- Current RCT evidence provides limited, inconsistent guidance on clinical predictors and moderators of pharmacological efficacy and tolerability in ADHD.
- Age, psychiatric comorbidities, sex, baseline symptom severity and ADHD presentation were most frequently examined and showed preliminary positive associations.
- Future studies should use individual participant data meta-analyses and well powered observational datasets to enable robust predictive models for treatment stratification.
CNS Drugs. 2026 Aug 13. doi: 10.1007/s40263-026-01323-7. Online ahead of print.
ABSTRACT
BACKGROUND AND OBJECTIVE: Randomised controlled trials (RCTs) have demonstrated that pharmacotherapy reduces symptoms of attention-deficit/hyperactivity disorder (ADHD) at a group level, but efficacy and tolerability vary across individuals. Clinico-demographic characteristics may act as predictors and/or moderators of treatment efficacy and tolerability, but systematic evidence remains limited. Therefore, we systematically analysed RCTs of ADHD medications to identify potential demographic and clinical predictors/moderators of efficacy and tolerability across the lifespan.
METHODS: Randomised controlled trials were identified from the MED-ADHD database ( https://med-adhd.org/ ), a repository of RCTs of medications for ADHD in children, adolescents and adults. The database is based on systematic searches of multiple electronic sources, including PubMed, BIOSIS Previews, CINAHL, the Cochrane Central Registry of Controlled Trials and EMBASE, and is also complemented by unpublished data obtained from manufacturers and study authors. We used the most recent (2026) version of MED-ADHD. The risk of bias was assessed using the revised Cochrane risk-of-bias tool (RoB 2).
RESULTS: Among the 171 RCTs screened, 62 assessed clinico-demographic factors as possible predictors or moderators of treatment efficacy and tolerability. Age was the most examined characteristic (56.4%) followed by, among the top five, psychiatric comorbidities (53.2%), sex (46.8%), baseline ADHD symptom severity (27.4%), and ADHD presentation (24.2%). Most RCTs reported no significant findings although there was preliminary evidence of associations with age (17.7%), psychiatric comorbidities (17.7%), baseline ADHD symptom severity (16.1%), sex (11.3%), and ADHD presentation (4.8%). Risk of bias was rated high in 37% of RCTs.
CONCLUSIONS: Stringent sample selection and reliance on group-level analyses may limit the power to detect predictors and moderators of treatment efficacy and tolerability in classical RCTs. Consequently, the current evidence provides limited and inconsistent guidance on clinically meaningful predictors/moderators. However, variables such as age, psychiatric comorbidities, sex, baseline severity and ADHD presentation have been the most frequently examined with positive findings and may hold promise. Future research should prioritise individual participant data meta-analyses of RCTs, alongside well-powered analyses of a comprehensive observational dataset, to more robustly investigate these associations and support treatment stratification through predictive models.
PMID:42595937 | DOI:10.1007/s40263-026-01323-7
Share Evidence Blueprint
Save to Google Notes
Search Google Scholar
Save as PDF
⭐ My Revision List
