Exercise vs Supplements for Brain Health: An Honest Comparison

We've evaluated two supplements with real cognitive evidence behind them: creatine and omega-3. Creatine's evidence is why we carried it for as long as we did, though it's since been discontinued from the Proco range; omega-3 remains part of the range today. In the interest of the same honesty we apply to every ingredient we evaluate, it's worth asking how that evidence stacks up against the single highest-evidence intervention for brain health: exercise. This isn't a marketing question for us — it's the question we'd want answered before spending money on a bottle instead of a pair of running shoes.

It's also a question most supplement companies never ask, because the honest answer complicates the pitch. We'd rather complicate the pitch than overstate what a capsule can do. So this article walks through what the exercise literature actually shows, how it was measured, how it stacks up next to the creatine and omega-3 evidence we've covered elsewhere, and what happens in the one trial that tested combining exercise with a supplement directly.

A note on availability: Proco discontinued its own Creatine Monohydrate in September 2026 and no longer sells it. The research in this article stands on its own — we just aren't the people selling you the product any more.

What exercise actually does, measured properly

A 2026 systematic review and three-level meta-analysis pooled 50 randomized controlled trials in 4,826 cognitively healthy older adults, comparing aerobic exercise and resistance training against controls.¹ The overall effect on executive function was real but modest (g=0.155, 95% CI 0.084–0.226, p<0.001), with working memory the strongest of the reported subdomains (g=0.208, 95% CI 0.093–0.322) and cognitive flexibility just clearing significance (g=0.150, 95% CI 0.001–0.299, p=0.049). Looking at exercise parameters one at a time, significant benefits appeared for a frequency of at least three sessions a week (g=0.188), a programme duration of 26 weeks or less (g=0.193), sessions of 60 minutes or less (g=0.190), and moderate-intensity exercise (g=0.254, 95% CI 0.117–0.390). Both aerobic training (g=0.130) and resistance training (g=0.232) worked on their own.

Two honest caveats belong with those numbers. Those parameter figures are separate subgroup analyses, not a recipe that stacks: the moderate-intensity subgroup's g=0.254 is the largest single figure in the set, but the review did not test moderate intensity against higher intensity head-to-head, so it cannot tell you that training harder is worse for your brain. And Egger's test indicated potential publication bias (t=2.630, p=0.009), which is a real reason to treat the pooled estimate as an upper bound rather than a floor. The authors rated the overall certainty of the evidence as moderate under GRADE — which, for this field, is unusually good.

What the dataset does support is more useful than a ranking anyway: benefits showed up consistently across every population subgroup examined — 60–69-year-olds and over-70s, more and less educated participants, mostly-female and mostly-male trials — and the largest subgroup effect was in people with 12 or fewer years of education (g=0.322). For someone deciding how to spend a limited number of weekly hours, "three sessions a week, an hour or less, at an intensity you can sustain" is a specific, evidence-consistent answer. It's also a very different kind of claim than "take more of this ingredient," which is usually where supplement dosing advice ends up.

The clearest evidence: exercise changes brain structure, not just test scores

Meta-analytic effect sizes are useful, but one older trial makes the mechanism concrete. In a year-long randomized controlled trial, 120 sedentary older adults were assigned to either a walking program (40 minutes, three days a week) or a stretching-and-toning control group.² The walking group's hippocampus — the brain region most associated with memory and most vulnerable to age-related shrinkage — grew by 2.12% on the left side and 1.97% on the right over the year. The stretching group's hippocampus shrank by roughly 1.4% over the same period, which is the typical rate of age-related decline; the paper's own framing is that exercise training effectively reversed one to two years of volume loss. The volume gains tracked with increases in serum brain-derived neurotrophic factor (BDNF), a protein involved in forming new neural connections. Caudate nucleus and thalamus volumes were unaffected, which is itself informative: the effect was regionally specific rather than a general imaging artifact.

That's a genuinely different kind of evidence than anything a supplement trial has produced for cognition: a structural brain change, in a controlled comparison, tied to a plausible biological mechanism. It's also why exercise anchors every serious clinical recommendation for healthy brain aging, while supplements sit further down the list as adjuncts — a hierarchy we go through in what the research actually shows on modifiable dementia risk and in the 45% figure, and where to start.

It's worth being precise about what this trial does and doesn't establish. It's one study, in 120 people, using MRI-measured hippocampal volume as a proxy for brain health rather than a hard clinical outcome like a dementia diagnosis. Effect sizes on the actual memory tests in that trial were modest — the reported cognitive gain was in spatial memory specifically — and not every subsequent exercise-and-brain-imaging study has replicated the volume finding at the same magnitude. But the direction has shown up across multiple imaging studies and reviews since, which is exactly the kind of convergent, mechanistically-grounded evidence that's largely absent from the supplement literature on cognition.

How that compares to the evidence behind our own products

Our creatine and cognition guide covers a 2024 meta-analysis of 16 RCTs (492 participants) that found significant effects on memory (SMD 0.31, 95% CI 0.18–0.44), attention time and processing speed — but no significant effect on overall cognitive function or executive function, and GRADE certainty was moderate for memory alone and low for everything else.³ Our omega-3 and cognition guide covers a 2025 dose-response meta-analysis of 58 trials, which estimated change in cognition per 2,000mg/day increment of omega-3 and identified 1,000–2,500mg/day as the range associated with the most consistent effects; certainty was low to moderate depending on the domain, and global cognition followed a curve that turned downward at the highest doses.⁴ Read side by side with the exercise data, the honest pattern is: exercise's effect size sits in a similar small-to-moderate range as the best-evidenced supplements — it isn't obviously larger on a cognitive test score — but it's supported by a bigger, more consistent evidence base, it has direct structural evidence behind it, and it comes with cardiovascular and metabolic benefits no supplement replicates.

