Vitamin D at 39, Brain at 55: The Framingham Signal

Most people don’t think about dementia at thirty-nine. That’s the right age to not think about it — the body still works, memory still complies, and Alzheimer’s feels like something from a distant future. But a new paper out of the University of Galway argues that thirty-nine is precisely when a lot gets decided. Emer McGrath’s team went back to nearly 800 participants of the Framingham Heart Study who had their vitamin D measured in midlife — and then, sixteen years later, scanned their brains with a PET tracer for tau protein. The people whose midlife vitamin D had been higher now had less tau in exactly the regions that Alzheimer’s attacks first. It’s one of those findings you want to shout from a rooftop and quietly tuck into a footnote at the same time.

Why Sixteen Years Is the Right Window

Illustration. Source: Unsplash

Almost every earlier study of vitamin D and cognitive health was done on people aged 65 and up — an age where the disease may already be quietly forming. That gave tangled results: was low vitamin D a cause? Or was aging itself dropping vitamin D? Or was a silently approaching disease making people less active, less sunlit, less nourished? McGrath’s team came at the question from the other end. They took the starting point in midlife — around 39 years old — and then waited. Sixteen years. That’s enough time for the first molecular traces of Alzheimer’s to appear in some of the subjects, while most remain clinically symptom-free.

The cohort: about 800 people from the Framingham Heart Study — the epidemiological project that started in a small Massachusetts town in 1948 and became the world’s main laboratory for cardiovascular and cognitive aging research. Roughly 15% of this subgroup had Irish ancestry, which is how the University of Galway team ended up involved. Every participant was dementia-free at baseline.

What did they measure sixteen years later? Two biomarkers that are now considered the most reliable early signals of Alzheimer’s, long before symptoms appear:

Tau protein — a structural protein that normally helps maintain the shape of neurons and their projections. In Alzheimer’s disease, tau changes conformation, clumps into tangles, and poisons the cell. Tau accumulation in specific brain regions is one of the earliest visible changes of the disease.

Amyloid beta — the other key marker. These are protein fragments that deposit as plaques between neurons. For decades amyloid was considered the primary culprit in Alzheimer’s; today that hypothesis is contested, but as an early marker it’s still valuable.

A PET scanner with a radioactive tracer lets you see exactly how much tau and amyloid have accumulated in specific brain regions — primarily the entorhinal cortex and hippocampus, which are hit first.

What They Found

The main result: people whose midlife vitamin D was above 30 ng/mL had less tau in their brains sixteen years later than those below that threshold. The association held after adjusting for age, sex, and depression symptoms. It wasn’t a smooth «more is better» curve — the threshold was essentially binary, and pushing above 30 ng/mL didn’t buy additional benefit. The effect on amyloid-beta was weaker and less consistent.

The press release doesn’t include specific correlation coefficients, p-values, or effect sizes — for those you need the full paper in Neurology Open Access. But the researchers did something many of their peers in similar stories skip: they immediately flagged that this is not a causal result.

«The study does not prove that vitamin D levels lower the level of tau in the brain or the risk of dementia; it only shows an association, ” the University of Galway press release states.

That sentence deserves unpacking — because right next to it sit two other studies that seem to say the opposite.

The Problem: RCTs Don’t See What Observation Sees

The first is D-Health, a randomized trial published in 2023. Almost 21,000 older Australians, randomly assigned to get 60,000 IU of vitamin D once a month or placebo. Five-year follow-up, rigorous cognitive testing, Telephone Interview for Cognitive Status (TICS) as the primary measure. The result: a group difference of 0.04 points on TICS, with a confidence interval from −0.14 to +0.23. The cognitive odometer didn’t move. The odds ratio for cognitive impairment was exactly 1.00 (CI 0.75 to 1.33). In plain English: vitamin D supplements in older adults did precisely nothing.

The second is VITAL, the American trial of 25,000 people taking 2,000 IU of vitamin D3 daily (or placebo) for 5.3 years. Global cognitive function — no significant difference between groups. Same zero.

How can both be true? The answer is the time window. Both D-Health and VITAL recruited people averaging over 65. In many of them, disease processes were already underway at the molecular level. Vitamin D, even if it really does protect the brain, can’t reverse time: if tau is already clumping in the entorhinal cortex, supplements can’t sweep it away. At best they might slow further deterioration — an effect too subtle for a 5-year RCT using cognitive questionnaires to catch.

McGrath’s paper formulates exactly this hypothesis: the critical window is midlife, not old age. By 65, much is already decided. At 39, it isn’t.

Why Vitamin D Could Plausibly Affect the Brain

The biology says the idea isn’t pulled out of thin air. Vitamin D receptors (VDRs) are widely expressed in neurons and glial cells. The hippocampus — the brain’s memory hub — is especially dense with them. In cell-culture experiments, vitamin D suppresses inflammatory responses by microglia, the brain’s immune «police, ” which damage neurons when chronically activated. It also regulates calcium transport, supports myelination, and, in some studies, reduces tau aggregation in neuronal cultures.

