Health information, not medical advice. Supplementation decisions belong between you and a clinician who understands plant-based nutrition. This is a long, layered piece: the two-minute answer is right here at the top, and it gets progressively more technical the further you read.

If you want the whole omega-3 picture in one place, this walks from the plain-language verdict down through the biochemistry, the trials, the risks, the measurement problem, and the expert disagreement. Jump to any section using the links, or just start reading.

The short answer: do vegans need to supplement DHA and EPA?

Is a DHA supplement necessary on a healthy vegan diet?

If you are a healthy adult eating a well-planned whole-food vegan diet, there is no good evidence that you need to supplement DHA/EPA. There is also no good evidence that a small dose hurts you. This is a genuinely low-stakes decision, and anyone telling you otherwise in either direction is going beyond the data.

That is the honest summary. Here is the slightly longer version.

What about omega-3 for the heart on a vegan diet?

You almost certainly do not need a supplement for your heart. The big long-chain omega-3 supplement trials mostly failed to move hard outcomes; the most extensive review found at most a slight reduction in coronary heart disease. Plant omega-3 (ALA) has its own favorable outcome data, including a slight reduction in cardiovascular events in that same review. And vegans already have excellent cardiovascular profiles to begin with. For the bigger picture on dietary fat and heart health, see our deeper dives on dietary fat, cholesterol, and seed oils.

What about omega-3 for the brain on a vegan diet?

Nobody actually knows, in either direction. The supplement trials also failed for the brain, including a 2026 trial that proved the DHA reached the brain and still found no benefit. But no one has ever run a proper long-term study on the brains of healthy long-term vegans, so “we do not know” is the truthful answer, not “you are fine” and not “you are at risk.”

What actually matters more than DHA?

B12 (non-negotiable), vitamin D, iodine, sleep, exercise, not smoking, blood pressure, and overall diet quality. Every serious dementia-prevention framework ranks these far above omega-3, and WHO’s 2026 guidelines go further, recommending against omega-3 supplements for reducing dementia risk. If you are spending worry-energy on DHA while your B12 status is unverified, you have the priorities backwards. Start with our full guide to vitamin B12, then the broader vegan nutrition overview.

Who has the strongest case for taking DHA anyway?

• Pregnant and breastfeeding women. This is the one place near-consensus exists.
• Infants and young children. See our guide to raising children vegan.
• APOE4 carriers. A plausible case given altered brain DHA handling, though a 2026 trial found supplemental DHA reached the brain equally well with or without APOE4 and helped neither group.
• People with a confirmed very low Omega-3 Index (under about 3%). Not because harm is proven, but because you are outside the range that all the reassuring vegan data was built on.
• Known FADS1/FADS2 variants that impair conversion.
• Older adults with cognitive concerns or family history.

If you skip the supplement, what should you do instead?

1. Eat ALA daily. One tablespoon of ground flaxseed or chia covers the baseline recommendation. Walnuts and hemp seeds work too. Flax and chia have the best ALA-to-omega-6 ratio.
2. Grind your flax. Whole flaxseeds pass through you largely intact.
3. Do not drown everything in omega-6 oils. ALA and linoleic acid (LA) compete for the same enzymes. You do not need to fear omega-6; it is essential and the “ratio panic” is overblown (more below). But cooking everything in sunflower, corn, safflower, or generic “vegetable oil” works against you. Olive, avocado, or no added oil are better defaults. More in our seed oils article.
4. Do not megadose ALA either. There is some evidence that very high ALA can compete with itself at a later step in the pathway.* Moderate and consistent beats heroic.
5. Lock down B12. This one is not optional, and it is plausibly implicated in the one piece of evidence that makes vegetarians look worse on brain outcomes (see the cohort section).

* Am I eating too much ALA from whole foods? Almost certainly not. This concern is aimed at people taking flaxseed oil by the tablespoon on top of everything else, not at people eating whole foods. A typical day looks like: about 1.6 to 2.4 g ALA from a tablespoon of ground flax, about 2.5 g from a palm of walnuts, and roughly 0.3 to 0.8 g of background ALA from greens, soy, beans, and other seeds. That totals roughly 4 to 5.5 g/day, about 3 to 4 times the Adequate Intake (1.6 g/day for men, 1.1 g/day for women): comfortably normal, and squarely inside the studied range. And Klein 2025 gave about 3 g/day of supplemental flaxseed oil on top of habitual diet for nine months and still raised erythrocyte DHA by 12.8 to 26%, with conversion unaffected by LA or AA status. The mechanism is real but the threshold is high: both Delta-6 desaturase steps use the same enzyme (shown directly for the rat enzyme), and a study in human liver cells found evidence that ALA competes with the final substrate for it, so at some intake the first step could crowd the last. In people that remains a hypothesis, and those cell-study authors themselves read their results as showing that DHA synthesis from ALA is simply tightly regulated. The genuinely useful takeaway: if you eat 4 to 5 g of ALA daily and your index is still low, substrate is not your bottleneck, conversion is, and eating more flax will not fix that. That points at FADS genotype as the thing worth checking. One optional refinement: walnuts carry roughly 4 times more linoleic acid than ALA (about 10.7 g LA vs 2.5 g ALA per ounce) while flax and chia run the other way, so if you ever wanted to nudge the balance you could weight slightly more toward flax and chia. But since LA status did not affect conversion in Klein 2025, file this under marginal, not “fix this,” and do not let it scare you off walnuts, which are an excellent whole food.

Why the answer is “probably not”

Has the study that would settle this actually been done?

No, and this is the uncomfortable truth both sides skip. The study that would settle it: a long-duration randomized trial of algal DHA versus placebo, in long-term well-planned vegans, with cognitive and brain-imaging endpoints, controlling for B12 status (homocysteine and methylmalonic acid, not just serum B12), sleep, exercise, social engagement, and APOE genotype. That study does not exist and probably never will, because dementia has a decades-long latency and the event rate is too low to power a feasible trial. So everything below is inference from adjacent evidence.

What do the randomized controlled trials show? (Tier 1, strongest evidence)

This is the strongest evidence type, and it points the same direction over and over: no cognitive benefit from supplementation in the populations tested.

• AREDS2 (Chew, JAMA 2015): about 3,500 older adults, 1 g/day EPA+DHA, 5 years. No effect on cognition.
• VITAL-Cog (Kang, 2022): 3,424 plus 794 adults 60+, 840 mg/day, 2 to 3 years. Essentially flat (annual global cognitive change of about minus 0.01 SD units, 95% CI minus 0.02 to 0.003).
• MAPT (Andrieu, Lancet Neurology 2017): 3 years, 800 mg DHA plus 225 mg EPA. Omega-3 alone was null.
• ADCS-DHA (Quinn, JAMA 2010): 402 patients with mild-to-moderate Alzheimer’s, 2 g/day DHA, 18 months. No slowing.
• OPAL (Dangour, AJCN 2010): 867 healthy older UK adults aged 70 to 79, 200 mg EPA plus 500 mg DHA daily, 2 years. Null on the primary memory outcome and all secondary cognitive outcomes.
• Shinto (JAMA Network Open 2024): 102 adults aged 75+ with low omega-3 status, 975 mg EPA plus 650 mg DHA daily, 3 years. The primary endpoint, white matter lesion progression, was null; a signal appeared only in APOE4 carriers and only on a secondary imaging measure of neuronal integrity.
• PreventE4 (Yassine et al., 2026): 365 at-risk older adults, low DHA intake, 47% APOE4 carriers, 2 g/day DHA, 2 years. Null despite proven brain delivery. Now peer-reviewed and published in eBioMedicine, June 2026.

Why is the PreventE4 trial such a big deal?

Because it closed a loophole every earlier null trial left open. The old objection was: sure, blood levels went up, but did the DHA actually reach the brain? PreventE4 made cerebrospinal fluid target engagement its primary endpoint. The CSF DHA/AA ratio rose with a between-group difference of 0.19 (95% CI 0.16 to 0.21, p below 0.0001), and it happened equally in APOE4 carriers and non-carriers (interaction p equals 0.71), which is notable because the whole premise of the trial was that carriers handle DHA differently.

And then nothing happened. Hippocampal volume did not differ (left minus 0.26%, 95% CI minus 1.43 to 0.91; right plus 0.32%, minus 0.47 to 1.11). Cortical thickness did not differ. The RBANS cognitive score showed a treatment difference of 0.22 (95% CI minus 1.05 to 1.49); both groups improved about 2.7 points from practice effects. The authors concluded that despite biochemical target engagement, no differences in cognition or brain structure were observed over 24 months, and that future research should prioritize brain DHA metabolism over further supplementation trials. Lead investigator Hussein Yassine put it plainly in press coverage: their results do not support fish oil supplements as a preventive measure against Alzheimer’s.

Note what this did to the “dietary DHA cannot reach the brain” argument. The paper describes the supplement only as 2 g/day of DHA in 400 mg capsules, without naming its source or chemical form, but nothing in it suggests the specialised LPC or phospholipid formulation that the transport-form hypothesis says is required, and it demonstrably raised CNS DHA in humans. The strong version of the transport-form claim (that ordinary supplement DHA has virtually no effect on brain DHA) rested on mouse data comparing LPC-DHA with free DHA; triglyceride DHA was not directly tested. In humans at 2 g/day it clearly does something. The corollary cuts the other way too: if DHA got into the CNS and nothing happened, that is an argument against supplementation being useful, not an argument that dietary DHA is inert.

Do vegans and vegetarians have worse heart or brain outcomes? (Tier 2, cohorts)

On the cardiovascular side, the picture is reassuring with one asterisk. In EPIC-Oxford (Tong, BMJ 2019), vegetarians had 22% lower ischaemic heart disease (HR 0.78, 95% CI 0.70 to 0.87) but 20% higher total stroke (HR 1.20, 1.02 to 1.40), about 3 extra strokes per 1,000 over 10 years, mostly haemorrhagic. The stroke association held up after adjustment for disease risk factors, though the heart-disease benefit was partly attenuated by them (the ischaemic heart disease hazard ratio moved from 0.78 to 0.90 once cholesterol, blood pressure, diabetes, and BMI were accounted for, which is itself a clue that the benefit runs largely through those pathways). The authors did not attribute the stroke signal to omega-3; low LDL and total cholesterol is the leading candidate, with B12 also in the mix. Crucially, this stroke signal is not replicated: the Taiwanese Tzu Chi cohort (Chiu, Neurology 2020) found the opposite, with lower stroke across the board (overall HR 0.52), and a 2022 systematic review found no overall vegetarian-stroke association.

On the brain side, the cohorts mostly favor plant-based eating. Tzu Chi (Tsai, 2022) found reduced dementia risk in vegetarians (HR 0.671). Adventist Health Study-1 (Giem 1993) reported that meat eaters in a small matched substudy were about twice as likely to develop dementia (relative risk 2.18), but the result fell short of statistical significance (p = 0.065), a larger unmatched substudy found no significant difference, and the authors labelled the findings preliminary. Our full breakdown of that population is in the Adventist Health Studies article, and the broader longevity picture is in what the longest-living people eat and our look at whether the Blue Zones are debunked.

What about the Adventist Health Study-2 neurological finding?

This is the strongest single piece of evidence against the “vegans are fine” position, and it should not be waved away. AHS-2 (Abris et al., AJCN 2024; 88,400 participants, 12,515 deaths, median follow-up 11 years) found vegetarians had lower mortality overall and specifically from renal failure, infectious disease, diabetes, and heart disease. But estimated at age 85, the neurological causes ran the other way: stroke (HR 1.17, 95% CI 1.02 to 1.33), dementia (HR 1.13, 95% CI 1.00 to 1.27), and Parkinson’s disease (HR 1.37, 95% CI 0.98 to 1.91). Note that the Parkinson’s interval crosses 1.0 and the dementia interval touches it, so only stroke is unambiguously significant in that combined group.

