Psychoactive medicine use and the risk of hip fracture in older people: A case-crossover study

Michael J. Leach, Nicole L. Pratt, Elizabeth E. Roughead

Research output: Contribution to journalArticleResearchpeer-review

14 Citations (Scopus)


Purpose: Many factors can increase the risk of hip fracture, including medicines. Traditional observational studies have assessed the risk, but results may be confounded by unmeasured factors. The case-crossover design is an alternative approach that controls for patient-specific, time-invariant confounding factors. This study aimed to assess associations between psychoactive medicines and hip fracture in the elderly using the case-crossover method. Methods: Data were sourced from the Australian Government Department of Veterans' Affairs healthcare claims database. The study population comprised beneficiaries aged over 65years who were hospitalised for hip fracture between 2009 and 2012. The case-crossover method was used to compare individuals' medicine exposure immediately before hip fracture (case window) with their exposure at an earlier time (control window). Conditional logistic regression was employed to quantify associations between medicine use and hip fracture. Alternative exposure windows were assessed in sensitivity analyses. Results: Eight thousand eight hundred twenty-eight patients were hospitalised for hip fracture. Their median age was 88years, and 63% were female. Odds of hip fracture were increased for opioids (odds ratio [OR]=1.62, 95% confidence interval [CI]=1.42-1.84), selective serotonin reuptake inhibitors (OR=1.54, 95% CI=1.28-1.86) and antipsychotics (OR=1.47, 95% CI=1.21-1.80). There was no significant association for tricyclic antidepressants in the primary analysis (OR=1.18, 95% CI=0.91-1.52), but there was a significant association in the sensitivity analysis using an alternative, earlier control window (OR=1.43, 95% CI=1.11-1.83). Conclusions: Increased risk of hip fracture in older people was found for opioids, selective serotonin reuptake inhibitors and antipsychotics. The choice of control window influenced the result for tricyclic antidepressants.

Original languageEnglish
Pages (from-to)576-582
Number of pages7
JournalPharmacoepidemiology and Drug Safety
Issue number6
Publication statusPublished - Jun 2015
Externally publishedYes


  • Case-crossover design
  • Elderly
  • Hip fracture
  • Pharmacoepidemiology
  • Psychoactive medicines
  • Sensitivity analysis

Cite this