Intranasal oxytocin shows no significant effect on alcohol-related outcomes in meta-analysis of six RCTs

Despite promising preclinical data, randomized controlled trials (RCTs) investigating intranasal oxytocin (OT) as an adjunctive treatment for Alcohol Use Disorder (AUD) have yielded mixed results. Current standard-of-care treatments for AUD often have limited efficacy or adherence, creating a need for novel therapeutic strategies. Oxytocin, a neuropeptide involved in social bonding and stress regulation, has been hypothesized to modulate reward pathways and reduce alcohol craving, making it a candidate for AUD intervention. This meta-analysis aimed to clarify its therapeutic potential by synthesizing existing RCT data.

Researchers conducted a multilevel random-effects meta-analysis of six eligible RCTs comparing intranasal oxytocin with placebo for alcohol-related outcomes. Hedges' g values were calculated and winsorised at |g| = 3 to mitigate extreme small-sample effects. Moderator analyses explored factors like outcome domain, OT dose, treatment duration, administration frequency, and clinical setting. Publication bias was assessed using multilevel PET-PEESE, Egger's test, trim-and-fill, and limit meta-analysis. Bayesian multilevel models were employed to examine average treatment effects and outcome variability.

The overall pooled effect of intranasal oxytocin was not statistically significant, with a Hedges' g of 0.34 (95% CI -0.48-1.17, p = 0.47). Substantial between-study heterogeneity was observed (Q(48) = 504.40, p < .001). Cook's distance identified one study (Pedersen et al., 2013) as statistically influential. After excluding this study, analyses on the remaining five studies still found no significant moderation by outcome domain, dose, duration, frequency, or setting. A nominally significant positive association with effect size was noted for year of publication in the restricted sample (β = 0.184, p = .042). Multiple publication bias diagnostics consistently indicated the absence of a systematic adjusted effect.

Bayesian multilevel analysis further confirmed the absence of a credible treatment effect (posterior mean μ = -0.005, 95% CrI -0.53-0.52), with robust Bayesian meta-analysis providing moderate support for the null hypothesis (BF₀₁ = 4.74). Variability analyses found no evidence that oxytocin increased outcome dispersion relative to placebo.