Stimulant Treatment in Children with Attention-Deficit/Hyperactivity Disorder Moderates Adolescent Academic Outcome

Abstract

Treatment with psychostimulant medication has been shown to improve scholastic functioning in children with attention-deficit/hyperactivity disorder (ADHD). However, the extent to which long-term academic gains are apparent in those having received such treatment remains elusive. This study examined prospectively the relationship of childhood stimulant treatment to academic functioning during adolescence. Children (n = 169) were initially recruited and diagnosed with ADHD when they were 7–11 years old. A subsample of those with childhood ADHD (n = 90) was reevaluated on average 9.13 (SD = 1.5) years later. Probands who did and did not receive treatment with stimulant medication were compared to each other and to a never-ADHD comparison group (n = 80) on three subtests from the Wechsler Individual Achievement Test-II (WIAT-II), as well as high school grade point average (GPA) and number of retentions in school as derived from school records. Analyses of covariance controlling for severity of childhood ADHD symptoms indicated that probands treated with psychostimulant medication achieved better academic outcomes, as measured by WIAT-II subtests and high school GPA, than those not treated with psychostimulants (p < .05). However, treated probands did not fare as well as the never-ADHD comparison group. Psychostimulant treatment for children with ADHD may benefit long-term adolescent academic performance, although the extent of improvement is likely to vary as a function of multiple factors.

A ttention-deficit/hyperactivity disorder (ADHD) is a common psychiatric disorder characterized by symptoms of hyperactivity, impulsivity, and inattention that begin in early childhood and continue to cause impairment into adolescence and young adulthood (Manuzza et al. 1997; Barkley et al. 2002; Biederman and Faraone 2005; Hechtman and Weiss 1983). Relative to their typically developing counterparts, children with ADHD have been shown in numerous studies to be at a significantly greater risk for poor academic performance, as indicated by higher rates of retention and special education placement, and the need for more frequent academic tutoring or adjunctive services (Faraone et al. 1993; Barkley et al. 1990).

Prospective studies of children with ADHD have demonstrated impaired academic achievement into adolescence (Fischer et al. 1990; Mannuzza et al. 1993). Mannuzza et al. (1997) prospectively examined the long-term academic outcomes of children with ADHD, and found that ADHD probands had significantly less schooling than controls when followed-up 15–21 years later; subsequent analyses indicated that this finding was not attributable to group differences in intellectual functioning. Other prospective studies corroborate these findings, with reports of an educational disadvantage for adults with ADHD (Weiss et al. 1985; Mannuzza et al. 1993; Lambert 1988; Rasmussen and Gillberg 2000). In addition, retrospective studies of adolescents and young adults with ADHD have shown poor academic achievement during childhood (Torgersen et al. 2006, Murphy and Barkley 1996)

There are likely many factors which contribute to poor academic performance in individuals with ADHD (Barkley 1997). One factor is comorbid learning disabilities (LD) (Halperin et al. 1984; August and Garfinkel 1990). Estimates of comorbid LD among children with ADHD vary widely, depending on the definition used, with conservative estimates indicating that at least 20–25% of children are affected (Hinshaw 1992). However, children with ADHD have poor academic performance above and beyond the potential effects of comorbid LD (Wu et al. 2002; August and Garfinkel, 1990), suggesting that poor academic outcome in ADHD may be more closely linked to inattention (Hinshaw 1992), executive dysfunction (Barkley 1997) and/or externalizing behavior problems (Loney et al. 1981).

Stimulant treatment during childhood has not been shown to improve academic achievement (e.g., reading scores) in children with LD alone (Gittelman et al. 1983). Nevertheless, there are several possible mechanisms by which stimulant treatment could facilitate long-terms gains in academic achievement in children with ADHD. First, stimulant treatment has been shown to affect brain regions involved in executive functioning, increasing frontal activation and thereby enhancing inhibitory control (Vaidya 1998) as well as sustained effort and attention on neuropsychological tests; presumably the same benefits should be achieved within the classroom setting and during homework times (Evans et al. 2001). Thus, successful treatment of ADHD with stimulant medication is likely to improve neurocognitive function in ways that would be expected to enhance readiness to acquire academic skills.

