Gene Variant Linked to Omega-3 Building Blocks Fails to Explain Variable EPA Response

A common variant in a gene that helps elongate omega-3 fatty acids shapes the baseline levels of certain long-chain fats in red blood cells, but it does not explain why people given identical high-dose EPA supplements end up with such different amounts of EPA in their circulation. That is the central conclusion of a secondary analysis of the seAFOod polyp prevention trial, published in the journal Genes & Nutrition. The study, led by Ge Sun and Mark A. Hull of the University of Leeds together with colleagues at the University of East Anglia and the University of Bradford, examined whether five single nucleotide polymorphisms in the ELOVL2 gene could account for the strikingly variable red blood cell EPA levels seen among trial participants who all received the same daily dose of the omega-3 fatty acid for a full year.

The question matters because the seAFOod trial, a placebo-controlled 2 × 2 factorial study of EPA capsules and aspirin for polyp prevention, had already documented wide inter-individual variability in red blood cell EPA levels during treatment with a standardised daily dose of 2000 mg of 99% pure EPA. Previous work by the same team showed that differences in baseline EPA level, capsule compliance, and EPA formulation contributed to that variable response. But the researchers suspected another factor: metabolism. EPA can be converted to n-3 docosapentaenoic acid, or n-3DPA, by the elongase enzyme ELOVL2, and n-3DPA is the intermediate product on the pathway toward docosahexaenoic acid, or DHA. In the earlier trial analysis, red blood cell n-3DPA rose in a bimodal pattern among participants randomised to EPA capsules, while DHA levels did not change during treatment. That pattern suggested that conversion of EPA to n-3DPA by ELOVL2 might be the key metabolic step behind the variable EPA levels in people receiving an identical dose.

Support for that hypothesis came from earlier genetic studies. Several investigations have reported associations between single nucleotide polymorphisms in the ELOVL2 gene, including rs953413, rs2236212, and rs3734398, and low circulating DHA levels. The Leeds team had also previously reported genotype-treatment interactions for polymorphisms in fatty acid desaturase 2 and in oxylipin-related genes such as PTGS1 and PTGS2 within the same trial. In that earlier genetic analysis of 647 seAFOod participants, the microfluidic multiplex SNP panel had already included five ELOVL2 variants: rs3734398, rs2236212, rs953413, rs3798713, and rs9393903, all of which satisfied Hardy-Weinberg equilibrium in the original cohort. The new analysis therefore tested whether ELOVL2 SNP genotypes were associated with the change in red blood cell EPA level in participants receiving EPA supplementation.

The study population consisted of 603 seAFOod trial participants who had both ELOVL2 genotype data and red blood cell n-3 HUFA data available at baseline. All five ELOVL2 SNPs satisfied Hardy-Weinberg equilibrium in this cohort and were in strong linkage disequilibrium, with four of them, rs3734398, rs2236212, rs953413, and rs3798713, showing R2 values greater than 0.95. Because the variants were so tightly correlated, the researchers focused on rs953413, the best characterised of the five, which is believed to be functional in human liver cells, the primary site of long-chain n-3 HUFA synthesis and metabolism. In the predominantly White European trial cohort, the rs953413 genotype frequencies were 30.8% GG, 50.1% GA, and 19.1% AA, a distribution consistent with that observed in Europeans in the 1000 Genomes Project.

The baseline comparison produced the study’s clearest finding. Before any trial intervention, minor (A) homozygotes displayed higher levels of n-3DPA, and also of DHA, in their red blood cells than major (G) allele carriers, with the difference reaching statistical significance for n-3DPA content. EPA levels themselves did not differ significantly by genotype at baseline. The corresponding median n-3DPA-to-EPA ratios, with interquartile ranges, were 2.55 in AA homozygotes, 2.34 in GA heterozygotes, and 2.34 in GG homozygotes, a difference that did not reach statistical significance. The researchers interpreted the higher n-3DPA in AA homozygotes as consistent with increased ELOVL2-dependent conversion of EPA to the longer-chain n-3DPA in that genotype group compared with G allele carriers.

To assess whether the variant influenced the response to supplementation, the team restricted the treatment analysis to participants who received EPA only, without concurrent aspirin, in order to avoid any potential indirect confounding effect of aspirin on n-3 HUFA metabolism. They analysed absolute red blood cell n-3 HUFA levels at each time point, as well as the change in percentage n-3 HUFA content between baseline and after either 6 or 12 months of treatment. Fatty acid levels had been quantified using liquid chromatography-tandem mass spectrometry, with a panel of nine fatty acids including EPA and n-3DPA, and relative n-3 HUFA levels expressed as the percentage of total measured fatty acids. Because the fatty acid data were not normally distributed, the researchers used non-parametric tests: the Kruskal-Wallis test for baseline comparisons across genotypes, and the pairwise Wilcox test with Benjamini-Hochberg adjustment for the change values, with statistical significance defined as p ≤ 0.05.

