NEW YORK (Reuters Health) – Stimulant medications show greater efficacy than nonstimulants for treating attention-deficit/hyperactivity disorder (ADHD) in adults, according to a new meta-analysis.

Study authors Dr. Stephen V. Faraone and Dr. Stephen J. Glatt note that stimulants have been the mainstay of ADHD pharmacotherapy for decades. Several nonstimulant medications have also been shown effective – tricyclic antidepressants, bupropion, modafinil, monoamine oxidase inhibitors, guanfacine, atomoxetine, and clonidine.

However, it’s been difficult to compare different drugs or drug classes due to the lack of head-to-head trials.

Drs. Faraone and Glatt, both from SUNY Upstate Medical University in Syracuse, New York, searched the medical literature for double-blind, placebo-controlled studies of ADHD in adults published after 1979, in which subjects were followed for at least 2 weeks. They identified 18 articles that involved more than 2000 subjects and evaluated 13 drugs.

In their report, which appeared online December 29 in the Journal of Clinical Psychiatry, effect size is expressed as the standardized mean difference, in which a 1-point difference is equivalent to 1 standard deviation on the outcome measure.

Five trials studied 4 long-acting stimulants (mixed amphetamine salts extended release, osmotic-release oral system methylphenidate, dexmethylphenidate extended release, and lisdexamfetamine dimesylate) in 637 participants. The mean effect size was 0.73.

The 3 short-acting stimulants evaluated in 7 trials (n = 459) were mixed amphetamine salts, dextroamphetamine, and methylphenidate. The effect size was 0.96.

Nine trials with 968 subjects examined the efficacy of 6 nonstimulants (atomoxetine, bupropion SR, modafinil, bupropion XL, paroxetine, and an investigational drug ABT-418). The mean effect size was 0.39.

There was significant heterogeneity among effect sizes, and evidence of publication bias, only in the trials of short-acting stimulants. Correction for publication bias revealed an effect size of 0.86, which did not differ significantly from that of long-acting stimulants.

There was also no significant difference between methylphenidate- and amphetamine-based medications.

On the other hand, the effect sizes for nonstimulant medications were significantly less than those for long-acting stimulants. After correcting for differences in study design, nonstimulants did not differ from short-acting stimulants.

Dr. Faraone and Dr. Glatt note that the effect sizes are robust and similar to those reported for medication treatment studies of children with ADHD. Only one medication, paroxetine, was worse than placebo.

Numbers needed to treat (NNT) ranged from 2 to 3 for long-acting stimulants, from 2 to 4 for short-acting stimulants, and from 2 to 5 for nonstimulants. Among nonstimulants, only modafinil had a NNT of less than 3.

The authors point out that the response rates are less than those reported in the literature, because the latter are not placebo adjusted.

“The NNT results…suggest that most patients will need more than 1 drug trial to achieve a positive outcome that cannot be attributed to a placebo effect.”

Still, in the absence of head-to-head trials, the researchers urge caution in comparing effects of different medications across studies.

The study was funded by a grant from Shire US, whose products include lisdexamfetamine dimesylate (Vyvanse), methylphenidate transdermal (Daytrana), methylphenidate hydrochloride (Equasym), and guanfacine (Intuniv).

Reference:
J Clin Psychiatry 2009.