Vitamin D and Cognition: What the VitaMIND Trial Found
A 620-person, 24-month randomised trial tested whether correcting mild to moderate vitamin D deficiency improves cognition in older adults. It found no benefit on any measure, cognitive, functional, or wellbeing.
This article is for educational and research purposes only. It does not constitute medical advice. Always consult a qualified healthcare professional regarding any health decisions.
Vitamin D deficiency has been linked to worse cognition in observational studies for over a decade, and vitamin D receptors are distributed widely enough through the brain that a protective mechanism has always been biologically plausible. What has been missing is a trial large and long enough to test whether correcting a common, mild deficiency actually changes cognitive outcomes, rather than simply correlating with them. A UK trial published in 2025 in the Journal of the American Medical Directors Association set out to close that gap, and its answer is a clean, well-powered no.
1. Why This Question Needed a Dedicated Trial
Severe vitamin D deficiency is already treated as a matter of course, for bone health and other established risks, and the evidence linking it to cognitive decline is strong enough that a trial deliberately withholding treatment from severely deficient people would be difficult to justify ethically. Mild to moderate deficiency, defined in this trial as a blood level under 50 nmol/L of 25-hydroxyvitamin D, sits in murkier territory. It is far more common (the paper cites 40.4% of UK adults meeting this threshold, rising further in winter), it usually goes undetected and untreated, and prior evidence for a cognitive benefit from correcting it was inconsistent: a systematic review of four earlier trials the VitaMIND authors cite found inconclusive results, but those trials ran for as little as six months or used combined multi-nutrient formulations that made vitamin D's specific contribution hard to isolate.
That gap, whether treating this milder, more prevalent form of deficiency changes cognitive trajectory, is what the VitaMIND trial (vitamin D for cognition) was built to answer directly, with a single-nutrient intervention, a long follow-up, and a large sample (Corbett A, Taylor R, Llewellyn D, et al. "Impact of Vitamin D Supplementation on Cognition in Adults With Mild to Moderate Vitamin D Deficiency: Outcomes From the VitaMIND Randomized Controlled Trial." Journal of the American Medical Directors Association 2025;26(8):105711).
2. Trial Design
VitaMIND was a two-arm, parallel, double-blind, 24-month randomised controlled trial run entirely remotely through the UK's PROTECT-UK online ageing cohort, itself a large pre-existing platform for online cognitive assessment. Adults over 50 with no dementia diagnosis were screened using a validated predictive algorithm for vitamin D deficiency risk, then randomised to either 4000 IU of vitamin D3 daily (the highest publicly available over-the-counter dose) or an identical placebo capsule.
A total of 620 participants were randomised, 310 to each arm, recruited between September 2020 and March 2021. All were required to meet criteria for age-associated cognitive decline, performing at least one standard deviation below the norm on at least one test in the PROTECT-UK cognitive battery at baseline. Mean age was 60.7 years and 75% of participants were female. In a subgroup of 49 participants who provided blood samples, median baseline vitamin D was 42 nmol/L, confirming the cohort genuinely sat in the mild to moderate deficiency range the trial targeted, not the more severe range where benefit is already well established.
The primary outcome was executive function and task switching, measured by a computerised version of the Trail Making Test Part B. Secondary outcomes covered spatial and numerical working memory, verbal reasoning, function (instrumental activities of daily living), behaviour, and wellbeing, all assessed at baseline, 6, 12, and 24 months.
3. Results: No Benefit, on Anything
At 24 months, vitamin D supplementation showed no significant benefit over placebo on the primary outcome (Cohen's d effect size 0.11, p=0.16), nor on any secondary measure: numerical working memory (p=0.38), spatial working memory (Paired Associate Learning p=0.84, Self-Ordered Search p=0.64), verbal reasoning (p=0.92), Switching Stroop (p=0.86), instrumental activities of daily living (p=0.86), behaviour (p=0.65), or wellbeing (p=0.92).
