Unexpected Effects of Methylphenidate

JOURNAL OF CHILD AND ADOLESCENT PSYCHOPHARMACOLOGY Volume 18, Number 3, 2008 © Mary Ann Liebert, Inc. Pp. 237–247 DOI: 10.1089/cap.2007.0140

Unexpected Effects of Methylphenidate in Attention- Deficit/Hyperactivity Disorder Reflect Decreases in Core/Secondary Symptoms and Physical Complaints Common to All Children

Mark D. Rapport, Ph.D., Michael J. Kofler, M.S., Maria M. Coiro, Ph.D., Joseph S. Raiker, B.S., Dustin E. Sarver, B.A., and R. Matt Alderson, M.S.

Abstract

Hypotheses concerning unexpected, psychostimulant-related effects reported in previous studies were examined by separating behavioral/physical complaints highly specific to methylphenidate (MPH) from those that (a) may mimic core/secondary symptoms of the disorder, or (b) are commonly reported by unmedicated children in the general population. Sixty-five children with attention-deficit/hyperactivity disorder (ADHD) participated in a double-blind, placebo-controlled, within-subject (crossover) experimental design and received a placebo and four MPH doses in counterbalanced order following baseline assessment. Behavioral and physical complaints were significantly higher under baseline relative to placebo and the four immediate-release MPH conditions (5 mg, 10 mg, 15 mg, and 20 mg) across three symptom categories: ADHD core/secondary symptoms; symptoms commonly reported in the general population, including unmedicated children with ADHD; and symptoms highly specific to MPH. No significant differences were found among active drug conditions. Past unexpected findings of psychostimulant effects in ADHD may be due to the inclusion of scale items that reflect core/secondary features of ADHD and normally occurring behavioral/physical complaints in children.

TTENTION-DEFICIT/HYPERACTIVITY DISORDER (ADHD) is a logical treatment for ADHD (Faraone et al. 2002; Grcevich et Acomplex and chronic disorder of brain, behavior, and al. 2001; Jensen 1999; Swanson and Volkow 2002), and its development whose behavioral and cognitive consequences reputation is well deserved based on traditional benchmarks, pervade multiple areas of functioning. Core features of the including breadth of effectiveness and overall response rate disorder involve difficulties with attention, impulsiveness, among affected individuals (Barkley, 2006; Denney and Rap- and hyperactivity (APA 2000) and are hypothesized to affect port 2001; Rapport et al. 1994). As with all therapies, how- behavioral and cognitive functioning to the extent that the ever, treatment emergent symptoms can and do occur. latter are dependent upon the former for successful execu- Side effects associated with psychostimulant treatment are tion (Rapport et al. 2001). Treatment of ADHD traditionally well documented in the literature and fall primarily into one involves using behavior or pharmacological therapy alone of three categories: cardiovascular effects (i.e., heart rate, or in combination (for reviews, see Chronis et al. 2004; Con- blood pressure), physical effects (i.e., weight and growth), ners et al. 2001; Jensen 1999). and physical and behavioral complaints. Recent reviews in- Pharmacological interventions (and particularly the psy- dicate that cardiovascular and physical effects associated chostimulants), however, are generally considered more cost with psychostimulant therapy are usually transient, dose de- effective and have the added benefit of affecting both be- pendent, readily resolved by discontinuing therapy, and fail havioral and cognitive domains throughout the day without to remain significant in long-term follow-up studies (Rap- the specific programming and oversight required by behav- port and Moffitt 2002). Psychostimulant-related physical and ior therapy (DuPaul and Eckert 1997; Gittelman-Klein and behavioral complaints, on the other hand, have received re- Klein 1976; MTA Cooperative Group 1999). Methylphenidate newed interest secondary to the unexpected effects reported (MPH) is by far the most commonly prescribed pharmaco- in recent, well-controlled studies. For example, parents