Now the part that almost never gets reported, and it matters a great deal if you are a vegan reading headlines about this study. The paper breaks the results down by specific diet, and for vegans alone, not one of the neurological findings reaches statistical significance. At age 85 the vegan estimates were stroke 1.22 (95% CI 0.98 to 1.53), dementia 1.20 (0.98 to 1.48), and Parkinson’s 1.43 (0.85 to 2.42). Every interval crosses 1.0. The findings that are significant belong to lacto-ovo-vegetarians: stroke 1.20 (1.04 to 1.39) and Parkinson’s 1.44 (1.01 to 2.04). So the widely repeated claim that “vegans have higher neurological death” is really a finding about vegetarians pooled together, driven by the dairy-and-egg-eating group, and the vegan-specific numbers are too imprecise to conclude anything.

A second detail complicates the DHA reading further. Pesco-vegetarians, who eat fish and therefore consume preformed EPA and DHA, showed the same elevated pattern at 85: stroke 1.04, dementia 1.17, Parkinson’s 1.35. They were not protected. If low DHA were the mechanism, the fish-eaters should have been the group that escaped it.

In fairness, the authors themselves nominate low omega-3 as one candidate explanation, writing that a possible mechanism for the higher stroke risk in older vegans and lacto-ovo-vegetarians is their near-zero intake and low tissue levels of long-chain omega-3 in the absence of supplementation. That deserves to be quoted rather than hidden. But they raise B12 in the same breath, noting that although B12 intake was high on average, 15.2% of vegans and 10.6% of lacto-ovo-vegetarians had intakes below 2.0 micrograms a day, and that B12 deficiency may itself impair conversion of shorter-chain to long-chain fatty acids. They also flag the omega-6 to omega-3 ratio, which predicted stroke mortality in this cohort.

The honest bottom line: a real signal worth taking seriously, the best argument the other side has, and one that is (a) not statistically significant in vegans specifically, (b) present in fish-eaters too, and (c) attributed by the authors to at least three competing mechanisms, only one of which is DHA. Their own closing word on it is that these results need further support and investigation.

Has anyone actually studied vegan brains directly? (Tier 3)

Barely. The closest thing is work from Thomas Sanders’s group in the UK, summarised in his 2009 review. They reported being “unable to find any evidence of cognitive impairment” in vegan men: on the National Adult Reading Test, a standard measure of pre-morbid IQ, 118 vegan men averaged 119 (range 101 to 128). The review gives that vegan figure on its own, without an omnivore comparison, so it shows an absence of obvious impairment rather than equivalence. The group also ran a small placebo-controlled trial in which 200 mg of DHA a day for three months raised the proportion of DHA in vegans’ plasma by 50 percent, published only as a conference abstract. The review’s own summary is that there is “no evidence of adverse effects on health or cognitive function with lower DHA intake in vegetarians.” It is thin evidence, but it points the reassuring direction.

What does ALA do on its own? (Tier 4)

This gets skipped constantly and should not be. Naghshi (BMJ 2021), a dose-response meta-analysis, found dietary ALA associated with reduced all-cause, cardiovascular, and coronary heart disease mortality, though also with a slightly higher risk of cancer mortality. Zamroziewicz (2017), a cross-sectional study of about 100 older adults, found that a blood pattern of ALA and two related precursors, stearidonic acid (SDA) and eicosatrienoic acid (ETE), predicted better fluid intelligence, with frontoparietal gray matter volume fully mediating the link, while the pattern of EPA, DPA and DHA did not. The Cardiovascular Health Study (Virtanen 2013) found its clearest brain-imaging signals for long-chain EPA and DHA, which were linked to fewer silent brain infarcts and better white matter grade; plasma ALA was associated only with modestly better markers of brain atrophy, and that finding did not hold up when ALA intake was analysed. And the 2026 UK Biobank early-onset dementia analysis (Sala-Vila, Harris et al.; n=217,122) found higher plasma total omega-3 associated with 36% lower early-onset dementia risk (HR 0.64, 95% CI 0.47 to 0.87), with the “non-DHA” fraction (ALA + EPA + DPA) significant across more quintiles than DHA alone. That last one supports the plant-omega-3 position, but it is observational, comes from a group that sells omega-3 testing, and “non-DHA” is a subtraction estimate.

What do the major health authorities say? (Tier 5)

The 2025 Academy of Nutrition and Dietetics vegetarian position paper states that EPA and DHA can be obtained through vegan algal supplements if desired or recommended, although this may not be a requirement for health or disease prevention, and notes that isotope studies suggest ALA conversion may be sufficient to maintain adequate DHA through plant-sourced ALA alone. WHO’s 2026 dementia-risk guidelines (second edition) recommend against omega-3 polyunsaturated fatty acid supplementation to reduce the risk of cognitive decline or dementia, a strong recommendation based on moderate-certainty evidence. The Alzheimer’s Association says there is insufficient evidence to recommend DHA for treating or preventing Alzheimer’s. And the Yu 2020 umbrella review (243 observational studies, 153 trials) gave its strongest evidence rating to factors such as education, cognitive activity, high homocysteine, midlife hypertension, diabetes, and high body mass index in later life; omega-3 does not appear among its recommendations.

The biochemistry: how ALA becomes DHA

Why are only ALA and LA “essential” fatty acids?

Your body cannot make omega-3s from scratch. It lacks the Delta-12 and Delta-15 desaturase enzymes that place double bonds at the n-3 and n-6 positions. Plants have them; we do not. That is why ALA (18:3n-3) and LA (18:2n-6) are the only two truly essential fatty acids, a point we cover more broadly in our guide to essential nutrients. Everything else in the omega-3 family is built from ALA through alternating desaturation (adding a double bond) and elongation (adding two carbons).

What is the actual ALA to DHA conversion pathway?

The chain runs: ALA (18:3n-3), then via Delta-6 desaturase (FADS2) to SDA (stearidonic acid, 18:4n-3), then via elongase (ELOVL5) to ETA (20:4n-3), then via Delta-5 desaturase (FADS1) to EPA (20:5n-3), then via elongase (ELOVL2) to DPA n-3 (22:5n-3), then elongated to a 24-carbon intermediate, desaturated again by FADS2, and finally shortened back to 22 carbons by one round of peroxisomal beta-oxidation to yield DHA (22:6n-3).

Two features of this pathway matter enormously and almost never get explained. First, the last step is genuinely strange: DHA is not made by simply elongating DPA. The chain is elongated to 24 carbons, desaturated again by the same Delta-6 enzyme used at step one (demonstrated with the rat enzyme), then shortened back to 22 carbons by beta-oxidation in the peroxisome. This is the Sprecher pathway, and it is why DHA synthesis is inefficient and tightly regulated: it requires routing the molecule through a different organelle. Second, the same FADS2 enzyme is needed twice, at step one and at the final desaturation. That is the mechanistic basis for the claim that too much ALA might impair DHA synthesis by saturating the enzyme at the first step.

Does omega-6 block omega-3 conversion?

Less than the internet suggests. The omega-6 pathway (LA to GLA to DGLA to arachidonic acid) runs on the same FADS1, FADS2, and ELOVL5 enzymes, which is the basis of the concern. Rett and Whelan 2011 found that decreasing dietary LA by up to 90% showed no significant correlation with tissue AA (p=0.39), and increasing LA up to six-fold likewise showed no correlation (p=0.72): tissue AA is remarkably resistant to dietary LA. The FAO 2010 report concluded there is no rationale for a specific omega-6-to-3 or LA-to-ALA ratio recommendation if absolute intakes are adequate. That said, some studies do show that lowering LA while keeping ALA up raises EPA and DHA. The honest position: the ratio is not the crisis it is made out to be, absolute ALA intake matters more, and there is no reason to fear omega-6 as a class. Get your ALA, do not cook everything in high-LA oils, stop there. We cover this fully in our seed oils and dietary fat articles.

Is it true that only 1% of ALA converts to DHA?

The figure is real but it is used misleadingly. The actual numbers: in Burdge’s tracer studies, young men converted ALA to EPA at roughly 8%, and DHA synthesis was so low it was undetectable, while young women converted to EPA at about 21% and to DHA at about 9%. Notably, Burdge himself has since argued, in a 2022 review explicitly labelled an opinion, that ALA interconversion is sufficient as a source of longer-chain omega-3s in humans. Across the wider literature the male DHA figure is usually quoted as a range from under 0.1% up to about 4%, which reflects pooling many studies rather than any single measurement. Infants convert better than adults.

Why the percentage is the wrong question: a tracer percentage tells you what fraction of one labeled dose ended up as DHA. Most ALA is beta-oxidized for energy or stored, which is not a failure, it is what fat is for. The percentage says nothing about whether total-body synthesis meets tissue demand. Domenichiello, Kitson and Bazinet (2015) argued formally that oral tracer studies likely underestimate synthesis, because labeled ALA is absorbed into tissues and adipose before reaching the liver where conversion happens; using steady-state infusion methods, synthesis rates come out substantially higher. A rat study from the same group, funded by the food company Bunge, found brain DHA accretion was no different whether the rats ate ALA or preformed DHA, and the ALA-fed rats made DHA at roughly 100 times the rate the brain accumulated it; that is animal data, but it points the same direction. Do the arithmetic: brain DHA consumption is about 3.8 mg/day, adult adipose stores an estimated 20 to 50 g of DHA, so even a fraction-of-a-percent conversion rate produces enough. The valid counter-argument: producing enough for the brain’s minimum is not the same as optimal tissue status, and “adequate” here means absence of overt deficiency, which is a low bar.

Do the intermediates matter, or can I just take DHA directly?

This is the “whole pie versus one slice” question. The fair version: some intermediates probably do have independent value. In Zamroziewicz’s cross-sectional work, the precursor pattern (ALA, SDA and ETE) tracked with fluid intelligence and frontoparietal gray matter volume while the EPA, DPA and DHA pattern did not. DPA n-3 is increasingly studied on its own. ALA has independent outcome data. That is a decent argument for making sure you eat ALA rather than relying on a capsule alone. It is not a good argument that taking preformed DHA is harmful. And the mirror claim, that “taking DHA only gives you DHA,” is false in practice, though the mechanism is subtler than usually stated. Compound-specific isotope analysis (Metherel and Bazinet, 2019) shows DHA supplementation reliably raises plasma EPA, but the extra EPA does not carry the isotopic fingerprint of the DHA, so it is not true backward conversion; the likeliest explanation is that DHA spares existing EPA from being used up. Either way, taking DHA does raise your EPA level, so the strict one-way-pathway argument does not deliver the conclusion its advocates want.

Where DHA lives in the body and what it does

Where is DHA concentrated in the body?

DHA is not spread evenly. Two tissues stand out. In the retina, DHA makes up about 30% of total fatty acids, roughly half again as much as in the next most concentrated tissue, and it is most densely packed in the photoreceptor disk membranes, whose structural lipids are 80 to 90% phospholipid and only 8 to 10% neutral lipid, mostly cholesterol. In the brain, DHA is concentrated in gray matter and especially at synapses: it makes up roughly 18 to 45% of the fatty acids in isolated nerve terminals and synaptic vesicles, compared with under 5% in myelin. Arachidonic acid, an omega-6, is also a major brain fatty acid, worth remembering when people frame this as omega-3 good, omega-6 bad.

How much DHA does the brain actually use per day?

This reframes everything. Umhau et al. (Journal of Lipid Research 2009) used carbon-11 DHA PET imaging in living humans and found net brain DHA incorporation of 3.8 plus or minus 1.7 mg/day, whole-brain turnover of 0.076% per day, and a half-life of DHA in the human brain of approximately 2.5 years. Your brain consumes under 4 mg of DHA per day, and what is in there now takes years to turn over. Against 20 to 50 g stored in adipose, that reservoir alone could supply the brain for well over a decade before accounting for any ongoing synthesis.