Moreover, children with ADHD show beneficial effects of treatment on aspects of functioning in the classroom (e.g., increased accuracy and productivity; decreased classroom disruptions), in home/recreational settings (Chacko et al. 2005; Pelham et al. 1993; Greenhill et al. 2002; Hechtman et al. 2004; Evans and Pelham 1991), as well as on tests of academic efficiency (Swanson et al. 1998; Swanson et al. 2004). As such, it seems reasonable to expect that these treatment-related improvements might translate into improved academic outcomes.

Finally, the fact that stimulants decrease disruptive behaviors should result in children spending more time in academically-oriented classes and with peers who are more motivated to perform well in school. Thus, acute stimulant treatment that is successful in reducing ADHD symptoms and related disruptive behaviors not only has direct effects on the child which are likely to facilitate learning via enhanced cognitive processing and improved effort and productivity, but also, secondarily, increase the likelihood that the child will remain in a social context that is more encouraging of academic success.

Nevertheless, despite the potential for stimulant treatment to affect brain functions in a way that might be expected to produce improved academic performance, and the hypotheses that acute treatment-related improvements should lead to longer term gains in academic function, it remains unclear whether long-term gains in academic skills are realized with treatment. An early review of 17 studies conducted between 1940 and 1976 (Barkley and Cunningham 1978) concluded that there was no evidence to suggest lasting positive effects of stimulant medication on academic outcome of children with ADHD. However, the duration of treatment in the studies reviewed ranged only from two weeks to “4–6 months,” with only two studies having treatment duration of more than 12 weeks. As such, it is not surprising that lasting gains in academic achievement were not evident.

A more recent review, which included studies through the 1980s and early 1990s (Schachar and Tannock 1993) reported somewhat more encouraging, albeit “inconclusive” results. This review concluded that short-term studies provide consistent evidence of increased academic productivity, but only six of 10 randomized trials found some evidence of long-term academic gains, none of which was particularly robust. In general, positive findings were identified on one or two subtests of among several administered. However, again, only one of these 10 studies had treatment duration of greater than one year. Among the non-randomized studies reviewed, only one, which administered multimodal treatment in an open, uncontrolled design (Satterfield et al. 1981), reported long-term academic gains.

Results of recent studies have suggested some beneficial effects of treatment on academic achievement, although the findings remain far from conclusive. Paternite, Loney, Salisbury and Whaley (1999) reported young adult outcome data for a group of 121 boys who were diagnosed with either hyperkinetic reaction of childhood, as defined by DSM-II, or “minimal brain dysfunction” when they were 4 through 12 years old; all were subsequently treated with methylphenidate. Participants had an average treatment duration of 30.4 months (range = 1–76 months) and all had terminated treatment prior to the follow-up assessment. After controlling for severity of childhood symptoms, neither medication response nor dosage was associated with academic outcome. However, after controlling for all of the aforementioned variables, duration of treatment was associated with significantly higher reading scores; a trend was also detected, such that longer treatment duration was associated with improved math scores.

The MTA study also assessed academic functioning in treated and untreated children. Following 14 months of treatment, children with ADHD who were randomized to receive the protocol-driven medication treatment were found to have improved academic achievement as measured by the Wechsler Individual Achievement Test (WIAT), compared to children assigned to receive psychosocial treatment without stimulant medication and those receiving treatment-as-usual in the community (MTA Cooperative Group 1999). However, in both the 24- and 36-month assessments (10- and 22-month follow-up from the end of the protocol-driven treatment), group differences in academic achievement were no longer evident (MTA Cooperative Group 2004; 2007). While it is difficult to know with certainty why the early findings in favor of improved academic achievement were not maintained, the loss of treatment benefits may be at least partially attributable to the fact that approximately half of the children in the “non-medicated group” were receiving psychostimulant treatment at the time of the 36-month follow-up, suggesting that the treatment groups were no longer distinct. Moreover, the MTA Study cannot, as of yet, elucidate the impact of childhood stimulant treatment on scholastic achievement and related outcome in later adolescence, as the participants in the most recent publications were only 10–13 years of age.

