Years of higher body weight were linked to faster decline in memory and thinking

Being heavy in midlife has been tied to poorer memory and thinking in later life in several earlier studies, but the picture has been messy. Many studies weighed people only once, at the start, and a single number says little about the decades a body actually spends at a given weight.

A team at the University of Georgia tried a different measure. Instead of one weight reading, they used a running average of body mass index (BMI, weight divided by height squared) over many years, which they call the cumulative average BMI. The idea is to capture the total "dose" of extra weight a person has carried.

What they did

The researchers used the Health and Retirement Study, a long-running US survey of people aged 50 and over who are interviewed every two years. They started with the 1996 wave and kept everyone who had cognitive scores and weight data and showed no sign of cognitive impairment or dementia at the start: 8,252 people, with an average age of 59. The average participant was followed for 17.5 years, and the data ran to 2020.

Height and weight were reported by the participants themselves. The average BMI at the start was 27.3, which counts as overweight.

Thinking skills were measured with short tests done at each wave:

• Memory: recalling a list of ten words right away and again after a delay.
• Executive function, the skills used for planning and keeping track: subtracting 7 again and again, and counting backwards.

The team combined these into an overall score and asked whether people with a higher long-term average BMI lost points faster. Their models took into account age, sex, ethnicity, education, insurance, employment, smoking, depression and the number of chronic diseases. They also tested time gaps of 2 to 16 years between the weight measure and the later decline, to see when the link was strongest.

What they found

• Faster decline with higher long-term BMI. A higher cumulative BMI went with a slightly faster yearly drop in the overall score, in memory and in executive function. The link held after each extra set of adjustments.
• Strongest after about eight years. The association was weak two years later, grew stronger and peaked at the eight-year gap, then faded somewhat at longer gaps. This was true for all three scores.
• Stronger in older people. In participants aged 65 and over, the link with faster decline was more than four times as strong as in those aged 50 to 65. It was similar in men and women and across education levels.
• Robust to checks. Results were much the same when people with chronic diseases at the start were left out, and when the team used a statistical weighting to account for people who died or dropped out.

The yearly differences themselves are tiny on the scale the team used (a few thousandths of a standard deviation per year). The researchers describe them as "modest in magnitude", but point out that small yearly differences can add up over decades.

What to keep in mind

• Association, not cause. The authors stress that the design shows links, not cause and effect. Things they could not account for, such as physical activity, head injuries and environmental exposures, may play a role.
• Self-reported weight. Weights people report themselves can be off, and shrinking height in old age can push BMI up a little. BMI also cannot tell fat from muscle or show belly fat, and an average can hide periods of fast weight change.
• Who stayed in the study. Many people from the original survey lacked the needed data, and those with longer follow-up had to survive and keep taking part. The authors say the findings apply mainly to people like those in the sample and that results at the longer time gaps need caution. At short gaps, early disease may itself change weight.
• No test of weight loss. The study did not test whether losing weight protects memory. The authors note that recent trials of the weight-loss drug semaglutide (a prescription medicine) in people with mild Alzheimer's disease did not show a cognitive benefit over two years, and suggest that much longer periods may matter.

Why it matters

Most work on weight and the brain looks at a snapshot. This study follows the same people for up to 24 years and uses their weight history rather than a single reading. It adds to the evidence that carrying a higher weight for many years goes together with somewhat faster decline in memory and thinking, especially after 65, and it suggests the link shows up with a delay of several years. Whether changing that weight history would change the outcome is a question only trials can answer.