Is DHA recycled in the body?

Yes, and heavily. When a membrane phospholipid is remodeled, phospholipase A2 cleaves DHA from the sn-2 position; it is not excreted but re-esterified back into new phospholipids by acyltransferases. It is a largely closed local loop. In the retina it is tighter still: shed photoreceptor disks are phagocytosed by retinal pigment epithelium cells, and their DHA is stored and efficiently recycled back to the photoreceptors. Practically, this means tissue DHA is not something you burn through and must constantly replace from diet, and it means any trial shorter than a few years is fighting the kinetics: one honest explanation for why the RCTs keep coming up null.

What does DHA actually do in cells?

Six double bonds make DHA extraordinarily flexible, which lets membranes deform (important for vesicle fusion at synapses and disk stacking in photoreceptors) and changes how embedded receptors and ion channels behave, underlying effects on neurotransmission including dopamine signaling. DHA is also the precursor to specialized pro-resolving mediators (neuroprotectin D1, D-series resolvins, maresins) that actively resolve inflammation rather than merely suppressing it, and to the endocannabinoid synaptamide. It also acts as a ligand for nuclear transcription factors including PPARs.

How does DHA get into the brain, and does the LPC form matter?

This is where the interesting fight is. Mfsd2a is a transporter at the blood-brain barrier that carries DHA in lysophosphatidylcholine (LPC) form, not as free fatty acid (Nguyen, Nature 2014); humans with MFSD2A loss-of-function mutations develop lethal microcephaly. So LPC transport is unambiguously real and important, especially in development. The strong claim (Sugasini and Subbaiah) rests on a mouse study in which dietary LPC-DHA more than doubled brain DHA while the same dose of free DHA did not; the authors argue ordinary triglyceride supplements behave like free DHA, though triglyceride DHA was not one of the forms they tested, which is the basis for arguing that standard algal or fish-oil supplements cannot help the brain.

Why the strong claim does not hold up as stated: it is mouse and rat data, with no completed human cognitive-outcome trial (two are underway, expected around 2029). Bazinet’s own lab found that plasma non-esterified DHA, not LPC-DHA, is the major pool supplying the rat brain, and a 2025 study from that group failed to reproduce brain DHA increases from dietary phospholipid carriers in mice. PreventE4 raised human CSF DHA with an ordinary DHA supplement; the paper does not specify its chemical form, but nothing in it describes an LPC or phospholipid preparation. And the specific claim that plasma LPC-DHA can be as high in someone with a low Omega-3 Index as a high one has no human confirmation; the closest human data (Ly et al., J Lipid Res 2023) found plasma LPC 22:6 did not rise with high-dose supplementation even as the index did, which is consistent with LPC-DHA being separately regulated but does not establish the cross-sectional claim. What survives is narrower but still important: the Omega-3 Index is not a validated proxy for brain DHA, so inferring brain deficiency from a low index is unproven.

Why do Alzheimer’s autopsies not consistently show lower brain DHA?

If low DHA drove Alzheimer’s, AD brains should be DHA-depleted. Some studies find that; many do not. Fraser 2010, described by its authors as the largest analysis of its kind (58 normal and 114 Alzheimer’s brains), found that DHA varied more in Alzheimer’s brains but mean levels were not significantly different. Igarashi 2011 found no difference in DHA across total phospholipids, though DHA was lower in two specific phospholipid classes. And Cunnane’s group (2012) noted that despite numerous studies it was still unclear whether omega-3 levels differ in Alzheimer’s plasma or brain; in their own matched samples, brain DHA was lower only in one phospholipid class in two cortical regions, by 12 to 14%. Possible explanations, none settled: low blood DHA in AD may be a consequence of neurodegeneration rather than a cause (reverse causation); blood DHA may simply mark overall diet quality; what is depleted may be oxidized DHA products rather than total DHA (AD autopsies do consistently show elevated lipid peroxidation, and lower magnesium); or brain DHA may be defended homeostatically until very late. What AD autopsies do consistently show points toward protecting DHA from oxidation rather than adding more.

Does cerebrospinal fluid reflect brain DHA?

Partly. CSF is a much better proxy than red blood cells because it is on the brain side of the barrier, but it is not neuronal membrane content. DHA in CSF has crossed into the CNS compartment and is equilibrating there; it does not prove incorporation into synaptic phospholipids. So a critic can fairly say PreventE4 showed CNS delivery without showing tissue incorporation. That is a legitimate gap and the last remaining refuge for the transport-form argument, but it is much narrower than “dietary DHA cannot reach the brain,” and it does not rescue the case for supplementing, since the null functional result is the harder problem for that side.

The part almost nobody discusses: getting DHA out of the liver

Read this section with a caveat in front of it. What follows is a chain of individually documented steps that has never been tested end to end in a human being. Every link has evidence behind it. Most of that evidence is in rats and mice. Nobody has shown that changing a person’s B12 status changes their brain DHA. If this article criticizes other people for running a mechanism past its evidence, the same standard has to apply here, so treat what follows as a coherent hypothesis worth knowing about rather than as established human physiology.

Is low DHA a manufacturing problem or a shipping problem?

Almost the entire public argument about vegans and DHA assumes the question is whether you can make enough. Conversion percentages, desaturase enzymes, tracer studies: all of it is about production. But there is a second question that almost never gets asked, and the animal evidence suggests it may matter more. Once DHA exists in your liver, how does it get out and reach your brain?

The answer is that it has to be packaged. DHA cannot simply float to the brain. It gets built into phosphatidylcholine, a phospholipid, and exported from the liver in that form. An enzyme then clips it into lysophosphatidylcholine-DHA, which is the molecule the brain’s dedicated transporter recognizes. Miss any step and the DHA stays put.

What does B12 actually have to do with omega-3?

Here the story gets interesting, and it corrects something that is widely repeated. When the Adventist Health Study authors raised B12 as a possible explanation for their neurological findings, they wrote that B12 deficiency “may reduce conversion from medium- to long-chain unsaturated fatty acids.” That framing is almost certainly wrong. None of the enzymes that convert ALA into DHA require B12. There is no cobalamin-dependent step anywhere in that pathway.

But there is a B12-dependent step in the packaging. B12 is required by methionine synthase, which makes methionine, which becomes S-adenosylmethionine, the body’s main methyl donor. One of the biggest consumers of those methyl groups is an enzyme called PEMT, which converts phosphatidylethanolamine into phosphatidylcholine in the liver. In B12-deficient rats, incorporation of methyl groups into liver phosphatidylcholine fell by roughly half, and the liver’s phospholipid balance shifted: more phosphatidylethanolamine, less phosphatidylcholine. The authors of that 1978 study ended with a line that reads as prescient now, speculating that such membrane changes “could conceivably have remote effects on other organs, such as the brain or spinal cord, the specialized functions of which are even more membrane dependent.”

Why does the PEMT pathway matter specifically for DHA?

Because the liver’s two routes to phosphatidylcholine do not make the same molecule. The CDP-choline route, which runs on the choline you eat, produces mostly medium-chain saturated species. The PEMT route produces a far more polyunsaturated set, and the species list includes stearoyl-DHA phosphatidylcholine directly. Purified PEMT prefers long-chain polyunsaturated substrate.

The knockout experiment makes the point sharply. Mice engineered to lack PEMT showed significantly depleted DHA in both liver and plasma phosphatidylcholine. Meanwhile DHA accumulated in their liver phosphatidylethanolamine. The DHA had not vanished. It was sitting in the wrong pool, unable to be packaged for export. The authors concluded that PEMT “is critical for mobilizing” DHA and arachidonic acid “from liver into plasma.”

And the detail that matters most for anyone hoping to eat their way around this: choline supplementation did not fix it. Choline restored liver phosphatidylcholine and plasma triglycerides, but the DHA depletion persisted independent of choline status. For this particular job, dietary choline is not a substitute for methylation capacity.

What happens after the liver exports it?

The final step needs an enzyme to convert the exported phosphatidylcholine into the transportable form. Mice lacking hepatic lipase had brain DHA of 13.5% of total fatty acids against 17.1% in their littermates. Mice lacking both hepatic and endothelial lipase fell to 8.3%. Deficiency of either lecithin cholesterol acyltransferase or endothelial lipase alone did nothing, which points to hepatic lipase as the main source of the brain’s supply.

Put together, the chain runs: B12 to methyl donors to PEMT to DHA-containing phosphatidylcholine to hepatic lipase to the transportable form to the brain’s transporter. If that chain is real in humans, it reframes the whole vegan question. The interesting variable would not be how much DHA you synthesize. It would be whether you can ship what you already have.

One detail from PreventE4 is relevant here. Every participant in that trial, in both arms, also received a standard vitamin B-complex supplement. That makes it a clean test of whether supplemental DHA can reach the brain compartment in people with adequate B vitamins, which it did. But it also means the trial cannot tell us whether low B12 status would impair that delivery, because nobody in it was allowed to be short. The question this section raises remains untested in humans.

It would also explain the Adventist findings without requiring vegans to be bad at making DHA, and it would mean the most important omega-3 intervention for a vegan is the one most of them are already doing: taking B12 seriously.

What happens to a cell that cannot get enough DHA?

The membrane does not simply go empty. The body substitutes the closest available molecule, which is a 22-carbon omega-6 fatty acid called docosapentaenoic acid, or 22:5n-6. It fits the same slot and preserves the membrane’s gross physical properties. What it does not do is work as well.

The cleanest test gave n-3-deficient rats preformed 22:5n-6 to see whether it could stand in for DHA. It could not: those animals took longer to escape in a water maze and showed defective spatial retention, and the authors concluded there is “a highly specific structural requirement” for DHA that its near-twin cannot satisfy. This is not a rodent curiosity, either. Infants fed vegetable-oil formulas show the same pattern, lower brain DHA with a compensatory rise in 22:5n-6.

A companion study repleting deficient rats produced two numbers worth remembering. Serum DHA recovered fast, reaching 73% of normal in one week. The brain did not: 19% recovery at one week, 35% at two, and full recovery only after eight. Meanwhile the omega-6 stand-in was cleared out faster than the real DHA came in, which tells you the body treats it as a placeholder rather than an equal. The practical lesson is that blood and brain run on completely different clocks, which is the mechanistic reason a blood test tells you so little about the organ you actually care about.

Could DHA be an independent risk factor that standard blood work misses?

This is a fair question and the answer is yes in principle, which is worth understanding properly.

The conventional cardiovascular panel measures a few distinct things. LDL and apoB measure atherogenic particle burden. Homocysteine tracks with endothelial injury. C-reactive protein measures inflammation. A person can score beautifully on all of them. But none of those numbers say anything about the fatty acid composition of a cell membrane, or about the electrical behaviour of heart muscle, and that is where omega-3 acts.

The historical tell is that the Omega-3 Index was not originally proposed as a marker of atherosclerosis. It was proposed as a predictor of sudden cardiac death, which is an arrhythmic event, not a plaque rupture. And the supporting observational data are striking: among men with no prior cardiovascular disease followed for up to 17 years, those in the top quartile of blood long-chain omega-3 had a relative risk of sudden death of 0.19 (95% CI 0.05 to 0.71) compared with the bottom quartile, with the trend holding after adjustment.

So the concept is sound. A distinct pathway exists, and the standard panel does not see it.

Two things keep this from becoming a reason to panic. First, an 81% risk reduction is an enormous observational effect, and if it were straightforwardly causal, the large supplementation trials should have found something. They did not. That gap places a ceiling on how much of the association can be causal, and it is the same ceiling that applies everywhere else in this article. Second, the membrane-electrophysiology mechanism is not uniformly protective: it is the same mechanism implicated in the atrial fibrillation signal at higher doses. A pathway that can steady a heart rhythm can also destabilise one.