Finally, a recent study (Barbaresi et al. 2007) found modest associations between stimulant treatment and academic outcome in a population-based birth cohort that included 370 youth with ADHD who were followed from school entry through high school graduation. Specifically, these investigators found that children treated with stimulants had lower rates of school absenteeism and were less likely to be retained. Further, a modest, but statistically significant correlation (r = .15, p = .012) was identified between daily dose of medication and last reading score. However, this last reading score was acquired at a mean age of 12.8 years, thus not really reflective of long-term outcome.

Thus, while the extant literature provides compelling evidence for short-term gains in academic productivity for children with ADHD who are being treated with stimulant medication, there is far more limited support for lasting academic gains into adolescence or young adulthood. This lack of clarity with regard to long-term outcome is at least in part due to the fact that ethical constraints preclude the possibility of randomized control studies; children cannot be assigned to no- or non-optimal-treatment groups for extended periods of time. Even studies that initially employed random assignment to group, such as the MTA Study (MTA Cooperative Group 2007), eventually lose the initially-determined group distinctions; fewer than two years after the termination of the controlled trial, approximately half of the children in the “Psychosocial-only” group had received treatment with stimulant medication. Thus, the elucidation of the long-term impact of treatment during childhood can only be assessed though less-than-perfect naturalistic follow-up studies, which are plagued by difficulties associated with non-random assignment to group, and in particular, self-selection as to who receives treatment.

Yet, even among the published follow-up studies, there are none that included all of the following: 1) participants who were initially diagnosed with ADHD using contemporary diagnostic criteria; 2) a substantial number of participants who never received any treatment with stimulant medication as well as many with several years of such treatment; and 3) academic outcome measures obtained during the late adolescent/young adult period. Further, almost all long-term follow-up research in the area of ADHD has been conducted with middle class Caucasian families.

The current study examined prospectively the relationship between stimulant treatment of ADHD during childhood and academic outcome in an ethnically-diverse longitudinal sample of adolescents/young adults (for ease of reading we will heretofore refer to the participants as “adolescents”) who were diagnosed with ADHD during the early to mid 1990s when they were 7–11 years-old. Because these children were not originally recruited into a treatment study, a substantial proportion never received any treatment with stimulant medication. Nevertheless, many did receive consistent and extended treatment lasting as long as 5–12 years.

Based on the facts that stimulant medication results in enhanced neurocognitive functioning (e.g., increased attention and inhibitory control), increased academic productivity and effort, and a greater likelihood that a child will remain in a social milieu that is encouraging of academic success, we hypothesized that those receiving treatment with stimulant medication during childhood would have improved academic outcomes in adolescence relative to those with no prior history of stimulant treatment, and would attain academic achievement more similar to never-ADHD controls.

Method

Participants

Probands were 90 adolescents (80 male; 88%), drawn from a group of 169 youth who were diagnosed with ADHD during childhood (mean age at baseline = 9.11 years, SD = 1.22 years) following a comprehensive psychiatric assessment in a research study conducted between 1990 and 1997. Of these 169 participants, 18 refused participation in the follow-up, one was known to be deceased, seven were incarcerated, and 53 were lost to follow-up. Those who were and were not assessed at follow-up did not differ significantly with regard to age at initial evaluation, rates of childhood comorbid diagnoses, Full Scale IQ, socio-economic status (SES), or ADHD behavior ratings at initial assessment (all p > .10).

