Abstract
BACKGROUND: Recent prescribing practices have shifted from benzodiazepines (BZs) toward non-GABAergic hypnotics, including dual orexin receptor antagonists (DORAs) and melatonin receptor agonists (MRAs). We examined whether hypnotics involved in suicide attempts changed over time in a context-dependent manner. METHODS: We conducted a multicenter retrospective cohort study of consecutive patients presenting with suicide attempts at three hospitals in Japan between April 2020 and March 2025. Hypnotics involved in attempts were identified from empty medication packages collected at presentation. Annual proportions of BZs and non-GABAergic hypnotics (OMs: DORAs and MRAs) were analyzed using Cochran-Armitage trend tests under three conditions: (1) all suicide attempts, (2) overdose-related attempts, and (3) overdose-related attempts involving a hypnotic. Additional analyses separated DORAs from MRAs. RESULTS: Among 1111 suicide attempt encounters, 648 were overdose-related. OM involvement increased significantly over time across all denominators. In contrast, BZ involvement declined significantly only among overdose-related attempts. When OMs were disaggregated, DORA involvement showed a significant upward trend using overdose-related attempts as the denominator (χ(2) = 7.3048, p = 0.006877) and using all suicide attempts as the denominator (χ(2) = 7.6384, p = 0.005714). MRA (ramelteon) involvement did not show significant temporal change in either analysis. Overall, the increase in OM involvement was primarily attributable to DORAs. CONCLUSION: Hypnotics involved in suicide attempts changed in a context-dependent manner during the study period. The increase in non-GABAergic hypnotics was driven by DORAs, whereas reductions in BZ involvement were detectable only in overdose-related contexts. These findings suggest that evolving hypnotic availability may influence the profile of medications involved in self-poisoning.