The treatment results were largely negative. After 6 months of EPA treatment, participants who were AA homozygotes showed an upward shift in the distribution of change in n-3DPA values compared with placebo users, and this shift appeared more pronounced than the corresponding distributions for GA heterozygotes or GG homozygotes, which overlapped more with the density curve for individuals who received placebo capsules only. However, the differences in change in n-3DPA values according to rs953413 genotype did not reach statistical significance. There was also no significant difference between the change in red blood cell EPA or DHA levels after 12 months of EPA treatment according to genotype. In short, the variant that appeared to shape baseline n-3DPA levels did not meaningfully alter how individuals responded to the supplement itself.

The authors place their findings in the context of a 2024 scoping review that identified six previous studies investigating the relationship between ELOVL2 polymorphisms and blood n-3 HUFA levels, four of which included rs953413 genotyping. In three of those six studies, the presence of at least one minor allele was associated with lower plasma EPA or DHA levels, while the remainder found no association. The new results align only partly with that literature, and the authors note an important methodological distinction: they measured red blood cell levels of n-3 HUFAs, whereas the majority of previous studies measured plasma levels. Variability across different fatty acid pools is highlighted by a comparison of the InCHIANTI cohort, which measured plasma and reported lower EPA levels according to rs953413 genotype, and the GOLDN cohort, which measured red blood cells and reported no difference.

The absence of an association between rs953413 genotype and the change in red blood cell EPA level during supplementation, the authors conclude, indicates that other host factors are more important determinants of tissue EPA levels during n-3 HUFA supplementation. These include age, sex, background dietary polyunsaturated fatty acid intake, and baseline circulating n-3 PUFA level. The relationship between genetic variants in other PUFA biosynthetic genes involved in the conversion of EPA to DHA, namely ELOVL5 and FADS2, and the response to EPA supplementation has yet to be addressed. The influence of rs953413 was evident for n-3DPA levels but not for EPA or DHA levels, an observation the authors consider consistent with the multi-step enzymatic synthesis of EPA and DHA from precursor alpha-linolenic acid, which involves two fatty acid desaturase genes and ELOVL5 in addition to ELOVL2.

The study has clear strengths and limitations. On the strength side, the researchers had a well-characterised cohort, standardised EPA dosing within a randomised trial, and direct measurement of red blood cell HUFA levels. On the limitation side, the modest size of the study limited the number of AA homozygotes available for comparison, the genetic analysis was restricted to five ELOVL2 SNPs, and the generalisability to racially diverse populations is reduced because the seAFOod trial cohort was predominantly White European. The work was performed as part of the STOP-ADENOMA study, approved by the London and Surrey Borders Research Ethics Committee and registered as ISRCTN05926847 as a sub-study of the seAFOod polyp prevention trial. Blood sampling was scheduled at trial entry before the investigational product was started, after 6 months of treatment, and at the end of the 12-month intervention period. The overall conclusion is measured: ELOVL2 SNP rs953413 minor homozygosity is associated with higher baseline red blood cell n-3DPA levels, but it does not significantly modify the change in circulating EPA level in response to high-dose EPA supplementation for 12 months, and the study therefore does not provide evidence that genetic variation in ELOVL2 contributes to the inter-individual variability in circulating EPA associated with high-dose supplementation.

Subject of Research: Association of the ELOVL2 rs953413 polymorphism with red blood cell n-3 HUFA levels and EPA supplementation response in the seAFOod polyp prevention trial

Article Title: The relationship between elongation of very long-chain fatty acids (ELOVL) 2 polymorphism rs953413 and blood n-3 HUFA levels: a secondary analysis of EPA treatment in the seAFOod polyp prevention trial

Article References: Sun, G., Davies, J. R., Mell, T., Harland, M., Saleh, R. N. M., Race, A. D., Loadman, P. M., Minihane, A. M., & Hull, M. A. (2026). The relationship between elongation of very long-chain fatty acids (ELOVL) 2 polymorphism rs953413 and blood n-3 HUFA levels: a secondary analysis of EPA treatment in the seAFOod polyp prevention trial. Genes & Nutrition, 21(1), Article 10. https://doi.org/10.1186/s12263-026-00797-w

Image Credits: AI Generated

DOI: 10.1186/s12263-026-00797-w

Keywords: ELOVL2, rs953413, EPA, n-3 HUFA, red blood cell fatty acids, seAFOod trial, polyp prevention, omega-3 supplementation, n-3DPA, DHA, genetic association, SNP