Researchers at the University of Illinois Urbana-Champaign have brought machine learning to bear on one of aging's most intimate questions: what shapes the mind as the body grows older? Among the many threads of lifestyle — diet, movement, weight, blood pressure, and time itself — the study found that age, blood pressure, and BMI carry the greatest predictive weight for cognitive performance, while diet and exercise, though genuinely meaningful, play a supporting role. The finding does not diminish the value of how we eat or move, but rather refines the map, suggesting that for many people, ma
Machine learning reveals age, blood pressure, BMI as top predictors of brain function
Diet and exercise might offset the cognitive burden of a high BMI
Mark
So the study found that age, blood pressure, and BMI predict cognitive performance better than diet and exercise. Does that mean diet and exercise don't matter?
Mimi
Not at all. The study found they matter less in terms of raw predictive power, but the researchers noted that diet and exercise can sometimes offset the negative effects of high BMI or other risk factors. It's more nuanced than a simple ranking.
Luke
But we should be careful here—the study measured performance on a single cognitive test, the flanker task. That's a specific measure of attention and inhibitory control. We don't know if these same predictors would rank the same way for memory, processing speed in other contexts, or long-term cognitive decline.
Mark
Why did machine learning reveal something that traditional statistics couldn't?
Mimi
Traditional statistics struggle when you're trying to evaluate many variables all at once and how they interact with each other. Machine learning can hold all those relationships in mind simultaneously and weight them against each other in ways that conventional approaches can't.
Luke
That's true, but it's worth noting that the study used 374 people across a wide age range. The algorithms were tested for predictive validity, but we should know more about how well these findings generalize to other populations or other measures of cognitive function.
Mark
What's the practical takeaway for someone worried about their brain health?
Mimi
The study suggests that managing blood pressure and maintaining a healthy weight are particularly important for cognitive function. But diet and exercise still play a role, and they might be especially valuable if they help you manage your weight and blood pressure.
Luke
And age is the strongest predictor, which we can't change. So the study is really about the modifiable factors—blood pressure, BMI, diet, and activity—and which ones seem to matter most. But this is one study on one cognitive task. It's a useful data point, not a final answer.
Le Pouls
- Cognitive decline is not random — machine learning has now ranked the lifestyle factors that most reliably predict how well a person can focus and filter distraction as they age.
- Age, diastolic blood pressure, and BMI emerged as the dominant forces in the model, outweighing diet quality and physical activity in their power to forecast performance on attention tests.
- Diet and exercise are not irrelevant — they may even interact with weight and other risk factors to partially offset cognitive burden — but their predictive signal is quieter than the physical health markers.
- The tension now lies between what cannot be changed, age, and what can — blood pressure and BMI — with machine learning offering a tool to personalize which interventions matter most for which individuals.
- Clinicians and researchers may soon be able to move beyond generic lifestyle advice, using algorithmic precision to identify the specific levers most likely to preserve cognitive sharpness for each person's unique profile.
Researchers at the University of Illinois Urbana-Champaign have brought machine learning to bear on one of aging's most intimate questions: what shapes the mind as the body grows older? Among the many threads of lifestyle — diet, movement, weight, blood pressure, and time itself — the study found that age, blood pressure, and BMI carry the greatest predictive weight for cognitive performance, while diet and exercise, though genuinely meaningful, play a supporting role. The finding does not diminish the value of how we eat or move, but rather refines the map, suggesting that for many people, managing the body's physical markers may be the most direct path to protecting the thinking mind.
A research team at the University of Illinois Urbana-Champaign set out to answer a deceptively simple question: among all the lifestyle factors thought to shape brain health — diet, exercise, weight, blood pressure, and age — which ones actually matter most? Using machine learning rather than conventional statistics, they were able to weigh all these variables simultaneously across 374 adults, aged 19 to 82, each of whom completed a flanker task, a well-established test of attention and the ability to ignore distraction.
What the algorithms revealed was a clear hierarchy. Age was the single strongest predictor of cognitive performance. Diastolic blood pressure came second, followed by BMI and systolic blood pressure. These physical health markers dominated the model's forecasting power in ways that traditional analysis might have obscured.
Diet and exercise did not vanish from the picture. Adherence to a high-quality diet correlated with better test performance, and physical activity predicted reaction time to a moderate degree. More intriguingly, the researchers found hints that exercise and diet might interact with body weight to partially offset its cognitive burden — suggesting lifestyle choices can sometimes soften the impact of metabolic risk.
