Medication use and falls in community- dwelling older persons*

Background: the association between injurious falls requiring a visit to the emergency department and various classes of medications was examined in a case-control study of community living persons aged 66 years and older. Methods: administrative databases from an urban health region provided the information used. Five controls for each case were randomly selected from community dwelling older persons who had not reported an injurious fall to one of the six regional emergency departments in the study year. Two series of analyses on medication use within 30 days of the fall were conducted using logistic regression, the Wrst controlling for age, sex, and median income, the second controlling for co-morbid diagnoses as well. Results: during the study year there were 2,405 falls reported by 2,278 individuals to six regional emergency depart-ments giving a crude fall rate of 31.6 per 1,000 population per year. The initial analysis identiWed seven medication classes that were associated with an increased risk of an injurious fall, while controlling for age, gender and income. However, with further analyses controlling for the additional effects of co-morbid disease, narcotic pain-killers (odds ratio 1.68), anti-convulsants (odds ratio 1.51) and anti-depressants (odds ratio 1.46) were signiWcant independent predictors of sustaining an injurious fall. Conclusion: these results are based on a Canadian population-based study with a large community sample. The stud