Background: Previous studies have confirmed the contribution of various medications towards falls in the older population. Recently questions were raised as to whether the chronic illnesses or drug use was more important.
Objective: We attempt to test the hypothesis that underlying medical illnesses are the cause of falls rather than medications.
Design: Cross-sectional.
Setting: Urban community in Hong Kong.
Subjects: 4,000 Ambulatory community-dwelling men and women aged 65 years or over.
Methods: Demographic data, fall history in the previous 12 months, medical diagnoses, current medications and self-rated health were recorded. Body measurements and neuromuscular function tests were performed. Medical diagnoses and their corresponding medications were tested simultaneously in a multivariate model.
Results: 789 (19.7%) Subjects reported at least one fall and 235 (5.9%) experienced two or more falls. After adjustment for age and sex, medications associated with any falls were aspirin, diabetic drugs, nitrates, NSAIDs, and paracetamol, and those associated with recurrent falls were calcium channel blockers, diabetic drugs, nitrates, NSAIDs, aspirin and statins. Only anti-diabetics and nitrate showed moderate and borderline significance in multivariate analyses for recurrent and any falls respectively (OR 2.9, P = 0.01; OR 1.5, P = 0.027). Other medications failed to show significant relationship with falls, while eye diseases, heart diseases and musculoskeletal pain showed variable associations.
Conclusion: The apparent association between many medications and falls was mediated through the underlying medical diagnoses and neuromuscular impairment. Anti-diabetics agents were associated with falls.