Investigating traditional and novel predictors of a single versus multiple fragility fractures in a large observational cohort

Amin, Hamzah and Khan, Muhammed Aqib and Bukhari, Marwan (2026) Investigating traditional and novel predictors of a single versus multiple fragility fractures in a large observational cohort. PLoS One, 21 (2). ISSN 1932-6203

Text (pone.0343413.pdf)
pone.0343413.pdf - Published Version
Available under License Creative Commons Attribution.
Download (876kB)

Abstract

Background: Osteoporosis is a systemic skeletal disease characterised by reduced bone mass and a distortion of bone microarchitecture. It is clinically problematic as it leads to fragility fractures which confers excess morbidity and mortality on patients. Up to 32% of individuals will experience recurrent fragility fractures within two years of an initial fracture, yet existing risk models focus on the risk of having a single fragility fracture at a time. We aim to identify predictors of multiple fragility fractures to help improve risk stratification. Methods: 43,801 patients referred for their first DXA scan in the northwest of England between June 2004 and February 2024 were analysed. Participants underwent lumbar spine and femoral scans to assess bone density and regional body composition. A generalized additive model reporting odds ratios was used to compare risk factors for a single versus multiple fragility fractures. Results: Of the referred population, 14,212 (32.4%) had a single fragility fracture and 3,731 (8.5%) had multiple. Female gender was associated with lower odds of multiple fractures (OR 0.88, 95% CI: 0.79–0.99), while increased odds were linked to family history of fractures (OR 1.22, 95% CI: 1.11, 1.35), secondary osteoporosis (OR 1.15, 95% CI: 1.05, 1.26), rheumatoid arthritis (OR = 1.29, 95% CI: 1.08, 1.53), glucocorticoid therapy (OR = 1.18, 95% CI: 1.00, 1.39), smoking (OR 1.27, 95% CI: 1.12, 1.45) and falls risk (OR 2.02, 95% CI: 1.54, 2.63). The combination of falls risk and alcohol consumption increased multiple fracture odds (OR 7.62, 95% CI: 2.77, 20.94). Left femoral T-score and body fat percentage showed significant non-linear effects (both p < 0.001). Conclusion: Multiple fragility fractures were associated with many traditional risk factors. We also identified a novel link between falls risk and alcohol consumption, as well as the significant associations with body composition.