Adolescents with childhood ADHD were compared to 80 control participants (71 male; 89%) who were recruited as adolescents. Controls had no prior history of ADHD and were similar to probands with respect to age, SES, and race/ethnicity. All participants were between 16 and 22 years of age, with a mean (SD) age of 18.41 (1.51) years. The families were of primarily lower to lower middle class status with a mean (SD) SES score of 42.30 (17.09) on a measure of socio-economic prestige (Nakao and Treas 1994). However, a range of socioeconomic levels was represented (range = 20–96). The sample was ethnically diverse with 23.5% describing themselves as Non-Hispanic Caucasian, 25.9% African American, 30.9% Hispanic or Latino, 1.2% Asian, and 18.8% of mixed or other ancestry.

Procedures

Baseline evaluation

Cognitive assessment

At baseline, probands were administered the Wechsler Intelligence Scale for Children-Revised (WISC-R; Wechsler 1974; n = 41) or the Wechsler Intelligence Scale for Children-Third Edition (WISC-III; Wechsler 1991; n = 49), depending upon when they were recruited, and the Wide Range Achievement Test Revised (WRAT-R; Jastak and Wilkinson 1984; n = 41) or the screener from the WIAT (Wechsler 1992; n = 49), again depending upon when they were recruited.

Psychiatric assessment

All children were medication free for a minimum of 30 days prior to entry into the study and the collection of parent and teacher ratings. Children were initially screened using parent and teacher ratings on the Child Behavior Checklist (CBCL; Achenbach 1991) and IOWA Conners' Teachers questionnaire (IOWA; Loney and Milich 1982), respectively. Those with teacher ratings on the Inattention/Overactivity scale of the IOWA Conners' in the “clinical” range (Loney and Milich 1982) were invited for possible participation in the study. Those eventually included met criteria for ADHD during childhood based on a structured clinical interview using the Diagnostic Interview Schedule for Children (DISC) Version 2.3 (Shaffer et al. 1989), which reflects diagnostic criteria from DSM-III-R (American Psychiatric Association 1987; n = 41) or Version 3.0 (Shaffer et al. 1996), which incorporates diagnostic criteria from DSM-IV (American Psychiatric Association 1994; n = 49). Those evaluated using the DISC 3.0/DSM-IV all met criteria for ADHD, Combined Type. DSM-III-R did not have distinct subtypes; however, those diagnosed using DSM-III-R and DSM-IV criteria did not differ significantly on any CBCL or IOWA ratings. As such, these two subsamples were quite similar. Individuals with a chronic medical condition requiring systemic medication, a diagnosed neurological disorder, schizophrenia, pervasive developmental disorder (PDD), Tourette disorder, or a Full Scale IQ below 70 were excluded from the baseline protocol.

Adolescent evaluation

Probands and their families were contacted to participate in a follow-up assessment on average 9.13 (SD = 1.5) years following their initial evaluation. Mean (SD) age at follow up was 18.4 (1.5) years. Community control participants were recruited by placing flyers and advertisements in areas where probands lived and were similar to probands with regard to sex, age, race/ethnicity and socioeconomic status. Individuals with a history of ADHD, an IQ less than 70, chronic medical conditions, schizophrenia, PDD, and/or diagnosed neurological disorders were excluded from participation in the control group.

Following a complete description of the study, for those under the age of 18 years, written informed consent was obtained from the parent(s) or legal guardian(s) and written informed assent was obtained from the adolescents. Written informed consent was obtained from participants over the age of 18. This study was approved by the Institutional Review Boards of the affiliated academic institutions.

