The effects of anti-osteoporotic medication on fracture healing and outcomes: a systematic review and meta-analysis.

Abstract

Bisphosphonates and teriparatide are used in the management of osteoporosis and reduction of fracture risk. However, their effects on fracture healing have been subject to debate, with concerns that the antiresorptive action of these drugs might impair bone remodelling and delay union. This systematic review and meta-analysis aimed to assess the effects of anti-osteoporotic medication on fracture union, clinical outcomes, complications, and patient-reported outcome measures for acute fractures in adults. We searched MEDLINE (1946 to Sept 24, 2025), Embase (1974 to Sept 26, 2025), Cochrane Library (1946 to Sept 26, 2025), Web of Science (1900 to Sept 26, 2025), and Scopus (2000 to Sept 26, 2025) for randomised controlled trials involving patients aged 18 years or older who had sustained a fracture and had commenced on or were continuing anti-osteoporotic medicine. The included studies compared at least one form of anti-osteoporotic medication with another anti-osteoporotic medication or a placebo. Outcomes that were common to the studies included rates of non-union and delayed union of fractures, complications, overall time to fracture union, and visual analogue pain scores. χ², τ², and I² tests were conducted to evaluate heterogeneity. For I² values of 50% or more, subgroup analysis was considered; if subgroup analysis was not feasible, random-effects modelling was used. For I² values less than 50%, fixed-effect modelling was used. Risk ratios (RRs) and mean difference were calculated with 95% CIs. Standardised mean difference was used if there was substantial heterogeneity between outcome measures. This review is registered with PROSPERO (CRD42021230018). We identified 28 trials reporting on 5085 patients (3513 [69·1%] women and 1090 [21·4%] men; sex was not reported for 482 [9·5%] patients) for inclusion in the systematic review, of which 12 trials were suitable for meta-analysis. From pooled meta-analyses, no statistically significant differences were found between groups receiving bisphosphonate versus control for time to union (standardised mean difference 0·39, 95% CI -0·65 to 1·42; Z=0·73; p=0·47), delayed union (RR 1·24, 95% CI 0·82 to 1·89; Z=1·01; p=0·31), or non-union rates (0·71, 0·14 to 3·72; Z=0·40; p=0·69). Pooled analysis showed a significant reduction in time to union with use of teriparatide compared with a control (mean difference -3·03, 95% CI -3·61 to -2·45; Z=10·32; p<0·0001). There were no significant differences between delayed union rates (RR 0·72, 95% CI 0·18 to 2·84; Z=0·48; p=0·63) and Radiological Union Score for Hip scores (mean difference 0·58, 95% CI -0·46 to 1·62; Z=1·10; p=0·27) for patients receiving teriparatide versus control. On the basis of these analyses, bisphosphonate therapy did not appear to affect the time to or rate of fracture healing. Teriparatide does not appear to delay healing and might reduce the time to union for osteoporotic fractures, although further prospective evidence is needed. Heterogeneous reporting and variation in drug agents and doses used limited the comparison of other variables. Standardised reporting in this area is recommended. None.