years), adolescents (12-19 years), adults (20-55), seniors (>55) and 28 29 women who were pregnant and/or breastfeeding. Pregnancy status was determined using 30 31 the variable pregnancy status at exam (RIDEXPREG) from the demographics data and the

32 http://bmjopen.bmj.com/ 33 34 variable pregnancy test (URXPREG) from the laboratory data. Breastfeeding status was 35 36 determined from the reproductive health data (RHQ200). Women of childbearing age was 37 38 defined as those 15-44 years of age, which corresponded to the age range of women who 39

40 on September 24, 2021 by guest. Protected copyright. 41 reported being pregnant and/or breastfeeding. Each life stage was also analyzed by gender 42 43 and race/ethnicity. Supplements containing fish oil, cod liver oil, salmon oil, krill oil and 44 45 46 DHA were referred to in this analysis as ‘LC omega-3 supplements’ and were identified 47 48 from participants report of dietary supplement use in the past 30 days. 49 50 Serum LC omega-3 fatty acids reflect short-term (weeks to months) intake of 51 52 53 dietary fat (27,28). To provide additional public health context to findings, comparisons 54 55 were made to fatty acid concentrations (LC omega-3s=2.49%), previously identified as 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 approximately equivalent to DGA dietary recommendations of 250 mg per day of EPA and 4 5 6 DHA (29,30). Comparisons were also made to serum fatty acid concentrations associated 7 8 with cardioprotection. Virtanen et al. (31) reported lower risk of atrial fibrillation; a 9 10 common cardiac arrhythmia among participants with serum DHA ≥2.85%. Reduced risk of 11 12 13 mortality, sudden cardiac arrest and other adverse cardiac events have been consistently 14 15 associated with higher red blood cell (RBC) concentrations of EPA+DHA, also known as the 16 For peer review only 17 Omega-3 Index (32). To compare serum NHANES data with the Omega-3 Index, an 18 19 20 equation relating the RBC-based metric with plasma EPA + DHA concentrations was 21 22 applied to calculate risk categories for cardiovascular disease (33,34). This equation was 23 24 chosen based on previous evidence that the fatty acid composition of serum and plasma are 25 26 27 similar (35). Application of the equation resulted in the following Omega-3 Index 28 29 categories: low risk >6.23%, intermediate risk 3.11-6.23%, or high risk <3.11%. 30 31 The following covariates were used to characterize the sample population: age,

32 http://bmjopen.bmj.com/ 33 34 gender, race/ethnicity (Non-Hispanic White, Hispanic/Mexican-American, Non-Hispanic 35 36 Black, Asian and other/unknown), and for adults: poverty income ratio (PIR; low: 0-1.85, 37 38 medium >1.85-3.50 and high: >3.50), education (high 39

40 on September 24, 2021 by guest. Protected copyright. 41 school), current smoking status (yes/no) and BMI. 42 43 Sample sizes in the text and tables are unweighted. However, estimates for means, 44 45 46 proportions, standard errors (SE) and 95% confidence intervals (CI) were weighted using 47 48 sampling weights for the fatty acid subsample. Variance estimate was determined using the 49 50 Taylor Series Linearization approach. Differences between groups were determined by 51 52 53 linear regression at p<0.05. Statistical analyses were performed using survey procedures 54 55 in STATA software version 14.2 (StataCorp, College Station, TX, USA). 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Patient and public involvement 4 5 6 Patients and the public were not directly involved in the design, conduct, reporting or 7 8 dissemination of this study. 9 10 11 12 13 Results 14 15 Demographic characteristics 16 For peer review only 17 A total of 2,261 participants were included (Table 1), 47.6% were male, 52.5% were 18 19 20 female; 41.8% were 3-19 years of age; 58.2% were 20 years of age and older; 16.5% were 21 22 Mexican American or Hispanic, 65.2% were non-Hispanic White, 12.1% were non-Hispanic 23 24 black and 4.55% were Asian. With respect to education, 5.36% had education below high 25 26 27 school, 11.2% had a high school education, 63.7% had some college or higher education, 28 29 and 19.7% had education in between (more than high school). Among adults, 20.2% were 30 31 current smokers; 9.90% had low PIR, 32.7% had medium PIR and 66.3% had high PIR. The

32 http://bmjopen.bmj.com/ 33 34 mean (SE) BMI among adults was 28.7 (0.34) kg/m2. LC omega-3 supplement use was low, 35 36 with only 7.3% (N=166) in the overall population reporting use in the prior 30 days. 37 38 Mean concentrations of omega-3 fatty acids 39

40 on September 24, 2021 by guest. Protected copyright. 41 Mean serum concentrations of EPA, DHA and LC omega-3s by age group are shown 42 43 in Table 2. Among children 3-19 years, mean (SE) EPA, DHA and LC omega-3s were 0.28% 44 45 46 (0.01), 1.07% (0.02), and 1.75% (0.03), respectively. Among adults 20 and older, mean 47 48 (SE) EPA, DHA and the sum of LC omega-3s were 0.61% (0.02), 1.38% (0.05) and 2.43% 49 50 (0.08), respectively. Concentrations of EPA, DHA and the sum of LC omega-3s were all 51 52 53 significantly lower (p<0.001) in children compared to adults. Mean (SE) concentrations of 54 55 EPA, DHA and the sum of LC omega-3s in early childhood were 0.26% (0.01), 1.01% (0.03), 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 1.65% (0.04); middle childhood: 0.29% (0.02), 1.09% (0.03), 1.78% (0.06); and 4 5 6 adolescents: 0.29% (0.02), 1.08% (0.03) and 1.76% (0.04). Compared to adolescents, the 7 8 sum of LC omega-3s were lower in early childhood (p=0.03). Similarly, EPA and DHA 9 10 tended to be lower in early childhood (p=0.06 and p=0.07). Mean (SE) concentrations of 11 12 13 EPA, DHA and the sum of LC omega-3s in adults 20-55 years old were 0.51% (0.02), 1.29% 14 15 (0.06), 2.23% (0.08), which were significantly lower than in seniors (all p<0.001): 0.79% 16 For peer review only 17 (0.06), 1.56% (0.06), 2.83% (1.00). The prevalence of LC omega-3 supplement use ranged 18 19 20 from low use N=3 (1.21%) in adolescents to high use N=84 (16.9%) in seniors. 21 22 Mean serum concentrations of EPA, DHA and the sum of LC omega-3s by life stage 23 24 and gender are shown in Supplementary Table 1 and by race/ethnicity among children and 25 26 27 adults in Supplementary Table 2. Among adults 20 and older, males had significantly lower 28 29 serum concentrations of EPA (p=0.01), DHA (p<0.001) and the sum of LC omega-3s 30 31 (p<0.001) compared to females. When further examined by life stage, men consistently

32 http://bmjopen.bmj.com/ 33 34 had lower concentrations beginning with adolescent males having significantly lower DHA 35 36 concentrations than females (p=0.03). Additionally, males age 20-55 and males >55 had 37 38 significantly lower DHA and sum OF LC omega-3 fatty acids than females 20-55 (p=0.001 39

