Very small and large waists linked to higher death risk
At or above the median waist size for sex, age and weight-for-height, each 5 cm increase was linked to 16% higher mortality in men, 14% in women.

Our read
Populations- Studied:
- Cohort study of 17,252,815 adults undergoing routine health screening, mean 13.9 years
- What changed:
- Death risk was higher at both waist extremes; above-median waists had 16% higher risk in men, 14% in women per 5 cm
- The catch:
- Pre-existing illness could partly explain the higher death risk at waist extremes
Editor of record: Boris Djordjevic · AgingResearch.News
South Korean adults aged 18–99—8,905,225 men and 8,347,590 women—were followed for an average of 13.9 years in a prospective cohort study. They had waist size, height and weight measured during routine health screening in 2009–2010. Researchers compared deaths from any cause across 15 categories of body mass index, a measure of weight relative to height. During follow-up, 955,725 men and 615,825 women died.
Both high and very low waist measurements were linked to higher mortality within every body mass index category. This pattern was more pronounced at older ages. Waist ranges and combinations of waist size and body mass index linked to the lowest mortality shifted upward with age.
Why it matters
Obesity management may benefit from age-specific waist thresholds that also account for body mass index, the authors suggest.
Caveats
The authors note that pre-existing illness and differences in who survived to older ages may partly explain the links, especially among the oldest adults. Findings from this Korean population may not apply elsewhere.
From the paper
- Result
… BMI 21–21.9 with WC 85–89 cm (HR 1.71, 95% CI 1.64 to 1.78) or BMI 20–20.9 with WC 80–84 cm (HR 1.78, 95% CI 1.71 to 1.84) showed higher mortality than BMI 29–29.9 with WC 85–89 cm (HR 1.21, 95% CI 1.10 to 1.33) or WC 100–104 cm (HR 1.55, 95% CI 1.46 to 1.64) …
Results, Combined analysis of waist circumference and BMI on mortality
- Limitation
Some of the observed associations, particularly the stronger J-curve among the oldest adults, may partly reflect selective survival, confounding by pre-existing illness, or competing risk, which cannot be excluded in an observational design.
Quoted word for word from the full text on jech.bmj.com.
The paper
Catholic Kwandong University; Yonsei University
Journal of Epidemiology and Community Health, Nov 17, 2026

Comments
Be the first to comment.