Higher BMI Lowers Fracture Risk — New Meta-Analysis Rewrites FRAX Calculations

Fracture risk assessment is central to osteoporosis management worldwide, and the FRAX tool — used by clinicians across more than 70 countries — incorporates BMI as one of its key input variables. The original FRAX BMI algorithm was derived from earlier, smaller datasets. This new meta-analysis was conducted specifically to update and validate the BMI component of FRAX using a far larger and more diverse evidence base, ensuring the tool's continued accuracy in contemporary populations facing rising obesity rates.

The study pooled individual participant data from 25 prospective cohort studies contributing data to the FRAX development consortium, encompassing over 300,000 men and women from multiple countries including the United States, United Kingdom, Australia, Sweden, France, and others. Participants had baseline measurements of height and weight (allowing BMI calculation), were followed prospectively for incident fractures, and had data on key covariates. The primary outcomes were major osteoporotic fracture (MOF — comprising hip, clinical spine, forearm, and humerus fractures) and hip fracture specifically. Cox proportional hazards models were used within each cohort, with results pooled using random-effects meta-analysis. Analyses were adjusted for age and, where available, femoral neck bone mineral density (BMD).

The core finding was a robust inverse relationship between BMI and fracture risk across the full BMI spectrum. For every 5 kg/m² increase in BMI, the hazard ratio for major osteoporotic fracture was approximately 0.94 (95% CI: 0.92–0.96) and for hip fracture approximately 0.90 (95% CI: 0.87–0.93), both highly statistically significant. This relationship was present in both men and women, and persisted — though was attenuated — after adjustment for femoral neck BMD, indicating that BMI exerts fracture-protective effects partly through bone density and partly through other mechanisms such as soft tissue padding and altered fall biomechanics. Importantly, the relationship was nonlinear at the extremes: very low BMI conferred substantially elevated risk, while the incremental protective benefit of each additional BMI unit diminished at higher BMI values.

When specific BMI categories were examined, individuals classified as underweight (BMI <18.5 kg/m²) faced dramatically elevated fracture risk — approximately 1.7–2.0 times higher than those with normal BMI (18.5–24.9 kg/m²) for hip fracture. Those in the obese category (BMI ≥30 kg/m²) had meaningfully lower risk than normal-weight individuals, with hazard ratios around 0.75–0.80 for hip fracture. However, the protective effect of obesity was less pronounced for non-hip fractures including ankle and lower leg fractures, where obesity may paradoxically increase risk — a nuance the updated FRAX model accounts for.

For clinical practice, these findings validate and refine the role of BMI in the FRAX algorithm. The updated coefficients will improve fracture risk estimates particularly at the extremes of BMI — for underweight elderly patients (where risk has likely been underestimated) and for obese patients (where risk may have been overestimated in some site-specific analyses). Clinicians should note that even though obesity is associated with lower hip fracture risk, it does not eliminate risk and should not preclude fracture risk assessment. The general public takeaway is that very low body weight in older age carries real bone fracture risk, reinforcing guidance around maintaining adequate nutrition and muscle mass rather than pursuing very low body weight.