40 on September 24, 2021 by guest. Protected copyright. 41 and p=0.01) and females >55 (p=0.03 and p=0.04). Race/ethnicity differences between 42 43 omega-3s were similar among children and adults: relative to non-Hispanic White 44 45 46 individuals, EPA was significantly lower in Mexican American/Hispanic individuals, DHA 47 48 was significantly lower in non-Hispanic Black individuals, and EPA, DHA and the sum of LC 49 50 omega-3s were all significantly higher among Asian individuals. Mean concentrations of 51 52 53 EPA, DHA and LC omega-3s are also shown in Supplementary Table 3 by Institute of 54 55 Medicine (IOM) age categories to facilitate comparison with dietary recommendations. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Reference concentrations of all 27 fatty acids additionally measured in the NHANES 2011- 4 5 6 2012 cycle are shown in Supplementary Table 4. 7 8 Mean (SE) serum concentrations of EPA, DHA and the sum of LC omega-3s by 9 10 pregnancy/breastfeeding status are shown in Table 3. There was no statistical difference in 11 12 13 the concentrations of EPA among pregnant/breastfeeding women compared to among 14 15 women of childbearing age (p=0.17). DHA and the sum of LC omega-3s among 16 For peer review only 17 pregnant/breastfeeding women which were not statistically different than those in women 18 19 20 of childbearing age. LC omega-3 supplement use was reported by two (9.52%) 21 22 pregnant/breastfeeding women compared to 21 (6.03%) women of childbearing age. 23 24 The prevalence of individuals with concentrations below those approximately 25 26 27 equivalent to the DGA recommendations is shown in Table 4 by gender and life stage. 28 29 Overall, 95.5% of children and 68.3% of adults age 20 years and older had LC omega-3 30 31 concentrations below 2.49%, corresponding to intake of LC omega-3 recommended by the

32 http://bmjopen.bmj.com/ 33 34 DGA. The prevalence of those below the DGA recommendations ranged from 97.5% of 35 36 females in early childhood to 52.3% of female seniors. Early childhood (age 3-5 years), 37 38 middle childhood (age 6-11 years and adolescents (age 12-19 years), had the largest 39

40 on September 24, 2021 by guest. Protected copyright. 41 prevalence of low LC omega-3 concentrations, with 97.4%, 94.8% and 95.5% of the 42 43 population in these age groups below LC omega-3 concentrations (<2.49%) recommended 44 45 46 by the DGA, respectively. Figure 1 depicts the prevalence of participants with serum DHA 47 48 and Omega-3 Index in previously established risk categories (31,33). The vast majority of 49 50 adults had DHA concentrations below that associated with lower risk of atrial fibrillation 51 52 53 (2.85%): 97% of all adults, 98% of males and 96.2% of females. Among all adults, 88.7% of 54 55 the population had an Omega-3 Index in the high cardiovascular risk category, 10.4% of the 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 adults were in the intermediate risk category and just 1.0% of the adults were in the low 4 5 6 risk category. The prevalence of those with an Omega-3 Index in the low cardiovascular 7 8 risk category ranged from a low of 0.34% of adult males to a high of 3.08% among senior 9 10 females (Supplementary Table 5). 11 12 13 14 15 Discussion 16 For peer review only 17 This study is one of the few to provide information on reference concentrations of 18 19 20 serum LC omega-3 (a biomarker of omega-3 nutritional status) in a large, nationally 21 22 representative sample of Americans from early childhood into later life. This is critical for 23 24 providing benchmarking data on omega-3 status and tracking progress (or lack thereof) towards 25 26 27 guidance from expert committees to increase intake and for the general public to consume 8 28 29 ounces per week of a variety of seafood, providing approximately 250 mg of EPA and DHA 30 31 per day (1,2). The mean EPA and DHA serum concentrations of 0.28% and 1.07% in children

32 http://bmjopen.bmj.com/ 33 34 (3-19 years), and 0.61% and 1.38% in adults (20 years+), reflects the persistently low 35 36 consumption of seafood and omega-3s. Using estimates of LC omega-3 concentrations 37 38 approximately equivalent to the DGA dietary intake recommendations for EPA and DHA, 39

40 on September 24, 2021 by guest. Protected copyright. 41 nearly all children in the study, and over 68% of adults had serum LC omega-3 42 43 concentrations below those associated with US Dietary recommendations for LC omega-3 44 45 consumption. When further examined by life stage and demographics, several potentially 46 47 48 vulnerable populations emerged that may warrant targeted promotion of EPA and DHA 49 50 consumption; specifically early childhood (ages 3-5 years), adult males, Mexican 51 52 American/Hispanic individuals, and non-Hispanic Black individuals tended to have 53 54 55 particularly low serum concentrations of LC omega-3 fatty acids. Pregnant and/or 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 breastfeeding women also appear to be a population of concern as LC omega-3 fatty acid 4 5 6 concentrations were similar to women of comparable age despite recommendations for 7 8 higher EPA and DHA consumption to support infant development (2). 9 10 Adults, particularly seniors, had the highest concentrations of serum EPA, DHA and 11 12 13 sum of LC omega-3s, although only around one-quarter of adults, and approximately half of 14 15 seniors had concentrations corresponding to DGA consumption recommendations. 16 For peer review only 17 However, these concentrations are not adequate for cardioprotection; less than 2% of 18 19 20 seniors had an Omega-3 Index in the low cardiovascular disease risk category. The notably 21 22 higher serum concentrations of LC omega-3s in seniors >55 relative to adults 20-55 years 23 24 mirrors findings from previous studies of LC omega-3 dietary intake (15,17) that report 25 26 27 higher intake and also greater use of omega-3 supplementation among seniors. Currently, 28 29 there is no established Adequate Intake (AI) for EPA or DHA; the AI for alpha-linolenic acid 30 31 for males and females is the same for those age 14 and older (36). However, epidemiologic

32 http://bmjopen.bmj.com/ 33 34 and physiologic evidence suggests that older adults have an increased LC omega-3 35 36 requirement due to the aging body, including brain atrophy and cognitive aging (37,38), 37 38 mobility, bone health, and sarcopenia (39–41). Despite LC omega-3 concentrations being 39

40 on September 24, 2021 by guest. Protected copyright. 41 highest in seniors, the implications with respect to adequacy to support the normal 42 43 physiological processes underlying healthy aging are unclear. 44 45 46 Our findings suggest similar DHA and sum of LC omega-3 fatty acids between 47 48 women who were pregnant and/or breastfeeding and women of reproductive age. Both 49 50 groups have potential implications for maternal health as well as for fetal and infant 51 52 53 development. Omega-3s, specifically DHA, are important for fetal and infant neuron and 54 55 retinal membrane development in late pregnancy and early life, essentially the first 1000 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 days of life (9,42,43). Omega-3 intake requirements in pregnancy have not been 4 5 6 established, however recommendations from a number of international organizations have 7 8 been put forth (14) and public health messaging on the importance of LC omega-3s intake 9 10 in maternal health has been promoted in recent years (13). DHA demands during 11 12 13 pregnancy are higher due to diversion to the fetus, especially in the last trimester as the 14 15 nervous system is rapidly developing, as well as expanded maternal cell mass and general 16 For peer review only 17 requirements of the placenta (44). Pregnant women can become low in DHA if dietary 18 19 20 intakes do not compensate for increased demands. Public health messaging has 21 22 encouraged women to increase LC omega-3 intake, especially DHA through a variety of 23 24 sources (low-mercury fish, and DHA found in supplements and fortified foods) (10). 25 26 27 However, the prevalence of LC omega-3 supplementation in pregnant/breastfeeding 28 29 women was just 9.5%. In addition, it was found that serum EPA tended to be lower among 30 31 pregnant/breastfeeding women, although not statistically significant, possibly reflecting

32 http://bmjopen.bmj.com/ 33 34 the small sample size. This may have relevance for maternal health as epidemiologic 35 36 evidence suggests that LC omega-3s play a role in prevention and/or treatment of 37 38 neuropsychiatric diseases (45). In recent years, a number of systematic reviews and meta- 39