The honest conclusion is narrower than either camp would like. There is a real, mechanistically distinct pathway that conventional biomarkers do not capture, which is a good reason not to treat a perfect lipid panel as proof that omega-3 status is irrelevant. But the trial evidence caps the size of the effect, and nothing here establishes that a healthy vegan eating plenty of ALA is on the wrong side of it.

Is this pattern unique to omega-3?

No, and this is where the story gets bigger than DHA. The same architecture shows up across several nutrients: the plant precursor is converted under feedback control, while the animal end-product enters already converted and bypasses the checkpoint.

Vitamin A is the clearest case. Retinoic acid induces a transcription factor called ISX, which represses both the transporter that absorbs beta-carotene and the enzyme that converts it. Good vitamin A status turns the tap down; poor status turns it up. That is why you cannot overdose on vitamin A from carrots but can from liver or supplements. Iron works the same way: non-heme absorption is governed by hepcidin according to your stores, while heme iron enters by a route that responds far less to need. And dietary DHA downregulates its own synthesis pathway, which is the same logic again.

Seen that way, “poor conversion” may be a description of a governor rather than a defect. But the framework cuts both ways and it is important to say so: B12 has no plant precursor and therefore no regulated pathway to trust, which is exactly why it is non-negotiable. Vegetarian populations also run a genuinely higher prevalence of iron deficiency, because a regulated route can under-deliver. The principle is not that plant forms are better. It is that the location of the checkpoint changes what can go wrong.

That pattern deserves more room than it can get here, and we treat it properly in its own article on precursors, end-products, and where the body puts its regulatory checkpoints.

The measurement problem: the Omega-3 Index

What is the Omega-3 Index actually validated for?

The Omega-3 Index is the percentage of EPA plus DHA in red blood cell membrane fatty acids. It was developed by William S. Harris, who also founded OmegaQuant, which sells the test, worth knowing when you weigh claims about what it means. Proposed categories: desirable 8 to 12%, intermediate 4 to 8%, undesirable below 4%. It is genuinely validated for one thing: correlation with coronary heart disease mortality across cohort studies, and it predicts all-cause mortality in the Framingham Offspring cohort. As an epidemiological risk marker in general, largely omnivorous populations, it has real support.

Does the Omega-3 Index reflect brain DHA?

It has never been validated for that, and this is the crucial distinction that gets silently elided. The human evidence here is mixed: the AHS-2 brain sub-study (Loong, 2023), an exploratory cross-sectional study of 40 older adults, found EPA, DHA, and the index all correlated with white matter volume, and EPA and the index also with entorhinal cortex volume, delayed recall, and processing speed; but none correlated with hippocampal volume or frontal cortical thickness, and DHA itself was not significantly associated with any of six cognitive tests. Add the autopsy literature above, the rodent finding that brain DHA correlated negatively with the erythrocyte index, and an infant study where RBC DHA explained only about a quarter of the variance in brain growth. So a “brain target” of 6% or 7% is an extrapolation from a cardiovascular-risk biomarker, not a validated threshold.

Is there really a 4.4 threshold for brain aging?

This number gets quoted a lot, so it deserves a close look. It comes from Tan, Harris, Beiser et al. (Neurology 2012), a cross-sectional analysis of 1,575 Framingham participants averaging 67 years old. The authors did use the word “threshold,” and they did report that the lowest quartile had smaller total brain volume and worse performance on several cognitive tests, with the discussion framing it as roughly two years of extra structural brain aging.

But four things are usually left out. First, the design is a single snapshot, and the authors state plainly that it “precluded examination of the relationship between RBC omega-3 PUFA levels and the development of clinical dementia.” It cannot show a rate of change or establish causation. Second, RBC DHA was not linearly related to any brain MRI measure; the “threshold” is a dichotomous comparison of the bottom quartile against the top three, and 4.4 percent is simply the 25th-percentile cutoff of that sample (the median was 5.3 and the 75th percentile 6.5), not a breakpoint derived from any spline or curve-fitting analysis. Third, the paper found no association between omega-3 status and its hippocampal surrogate measure or silent brain infarcts in any model, and the authors concluded the underlying mechanism looked more vascular than neurodegenerative, which is a rather different story from the one usually told with this study. Fourth, it has never been independently replicated as a threshold value. The pattern that the lowest quartile fares worst does recur (Satizabal 2022, Pottala 2014), but no study has validated 4.4 as an inflection point.

Two popular embellishments are also worth flagging. The vivid “about a teaspoon less brain matter” line appears in neither the paper nor the associated press release. And the claim that an elderly brain “ends up smaller than that of two- to three-year-old children” misstates Courchesne 2000, which documents decline from an adolescent peak, not shrinkage below toddler volume.

None of this means the association is fake. It means the number is a quartile cutoff from one cross-sectional study, not a validated danger line, and it should not be used to make anyone panic.

What is a typical Omega-3 Index for a vegan?

Sarter 2014/2015 found vegans averaged 3.7%, with 64% below 4% and 27% below 3%, on an average ALA intake of 3.4 g/day (more than double the AI). Notably the vegan mean of 3.7% was similar to that of omnivore US soldiers who were not taking omega-3 supplements, so “low vegan” is really “low non-fish-eater.” AHS-2 found a vegan mean of 3.7%, with neither ALA intake nor the LA-to-ALA ratio associated with the index. EPIC-Norfolk (Welch 2010) found the precursor-product ratio was higher in non-fish-eaters, evidence of upregulated conversion, though the vegan subgroup there was very small.

Can eating ALA raise your Omega-3 Index?

Modestly and slowly. Lane 2022 found ALA supplementation does not reliably increase the index while algal EPA/DHA does. But Klein 2025 (nine months, about 3 g/day flaxseed oil) did raise erythrocyte DHA by 12.8 to 26% across all diet groups, and Ezaki 1999 (20 elderly Japanese adults eating about 3 g a day of extra ALA from perilla cooking oil) found serum EPA and DHA rose only later, at 10 months, with DHA going from 5.3 to 6.4% of serum lipids, roughly a 21% increase. So duration matters, and ALA can move DHA, just usually not enough to take a 2% index to 8%.

What Omega-3 Index should a vegan aim for?

Here is the honest impasse: because RBC concentrations are not a validated brain marker, there is no defensible brain-based target. Eight percent has cohort support for cardiovascular risk in omnivores. Any brain target you see quoted is an extrapolation. And for vegans specifically, no target has ever been tested against outcomes. If you set a numeric goal, be clear you are doing it on cardiovascular or precautionary grounds, not because a brain threshold exists.

How much EPA and DHA does it take to raise the index?

Walker et al. (AJCN 2019) pooled 14 trials into a predictive model. To reach a mean index of 8% in 13 weeks with triglyceride-form EPA+DHA: about 2,200 mg/day from a 2% baseline, about 1,500 mg from a 4% baseline, about 750 mg from a 6% baseline. Ethyl-ester form needs roughly 40% more. Two caveats: these doses far exceed the 250 mg most plant-based clinicians recommend and sit above the 1,000 to 1,500 mg threshold where atrial fibrillation risk becomes measurable, and the lifetime dose needed to maintain 8% is much lower (average Japanese intake is 800 to 1,000 mg/day with a population mean index above 8%). Flock 2013 found dose alone explained 68% of the variability in response; adding body weight, baseline index, age, sex, and physical activity raised that only to 78%. In Sarter’s vegan trial, 254 mg/day moved the group mean from 3.1% to 4.8% over three to four months (the paper gives the duration as both 12 weeks and 4 months).

Risks of supplementing

Does omega-3 supplementation cause atrial fibrillation?

This is the best-established risk, and it is clearly dose-dependent. Gencer et al. (Circulation 2021), pooling 7 cardiovascular outcome trials, found omega-3 raised atrial fibrillation risk by about 25% overall (HR 1.25, 95% CI 1.07 to 1.46), and split it cleanly by dose: at or below 1 g/day the hazard ratio was 1.12 (95% CI 1.03 to 1.22), roughly a 12% increase, while above 1 g/day it was 1.49 (95% CI 1.04 to 2.15). The test for a difference between those two tiers was highly significant, and a dose-response model found risk rising about 11% per additional gram per day. The low-dose figure is also the fragile one: when the authors swapped in more complete atrial fibrillation data from one of the low-dose trials, the low-dose estimate fell to 1.08 (95% CI 0.99 to 1.17) and was no longer statistically significant, while the high-dose result held. So the evidence for any risk at the doses vegans typically take is weak, and the evidence for risk above 1 gram a day is solid. A larger 2026 meta-analysis (35 trials, 37 datasets, 114,592 participants) found significant risk only in the high-cardiovascular-risk, high-dose group above 1,500 mg/day, with a pooled odds ratio of 1.48 (95% CI 1.21 to 1.81) and an absolute risk difference of 0.8%, while the other three subgroups were non-significant. One caveat: its authors have omega-3 commercial interests, one as chief medical officer of a company selling omega-3 products and another, the senior author, holding stock in the company that sells the omega-3 index test. It is peer-reviewed, but weigh it knowing who produced it. UK Biobank (Chen 2024), an observational study of 415,737 adults, found regular fish-oil use in people without cardiovascular disease associated with a higher risk of developing atrial fibrillation (HR 1.13, 95% CI 1.10 to 1.17) and stroke (1.05, 1.00 to 1.11), while in people who already had cardiovascular disease it was associated with slower progression, for example from atrial fibrillation to major cardiovascular events (0.92, 0.87 to 0.98). Practical translation: at 200 to 500 mg/day you are well below where this shows up; at 2,000-plus mg/day, the dose you would need to take a 2% index to 8%, you are in the zone where it does.

Does omega-3 help if you already have atrial fibrillation?

Commonly believed, but the evidence does not support it. A meta-analysis of 4 RCTs (1,268 participants) found no effect on preventing AF recurrence (HR 1.13, 95% CI 0.96 to 1.33), and the FORWARD trial (1 g/day for a year in patients with recurrent symptomatic AF) did not reduce recurrent AF and found no difference in the composite of mortality, stroke, MI, embolism, heart failure, or severe bleeding. The one supportive study was a small post-ablation nested case-control, not a randomized trial. There is an observational, EPA-specific signal that EPA correlates inversely with ischemic brain infarcts in AF patients, but AF patients have been underrepresented in trials.

Does DHA raise LDL cholesterol?

Yes, modestly, and it is DHA-specific. Theobald 2004 found about 0.7 g/day DHA raised LDL cholesterol by 7%. Bernstein 2012 (a meta-analysis of 11 algal-oil trials, median dose 1.68 g/day) found LDL rose by about 0.23 mmol/L, roughly 8%, while triglycerides fell and HDL rose. The ComparED study (2018) found high-dose DHA had more pronounced LDL effects than high-dose EPA. This is partly offset because DHA shifts LDL toward larger, less atherogenic particles, but if your cardiovascular strategy is built around keeping LDL low, as many whole-food plant-based approaches are, it is a genuine consideration at higher doses. See our cholesterol article.

Does omega-3 or fish oil cause prostate cancer?

