Higher body weight linked to lower risk of death from dementia

People who were overweight or obese when dementia emerged had lower mortality than those in the normal weight range. The finding persisted after researchers examined several biases that might otherwise explain this counterintuitive pattern.

The observational analysis, led by investigators at Columbia University’s Mailman School of Public Health and NYU’s School of Global Public Health, followed 4,523 adults. Published September 30 in the Journal of Alzheimer’s Disease, it suggests that weight at dementia onset may offer useful information about prognosis.

The results concern an association among people who developed dementia. They do not establish that excess weight caused longer survival, and they do not test whether deliberately gaining weight would help. The authors instead emphasize nutrition and the possible value of weight status alongside other prognostic factors.

The researchers used Health and Retirement Study data collected from 1998 through 2022. This ongoing, nationally representative U.S. study follows adults aged 51 and older. Everyone included in the analysis was dementia-free on entering the study and developed dementia later.

That design allowed the team to examine weight before and at dementia onset, then track subsequent survival. The long observation window offered information beyond a single weight measurement. It also supported analyses of weight changes that could influence the apparent relationship with mortality.

Compared with normal weight, overweight was associated with a 16% lower mortality hazard. This category covered a body mass index, or BMI, of 25 to 29.9. Class I obesity, defined as BMI of 30 to 34.9, was associated with a 27% lower hazard.

For class II and III obesity combined, representing BMI of at least 35, the reduction was 22%. The associations remained after adjustment for demographic characteristics, smoking and other chronic conditions. The pattern did not show progressively lower mortality with every increase in weight category.

The class I obesity estimate had a hazard ratio of 0.73, with a 95% confidence interval of 0.62 to 0.86. A hazard ratio compares mortality rates over follow-up. It does not mean that participants lived 27% longer or gained a specified number of years.

Obesity is associated with illnesses including heart disease and diabetes. Higher weight in midlife has also been associated with a greater likelihood of developing dementia later. Those findings address a different question from survival after dementia has emerged.

Weight can change substantially as the disease develops. Earlier research by Yuan Zhang and Virginia Chang found that weight loss began at least a decade before dementia onset. It accelerated in the years immediately preceding onset and continued afterward.

This creates a challenge for studies that classify people using only their current weight. Someone in the normal BMI range may previously have had a higher BMI and lost considerable weight. That weight loss could reflect disease processes associated with poorer survival.

Such a pattern could make higher weight appear beneficial even if weight itself provides no protection. The new analysis examined this possibility rather than assuming that all people in a BMI category shared similar histories. The researchers reported that weight-loss-related bias did not appear to be a major explanation for their results.

Another concern arises when researchers study only people who develop a particular disease. Restricting a sample in this way can create misleading relationships between characteristics that influence disease risk and subsequent outcomes. In this case, that could distort the relationship between BMI and mortality.

The authors also considered survivor bias. People with obesity may die from other conditions before reaching dementia onset. Those who remain available for study could therefore represent a comparatively healthier subset of the original population with obesity.

If that selection were strong enough, it might help explain the lower mortality observed after dementia emerged. Additional analyses evaluated both disease-related selection and selective survival. Neither appeared to be a major contributor to the association the researchers observed.

“Our analyses suggest that the major sources of bias we examined do not explain the survival advantage associated with higher weight,” Zhang said. He is an assistant professor of sociomedical sciences at Columbia Mailman and the Butler Columbia Aging Center.

The conclusion applies to the biases investigated. It does not establish that every possible source of distortion has been eliminated. Statistical adjustment and sensitivity analyses strengthen interpretation, but an observational association still cannot establish the effect of a weight-changing intervention.

The researchers discuss physiological reasons that higher weight might be associated with better survival after dementia onset. Greater energy stores could provide reserves during periods of wasting or inadequate food intake. More muscle mass is another possible contributor.

These are proposed explanations, not mechanisms demonstrated by the analysis. The study leaves unanswered why the higher-BMI groups had lower mortality. It also does not establish which aspect of body composition, if any, accounts for the pattern.

Eating difficulties can become an important part of dementia. Changes in the brain may reduce appetite or initiative, while impaired independence can make feeding more difficult. Increased energy expenditure may also contribute to weight loss and nutritional problems.

“Nutrition can be a big problem for people with dementia,” said Chang, the study’s first author and an associate professor at NYU. These difficulties help explain why the authors consider nutritional status relevant to interpreting survival. They do not prove that increasing BMI would correct the underlying problems.

Weight status may help describe prognosis without being a direct cause of the outcome being predicted. The authors suggest that BMI at dementia onset could contribute information alongside other factors. A group-level association, however, does not determine how long an individual will live.

Chang explicitly cautioned against interpreting the finding as a reason to gain weight. “This doesn’t mean that patients with dementia should attempt to become overweight or obese, as our study was not designed to test that question,” she said.

The study provides evidence that the relationship between weight and health depends on when it is measured and which outcome is examined. Its strongest conclusion is narrower than a claim that obesity protects against dementia. Higher BMI at onset was associated with lower subsequent mortality, while the causal explanation remains unresolved.

These resources examine weight changes around dementia onset, nutritional care and the broader evidence on risk and prognosis.

Time Path of Weight Status Before and After Incident Dementia: Tracks BMI over time and documents weight loss beginning years before dementia onset. (Journal of Aging and Health, 2024)

ESPEN guideline on nutrition and hydration in dementia – Update 2024: Reviews evidence and recommendations for assessing and supporting nutrition and hydration in people with dementia. (Clinical Nutrition, 2024)

Time to nursing home admission and death in people with dementia: systematic review and meta-analysis: Synthesizes survival research and examines factors that influence prognosis after dementia diagnosis. (BMJ, 2025)

Predictors for survival in patients with Alzheimer’s disease: a large comprehensive meta-analysis: Examines demographic, clinical and biological factors associated with survival in Alzheimer’s disease. (Translational Psychiatry, 2024)

Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission: Places obesity and other risk factors within a broader review of dementia prevention and care across the life course. (The Lancet, 2024)

Research findings are available online in the Journal of Alzheimer’s Disease.

The original story "Higher body weight linked to lower risk of death from dementia" is published in The Brighter Side of News.

Like these kind of feel good stories? Get The Brighter Side of News' newsletter.