40 on September 24, 2021 by guest. Protected copyright. 41 analyses have concluded that LC omega-3 supplementation with a higher EPA to DHA ratio, 42 43 with at least 1 gram of EPA, in conjunction with anti-depressive therapy, may be more 44 45 46 beneficial than either monotherapy alone (46–48). Associations with postpartum 47 48 depression specifically are unclear, but warrant further investigation considering the 49 50 prevalence of postpartum depression (49). The estimates for pregnant and breastfeeding 51 52 53 women, met analytic and reporting criteria for NHANES (22), but nonetheless, the sample 54 55 size was small and caution is warranted for interpretation of the findings. The findings are 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 however consistent with a study of EPA and DHA intake in NHANES which have pooled the 4 5 6 samples across several cycles and found EPA and DHA intake in pregnant women did not 7 8 differ from non-pregnant women in terms of being generally inadequate (17). 9 10 A strength of this study is the measurement of LC omega-3s in serum rather than 11 12 13 estimates from dietary intake which minimizes error from self-reporting. However, serum 14 15 fatty acids represent short term dietary consumption (27,28) and the results may therefore 16 For peer review only 17 not reflect longer term status. The large population allowed us to examine reference 18 19 20 concentrations for multiple life stages and demographic categories which was weighted to 21 22 be nationally representative of the US, although sample sizes for some subgroups e.g. 23 24 pregnant and breastfeeding women were limited as described above. The present study 25 26 27 provides the most contemporaneous data in a US nationally representative population, but 28 29 8 to 9 years has elapsed between biospecimen collection and the present study since the 30 31 NHANES serum fatty acid data only became publicly available in 2020. It is therefore

32 http://bmjopen.bmj.com/ 33 34 possible that concentrations reported herein differ from those currently in the US. Data on 35 36 dietary intake of EPA and DHA in pregnant women and women of childbearing age in 37 38 NHANES 2001-2014 suggest slight increases in intake across 2-year survey periods (14). 39

40 on September 24, 2021 by guest. Protected copyright. 41 Unfortunately, it is not possible to assess trends in circulating LC omega-3 fatty acids as 42 43 NHANES infrequently collects circulating fatty acid data. Fatty acids were last measured in 44 45 46 plasma collected in the 2003-2004 NHANES cycle which captured a different number of 47 48 fatty acids (24 versus 30 in 2011-2012), and was limited to individuals age 20 and older 49 50 (29). The absence of standardized guidelines or definitions of adequate or optimal serum 51 52 53 LC omega-3 concentrations to support overall health also precluded our ability to provide 54 55 inferences from these results. Further limitations include that Omega-3 Index risk 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 categories were determined from an equation relating concentrations in RBCs to plasma 4 5 6 which may introduce bias in our estimates and the cross-sectional nature of NHANES limits 7 8 the ability to study possible causal relationships and their resultant physiological function. 9 10 Lastly, serum concentrations of LC omega-3 fatty acids predominately reflect exogenous 11 12 13 dietary intake due to very limited conversion of short chain omega-3 fatty acids to LC 14 15 omega-3 fatty acids (50). However, polymorphisms that affect desaturase enzymes that 16 For peer review only 17 convert alpha-linolenic acid to LC omega-3s may also influence circulating LC omega-3 18 19 20 concentrations. The presence (or absence) of such polymorphisms may confound 21 22 differences in LC omega-3 between demographic groups. 23 24 Conclusions 25 26 27 In conclusion, this study provides reference ranges on serum LC omega-3 fatty acids, a 28 29 biomarker of status, across all life stages in a nationally representative population, information 30 31 that until now, was critically lacking. The findings demonstrate overall low serum concentrations

32 http://bmjopen.bmj.com/ 33 34 of LC omega-3 fatty acids including both EPA and DHA and the sum of LC omega-3 fatty acids 35 36 across all life stages. This data is supported by findings from previous dietary intake analyses 37 38 (14–16). Owing to the role of LC omega-3s in brain development, there is particular need for 39

40 on September 24, 2021 by guest. Protected copyright. 41 public health strategies aimed at increasing LC omega-3 intake in children and women of 42 43 childbearing age including those who are pregnant, breastfeeding or looking to become pregnant. 44 45 Based on life stages, gender and demographic factors, we demonstrate that children, males, 46 47 48 Mexican American/Hispanic and non-Hispanic Black individuals have lower EPA and/or 49 50 DHA serum concentrations than adults, females, and non-Hispanic White individuals, 51 52 respectively. This further demonstrates the need for public health awareness and action. 53 54 55 LC omega-3 fatty acids play a critically important role in many aspects of human health across 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 the lifespan. Healthcare practitioners need to ensure their patients are consuming enough LC 4 5 6 omega-3 fatty acids in their daily diet, and if they are not, they need to consider the role of 7 8 supplementation for their patients to ensure nutrient gaps are being met. Furthermore, there is 9 10 increased need for guidance and education for healthcare practitioners to ensure they are 11 12 prepared to make recommendations needed that include specific ways to increase EPA and DHA 13 14 15 intake for the entire US population, with particular attention to vulnerable populations. 16 For peer review only 17 18 19 Figure 1 legend: 20 21 22 Comparison to long chain omega-3 concentrations from prior studies were used to approximate 23 24 1) the prevalence of participants with omega-3 concentrations equivalent to omega-3 dietary 25 26 intake recommendations by the DGA (30), 2) the lower range of the tertiles of serum DHA 27 28 29 associated with lower risk of atrial fibrillation (31), and 3) cardiovascular risk categories 30 31 defined by the Omega-3 Index (33,34).

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

32 http://bmjopen.bmj.com/ 33 4 2. Dietary Guidelines Advisory Committee. Scientific Report of the 2020 Dietary Guidelines 34 5 Advisory Committee, Advisory Report to the Secretary of Agriculture and the Secretary of 35 6 Health and Human Services. Washington, DC: US Department of Agriculture, Agricultural 36 7 Research Service; 2020. Cited 2020 Nov 10. Available from: 37 38 8 https://www.dietaryguidelines.gov/sites/default/files/2020- 39 9 07/ScientificReport_of_the_2020DietaryGuidelinesAdvisoryCommittee_first-print.pdf

40 on September 24, 2021 by guest. Protected copyright. 41 10 3. Bowen KJ, Harris WS, Kris-Etherton PM. Omega-3 Fatty Acids and Cardiovascular 42 11 Disease: Are There Benefits? Curr Treat Options Cardiovasc Med. 2016 Nov;18(11):69. 43 44 12 4. Djoussé L, Gaziano JM, Buring JE, Lee I-M. Dietary omega-3 fatty acids and fish 45 13 consumption and risk of type 2 diabetes. Am J Clin Nutr. 2011 Jan;93(1):143–50. 46 47 48 14 5. Kaushik M, Mozaffarian D, Spiegelman D, Manson JE, Willett WC, Hu FB. Long-chain 49 15 omega-3 fatty acids, fish intake, and the risk of type 2 diabetes mellitus. Am J Clin Nutr. 50 16 2009 Sep;90(3):613–20. 51 52 17 6. Wang DD, Li Y, Chiuve SE, Stampfer MJ, Manson JE, Rimm EB, et al. Association of 53 18 Specific Dietary Fats With Total and Cause-Specific Mortality. JAMA Intern Med. 2016 54 19 01;176(8):1134–45. 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 20 7. Liu JC, Conklin SM, Manuck SB, Yao JK, Muldoon MF. Long-Chain Omega-3 Fatty 4 21 Acids and Blood Pressure. Am J Hypertens. 2011 Oct 1;24(10):1121–6. 5 6 7 22 8. Ueshima H, Stamler J, Elliott P, Chan Q, Brown IJ, Carnethon MR, et al. Food Omega-3 8 23 Fatty Acid Intake of Individuals (Total, Linolenic Acid, Long-Chain) and Their Blood 9 24 Pressure: INTERMAP Study. Hypertension. 2007 Aug;50(2):313–9. 10 11 25 9. Jensen CL. Effects of n−3 fatty acids during pregnancy and lactation. Am J Clin Nutr. 2006 12 26 Jun 1;83(6):1452S-1457S. 13 14 27 10. Greenberg JA, Bell SJ, Ausdal WV. Omega-3 Fatty Acid supplementation during 15 28 pregnancy. Rev Obstet Gynecol. 2008;1(4):162–9. 16 For peer review only 17 18 29 11. Gelenberg AJ, Freeman MP , Markowitz JC, Rosenbaum JF, Thase ME, Trivedi MH, et al. 19 30 Practice Guideline for the Treatment of Patients With Major Depressive Disorder, Third 20 31 Edition. page 51 and page 92. 2010. Available from: 21 32 http://psychiatryonline.org/pb/assets/raw/sitewi de/practice_guidelines/guidelines/mdd.pdf. 22 23 33 12. Rimm EB, Appel LJ, Chiuve SE, Djoussé L, Engler MB, Kris-Etherton PM, et al. Seafood 24 34 Long-Chain n-3 Polyunsaturated Fatty Acids and Cardiovascular Disease: A Science 25 26 35 Advisory From the American Heart Association. Circulation. 2018 03;138(1):e35–47. 27 28 36 13. Joint FAO/WHO Expert Consultation on Fats and fatty acids in human nutrition: report of 29 37 an expert consultation. Rome, Italy: Food and Agriculture Organization, 2010. 30 31 38 14. Zhang Z, Fulgoni VL, Kris-Etherton PM, Mitmesser SH. Dietary Intakes of EPA and DHA