Department of Psychology, University of Central Florida, Orlando, Florida

237 238 RAPPORT ET AL. report that their children experience fewer behavioral (Gor- havioral complaints commonly reported by children in- man et al. 2006) and physical complaints such as day- clude musculoskeletal pain, back pain, and dizziness dreaming, irritability, anxiety, staring, and nail biting under (Campo and Fritsch 1994). MPH relative to placebo or baseline (Ahman et al. 1993; Fin- Physical and behavioral complaints in children are not dling et al. 2001; Greenhill et al. 2001b; Short et al. 2004). limited to Western culture (cf. Belmaker et al. 1985), and Teachers also report more severe physical and behavioral extant evidence suggests that the frequency and type of complaints for children under placebo relative to low MPH physical and behavioral complaints vary with age and gen- doses (Barkley et al. 1990; DuPaul et al. 1996; Fischer and der (Last 1991). For example, recurrent abdominal pain ap- Newby 1991). pears to be more common in early childhood, whereas Two hypotheses have been proposed to account for the headaches, limb pain, and polysymptomatic complaints in unexpected findings: (1) Adult raters (parents, teachers) con- general increase with age (Achenbach 1989). Psychiatric fuse some symptoms of ADHD (e.g., daydreaming and star- disability confers additional risk for physical and behav- ing) with drug-related side effects (Fine and Johnson 1993; ioral complaints. Twenty percent to 69% of children and Firestone et al., 1998); and (2) ratings under no medication adolescents with at least one psychiatric diagnosis experi- conditions (i.e., baseline, placebo) reflect normal levels of ence physical and behavioral complaints (Egger et al. 1999; physical and behavioral complaints for boys in general, in- Masi et al. 2000; Taylor et al. 1996), particularly children cluding unmedicated children with ADHD, that may occur with anxiety and mood disorders (Carlson and Kashani less frequently as a function of treatment (Rapport et al. 1998; Pine 2002). Children with externalizing disorders are 2002). Both hypotheses are possible because of the nonspe- also at greater risk than non-psychiatric controls for phys- cific nature of side effect rating scales. Most scales are de- ical and behavioral complaints. Boys with conduct disor- signed to assess a wide range of possible symptoms rather der report twice as many headaches as boys without the than emergent symptoms associated with a specific phar- disorder, and unmedicated children with ADHD complain macological regimen. more frequently of stomachaches (Egger et al. 1999), poly- The rater confusion hypothesis can be addressed by ex- dipsia, and polyuria (Mitchell et al. 1987) relative to non- amining the type of physical and behavioral complaints re- psychiatrically-disabled children. ported under no-drug conditions, and changes from this Collectively, the physical and behavioral complaint liter- state that occur under placebo and active drug conditions ature indicates that moderate to elevated complaints of head- (Barkley et al. 1990; Greenhill et al. 2001). Results indicating aches, low energy, sore muscles, abdominal discomfort, daily treatment-related decreases in only complaints that mimic fatigue, musculoskeletal pain, back pain, and dizziness are core/secondary features of ADHD (e.g., inattentiveness, dif- common among children, and that a diagnosis of ADHD con- ficulty concentrating, staring, daydreaming), coupled with fers additive risk for stomachaches, polydipsia, and polyuria. no change or a worsening of complaints not typically asso- These findings suggest that a substantial number of physi- ciated with the disorder (e.g., cardiovascular, gastrointesti- cal and behavioral complaints attributed to medication in nal, and pseudoneurological symptoms), would lend em- past studies may have been confounded with high base rate pirical support to the rater confusion hypothesis. The occurrences of these behaviors in children with ADHD. This pre-existing/normal level complaints hypothesis, on the hypothesis can be examined by comparing base rate occur- other hand, can be examined by juxtaposing the psycho- rences of the above physical and behavioral complaints re- stimulant emergent symptom literature with extant litera- ported in the literature with those reported under placebo ture concerning the prevalence of physical and behavioral and active medication conditions, as previously suggested complaints in samples of non-referred and clinically diag- by Barkley and colleagues (1990) and more recently by nosed children. Greenhill and colleagues (2001a). A stable or decreasing fre- The occurrence of physical and behavioral complaints in quency of complaints under placebo and active medication community samples of non-referred children and adoles- conditions relative to baseline levels would support the pre- cents is well documented. For example, nearly half of the existing/normal level complaints hypothesis. 540 children and adolescents in a community sample at- Other types of emergent symptoms specific to MPH ther- tending third through twelfth grade reported at least one apy (e.g., cardiovascular symptoms, upset stomach, reduced physical symptom during the preceding two weeks (Gar- appetite)—other than those commonly reported by children ber et al. 1991), with the highest frequency of complaints or potentially confused with core and secondary symptoms involving headaches (25%), low energy (23%), sore muscles of ADHD—may worsen as a function of increasing MPH (21%), and abdominal discomfort (17%). Literature reviews dose. These can be extracted from side effect rating scales confirm that headaches represent the most commonly re- and examined separately to determine whether changes in ported painful physical or behavioral symptom, and occur frequency occur as a function of expectancy (placebo) or ac- at least weekly in 10% to 30% of children in community tive drug over and above base rate levels (i.e., true emergent samples (Campo and Fritsch 1994). Epidemiological stud- symptoms). ies corroborate these findings, wherein 50% to 80% of chil- The present study examined rival hypotheses that may ac- dren report some type of headache within the past year count for past reports of unexpected reductions in physical (Barea et al. 1996), and upwards to 15% experience head- and behavioral complaints in children with ADHD receiv- aches of a migrainous nature (Abu-Arefeh and Russell ing psychostimulant therapy. Adverse events endorsed by 1996). Recurrent abdominal pain is also common (10% to parents and children under baseline, placebo, and four MPH 25% of school-age children and adolescents), and nearly conditions (5 mg, 10 mg, 15 mg, 20 mg) were allocated to 15% of children complain of daily fatigue (Apley 1975; Gar- one of three categories using the empirically-driven rational ber et al. 1991; Linna et al. 1991). Other physical and be- approach described below—behavioral/physical complaints ADHD PHYSICAL/BEHAVIORAL COMPLAINTS 239 that (1) may mimic core/secondary features of ADHD; (2) were packaged in colored gelatin capsules by the clinic’s are commonly reported by children in the general popula- pharmacist to avoid detection of dose and taste. Capsules tion, including unmedicated children with ADHD; or (3) are were sealed in individual, daily envelopes to help control for commonly associated with psychostimulant treatment (i.e., accurate administration. true emergent symptoms)—then subjected to analysis. After baseline data collection (first week), parents were given one week of medication in predated envelopes at a single dose level (i.e., placebo, 5 mg, 10 mg, 15 mg, or 20 mg). Method Single morning doses (as opposed to twice per day) were ad- Participants ministered to maintain experimental control, as it was not possible to ensure that medication would be appropriately Sixty-five children (58 boys, 7 girls) referred by commu- administered at school at an established time during the day. nity psychiatrists, pediatricians, and school personnel met This procedure continued until each child received every the following inclusion criteria and participated in the study: dose for 6 consecutive days. All weekly dose changes oc- (a) diagnosis of ADHD based on parent clinical interview us- curred on Sundays (i.e., no capsules were administered on ing the Schedule for Affective Disorders and Schizophrenia Saturdays) to allow for “washout” and to control for possi- for School-Age Children (K-SADS; Kaufman et al. 1997); (b) ble rebound/carryover effects. Parents were instructed to problems in at least 50% of the situations on Barkley’s (2006) give their child a capsule each morning, 30 minutes before Home Situations Questionnaire; (c) a maternal rating at least breakfast. Both used and unused envelopes were returned two standard deviations above the mean on the Werry- on a weekly basis to assess medication compliance. Medica- Weiss-Peters Activity Scale (Routh et al. 1974); (d) a teacher tion was properly administered nearly 100% of the time. rating of at least two standard deviations above the mean on Behavioral and physical complaint items were adapted the Abbreviated Conners’ Teacher Rating Scale (ACTRS; from the National Institute of Mental Health Subject’s Treat- Conners et al. 1998); (e) absence of conduct disorder; and (f) ment Emergent Symptoms Scale (STESS; Guy 1976). The absence of gross neurological, sensory, or motor impairment STESS was initially devised for adult raters and consists of as determined by pediatric examination. The University In- questions concerning the occurrence and severity of a broad ternal Review Board approved the study, parents provided range of possible physical and behavioral complaints and informed consent, and children provided assent to partici- emergent symptoms associated with pharmacotherapy. A pate in the study. Children were from 6 to 11 years of age child version was created using the same items, but re- (M 8.56, SD 1.25) and fell within the average range of worded to ask children (1) whether they had experienced a intelligence (M 102.8, SD 10.0) based on the Peabody particular problem within the past