Methylphenidate Versus Dexamphetamine in Children With Attention Deficit Hyperactivity Disorder: A Double-blind, Crossover Trial Daryl Efron, FRACP; Frederick Jarman, FRACP; and Melinda Barker, Grad Dip Ed Psych ABSTRACT. Objective. To compare methylphenidate behavioral, academic, and social functioning. Many (MPH) and dexamphetamine (DEX) in a sample of chil- well-designed, placebo-controlled studies have dem- dren with attention deficit hyperactivity disorder onstrated beyond doubt the benefits of stimulants in (ADHD). the vast majority of children with ADHD.2–4 In a Method. A total of 125 children with ADHD received review of 110 studies on the effects of stimulant both MPH (0.3 mg/kg twice daily) and DEX (0.15 mg/kg drugs on more than 4200 children with ADHD, twice daily) for 2 weeks a double-blind, crossover study. 4 ; Outcome measures were Conners’ Parent Rating Scale– Barkley found that 75% of subjects were regarded Revised, Conners’ Teacher Rating Scale–Revised, a Par- as improved on stimulants. The mean placebo re- ent Global Perceptions questionnaire, the Continuous sponse was 39%. Performance Test, and the Barkley Side Effects Rating Methylphenidate (MPH) and dexamphetamine Scale. (DEX) are the two stimulants prescribed most fre- Results. There were significant group mean im- quently and have been shown to have similar types provements from baseline score on all measures for of positive effects in children with ADHD. However, both stimulants. On the Conners’ Teacher Rating Scal- it is not known whether one is more efficacious than e–Revised, response was greater on MPH than DEX on the other in terms of probability of producing a the conduct problems and hyperactivity factors, as well positive response, magnitude of response, quality of as on the hyperactivity index. On the Conners’ Parent Rating Scale–Revised, anxiety was the only factor to improved performance, or side-effect profile. Some differ significantly, in favor of MPH. Parents rated authors suggest that the two stimulants are equally 5 73% of subjects as globally improved on MPH and 69% effective or that “there is little to choose between improved on DEX, compared with baseline. Overall, them,”6 whereas a number of clinicians have the 46% of parents chose MPH as the preferred drug, com- impression that MPH is the more efficacious of the pared with 37% who chose DEX. On the Continuous two and has fewer associated adverse effects. MPH is Performance Test, there was no difference in the num- often designated the drug of first choice in psycho- ber of correct responses or errors between the two pharmacology texts, despite the absence of support- drugs. ing evidence.7 Conclusions. Most children with ADHD improve Clinical experience suggests that although most significantly on both MPH and DEX. There was a slight advantage to MPH on most measures. Pediatrics children respond equally well to either of these 1997;100(6). URL: http://www.pediatrics.org/cgi/ stimulants, a subgroup of children seem to re- content/full/100/6/e6; attention deficit hyperactivity spond better to one than the other. However, there disorder, stimulant medication, methylphenidate, dex- has been surprising little research published exam- amphetamine. ining the question of relative efficacy and toxicity of the two most commonly used drugs in child- ABBREVIATIONS. ADHD, attention deficit hyperactivity disor- hood behavior disturbance. No advantage to either der; MPH, methylphenidate; DEX, dexamphetamine; DSM–IV, drug has been demonstrated to date in the sparse Diagnostic and Statistical Manual of Mental Disorders, 4th ed; literature directly comparing MPH with DEX. CBCL, Child Behavior Checklist; TRF, Teacher Report Form; Hence, the choice of drug is often made on the CPRS–R, Conners’ Parent Rating Scale–Revised; CTRS–R, Con- basis of previous anecdotal experience, trial and ners’ Teacher Rating Scale–Revised; CPT, Continuous Perfor- mance Test; SERS, side effect rating scale. error, and/or cost. In the present study, we set out to compare systematically MPH and DEX in a sam- ple of children with ADHD. timulant medication is the most effective treat- ment for children with attention deficit hyper- METHODS activity disorder (ADHD).1,2 Stimulants have S Subjects been shown to induce short-term enhancement of Subjects were selected from ambulatory patients referred to the Royal Children’s Hospital, Melbourne, Australia, for an assess- From the Centre for Community Child Health and Ambulatory Paediatrics, ment for possible ADHD. Referral sources included pediatricians, Royal Children’s Hospital, Melbourne, Australia. family practitioners, school nurses, and psychologists. In addition, Received for publication May 1, 1997; accepted Jul 14, 1997. many parents self-referred by calling the hospital, usually at the Reprint requests to (D.E.) Centre for Community Child Health and Ambu- suggestion of a relative, friend, the