EVIDENCE LIBRARY
Fracture Cohort Associated Recent and Past Benzodiazepine Use With Mortality but Not Current Use
A Dutch cohort found a counterintuitive timing pattern after osteoporotic fracture. The authors considered survivor selection rather than concluding that current use was protective.
A Dutch cohort found a counterintuitive timing pattern after osteoporotic fracture. The authors considered survivor selection rather than concluding that current use was protective.
How exposure was classified
The study followed 4,854 older patients after a first major osteoporotic fracture, including 1,766 hip fractures. Follow-up was divided into thirty-day periods, with current, recent, and past medication exposure defined relative to each period.
Benzodiazepine dispensing increased during the first year after fracture. Recent and past use were associated with all-cause mortality, while current use did not show that association.
The timing pattern limits a simple mechanism
A missing current-use association cannot be interpreted automatically as a beneficial drug effect. People who remain on treatment may differ from those who stop or survive into later exposure categories. The authors discussed healthy-survivor bias as a possible explanation.
Lifestyle factors such as smoking and body mass were unavailable. Cause-specific death information was also lacking, so the study could not assign deaths to sedation, falls, respiratory effects, or another mechanism.
Use in testimony
Describe the exposure windows and all-cause endpoint explicitly. The paper illustrates the importance of time-varying treatment classification and selection. It cannot establish that stopping a benzodiazepine caused death or that continued use improved survival after an individual fracture.
Source paper
Original analysis of the supplied journal article, identified as PMC6422967.