32 39 Omega-3 Fatty Acids among US Childbearing-Age and Pregnant Women: An Analysis of http://bmjopen.bmj.com/ 33 40 NHANES 2001-2014. Nutrients. 2018 28;10(4). 34 35 41 15. Richter CK, Bowen KJ, Mozaffarian D, Kris-Etherton PM, Skulas-Ray AC. Total Long- 36 37 42 Chain n-3 Fatty Acid Intake and Food Sources in the United States Compared to 38 43 Recommended Intakes: NHANES 2003–2008. Lipids. 2017 Nov;52(11):917–27. 39

40 44 16. Papanikolaou Y, Brooks J, Reider C, Fulgoni VL. U.S. adults are not meeting on September 24, 2021 by guest. Protected copyright. 41 45 recommended levels for fish and omega-3 fatty acid intake: results of an analysis using 42 46 observational data from NHANES 2003–2008. Nutr J. 2014 Dec;13(1):31. 43 44 47 17. Thompson M, Hein N, Hanson C, Smith LM, Anderson-Berry A, Richter CK, et al. Omega- 45 48 3 Fatty Acid Intake by Age, Gender, and Pregnancy Status in the United States: National 46 − 47 49 Health and Nutrition Examination Survey 2003 2014. Nutrients. 2019 Jan 15;11(1). 48 49 50 18. Villegas R, Takata Y, Murff H, Blot WJ. Fish, omega-3 long-chain fatty acids, and all- 50 51 cause mortality in a low-income US population: Results from the Southern Community 51 52 Cohort Study. Nutr Metab Cardiovasc Dis. 2015 Jul;25(7):651–8. 52 53 53 19. Berry JD, Prineas RJ, van Horn L, Passman R, Larson J, Goldberger J, et al. Dietary Fish 54 54 Intake and Incident Atrial Fibrillation (from the Women’s Health Initiative). Am J Cardiol. 55 56 55 2010 Mar;105(6):844–8. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 56 20. Thompson FE, Kirkpatrick SI, Subar AF, Reedy J, Schap TE, Wilson MM, et al. The 4 57 National Cancer Institute’s Dietary Assessment Primer: A Resource for Diet Research. J 5 6 58 Acad Nutr Diet. 2015 Dec;115(12):1986–95. 7 8 59 21. Archer E. The NHANES dietary data are physiologically implausible and inadmissible as 9 60 scientific evidence. Am J Clin Nutr. 2017;106(3):951–2. 10 11 61 22. Johnson CL, Paulose-Ram R, Ogden CL, Carroll MD, Kruszon-Moran D, Dohrmann SM, 12 62 et al. National health and nutrition examination survey: analytic guidelines, 1999-2010. 13 63 Vital Health Stat 2. 2013 Sep;(161):1–24. 14 15 64 23. Lagerstedt SA, Hinrichs DR, Batt SM, Magera MJ, Rinaldo P, McConnell JP. Quantitative 16 For peer review only 17 65 Determination of Plasma C8–C26 Total Fatty Acids for the Biochemical Diagnosis of 18 66 Nutritional and Metabolic Disorders. Mol Genet Metab. 2001 May;73(1):38–45. 19 20 67 24. Sun Y, Koh HWL, Choi H, Koh W-P, Yuan J-M, Newman JW, et al. Plasma fatty acids, 21 68 oxylipins, and risk of myocardial infarction: the Singapore Chinese Health Study. J Lipid 22 69 Res. 2016;57(7):1300–7. 23 24 70 25. Wu JHY, Marklund M, Imamura F, Tintle N, Ardisson Korat AV, de Goede J, et al. 25 26 71 Omega-6 fatty acid biomarkers and incident type 2 diabetes: pooled analysis of individual- 27 72 level data for 39 740 adults from 20 prospective cohort studies. Lancet Diabetes 28 73 Endocrinol. 2017;5(12):965–74. 29 30 74 26. Imamura F, Fretts A, Marklund M, Ardisson Korat AV, Yang W-S, Lankinen M, et al. 31 75 Fatty acid biomarkers of dairy fat consumption and incidence of type 2 diabetes: A pooled

32 76 analysis of prospective cohort studies. PLoS Med. 2018;15(10):e1002670. http://bmjopen.bmj.com/ 33 34 77 27. Holman RT. Control of polyunsaturated acids in tissue lipids. J Am Coll Nutr. 35 36 78 1986;5(2):183–211. 37 38 79 28. Kwon JS, Snook JT, Wardlaw GM, Hwang DH. Effects of diets high in saturated fatty 39 80 acids, canola oil, or safflower oil on platelet function, thromboxane B2 formation, and fatty

40 81 acid composition of platelet phospholipids. Am J Clin Nutr. 1991 Aug;54(2):351–8. on September 24, 2021 by guest. Protected copyright. 41 42 82 29. Murphy RA, Yu EA, Ciappio ED, Mehta S, McBurney MI. Suboptimal Plasma Long Chain 43 83 n-3 Concentrations are Common among Adults in the United States, NHANES 2003-2004. 44 84 Nutrients. 2015 Dec 9;7(12):10282–9. 45 46 47 85 30. Sun Q, Ma J, Campos H, Hankinson SE, Hu FB. Comparison between plasma and 48 86 erythrocyte fatty acid content as biomarkers of fatty acid intake in US women. Am J Clin 49 87 Nutr. 2007 Jul;86(1):74–81. 50 51 88 31. Virtanen JK, Mursu J, Voutilainen S, Tuomainen T-P. Serum Long-Chain n-3 52 89 Polyunsaturated Fatty Acids and Risk of Hospital Diagnosis of Atrial Fibrillation in Men. 53 90 Circulation. 2009 Dec 8;120(23):2315–21. 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 91 32. von Schacky C. Omega-3 Index and Cardiovascular Health. Nutrients. 2014 Feb 4 92 21;6(2):799–814. 5 6 7 93 33. Harris WS. The omega-3 index as a risk factor for coronary heart disease. Am J Clin Nutr. 8 94 2008 Jun 1;87(6):1997S-2002S. 9 10 95 34. Chaudhary R, Saadin K, Bliden KP, Harris WS, Dinh B, Sharma T, et al. Risk factors 11 96 associated with plasma omega-3 fatty acid levels in patients with suspected coronary artery 12 97 disease. Prostaglandins Leukot Essent Fatty Acids. 2016 Oct;113:40–5. 13 14 98 35. Armstrong JM, Metherel AH, Stark KD. Direct Microwave Transesterification of Fingertip 15 99 Prick Blood Samples for Fatty Acid Determinations. Lipids. 2008 Feb;43(2):187–96. 16 For peer review only 17 18 100 36. Panel on Macronutrients, Panel on the Definition of Dietary Fiber, Subcommittee on Upper 19 101 Reference Levels of Nutrients, Subcommittee on Interpretation and Uses of Dietary 20 102 Reference Intakes, Standing Committee on the Scientific Evaluation of Dietary Reference 21 103 Intakes, Food and Nutrition Board, et al. Dietary Reference Intakes for Energy, 22 104 Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. Washington, 23 105 D.C.: National Academies Press; 2005. Cited 2020 Apr 20. Available from: 24 25 106 https://www.nap.edu/catalog/10490 26 27 107 37. Issa AM, Mojica WA, Morton SC, Traina S, Newberry SJ, Hilton LG, et al. The Efficacy of 28 108 Omega–3 Fatty Acids on Cognitive Function in Aging and Dementia: A Systematic 29 109 Review. Dement Geriatr Cogn Disord. 2006;21(2):88–96. 30 31 110 38. Denis I, Potier B, Heberden C, Vancassel S. Omega-3 polyunsaturated fatty acids and brain