Psychoeducational assessment/academic outcomes

Intellectual abilities were assessed during adolescence using the Wechsler Adult Intelligence Scale Third Edition (WAIS-III; Wechsler 1997). Three subtests from the WIAT-II (Wechsler 2001)—Word Reading, Pseudoword Decoding, and Numerical Operations—were used to assess academic achievement. Parents and adolescents were interviewed about the adolescent's school history, and parents also completed a form which included inquiries about the number of times their child repeated a grade in school, as well as the actual grade(s) repeated. Additionally, we were able to acquire transcripts from the last high school the participant attended for a portion of the sample (43 probands and 34 controls). Probands for whom transcripts were and were not retrieved did not differ significantly with regard to childhood or adolescent IQ, academic achievement scores, or parent and teacher (childhood only) ratings of behavior. However, among controls, those for whom high school records were acquired had significantly higher IQ scores (100.9 vs, 92.8, p = .015). Because high schools vary with regard to grading systems (i.e., letter grades vs. different types of number grades) grade point averages were converted to a five-point scale to facilitate comparisons among varied types of grade point systems (see Table 1). Those probands who were currently being treated with stimulant medication for ADHD (n = 15) did not take their medication on the day of the assessment.

Table 1.

Grade Point Average Conversion Chart ^a

Converted 5-point Scale	4-Point Grade Scale	Letter-Grade Scale	100-Point Grade
1=	≤1.0 =	F=	≤65
2=	1.1–2.0 =	D=	66–75
3=	2.1–3.0 =	C=	76–85
4=	3.1–3.5 =	B=	86–95
5=	3.6–4.0 =	A=	96–100

Grade point averages converted to a 5-point scale to facilitate comparisons among varied types of grade point systems.

Medication status

Parents were administered a “Services Received Interview” by a member of the study team to acquire a complete history of any type of treatment the participant had ever received, inclusive of any type of counseling, therapy and/or pharmacotherapy. In addition, probands were separately interviewed about treatments they had received. Specific information regarding duration, type, and age of treatment was systematically queried. This information was supplemented by a review of childhood (i.e., baseline) records for ADHD probands only, which also included information regarding childhood medication status and history (i.e., treatment received prior to the baseline assessment). Based upon all available data, a narrative was constructed detailing the treatment history of each participant. Based upon the narrative, those with a childhood history of ADHD were divided into two subgroups; medicated (ADHD-M) and unmedicated (ADHD-U). To be included in the ADHD-M group, adolescents were required to have received stimulant medication treatment and complied with that treatment consistently for a minimum of one year (n = 48, 90% male). The mean (SD, range) number of years children in the ADHD-M group received stimulant medication treatment was 5.33 (3.02, 1–12) years. The ADHD-U group (n = 42, 86% male) consisted of those who were either stimulant naive (n = 27), or were briefly or inconsistently treated with stimulant medication (n = 15), never reported to exceed one year.

Data analysis

A multivariate analysis of variance (MANOVA) was employed to test the a priori hypotheses that ADHD-M probands would achieve significantly higher academic scores than ADHD-U probands, and that ADHD-M probands would attain similar academic achievement to control participants. A significant MANOVA was followed by one-way ANOVAs that used the Bonferroni-Holm procedure to correct for multiple comparisons, and post-hoc Tukey's Honestly Significant Difference (HSD) contrasts to determine individual group differences. Dependent variables for the MANOVA were the Word Reading, Pseudo-word Decoding, and Numerical Operations scores from the WIAT-II (Wechsler 2001). A one-way ANOVA was used to assess differences among groups with regard to high school grade point average using the GPA conversion method presented in Table 1. Post-hoc analyses were conducted with Tukey's HSD tests to control for multiple comparisons on all analyses of variance. Due to the fact that data regarding the number of grade repetitions were positively skewed, a non-parametric Kruskal-Wallis test was used to examine group differences, with Mann-Whitney U tests subsequently conducted to assess pairwise comparisons. In addition a chi-square analysis was conducted to examine the extent to which ADHD-U, ADHD-M, and controls differed with respect to the dichotomous incidence of grade repetition (i.e., whether or not the participant ever repeated a grade).

Because the ADHD-M and ADHD-U groups differed significantly in childhood with regard to their age at initial evaluation as well as ratings on the IOWA Inattention/Overactivity scale, we reanalyzed these data comparing the two proband subgroups utilizing these variables as covariates. Controls could not be included in these analyses because they were recruited during adolescence.