three days, and (2) if so, Picture Vocabulary Test (Dunn and Dunn 1997). They were the degree of severity using the original version’s 4-point re- all Caucasian and from families of low to middle socioeco- sponse format (“not at all,” “just a little,” “pretty much,” nomic status (Hollingshead 1975). Eight had experienced “very much”). Additional details of the development of the brief trials of stimulant therapy within the previous 4 years. child version are reported in Rapport et al. (2002). The re- None were prescribed psychostimulants immediately prior vised child version has been shown to be sensitive to med- to beginning the current study. ication status and dosage changes (Rapport et al. 2002). Several children showed symptoms of—but did not meet Parents were asked to respond to each item during each formal criteria for—mood, anxiety, oppositional defiant dis- weekly visit throughout the study. Children were adminis- order, and conduct disorders. Increasing the heterogeneity tered the items during each weekly visit by trained gradu- of ADHD samples by including children comorbid for other ate research assistants who read each item aloud to the child disorders would impair interpretation of research findings and recorded their verbal response. Parents and children as argued elsewhere (e.g., Vaessen and Van der Meere 1990). were asked about the presence and severity of each symp- All selected children met criteria for ADHD-Combined Type tom during the preceding three days. Graduate assistants and were attending general education elementary school were blind to children’s medication status. Completed ques- classrooms, although several received concurrent special ed- tionnaires reflected physical and behavioral complaints ex- ucation services. perienced by children within the past three days and associated with a single experimental condition (baseline, Experimental design and procedures placebo, or one of the four MPH conditions). A double-blind, placebo-controlled, within-subject (cross- Items were judged to fall into one of three categories us- over) experimental design was used. Children with ADHD ing an empirically-driven rational approach (Clark and Wat- received a placebo and each of four active immediate-release son 1995). Four independent judges reviewed the empirical MPH doses following baseline assessment. Order of dose ad- literature described above and determined whether each ministration was counterbalanced and determined by ran- item reflected behavioral/physical complaints that (1) may dom assignment such that an equal number of children re- mimic core and secondary features of ADHD (Ahman et al. ceived each dose during a given week of the study. MPH 1993; Barkley et al. 1990; Beidel et al. 1991; APA 2000); (2) was prescribed by each child’s physician in the following are common to pediatric age children (Belmaker et al. 1985; doses: placebo, 5 mg (range .10 to 0.26 mg/kg), 10 mg Campo and Fritsch 1994; Garber et al. 1991), including un- (range 0.20 to 0.52 mg/kg), 15 mg (range 0.29 to 0.79 medicated children with ADHD (Egger et al. 1999; Mitchell mg/kg) and 20 mg (range 0.39 to 1.1 mg/kg). Fixed doses et al. 1987); or (3) do not fit into the preceding categories and were used to reflect typical pediatric practice because re- are associated with psychostimulant treatment (e.g., nau- sponse to MPH is independent of body mass (Rapport and sea/upset stomach, dry mouth, decreased appetite). Intra- Denney 1997; Swanson et al. 1991). MPH and placebo doses class correlation and Fleiss’ kappa were computed to test the 240 RAPPORT ET AL. reliability of the revised scale categories (Nunnally and Bern- more than one component reached statistical significance stein 1994). Intraclass correlation was computed using the (Keppel 1991). Shrout and Fleiss (1979; Case 2) method to determine the Trend analysis of parent endorsement of physical and be- generalizability of the four judges’ item categorization to the havioral complaints excluding the placebo condition (i.e., larger population of potential judges (i.e., others that have baseline, 5 mg, 10 mg, 15 mg, 20 mg) revealed significant lin- reviewed the treatment emergent and physical complaints ear F (4, 256) 37.83, p .001, quadratic F (4, 256) 20.77, literatures), ICC(2, 4) .95, p .0005. Fleiss’ kappa for more p .001, and cubic F (4, 256) 9.31, p .001 trends. The lin- than two raters (Siegel and Castellan 1988) was computed ear trend accounted for the greatest proportion of variance using SPSS syntax (King 2004), and provided further evi- (15%), whereas 6% and 3% of unique variance was accounted dence of reliability, .86. Individual disagreements were for by the two higher-order trends, respectively. This find- discussed and resolved based on empirical literature. Items ing indicates that parental report of physical and behavioral judged to belong to each category are shown in Tables 1 complaints that mimic core/secondary features of the dis- and 2. order evince a moderately abrupt decrease between baseline and MPH 5-mg, with minimal variation in frequency under Results higher dose conditions (see Fig. 1). Only the linear trend was significant for the placebo-active drug trend analysis F (4, All tests were two-tailed with alpha level set at .05. A 256) 12.35, p .001, and the small proportion of variance three-tier data analytic strategy was used to examine the accounted for by the model (5%) indicates that the primary study’s two primary hypotheses: (Tier I) that adult raters change in complaint frequency results from changes between confuse some features of ADHD with drug-related side ef- baseline and active drug conditions. fects; and (Tier II) that baseline/placebo measures reflect Collectively, the preceding analyses indicate a significant normal levels of physical and behavioral complaints for boys reduction in parental report of physical and behavioral com- in general including unmedicated children with ADHD. Tier plaints that mimic core/secondary ADHD features under III analyzed endorsement of remaining emergent symptoms placebo relative to baseline, and under active MPH condi- associated with psychostimulants to determine whether pre- tions relative to both baseline and placebo, with no signifi- viously reported unexpected effects remain evident after re- cant variation in complaint frequency among active drug moving ADHD-like and common behavior and physical conditions. These findings are consistent with the hypothe- complaints. Mean severity and percentage of parents and sis that adult raters confuse some features of ADHD with children endorsing each symptom are shown in Tables 1 drug-related side effects. and 2. Tier II analysis: Preexisting/normal level hypothesis Tier I analysis: Rater confusion hypothesis Parent and child reports of physical and behavioral com- Parental report of physical and behavioral complaints that plaints common to pediatric age children were analyzed us- may mimic core/secondary features of ADHD were ana- ing a 2 (child and parent raters) 6 (conditions: baseline, lyzed using a repeated measures ANOVA (conditions: base- placebo, 5 mg, 10 mg, 15 mg, 20 mg) mixed-model ANOVA line, placebo, 5 mg, 10 mg, 15 mg, 20 mg). The analysis ex- to examine the preexisting/normal level hypothesis. This hy- amines the hypothesis that adult raters confuse ADHD pothesis posits that the base rate of physical and behavioral core/secondary features as MPH side effects. The main ef- complaints commonly reported by children (including un- fect for drug condition was significant, F (5, 640) 16.34, p medicated children with ADHD) must be accounted for by .001. Tukey post hoc analysis revealed that parent endorse- examining baseline frequencies to separate them from actual ment of physical and behavioral complaints that mimic treatment emergent symptoms associated with MPH. core/secondary ADHD features under baseline were signif- The main effect for drug condition was significant, F (5, icantly higher than complaints endorsed under placebo and 640) 4.42, p .001. Examination of Fig. 2 suggests that the all active medication conditions (all p .05). Additionally, significant decrease in endorsements from baseline to behavioral complaints were significantly higher in the placebo (p .05) and all active MPH conditions (all p-val- placebo condition relative to the four MPH conditions (all ues .05) is related primarily to decreased child endorse- p .05). No significant differences emerged among the four ments. However, neither the main effect for rater, F (1, 128) MPH doses (see Fig. 1). 2.84, ns, nor the rater by drug condition interaction, F (5, Analysis of trend was performed to examine the shape of 640) 2.22, ns, was significant. the relationship between parent ratings of physical/behavioral complaints and MPH dose. Parent endorsements under Tier III analysis: Side effects commonly attributed to baseline and active medication conditions were examined in psychostimulant treatment the initial analysis (i.e., the placebo condition was excluded). The ensuing analysis was conducted using placebo and ac- Parent and child reports of physical and behavioral com- tive drug conditions with baseline excluded.1 The propor- plaints commonly attributed to psychostimulant therapy 2 tion of treatment variance (R trend) was computed for each (true emergent symptoms) were analyzed using a 2 (child significant trend to determine the relative contribution of and parents raters) 6 (conditions: baseline, placebo, 5 mg, each trend component (e.g., linear, cubic, quadratic) when 10 mg, 15 mg, 20 mg) Mixed-model ANOVA. The rater by drug interaction, F (5, 640) 2.94, p .05, and main effect 1 for drug were significant, F (5, 640) 10.66, p .001. There A basic assumption of trend analysis is that “levels” of experimental conditions be of similar magnitude, which precludes simul- was no significant main effect for rater, F (1, 128) .099, ns. taneous use of baseline and placebo in the same analysis. Tukey post hoc analysis indicated that children endorsed TABLE 1. MEAN SEVERITY AND PERCENTAGE OF PARTICIPANTS ENDORSING EACH BEHAVIORAL AND PHYSICAL SYMPTOM UNDER BASELINE , PLACEBO , AND ACTIVE METHYLPHENIDATE CONDITIONS —CHILD RATINGS

