Results
Omega-3 supplementation was associated with significantly accelerated cognitive decline, as evidenced by a faster decrease in MMSE scores (β = -0.266, p < 0.001) and a faster increase in both ADAS-Cog13 (β = 0.823, p < 0.001) and CDR-SB scores (β = 0.205, p < 0.001). This association was not mediated by Aβ deposition, tau pathology, or gray matter atrophy. Instead, longitudinal FDG hypometabolism within AD-vulnerable regions served as a significant mediating pathway, accounting for 30.8%, 40.8%, and 19.0% of the total effect on the decline in MMSE, ADAS-Cog13, and CDR-SB, respectively.
Conclusions
Omega-3 supplementation may be associated with accelerated cognitive decline in older adults, potentially through adverse effects on cerebral synaptic function rather than classical AD proteinopathies. These findings challenge the prevailing view of omega-3 as uniformly beneficial and highlight the need for a cautious reassessment of its widespread use for cognitive protection.
ChatGPT:
Journal: serious / legitimate (Impact Factor 7.8 and CiteScore 10.0).
Research team: real AD-neurology group with relevant publication footprint, but not a world-leading household-name omega-3 trial group.
Paper: credible enough to take seriously, but the result should still be treated as an observational signal, not as causal proof.
Main takeaway: interesting signal, but I would not treat it as proof that omega-3 causes cognitive decline.
Reasons:
Confounding by indication is a big concern.
People who choose omega-3 supplements may already be more worried about memory, have subtle symptoms, have vascular/metabolic issues, or be following advice after health deterioration. The authors say they found “no pre-existing differences in cognitive or FDG-SUVR trajectories before initiation,” but that does not fully eliminate residual confounding.
ADNI is not a general-population cohort.
ADNI is enriched for people at risk of Alzheimer’s disease, with detailed imaging and cognitive follow-up. Results may not generalize to healthy 40–70-year-olds taking modest fish oil.
“Omega-3 supplementation” is probably heterogeneous.
Observational datasets often lack precise information on EPA/DHA dose, brand, oxidation quality, duration, adherence, baseline omega-3 index, fish intake, indication, and co-supplements. Those matter a lot.
The result conflicts with much of the broader prior expectation, but not cleanly with RCT evidence.
Prior randomized evidence has generally been weak/neutral rather than strongly positive. Cochrane found no cognitive benefit for omega-3 in dementia over six months, and a Nutrition Reviews meta-analysis in non-demented adults found no effect on global cognition but a mild memory benefit.
Dietary omega-3 and supplements should be separated.
Observational studies of dietary fish/omega-3 often look more favorable than supplement trials. A 2023 review found dietary intake and biomarkers were associated with lower dementia/cognitive-decline risk, while supplement RCT evidence remains much less convincing.
These results are quite surprising, especially considering the recent ITP results. I’m starting to worry that my supplement stack might be doing more harm than good.
What is your personal opinion regarding EPA? Is it still worth it to keep it on my stack despite reta and pitava? You are pretty much the Omega 3 expert
Based on the evidence I found, DHA supplementation seems detrimental.
For EPA-only supplementation, the strongest evidence we have is REDUCE-IT 4g/day: 25% risk reduction https://jamanetwork.com/journals/jama/fullarticle/2773120 (btw, the STRENGTH trial added DHA to EPA and found no benefits. So DHA might have erased EPA’s benefits.)
JELIS 1.8 g/day found a 19% risk reduction, consistent with REDUCE-IT.
For depression, we also have some positive signals in meta-reviews.
So I guess, if you already have apoB, blood pressure, and glucose under control, then adding up to 4 g/day EPA-only (Vazkepa/Vascepa) can be good? I don’t take it because the evidence isn’t super strong (some people criticize REDUCE-IT for the placebo chosen, also those were very stick population, what is it like in healthy young people?) and I prefer to limit my stack.
Another point: REDUCE-IT looked at Tg 135–499 mg/dL; LDL 41–100 mg/dL and EVAPORATE Tg 135–499 mg/dL; LDL >40 to ≤115 mg/dL.
Are EPA’s benefits replicable to people with normal Tg?
JELIS was total cholesterol ≥250 mg/dL. But it was open-label (not double blind), without placebo, with low-intensity statins, in Japanese people (high-fish-intake population).
Conclusion: if you have Tg > 135 mg/dL, 4 g/day Vazkepa makes sense. If Tg < 135 mg/dL and low apoB: unclear…
