Waist size flags mortality risk that BMI alone misses in seniors

Clinical takeaway: BMI alone may understate mortality risk in older patients with central adiposity. Measuring waist circumference identifies a higher-risk group, particularly patients whose BMI falls in the normal range. 

After age 65, BMI and actual adiposity drift apart as lean mass declines and fat redistributes toward the abdomen. Waist measurement would help to capture the associated increased mortality risk, but this recommendation to pair BMI with a measure of central adiposity is not new.  

The 2025 Lancet Commission on Clinical Obesity called for confirming excess adiposity with a second anthropometric criterion, and a prior consensus statement argued for treating waist circumference as a vital sign. But neither could refer to nationally representative US data on the two measures together in adults over 65. This analysis supplies that, based on almost 7,000 Medicare beneficiaries in a nationally representative cohort study followed from 2011 through 2024. 

Among participants with a normal BMI, a high waist circumference, over 40 inches in males and over 35 in females, was tied to 33% higher mortality risk in males and 23% higher in females, compared with normal BMI and a low waist measurement. Waist size carried an independent effect across the full cohort, with high waist circumference associated with a 24% higher risk in both sexes after adjustment for BMI. Risk rose as waist size increased, though estimates at the largest measurements rested on small subgroups. 

Underweight, not obesity, was associated with the highest mortality risk. It was tied to roughly double the mortality risk in both sexes, rising to nearly threefold in the small group of females who were underweight with a high waist size. 

An inverse BMI signal was concentrated in males, who showed 46% lower mortality risk when overweight and 51% lower with class I or II obesity. Females leaned the same direction, with 27% lower risk when overweight, but that effect did not hold up in the joint analysis. 

The analysis used the National Health and Aging Trends Study, with BMI and waist circumference updated at each annual round rather than measured once, so estimates tracked shifting body composition across the follow-up. Cox models adjusted for age, race, education, income, smoking, homebound status, and the complementary adiposity measure, and deaths within the first year of enrollment were excluded to limit reverse causation. 

BMI category alone sorted these patients poorly, in both directions. Some with a normal BMI carried elevated risk, and some classified as overweight did not. The authors attribute this to BMI's blindness to body composition rather than to any protective effect of adiposity. 

"BMI and waist circumference provide complementary information about health risk in older adults," said study co-author Bryan Blissmer, PhD, of the University of Rhode Island. "Our findings suggest that considering both measures together, rather than relying on BMI alone, may improve mortality risk assessment and support more informed clinical decision-making." 

Source: Xu F, et al. (2026 Aug 19) J Am Geriatr Soc. Time-varying body mass index, waist circumference and all-cause mortality in US adults aged 65 or older