Among Americans over 65 with a BMI considered normal, having a high waist circumference was associated with a 33% greater risk of mortality in men and 23% in women. On the scale, in short, they ended up in the same group. The tape measure told something more.
The data comes from a study published in Journal of the American Geriatrics Societywhich analyzed 6,905 people age 65 and older included in the National Health and Aging Trends Study, a representative sample of U.S. seniors enrolled in Medicare. The researchers used data collected between 2011 and 2024, updating both BMI and waist circumference over time instead of relying on a snapshot taken only at the beginning.
The result puts some cracks in the habit of looking at BMI as if it were enough on its own. Especially when the years increase and, together with them, the body also changes.
Higher BMI was associated with lower mortality. With several asterisks
BMI, or body mass index, relates weight and height. It is quick, cheap and very useful in population studies, but it does not distinguish a kilo of muscle from a kilo of fat and does not say where that fat is located.
In the American sample this limitation is seen quite well. Compared to men with normal BMI, those classified as overweight had a 46% lower model-estimated mortality risk; for men with class I and II obesity, that is, with a BMI between 30 and less than 40, the reduction reached 51%. In women, overweight was associated with a 27% lower risk, while for class I and II obesity the difference was not statistically significant.
Underweight went in the opposite direction: the risk was almost double in both men and women.
It is the terrain on which the so-called “obesity paradox” has been discussed for years, the apparent survival advantage observed in some groups of elderly people with higher BMI. The data does not authorize us to prescribe a few extra kilos as an elixir of long life. The same authors recall that existing illnesses, involuntary weight loss, sarcopenia, selective survival and reverse causality can weigh heavily: a person can lose weight because they are ill, making their lower weight appear as the cause of a risk that had already increased.
There is also a very down-to-earth problem. As we age, we lose muscle mass, height may decrease and fat tends to redistribute. The BMI continues to make its division between kilos and square meters as if nothing had happened.
The waist adds the missing piece
Waist circumference instead looks at fat distribution. In the study it was considered elevated above 40 inches in men, about 102 centimeters, and above 35 inches in women, about 89 centimeters.
After also taking BMI into account, a high circumference was associated with a 24% higher risk of mortality in both sexes. In exploratory analyses, risk also tended to increase when moving to higher circumference bands.
The most interesting case remains that of normal BMI with a high waist size. Compared to people with a normal BMI and circumference below the threshold, the estimated risk rose by 33% in men and 23% in women.
Two people can therefore have the same BMI and very different fat distribution. A difference that the weight-height calculation alone simply does not have the tools to see.
The study has some important methodological advantages: waist circumference was measured by trained personnel and BMI and waist size were updated during follow-up. Weight and height, however, were reported by the participants. The authors excluded those who had died within the first year to reduce the risk that advanced disease and weight loss before death would alter the results, but could not completely eliminate this problem. And we are still talking about an observational study: it measures associations, it does not demonstrate that a certain circumference directly causes a death.
BMI is already losing its monopoly
American research does not come alone. Another study published on BMC Public Health compared six body markers in 89,745 people aged 45 or older from six large longitudinal studies conducted in 25 countries.
Over a median follow-up of ten years, 16,593 deaths were recorded. BMI showed a weak, inverse association with mortality, which attenuated when deaths occurring in the first two years were excluded. Some indicators incorporating abdominal fat distribution, especially ABSI and taper index, maintained more consistent positive associations.
The direction had already been formalized in 2025 by the International Commission on Clinical Obesity published on The Lancet Diabetes & Endocrinology: BMI can serve as a screening tool and in population studies, while to evaluate the individual it should be accompanied by at least one other measure of adiposity, such as waist circumference, waist-hip ratio or waist-height ratio.
For the elderly the message is even more concrete. A number on the scale can stay the same as your muscles, height and fat distribution change. BMI continues to be useful. Except that, after the age of 65, showing up alone for the appointment risks leaving a rather important part of the body out the door.
