by James O’Hanlon, Ph.D., Psychopharmacologist

Children with ASD are commonly hyperactive. This symptom is disruptive at home, limits learning in school and interferes with behavioral therapy. Of course many typically developing children with the diagnosis of Attention Deficit/Hyperactivity Disorder (ADHD) present the same symptoms, which besides physical restlessness include deficient attention, distractibility and impulsivity. On the basis of appearances, many think that the cause of hyperactivity is the same in both disorders and therefore efficacious medication in one should be as efficacious in the other. Appearances can be deceptive. Trials of three drugs approved for treating hyperactivity in ADHD were recently undertaken with individuals, aged 5-17 yr., who suffer from ASD. Collectively, the results indicate that drugs approved for managing hyperactivity in ADHD are not always appropriate for treating the same symptoms in ASD.

Drugs approved in the USA for the treatment of ADHD include various amphetamines (Adderall, Vyvanse, others), methylphenidate (Concerta, Focalin, Ritalin, others), atomoxetine (Strattera) and guanfacine (Tenex, Intuniv). The amphetamines need not concern us here. The last assessments of amphetamine in children with autism were made in the late 1970’s. At that time autism was thought to be a psychotic disorder and amphetamines were said to cause a “worsening of psychosis”. The drugs also exacerbated hyperactivity in the majority of children treated. The results were so daunting that no one wished to pursue studies of amphetamine in the ASD population. But amphetamines are pharmacologically unique and drugs with different mechanisms of therapeutic activity are now available for treating ADHD. The question of whether the benefits of three of them – methylphenidate, atomoxetine and guanfacine – extend to ASD, has recently been answered in separate trials involving children with that disorder, aged 5-17 yr. 1, 2, 3

The methylphenidate trial was conducted in three stages. First was a sensitivity check wherein all 72 participants were given day-long exposures to each of three therapeutic doses. Six became agitated and dropped out. The remaining 66 children entered a double-blind, crossover trial wherein each dose and placebo was given for a week. Again, some children dropped out due to side effects and others did not respond favorably to any methylphenidate dose. The 35 individuals who responded favorably to one or more doses continued taking individualized “optimal doses” for the next 8 weeks. The atomoxetine and guanfacine trials followed a double-blind, parallel group design. Respectively, 97 and 62 children were randomly assigned to drug or placebo groups in approximately equal numbers. Atomoxetine doses were gradually increased over 3 weeks to maximum dose that continued over the next 5 weeks. A similar dose escalation was possible in the 8-week guanfacine trial but the progression could be reversed by returning to the preceding dose to cope with emerging side effects. Drowsiness, fatigue, emotionality or irritability led to dose reduction in 30% of the group with the rest going on to maximum dose.

Statistical tests showed that by the end of the trial, each drug was “significantly” more effective than placebo for reducing hyperactivity. But statistical significance only means that the final mean difference between drug and placebo is unlikely to be due to chance. Another statistic, “effect size”, is of greater clinical relevance because it measures the magnitude of the mean difference relative to individual differences that existed prior to treatment.  Effect size typically varies in drug trials from about 0.2 to 2.00, in practical terms, from a clinically irrelevant effect to an overwhelmingly strong indication of drug efficacy. Methylphenidate in the optimum dose and atomoxetine had similar effect sizes of about 0.9 – good but not exceptional. Guanfacine had an excellent effect size of 1.67. Another indication of efficacy is the relative frequency of individuals, called “responders”, whose reduction in target symptoms exceeds a standard criterion during drug treatment. Responder frequency was only 21% in the atomoxetine trial but about 50% did in both the methylphenidate and guanfacine trials.

The choice among alternative medications is not only determined from their respective efficacies. It also depends upon comparison of their side effect profiles. As generally defined, serious side effects range from those requiring temporary hospitalization to those constituting a threat to life. Most side effects occurring in clinical trials are mild-moderate in severity and abate as treatment continues. They are still distressing to the participants and in the present context, to their parents as well. The net cumulative effect of all side effects is said to reduce a drug’s  “tolerability”. The simplest way to compare different drugs’ tolerabilities is from the respective percentages of individuals who discontinued treatment due to side effects. In these three trials, 18% discontinued methylphenidate, 13% guanfacine and 2% atomoxetine. One guanfacine side effect was judged serious: i.e., physical aggression of a 6 ¾ -year-old boy, requiring hospitalization for 3 days.

Guanfacine emerged as the most efficacious drug for managing hyperactivity in children with ASD. Yet almost as much attention was given for adjusting guanfacine doses as for methylphenidate to achieve a balance between efficacy and tolerability. Neither drug can be effectively used without careful medical monitoring and dose adjustment. In contrast, atomoxetine was the least efficacious though well tolerated across its pediatric dose range. It is important to note that the same order of efficacy does not apply to typically developing children with ADHD. For them, methylphenidate (or amphetamine) is recognized as most effective drug for reducing hyperactivity, atomoxetine less and guanfacine least of all. It seems the pathophysiological processes that cause hyperactivity in ASD and ADHD are different.

1 Research Units on Pediatric Psychopharmacology Autism Network. Archives of General Psychiatry 2005;62:1266-1274

2  Harfterkamp M et al. Journal of the American Academy of Child & Adolescent Psychiatry 2012;51:733-741

3  Scahill L et al. American Journal of Psychiatry 2015;172:1197-1206