Contested, not resolved, but weaker than the headlines. Brasky 2013 (SELECT trial) found higher plasma phospholipid long-chain omega-3 associated with prostate cancer across grades: comparing highest to lowest quartile, total cancer HR 1.43 (95% CI 1.09 to 1.88), low-grade 1.44 (1.08 to 1.93), and high-grade 1.71 (1.00 to 2.94). Note that the high-grade interval touches 1.0, and the study had 834 cases of which 156 were high-grade. There was no separate analysis of advanced-stage disease. But nobody in SELECT was given fish oil, the team collected no fish-intake data, and Brasky himself said they did not know the source. The absolute blood differences between cases and controls were tiny, fractions of a percentage point of plasma phospholipid, and at levels typical of people who do not take fish oil at all. Reverse causation is plausible: 30 to 40% of men who developed cancer had baseline PSA over 3 versus 7% of controls, and pre-cancerous tissue shows altered fatty acid metabolism. Japanese men eat about 8 times more fish than American men, with an index of 8 to 10%, yet historically have far lower prostate cancer mortality. The same team had previously reported fish-oil supplement use was not associated with increased risk, dietary intake studies generally show no effect on incidence, and fish intake is associated with reduced prostate cancer mortality. Honest status: an unreplicated biomarker association with a credible reverse-causation explanation, not something that should drive your decision.

Does DHA oxidize and cause harm?

DHA has six double bonds, making it the most peroxidizable fatty acid in the body, and fish oil supplements have been documented to contain oxidized lipids, sometimes exceeding set limits. But the leap from “DHA is peroxidizable” to “DHA supplementation causes oxidative harm in vivo” is mechanistic speculation, the same species of overreach the anti-supplement camp rightly criticizes elsewhere. Practical response: buy fresh, buy from companies that test, store properly, do not megadose, and eat an antioxidant-rich diet, which vegans generally do. High doses (900 mg EPA plus 600 mg DHA and up) may also reduce immune function and increase bleeding time, per the Institute of Medicine and FDA, but neither is relevant at 250 to 500 mg.

Practical supplement questions

Does it matter that algae DHA is triglyceride form and not phospholipid?

Yes for absorption, no for the brain-transport argument. Forms you will encounter: natural triglyceride (in fish and algae), ethyl ester (cheaper, made during concentration), re-esterified triglyceride, free fatty acid, phospholipid (krill, not vegan), and monoacylglycerol. Triglyceride form appears to absorb somewhat better than ethyl ester: in Walker’s dose model, ethyl esters needed roughly 40% more to reach the same index. In a small single-dose crossover study of 12 men, Schuchardt 2011 found krill phospholipid highest, then triglyceride, then ethyl ester, though the differences were not statistically significant. Most algal oil is already triglyceride form, which is good, and Arterburn showed algal-oil capsules and cooked salmon were nutritionally equivalent DHA sources. The important non-answer: better systemic absorption is not the same as better brain delivery, and PreventE4 already showed that an ordinary DHA supplement, not a specialised phospholipid form, reaches human CSF. So buy triglyceride over ethyl ester for value, but do not imagine you are solving the brain-transport question by doing so. The tip that matters more than form: take it with a fat-containing meal.

Is sublingual or liquid omega-3 better absorbed than capsules?

No good evidence supports sublingual for omega-3. Fatty acids are absorbed via the intestinal lymphatic system in chylomicrons, which is not something the sublingual mucosa does well for a large lipid; sublingual makes sense for B12, not for omega-3. Liquid versus capsule makes no meaningful absorption difference, though liquid is usually cheaper per milligram and oxidizes faster once opened, so refrigerate it.

How do I compare omega-3 products with different doses and ratios?

Read the EPA plus DHA content, not the oil content. A “1,000 mg algae oil” softgel might contain 300 mg of actual EPA plus DHA. Find milligrams of EPA plus DHA per serving, divide by capsules per serving to get milligrams per capsule, then divide product price by total milligrams to get cost per milligram, and compare that across products. On ratio: most algal products are DHA-dominant, some DHA-only, a few balanced. Given that several lines of evidence point toward EPA being at least as relevant as DHA and that DHA is the one raising LDL and carrying the AF and prostate signals, a product containing both is probably a more sensible default than DHA-only. Look for third-party testing, a published oxidation value, minimal fillers, opaque packaging, and a reasonable expiry. Tank-grown algae avoids the ocean-contaminant problem entirely, which is the strongest argument for algae over fish oil regardless of everything else here.

How much DHA and EPA should a vegan take?

There is no evidence-based answer, because the outcome that would define “enough” has never been measured in this population. What exists: Jack Norris of VeganHealth.org, who writes that he doesn’t think there’s sufficient evidence that vegans require a dietary source of EPA or DHA, suggests that anyone wanting to be extra cautious either eat an additional 2,000 mg of ALA a day or take 200 to 300 mg of DHA a day (he formerly suggested 200 to 300 mg every 2 to 3 days); Greger at 250 mg/day; Fuhrman aiming for an index of 7 to 9% (which means well over 1 g/day from a low baseline); and about 200 mg preformed DHA as the general pregnancy guideline. The rough consensus among people who recommend it at all is 250 mg/day, which sits well below any risk threshold, and Sarter found that a nearly identical dose, 254 mg a day, moved a vegan group mean from 3.1% to 4.8% over three to four months.

The expert disagreement: eight voices, one unsettled question

This topic has attracted an unusually wide spread of credentialed opinion, and laying the voices side by side is more useful than picking a favorite. They disagree mostly about two things: how far above a 4 percent omega-3 index you should try to get, and whether observational biomarker data or randomized outcome data should drive the decision. Money matters too, so conflicts are flagged throughout. Here is the spectrum, from “it’s a scam” to “vegans clearly need it”: Geoff Palmer, Tim Radak, the Sherzais, Michael Greger, Philip Calder, Simon Hill, Rhonda Patrick, and Joel Fuhrman.

Is Geoff Palmer right that DHA supplements are a scam?

Geoff Palmer, founder of the plant-based sports-nutrition brand Clean Machine, has become the most prominent voice arguing that DHA supplements, fish and algae alike, are “100 percent a scam” and that vegans should take none. He has no scientific or clinical credentials, and there is a conflict worth stating plainly up front: Clean Machine sells Ahiflower oil, a plant ALA and SDA product that is a direct commercial substitute for the DHA supplements he attacks, and his reference document thanks an executive of the company that is Ahiflower’s exclusive supplier. That does not make him wrong, but it is the same kind of conflict this article flags for Fuhrman and Patrick, and it is larger than either.

What he gets right, and where even that gets slippery. Two kernels are real. First, the mechanism of retroconversion: Palmer says the ALA-to-DHA pathway runs one way and that DHA does not convert backward into EPA, and on the narrow enzymatic point the best human evidence (Metherel and Bazinet’s isotope work) agrees. But watch what he builds on it. His actual words are “if you take DHA, that’s all you get,” and that outcome claim is false: in those same studies, supplementing DHA raised plasma EPA by roughly 130 percent. The pathway doesn’t reverse, but your EPA level climbs anyway, because DHA spares existing EPA from being used up. So he is right about the plumbing and wrong about the water: taking DHA does not leave your EPA untouched. Second kernel, cleaner: he is right that the red blood cell omega-3 index does not measure brain DHA, a point the dolphin data and PreventE4’s own spinal-fluid design both support. On that one he is ahead of the sloppier pro-supplement talking points.

Where he overshoots. On almost every claim he starts from a real citable kernel and then sprints past it into an absolute, commercially convenient conclusion his own cited researchers do not endorse. “The brain can’t use dietary DHA at all” rests on mouse data from a lab with a competing-product interest, and PreventE4 showed an ordinary DHA supplement does reach the human central nervous system. “90 percent of conversion is invisible to blood tests” is his number, not a finding in any paper. His prostate-cancer claim misstates the study: the 71 percent figure was for high-grade cancer incidence, not death, and his proposed membrane-fluidity mechanism appears nowhere in the literature and is contradicted by studies showing DHA suppresses androgen-receptor signaling. “Inuit were dying at 35 because of DHA” inverts the history: their short lifespans were driven by infection, accident, and infant mortality, and life expectancy rose as the marine diet was abandoned. He also argues that LDL cholesterol itself does not cause heart disease, only oxidised LDL does, which is ironically a core low-carb and carnivore talking point. The European Atherosclerosis Society’s consensus review, drawing on more than 200 studies, over 2 million participants and more than 150,000 cardiovascular events, concluded that the evidence “unequivocally establishes that LDL causes” atherosclerotic cardiovascular disease. Its strongest line of evidence is genetic: people who inherit variants that lower LDL from birth have correspondingly less heart disease, which isolates LDL itself rather than any oxidised fraction. Most of that panel disclosed financial ties to makers of cholesterol-lowering drugs, which is worth knowing, but the conclusion rests on genetic data and on trials of several different drug classes whose results converge, not on the panel’s say-so. Oxidation is real, but it is a step in how LDL does its damage, not an alternative to LDL causing it: a particle has to be trapped in the artery wall before it oxidizes there, as our cholesterol article explains.

The dangerous one, with an important update. In his Plant Based News interview Palmer claimed B12 is “found in plants” and called the B12 requirement “garbage,” citing a hydroponic-lettuce study in which lettuce absorbed B12 only because it was artificially injected into the growing medium, at an absorption rate of hundredths of a percent. That is a horticultural proof of concept, not a dietary source. But in a later long-form debate with the dietitian Dr. Matthew Nagra, Palmer sharply walked this back, saying “please don’t listen to any influencer, me or anybody else, that says you don’t need to take B12,” calling that message “reckless” and “dangerous,” and reframing his point as a speculative theory about ancestral soil rather than current advice. Take the corrected version as his real position: vegans should supplement B12. The walkback matters, and it is a pattern, because Palmer himself concedes he overstates.

What happened when a dietitian debated Palmer directly?

The most useful test of Palmer’s case is the roughly five-hour debate in which Dr. Matthew Nagra, an evidence-focused plant-based dietitian, went through the claims one by one. Two things stand out. First, Palmer opened by conceding the game: he admitted he says “does” when he should say “may,” that his case is “theoretical and mechanistic” because “the data doesn’t exist,” and that he states things “as conclusions” while really raising “questions.” That is a remarkable admission from someone who told a larger audience that DHA is “100 percent a scam.” It means his strongest claims are, by his own account, hypotheses dressed as findings.

Second, the specifics did not survive scrutiny. On the central “DHA blunts EPA and causes harm” claim, Nagra showed Palmer was stitching together separate studies with different doses, different comparators, and different populations, then treating the contrast as if a single trial had pitted EPA against EPA-plus-DHA, which no study in his stack actually did. When Palmer leaned on a 37 percent “increased risk” figure, Nagra pointed out it was not statistically significant, with a confidence interval running from a 15 percent reduction to a doubling, meaning the data could not even establish the direction of the effect. On oxidation, Palmer’s key citation was a paper literally titled “Why Fish Oil Fails,” and the provenance of that paper matters. It was published in the Journal of Lipids in 2014 by Brian Peskin, and the journal retracted it later that year over an undeclared competing interest: Peskin was tied to a company holding patents on products that compete directly with fish oil. He is not a researcher, and Texas regulators fined him in 2003 over false claims that included misrepresenting himself as holding a PhD. So the marquee citation for the oxidation argument is a retracted paper by a supplement marketer with an undisclosed conflict. Nagra made the same point this article makes about the rest of Palmer’s mechanistic case: a laboratory oxidation finding is not a human outcome, and if isolated mechanisms settled things you could equally “prove” that exercise is bad because it too transiently raises LDL’s susceptibility to oxidation. Nagra’s closing framing was that Palmer’s position is grounded in mechanism and possibility, his own in outcome data, and that when the questions lack clear answers the honest move is caution, not certainty.

Net verdict: Palmer is a useful corrective on two narrow technical points and wrong, overstated, or, by his own admission, unfalsifiable on the headline claims that drive his “scam” conclusion. He walks the most dangerous one (B12) back when pressed, which is to his credit but also tells you how much weight the original framing could bear. None of his named researchers, including Radak, endorse the leap from “the biology is more complex than fish-oil marketing suggests” to “take no DHA.”

What does Dr. Tim Radak say about vegans and DHA?

