Long-acting intranasal insulin for the treatment of delirium—a randomised clinical trial

Background Delirium affects up to 25% of hospitalised older patients; however, there are no effective pharmacological treatments. Accumulating evidence of brain insulin resistance and altered cerebral glucose metabolism during delirium present a promising therapeutic target. Objective To assess the safety and efficacy of intranasal insulin in the treatment of delirium. Design Single-centre, randomised, double-blind, placebo-controlled trial. Setting Two geriatric medicine wards in a tertiary hospital. Participants 100 patients aged over 64 years presenting to hospital and admitted under geriatric medicine with delirium. Interventions Participants were randomised in a 1:1 ratio to receive 20 IU of long-acting insulin or placebo intranasally twice daily until delirium resolution, hospital discharge or intervention futility defined by prespecified criteria. Main outcome The primary outcome was delirium duration, assessed daily using the Confusion Assessment Method. Secondary outcomes included acute length of stay (LOS), delirium severity, antipsychotic use, hospital complications and mortality. Results The intention-to-treat analysis included 97 participants [intranasal insulin n = 48, control n = 49; mean (SD) age, 87.6 (7.0) years; 63% female]. Baseline characteristics were similar between groups. Median delirium duration [days (IQR)] was 4.8 [2.9, 9.2] for intranasal insulin and 6.8 [4.0, 9.8] for the control (HR 0.7, 95% CI 0.43–1.15; P = .16). Median acute LOS (days) was 7.9 [4.6, 14.5] for intranasal insulin and 12.9 [6.9, 16.8] for the control (HR 0.56, 0.35–0.89; P = .014). No significant differences were observed in other secondary outcomes. Intranasal insulin demonstrated favourable tolerability. Overall, 86% of the participants were compliant with the intervention (≥80% of doses). Prespecified subgroup analysis revealed an age-related response, with participants aged ≤88 years showing shorter delirium duration with intranasal insulin [n = 46; intranasal insulin: median 3.9 (IQR 2.9, 6.9) days vs control: 7.0 (4.7, 9.7); HR 0.34, 0.16–0.74; P = .006], whereas no difference was observed in participants aged >88 years [n = 51; intranasal insulin 5.4 (2.9, 11.1) vs control 4.9 (2.6, 12.9); HR 0.87, 0.39–1.94; P = .73]. Conclusion and relevance This is the first study of intranasal insulin for delirium treatment. The reduced LOS combined with the observed age-related effects warrants further investigation into the clinical potential of intranasal insulin in managing delirium in older patients. Trial registration ACTRN 12618000318280.