Behavioral/Physical Complaints1 Baseline Placebo 5 mg 10 mg 15 mg 20 mg

Core/secondary symptoms of ADHD Difficulty sitting still .47(25%) .16(6.6%) .15(8.2%) .21(9.5%) .20(7.8%) .06(3.2%) Difficulty sleeping .86(39.1%) .24(12.9%) .25(14.1%) .27(14.1%) .33(14.1%) .24(9.7%) Decreased sleep .26(18.5%) .21(9.7%) .34(12.5%) .22(10.9%) .11(4.7%) .13(6.7%) Crying .17(7.7%) .16(6.5%) .05(1.5%) .08(3.2%) .08(3.1%) .00(0.0%) Difficulty with attention .41(15.9%) .25(11.5%) .10(4.8%) .16(6.3%) .09(6.3%) .02(1.7%) More talkative .98(44.4%) .37(15%) .44(18%) .41(17.5%) .27(12.7%) .36(13.1%) Difficulty with sports .25(10.8%) .08(3.3%) .05(1.6%) .03(1.6%) .05(1.6%) .03(1.6%) Difficulty with parent relationships .13(9.4%) .21(11.1%) .09(6.2%) .13(6.3%) .03(3.1%) .03(3.3%) Difficulty with peer relationships .53(25%) .24(12.7%) .18(8.1%) .16(6.3%) .21(8.2%) .13(6.5%) Anger .54(24.6%) .13(6.5%) .08(3.1%) .17(6.3%) .20(11.1%) .10(4.8%)

Complaints common to all children Stomach aches .46(33.8%) .21(11.3%) .23(13.8%) .25(14.1%) .25(15.9%) .25(12.7%) Cramps .31(20%) .18(9.8%) .18(6.5%) .13(6.3%) .16(9.7%) .23(9.7%) Headaches .50(23.4%) .18(9.7%) .18(9.2%) .38(18.8%) .25(14.1%) .21(11.1%) Dizziness .23(13.8%) .11(6.6%) .02(1.6%) .08(4.7%) .02(1.6%) .16(9.7%) Tiredness/fatigue .60(32.3%) .37(16.1%) .33(17.2%) .38(15.6%) .36(18.8%) .18(11.5%) Muscle aches .34(25%) .09(6.3%) .06(4.6%) .06(3.1%) .09(6.2%) .12(6.2%)

Emergent symptoms Less eating .23(20%) .20(10%) .11(4.8%) .30(10.9%) .16(6.3%) .14(7.9%) More drinking 1.22(53.1%) .32(11.7%) .21(9.7%) .27(9.4%) .32(12.7%) .26(11.3%) Dry mouth .59(35.9%) .37(14.5%) .31(13.8%) .20(10.9%) .27(14.1%) .21(11.5%) More bowel movements .09(6.3%) .12(6.7%) .11(4.7%) .17(6.3%) .05(1.6%) .03(3.3%) Fewer bowel movements .36(21.9%) .31(13.1%) .20(7.8%) .17(7.9%) .19(7.8%) .12(5.0%) Harder bowel movements .05(4.9%) .11(4.8%) .11(4.8%) .05(1.6%) .00(0.0%) .02(1.7%) Softer bowel movements .40(21.7%) .24(12.9%) .11(4.8%) .36(12.5%) .19(6.3%) .15(6.8%) Sick to stomach .36(25%) .21(11.3%) .18(9.2%) .27(14.1%) .25(12.7%) .23(9.7%) Throw up .19(14.3%) .10(4.8%) .02(1.6%) .08(3.1%) .13(6.3%) .03(3.2%) More bedwetting .11(4.7%) .11(4.8%) .08(3.1%) .11(4.7%) .11(4.7%) .06(3.2%) Polyurea .25(12.3%) .11(6.6%) .09(4.7%) .11(4.7%) .00(0.0%) .07(3.3%) Decreased urine .31(23.1%) .23(8.2%) .15(7.7%) .19(6.3%) .16(7.9%) .13(6.6%) Painful urination .08(4.6%) .10(6.5%) .00(0.0%) .06(3.2%) .06(3.1%) .00(0.0%) Skin itching .41(28.1%) .25(9.8%) .20(7.8%) .28(12.5%) .19(12.5%) .18(6.5%) Rash .12(9.2%) .11(5.0%) .11(6.3%) .11(6.3%) .10(6.5%) .06(3.2%) Difficulty with harder words .22(11.1%) .13(8.2%) .00(0.0%) .00(0.0%) .05(1.6%) .00(0.0%) More shaky .17(9.4%) .11(4.9%) .06(3.1%) .16(6.3%) .02(1.6%) .13(8.3%) Harder to do things with hands .08(7.7%) .16(8.1%) .11(4.9%) .09(4.7%) .05(1.6%) .07(6.6%) More bad dreams .22(10.8%) .10(4.8%) .15(6.2%) .06(3.1%) .05(1.6%) .11(4.8%) More energy 1.24(55.6%) 1.18(48.4%) .82(36.1%) .76(27.4%) .78(32.8%) .89(32.8%) Unhappy .20(10.8%) .06(3.2%) .16(6.3%) .10(4.8%) .05(1.6%) .00(0.0%) Feeling “worse” .04(3.8%) .06(3.2%) .14(7.9%) .08(4.8%) .02(1.6%) .03(3.2%)

Note: Values in each column reflect symptom mean severity score. Values in parentheses reflect percentage of participants endorsing symptom 1Moderate increases in parent or child ratings of select emergent symptoms were reported for some children under the two highest doses (15 mg and 20 mg): feeling sick to stomach, n 11; itchy skin, n 10; dry mouth, n 9; and reduced appetite, n 6). ADHD, attention-deficit/hyperactivity disorder. TABLE 2. MEAN SEVERITY AND PERCENTAGE OF PARTICIPANTS ENDORSING EACH BEHAVIORAL AND PHYSICAL SYMPTOM UNDER BASELINE , PLACEBO , AND ACTIVE METHYLPHENIDATE CONDITIONS —PARENT RATINGS