Radak (DrPH, MPH, RDN; public health nutritionist, former Director of Nutrition at the Physicians Committee for Responsible Medicine; plant-based advocacy background, no commercial product) lands on “probably not needed.” His strongest contribution is the most rigorous treatment of confounding in this literature: during the decades these cohorts ran, people were told to swap red meat for fish, so a high omega-3 index often marks the person who followed health advice, and the bottom quintile often marks a poor overall diet. He documents dozens of uncontrolled variables and correctly notes WHO and the Alzheimer’s Association do not recommend supplementation. Where the case is weaker: the scrutiny runs in one direction. He forensically dismantles the one positive brain-volume trial (Witte 2014) without applying comparable scrutiny to the findings that support his own position, leans on the precursor-product result from Welch 2010 more heavily than that study’s very small vegan subgroup can bear, and makes a hundred-fold units error on the AHS-2 adipose DHA figure, writing 12% where Miles 2019 reports 0.12%. The underlying point there still stands, since vegans clearly do store DHA in adipose tissue; it is the magnitude that is off. He also leans at times on the same kind of mechanistic reasoning he criticizes elsewhere, particularly around DHA oxidation. His actual bottom line is moderate: an algae supplement could be considered, the research does not suggest benefit or need, except for FADS carriers. That is “not needed,” not “avoid.”

What does Dr. Joel Fuhrman say about vegans and DHA?

Fuhrman (MD) says vegans need it. Note the conflict: he sells both an EPA/DHA product and the omega-3 index test, and the article recommending both links to his own store, a more concrete conflict than Radak’s advocacy background. His strongest contribution is the AHS-2 mortality data, and he is right that the vegan index sits just under 4% and that ALA does not reliably raise it. Where he is weak: he essentially does not engage the null RCT literature (AREDS2, VITAL, MAPT, Cochrane were all available; his January 2026 article predates PreventE4 but not those), calls the mechanistic link “undeniable,” describes a vegan brain as “stripped of its primary defense,” dismisses the B12 explanation for AHS-2 in one sentence, and cites a Parkinson’s/pesticide study whose protective factor was actually ALA, the plant omega-3, to argue for EPA/DHA. His 7 to 9% target is a goal he concedes may be unattainable.

Why is DHA not in Dr. Greger’s Daily Dozen?

Because the Daily Dozen is a whole-food checklist, and supplements live on a separate list. This is commonly misread as Greger not recommending DHA. He does: his Optimum Nutrition Recommendations include 250 mg/day of pollutant-free (yeast- or algae-derived) long-chain omega-3s, alongside B12, vitamin D, and iodine. Ground flaxseed is in the Daily Dozen and covers ALA. His stated rationale is specifically brain, explicitly not heart: in his own words this would be predominantly for brain health, not heart health, since the long-chain omega-3s have disappointingly not reduced cardiovascular risk. He cites Sarter’s finding that a supplement of roughly that size (254 mg a day in the study) moved vegans from 3.1% to 4.8% and frames it as precaution: our bodies can elongate ALA, but can they make enough for optimal brain health, and until we know more, take 250 mg. That is arguably the most defensible of the four positions: explicitly precautionary rather than evidence-of-benefit, honest about the cardiovascular failure, and at a dose below any risk threshold.

Does Greger’s brain-atrophy argument still hold up?

Partly, and it is worth separating the pieces. His argument runs: low omega-3 status correlates with brain atrophy, there is a threshold around 4.4, most vegans fall below it, one trial showed supplementation slows atrophy, therefore take 250 mg.

The correlation link holds. Framingham, WHIMS, and the 2022 midlife Framingham analysis all find the lowest omega-3 group faring worst on brain measures. The vegan-status link holds too, straight from Sarter. Those parts are accurate.

The two weakest links are the threshold and the trial. As covered above, 4.4 is a quartile cutoff rather than a validated breakpoint. And the single supporting RCT, Witte 2014, has never been replicated in the decade-plus since, despite being the linchpin of the causal claim. Meanwhile the evidence has moved against that link: AREDS2, MAPT, VITAL-Cog, and Shinto 2024 were all null on cognition or brain imaging, and PreventE4 in 2026 pushed participants from an index of about 4.9 all the way to about 11, proved the DHA reached the central nervous system, and still found no difference in hippocampal volume, cortical thickness, or cognition.

One honest caveat rescues his specific claim from outright falsification. PreventE4 participants started at about 4.9, which is above his 4.4 line, so the trial tested whether pushing already-adequate people higher helps, not whether rescuing below-threshold people helps. No trial has ever tested that second question with brain-structure endpoints. So his hypothesis is best described as unfalsified but also unsupported by any randomized evidence, and the honest framing of a 250 mg dose is insurance against a documented low biomarker, not a proven way to protect your brain.

What do the Sherzais, the vegan neurologists, recommend?

Dean and Ayesha Sherzai are neurologists who co-direct an Alzheimer’s prevention program, wrote The Alzheimer’s Solution, and raised two children vegan from conception. Because they are plant-based brain specialists, their position carries particular weight, and it is notably measured. They are food-first: eat omega-3-rich whole foods like chia and flax, cut the saturated and omega-6 fat that impairs conversion, and supplement algae-based DHA optionally, mainly at high-stakes life stages. As Dean Sherzai put it, “eat foods rich in omega-3s like chia and flax seeds and you should be fine, but even then, if you’re worried, especially in certain times of life like developing brain, children, pregnancy, or aging brain, take a supplement, algae-based.” Their book names a figure of at least 250 mg DHA per day for those who choose to supplement. Ayesha Sherzai supplemented during pregnancy but currently takes only B12 plus a careful diet.

Two things make them useful in this lineup. First, their own peer-reviewed systematic review (2023) is the most sober document any plant-based advocate here has produced: it found “inconsistent effects” of supplementation and a likely “threshold effect” in which people already meeting their needs gained nothing, with a trend toward benefit only in those already declining. Second, they have no supplement line and no testing-company sponsorship that I could find, which sets them apart from Fuhrman. They frame Alzheimer’s as largely modifiable by lifestyle and treat any single supplement as minor next to diet pattern, blood pressure, glucose, and sleep. Their practical stance sits almost exactly alongside Greger’s: a small algal dose as reasonable insurance, prioritized for pregnancy and the aging brain, not a mandate.

What does Professor Philip Calder say?

Philip Calder is one of the most-cited academic lipid scientists in the world and a useful non-vegan, non-commercial anchor. He is careful with the word “essential”: strictly, only ALA and linoleic acid are essential, because vegan populations “eat zero EPA and DHA and they live,” so DHA is best called conditionally essential, an optimization rather than a survival requirement. He is candid that ALA-to-DHA conversion is low (under about 5 percent), that high omega-6 intake worsens it, and that you cannot reach the EPA and DHA levels associated with benefit by eating ALA alone. His practical read: preformed EPA and DHA are helpful, most people don’t get enough, and if you don’t eat fish you should both mind your omega-6 intake and consider a supplement, aiming for a modest lifelong intake rather than short high doses.

What makes Calder valuable is his evenhandedness on the surrounding fights. He pushes back hard on omega-6 fear, noting that people with the highest blood linoleic acid have the lowest cardiovascular and diabetes risk, and he cautions that raising omega-3 by slashing omega-6 is not automatically a net win because omega-6 does essential work too. He also frames omega-3 as “a nutrient, not a drug,” arguing the null supplement trials asked too much: a small dose, added late, on top of multiple medications, over a few years, cannot mimic a lifetime of adequate status. That framing is the single best explanation on offer for why the observational and trial evidence point different directions, and it applies to the vegan question directly.

What does Dr. Rhonda Patrick say about omega-3?

Patrick (PhD, biomedical science) is the non-vegan voice here, and including her is useful precisely because she is working from the same biomarker literature as the plant-based advocates but reaching a much more aggressive conclusion. Her position: aim for an Omega-3 Index of 8 to 11 percent, which she describes as the range with the greatest health benefits.

What she gets right deserves credit. Her treatment of supplement forms is accurate and matches the independent evidence: triglyceride and phospholipid forms need pancreatic lipase, ethyl esters need an additional enzyme and therefore a fatty meal, and free fatty acid forms need no enzymatic digestion at all. She reports the Walker 2019 dosing model correctly. Her coverage of FADS and ELOVL genetics, sex differences in conversion, and the finding that the index rises with age is all sound. She reports the atrial fibrillation dose split honestly rather than burying it, noting that under 1 gram raised risk 12 percent while over 1 gram raised it 49 percent. And notably, she concedes that the Omega-3 Index may not be the best marker for the brain, citing work suggesting plasma BDNF better reflects brain concentrations. That is a real point of agreement with the plant-based skeptics.

Where the argument strains. Her page is built around roughly ten pull-quotes from William Harris, who developed the Omega-3 Index and owns the company that sells the test, which is a heavier reliance on a single commercially interested source than any of the vegan authors. The 8 to 11 percent target is sourced to a 2004 paper proposing the index in the first place, which is thin support for calling it robust. And there is an unresolved internal tension: she targets 8 to 11 percent, correctly notes this requires 1,750 to 2,500 mg per day, and correctly notes that over 1 gram per day raises atrial fibrillation risk by 49 percent, without reconciling the three. Her weakest material is the COVID section, which leans on a 101-person trial reporting a survival difference too large to be plausible and on computer modeling of the viral spike protein.

Why does Rhonda Patrick recommend fish over algae supplements?

She does acknowledge algae. She notes algal sources suit people who avoid fish and cites the systematic review showing algal supplementation raises DHA and the omega-3 index in vegetarians. But the framing subordinates it: the article states it focuses primarily on marine-derived omega-3s, and the main comment on algal products is that they provide lower concentrations than similarly priced fish oil. Her enthusiasm goes toward salmon roe as a concentrated phospholipid source.

The reason matters, and it produces the most interesting collision in this whole debate. Her stated rationale is that APOE4 carriers respond well to DHA from fish but less well to most supplements, because fish provides DHA in phospholipid form, which determines whether it becomes free DHA or LPC-DHA for transport across the blood-brain barrier. She cites her own 2019 hypothesis paper for this.

That is the same mechanistic argument the plant-based side uses, aimed in the opposite direction. The vegan version says the LPC transport route means ordinary supplements cannot enrich the brain, so rely on ALA and endogenous conversion. Patrick’s version says the LPC transport route means non-phospholipid supplements cannot enrich the brain, so eat fish and roe. Identical premise, opposite conclusion. That alone should make anyone cautious about how much weight the mechanism can carry.

And PreventE4 undercuts both versions, hers quite directly. The trial gave an ordinary DHA supplement, not described as a phospholipid form, to a cohort that was 47 percent APOE4 carriers. Her hypothesis predicts carriers should be the group that fails to get it into the brain. Instead, cerebrospinal fluid DHA rose essentially equally in carriers and non-carriers, with an interaction p-value of 0.71. The transport-form part of the hypothesis did not fare well against the trial, with one honest caveat: the paper does not state the supplement’s exact chemical form, so a sceptic could still ask for that detail. Her page was last updated in July 2026, about a month after PreventE4 published, and does not address it.

What does Simon Hill (Plant Proof) recommend?

Simon Hill is a plant-based nutrition communicator whose position is the most pro-supplementation of the plant-based voices here, closer to Harris and Patrick than to Greger or the Sherzais. He targets an omega-3 index of 8 to 12 percent and recommends 1 to 1.5 grams a day of combined DHA and EPA, algae oil preferred, in triglyceride form, taken with a meal, with a baseline index test before starting and a retest after four months. He argues that vegans and vegetarians can run an index as low as 2 to 3 percent, that ALA from flax and chia is unlikely to be sufficient for most people, and that the atrial fibrillation signal is real but dose-dependent and outweighed, for people without cardiovascular disease, by the cardiovascular benefit he infers from the VITAL low-fish subgroup.

