Eur Arch Psychiatry Clin Neurosci. 2026 May 19. doi: 10.1007/s00406-026-02264-4. Online ahead of print.
ABSTRACT
This systematic review and meta-analysis aimed to investigate the efficacy and safety of adding nuclear factor erythroid 2-related factor 2 (Nrf2) activators to antipsychotic treatments for individuals with schizophrenia. The primary outcome was overall symptom improvement. Other efficacy outcomes involve the improvement of positive symptoms, negative and depressive symptoms, Positive and Negative Syndrome Scale general subscale (PANSS-G) scores, and cognitive function composite scores and individual cognitive domain scores (only for sulforaphane). Further, we included all-cause discontinuation, adverse event-related discontinuation, incidence of individual adverse events, and changes in blood inflammatory biomarkers, lipid profiles, and glucose levels. Our review included 16 randomized-controlled trials (n = 953, mean age = 35.27 years; 66.74% male). Among these, 15 were placebo-controlled, while the remaining study did not utilize a placebo. This systematic review comprised six studies on ascorbic acid, four on curcumin, two on resveratrol, and four on sulforaphane. Curcumin showed marginal superiority over placebo for the primary outcome (standardized mean difference = - 0.53, 95% confidence interval: -1.08, 0.02); however, non-significant subgroup results (all p > 0.05) with wide confidence intervals across formulations and co-interventions suggest that these findings lack robustness. Curcumin significantly outperformed placebo in reducing positive symptoms, negative symptoms, and PANSS-G scores but did not differ from placebo for other outcomes. Sulforaphane showed a lower all-cause discontinuation rate than placebo, suggesting potential efficacy and better treatment adherence. Additionally, sulforaphane reduced total cholesterol, low-density lipoprotein cholesterol, and triglyceride levels, though no significant differences were found in other outcomes. Meta-analysis of efficacy outcomes for ascorbic acid and resveratrol could not be conducted due to insufficient data.
PMID:42154264 | DOI:10.1007/s00406-026-02264-4