Does diet soda cause dementia? The honest answer

Short answer: Nobody knows — and the studies genuinely contradict each other. Some large studies link daily diet soda to a higher rate of dementia. The largest one, with the most cases, found no link at all. And in the newest study, the association disappeared once researchers set aside people who already had obesity or diabetes — which suggests diet soda may be a marker of metabolic problems more than a cause of brain problems. What the evidence does not support is the thing most people assume: that switching from regular soda to diet soda is a brain-health upgrade. There's no good evidence that it is. Water is the boring, low-risk answer.

This is educational, not medical advice. If you have diabetes, prediabetes, high blood pressure or are managing your weight, your beverage choices sit inside a bigger clinical picture — discuss changes with your clinician or dietitian rather than acting on a headline.

The four studies that matter

Every couple of years a study lands, a headline says diet soda triples your dementia risk, and millions of people look guiltily at the can on their desk. The headlines describe real research — but each one is reported as if it settled the question, when in fact each new study has pointed in a different direction. Here are the four studies the headlines are built on, with the caveats the coverage dropped. They are not the entire literature — other cohorts have reported findings in both directions — but they are the ones the public argument runs on.

2017 — Framingham: the study that started it

Researchers followed the Framingham Heart Study Offspring cohort and found that people who drank artificially sweetened soft drinks daily had roughly three times the rate of ischaemic stroke (hazard ratio 2.96, 95% CI 1.26–6.97) and of Alzheimer's disease dementia (hazard ratio 2.89, 95% CI 1.18–7.07) compared with people who drank none (Pase et al., Stroke, 2017).

Two details rarely make the coverage. The dementia analysis rested on 1,484 people, and a confidence interval running from 1.18 to 7.07 means "somewhere between slightly raised and sevenfold" — the study could not pin the number down. And sugar-sweetened drinks showed no association at all, which is strange: if sugary drinks are bad for the vascular system — and they are — the sugar arm should look worse, not cleaner. When the supposedly healthier product looks worse than the sugary one, the likely explanation isn't chemistry. It's who chooses which drink.

2025 — the largest analysis found nothing

In 2025, researchers pooled six major US cohort studies — the Health and Retirement Study, ARIC, the Chicago Health and Aging Project, the Rush Memory and Aging Project, and both Framingham cohorts — covering 10,974 older adults, 116,067 person-years and 2,445 actual cases of dementia. That is, by a wide margin, the largest case count anyone has brought to this question.

The result: nothing. Artificially sweetened beverages, hazard ratio 1.00 per serving per week (95% CI 0.99–1.01). Highest versus lowest consumption, hazard ratio 1.00 (0.83–1.21). Sugar-sweetened beverages were null too. The findings held across every subgroup the authors tested — age, sex, body mass index, and presence of chronic disease (Chen et al., JAMA Psychiatry, 2025).

One important limit, which the authors raise themselves: this study measured late-life intake, in people averaging 73 years old. It cannot tell you what drinking a litre of diet cola a day through your forties does. They explicitly call for research on earlier-life exposure.

2025 — a midlife study found faster cognitive decline

Which is roughly what the next study looked at. Researchers followed 12,772 Brazilian adults, average age 52, for about eight years, measuring intake of seven low- and no-calorie sweeteners: aspartame, saccharin, acesulfame-K, erythritol, xylitol, sorbitol and tagatose.

People in the highest intake group declined on global cognition about 62% faster than those in the lowest — the researchers translated that into roughly 1.6 extra years of brain ageing over the study period. The middle group declined about 35% faster. Tagatose was the one sweetener with no association (Gonçalves et al., Neurology, 2025).

Two subgroup findings make it murkier. The effect appeared only in people under 60 — above 60, nothing, which is the opposite of what you'd expect from a slow cumulative toxin. And it was stronger in people with diabetes, who are of course far more likely to be drinking artificially sweetened drinks in the first place. Note too what was measured: cognitive decline on tests, not diagnosed dementia.

2026 — the newest study, and the sentence buried inside it

In January 2026, the Northern Manhattan Study published the freshest data we have. It followed 947 dementia-free adults, average age 64, in a deliberately multi-ethnic New York cohort, for about a decade. Around one in five developed dementia.

Each additional diet soda per day was associated with a 39% higher rate of dementia (adjusted incidence rate ratio 1.39, 95% CI 1.09–1.75). People drinking more than one a day had a 4.15-fold higher rate than those drinking one or fewer (95% CI 1.81–9.49). Regular soda showed only a non-significant trend (Farhi et al., Journal of Alzheimer's Disease, 2026).

That 4.15 figure is what the headlines are built on. Here is what sits three paragraphs down in the same paper: the association was not apparent after excluding those with obesity or diabetes.

Take out the people who already had metabolic disease, and the diet-soda signal vanished. The study's senior author, Dr Hannah Gardener, framed the open question precisely: the research still has to work out whether diet soda is "a driver or a marker of underlying health risks," and whether the link is causal at all (University of Miami).

(A note on numbers, since you may see this study quoted elsewhere: the widely-repeated "3.91-fold" figure comes from the earlier conference abstract, based on a slightly different sample. The peer-reviewed paper reports 4.15. We've used the published one.)