He discloses that he has no omega-3 product or testing company of his own, though he has sponsor and affiliate relationships with other supplement brands (among them a prebiotic he co-formulated, plus a plant-based multivitamin and other wellness products), so weight his advice knowing he profits from supplements broadly but not from omega-3 specifically. His dose sits at the high end of what this article’s evidence supports: reaching an 8 percent index from a low baseline genuinely does take the 1 to 2 gram range, but that is also the dose zone where the AFib and LDL trade-offs begin, and where the null cognitive trials found no benefit despite hitting those levels. His “measure and optimize to a target” philosophy is coherent, but it rests on treating the index as a goal worth chasing, which the brain evidence does not establish and the cardiovascular evidence supports only in general, largely omnivorous populations.

What do all these experts agree on?

More than the shouting suggests. Nearly everyone here accepts that the vegan omega-3 index runs around 4 percent, that high-dose omega-3 carries an atrial fibrillation signal, and that the Omega-3 Index is an imperfect marker for the brain specifically. The plant-based experts, even Palmer, agree that B12 is non-negotiable (Palmer being the lone, and dangerous, exception), and that flax, chia, and walnuts belong in the daily diet. Almost no one is telling healthy vegans to eat fish.

The disagreement is really about two things: how far above 4 percent to chase, and whether observational biomarker data or randomized outcome data should drive the call. It sorts the eight voices into three camps. The scam and skeptic camp (Palmer, Radak) weights the trials and the confounding critique and lands on “not needed” or “avoid.” The precautionary camp (the Sherzais, Greger, and Calder in practice) treats a small algal dose as cheap insurance, prioritized for pregnancy and the aging brain, without claiming proven benefit. The optimize-a-target camp (Simon Hill, Patrick, Fuhrman) weights the biomarker data heavily and pushes for an 8-to-12 percent index requiring roughly 1 to 2 grams a day, which is also exactly the dose zone where the AFib and LDL trade-offs appear and where the null cognitive trials found nothing.

The practical distance from the skeptics to the precautionary camp is one small 250 mg capsule. The distance from there to the optimizers is about eight capsules and a different theory of evidence. The real crux, still unresolved, is whether a 4 percent index in an otherwise healthy person is a benign set point or a quiet insufficiency with a decades-long fuse. Nobody has the evidence to close it. The sharpest illustration is watching the same LPC-DHA transport mechanism get deployed by Palmer to argue for flaxseed and by Patrick to argue for salmon roe in the same year, while the Adventist Health Study-2 team, whose corresponding author is Gary Fraser, has published a signal of higher neurological mortality among the oldest vegetarians that no one can yet explain.

What would actually settle this

What study would answer the vegan DHA question for good?

A long-duration randomized trial of algal EPA/DHA versus placebo in long-term well-planned vegans, with cognitive and brain-structural endpoints, controlling for B12 status (homocysteine and MMA, not just serum B12), sleep, exercise, social engagement, and APOE genotype. It will probably never happen: dementia latency is measured in decades, early-onset dementia is too rare to power, the vegan population is small, and nobody has a commercial incentive to fund a trial whose likely result is “the cheap capsule does not matter.” What is actually in progress: LPC-DHA cognition trials (University of Cincinnati and others, expected around 2029), the first real human test of the transport-form hypothesis, plus continued follow-up of the AHS-2 and UK Biobank cohorts. What would change the picture: a positive LPC-DHA trial, replication of the AHS-2 neurological signal in an independent vegan cohort with B12 properly measured, or a trial in genuinely low-status individuals showing benefit.

The takeaway

So, bottom line, do healthy vegans need to worry about DHA?

If you eat a well-planned whole-food vegan diet, get your ALA from flax, chia, and walnuts, do not drown everything in omega-6 oils, keep your B12 verified with MMA and homocysteine rather than just serum B12, exercise, sleep, stay socially and cognitively engaged, and keep your blood pressure and metabolic health in good shape, you are doing far more for your brain and heart than omega-3 capsules have been shown to do for generally healthy people in randomized trials.

If you want to take 250 mg of algal EPA/DHA anyway as cheap insurance against a question nobody has answered, that is a defensible choice with a very small downside. If you do not, that is also defensible, and the evidence is arguably on your side. What is not defensible is the confidence on either end: the people telling you DHA is dangerous and you should avoid it, and the people telling you your brain is being stripped of its defenses. Both are running well past what anyone actually knows. Talk to a physician or registered dietitian who understands plant-based diets, especially if you are pregnant or planning to be, know your APOE status, are over 65, or have a confirmed very low index.

For the rest of the picture on eating this way, see our guides to vegan nutrition, B12, dietary fat, cholesterol, seed oils, plant toxins and antinutrients, raw diets, losing weight, and raising children vegan. And if you came here through the ethics-and-environment side of the conversation, our pieces on crop deaths, the pesticide argument, the “livestock eat food we can’t” claim, regenerative farming, rewilding, coconut oil vs animal fat, the Adventist Health Studies, the longest-living people, the Blue Zones, the Saul Newman Blue Zones fact-check, and the meat and bowel cancer link go deeper.

References and further reading

Links open in a new window. Where a study was discussed above, the primary source, DOI, or PubMed record is linked here.

Supplementation trials: brain and cognition

• Chew EY, et al. AREDS2 cognitive outcomes. JAMA 2015. PubMed
• Kang JH, et al. VITAL cognitive ancillary. Alzheimer’s & Dementia TRCI 2022. PubMed
• Andrieu S, et al. MAPT trial. Lancet Neurology 2017. PubMed
• Quinn JF, et al. ADCS DHA trial in Alzheimer’s. JAMA 2010. PubMed
• Dangour AD, et al. OPAL trial. Am J Clin Nutr 2010. PubMed
• Shinto L, et al. Omega-3 and white matter, JAMA Network Open 2024. PubMed
• Yassine HN, Ghasem Pour S, Juarez M, et al. CNS target engagement of high-dose DHA supplementation in older adults at risk for dementia: a randomised, double-blind, placebo-controlled trial. eBioMedicine 2026;129:106316. PubMed
• Sydenham E, Dangour AD, Lim WS. Omega-3 for prevention of cognitive decline. Cochrane 2012. PubMed
• Witte AV, et al. Long-chain omega-3 and brain function/structure. Cereb Cortex 2014. PubMed

Cardiovascular outcome trials and reviews

• Abdelhamid AS, et al. Omega-3 for cardiovascular disease. Cochrane 2020. PubMed
• Manson JE, et al. VITAL trial (marine n-3). N Engl J Med 2019. PubMed
• ASCEND Study Group. n-3 fatty acids in diabetes. N Engl J Med 2018. PubMed
• Naghshi S, et al. Dietary ALA and mortality, dose-response meta-analysis. BMJ 2021. PubMed

Atrial fibrillation

• Gencer B, et al. Omega-3 supplements and atrial fibrillation risk. Circulation 2021. PubMed
• Abuknesha NR, O’Keefe JH, Qian F, Tintle NL, Harris WS, et al. Effects of omega-3 fatty acid treatment on risk for atrial fibrillation: an updated meta-analysis of 35 trials including 114,592 individuals. Circ Arrhythm Electrophysiol 2026. PubMed
• Jiang Y, Tan HC, Tam WWS, Lim TW, Wang W. A meta-analysis on omega-3 supplements in preventing recurrence of atrial fibrillation. Oncotarget 2018;9:6586-6594. PMC
• Chen G, Qian ZM, Zhang J, et al. Regular use of fish oil supplements and course of cardiovascular diseases: prospective cohort study. BMJ Medicine 2024;3:e000451. DOI
• Macchia A, et al. Omega-3 fatty acids for the prevention of recurrent symptomatic atrial fibrillation: results of the FORWARD trial. J Am Coll Cardiol 2013. DOI

LDL cholesterol and lipid effects

• Theobald HE, Chowienczyk PJ, Whittall R, Humphries SE, Sanders TAB. LDL cholesterol-raising effect of low-dose docosahexaenoic acid in middle-aged men and women. Am J Clin Nutr 2004;79:558-563. PubMed
• Bernstein AM, et al. Algal oil DHA meta-analysis. J Nutr 2012. PubMed
• Allaire J, et al. ComparED study (DHA vs EPA, LDL). J Clin Endocrinol Metab 2018. PubMed

Prostate cancer

• Brasky TM, et al. Plasma phospholipid fatty acids and prostate cancer in SELECT. J Natl Cancer Inst 2013. PubMed
• McCarty MF, DiNicolantonio JJ. Omega-3 and prostate cancer: examining the pertinent evidence. Mayo Clin Proc 2014. Mayo Clinic Proceedings
• Dietary omega-3 and prostate cancer progression, PLCO analysis 2025. PMC

Vegetarian and vegan cohorts

• Tong TYN, et al. EPIC-Oxford, heart disease and stroke over 18 years. BMJ 2019. PubMed
• Chiu THT, et al. Vegetarian diet and stroke, Tzu Chi cohort. Neurology 2020. PubMed
• Tsai J, et al. Vegetarian diet and dementia, Tzu Chi cohort. Nutrients 2022. Nutrients
• Abris GP, et al. Cause-specific mortality in vegetarians, AHS-2. Am J Clin Nutr 2024. PubMed
• Miles FL, et al. Plasma, urine, and adipose tissue biomarkers of dietary intake differ between vegetarian and non-vegetarian diet groups in the Adventist Health Study-2. J Nutr 2019. DOI
• Harris WS, Pawlak R, et al. Dietary and erythrocyte PUFAs in AHS-2 diet groups. PLEFA 2025. PubMed
• Welch AA, et al. EPIC-Norfolk, precursor-product ratio of ALA to long-chain n-3. Am J Clin Nutr 2010. PubMed
• Loong S, Barnes S, Gatto NM, Chowdhury S, Lee GJ. Omega-3 fatty acids, cognition, and brain volume in older adults. Brain Sci 2023;13(9):1278. Brain Sciences

Conversion, tracer studies, and the pathway

• Domenichiello AF, Kitson AP, Bazinet RP. Is DHA synthesis from ALA sufficient to supply the adult brain? Prog Lipid Res 2015. PubMed
• Barcelo-Coblijn G, Murphy EJ. ALA and its conversion to longer chain n-3. Prog Lipid Res 2009. PubMed
• Baker EJ, et al. Metabolism and functional effects of plant-derived omega-3 in humans. Prog Lipid Res 2016. PubMed
• FAO. Fats and fatty acids in human nutrition: report of an expert consultation. FAO Food and Nutrition Paper 91. Rome: FAO; 2010.
• Rett BS, Whelan J. Increasing dietary linoleic acid does not increase tissue arachidonic acid. Nutr Metab 2011. PMC
• Klein V, et al. ALA supplementation and long-chain n-3 profiles across diets (NuEva). Front Nutr 2025. Frontiers in Nutrition

