Higher BMI After 65 Tracked Lower Death Risk. Waist Didn’t

Older men and women with arms outstretched during an exercise class in a community hall
Older adults in an exercise class at the Rockville Senior Center in Maryland, October 1988 (file photo). Photo: Bill Branson, National Cancer Institute, via Wikimedia Commons, public domain (cropped)

Overweight American men over 65 died at roughly half the rate of their normal-weight peers over 14 years. That number comes from a study of 6,905 older adults in the Journal of the American Geriatrics Society. It is also the number most likely to be misread.

The same study found something else. A large waist raised the risk of death at every BMI, a normal one included. The authors go out of their way to say the results are not evidence that body fat protects anyone.

What the study measured, and in whom

Fang Xu, Bryan Blissmer and four colleagues at the University of Rhode Island and the University of Connecticut used the National Health and Aging Trends Study. It follows a nationally representative sample of Medicare beneficiaries. The team tracked 6,905 people aged 65 and over from 2011 to 2024. Height, weight and waist were updated every year rather than fixed at the start (doi:10.1111/jgs.70600).

Against normal-weight men, overweight men had a hazard ratio of 0.54 for death from any cause. Men with class I or II obesity came in at 0.49. Women showed the same pattern, more mildly. Underweight people of both sexes died sooner.

Waist told a different story. A waist over 40 inches in men, or 35 in women, carried a hazard ratio of 1.24. That held after the researchers accounted for BMI. Normal BMI with a big waist beat normal BMI with a small one in only one direction: 1.33 in men, 1.23 in women.

The authors warn against the obvious reading

The paper’s conclusion is about measurement, not diet. It points to “the adverse association of central adiposity and the limitations of BMI-only classification for mortality risk stratification in older adults.”

In the discussion the authors list what could produce an apparent survival benefit with no biology behind it. They name reverse causation, unmeasured illness and survivorship bias. The results, they write, “should not be interpreted as evidence that excess adiposity is biologically protective.”

Two limitations matter. Participants reported their own height and weight, which the authors concede may misclassify people; trained staff did measure the waists. The team also excluded early deaths, but still accepts that reverse causation is possible. That is the loop where illness drives weight down, rather than low weight driving death.

“BMI should not be interpreted in isolation when evaluating older adults,” Blissmer told Newsweek. Clinicians should watch for unintentional weight loss or a falling BMI, he added. Those often signal lost muscle, frailty or poor nutrition.

Why the obesity paradox is still disputed

Researchers have argued about this inverse association for two decades. The argument turns on study design.

The largest counterweight is a 2016 meta-analysis in The Lancet, pooling individual data from 239 studies on four continents. Its authors built it to strip out the two best-known distortions. They used only never-smokers, dropped anyone with chronic disease at recruitment, and ignored the first five years of follow-up. That left 3.95 million people and 385,879 deaths (PMID 27423262).

There, mortality sat lowest at a BMI of 20 to 25 and climbed above it. The hazard ratio ran 1.07 in the 25 to 27.5 band and 1.20 from 27.5 to 30. Across the three obesity grades it reached 1.45, 1.94 and 2.76.

Age did weaken the effect. Each 5-point rise in BMI above 25 carried a hazard ratio of 1.52 when measured at 35 to 49. At 70 to 89 it was 1.21. Weaker, but still above 1.

Put the two designs side by side and the disagreement becomes legible. The Rhode Island team adjusted for ever-smoking and excluded early deaths. The Lancet collaboration threw out smokers entirely, threw out the already-ill, and threw out five years of deaths. The harder a study scrubs for sickness and smoking, the smaller the paradox gets. It does not vanish in the oldest groups, which is why the question stays open.

What this does and does not tell a 70-year-old

Nothing here recommends putting on weight. A high BMI at 70 can mean retained muscle and bone, which helps people survive pneumonia or a hip fracture. It can equally mean abdominal fat, which does the opposite. BMI cannot tell those apart. A tape measure around the navel can, and it costs nothing.

The practical reading is narrow. Falling weight in an older adult is a warning worth investigating, not a success. A thickening waist is worth acting on even when the scale looks fine. “Normal BMI” is not a clean bill of health at any age.

American health advice keeps getting rewritten around age rather than one threshold. See the new prostate screening guidance on who needs a PSA test and when, or the open questions about patients who stop taking GLP-1 drugs and regain weight. Weight keeps proving a poor single number, which is partly why insulin supply pressure as drugmakers chase GLP-1s reaches the same patients.

The Rhode Island group has a second paper out this year, in Frontiers in Public Health, on what yearly changes in these measurements predict. For anyone over 65 that is the sharper question. Not where the number sits, but which way it is moving.

Exit mobile version