So what's actually going on?

Line the four studies up and a pattern appears that has nothing to do with sweeteners:

The thread that keeps reappearing is not the drink. It is metabolic health. In the Northern Manhattan Study, excluding participants with obesity or diabetes made the association disappear; in the Brazilian cohort the association was stronger in people with diabetes. That is suggestive, not settled — only one of these studies tested the exclusion directly.

That points hard at what researchers call confounding by indication. People do not drink diet soda at random. They switch to it because they've gained weight, because they've been told to cut sugar, because they've been diagnosed with prediabetes or type 2 diabetes. Those conditions are established dementia risk factors in their own right, and they were there first. Statistical adjustment only goes so far: a person on their third diet cola of the day is telling you something about their health history that no questionnaire fully captures.

This does not mean diet soda is proven harmless. It means the honest state of the evidence is unresolved, and the strongest available reading is that the drink is at least as much a signal as a cause. Anyone telling you otherwise — in either direction — is ahead of the data.

The comparison most people actually care about

Most people asking this aren't choosing between diet soda and water. They're choosing between diet and regular, and want to know if they picked right. The uncomfortable answer: there's no good evidence that the swap protects your brain. In Framingham the diet arm looked worse; in the largest pooled analysis both were null; nobody has run a randomised trial and nobody will.

What is reasonably well established is the sugar side of the ledger — sugary drinks are linked to weight gain, insulin resistance and type 2 diabetes, and type 2 diabetes is a recognised dementia risk factor. We covered that in sugary drinks and dementia risk. So: regular soda has a clearer downside, diet soda an unresolved one, and neither has an upside. The dull option is the safest one: water carries none of the open questions sweetened drinks do. Unsweetened tea and plain coffee are reasonable alternatives, though the evidence on how they relate to dementia risk is mixed rather than settled.

What the guidelines say (and don't)

In July 2026 the World Health Organization published the second edition of its dementia risk-reduction guidelines. Artificial sweeteners are not on the list, in either direction — there isn't enough evidence to put them there (WHO, July 2026; we broke down what the guidelines actually changed separately).

What is on the list is the stuff underneath the diet-soda question: managing diabetes, obesity, high blood pressure and cholesterol, plus activity, diet, hearing, alcohol, smoking and sleep. Those rest on far stronger evidence than any beverage study — and they are exactly the factors that made the diet-soda association appear and disappear.

What to actually do about it

A proportionate response — not a panic, and not a shrug.

• Don't treat the swap as a win. Moving from regular to diet hasn't been shown to be a brain-health measure. The useful next step is downward in total, not sideways.
• Cut the frequency. The signals turn up at daily-level intake: one or more artificially sweetened drinks a day in Framingham, more than one a day in the Northern Manhattan Study, and the highest tertile of total sweetener intake (about 191 mg/day, roughly one can of diet soda's worth of aspartame) in the Brazilian cohort — where the middle tertile also declined faster than the lowest. Drinking less often has not been singled out as the problem, but none of these studies set out to test occasional intake, so treat that as an absence of evidence rather than evidence of safety.
• Put water in the gap — unsweetened tea or plain coffee are reasonable too. None of them is a proven brain-health intervention; the point is simply that they don't carry the open question sweetened drinks do.
• Deal with what's underneath. If the reason you drink diet soda is weight, blood sugar or a sugar habit you're managing, that is the dementia-relevant variable — a far bigger lever than the can. Get your blood pressure, blood sugar and cholesterol checked and discuss the numbers with your clinician.
• Don't stop a treatment over this. Nothing here is a reason to change diabetes or blood-pressure medication. That's a conversation with your clinician, not a conclusion from a nutrition study.

The bigger picture

The diet-soda question illustrates a habit worth breaking: hunting for the single item that is secretly ruining your brain. It's appealing because it's actionable — cut one thing and you're done. The dietary signals that do hold up are about patterns rather than single items, which is the thread running through the foods most consistently linked to higher dementia risk.

The real picture is less dramatic and more useful. Up to 45% of dementia risk worldwide is attributed to modifiable factors — blood pressure, hearing, blood sugar, activity, alcohol, smoking, sleep, social connection, vision, cholesterol and more — and which of those is loose in your life differs from your neighbour's. That's a population estimate under ideal conditions, not a promise to any individual, but none of it hinges on what's in your fridge door.

How we research and review our content →

This article is educational and is not medical advice, diagnosis, or treatment. The studies cited — the 2017 Framingham Stroke analysis, the 2025 JAMA Psychiatry pooled cohort analysis, the 2025 Neurology sweetener and cognitive-decline study, the 2026 Northern Manhattan Study, and the 2026 WHO dementia risk-reduction guidelines — are observational and describe associations in general populations, not Solenna specifically; they conflict with each other and none establishes that diet soda or artificial sweeteners cause dementia. Solenna is not a medical device and does not diagnose, prevent, treat, or cure Alzheimer's disease or any other condition. Risk reduction means lowering probability, not eliminating it; individual results vary and no outcome is guaranteed. If you are managing diabetes, blood pressure, cholesterol or weight, discuss any dietary change with a qualified healthcare professional, and never start, stop, or change a medication on your own.