Genetic Loci Associated with Plasma Phospholipid n-3 Fatty AcidsA Meta-Analysis of Genome-Wide Association Studies from the CHARGE Consortium

For decades, nutritionists told the public that eating fish wasn't strictly necessary. Your body could make its own omega-3s from plants by converting alpha-linolenic acid, or ALA, from flaxseed and walnuts into the long-chain fatty acids your heart and brain actually need. Simply put, a recent study examined the genomes of nearly nine thousand people and found that for a significant portion of the population, that conversion barely happens at all. And the reason lies in their DNA. The omega-3s we are discussing — eicosapentaenoic acid, or EPA, docosapentaenoic acid, or DPA, and docosahexaenoic acid, or DHA — are linked to a lower risk of sudden cardiac death, heart failure, metabolic syndrome, depression, and dementia. The body can synthesize them from plant-derived ALA through a series of chemical transformations. Desaturation enzymes reshape the molecule, elongase enzymes extend its carbon chain, and at the end of the process, you get DHA. However, tracer studies indicate that less than five percent of dietary ALA actually makes it through that full conversion. Most is burned for energy or stored as fat. Consequently, blood omega-3 levels vary tremendously between individuals, and diet alone does not explain why. Lemaître and colleagues aimed to investigate what else contributes to this variation. Their method was a genome-wide association study, or GWAS, which scans millions of genetic markers throughout the entire genome without assuming in advance where the interesting signals will emerge. The CHARGE Consortium meta-analysis pooled data from five population-based cohorts: ARIC, CHS, CARDIA, InCHIANTI, and MESA, totaling eight thousand eight hundred sixty-six participants of European ancestry. They measured all four major n-3 polyunsaturated fatty acids in blood plasma, expressed as a percentage of total fatty acids. Then they tested whether common single-nucleotide polymorphisms, or SNPs, which are single-letter differences in the genetic code, were associated with higher or lower levels of each fatty acid. Genome-wide significance was set at a p-value below five times ten to the negative eighth. What they discovered was a narrative in two acts, both unfolding within the metabolic pathway itself. The first act pertains to the FADS1 and FADS2 genes, which encode the desaturase enzymes responsible for the initial conversion steps. Minor alleles at SNPs in this cluster were associated with higher ALA at a p-value of three times ten to the negative sixty-fourth, and with lower EPA at five times ten to the negative fifty-eighth, as well as lower DPA at four times ten to the negative one hundred and fifty-fourth. Those are remarkable levels of statistical confidence. Carriers of these minor alleles tend to accumulate ALA — the plant substrate — because their desaturase enzymes are not converting it efficiently into the longer-chain products downstream. The locus explained three point eight percent of the variance in ALA levels, two point zero percent of EPA variance, and eight point six percent of DPA variance across the cohort. The team went further. They examined whether genotype actually modified the conversion response to ALA — in other words, does having these variants alter how much EPA you produce when ALA levels increase? For a representative FADS2 single-nucleotide polymorphism, rs1535, the interaction was statistically significant, with a p-value of nine point three times ten to the negative seventh. The data is revealing. For every one standard deviation increase in plasma ALA — about zero point zero five percentage points of total fatty acids — EPA rose by zero point zero eighty-six percentage points in individuals with no copies of the minor allele. In heterozygotes, it rose by only zero point zero sixty-three. In individuals with two copies of the minor allele, it increased by just zero point zero thirty-six. Each copy of the variant roughly halved the conversion gain. The same substrate, moving through a less efficient enzyme, produces a smaller output. This is a gene-diet interaction that is visible in blood. The second act unfolds further down the pathway, at the ELOVL2 gene, which encodes an elongase — the enzyme that stretches EPA and DPA into DHA. Here the pattern shifts in a revealing manner. Minor alleles at ELOVL2 SNPs were associated with higher EPA at a p-value of two times ten to the negative twelfth, higher DPA at one times ten to the negative forty-third, but lower DHA at one times ten to the negative fifteenth. Intermediates accumulate, while the final product falls short. This aligns perfectly with the expectation that the elongase step is the rate-limiting factor at this stage. When the team statistically adjusted DHA associations for EPA levels, the ELOVL2 signal sharpened considerably — the p-value for one key SNP improved from ten to the negative fifteenth to ten to the negative thirty-ninth — highlighting the close biochemical connection between these metabolites. Together, FADS and ELOVL2 present a coherent story of two independent bottlenecks in a single pathway. Impair the desaturase step with FADS variants, and ALA accumulates while EPA and DPA levels drop. Impair the elongase step with ELOVL2 variants, and EPA and DPA accumulate while DHA levels decrease. Common genetic variation controls the flow rate at each intersection, and millions of people carry variants at one or both locations. Then the genome surprised the researchers with an unexpected finding. Alongside the anticipated pathway genes, SNPs in GCKR — the glucokinase regulator gene, known for its role in glucose metabolism, not fat metabolism — showed a genome-wide significant association with DPA at a p-value of nine times ten to the negative ninth. The most associated marker was rs780094, and each copy of the minor allele was linked to zero point zero seventeen percentage points higher DPA. The team confirmed this wasn't merely a reflection of circulating lipid levels: adjusting for triglycerides, high-density lipoprotein, and low-density lipoprotein did not significantly alter the result. The biological connection linking glucose regulation to a long-chain omega-3 intermediate remains to be clarified, but the signal is genuine. The genome, it turns out, does not adhere to the boundaries we set around metabolic pathways. Now, all of this was established primarily in individuals of European ancestry. The natural follow-up question is whether these genetic effects apply more broadly. Lemaître and colleagues tested two representative SNPs — rs174548 in FADS1 and rs3734398 in ELOVL2 — in samples of African, Chinese, and Hispanic ancestry. For FADS1, the story was generally consistent. Across all four ancestry groups, the same allele was linked to higher ALA and lower long-chain n-3s. Effect sizes were similar in direction, although individual cohorts sometimes lacked the statistical power to reach significance on their own — a sample size issue, not a signal issue. ELOVL2 revealed a more complex narrative. In African-ancestry samples, associations with EPA, DPA, and DHA were similar to those in Europeans. However, among Chinese participants, the minor allele at rs3734398 was nearly fixed — present in ninety-two percent of the population — leaving almost no variation to detect any association. Among Hispanic participants, the allele was linked to higher DPA and lower DHA but not with EPA. When an allele is nearly universal in a population, it cannot explain differences within that population, regardless of whether it has functional consequences. The finding related to ELOVL2 may have universal effects but is only observable as an association in populations where it remains variable. The findings concerning FADS, by contrast, appear to be more genuinely applicable across different populations. Allele frequencies differ significantly — twenty-nine percent in Europeans, twenty-one percent in Africans, fifty-two percent in Hispanics, fifty-eight percent in Chinese — and in each case, the same directional association holds. This consistency across ancestries bolsters the argument that variation in FADS1 and FADS2 is genuinely causal, not merely a statistical artifact of population structure. What this body of findings means in practical terms is that dietary advice focused solely on food intake may overlook a biological layer. If you possess minor alleles at FADS2, a one-standard-deviation increase in dietary ALA generates roughly half the EPA gain compared to someone without those variants. Your genome is partially determining how effectively your diet is converted into the fatty acids your tissues actually utilize. This doesn't render diet irrelevant — quite the opposite. Fish remains the most direct source of EPA and DHA. However, two individuals consuming identical diets can experience significantly different circulating levels of the same fatty acids, and common genetic variation contributes to why this occurs. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.