32 111 aging: Curr Opin Clin Nutr Metab Care. 2015 Mar;18(2):139–46. http://bmjopen.bmj.com/ 33 34 112 39. Reinders I, Murphy RA, Song X, Visser M, Cotch MF, Lang TF, et al. Polyunsaturated 35 36 113 fatty acids in relation to incident mobility disability and decline in gait speed; the Age, 37 114 Gene/Environment Susceptibility-Reykjavik Study. Eur J Clin Nutr. 2015 Apr;69(4):489– 38 115 93. 39

40 116 40. Harris TB, Song X, Reinders I, Lang TF, Garcia ME, Siggeirsdottir K, et al. Plasma on September 24, 2021 by guest. Protected copyright. 41 117 phospholipid fatty acids and fish-oil consumption in relation to osteoporotic fracture risk in 42 118 older adults: the Age, Gene/Environment Susceptibility Study. Am J Clin Nutr. 2015 43 119 May;101(5):947–55. 44 45 46 120 41. Dupont J, Dedeyne L, Dalle S, Koppo K, Gielen E. The role of omega-3 in the prevention 47 121 and treatment of sarcopenia. Aging Clin Exp Res. 2019 Jun;31(6):825–36. 48 49 122 42. Middleton P, Gomersall JC, Gould JF, Shepherd E, Olsen SF, Makrides M. Omega-3 fatty 50 123 acid addition during pregnancy. Cochrane Database Syst Rev. 2018 15;11:CD003402. 51 52 124 43. Carlson SE, Gajewski BJ, Valentine CJ, Rogers LK, Weiner CP, DeFranco EA, et al. 53 125 Assessment of DHA on reducing early preterm birth: the ADORE randomized controlled 54 55 126 trial protocol. BMC Pregnancy Childbirth. 2017 13;17(1):62. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 127 44. Rees A-M, Austin M-P, Owen C, Parker G. Omega-3 deficiency associated with perinatal 4 128 depression: case control study. Psychiatry Res. 2009 Apr 30;166(2–3):254–9. 5 6 7 129 45. Larrieu T, Layé S. Food for Mood: Relevance of Nutritional Omega-3 Fatty Acids for 8 130 Depression and Anxiety. Front Physiol. 2018;9:1047. 9 10 131 46. Liao Y, Xie B, Zhang H, He Q, Guo L, Subramaniapillai M, et al. Efficacy of omega-3 11 132 PUFAs in depression: A meta-analysis. Transl Psychiatry. 2019 05;9(1):190. 12 13 133 47. Hallahan B, Ryan T, Hibbeln JR, Murray IT, Glynn S, Ramsden CE, et al. Efficacy of 14 134 omega-3 highly unsaturated fatty acids in the treatment of depression. Br J Psychiatry J 15 135 Ment Sci. 2016;209(3):192–201. 16 For peer review only 17 18 136 48. Wojcicki JM, Heyman MB. Maternal omega-3 fatty acid supplementation and risk for 19 137 perinatal maternal depression. J Matern Fetal Neonatal Med. 2011 May;24(5):680–6. 20 21 138 49. Browne JC, Scott KM, Silvers KM. Fish consumption in pregnancy and omega-3 status 22 139 after birth are not associated with postnatal depression. J Affect Disord. 2006 Feb;90(2– 23 140 3):131–9. 24 25 141 50. Plourde M, Cunnane SC. Extremely limited synthesis of long chain polyunsaturates in 26 27 142 adults: implications for their dietary essentiality and use as supplements. Appl Physiol Nutr 28 143 Metab Physiol Appl Nutr Metab. 2007 Aug;32(4):619–34. 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

32 http://bmjopen.bmj.com/ 33 and advised on interpretation of the data. RAM compiled the dataset and performed the statistical 34 35 analysis. RAM drafted the manuscript. KM, PPD, SE, and SHM review the manuscript, provided 36 37 38 amendment and approved the final version. RAM had full access to study data and takes 39

40 responsibility for its accuracy and the integrity of the analysis. on September 24, 2021 by guest. Protected copyright. 41 42 Funding: N/A 43 44 45 Competing interests: RAM carried out the work presented in the manuscript as a consultant for 46 47 Pharmavite. PPD, SE, KM and SHM are employees of Pharmavite, LLC, manufacturers and 48 49 suppliers of omega-3 nutritional lipids. 50 51 52 Patient and public involvement: Patients and/or the public were not involved in the design, or 53 54 conduct, or reporting, or dissemination plans of this research. 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Patient consent for publication: Not required 4 5 6 Ethics approval: The NCHS Ethics Review Board protects the rights and welfare of NHANES 7 8 participants. The NHANES protocol complies with the U.S. Department of Health and Human 9 10 Services’ Policy for Protection of Human Research Subjects. NCHS IRB/ERC Protocol number: 11 12 2011-17. This study was exempt from Institutional Review Board at the University of British 13 14 15 Columbia related to the use of publicly available data for research and publication. 16 For peer review only 17 Provenance and peer review: Not commissioned; internally peer reviewed by Pharmavite 18 19 Data availability statement: Data are available in a public, open access repository. The dataset 20 21 22 used for this study was generated from data publicly released by the National Health and 23 24 Nutrition Examination Survey (NHANES). 25 26 Word count: 3,821 27 28 29 30 144 31

32 145 http://bmjopen.bmj.com/ 33 34 146 35 36 147 37 38 148 39

40 149 on September 24, 2021 by guest. Protected copyright. 41 42 43 150 44 45 151 46 47 152 48 49 153 50 51 154 52 53 54 155 55 56 156 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 157 4 5 158 6 7 8 159 9 10 160 11 12 161 13 14 162 15 16 163 For peer review only 17 18 19 164 20 21 165 22 23 166 24 25 167 Table 1. Characteristics of study participants in the NHANES 2011-2012 fatty acid 26 168 sample 27 28 Characteristics N (%) 29 30 Gender Male 1098 31 (47.6)

32 Female 1163 http://bmjopen.bmj.com/ 33 (52.5) 34 Age group Children, 3-19yrs 945 (41.8) 35 1316 36 Adults, 20yrs+ 37 (58.2) 38 Early childhood, 3- 179 (2.40) 39 5yrs