The study's lead researcher, Naiman Khan, noted that diets like the Mediterranean, DASH, and MIND have well-established links to cognitive preservation. Yet when all factors are weighed together, dietary benefits, while real, are outpaced in predictive strength by physical health markers. The finding reframes rather than dismisses the role of food and movement.
The deeper promise lies in personalization. Machine learning could help clinicians identify which combination of factors — blood pressure management, weight, diet, or activity — is most likely to protect a given individual's cognitive function as they age. Age cannot be altered, but blood pressure and BMI can be managed, and knowing which lever to pull first may prove to be the most valuable insight of all.
Researchers at the University of Illinois Urbana-Champaign have used machine learning to sort through the tangle of lifestyle factors that shape how well our brains work as we age. The question they posed was straightforward: among diet, exercise, weight, blood pressure, and age itself, which ones matter most for cognitive performance? The answer, published in The Journal of Nutrition, turned out to be more precise than traditional statistics could have revealed.
The study analyzed data from 374 adults ranging from 19 to 82 years old. Each participant completed a flanker task—a standard cognitive test that measures how quickly and accurately someone can focus on a central object while ignoring distracting information around it. The test is well-established in neuroscience as a measure of attention and inhibitory control, the mental ability to filter out noise and stay on target. Alongside their test performance, researchers collected information about each person's age, body mass index, blood pressure readings, physical activity levels, and dietary patterns.
The machine learning algorithms the team deployed could evaluate all these variables simultaneously in ways that conventional statistical methods cannot. What emerged from the analysis was a clear hierarchy. Age proved to be the single strongest predictor of how someone performed on the flanker task. Diastolic blood pressure came second, followed by BMI, and then systolic blood pressure. These three physical markers—age, blood pressure, and weight—dominated the model's ability to forecast cognitive performance.
Diet and exercise did not disappear from the picture, but they played a supporting role. Adherence to the Healthy Eating Index, a measure of overall diet quality, correlated with better performance on the test, though less strongly than the physical health markers. Physical activity emerged as a moderate predictor of reaction time. More intriguingly, the researchers found evidence that exercise and diet might interact with body weight and other factors to influence cognitive function—suggesting that a person's lifestyle choices could sometimes offset the cognitive burden of a high BMI or other risk factors.
Naiman Khan, the professor of health and kinesiology who led the work alongside Ph.D. student Shreya Verma, noted that previous research has already linked specific diets to cognitive preservation. The Mediterranean diet, the DASH diet (Dietary Approaches to Stop Hypertension), and the MIND diet—which combines elements of the other two—have all been associated with protection against cognitive decline and dementia. Similarly, diets rich in antioxidants, omega-3 fatty acids, and vitamins have shown associations with better brain function. Yet this new study, by weighing all factors together, revealed that these dietary benefits, while real, are outweighed in predictive power by the physical health measures.
The implications point toward a more personalized approach to cognitive health. Khan suggested that machine learning could help clinicians and researchers tailor strategies for aging populations, for people with metabolic risks, and for anyone seeking to enhance brain function through lifestyle changes. The precision that algorithms bring to analyzing large datasets with multiple variables could surface patterns that remain hidden in conventional analysis. Rather than offering generic advice about diet and exercise, future interventions might be calibrated to an individual's specific profile—recognizing that for some people, managing blood pressure or weight might yield the most cognitive benefit, while for others, dietary changes could be the lever that moves the needle.
The study does not suggest that diet and exercise are unimportant. It suggests, instead, that the relationship between lifestyle and brain health is more intricate than a simple checklist. Age cannot be changed, but blood pressure and BMI can be managed. The question now is whether interventions targeting these modifiable factors, informed by machine learning's ability to see which combinations matter most for which people, might help preserve the cognitive sharpness we rely on as we grow older.
Citations marquantes
Machine learning can evaluate a host of variables at once to identify those that align most closely with cognitive performance, in ways that standard statistical approaches cannot embrace all at once.— Naiman Khan, professor of health and kinesiology, University of Illinois Urbana-Champaign
Physical activity emerged as a moderate predictor of reaction time, with results suggesting it may interact with other lifestyle factors such as diet and body weight to influence cognitive performance.— Naiman Khan
Envie de l'histoire complète ?
Lire l'original sur News-Medical.net ↗
Technology