The null result held up under scrutiny rather than resting on a single analysis. Pre-specified subgroup analyses by age, sex, and deficiency severity found no significant treatment effect in any subgroup on the primary outcome. A per-protocol analysis restricted to participants with confirmed tablet-taking adherence found essentially the same null result (mean difference 6174, 95% CI -17,095 to 4747, p=0.266). Adherence itself was not in question: median compliance was 98% among participants with detailed tablet-count data, and blood testing in a subsample confirmed vitamin D status improved significantly in the treatment group by 24 months (p under 0.001), meaning the trial delivered its intervention successfully and still found nothing. No treatment-related serious adverse events were reported.
The authors' own framing is direct: "This study provides robust evidence that vitamin D supplementation does not result in cognitive benefits in individuals with mild to moderate deficiency," while noting that "supplementation may be of value for other clinical indications and in people with severe deficiency."
4. What This Does and Doesn't Rule Out
This trial specifically addresses mild to moderate deficiency (25 to 50 nmol/L) in adults already showing early, subclinical cognitive decline. It does not test, and does not contradict, the separate and better-established case for correcting severe deficiency (under 25 nmol/L), which the authors explicitly excluded as insufficiently represented in this cohort and note was not the trial's research question. It also does not speak to vitamin D's other roles, in bone health, muscle function, or mood, which rest on different evidence bases entirely.
The trial's population, recruited through an existing online research cohort, skewed toward White, female, and highly educated participants who were already digitally engaged, a limitation the authors acknowledge directly and one that means the findings may not generalise cleanly to the full UK population, particularly given known variation in vitamin D metabolism across ethnic groups. The 24-month duration is also finite. The authors raise, without much confidence, the possibility that a longer trial might show something different, while noting this would raise separate practical questions about the feasibility of decades-long public health supplementation programmes.
Within those bounds, the result is about as clean a null as trial design allows: adequately powered, confirmed biological delivery of the intervention, and a p-value nowhere near significance on the primary outcome or any of the ten-plus secondary measures tested.
5. Part of a Broader Pattern
This result sits alongside a growing list of single-nutrient supplementation trials that have failed to translate a plausible mechanism and consistent observational association into a measurable cognitive benefit in the specific population and duration tested. Our review of the 2026 DHA target-engagement trial found the same pattern for omega-3 fatty acids: confirmed biological delivery, and no cognitive or structural benefit over 24 months. Our broader Alzheimer's prevention research review covers why modifiable-risk-factor interventions with a clearer behavioural mechanism, like the FINGER trial's multidomain lifestyle approach, have generally fared better in trials than single-nutrient supplementation aimed at a population without severe, clinically established deficiency.
None of this means vitamin D is irrelevant to brain ageing as a biological system; receptor distribution and mechanistic plausibility are not in dispute. It does mean the specific, narrower claim, that correcting a common mild deficiency will measurably slow cognitive decline, no longer has trial evidence behind it, at least not at this dose, in this population, over this duration.
6. Summary
The VitaMIND trial randomised 620 adults with mild to moderate vitamin D deficiency and early cognitive decline to 4000 IU daily vitamin D3 or placebo for 24 months. Vitamin D levels rose as expected in the treatment group, confirming the intervention was delivered, but no significant benefit emerged on the primary outcome of executive function or on any of the secondary measures of memory, reasoning, function, behaviour, or wellbeing, including in pre-specified subgroup and per-protocol analyses. The trial does not address severe deficiency, where treatment is already standard practice for other reasons, but it provides robust evidence against including vitamin D supplementation in cognitive-decline risk-reduction guidance for people with milder, more common deficiency. It adds to a pattern, also seen in a recent high-dose DHA trial, of biologically plausible single-nutrient interventions confirming target delivery while still failing to move cognitive outcomes in rigorously controlled, adequately powered trials.
Primary source: Corbett A, Taylor R, Llewellyn D, Ranson JM, Hampshire A, Pickering E, Palmer A, Aarsland D, Cader D, Frost D, Ballard C. "Impact of Vitamin D Supplementation on Cognition in Adults With Mild to Moderate Vitamin D Deficiency: Outcomes From the VitaMIND Randomized Controlled Trial." Journal of the American Medical Directors Association 2025;26(8):105711. DOI: 10.1016/j.jamda.2025.105711. PMID: 40480279. Trial registration: ISRCTN79265514. This article summarises a single published trial and is not a substitute for individualised medical advice.