To establish association of specific central nervous system (CNS) drugs with recurrent falling.

Observational cohort study conducted in seven nursing homes (NH) for 822 residents in city of Krakow, Poland. Prospective calendar-recorded registration of all falls during the study period (18 months) was performed by nurses in one hour after each fall. Among other data, age, gender and drugs taken in the day of fall were collected. Poisson regression model was used to establish drugs enhancing the rate of falling.

During the study period, 289 falls were registered in 162 nursing home residents: 61 falls in 29 residents of NH for psychiatric patients and 228 falls in 133 residents of NHs for older and chronically ill people. According to univariate analysis, the rate of falling increased with use of antiepileptics (cIRR=1.60; P<0.001), number of CNS drugs (crude incident rate ratio [cIRR]=1.16; P<0.001), and for residents of NH for psychiatric patients (cIRR=1.37; P<0.03)–despite their lower average age (61.9 vs. 78.4; P<0.001). After adjusting for age, gender, number of other than CNS drugs, type of NH and CNS drugs only antiepileptics (IRR=1.65; P<0.003) significantly increased the rate of falling. Anti-Parkinson drugs were near to influence it (IRR=1.41; 95% confidence interval=0.989–2.007; P=0.058).

NH residents taking antiepileptic drugs, and especially those taking more than one CNS drug or stay in NHs for persons with psychiatric illness, need more professionals’ attention due to a higher risk of recurrent falling.