Does adding a supplement on top of exercise help?

This is the question we actually care about: if you're already exercising, does layering a supplement on top move the needle further? A 24-week randomized, single-masked, placebo-controlled trial in non-demented older Japanese adults tested exactly that.⁵ Participants were assigned to no exercise plus placebo, exercise (150 minutes/week of resistance and aerobic training) plus placebo, or the same exercise program plus a long-chain polyunsaturated fatty acid supplement — DHA 300mg, EPA 100mg and arachidonic acid 120mg, 520mg/day in total. In the per-protocol analysis (n=76), there were no significant differences between the exercise groups and the control group on any of the neuropsychological tests. That's a broader null than "the supplement added nothing": in this trial, neither exercise arm separated from control on the cognitive measures. The one exception was a subgroup of participants whose skeletal muscle mass index fell at the sarcopenia cut-off, where the exercise-plus-supplement group showed better selective attention and working memory than controls — a plausible signal, but a subgroup finding in a small trial, not the headline result.

We think that result deserves more attention than it usually gets, because it's the closest thing to a direct test of the claim this whole article is built around. It isn't a case for skipping omega-3 — sarcopenia risk is common enough in older adults that the subgroup effect is clinically relevant, and the supplement's other benefits aren't in question here. But it also isn't the clean "exercise works, supplement adds nothing" story it's sometimes summarised as. It's a small trial that didn't detect much from anything, which is its own useful lesson about how hard these effects are to measure over six months.

Bottom line

If you only have the budget or willpower for one intervention, the evidence says start with training, not a bottle.

None of this is an argument against creatine or omega-3 — the trial evidence for both is real, if narrower than the headlines suggest, which is why we carried creatine for as long as we did and why omega-3 remains in the range today. It's an argument against treating either as a substitute for the thing with the strongest, most consistent evidence in this entire category: a structured exercise routine that, in at least one controlled trial, physically grew the brain region most tied to memory. Exercise is the intervention with a plausible mechanism (BDNF and structural growth) and a much larger combined trial base than any single supplement can claim. Supplements are the reasonable next layer on top — not the substitute underneath. The same reasoning is why we don't stock most of what gets marketed for cognition at all, which we set out in why we don't sell brain blends.

If we're being consistent with how we evaluate every ingredient we sell, we have to apply the same scrutiny to the intervention that isn't a product. Exercise doesn't have a company behind it funding awareness campaigns, so it rarely gets framed as "the highest-evidence option" the way a new supplement launch does — but on the numbers, it is. Our practical recommendation hasn't changed: get the training in first, at a moderate, sustainable intensity, three or more times a week, in sessions of an hour or less. Layer omega-3 on top if the specific evidence for your situation — low dietary intake, a documented gap — makes that a reasonable next step, not as a way to skip the harder part. For a different outcome and a different leaderboard, see what the trials show for exercise and depression and which type of exercise actually helps anxiety — the best exercise for mental health depends on which outcome you're targeting.

A note on our own dose, stated plainly: Proco Omega-3 provides 180mg EPA and 120mg DHA per softgel, and at the label's suggested use of one softgel twice daily that comes to 600mg of combined EPA and DHA a day. The dose-response meta-analysis cited above estimated effects per 2,000mg/day increment and identified 1,000–2,500mg/day as the range with the most consistent signal. Our serving sits below that. It is a fully disclosed maintenance-range dose — and it is not "dosed to the studied range" for cognition, which is what an earlier version of this page claimed. Our EPA/DHA dosage guide works through what different intakes actually correspond to.

This article is for informational purposes only and has not been evaluated by the FDA. It is not intended to diagnose, treat, cure, or prevent any disease. Speak with a healthcare provider before starting any new supplement, especially if you take medication or have an existing health condition.

References

1. Effects of exercise on executive function in cognitively healthy older adults: a systematic review and three-level meta-analysis. BMC Public Health, 2026;26:1762. link.springer.com
2. Erickson KI, et al. Exercise training increases size of hippocampus and improves memory. PNAS, 2011;108(7):3017–3022. pmc.ncbi.nlm.nih.gov
3. Xu C, Bi S, Zhang W, Luo L. The effects of creatine supplementation on cognitive function in adults: a systematic review and meta-analysis. Frontiers in Nutrition, 2024;11:1424972. ncbi.nlm.nih.gov
4. A systematic review and dose-response meta-analysis of omega-3 supplementation on cognitive function. Scientific Reports, 2025;15:30610. nature.com
5. Effects of combining exercise with long-chain polyunsaturated fatty acid supplementation on cognitive function in the elderly: a randomised controlled trial. Scientific Reports, 2020;10:12906. nature.com

Proco Omega-3 is a fully disclosed 180mg EPA + 120mg DHA per softgel — a reasonable layer on top of training, not a replacement for it, and not dosed to match the cognition trials. See the range →