child’s teacher, or the state latory Paediatrics, Royal Children’s Hospital, Flemington Rd, Parkville, ADHD parent support group. Victoria, 3052, Australia. Criteria for enrollment in the trial were 1) age between 5 and 15 PEDIATRICS (ISSN 0031 4005). Copyright © 1997 by the American Acad- years; 2) satisfy Diagnostic and Statistical Manual of Mental Dis- emy of Pediatrics. orders, 4th ed (DSM–IV) criteria for ADHD.8 The DuPaul ADHD http://www.pediatrics.org/cgi/content/full/100/6/Downloaded from www.aappublications.org/newse6PEDIATRICS by guest on October Vol. 2, 1002021 No. 6 December 1997 1of7 rating scale9 a was used in which each DSM–IV ADHD symptom number of commission errors (false alarms) is thought to reflect was marked on a four-point scale: “never or rarely,” “sometimes,” the degree of impulsivity. “often,” or “very often”; only symptoms rated often or very often Subjects attempted the CPT at baseline and again on the final were considered present and counted toward the diagnosis; 3) T day of each medication cycle, 1 to 2 hours after ingestion of their score of at least 1.5 SD units above the mean on the attention morning dose (ie, at a time coinciding with peak behavioral problems scale of the Child Behavior Checklist (CBCL)10 or effects). Teacher Report Form (TRF)11; 4) no history of intellectual disabil- Side effects were evaluated using the Barkley Side Effects Rat- ity, gross neurologic abnormality, or Tourette’s syndrome; and 5) ing Scale (SERS). This questionnaire assesses the frequency and decision made to undertake stimulant medication trial on clinical severity of 17 common side effects of stimulants, rated on a scale grounds. from 0 (absent) to 9 (severe).9 Because children with ADHD often display some of these apparent side effects before receiving med- Procedure ication, this questionnaire was also administered at baseline, so This study used a double-blind, crossover design. Subjects were that true medication effects were able to be measured. IQ was estimated using a short form of the Wechsler Intelli- randomized to receive either DEX or MPH for the first 2 weeks of 14 the study. After a 24-hour washout period, they were crossed over gence Scale for Children, 3rd ed, comprising two verbal (simi- to receive the other stimulant for the 3rd and 4th weeks. Each drug larities and vocabulary) and two performance (block design and was administered twice a day, after breakfast and after lunch, at a object assembly) scale subtests. standardized dose. The dose was 0.15 mg/kg/dose for DEX and 0.3 mg/kg/dose for MPH, rounded off to the nearest capsule size. Data Analysis Both drugs were presented in identical form, as a crushed powder The methods used to analyze the CPRS–R and CTRS–R were in opaque gelatin capsules (2.5 mg for DEX and 5 mg for MPH). identical. Data were analyzed by factor. The principal measure The investigators, families, subjects, and teachers were blind to the used was the difference in T score from baseline to the end of the randomization order throughout the study period. The study treatment period (ie, baseline T score 2 treatment T score). protocol was approved by the Ethics in Human Research Com- Initially within-subject analysis of variance (ANOVA) for re- mittee of the Royal Children’s Hospital, and written informed peated measures was computed for the means of all factors (ie, consent was obtained from parents. comparison of scores at three trial conditions: baseline, DEX, and MPH). Because all F values were significant, planned comparisons Measures were conducted to define differences, using paired-samples t tests The following four principle measures of response to stimulant (Table 1). T medication were used in this study. The change in score for each factor was then submitted to the Hills and Armitage analysis for data from crossover trials.15 Data 12 were first tested for period effects and treatment-by-period inter- Conners’ Parent Rating Scale–Revised (CPRS–R) action effects, and finally the magnitudes of the effects of the two This 48-item questionnaire yields five factors: conduct prob- interventions were compared. lems, learning problems, psychosomatic, impulsive–hyperactive, Subjects were then classified into responders and nonre- and anxiety. In addition, a composite hyperactivity index has been sponders to each stimulant. This enabled a categorical analysis of derived from the 10 items with the highest loading from the factor response, using the same three measures. On the CPRS–R and scales. Each item is rated on a four-point scale (not at all 5 0, just CTRS–R, subjects were grouped according to the change in hyper- a little 5 1, pretty much 5 2, and very much 5 3). Raw scores for activity index T score from the baseline score. The cutoff point was each factor are transformed by age and sex into T scores, with a chosen so as to classify response in a meaningful way, both mean of 50 and an SD of 10.