The epidemiology backs this up: earlier Framingham work (with overlap in the current author list) already showed an association between baseline vitamin D and dementia risk nine years later. And Mendelian randomization — a method that uses genetic variants to simulate a randomized trial — hints at a causal effect on cognitive function, but mainly in people with true deficiency (below 25 nmol/L). Above that threshold, more vitamin D changes little.

This reconciles the contradiction. Observational studies (including the new McGrath paper) are catching a real biological signal — but only in people who genuinely had a deficiency. The RCTs, recruiting mostly people with normal or near-normal vitamin D, catch nothing because their participants didn’t have a deficit to fix. It isn’t «nothing works» — it’s «supplementation works when you were short, and not otherwise.»

Critical Analysis: The Sober List of Caveats

The paper is published in Neurology Open Access, a peer-reviewed journal of the American Academy of Neurology. Not Nature, but not a grey preprint either. Cohort size and follow-up duration are its strengths.

The main methodological issue with this and similar studies is residual confounding. Vitamin D levels correlate with many other things: active lifestyle, more sun exposure, better diet, higher income, better access to healthcare. Each of those independently protects the brain. Even after adjusting for age, sex, and depression, a shadow of these unmeasured variables can linger. The open press release doesn’t detail the full adjustment set, and this is where a skeptic has to put a question mark.

The second issue is a single vitamin D measurement at baseline. Blood vitamin D shifts seasonally, yearly, with diet changes. One point is one snapshot. It may reflect a person’s central tendency — or it may be a random outlier. Over sixteen years of follow-up, one point inevitably loses fidelity: people live, age, change habits.

The third is that the Framingham cohort isn’t ethnically diverse. Most participants are white Americans, some with Irish ancestry. Vitamin D and Alzheimer’s may behave differently in populations with different baseline deficiency rates, different APOE genotype frequencies, or different annual sunlight exposure. Generalizing the finding to all humanity is premature.

Finally, APOE4. That’s the strongest known inherited risk factor for Alzheimer’s — carriers have 3 to 12 times higher risk. The press release doesn’t say whether results were stratified by APOE status. If they weren’t, that’s a notable gap — because vitamin D biology may work differently in carriers and non-carriers.

What This Means for a 39-Year-Old Reader

The answer is undramatic, but actionable. First, check your vitamin D level. It’s a simple blood test, inexpensive in most clinics. If you’re below 30 ng/mL (or 75 nmol/L), you have a clear target. Second, if you’re deficient, fix it — through three routes: more sun (15–20 minutes of bare skin exposure 3–4 times a week, latitude-adjusted), foods rich in vitamin D (fatty fish, eggs, fortified milk), or supplements (1,000–2,000 IU per day is a safe starting dose for most adults). Doses above 4,000 IU without medical supervision aren’t recommended — hypervitaminosis D is real and unpleasant.

Second — don’t expect miracles from pills if your level is already normal. The RCT data is unambiguous here. More vitamin D than you need doesn’t make you smarter and doesn’t protect the brain further. This isn’t a vitamin god you can appease with sacrifice. It’s a nutrient that needs to be present in sufficient amounts — no more.

McGrath’s paper adds an important detail to the picture: brain health strategy should start in midlife, not old age. What you do at 39 will or won’t bear fruit by 55. Thirty-nine isn’t old. It’s a window of opportunity closing slowly, but inevitably.

Frequently Asked Questions

Should I start taking vitamin D right now?

Measure first. If you’re deficient (below 30 ng/mL or 75 nmol/L), yes — correcting the deficit is justified and has evidence behind it. If your level is normal, supplements won’t give you anything extra, and that’s confirmed by RCTs on 46,000 people.

Why did D-Health and VITAL RCTs show no benefit if observational studies do?

Most likely because both trials enrolled people 65 and older, when the window for brain protection was already closing. McGrath’s work points to midlife as the critical window. RCTs on people aged 35–50 that would test this hypothesis directly haven’t been done yet.

What’s a «safe» blood vitamin D level?

Most specialists agree that 30 ng/mL (75 nmol/L) is a reasonable floor and 40–60 ng/mL is an optimal range. Above 100 ng/mL is excess that can cause harm. There’s no goal of «push it to the maximum.»

Is tau protein already an Alzheimer’s diagnosis?

No. Tau accumulation begins 10–20 years before clinical symptoms. Elevated tau on a PET scan doesn’t mean you have dementia — it’s a marker of risk and early pathology. Many people with «positive» tau scans live into deep old age without symptoms.

Tau and amyloid — which matters more?

That’s actively debated. The classical «amyloid hypothesis» dominated for thirty years, but drugs that remove amyloid produced only modest clinical effects. Today more evidence supports tau as the better correlate of functional decline. That’s why newer studies, including this one, focus on tau.