A prospective cohort study of 139,213 U.S. postmenopausal women aged 50 to 79 years found that incorporating waist circumference thresholds stratified by body mass index modestly improved long-term mortality risk prediction. The findings support routine clinical measurement of waist circumference in all body mass index categories, including among patients already classified as obese.

Participants were followed for up to 24 years as part of the Women’s Health Initiative, during which 69,297 deaths occurred. Researchers developed three nested risk models to predict 10- and 20-year mortality:

• A mortality model used standard health predictors from the American Heart Association’s Life’s Essential 8 (eg, smoking status, physical activity, sleep, diet, and blood pressure);

• A BMI model, which added body mass index (BMI) categories; and

• A BMI-WC model, which further stratified BMI categories by BMI-specific waist circumference (WC) thresholds recommended by the International Atherosclerosis Society and International Chair on Cardiometabolic Risk (IAS/ICCR).

The thresholds define a “large” WC as 80 centimeters or more for women with a normal BMI ranging from 18.5 up to 25 kg/m², increasing by BMI category to 115 centimeters or more for women with a BMI of 40 kg/m² or higher. In multivariable-adjusted models, mortality risk was consistently higher for women with a large WC compared with their counterparts of the same BMI who had a normal WC.

Women with a normal BMI but a large WC had a significantly higher risk of death (hazard ratio [HR], 1.17; 95% confidence interval [CI], 1.12 to 1.21) compared with those who had both normal BMI and normal WC. This level of risk was similar to that of women with class 1 obesity (BMI from 30 to less than 35 kg/m²) and normal WC (HR, 1.12; 95% CI, 1.08 to 1.16). The group at highest risk—women with a BMI of 40 kg/m² or higher and a large WC of 115 centimeters or more—had an HR of 1.98 (95% CI, 1.82 to 2.14).

In the first validation cohort (n=48,335), which had a high prevalence of overweight or obesity, the addition of WC thresholds improved the 10-year c-statistic from 60.7% to 61.3% (an absolute gain of 0.7%) and yielded a continuous net reclassification improvement (NRI) of 20.4% (95% CI, 17.3%–23.6%). In the second validation cohort (n=23,104), which included more Black (12.7%) and Hispanic (9.1%) women, NRI was 12.3% (95% CI, 8.5%–16.0%), although the c-statistic increase was smaller (0.3%) and was not statistically significant.

The authors note that these improvements in predictive performance were modest, likely due to the relatively small magnitude of the hazard ratios across WC categories, most of which were under 2. Nonetheless, the enhancements are meaningful at a population level, particularly given the high prevalence of central adiposity and obesity in older U.S. women.

WC was measured at the narrowest waist, which differs from World Health Organization and U.S. National Institutes of Health guidelines recommending measurements at the iliac crest or other bony landmarks. However, a prior systematic review cited by the authors concluded that measurement technique does not meaningfully affect associations with mortality risk.

Current U.S. guidelines advise against measuring WC in individuals with a BMI of 35 kg/m² or higher, based on the assumption that it provides no additional value. These new findings challenge that position. The study found that elevated WC added independent prognostic information across all BMI categories, including higher levels of obesity.

The authors conclude that integrating BMI-specific WC thresholds into routine clinical assessments could help better stratify obesity-related mortality risk and guide interventions—especially for older women.

Disclosures and funding information are available in the full published study.

Source: Annals of Internal Medicine