Waist, waist-height-ratio vs body mass index and the risks of multiple diseases: a cohort study with replication

Importance

Body mass index (BMI) is widely used to assess obesity-related health risks, but waist circumference (WC) and waist-to-height ratio (WHtR) are thought to better capture central adiposity and associated disease risk.

Objective

To compare the associations of BMI, WC, and WHtR with the risk of 78 incident conditions, multiple long-term conditions (MLTCs), and mortality, and to evaluate whether waist-based measures improve risk identification over BMI and, if so, over which BMI range.

Design, Setting, and Participants

This cohort study analyzed data from 495,911 adults in the UK Biobank with a mean follow-up of 13.7 years. Findings were replicated in the Whitehall II study (n = 7,973; mean follow-up, 28.6 years).

Exposures

Obesity defined by BMI, WC, and WHtR.

Main Outcomes and Measures

Associations of each adiposity measure with 78 incident conditions, MLTCs, and mortality were assessed using Cox proportional hazards models. Population attributable fractions (PAFs) were calculated to estimate the proportion of outcomes attributable to obesity defined by each measure.

Results

Obesity defined by BMI, WC, or WHtR was associated with increased risk for most outcomes. WHtR-defined obesity generally showed the strongest associations, followed by WC and BMI, with some exceptions (e.g., osteoarthritis). However, PAFs for MLTCs for WC and WHtR were <1% percentage points higher than for BMI. Notably, meaningful gains in PAFs were observed only among individuals with BMI between 25.0 and 29.9. These findings were replicated in the Whitehall II cohort.

Conclusions and Relevance

Replacing BMI with waist-based measures did not meaningfully improve risk identification for MLTCs in the general population when BMI was ≥30. However, waist-based measures may offer additional predictive value among individuals with BMI <30.

Key Points