OBJECTIVE: Non-GABAergic hypnotics introduced into clinical practice within the last two decades are pharmacologically presumed to have a lower risk of falls, but clinical investigations are scarce. We aimed to evaluate the risk of falls associated with different classes of hypnotics, namely benzodiazepines, ramelteon, suvorexant, and trazodone, in hospitalized individuals at a general hospital. METHOD: In this retrospective cohort study, data on the incidence of falls, hypnotic use, age, sex, body mass index (BMI), activities of daily living (ADL) score, presence of surgery, emergency admission, and ambulance use were collected for hospitalized patients aged 20 years or older who had been discharged from a tertiary general hospital in Japan between April 1, 2014, and March 31, 2019. The Cox proportional hazards model was used to examine factors associated with falls, adjusting for other demographics as covariates. RESULTS: Among 28,029 patients, 383 falls occurred in 322 patients. Ramelteon or suvorexant was not associated with an elevated incidence of falls (adjusted hazard ratio [aHR], 0.78
95 % CI, 0.34 to 1.81 and 0.44
95 % CI, 0.13-1.46, respectively), in contrast to benzodiazepines (aHR, 2.17
95 % CI 1.67-2.83) or trazodone (aHR, 1.96
95 % CI 1.25-3.07). Advanced age, lower BMI, wheelchair dependency, non-surgical status, absence of emergency admissions, and ambulance use were also associated with the elevated incidence of falls. CONCLUSIONS: In hospitalized patients at general hospitals, ramelteon and suvorexant may not increase the risk of falls, while the use of benzodiazepines and trazodone requires careful attention to minimize this risk.