Use of Gabapentinoids and Risk of Hip Fractures: A Population-based Case-crossover and Case-case-time-control Study

Background: Use of gabapentinoids has been increasing globally, along with reports of misuse and poisoning. Population at risk for hip fractures is also increasing with our ageing population. However, the risk of hip fractures associated with gabapentinoids remains controversial with conflicting results.

Objectives: To investigate the association between gabapentinoids and the risk of hip fractures using self-controlled study design

Methods: We conducted a case-crossover study in all patients admitted for hip fracture in Victoria, Australia from 2013-18 who were dispensed ≥1 gabapentinoid prior to admission. Data was extracted from our linked population-based datasets, Victorian Admission Episodes Dataset and Pharmaceutical Benefits Scheme. The index date was defined as admission for first hip fracture; the index period as 1 to 60 days before index date; the reference period as 121 to 180 days before index date. Conditional logistic regression was used to estimate the odds ratio (OR) and 95% confidence interval (CI) for dispensing of gabapentinoids in the index period compared to the reference period. To adjust for the exposure-time trend, we performed case-case-time-control study by matching each index case with 5 “future cases” (i.e. first hip fracture 60 to 365 days after the index date) of the same age and sex. The exposure-time trend-adjusted OR was estimated using an interaction term. Model was also adjusted for time-varying exposures of other central nervous system active drugs. Sensitivity analyses were conducted by extending washout period between index and reference periods from 60 days to 90, 120, and 150 days.

Results: Among 3,124 patients dispensed gabapentinoids prior to admission for hip fractures, 644 patients (36% aged ≥85 years, 70% female, 97% used pregabalin) had crossover in exposure across index and reference periods (i.e. dispensed gabapentinoids in index (n=425) or reference (n=219) periods alone). The case-crossover analysis showed an association between dispensing gabapentinoids and a higher risk of hip fractures (OR 1.94, 95% CI 1.65-2.28). After adjusting for exposure-time trend, the risk of hip fractures was attenuated but remained elevated (OR 1.50, 95% CI 1.24-1.81). Additional adjustment for time-varying covariates produced similar estimate (OR 1.45, 95% CI 1.20-1.75). In the age-stratified analysis, higher risk was found in patients aged ≥85 years (OR 1.83, 95% CI 1.35-2.50), while the risk became insignificant for patients aged < 85 years (OR 1.18, 95% CI 0.93-1.50). Sensitivity analyses showed our results were robust.

Conclusions: Use of gabapentinoids was associated with a higher risk of hip fractures. Furthermore, patients aged 85 years or older were more susceptible to the risk of hip fractures associated with gabapentinoids.