Behavioral/Physical Complaints1 Baseline Placebo 5 mg 10 mg 15 mg 20 mg

Core/secondary symptoms of ADHD Clumsiness .21(15%) .15(11.3%) .00(0.0%) .03(3.3%) .13(7.9%) .00(0.0%) Difficulty sleeping .08(4.7%) .13(11.3%) .06(6.2%) .27(20.6%) .10(10.2%) .13(8.3%) Decreased sleep .17(14.1%) .14(11.1%) .08(7.7%) .16(14.8%) .13(11.5%) .15(15.3%) Crying .46(34.9%) .26(19.4%) .28(23.1%) .32(21.7%) .35(23.8%) .28(21.7%) Daydreaming .23(19.4%) .11(9.8%) .09(9.4%) .02(1.7%) .07(5.0%) .11(6.6%) Difficulty with attention .87(46.8%) .83(45.0%) .42(29.2%) .31(21.3%) .23(14.8%) .24(13.6%) More talkative .63(36.5%) .60(38.7%) .43(29.2%) .58(33.3%) .48(30.6%) .37(27.1%) Difficulty with sports .25(12.3%) .17(7.8%) .05(4.9%) .00(0.0%) .08(4.8%) .05(3.2%) Difficulty with parent relationships .91(56.3%) .67(43.8%) .53(40%) .45(32.3%) .43(25.4%) .48(33.9%) Difficulty with peer relationships 1.12(61.5%) .70(42.2%) .49(34.4%) .48(33.9%) .44(25.4%) .40(29%) Anger 1.29(69.2%) .69(48.4%) .60(41.9%) .47(38.7%) .54(33.3%) .47(29%) Complaints common to all children Stomach aches .33(26.6%) .27(18.8%) .20(18.8%) .30(25.4%) .22(17.5%) .32(27.4%) Cramps .05(4.8%) .08(4.8%) .12(9.2%) .05(4.8%) .02(1.6%) .08(4.8%) Headaches .39(37.5%) .19(18.8%) .25(20%) .24(17.5%) .17(14.1%) .19(14.5%) Dizziness .00(0.0%) .05(3.2%) .00(0.0%) .03(3.2%) .03(1.6%) .11(8.1%) Tiredness/fatigue .15(10.8%) .05(4.8%) .20(18.8%) .16(16.1%) .20(15%) .21(17.7%) Muscle aches .09(7.8%) .06(3.1%) .05(4.8%) .03(3.2%) .00(0.0%) .15(8.1%) Emergent symptoms Less eating .12(10.8%) .17(15.9%) .25(20%) .30(22.2%) .37(23.8%) .38(27%) More drinking .14(10.8%) .13(9.5%) .22(15.6%) .15(11.5%) .27(22.2%) .16(11.5%) Dry mouth .05(4.9%) .10(9.8%) .15(11.7%) .08(6.5%) .10(6.8%) .13(10%) Constipation .05(4.7%) .02(1.6%) .00(0.0%) .03(3.2%) .03(3.2%) .00(0.0%) Diarrhea .09(9.4%) .11(7.9%) .11(7.9%) .16(11.3%) .10(4.8%) .08(6.3%) Nausea .03(1.6%) .08(6.3%) .05(4.6%) .06(4.8%) .08(6.3%) .16(11.3%) Bedwetting .17(10.8%) .11(7.9%) .00(0.0%) .10(4.8%) .11(7.8%) .05(3.2%) Polyurea .08(6.3%) .00(0.0%) .10(7.9%) .07(5%) .08(4.9%) .08(6.6%) Decreased urine .02(1.6%) .00(0.0%) .00(0.0%) .05(3.3%) .00(0.0%) .00(0.0%) Painful urination .00(0.0%) .02(1.6%) .00(0.0%) .00(0.0%) .02(1.6%) .02(1.6%) Skin itching .09(6.3%) .19(14.3%) .09(7.7%) .10(9.7%) .16(14.1%) .15(14.5%) Rash .03(3.1%) .05(3.1%) .06(6.2%) .03(3.2%) .13(9.4%) .10(8.1%) Light sensitivity .05(4.8%) .06(6.5%) .05(4.9%) .07(4.9%) .05(4.9%) .05(3.5%) Difficulty with balance .18(11.5%) .08(6.5%) .03(3.2%) .03(3.3%) .06(3.2%) .02(1.7%) Difficulty with speech .09(6.3%) .08(6.3%) .05(4.7%) .02(1.6%) .06(3.2%) .03(3.3%) Shaky hands .03(3.1%) .03(3.3%) .00(0.0%) .02(1.7%) .05(3.2%) .05(3.3%) More bad dreams .02(1.6%) .05(4.9%) .05(3.2%) .05(4.9%) .03(3.2%) .07(4.9%) Difficult to please .58(35.4%) .48(38.1%) .34(29.2%) .37(24.2%) .27(21%) .18(16.4%) More serious .26(25.8%) .30(25%) .39(37.5%) .38(28.6%) .48(34.4%) .48(33.9%) Silliness .73(50.8%) .48(32.8%) .31(24.6%) .49(33.3%) .39(29.7%) .30(27.9%) More energy .69(36.9%) .60(39.7%) .35(24.6%) .50(30.6%) .29(20.6%) .26(16.4%) Less Energy .02(1.6%) .06(4.8%) .22(15.4%) .18(13.1%) .22(14.3%) .26(21%) More withdrawn .06(6.2%) .05(4.8%) .06(6.2%) .03(3.3%) .15(9.7%) .16(11.3%) Unhappy .41(32.8%) .24(17.5%) .16(15.6%) .18(12.9%) .21(14.3%) .16(16.1%) Happiness .52(31.3%) .44(26.2%) .37(26.2%) .47(32.2%) .44(31.7%) .33(23.3%) More repetitive behaviors .44(25.8%) .23(17.7%) .09(6.3%) .18(16.4%) .08(8.1%) .17(8.3%) More self-harm behaviors .03(1.5%) .00(0.0%) .00(0.0%) .00(0.0%) .03(1.6%) .00(0.0%) Feeling “worse” .44(20.6%) .58(33.9%) .25(20.6%) .23(16.4%) .22(13.3%) .16(10.5%)

STESS side effect rating scale (Rapport et al. 2002) were sep- arated into specific categories using an empirically-driven rational approach to investigate predictions that some physical and behavioral complaints represent ADHD primary/ secondary features, some reflect non-medication-related complaints reported by children in general, and others are true emergent symptoms specific to psychostimulant treatment. Child endorsements of both common physical complaints and true emergent symptoms showed significant reductions from baseline to active immediate-release MPH conditions, and were of a sufficient magnitude to be considered clinically meaningful (range 27% to 62% reduction). Significant reductions in symptom endorsements were also observed between placebo and active MPH conditions for parent ratings that reflect ADHD core/secondary features (e.g., sitting still, difficult peer relationships), and for child symptom ratings specific to MPH treatment (e.g., stomachaches, reduced appetite). These changes were smaller in magnitude than those observed for the baseline-MPH contrasts, and suggest that part of the observed effect is associated with expectancy, FIG. 1. Mean number of physical and behavioral symp- “cry for help,” regression to the mean, or other factors re- toms endorsed by parents that may reflect core or secondary lated to placebo phenomena. No significant differences features of attention-deficit/hyperactivity disorder under emerged among the four active MPH doses for child and baseline, placebo, and active methylphenidate (MPH) con- parent endorsements across the three categories. ditions. Our results are consistent with those reported in past studies that included baseline measures of psychostimulant effects. For example, Ahman and colleagues (1993) reported reduced frequencies in four behavioral symptoms assessed significantly more physical and behavioral complaints com- by the Stimulant Drug Side Effects Rating Scale (Barkley monly attributed to psychostimulants under baseline rela- 2006) under their high dose MPH condition (dose range tive to placebo and all four active MPH conditions (all p .3 mg/kg to .6 mg/kg). Efron and colleagues (1997) reported .05), and under placebo relative to 5 mg, 15 mg, and 20 mg conditions (all p .05). No significant differences emerged among the active MPH conditions (see Fig. 3). Parent endorsements showed a similar pattern of results—significantly more physical and behavioral complaints were reported under baseline relative to 5 mg and 20 mg MPH (both p .05). No other contrasts were significant. Collectively, these results indicate that the significant interaction effect was due to initially higher child relative to adult complaint endorsements under baseline, coupled with lower complaint frequency under the MPH conditions. Trend analysis reveals that child ratings are best characterized by a linear decrease in physical and behavioral complaints, with the largest decrease occurring between baseline and active medication conditions. Parent ratings are best characterized by a cubic decrease in complaints that accounts for less than 2% of change—indicating minimal change across the experimental conditions. Group-level analysis was supplemented by examination of individual child rating scale results, and revealed moderate increases in select emergent symptoms for some children (e.g., feeling sick to stomach, n 11; itchy skin, n 10; dry mouth, n 9; and reduced appetite, n 6) under the two highest doses (15 mg and 20 mg).

