Short-acting benzodiazepines have been touted as safer for use in the elderly because their longer-acting counterparts are clearly associated with a higher risk of falls and a higher rate of injury after a fall. However, little information is available about the relative safety of benzodiazepines with a shorter elimination half-life. Ray and colleagues conducted this historical cohort study to determine whether there was a difference in the rate of falls between nursing home residents taking long elimination half-life or short elimination half-life benzodiazepines.

Data from previous studies evaluating the role of antidepressants on the rate of falls in nursing home residents were reviewed. The study population included 2,510 patients who were 65 years or older in 53 nursing homes in Tennessee. Information on medication use was abstracted from medication administration records. The categories of benzodiazepine use included none (none during the past week), recent (none taken on the current day but taken at least once during the past week) and current (taken on the day in question). Benzodiazepines were further classified by elimination half-life, dosage and duration of use. The long elimination half-life benzodiazepines (at least 24 hours) included diazepam, chlordiazepoxide, clorazepate and flurazepam. Benzodiazepines with an intermediate elimination half-life (12 to 23 hours) included lorazepam, alprazolam, clonazepam and estazolam, and those with a short elimination half-life (less than 12 hours) included temazepam, oxazepam and triazolam. Zolpidem, although not a benzodiazepine, was also included as a short-acting agent because its effect on the gamma-aminobutyric acid receptor complex causes it to produce psycho-motor impairment similar to that occurring with benzodiazepine use. Use of physical restraints and the number of falls during the previous 90 days were abstracted from medical records. Cognitive impairment, incontinence and the ability to perform other daily-living activities were also recorded. Falls were classified as daytime falls if they occurred between 7:00 a.m. and 7:59 p.m., with all other falls classified as nighttime falls.

One third of the 2,510 residents assessed had taken at least one benzodiazepine during the mean follow-up period of 225 days. Patients who received benzodiazepines were nearly twice as likely to fall as those who did not receive this class of medication; even after adjusting for this difference, those residents who received benzodiazepines had a 44 percent greater fall rate than those who did not. During the 1,544 person-years of follow-up, 3,706 falls occurred, of which 116 were associated with fractures and 83 with lacerations requiring sutures, and there were 23 other major injuries. For the 3,507 falls for which the time of occurrence was recorded, the rate of falls was 173 per 100 during the day and 54 per 100 during the night. Increasing the dose of benzodiazepine was positively correlated with an increased rate of falls. The risk of falling was highest during the week following the beginning of benzodiazepine therapy but remained elevated after one month of treatment. Although the risk of falls increased with the increase in the elimination half-life of the benzodiazepine selected, even the short elimination half-life agents were associated with a twofold increased risk of falls during the night.

The authors conclude that benzodiazepine therapy should be used cautiously in the elderly because even the short elimination half-life agents are associated with a significant increase in the risk of falling in this population. Even when the benzodiazepine prescribed as a sedative is short-acting, particular steps should be taken to prevent falls.