Sleep-Associated Adverse Events During Methylphenidate Treatment of Attention-Deficit/Hyperactivity Disorder: A Meta-Analysis.

OBJECTIVE

Sleep disturbances are a feature of attention-deficit/hyperactivity disorder (ADHD) and an adverse event (AE) of methylphenidate treatment. The authors sought to clarify methylphenidate-associated sleep problems and how studies are affected by confounding factors.

DATA SOURCES

Published studies in English collected via online databases and unpublished data from www.clinicaltrials.gov and US Food and Drug Administration websites. Sources were searched from inception to August 2017.

STUDY SELECTION

Included were blinded placebo-controlled studies of youth with ADHD conducted in naturalistic settings, leading to 35 studies yielding 75 observations of sleep-related AEs. These studies comprised 3,079 drug-exposed and 2,606 placebo-treated patients.

DATA EXTRACTION

Two PhD-level reviewers reviewed each study for inclusion. Four PhD/PharmD-level reviewers extracted data in duplicate. Discrepancies were resolved by discussion or, if needed, by the senior author.

RESULTS

Increased pooled relative risks (RRs) were found for methylphenidate-associated sleep-related AEs for insomnia (general), initial insomnia, middle insomnia, combined insomnia, and sleep disorder. Several sample or study design features were significantly associated with the RR for sleep-related AEs and the methylphenidate formulation studied (P < .05). After correction for confounding variables, significant differences among drugs were found for initial insomnia, insomnia (general), and sleep disorder (P < .0001) as the other categories could not be tested due to insufficient studies. The findings also show that the RR and its interpretation are constrained by the placebo AE rate.

CONCLUSIONS

Several types of insomnia and sleep problems are associated with methylphenidate treatment. Study design and sample features influence the RR statistic. By showing that the rate of placebo AEs impacts the RR, this study provides the field with a useful covariate for adjusting RR statistics.