Methylphenidate for attention-deficit/hyperactivity disorder: Too much of a good thing?

Methylphenidate has the longest history of any medication in child psychiatry. Its behavioural effects were noticed serendipitously when it was used to treat post-encephalitic headaches, and it soon became widely accepted as a treatment for hyperactive reaction of childhood in the 1960s. Nowadays, it is the most prescribed drug for the treatment of attention-deficit/hyperactivity disorder (ADHD). Consequently, the use of methylphenidate for ADHD is one of the most frequently studied treatments in medicine – with many primary studies and meta-analytic reviews. However, despite this long history, its use in children and adolescents in particular remains controversial.

Thus, the recent publication of a Cochrane review on the topic (Storebo et al., 2015) is timely, and one would hope that it would shed light rather than more heat on this subject. The review’s overall conclusion that ‘methylphenidate may improve ADHD symptomatology’ (italics added) is measured and cautious, whereas, in contrast, the authors’ other conclusions are less ambiguous.

First, despite there being 38 parallel group trials and 147 cross-over trials involving over 12,000 participants, the evidence was rated to be of very low quality. The reviewers departed from usual Cochrane protocol by adding the item ‘vested interest’ to the suite of quality measures. Studies were deemed low quality if they performed poorly in one domain. Close examination of the review shows it was most often ‘vested interest’ that resulted in studies being rated low in quality. In addition, the data had a serious risk of bias, there was under-reporting of relevant outcome data and there was significant heterogeneity in the results of trials. If performed well, a Cochrane review process should be able to account for reporting bias. In the Storebo review, the reviewers could have conducted a sensitivity analysis to determine whether vested interest did have an impact on the estimates of effect size, but this was not done.

Second, and of more concern, there were no eligible trials that examined the effects of more extended exposure on children’s general behaviour. In fact, the average duration of the 38 parallel trials was only 75 days. In practice, ADHD is considered to be a chronic illness, and therefore, methylphenidate is usually prescribed long-term and, typically, adolescents and children are administered the medication for at least 6 months (Hodgkins et al., 2011). The lack of evidence for extended efficacy is a major concern, although gathering such evidence is a considerable challenge (Hazell, 2011). Another systematic review of long-term outcomes for the management of ADHD in children that presumably used more relaxed inclusion criteria than the Cochrane review is that by Parker et al. (2013). The authors concluded that besides the paucity of data, there was no evidence to suggest that short-term effects of treatment were maintained. Over half of the total number of participants in the review were from the Multimodal Treatment Study of Children with ADHD (MTA) – a 14-month randomised controlled trial (RCT) in which intensive behavioural, medication and multimodal treatment arms were compared relative to each other and a treatment as usual community intervention. The MTA study initially suggested the short-term superiority of well-delivered medication treatment for ADHD and significantly influenced US treatment guidelines for ADHD.

Regarding long-term side effects, the Cochrane review reports evidence of a 29% increase in the overall risk of non-serious adverse events, most commonly sleep problems and decreased appetite. But the overall quality of this evidence is very low, and the authors considered it likely that the real increase in adverse events is higher. Generally, the use of methylphenidate does not appear to be associated with an increased occurrence of serious adverse events, but again the evidence is very poor and remarkably only 4.9% of trials captured relevant data. The main persistent effect of methylphenidate treatment in the MTA trial appears to be a nearly 1 in (2 cm) loss of height in children on extended dosing. The height loss reported in the MTA study is an obvious concern.

In their interpretation of the findings, the reviewers assert the studies were easily unblinded because methylphenidate has recognisable side effects. While plausible, the claim was not substantiated, and interestingly, the reviewers did not examine the studies to find out whether teachers determine the participant’s treatment allocation at greater than chance levels. It is very possible that these types of data were not collected from teachers, given the challenges of obtaining a full set of questionnaires from them.

Of note, the MTA authors are now more circumspect, reporting that the initial advantages for optimally medicated participants had halved 10 months later and disappeared a year after this. This pattern of initial superiority of medication-based treatment that tapered and then evaporated has continued across all subsequent years of follow-up. The MTA authors now state that medication may not be an extended hoped for panacea and that a ‘reconsideration’ of the MTA findings may be necessary (Hinshaw et al., 2015). This and the new review by Storebo and colleagues should make us reflect again on the benefits versus risks of prescribing methylphenidate for children with ADHD. While the evidence base appears large, the quality of evidence is poor. Importantly, the evidence base is only for short-term use of methylphenidate and suggests, at best, a modest effect. The low quality of evidence means we remain uncertain of the reliability of the estimates of treatment benefit. In addition, methylphenidate is associated with increased risks of adverse events, and these events are likely to be underestimated.

Of more concern is that there is an absence of evidence to support the long-term prescription of methylphenidate. This is how methylphenidate is most commonly used in clinical practice. Children may show short-term reduction in symptoms but then remain on medication for years. The few long-term studies suggest that extended treatment is not associated with educational benefits or improvements in a child’s quality of life and may be associated with an increase in the risk of harm. As such, clinicians should heed guideline recommendations to prescribe methylphenidate only after the diagnosis of ADHD has been established with a high degree of confidence and certainty and to regularly review the need for treatment continuation.