Discussion FIG. 2. Mean severity ratings of parent and child endorse- The present study examined competing hypotheses re- ments of physical and behavioral complaints commonly re- garding past reports of unexpected reductions in physical ported by children (including those with attention- and behavioral complaints associated with MPH therapy for deficit/hyperactivity disorder) in the general population. children with ADHD. Symptoms on the child version of the MPH methylphenidate; BL baseline; PL placebo. 244 RAPPORT ET AL.

tive to baseline by child endorsements. Mean child complaint severity decreased between 44% and 56% from baseline to active MPH conditions. A proportion of this change appears to be related to expectancy effects associated with placebo phenomena in child ratings (i.e., changes from baseline to placebo for child ratings was 52%). These effects are traditionally attributed to a person’s belief in a treatment’s potential efficacy, but may also be mediated by changes in emotional state, perception, and behavioral improvement (Stewart-Williams 2004). Placebo effects, however, only partially account for the significant reductions in complaints from baseline to MPH. A halo effect may explain the similar magnitude reductions from placebo to active medication conditions across all three symptom categories. In other words, active medication re- duces impairing symptoms (reflected in the Tier I analyses), which may result in parents viewing their children as health- ier in general (as reflected in the Tiers II and III analyses). The reduction in children’s endorsements may also reflect improvement in associated areas of daily functioning, such as improved attention, academic performance, and class- FIG. 3. Mean number of physical and behavioral symp- room conduct associated with both placebo and active med- toms endorsed by parents and children that are commonly ication conditions (e.g., DuPaul and Rapport 1993; MTA Co- attributed to methylphenidate (MPH) therapy (true side ef- operative Group 1999; Pelham et al. 1990). This potential halo fects). effect merits consideration and requires a carefully designed protocol to determine whether improved behavioral and/or academic functioning serve as significant mediators for chil- similar results—a decrease in frequency and severity of side dren’s emergent symptoms (Greenhill et al. 2001b). effects relative to baseline when comparing MPH to dexam- Similar results were found when examining child en- phetamine at .3 mg/kg and .15 mg/kg, respectively. Placebo- dorsement of symptoms highly specific to psychostimulant controlled dose-response investigations without baseline medication. Children endorsed a significantly higher sever- measures report similar findings (e.g., Firestone et al. 1998; ity of complaints for this category under baseline relative to Fischer and Newby 1991; Short et al. 2004). Collectively, these placebo (a 34% decrease) and all four active MPH conditions findings suggest that several items included on emergent rat- (decreases between 43% and 59%); and under placebo rela- ing scales likely reflect phenotypical, behavioral features of tive to 5mg (25% decrease), 15-mg (31% decrease), and 20- ADHD rather than true emergent symptoms related to MPH mg (38% decrease) conditions. Parent endorsements showed therapy. Parent and teacher ratings of these scale items have fewer significant contrasts, with higher severity of physical previously been attributed to rater confusion. However, an and behavioral complaints reported under baseline relative equally plausible explanation is that raters are accurately re- to 5mg (28% decrease) and 20mg (27% decrease) MPH con- flecting observed behavior changes in children consistent ditions only. Idiographic examination revealed moderate in- with the vast literature on psychostimulant response, and increases for some children in nausea (n 11), itchy skin (n vestigators have misinterpreted these ratings by relying on 10), dry mouth (n 9), and reduced appetite (n 6) under total rather than item or factor scale scores. This explanation the two highest doses (15mg and 20 mg). also holds for scales developed specifically to monitor emergent symptoms associated with psychostimulant therapy, Limitations such as the Stimulant Drug Side Effects Rating Scale (Barkley 2006), which contains multiple items that mirror core and Although all three classes of potential emergent symptoms secondary features of ADHD (e.g., trouble sleeping, prone- showed significant decreases from baseline, several caveats ness to crying, daydreams). merit consideration. Increased frequency and/or severity of Children’s complaints of physical discomfort on a weekly emergent symptoms reported by or observed in children re- and reoccurring basis are well documented in the literature ceiving psychostimulant therapy are probable to the extent (Egger et al. 1999; DuPaul et al. 1996; Mitchell et al. 1987). that dosing regimens differ from the parameters reported Our results corroborate extant research in demonstrating herein, particularly in those symptoms highly specific to that children with ADHD share this propensity for experi- MPH. That is, children receiving multiple doses per day, sin- encing weekly discomfort such as headaches, feeling tired, gle doses exceeding 20 mg, different MPH formulations, and muscle aches, stomachaches, and occasional dizziness under MPH over a longer duration of time are likely to experience no-medication conditions. The results provide a compelling a higher frequency and/or severity of emergent symptoms rationale for recognizing that children, like adults, regularly (Gadow and Sverd 2006). The immediate-release MPH for- experience physical discomfort that must be recognized and mulary is currently used less frequently than newer formu- accounted for prior to initiating a medication trial. An un- lations, and the generalizability of the present findings to expected result, however, was the significant decrease in long-duration, sustained-release, and other variants is these complaints under placebo and MPH conditions rela- unknown. Preschool children with ADHD (Connor 2002; ADHD PHYSICAL/BEHAVIORAL COMPLAINTS 245