40 Middle childhood, 6- 519 (6.52) on September 24, 2021 by guest. Protected copyright. 41 11yrs 42 Adolescents, 12- 247 (10.3) 43 44 19yrs 45 Adults, 20-55yrs 818 (53.9) 46 Seniors, >55yrs 498 (26.9) 47 Race/ethnicity Mexican 605 (16.5) 48 American/Hispanic 49 Non-Hispanic White 729 (65.2) 50 51 Non-Hispanic Black 592 (12.1) 52 Asian 265 (4.55) 53 Other/unknown 70 (1.66) 54 Education-adults High school 279 (19.7) 4 5 Some college or 734 (63.7) 6 above 7 Current smoking Yes 265 (20.2) 8 No 1,049 9 (79.8) 10 Poverty Income Ratio Low 0-1.85 20 (9.90) 11 12 Medium >1.85-3.50 423 (32.7) 13 High >3.50 712 (66.3) 14 BMI-adults Mean, kg/m2 28.7 (0.34) 15 LC omega-3 supplement use, N Yes 166 (7.30) 16 (%) For peer review only 17 169 Sample sizes are unweighted however, estimates for means, proportions and standard 18 19 170 errors are weighted. Poverty to income ratio is defined as total family income relative to 20 171 poverty guidelines. Long chain (LC) omega-3 supplement use defined from self-reported 21 172 supplement use (fish oil, cod liver oil, salmon oil, krill oil and DHA) in prior 30 days. 22 23 173 24 174 25 175 26 27 176 28 177 29 178 30 179 31 180 32 181 http://bmjopen.bmj.com/ 33 182 34 35 36 183 Table 2. Mean serum long chain omega-3 fatty acid concentrations (% of total fatty acids) 37 184 by age group in NHANES 2011-2012 38 39 Age group Variable N Mean (SE) 95%CI P a

40 Children EPA 945 0.28 (0.01) 0.25-0.31 <0.001 on September 24, 2021 by guest. Protected copyright. 41 3-19yrs DHA 945 1.07 (0.02) 1.03-1.10 <0.001 42 Sum LC omega-3 945 1.75 (0.03) 1.68-1.83 <0.001 43 Adults EPA 1,316 0.61 (0.02) 0.56-0.66 Ref 44 20yrs+ DHA 1,316 1.38 (0.05) 1.27-1.49 Ref 45 46 Sum LC omega-3 1,316 2.43 (0.08) 2.28-2.59 Ref 47 Early childhoodb EPA 179 0.26 (0.01) 0.24-0.27 0.06 48 3-5yrs DHA 179 1.01 (0.03) 0.96-1.07 0.07 49 Sum LC omega-3 179 1.65 (0.04) 1.58-1.73 0.03 50 Middle 0.99 51 b EPA 519 0.29 (0.02) 0.24-0.34 52 childhood 53 6-11yrs DHA 519 1.09 (0.03) 1.02-1.16 0.76 54 Sum LC omega-3 519 1.78 (0.06) 1.65-1.90 0.78 55 Adolescents EPA 247 0.29 (0.02) 0.25-0.32 Ref 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 12-19yrs DHA 247 1.08 (0.03) 1.02-1.13 Ref 4 5 Sum LC omega-3 247 1.76 (0.04) 1.78-1.84 Ref c 6 Adults EPA 818 0.51 (0.02) 0.47-0.55 Ref 7 20-55yrs DHA 818 1.29 (0.06) 1.18-1.41 Ref 8 Sum LC omega-3 818 2.23 (0.08) 2.08-2.39 Ref 9 Seniors EPA 498 0.79 (0.06) 0.67-0.92 <0.001 10 >55yrs DHA 498 1.56 (0.06) 1.44-1.68 <0.001 11 12 Sum LC omega-3 498 2.83 (1.00) 2.62-3.04 <0.001 13 185 Comparisons are a) children 3-19 years versus adults 20+years, b) early childhood and middle 14 186 childhood versus adolescents, and c) adults versus seniors. Sum LC omega-3 represents 15 187 EPA+DPA+DHA 16 188 For peer review only 17 189 18 19 190 20 191 21 192 22 193 23 194 24 195 25 196 26 27 197 28 198 29 199 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 Table 3. Mean serum long chain omega-3 fatty acid concentrations (% of total fatty acids) 39 in women who were pregnant/breastfeeding (N=21) and women of childbearing age 40 (N=319) on September 24, 2021 by guest. Protected copyright. 41 42 95% CI for Women of 95% CI for P Pregnant/ 43 Mean childbearing Mean breastfeeding 44 age 45 46 Fatty acid Mean (SE) Mean (SE) 47 EPA 0.34 (0.07) 0.19-0.49 0.47 (0.03) 0.39-0.53 0.17 48 DHA 1.52 (0.08) 1.35-1.68 1.37 (0.07) 1.21-1.52 0.10 49 Sum LC omega-3 2.18 (0.15) 1.87-2.49 2.24 (0.10) 2.02-2.46 0.73 50 Sample sizes are unweighted, however, estimates for means and SE are weighted. Women 51 52 of childbearing age are those, age 15-44 years. 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

40 on September 24, 2021 by guest. Protected copyright. 41 42 43 200 Table 4. Prevalence, N (%) of EPA+DHA fatty acid concentrations <2.49%, which is 44 201 approximately equivalent to DGA dietary recommendations for EPA and DHA intake by 45 202 gender and age group in NHANES 2011-2012. 46 47 Children, Adults, 20 Early Middle Adolescent Adults Seniors 48 3-19 yrs yrs+ childhoo childhoo s 49 d d 50 51 <2.49% EPA + DHA 52 All 880 856 168 488 224 (95.5) 593 263 53 (95.5) (68.3) (97.4) (94.8) (76.0) (52.9) 54 Male 429 445 85 (97.3) 239 105 (97.6) 313 132 55 (96.6) (72.4) (94.7) (80.7) (53.7) 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Femal 451 411 83 (97.5) 249 119 (93.6) 280 131 4 5 e (94.5) (64.6) (94.9) (71.4) (52.3) 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

Page 31 of 38 BMJ Open A 100

1 90 2 80 3 4 70 5 97.1 98.0 96.2 6 60 7 50 8 9 40 Percent 10 For peer review only 30

BMJ Open Page 32 of 38

1 2 3 4 Supplementary Table 1. Mean long chain omega-3 fatty acids (% of total fatty acids) by gender and age group in NHANES 2011- 5 2012 6 7 Males Females P 8 Age group Variable N Mean (SE) 95%CI N Mean (SE) 95% CI 9 10 Children, 3-19yrs EPA 457 0.30 (0.02) 0.26-0.33 488 0.27 (0.01) 0.25-0.30 0.45 11 DHA 457 1.05 (0.02) 1.00-1.10 488 1.10 (0.03) 1.04-1.16 0.18 12 sum LC omegaFor-3 457peer 1.75 (0.04) review1.67-1.84 488 1.75only (0.04) 1.66-1.84 0.98 13 Adults, 20yrs+ EPA 641 0.55 (0.03) 0.50-0.61 675 0.65 (0.03) 0.58-0.73 0.01 14 DHA 641 1.27 (0.05) 1.17-1.37 675 1.49 (0.07) 1.34-1.63 <0.001 15 sum LC omega-3 641 2.27 (0.07) 2.11-2.42 675 2.58 (0.09) 2.39-2.78 <0.001 16 Early childhood EPA 90 0.26 (0.01) 0.23-0.29 89 0.25 (0.01) 0.23-0.28 0.64

