Twenty-nine studies were included.
No randomised controlled trials were included. Areas of difference between studies included settings; sample size; characteristics of population; years of data collection; study design; and temporal relationship between medication ascertainment, fall ascertainment, and the index fall. 24% of studies did not provide a definition of a fall.
One or more falls for cardiac drugs:
Type 1A anti-arrythmics (10 studies, 2779 subjects) OR = 1.59 (95%CI: 1.02, 2.48); Digoxin (17 studies, 6121 subjects) OR = 1.22 (95%CI: 1.05, 1.42); ACE inhibitors (10 studies, 3612 subjects) OR = 1.20 (95%CI: 0.92, 1.58); Centrally acting anti- hypertensives (11 studies, 3422 subjects) OR = 1.16 (95%CI: 0.87, 1.55); Diuretics (26 studies, 20217 subjects) OR = 1.08 (95%CI: 1.02, 1.16); Thiazide diuretics (12 studies, 12475 subjects) OR = 1.06 (95%CI: 0.97, 1.16); Loop diuretics (11 studies, 2987 subjects) OR = 0.90 (95%CI: 0.73, 1.12); Nitrates (14 studies, 4967 subjects) OR = 1.13 (95%CI: 0.95, 1.36); Beta blockers (18 studies, 5551 subjects) OR 0.93 (95%CI: 0.77, 1.11); Calcium channel blockers (13 studies, 3929 subjects) OR = 0.94 ( 95%CI: 0.77, 1.14). No statistical heterogeneity was shown. One or more falls for analgesic drugs: Narcotics (13 studies, 4537 subjects) OR = 0.97 (95%CI: 0.78, 1.20 ); NSAIDs (13 studies, 4415 subjects) OR = 1.16 (95%CI: 0.97, 1.38); Aspirin (9 studies, 2269 subjects) OR = 1.12 (95%CI: 0.80, 1.57); Non-narcotics (9 studies, 2528 subjects) OR = 1.09 (95%CI: 0.88, 1.34). No statistical heterogeneity was shown.
No effect on OR was noted after stratification for the following factors: subject residence in the community or an institution; percentage of fallers 35% or greater in community studies; mean age of subjects 75 years or over; ascertainment of medications and falls at the time of the fall; and study design. A reduction in the size of ORs over time was found only for diuretics correlation coefficient (r) = -0.55, P = 0.004. Diuretic use by recurrent fallers compared with single or non fallers OR = 1.17 (95%CI:1.04, 1.32).
Very few studies adjusted for confounders. It was not possible to compare crude with adjusted ORs for particular medications.
Medications significantly associated with falls: Digoxin (9 cohort studies) RR = 1.14 (95%CI: 1.01, 1.29) vs OR = 1.29 (95%CI: 1.01, 1.65); Type 1A anti-arrythmics (5 cohort studies) RR = 0.99 (95%CI: 0.69, 1.42) vs OR = 0.95 (95%CI: 0.46, 1.97); Diuretics (9 cohort studies) RR = 1.03 (95%CI: 0.97, 1.09) vs OR = 1.05 (95%CI: 0.96, 1.14). Subjects reporting the use of more than 3 or 4 medications of any type were at increased risk of recurrent falls.