Results

Childhood and adolescent characteristics

As shown in Table 2, the ADHD-U and ADHD-M groups differed with regard to age at initial evaluation and teacher ratings of Inattention/Overactivity in childhood, but did not differ in parent ratings of behavior or measures of childhood academic achievement. Consistent with dimensional externalizing ratings shown in Table 2, rates of oppositional defiant disorder (ODD, 48% vs. 54%) and conduct disorder (CD; 26% vs. 31%) did not differ significantly between the ADHD-U and ADHD-M groups (both p > .50). Further, as indicated in Table 3, there were no significant group differences between ADHD-U, ADHD-M, and control participants during adolescence with respect to age, SES, FSIQ, and race/ethnicity. As expected, controls differed significantly from both groups of probands with regard to ADHD symptom severity. However, as shown in Table 3, the ADHD-U and ADHD-M groups did not differ significantly in adolescent symptom severity as indicated by self and parent reports.

Table 2.

Childhood Characteristics of Probands as a Function of Medication Treatment Status

Variable	ADHD-U n = 48 Mean (SD)	ADHD-M n = 42 Mean (SD)	t	p
Age at initial evaluation	9.42 (1.23)	8.80 (1.15)	−2.50	0.01
FSIQ (Childhood)	90.51 (13.66)	96.45 (14.71)	1.95	NS^a
WRAT-R/WIAT Math Score	91.14 (12.86)	95.09 (15.71)	1.29	NS
WRAT-R/WIAT Reading Score	91.36 (15.28)	94.3 (19.67)	0.781	NS
CBCL Externalizing T-Score	67.7 (10.87)	71.19 (10.76)	1.47	NS
CBCL Internalizing T-Score	65.43 (11.79)	65.05 (11.08)	−0.15	NS
CBCL Aggression T-Score	69.28 (13.2)	73.51 (13.91)	1.42	NS
CBCL Delinquency T-Score	66.6 (9.41)	68.79 (9.92)	1.03	NS
CBCL Attention T-Score	70.6 (9.98)	73.93 (9.6)	1.55	NS
IOWA Aggression	8.1 (4.62)	7.89 (4.73)	−0.21	NS
IOWA Inattention/Overactivity	10.20 (3.47)	12.25 (2.62)	3.07	0.003

p > .05.

ADHD-U = attention-deficit/hyperactivity disorder-unmedicated; ADHD-M = attention-deficit/hyperactivity disorder-medicated; FSIQ = Full Scale IQ; WRAT-R = Wide Range Achievement Test-Revised; WIAT = Wechsler Individual Achievement Test; CBCL = Child Behavior Checklist; IOWA = Inattention-Overactivity with Aggression; NS = not significant.

Table 3.

Adolescent Characteristics of Participants as a Function of Group and Medication Treatment Status

	ADHD-U (unmedicated) n = 42	ADHD-M (medicated) n = 48	Controls n = 80	F/x ²	p
Gender (M/F)	36/6	43/5	71/9	0.36	NS^a
Age: Mean (SD)	18.24 (1.35)	18.20 (1.38)	18.63 (1.64)	1.61	NS
SES: Mean (SD)	41.6 (17.16)	45.75 (17.76)	40.59 (16.56)	1.42	NS
FSIQ: Mean (SD)	90.02 (11.73)	95.37 (16.48)	96.26 (14.91)	2.57	NS
Race				8.49	NS
% Caucasian	19.0 (8/42)	31.3 (15/48)	21.3 (17/80)
% African American	31.0 (13/42)	16.7 (8/48)	28.7 (23/80)
% Hispanic or Latino	26.0 (11/42)	33.0 (16/48)	31.0 (25/80)
% Asian	0.0 (0/42)	0.0 (0/48)	2.5 (2/80)
% Mixed/other	23.8 (10/42)	18.8 (9/48)	16.3 (13/80)
ADHD checklist, self-report^b	14.93 (10.8)	19.54 (13.5)	4.66 (5.4)	38.97	<.001
ADHD checklist, parent^b	21.14 (13.6)	26.64 (13.7)	5.74 (7.7)	57.15	<.001

p > .05.