17 http://bmjopen.bmj.com/ 18 3-5yrs DHA 90 1.06 (0.04) 0.98-1.14 89 0.97 (0.04) 0.89-1.04 0.73 19 sum LC omega-3 90 1.71 (0.06) 1.59-1.83 89 1.60 (0.05) 1.50-1.70 0.15 20 Middle childhood EPA 255 0.31 (0.03) 0.25-0.36 264 0.27 (0.02) 0.23-0.31 0.02 21 6-11yrs DHA 255 1.10 (0.05) 1.00-1.21 264 1.08 (0.03) 1.01-1.15 0.63 22 sum LC omega-3 255 1.83 (0.08) 1.66-2.00 264 1.73 (0.05) 1.62-1.84 0.17 23 Adolescents EPA 112 0.30 (0.02) 0.25-0.34 135 0.28 (0.02) 0.25-0.31 0.40 24 12-19yrs DHA 112 1.01 (0.03) 0.95-1.07 135 1.14 (0.05) 1.04-1.23 0.03

25 on September 24, 2021 by guest. Protected copyright. 26 sum LC omega-3 112 1.71 (0.05) 1.62-1.81 135 1.80 (0.06) 1.67-1.93 0.25 27 Adults EPA 404 0.49 (0.02) 0.46-0.53 414 0.53 (0.03) 0.46-0.60 0.29 28 20-55yrs DHA 404 1.18 (0.05) 1.08-1.28 414 1.40 (0.07) 1.25-1.55 0.001 29 sum LC omega-3 404 2.11 (0.06) 1.99-2.24 414 2.35 (0.10) 2.13-2.57 0.01 30 Seniors EPA 237 0.68 (0.06) 0.56-0.80 261 0.88 (0.10) 0.67-1.09 0.11 31 >55yrs DHA 237 1.47 (0.05) 1.36-1.58 261 1.64 (0.08) 1.47-1.80 0.03 32 33 sum LC omega-3 237 2.62 (0.11) 2.39-2.85 261 3.00 (0.15) 2.68-3.32 0.04 34 p-values represent differences between males and females within a respective life stage. 35 36 37 38 39 40 41 42 43 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from

Page 33 of 38 BMJ Open

1 2 3 4 Supplementary Table 2. Mean long chain omega-3 fatty acids (% of total fatty acids) by race/ethnicity in NHANES 2011-2012. 5 6 Children, 3-19 yrs Adults, 20 yrs+ 7 Race/ethnicity Fatty acid N Mean (SE) 95%CI P N Mean (SE) 95%CI P 8 Mexican American/Hispanic EPA 312 0.23 (0.01) 0.21-0.26 0.01 293 0.44 (0.02) 0.40-0.47 <0.001 9 DHA 1.02 (0.03) 0.97-1.08 0.72 1.26 (0.04) 1.20-1.45 0.29 10 sum LC omega-3 1.62 (0.04) 1.54-1.71 0.08 2.10 (0.05) 1.99 -2.21 0.01 11 Non-Hispanic White EPA 202 0.30 (0.02) 0.26-0.34 Ref 527 0.63 (0.04) 0.55-0.70 Ref 12 ForDHA peer 1.04review (0.03) 0.98-1.10 onlyRef 1.32 (0.06) 1.20-1.45 Ref 13 14 sum LC omega-3 1.75 (0.05) 1.65-1.85 Ref 2.40 (0.09) 2.21-2.60 Ref 15 Non-Hispanic Black EPA 269 0.29 (0.02) 0.25-0.32 0.67 323 0.55 (0.03) 0.49-0.61 0.08 16 DHA 1.20 (0.03) 1.14-1.26 <0.001 1.55 (0.06) 1.43-1.68 0.01

17 sum LC omega-3 1.88 (0.05) 1.78-1.97 0.02 2.55 (0.09)http://bmjopen.bmj.com/ 2.36-2.73 0.19 18 Asian EPA 98 0.37 (0.02) 0.32-0.42 0.04 167 0.94 (0.07) 0.79-1.10 0.003 19 DHA 1.40 (0.05) 1.30-1.51 <0.001 2.13 (0.09) 1.95-2.30 <0.001 20 21 sum LC omega-3 2.18 (0.06) 2.06-2.30 <0.001 3.54 (0.15) 3.22-3.87 <0.001 22 Other/unknown EPA 41 0.27 (0.02) 0.22-0.32 0.29 29 0.60 (0.06) 0.47-0.73 0.75 23 DHA 1.03 (0.05) 0.95-1.11 0.81 1.59 (0.21) 1.15-2.04 0.22 24 sum LC omega-3 1.68 (0.06) 1.56-1.80 0.38 2.67 (0.26) 2.12-3.21 0.35

25 p-values represent differences between race/ethnicity within a respective life stage. For example, relative to on September 24, 2021 by guest. Protected copyright. non-Hispanic white 26 children, Asian children had higher EPA (P=0.04). 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 34 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 4 Supplementary Table 3. Mean long chain omega-3 fatty acids (% of total fatty acids) by 5 IOM age groups in NHANES 2011-2012. 6 7 Age group Variable N Mean (SE) 95%CI 8 1-3yrs EPA 55 0.25 (0.02) 0.20-0.31 9 10 DHA 55 0.98 (0.04) 0.88-1.07 11 sum LC omega-3 55 1.60 (0.08) 1.44-1.76 12 4-8yrs EPA 382 0.27 (0.01) 0.25-0.28 13 DHA 382 1.04 (0.02) 1.00-1.08 14 sum LC omega-3 382 1.70 (0.03) 1.65-1.76 15 9-13yrs EPA 325 0.29 (0.03) 0.23-0.34 16 For peerDHA review325 only1.13 (0.04) 1.05-1.20 17 18 sum LC omega-3 325 1.82 (0.07) 1.67-1.96 19 14-18yrs EPA 145 0.29 (0.02) 0.26-0.33 20 DHA 145 1.05 (0.03) 0.99-1.11 21 sum LC omega-3 145 1.74 (0.04) 1.66-1.83 22 19-30yrs EPA 258 0.41 (0.03) 0.35-0.48 23 DHA 258 1.27 (0.08) 1.11-1.43 24 sum LC omega-3 258 2.09 (0.10) 1.88-2.30 25 26 31-50yrs EPA 425 0.52 (0.02) 0.48-0.56 27 DHA 425 1.26 (0.05) 1.15-1.36 28 sum LC omega-3 425 2.21 (0.07) 2.06-2.36 29 51-70yrs EPA 409 0.76 (0.06) 0.63-0.89 30 DHA 409 1.49 (0.06) 1.36-1.61 31 sum LC omega-3 409 2.71 (0.11) 2.49-2.94

32 http://bmjopen.bmj.com/ 33 >70yrs EPA 156 0.80 (0.09) 0.62-0.98 34 DHA 156 1.73 (0.11) 1.51-1.96 35 sum LC omega-3 156 3.02 (0.20) 2.61-3.43 36 37 38 39

