Overweight and obese people were about 30 percent less likely to develop dementia 15 years later than people of a healthy weight. Conversely, underweight people were 34 percent more likely to develop dementia than those whose weight was normal, according to the study authors.
"Our findings were unexpected, that obese and overweight people would be protected," said lead researcher Dr. Nawab Qizilbash, from OXON Epidemiology Ltd. in Madrid, Spain.
However, the retrospective study was only able to show an association between obesity and a reduced risk of dementia, not a cause-and-effect relationship.
And Qizilbash added that people shouldn't take these preliminary findings as a license to gain weight in hopes of preventing dementia since the study also showed a predictable increase in premature death risk from being overweight or obese.
"Even if there were to be a protective effect on dementia from being overweight or obese, you may not live long enough to get the benefit," Qizilbash said.
He added that the findings need further study. "We don't have a biological explanation for the association we observed," he said.
However, Qizilbash thinks the results might point the way toward new treatments or ways to prevent Alzheimer's disease and other dementias.
"Our results should open new avenues for research for protective factors for dementia. They may be used to provide insight to looking for a mechanism and developing new treatments for dementia," he said. "In addition, doctors, public health scientists and policy makers may need to rethink how to identify who is at high risk for dementia."
The report was published online April 10 in The Lancet Diabetes & Endocrinology.
Deborah Gustafson, a professor of neurology at SUNY Downstate Medical Center in New York City, and author of an accompanying editorial in the journal, questioned the study's methods.
Specifically, she wondered whether weight measured in middle age actually reflects the risk of dementia 15 years later. Since there are genetic components to both weight and dementia, the role of weight by itself isn't clear, she said.
"Despite the high number of participants [in the study], it is not the final word, given the methodological questions," Gustafson said. "More research is needed, as well as clarification of these results."
In the study, Qizilbash and colleagues analyzed nearly 20 years of medical records on almost 2 million British adults, whose average age was 55 at the start of the study. During 15 years of follow-up, 45,500 of the participants developed dementia.
The associations between weight and dementia held up even after the researchers took into account when the participants were born or their age at diagnosis. And factors such as drinking and smoking, which are known risk factors for dementia, made little difference in the results, the researchers added.
Dr. Malaz Boustani, an associate professor of medicine at the Indiana University School of Medicine and a spokesperson for the American Federation for Aging Research, suggested that "if a biological mechanism can be found to support these findings, it might lead to new ways of treating or preventing dementia."