Most of us have a well-rehearsed speech for patients with elevated blood sugar. We talk about the heart. The kidneys. The eyes. The feet. We walk through the whole organ-by-organ checklist a glucose reading can quietly damage over the years. What most of us, myself included, haven’t been saying nearly often enough is anything about the brain, and a study published this summer in Molecular Psychiatry is a good reason to start.
Researchers combined MRI scans, genetic data, and blood metabolite profiles from tens of thousands of UK Biobank participants to train a model that estimates a person’s brain age directly from imaging, independent of the calendar. They then compared that estimate to each person’s actual age and asked which of dozens of blood metabolites best explained the gap in people whose brains looked older than their birth certificate said they should.
One result stood far above the rest. Blood glucose showed the strongest association with accelerated brain aging of any metabolite in the study, stronger than cholesterol, stronger than inflammatory markers, stronger than anything else the researchers measured. They then used Mendelian randomization to test whether the relationship was likely causal rather than incidental, and the genetic evidence pointed toward glucose actively driving the aging process, not just tracking alongside it. Higher glucose was associated with smaller volume across eighty separate brain regions and with higher rates of dementia, Alzheimer’s disease, vascular dementia, Parkinson’s disease, stroke, depression, and anxiety.
I want to be precise about what this does and doesn’t show, since our colleagues deserve the same caution we’d want applied to our own work before we change how we counsel patients. This is an observational and genetic analysis, not a clinical trial proving that lowering glucose reverses brain aging or prevents a dementia diagnosis, and no single glucose reading determines any one patient’s trajectory. Blood pressure, sleep, physical activity, smoking, and education all shape brain aging too. But out of everything the researchers measured, in a dataset that size, glucose was the strongest modifiable signal by a wide margin, and the pattern showed up well before a diabetes diagnosis, across the full range from normal glucose through prediabetes.
For decades, we’ve treated glycemic control mainly as diabetes prevention, protecting the pancreas, avoiding a diagnosis, hitting an A1C target. This dataset is a reminder that the brain is a metabolic organ too, one that consumes a disproportionate share of the body’s glucose and is bathed in whatever concentration is circulating, hour after hour, for a lifetime.
Two levers we can offer patients don’t require a new prescription. The first is what’s on the plate. In randomized trials, including two my colleagues and I ran, shifting the diet toward plant foods, vegetables, legumes, whole grains, in place of meat and dairy, improved insulin sensitivity and glycemic control more than the standard comparison diet, independent of weight change. A low-fat vegan diet improved glycemic control and cardiovascular risk factors in patients with type 2 diabetes, and a vegetarian diet outperformed a conventional diabetic diet on insulin resistance and oxidative stress markers in a separate trial. Neither required cutting out animal foods entirely; the benefit tracked with the direction and degree of the shift, not an all-or-nothing switch.
The second is timing. Insulin sensitivity isn’t constant across the day; it’s highest in the morning and declines as the day goes on, so identical meals produce different glycemic responses depending on the hour they’re eaten. In a randomized trial of early time-restricted eating, compressing food intake into the earlier part of the day improved insulin sensitivity, blood pressure, and a marker of oxidative stress in men with prediabetes, with no change in weight. In a trial my team ran in patients with type 2 diabetes, eating two larger meals, breakfast and lunch, instead of six small meals spread across the day improved insulin sensitivity and reduced liver fat, again independent of total calories consumed.
Both levers hit the same target, blood glucose, from different angles: what reaches the bloodstream, and when it arrives relative to the body’s own metabolic clock. Neither requires a device, a prescription, or perfection to produce a measurable effect, which makes them easy to fold into a visit that’s already short on time.
None of this makes brain aging simple, and no single dietary change erases the role of genetics, blood pressure, sleep, or the rest of what we already counsel on. But of everything researchers could measure in tens of thousands of brains, glucose came out as the strongest modifiable signal they found. We’ve spent years asking patients what they’re eating for the sake of their hearts and their waistlines. This study is a good reason to start asking for the sake of their minds too, and to have that conversation before diabetes is on the chart, not after.
Hana Kahleova is an endocrinologist.