ADHD-U = ADHD-M > Controls; p < .05.

ADHD-U = attention-deficit/hyperactivity disorder-unmedicated; ADHD-M = attention-deficit/hyperactivity disorder-medicated; Gender M = male; Gender F = female; SES = socioeconomic status; FSIQ = Full Scale IQ; NS = not significant.

Academic achievement

WIAT-II

The MANOVA applied to subtests from the WIAT-II was statistically significant (Wilks' Lambda = .885, p = 0.003, η _p ² = .06). Subsequent ANOVAs revealed statistically significant group differences with respect to all three subtest scores Word Reading: F (2, 164) = 9.73, p < 0.001, η _p ² = .11; Pseudoword Decoding, F (2, 164) = 9.22, p < 0.001, η _p ² = .10; Numerical Operations: F (2, 164) = 4.07, p = 0.019, η _p ² = .05). As depicted in Figure 1, significant linear trends characterized all three academic achievement measures such that controls performed the best and the ADHD-U group performed most poorly. Follow-up post-hoc analyses, utilizing Tukey's HSD procedure indicated that, for Word Reading, the control group scored significantly higher than ADHD-U participants (p < 0.001), and marginally higher than ADHD-M participants (p = 0.06). Although the ADHD-M group did score higher than the ADHD-U group, this difference did not reach statistical significance (p > .10). For Pseudoword Decoding, the control group scored significantly higher than the ADHD-U participants (p < 0.001), but not significantly higher than the ADHD-M participants (p = 0.15). Although the ADHD-M group scored higher than the ADHD-U group, this difference fell just short of statistical significance (p = 0.06). Finally, for the Numerical Operations subtest, the control group scored significantly higher than the ADHD-U, but not ADHD-M participants (p = 0.01 and p = 0.45, respectively); ADHD-M and ADHD-U participants did not differ significantly from one another (p = 0.28).

FIG 1.

WIAT-II Test Scores. A significant linear trend is observed for all three subtests of academic achievement measures such that controls performed the best and the ADHD-U group performed most poorly (all F's > 7.70, p < .01). WIAT-II = Wechsler Individual Achievement Test-Second Edition.

Because the ADHD-U group tended to have lower adolescent FSIQ scores (p = .08) as compared to the controls and the ADHD-M group, another MANOVA was conducted using adolescent FSIQ as a covariate. Significant results remained (Wilks-Lambda = 0.890, p = 0.005, η _p ² = .06), with univariate analyses revealing significant differences for both Word Decoding, (F (2, 162) = 8.55, p < 0.001, η _p ² = .10) and Pseudoword Reading (F (2, 162) = 7.20, p = 0.001, η _p ² = .08); group differences in Numerical Operations no longer reached statistical significance (F (2, 162) = 1.81, p = 0.167, η _p ² = .02).

Analyses of covariance (ANCOVA) comparing ADHD-U and ADHD-M probands, controlling for age at initial evaluation and IOWA Inattention/Overactivity ratings from childhood, revealed statistically significant differences with respect to all three adolescent achievement test scores: Word Reading: F (1, 78) = 6.36, p = .014, η _p ² = .075; Pseudoword Decoding, F (1, 78) = 6.71, p = .011, η _p ² = .079; Numerical Operations: F (1, 78) = 4.86, p = .03, η _p ² = .059, such that those who were previously medicated performed better on all three measures.