40 on September 24, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 Supplementary Table 4. Fatty acids (% of total fatty acids) measured in NHANES 2011-2012 4 5 Children, 3-19 Adults, 20+ 6 Fatty acid Mean (SE) 95% CI Mean (SE) 95% CI 7 8 Capric acid, 10:0 0.02 (0.00) 0.02-0.02 0.03 (0.00) 0.03-0.04 9 Lauric acid, 12:0 0.10 (0.01) 0.09-0.11 0.15 (0.01) 0.13-0.17 10 11 Myristic acid, 14:0 1.10 (0.02) 1.07-1.13 1.09 (0.03) 1.03-1.15 12 Pentadecanoic acid, 15:0 0.21 (0.00) 0.20-0.22 0.21 (0.00) 0.20-0.22 13 Palmitic acid, 16:0 24.1 (0.07) 23.9-24.2 23.1 (0.13) 22.8-23.3 14 15 Margaric acid, 17:0 0.27 (0.00) 0.26-0.27 0.28 (0.00) 0.27-0.28 16 Stearic acid, 18:0 For peer 5.80review (0.02) 5.76 -only5.84 6.29 (0.06) 6.16-6.42 17 18 Arachidic acid, 20:0 0.20 (0.00) 0.20-0.21 0.22 (0.00) 0.21-0.22 19 Docosanoic acid, 22:0 0.58 (0.01) 0.56-0.60 0.61 (0.01) 0.59-0.62 20 Tricosanoic acid, 23:0 0.25 (0.00) 0.25-0.26 0.26 (0.00) 0.26-0.27 21 22 Lignoceric acid, 24:0 0.50 (0.01) 0.48-0.51 0.52 (0.01) 0.50-0.53 23 Myristoleic acid, 14:1n-5 0.07 (0.00) 0.06-0.07 0.07 (0.00) 0.06-0.07 24 25 Palmitoleic acid, 16:1n-7 2.03 (0.03) 1.97-2.10 1.56 (0.03) 1.49-1.62 26 cis-Vaccenic acid, 18:1n-7 1.24 (0.01) 1.21-1.26 1.10 (0.01) 1.08-1.13 27 28 Oleic acid, 18:1n-9 18.2 (0.09) 18.0-18.4 17.3 (0.13) 17.0-17.6 29 Eicosenoic acid, 20:1n-9 0.12 (0.00) 0.12-0.12 0.12 (0.00) 0.12-0.12 30 Nervonic acid, 24:1n-9 0.74 (0.01) 0.72-0.77 0.74 (0.01) 0.72-0.75 31

40 Eicosatrienoic acid, 20:3n-9 0.07 (0.00) 0.06-0.07 0.06 (0.00) 0.05-0.06 on September 24, 2021 by guest. Protected copyright. 41 42 Arachidonic acid, 20:4n-6 7.35 (0.07) 7.20-7.50 6.87 (0.08) 6.71-7.04 43 Docosatetraenoic acid, 22:4n-6 0.22 (0.00) 0.21-0.23 0.26 (0.00) 0.25-0.27 44 45 Docosapentaenoic acid, 22:5n-6 0.17 (0.00) 0.16-0.17 0.22 (0.00) 0.21-0.23 46 Data shown are weighted means and 95% confidence intervals. Fatty acids shown represent 32 47 of the 35 fatty acids measured in NHANES. Together with EPA, DHA and DPA (shown in the 48 manuscript tables) were summed to generate total fatty acids to derive the relative concentration 49 (%) of fatty acids. 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from 1 2 3 4 Supplementary Table 5. Prevalence, N (%) of omega-3 fatty acid concentrations below those 5 identified as having cardiovascular benefits by gender and age group in NHANES 2011-2012. 6 7 8 Adults, 20 yrs+ Adults Seniors 9 DHA ≥2.85%, lower risk of atrial fibrillation 10 All 61 (2.93) 27 (2.07) 34 (4.65) 11 Male 19 (1.98) 8 (1.39) 11 (3.30) 12 13 Female 42 (3.80) 19 (2.72) 23 (5.70) 14 Omega-3 Index >6.23%, Low risk 15 All 20 (1.0) 10 (0.57) 10 (1.86) 16 Male For peer7 (0.34) review4 (0.37) only 3 (0.29) 17 Female 13 (1.60) 6 (0.78) 7 (3.08) 18 Omega-3 Index 3.11-6.23%, Intermediate risk 19 20 All 138 (10.4) 61 (6.59) 77 (17.9) 21 Male 48 (7.58) 21 (4.61) 27 (14.3) 22 Female 90 (12.9) 40 (8.52) 50 (20.7) 23 Omega-3 Index <3.11%, High risk 24 All 1158 (88.7) 747 (92.8) 411 (80.2) 25 26 Male 586 (92.1) 379 (95.0) 207 (85.4) 27 Female 572 (85.6) 368 (90.7) 204 (76.2) 28 Comparison to long chain omega-3 levels from prior studies were used to approximate 1) the 29 30 prevalence of participants with omega-3 levels equivalent to omega-3 dietary intake 31 recommendations by the DGA (24), 2) the lower range of the tertiles of serum DHA associated

32 with lower risk of atrial fibrillation (31), and 3) cardiovascular risk categories defined by the http://bmjopen.bmj.com/ 33 34 Omega-3 Index (32,33). 35 36 37 38 39

32 http://bmjopen.bmj.com/ 33 Setting 5 Describe the setting, locations, and relevant dates, including periods of 34 35 recruitment, exposure, follow-up, and data collection 36 37 Participants 6 (a) Cohort study—Give the eligibility criteria, and the sources and methods of 38 selection of participants. Describe methods of follow-up 39 Case-control study—Give the eligibility criteria, and the sources and methods of 40 on September 24, 2021 by guest. Protected copyright. 41 case ascertainment and control selection. Give the rationale for the choice of 42 cases and controls 43 44 Cross-sectional study—Give the eligibility criteria, and the sources and methods of 45 selection of participants 46 47 (b) Cohort study—For matched studies, give matching criteria and number of 48 exposed and unexposed 49 50 Case-control study—For matched studies, give matching criteria and the number 51 52 of controls per case 53 54 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and 55 effect modifiers. Give diagnostic criteria, if applicable 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 38 BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from Section and Item Item Reported on 1 Recommendation 2 No. Page No. 3 Data Sources/ 8* For each variable of interest, give sources of data and details of methods of 4 Measurement assessment (measurement). Describe comparability of assessment methods if 5 there is more than one group 6 7 Bias 9 Describe any efforts to address potential sources of bias 8 9 Study Size 10 Explain how the study size was arrived at 10 11 Quantitative Variables 11 Explain how quantitative variables were handled in the analyses. If applicable, 12 describe which groupings were chosen and why 13 14 Statistical Methods 12 (a) Describe all statistical methods, including those used to control for 15 16 Forconfounding peer review only 17 18 (b) Describe any methods used to examine subgroups and interactions 19 20 (c) Explain how missing data were addressed 21 22 (d) Cohort study—If applicable, explain how loss to follow-up was addressed 23 24 Case-control study—If applicable, explain how matching of cases and controls was 25 addressed 26 27 Cross-sectional study—If applicable, describe analytical methods taking account of 28 sampling strategy 29 30 (e) Describe any sensitivity analyses 31

40 on September 24, 2021 by guest. Protected copyright. 41 (c) Consider use of a flow diagram 42 43 Descriptive Data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) and 44 information on exposures and potential confounders 45 46 (b) Indicate number of participants with missing data for each variable of interest 47 48 (c) Cohort study—Summarise follow-up time (eg, average and total amount) 49 50 Outcome Data 15* Cohort study—Report numbers of outcome events or summary measures over 51 time 52 53 Case-control study—Report numbers in each exposure category, or summary 54 measures of exposure 55 56 Cross-sectional study—Report numbers of outcome events or summary measures 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 38 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-043301 on 10 May 2021. Downloaded from Section and Item Item Reported on 1 Recommendation 2 No. Page No. 3 Main Results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates 4 and their precision (eg, 95% confidence interval). Make clear which confounders 5 were adjusted for and why they were included 6 7 (b) Report category boundaries when continuous variables were categorized 8 9 (c) If relevant, consider translating estimates of relative risk into absolute risk for a 10 meaningful time period 11 12 Other Analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and 13 14 sensitivity analyses 15 16 Discussion For peer review only 17 18 Key Results 18 Summarise key results with reference to study objectives 19 20 Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or 21 imprecision. Discuss both direction and magnitude of any potential bias 22 23 Interpretation 20 Give a cautious overall interpretation of results considering objectives, limitations, 24 multiplicity of analyses, results from similar studies, and other relevant evidence 25 26 Generalisability 21 Discuss the generalisability (external validity) of the study results 27 28 Other Information 29 30 Funding 22 Give the source of funding and the role of the funders for the present study and, if 31 applicable, for the original study on which the present article is based