"Low-Calorie Sweeteners and Faster Brain Aging: What a 12,772-Person Study Actually Shows"

A diet soda is sold as a free choice: all of the sweetness, none of the sugar, none of the cost. A study published in Neurology, from the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil), suggests the cost may not be zero after all — not in calories, but in the speed at which thinking and memory decline. The finding has been circulating widely this week, and it is worth reading carefully, because what it shows and what it does not show are almost equally important.

The researchers followed 12,772 civil servants, average age just under 52, across three waves of testing between 2008 and 2019 — roughly eight years of follow-up. Participants filled in food-frequency questionnaires from which the team calculated daily intake of seven low- and no-calorie sweeteners: aspartame, saccharin, acesulfame-K, erythritol, xylitol, sorbitol and tagatose. They then took six cognitive tests covering memory, verbal fluency and overall thinking. The average intake across the whole group was about 92 milligrams a day. The lowest third of consumers averaged 20 milligrams; the highest third averaged 191 milligrams, roughly the amount of sweetener in a single can of diet soda, according to news reports of the study.

Among participants younger than 60, those in the highest intake group showed a 62 percent faster decline in overall cognitive ability than those in the lowest group. The researchers translated that gap into about 1.6 years of extra cognitive aging. The decline showed up most clearly in memory and verbal fluency. Six of the seven sweeteners were individually linked to faster decline in global cognition; tagatose, the seventh, was not. Sorbitol was the most commonly consumed sweetener in the group. The association also held for people with diabetes, where it showed up in memory as well as overall cognition, and for people without diabetes, where it showed up in verbal fluency.

There are two details that keep this study from being a simple scare story. First, the association appeared only in participants under 60; among those aged 60 and older, sweetener intake was not linked to faster decline. That split is strange if the mechanism were straightforward damage, and the authors do not fully explain it — it may reflect survival effects, different consumption patterns at older ages, or simply that the younger group had more decline left to measure. Second, this is an observational study, and the authors say so plainly. Diets were self-reported, some participants dropped out over eight years, and people who drink a lot of diet soda also tend to eat more ultra-processed food, carry more weight, and live with more diabetes — all of which affect cognition. The models adjusted for age, sex, blood pressure and cardiovascular factors, but residual confounding, the polite term for everything the model could not capture, remains.

That is why my reading of this study is neither dismissal nor alarm. A 62 percent faster decline sounds dramatic until you remember it is 62 percent faster than a slow baseline: normal cognitive aging in your fifties is gradual, and 1.6 years of extra aging is a difference in trajectory, not a diagnosis. But the sample is large, the follow-up is long, the dose pattern is consistent across six different sweeteners, and the result lines up with earlier cohort work linking artificially sweetened drinks to stroke and dementia risk. It is the kind of finding that deserves a replication attempt, not a shrug.

So here is the prediction I am putting on the record with this article. By 31 December 2027, a peer-reviewed study in a cohort outside Brazil, or a meta-analysis pooling more than one cohort, will report a same-direction association: higher low- and no-calorie sweetener consumption linked to faster cognitive decline, statistically significant for global cognition or memory. I put that at 50 percent. The case for it is the size of the effect, the size of the cohort, and the attention this study is now receiving, which gives other research groups every incentive to test the claim in their own data. The case against it is that observational nutrition findings replicate poorly, that the under-60-only pattern is a warning sign, and that a year is a short time for a new cohort analysis to be designed, run, peer-reviewed and published. A meta-analysis of existing cohorts is the faster route, and that is the most likely way this prediction comes true — or fails, if the pooled result washes out.

For a reader deciding what to do on Monday morning, the honest summary is narrower than the headlines. This study does not prove that a diet soda damages your brain. It does show, in a large group followed for eight years, that the heaviest consumers of these sweeteners lost ground faster than the lightest consumers, and that the difference was concentrated in people under 60. If you drink one diet soda a day, you are roughly in the highest group this study measured. That is a fact worth knowing, and it is not a reason to panic — it is a reason to watch what the next study says, because this is exactly the kind of question where one cohort should never be the last word.