Review of: Cortese S et al, Lancet 208;5:727–738
Type of study: Meta-analysis

With so many medications available to treat ADHD, wouldn’t it be nice to know if some are better than others? In this comprehensive meta-analysis, researchers sought to compare the relative efficacy and tolerability of both stimulant (methylphenidate and amphetamines) and non-stimulant (atomoxetine, bupropion, modafinil, clonidine, and guanfacine) medications for ADHD in children and adults.

The investigators combed through published and unpublished databases and located 82 double-blind, randomized controlled trials in children and adolescents, and 52 such trials in adults. Together, they included over 10,000 children and adolescents, and over 8,000 adults. The primary outcome was change in ­clinician-rated ADHD symptoms, while teacher ratings were also evaluated for children. “Tolerability” was defined as the percentage who dropped out because of side effects, while the broader term “acceptability” referred to those who dropped out for any reason. Outcomes were evaluated through 12 weeks of treatment.

In children and adolescents, all medications were superior to placebo. Amphetamines emerged as the most effective ADHD medication, superior to modafinil, guanfacine, atomoxetine, and methylphenidate. Methylphenidate was superior to atomoxetine. Based on teacher ratings, only methylphenidate and modafinil separated from placebo (none of the amphetamine trials included teacher ratings). With respect to tolerability, amphetamines and guanfacine both displayed significantly more adverse effects than placebo; amphetamines also significantly increased diastolic blood pressure. Methylphenidate was better-tolerated than the amphetamines, and it was the only medication with better acceptability than placebo.

In adults, amphetamines emerged not only as the most efficacious agents but also the only ones with better acceptability than placebo. Methylphenidate, atomoxetine, and bupropion all had similar effect sizes. Clonidine and guanfacine did not have data in adults, and modafinil was ineffective in this population, despite having positive results in children. Tolerability was similar among the agents. In contrast to their effects on children, amphetamines did not increase diastolic blood pressure in adults. Overall, ADHD medications were less efficacious and less well-tolerated in adults than in children and adolescents.

What are the weaknesses? There was a dearth of head-to-head trials, so these comparisons could only be made indirectly. The dropout rate was used as a proxy for acceptability, and this is a rough estimate. Finally, while the large sample sizes instill greater confidence in the results, they also risk finding significant differences that may not necessarily be clinically meaningful.

TCPR’s Take
It’s rare for one medication to stand out in its class, but the amphetamines clearly emerged as the most effective option in both children and adults. That does not mean they should always be first choice, though. Methylphenidate was a more tolerable option in children, and there will always be patients who respond better to the methylphenidate varieties. Non-­stimulant options take longer to work, but they performed fairly well in this ­meta-analysis, sometimes rivaling methylphenidate’s benefits. The only failure was modafinil, which worked in children but not adults.