Introduction This study systematically evaluated the safety profile of esketamine for treatment-resistant depression through a meta-analysis, focusing on dose-dependent adverse events and associated risk factors to inform precision dosing. Methods PubMed, Embase, the Cochrane Library, the Mainland China Biomedical Literature Database (CBM), the China National Knowledge Infrastructure (CNKI) and Wanfang databases were searched from inception to March 2025. Randomized controlled trials evaluating esketamine for treatment-resistant depression were included. Primary outcomes included the incidence of adverse events, discontinuation due to adverse events, and clinical response or remission. Statistical analysis was conducted using RevMan 5. 4. 1, with subgroup analyses by dosage, administration route, and geographic region (Mainland China vs. International multi-regional). Results Nine randomized controlled trials involving 1, 449 patients were included. Dosages ranged from 28 to 84 mg for nasal spray and 0. 20–0. 40 mg/kg for intravenous injection. Esketamine significantly increased the risk of nine adverse events, including nausea, dissociation, dizziness, vertigo, elevated blood pressure, and somnolence, compared with controls (P 0. 05). Risks were strongly dose-dependent: the high-dose group (≥56 mg or 0. 40 mg/kg) showed a greater risk than the low-dose group (≤28 mg or 0. 20 mg/kg), with RR for nausea of 3. 72 versus 1. 69 and RR for dissociation of 10. 65 versus 3. 27. Patients in International multi-regional studies also had higher risks of nausea, somnolence, and headache than those in Mainland China studies. Although esketamine improved the clinical response rate (RR = 1. 94), it increased treatment discontinuation due to adverse events by 2. 22-fold (P = 0. 025). Discussion Esketamine improves symptoms in patients with treatment-resistant depression but significantly increases dose-dependent adverse events. Clinical use should adopt personalized dosing strategies that balance efficacy and tolerability based on individual patient profiles. Systematic Review Registration https: //www. crd. york. ac. uk/prospero/displayᵣecord. php? RecordID=1024830, identifier CRD420251024830.
Qu et al. (2026) studied this question.