We retrospectively analysed effects of bisphosphonates (BPs) on mortality in Austrian hip fracture patients. For 31 668 patients ≥50 years sustaining a hip fracture in Austria between July 2008 and December 2010, information on survival with follow-up until June 2011 and on prescription of BPs between July 2007 and June 2011 was available. Using Cox and logistic regression analysis, cumulative all-cause mortality among patients who started treatment before or after fracture was compared with that among age- and sex-matched hip fracture patients without anti-osteoporotic medication. The minimum prescription interval was set at half a year, and matched subjects had to be alive during the prescription interval of his/her assigned treated subject. Compared with female patients receiving no anti-osteoporotic prescription, women who initiated BPs before first fracture (n=8,868) displayed unaltered short-term mortality (hazard ratio (HR) at 90 days after fracture: 0.91 (95%-CI: 0.761.09, P=0.30)) but decreased long-term mortality (odds ratios (ORs) at one year and 3 years post-fracture, respectively: 0.70 (0.620.79, P<0.0001), 0.68 (0.610.76, P<0.0001)). Women starting BPs after first fracture (n=3,216) exhibited relative HRs of 0.29 (0.160.55, P<0.001) and 0.39 (0.290.52, P<0.0001) 1 year and 3 years post-fracture, respectively. For males using BPs already before fracture (n=837), no statistically significant reduction in mortality emerged, however, lowered mortality at one year post-fracture was observed for men treated only after fracture (n=633) (HR 0.12 (0.020.88), P<0.05). Among hip fracture patients using BPs, mortality was reduced predominantly in females. The smaller effect of BPs on pre-fracture users relative to post-fracture users survival might reflect a selection bias inherent to this observational study with more co-morbidity among BP users than non-users. However, the high extent of mortality reduction found in post-fracture BP users portends a causal relationship with anti-resorptive treatment with BPs.
Disclosure: The authors declared no competing interests.