Kollins 2004), and those whose presentation includes other epidemiology, and longitudinal research. In: Children at Risk: clinical features such as developmental delays (Aman et al. Assessment, Longitudinal Research, and Intervention. Edited 1991; Handen et al. 1992), may also experience a higher fre- by Brambring M, Losel F. Oxford, England, Walter De quency and/or severity of psychostimulant-related emer- Gruyter, 1989, pp 24–45. gent symptoms. Because baseline assessments necessarily Ahman PA, Waltonen SJ, Olson KA, Theye FW, Van Erem AJ, preceded placebo and active medication conditions, and no LaPlant, RJ: Placebo controlled evaluation of Ritalin side ef- age-matched control group was included, the influence of fects. Pediatrics 91:1101–1106, 1993. halo, order, or regression to the mean effects on the signifi- Aman MG, Marks RE, Tubott SH, Wilsher CP, Merry SN: Clin- cant decreases in symptom severity between baseline and ical effects of methylphenidate and thioridazine in intellectu- ally subaverage children. J Am Acad Child Adolesc Psychia- placebo cannot be entirely eliminated. However, the pattern try 30:246–256, 1991. of symptom reductions from the counterbalanced placebo to American Psychiatric Association: Diagnostic and Statistical active medication conditions was similar to the pattern Manual of Mental Disorders (4th ed.). Washington, DC, Amer- shown from baseline (albeit smaller magnitude). Carryover ican Psychiatric Association, 2000. effects are unlikely due to counterbalancing and the inclu- Apley J: The child with Abdominal Pains (2nd ed.). London, sion of a “washout” day between all conditions. In addition, Blackwell, 1975. four days elapsed between the administration of the pre- Barea LM, Tannhauser M, Rotta NT: An epidemiologic study of ceding week’s final dosage and symptom ratings for the cur- headache among children and adolescents of southern Brazil. rent dosage. A final caveat involves the applicability of Cephalalgia 16:545–550, 1996. group-level results for titrating and monitoring psychostim- Barkley RA, McMurray MB, Edelbrock CS, Robbins K: Side ef- ulant regimens for individual children. Generalization can fects of methylphenidate in children with attention deficit hy- never be assumed, and is always limited by the highly idio- peractivity disorder: A systematic, placebo-controlled evalu- syncratic treatment response observed within and across be- ation. Pediatrics 86:184–192, 1990. havioral, cognitive, and emergent symptom domains in chil- Barkley RA: Attention Deficit Hyperactivity Disorder: A Hand- dren with ADHD (Rapport and Kelly 1991). book for Diagnosis and Treatment (3rd edition). New York, Guildford Press, 2006. Conclusions Beidel DC, Christ MA, Long PJ: Somatic complaints in anxious children. J Abnorm Child Psychol 19:659–670, 1991. Collectively, our findings point to a clear need to develop Belmaker E, Espinoza R, Pogrund R: Use of medical services by psychometrically sound treatment emergent symptom rat- adolescents with non-specific somatic symptoms. Int J Ado- ing scales for purposes of monitoring physical and behav- lesc Med Health 1:150–156, 1985. ioral complaints in children treated with psychostimulants Campo JV, Fritsch SL: Somatization in children and adolescents. (Greenhill et al. 2001b). Special care is warranted in word- J Am Acad Child Adolesc Psychiatry 33:1223–1235, 1994. ing scale items to ensure that their content differentiates be- Carlson GA, Kashani JH: Phenomenology of major depression tween emergent symptoms and core/secondary symptoms from childhood through adulthood: Analysis of three studies. of the disorder. This can be accomplished through extensive Am J Psychiatry 145:1222–1225, 1988. item refinement and submitting items to expert judges for Chronis, AM, Chacko, A, Fabiano, GA, Wymbs, BT, & Pelham, WE: Enhancements to the behavioral parent training para- review (see Clark and Watson 1995). The development of a digm for families of children with ADHD: Review and future separate factor scale that contains typical daily complaints directions. Clin Child Fam Psychol Rev 7: 1–27, 2004. endorsed by children based on extant research may also be Clark LA, Watson D: Constructing validity: Basic issues in ob- desirable. Separating these complaints from drug-related jective scale development. Psychol Assess 7:309–319, 1995. emergent symptoms affords practitioners improved ability Connor DF: Controlled trial of methylphenidate in preschool to assess the extent of a child’s usual complaint frequency children with minimal brain dysfunction. Int J Ment Health, and severity, and to determine whether it changes with treat- 4:61–74, 2002 ment as evidenced herein. Including items that help differ- Conners CK, Epstein JN, March JS: Multimodal treatment of entiate improved attention from staring or constricted at- ADHD in the MTA: An alternative outcome analysis. J Am tention are also desirable. Finally, the administration of Acad Child Adolesc Psychiatry 40:159–167, 2001. treatment emergent scales prior to initiating therapy must be Conners CK, Sitarenios G, Parker JDA, & Epstein JN: Revision considered a de rigueur component of clinical management and restandardization of the Conners Teacher Rating Scale for establishing a true baseline of child complaints (Barkley (CTRS-R): Factor structure, reliability, and criterion validity. et al. 1990; Greenhill et al. 2001a). J Abnorm Child Psychol 26: 279–291, 1998. Denney CB, Rapport MD: Cognitive pharmacology of stimulants Disclosures in children with ADHD. In: Stimulant Drugs and ADHD: Ba- sic and Clinical Neuroscience. Edited by Solanto MV, Amsten Drs. Rapport and Coiro and Mr. Kofler, Mr. Raiker, Mr. AFT. New York, Oxford University Press, 2001, pp 283–302. Alderson, and Mr. Sarver have no financial, corporate, or Dunn LM, Dunn LM: Peabody Picture Vocabulary Test-Third commercial relationships to disclose. Edition. Circle Press, Minn.: American Guidance Service, 1997. DuPaul GJ, Rapport MD: Does methylphenidate normalize the References classroom performance of children with attention deficit disorder? J Am Acad Child Adolesc Psychiatry 32:190–198, 1993. Abu-Arefeh I, Russell G: Recurrent limb pain in school. Arch DuPaul GJ, Anastopoulos AD, Kwasnick D, Barkley RA, Mc- Dis Child 74:336–339, 1996. Murray MB: Methylphenidate effects on children with atten- Achenbach TM: Empirically based assessment of child and ado- tion deficit hyperactivity disorder: Self-report symptoms, side lescent disorders: Implications for diagnosis, classification, effects and self-esteem. J Atten Disord 1:3–15, 1996. 246 RAPPORT ET AL.

