DOI: 10.3390/jcm15166224 ISSN: 2077-0383

Antiresorptive Use and Risk of Hip Fracture in Older Women with Osteoporosis: Findings from a Real-World National Database of 21,332 Women

Bernardo A. Cedeno-Veloz, Juan Erviti, Marta Gutiérrez-Valencia, Leire Leache, Luis Carlos Saiz, Alba M. Rodríguez García, Nicolás Martínez-Velilla

Background: Evidence on the real-world effectiveness of antiresorptive therapy for hip fracture prevention in very old women remains limited. We aimed to assess the association between antiresorptive use and incident hip fracture among age ≥ 75 older women with osteoporosis. Methods: We conducted a population-based nested case–control study using the Base de Datos para la Investigación Farmacoepidemiológica en Ámbito Público (BIFAP), a Spanish research database. Women aged 75 years or older with osteoporosis and no previous antiresorptive use at cohort entry were followed from 2010 to 2022. Incident hip fracture cases were matched to risk-set controls by age, autonomous region, and calendar year of database registration. Antiresorptive exposure was classified as current, recent, past, or never use. Conditional logistic regression was used to estimate adjusted odds ratios (aORs) and confidence intervals. Results: The study included 2057 incident hip fracture cases and 19,275 matched controls. Compared with never users, current antiresorptive users had lower odds of hip fracture (aOR 0.76; 0.65–0.87), whereas past users had higher odds (aOR 1.91; 1.57–2.33). Ever use was not associated with hip fracture risk compared with never use (aOR 1.01; 0.90–1.13). Conclusions: In this real-world nested case–control study of women aged ≥75 years with osteoporosis, current antiresorptive use was associated with lower odds of hip fracture, whereas past use was associated with higher odds. Because treatment initiation, persistence, and discontinuation are strongly influenced by baseline fracture risk, osteoporosis severity, frailty, and clinical surveillance, these findings should be interpreted as associations rather than causal treatment effects.

More from our Archive