Brain DHA, transport, and turnover

• Burdge GC, Wootton SA. Conversion of alpha-linolenic acid to eicosapentaenoic, docosapentaenoic and docosahexaenoic acids in young women. Br J Nutr 2002;88:411-420. DOI
• Burdge GC, Jones AE, Wootton SA. Eicosapentaenoic and docosapentaenoic acids are the principal products of alpha-linolenic acid metabolism in young men. Br J Nutr 2002;88:355-363. DOI
• Burdge GC. Alpha-linolenic acid interconversion is sufficient as a source of longer chain omega-3 polyunsaturated fatty acids in humans: an opinion. Lipids 2022;57:267-287. DOI
• D’andrea S, et al. The same rat Delta6-desaturase not only acts on 18- but also on 24-carbon fatty acids in very-long-chain polyunsaturated fatty acid biosynthesis. Biochem J 2002;364:49-55. DOI
• Portolesi R, Powell BC, Gibson RA. Competition between 24:5n-3 and ALA for Delta6 desaturase may limit the accumulation of DHA in HepG2 cell membranes. J Lipid Res 2007. DOI
• Sprecher H. Metabolism of highly unsaturated n-3 and n-6 fatty acids. Biochim Biophys Acta 2000;1486:219-231. PubMed
• Ezaki O, et al. Long-term effects of dietary alpha-linolenic acid from perilla oil on serum fatty acids composition and on the risk factors of coronary heart disease in Japanese elderly subjects. J Nutr Sci Vitaminol 1999;45:759-772. DOI
• Lane KE, Wilson M, Hellon TG, Davies IG. Bioavailability and conversion of plant based sources of omega-3 fatty acids: a scoping review to update supplementation options for vegetarians and vegans. Crit Rev Food Sci Nutr 2022. DOI
• Umhau JC, et al. Imaging incorporation of circulating DHA into human brain. J Lipid Res 2009. Journal of Lipid Research
• Nguyen LN, et al. Mfsd2a is a transporter for DHA. Nature 2014. PubMed
• Guemez-Gamboa A, et al. MFSD2A mutations and microcephaly. Nat Genet 2015. PubMed
• Sugasini D, et al. Dietary LPC-DHA and brain DHA enrichment. Sci Rep 2017. Scientific Reports
• Ly J, et al. Circulating phosphatidylcholines as surrogate biomarkers of the omega-3 index. J Lipid Res 2023. Journal of Lipid Research
• Domenichiello AF, Chen CT, Trepanier MO, Stavro PM, Bazinet RP. Whole body synthesis rates of DHA from alpha-linolenic acid are greater than brain DHA accretion and uptake rates in adult rats. J Lipid Res 2014;55:62-74. DOI

The Omega-3 Index and dose-response

• Harris WS, Del Gobbo L, Tintle NL. The Omega-3 Index and CHD mortality. Atherosclerosis 2017. PubMed
• Harris WS, Tintle NL, Etherton MR, Vasan RS. Erythrocyte long-chain omega-3 fatty acid levels are inversely associated with mortality and with incident cardiovascular disease: the Framingham Heart Study. J Clin Lipidol 2018;12:718-727 (corrigendum 2020;14:740). DOI
• Flock MR, et al. Determinants of erythrocyte omega-3 content, dose-response RCT. J Am Heart Assoc 2013. JAHA
• Walker RE, et al. Predicting the effects of supplemental EPA and DHA on the omega-3 index. Am J Clin Nutr 2019. AJCN
• Sarter B, et al. Blood DHA and EPA in vegans, and effects of an algal supplement. Clin Nutr 2015. PubMed

Supplement form and bioavailability

• Schuchardt JP, et al. EPA and DHA incorporation from fish oil vs krill oil. Lipids Health Dis 2011. PMC
• Arterburn LM, Oken HA, Bailey Hall E, Hamersley J, Kuratko CN, Hoffman JP. Algal-oil capsules and cooked salmon: nutritionally equivalent sources of docosahexaenoic acid. J Am Diet Assoc 2008;108:1204-1209. DOI

Tissue distribution and roles

• SanGiovanni JP, Chew EY. Omega-3 in the retina. Prog Retin Eye Res (review). PMC
• Lacombe RJS, Bazinet RP. DHA in the brain (structure and function review). 2019. ScienceDirect

Autopsy and the DHA-Alzheimer’s puzzle

• Cunnane SC, et al. Plasma and brain fatty acids in Alzheimer’s disease. 2012/2013. PubMed
• Igarashi M, et al. Disturbed choline plasmalogen and phospholipid fatty acid concentrations in Alzheimer’s disease prefrontal cortex. J Alzheimers Dis 2011. DOI
• Fraser T, et al. Fatty acid composition of frontal, temporal and parietal cortex in Alzheimer’s. 2010. PubMed

Liver export, LPC-DHA assembly, and the omega-6 substitute

• Akesson B, Fehling C, Jagerstad M. Effect of vitamin B12 deficiency on phosphatidylethanolamine methylation in rat liver. Br J Nutr 1978. DOI
• DeLong CJ, Shen YJ, Thomas MJ, Cui Z. Molecular distinction of phosphatidylcholine synthesis between the CDP-choline pathway and phosphatidylethanolamine methylation pathway. J Biol Chem 1999. PubMed
• Watkins SM, Zhu X, Zeisel SH. PEMT activity and dietary choline regulate liver-plasma lipid flux and essential fatty acid metabolism in mice. J Nutr 2003. PubMed
• Sugasini D, Yang P, Ng D, et al. Potential role of hepatic lipase in the accretion of docosahexaenoic acid by the brain. BBA Mol Cell Biol Lipids 2021. PubMed Central
• Lim SY, Hoshiba J, Moriguchi T, Salem N Jr. n-6 docosapentaenoic acid substitution for DHA leads to a loss in spatial task performance. J Neurochem 2005. PubMed
• Moriguchi T, Loewke J, Garrison M, Catalan JN, Salem N Jr. Reversal of docosahexaenoic acid deficiency in the rat brain, retina, liver, and serum. J Lipid Res 2001. Journal of Lipid Research
• Albert CM, Campos H, Stampfer MJ, et al. Blood levels of long-chain n-3 fatty acids and the risk of sudden death. N Engl J Med 2002. NEJM
• Lobo GP, Hessel S, Eichinger A, et al. ISX is a retinoic acid-sensitive gatekeeper that controls intestinal beta-carotene absorption and vitamin A production. FASEB J 2010. DOI
• Ganz T. Systemic iron homeostasis. Physiol Rev 2013. PubMed

Health authority positions

• World Health Organization. Risk reduction of cognitive decline and dementia: WHO guidelines, second edition. Geneva: WHO; 2026. ISBN 978-92-4-012355-7.
• Yu JT, et al. Evidence-based prevention of Alzheimer’s disease: systematic review and meta-analysis of 243 observational prospective studies and 153 randomised controlled trials. J Neurol Neurosurg Psychiatry 2020. DOI
• Alzheimer’s Association. Alternative treatments: omega-3 fatty acids. alz.org. Alzheimer’s Association

Vegan and vegetarian brain and cognition studies

• Sanders TAB. DHA status of vegetarians. Prostaglandins Leukot Essent Fatty Acids 2009;81:137-141. DOI
• Giem P, Beeson WL, Fraser GE. The incidence of dementia and intake of animal products: preliminary findings from the Adventist Health Study. Neuroepidemiology 1993;12:28-36. PubMed
• Zamroziewicz MK, Paul EJ, Zwilling CE, Barbey AK. Determinants of fluid intelligence in healthy aging: omega-3 polyunsaturated fatty acid status and frontoparietal cortex structure. Nutr Neurosci 2017. DOI
• Virtanen JK, et al. Circulating omega-3 polyunsaturated fatty acids and subclinical brain abnormalities on MRI in older adults: the Cardiovascular Health Study. J Am Heart Assoc 2013. DOI

Practitioner positions

• Norris J. Omega-3s. VeganHealth.org. VeganHealth.org
• Greger M. Optimum Nutrition Recommendations. NutritionFacts.org. NutritionFacts
• Greger M. Should Vegans Take DHA to Preserve Brain Function? NutritionFacts
• Fuhrman J. Why Vegans Need DHA and EPA. DrFuhrman.com 2026. DrFuhrman.com
• Radak T. Essential fatty acids, diet, and lifestyle factors influencing brain health. radaktim.wixsite.com
• Patrick RP. Omega-3 fatty acids (topic overview). FoundMyFitness. FoundMyFitness
• Patrick RP. Role of phosphatidylcholine-DHA in preventing APOE4-associated Alzheimer’s disease. FASEB J 2019. PubMed
• Sherzai AZ, Sherzai AN, Sherzai D. A systematic review of omega-3 consumption and neuroprotective cognitive outcomes. Am J Lifestyle Med 2023. PubMed
• Sherzai D, Sherzai A. Omega-3s and Brain Health (interview). The Proof with Simon Hill. theproof.com
• Hill S. Omega-3 vs Omega-6: The Ultimate Guide to Essential Fatty Acids (masterclass, with Calder, Fuhrman, the Sherzais, Fraser, Harris). The Proof. theproof.com
• Calder PC, Innes JK. Marine omega-3 fatty acids for cardiovascular health: an update for 2020. Int J Mol Sci 2020. PubMed

Retroconversion, feedback inhibition, and the Palmer claims

• Metherel AH, Irfan M, Klingel SL, Mutch DM, Bazinet RP. Compound-specific isotope analysis reveals no retroconversion of DHA to EPA but substantial conversion of EPA to DHA following supplementation. Am J Clin Nutr 2019. PubMed
• Metherel AH, Bazinet RP. Dietary DHA downregulates liver DHA synthesis by inhibiting eicosapentaenoic acid elongation. J Lipid Res 2024. PubMed
• Ference BA, Ginsberg HN, Graham I, et al. Low-density lipoproteins cause atherosclerotic cardiovascular disease. 1. Evidence from genetic, epidemiologic, and clinical studies. A consensus statement from the European Atherosclerosis Society Consensus Panel. Eur Heart J 2017;38(32):2459-2472. PubMed
• Harris WS, Schmitt TL. Unexpected similarity in RBC DHA and AA levels between bottlenose dolphins and humans. PLEFA 2014. PubMed
• Bito T, et al. Production of cyanocobalamin-enriched lettuce grown using hydroponics. J Agric Food Chem 2013. PubMed
• Nagra M, Palmer G. Omega-3 Supplements: Helpful or Harmful? (debate). YouTube
• Peskin BS. Why fish oil fails. J Lipids 2014. RETRACTED by the journal over an undeclared competing interest. This is Palmer’s oxidation citation. PubMed
• Retraction notice: Retracted: Why Fish Oil Fails. J Lipids 2014. PubMed Central

The omega-3 index threshold and brain-imaging studies

• Tan ZS, Harris WS, Beiser AS, et al. Red blood cell omega-3 fatty acid levels and markers of accelerated brain aging. Neurology 2012. PubMed
• Satizabal CL, Himali JJ, Beiser AS, et al. RBC omega-3 fatty acids, MRI markers, and cognition in midlife: Framingham. Neurology 2022. PubMed
• Courchesne E, Chisum HJ, Townsend J, et al. Normal brain development and aging: quantitative MR imaging. Radiology 2000. PubMed
• Harris WS, von Schacky C. The Omega-3 Index: a new risk factor for death from coronary heart disease? Prev Med 2004. PubMed
• Harris WS, Pottala JV, Varvel SA, et al. Erythrocyte omega-3 increases and linoleic acid decreases with age: 160,000 patients. PLEFA 2013. PubMed
• Benton D, Donohoe RT, Clayton DE, Long SJ. Supplementation with DHA and psychological functioning of young adults. Br J Nutr 2013. PubMed
• Stonehouse W, Conlon CA, Podd J, et al. DHA supplementation improved memory and reaction time in healthy young adults. Am J Clin Nutr 2013. PubMed

Related reading on VeganLinked

• Vegan nutrition overview
• Essential nutrients on a vegan diet
• Vitamin B12
• Dietary fat
• Cholesterol
• Seed oils
• The Adventist Health Studies
• What the longest-living people eat
• Are the Blue Zones debunked?
• Saul Newman vs the Blue Zones, a fact-check
• Plant toxins and antinutrients
• Raw diets
• Losing weight on a vegan diet
• Raising children vegan
• Crop deaths
• The pesticide argument
• Livestock and food we cannot eat
• Regenerative farming
• Rewilding and regenerative claims
• Coconut oil vs animal fat
• Meat and bowel cancer

Last updated: Tuesday, October 6, 2026, 12:53 PM UTC.