DuPaul GJ, Eckert TL: Interventions for students with Attention- Guy W: ECDEU Assessment Manual for Psychopharmacology Deficit/Hyperactivity Disorder: One size does not fit all. (revised). Rockville, Maryland, Department of Health, Edu- School Psych Rev 26:369–381, 1997. cation and Human Welfare Publication No. (ADM) 76–338, Efron D, Jarman F, Barker M: Side effects of methylphenidate 1976. and dexamphetamine in children with attention deficit hy- Handen BL, Breaux AM, Janosky J, McAuliffe S, Feldman H, peractivity disorder: A double-blind, crossover trial. Pedi- Gosling H: Effects and noneffects of methylphenidate in chil- atrics 100:662–666, 1997. dren with mental retardation and ADHD. J Am Acad Child Egger H, Costello EJ, Angold A: Headaches and psychopathol- Adolesc Psychiatry 31:455–461, 1992. ogy in children and adolescents. J Am Acad Child Adolesc Hollingshead AB: Four Factor Index of Social Status. New Haven Psychiatry 37:951–958, 1999. (Connecticut), Yale University, 1975. Faraone SV, Biederman J, Roe C: Comparative efficacy of adder- Jensen PS: Fact versus fancy concerning the multimodal treat- all and methylphenidate in attention-deficit/hyperactivity ment study for attention-deficit hyperactivity disorder. Can J disorder: A meta-analysis. J Clin Psychopharmacol 22:468– Psychiatry 44:975–980, 1999. 473, 2002. Kaufman J, Birmaher B, Brent D, Rao U, Flynn C, Moreci P, Findling RL, Short EJ, Manos MJ: Developmental aspects of psy- Williamson D, Ryan N: Schedule for Affective Disorders and chostimulant treatment in children and adolescents with at- Schizophrenia for School-Age Children-Present and Lifetime tention-deficit/hyperactivity disorder. J Am Acad Child Version (K-SADS-PL): Initial reliability and validity data. J Adoelsc Psychiatry 40:1441–1447, 2001. Am Acad Child Adolesc Psychiatry 36:980-988, 1997. Fine S, Johnston C: Drug and placebo side effects in Keppel G: Design and Analysis: A Researcher’s Handbook methylphenidate-placebo trial for attention deficit hyperac- (3rd ed.). Upper Saddle River (New Jersey), Prentice Hall, tivity disorder. Child Psychiatry Hum Dev 24: 25–30, 1993. 1991. Firestone P, Musten LM, Pisterman S, Mercer J, Bennett S: Short- King JE: Software solutions for obtaining kappa-type statistic for term side effects of stimulant medication are increased in use with multiple raters. Dallas (Texas), Southwest Educa- preschool children with attention deficit hyperactivity disor- tional Research Association Annual Meeting, 2004. der: A double-blind placebo-controlled study. J Child Adolesc Kollins SH, the PATS Study Group: Preschool ADHD Treatment Psychopharmacol 8:13–25, 1998. Study. Phoenix (Arizona), National Institute of Mental Health, Fischer M, Newby RF: Assessment of stimulant response in New Clinical Drug Evaluation Unit (NCDEU) Annual Meet- ADHD children using a refined multimethod clinical proto- ing, 2004. col. J Clin Child Adolesc Psychol 20:232–244, 1991. Last C: Somatic complaints in anxiety disordered children. J Gadow J, Sverd J: Attention Deficit Hyperactivity Disorder, Anxiety Disord 5:125–138, 1991. Chronic Tic Disorder, and Methylphenidate. In: Advances in Linna S, Moilanen I, Keistinen H, Ernvall ML: Prevalence of psy- Neurology: Tourette Syndrome Vol. 99. Edited by Walkup J, chosomatic symptoms in children. Psychother Psychosom Mink J, Hollenbeck P. Philadelphia, Lippincott, 2006, pp 56:85–87, 1991. 197–207. Masi G, Favilla L, Millepiedi S, Mucci M: Somatic symptoms in Garber J, Walker L, Zeman J: Somatization symptoms in a com- children and adolescents referred for emotional and behav- munity sample of children and adolescents: Further valida- ioral disorders. Psychiatry 63:140–149, 2000. tion of the Children’s Somatization Inventory. Psychol Assess Mitchell EA, Aman MG, Turbott SH, Manku M: Clinical char- 3:588–595, 1991. acteristics and serum essential fatty acid levels in hyperactive Gittelman-Klein R, Klein DF: Methylphenidate effects in learn- children. Clin Pediatr (Phila) 26:406–411, 1987. ing disabilities: Psychometric changes. Arch Gen Psychiatry, Multimodal Treatment Study of Children with ADHD Cooper- 33:655–664, 1976. ative Group: Moderators and mediators of treatment response Gorman EB, Klorman R, Thatcher JE, Borgstedt AD: Effects of for children with attention-deficit/hyperactivity disorder: The methylphenidate on subtypes of attention-deficit/hyperactiv- Multimodal treatment study of children with attention- ity disorder. J Am Acad Child Adolesc Psychiatry 45:808–816, deficit/hyperactivity disorder. Arch Gen Psychiatry 56:1088– 2006. 1096, 1999. Grcevich S, Rowane WA, Marcellino B, Sullivan-Hurst S: Ret- Nunnally JC, Bernstein IH: Psychometric Theory (3rd ed.). New rospective comparison of Adderall and methylphenidate in York, McGraw-Hill, 1994. the treatment of attention-deficit/hyperactivity disorder. J Pelham WE, McBurnett K, Harper GW, Milich R, Murphy DA, Child Adolesc Psychopharmacol, 11:35–41, 2001. Clinton J, Thiele C: Methylphenidate and baseball playing in Greenhill L, Swanson JM, Vitiello B, Davies M, Clevenger W, ADHD children: Who’s on first? J Consult Clin Psychol Wu M, Arnold LE, Abikoff HB, Bukstein OG, Conners CK, El- 58:130–133, 1990. liott GR, Hechtman L, Hinshaw SP, Hoza B, Jensen PS, Krae- Pine DS: Treating children and adolescents with selective sero- mer HC, March JS, Newcorn JH, Severe JB, Wells K, Wigal T: tonin reuptake inhibitors: How long is appropriate? J Child Impairment and deportment responses to different Adolesc Psychopharmacol 2:189–203, 2002. methylphenidate doses in children with ADHD: The MTA Rapport MD, Kelley KL: Psychostimulant effects on learning and titration trial. J Am Acad Child Adolesc Psychiatry, cognitive function: Findings and implications for children 40:180–187, 2001a. with attention deficit hyperactivity disorder. Clin Psychol Rev Greenhill LL, Vitiello B, Abikoff H, Levine J, March JS, Riddle 11:61–92, 1991. MA, Capasso L, Cooper T, Davies M Fisher P, Findling RL, Rapport MD, Denney C, DuPaul GJ, Gardner MJ: Attention Fried J, Labellarte M, McCracken J, McMahon DJ, Robinson J, deficit disorder and methylphenidate: Normalization rates, Skrobala AM, Scahill L, Walkup J, Zito J: Improving the meth- clinical effectiveness, and response prediction in 76 children. ods for evaluating the safety of psychotropic medications in J Am Acad Child Adolesc Psychiatry 33:882–893, 1994. children and adolescents. Current Therapeutic Res 62:873–884, Rapport MD, Denny C: Titrating methylphenidate in children 2001b. with attention-deficit/hyperactivity disorder: Is body mass ADHD PHYSICAL/BEHAVIORAL COMPLAINTS 247

predictive of clinical response? J Am Acad Child Adolesc Psy- Swanson JM, Cantwell D, Lerner M, McBurnett K, Hanna G: Ef- chiatry 36:523–530, 1997. fects of stimulant medication on learning in children with Rapport MD, Chung K-M, Shore G, Issacs P: A conceptual model ADHD. J Learn Disabil 24:219–230, 1991. of child psychopathology: Implications for understanding at- Swanson JM, Volkow ND: Pharmacokinetic and pharmacody- tention deficit hyperactivity disorder. J Clin Child Adolesc namic properties of stimulants: Implications for the design of Psychol 30:48–58, 2001. new treatment for ADHD. Behav Brain Res 130:73–78, 2002. Rapport MD, Moffitt C: Attention-deficit/hyperactivity disor- Taylor DC, Szatmari P, Boyle MH, Offord DR: Somatization and der and methylphenidate: A review of height/weight, car- the vocabulary of everyday bodily experiences and concerns: diovascular, somatic complaint side effects. Clin Psychol Rev A community study of adolescents. J Am Acad Child Adolesc 22:1107–1131, 2002. Psychiatry 35:491–499, 1996. Rapport MD, Moffit C, Randall R: Attention-deficit/hyperac- Vaessen W, & Van der Meere JJ: Issues in the selection of hy- tivity disorder and methylphenidate: A dose response analy- peractive/ADD(H) children for experimental clinical studies. sis and parent-child comparisons of somatic complaints. J At- In Developmental Biopsychology: Experimental and Obser- ten Disord 6:15–24, 2002. vational Studies in Groups “at Risk”. Edited by A. F. Kalver- Routh DK, Schroeder CS, O’Tauma L: Development of activity boer. Ann Arbor: University of Michigan Press, 1990, pp. level in children. Dev Psychol 10:163–168, 1974. 21–37. Short EJ, Manos MJ, Findling RL, Schubel EA: A prospective study of stimulant response in preschool children: Insights from ROC analyses. J Am Acad Child Adolesc Psychiatry 43:251–259, 2004. Address reprint requests to: Shrout PE, Fleiss JL: Intraclass correlations: Uses in assessing Mark D. Rapport, Ph.D. rater reliability. Psychol Bull 2: 420–428, 1979. Department of Psychology Siegel S, Castellan NJ: Nonparametric Statistics for the Behav- 4000 Central Florida Blvd. ioural Sciences (2nd ed.). New York, McGraw-Hill, 1988. Orlando, Fl 32816 Stewart-Williams S: The placebo puzzle: Putting together the pieces. Health Psychol 23:198–206, 2004. E-mail: [email protected]