In addition, to shed light on the extent to which the ADHD-U and ADHD-M groups differentially changed in academic achievement over development, we conducted separate two-way (Group × Time) ANOVAs with reading and math scores serving as the dependent variables. These findings, however, must be interpreted with caution due to the fact that different tests (some were administered the WRAT-R and some the WIAT in childhood) or different versions of the same test (WIAT vs. WIAT-II) were used in childhood and adolescence. Overall, there was a significant drop in standard scores over time [Reading: F (1,85) = 7.71, p = .007, η _p ² = .083; Math: F (1,85) = 8.16, p = .005, η _p ² = .088] for both groups. There were no significant main effects for Group or Group × Time interactions (all p > .05); however, the ADHD-U group tended to perform more poorly in math [F (1,85) = 3.34, p = .071, η _p ² = .038] and to have a greater diminution in reading scores over time [F (1,85) = 3.27, p = .074, η _p ² = .037].

Further, additional analyses examining the impact of comorbid ODD or CD on WIAT-II scores yielded no significant group differences or Group × Medication interactions. Finally, among the probands, reading achievement, as measured by the WIAT-II, was not related to adolescent ADHD symptom severity, as measured by an ADHD DSM-IV Checklist (Swanson, 1992) completed by both the proband and his/her parent (all p > .10). However, poorer math scores were associated with greater symptoms as rated by the proband (r = −.24, p < .05) and his/her parent (r = −.29, p < .01).

Grade point average

There was a significant difference among groups (one-way ANOVA) with regard to high school GPA (F (2, 74) = 6.43, p = 0.003, η _p ² = .15); a significant linear trend was found such that controls performed best and ADHD-U participants performed most poorly (F = 12.82, p < .001; see Fig. 2). Post-hoc analyses revealed that the control group achieved a significantly higher GPA than the ADHD-U group, but not ADHD-M participants (p = 0.002 and p = 0.236, respectively). Although the ADHD-M group achieved a higher GPA than the ADHD-U group, the difference did not reach statistical significance (p = 0.18). A FSIQ-controlled ANCOVA comparing groups with regard to GPA just fell short of statistical significance (F (2, 73) = 2.63, p = 0.08, η _p ² = .07).

FIG 2.

Adolescent grade point average. A significant linear trend is observed for grade point average such that controls performed the best and the ADHD-U group performed most poorly (F = 12.82, p < .01). GPA = grade point average; ADHD-U = attention-deficit/hyperactivity disorder-unmedicated; ADHD-M = attention-deficit/hyperactivity disorder-medicated.

Analyses comparing ADHD-U and ADHD-M participants, controlling for age at initial evaluation and IOWA Inattention/Overactivity ratings, revealed a significant difference between groups with regard to high school grade point average (GPA) (F (1,36) = 4.83, p = .035, η _p ² = .118), such that the medicated group received higher scores. Additional analyses examining the impact of comorbid ODD or CD on GPA yielded no significant group differences or Group × Medication interactions.

Grade repetitions

Mean ranks for number of grade repetitions revealed a significant difference among groups (Kruskal-Wallis = 14.98, p = 0.001). Follow-up Mann-Whitney U tests showed that the control group differed significantly in number of grade repetitions from both ADHD-U and ADHD-M participants (p < 0.001 and p = 0.004, respectively), who did not differ significantly from each other (p = 0.504). There was a significant difference (Chi-square) between the control group and both ADHD groups (p < .001), with regard to ever having repeated a grade; the two ADHD groups did not differ from one another (see Fig. 3).

FIG 3.

Ever repeated a grade. Irrespective of medication treatment, adolescents with childhood ADHD were more likely to have been retained in school than controls (p < 0.001). ADHD-U = attention-deficit/hyperactivity disorder-unmedicated; ADHD-M = attention-deficit/hyperactivity disorder-medicated.

Discussion

This longitudinal study examined the relationship of stimulant treatment of ADHD in childhood to academic outcome during late adolescence. Consistent with previous studies (Frick et al. 1991; Faraone et al. 1993; Barkley et al. 1990), adolescents with childhood ADHD in general had lower academic achievement during adolescence relative to a never-ADHD comparison group. However, academic outcome in those with ADHD appeared to be at least in part mediated by a history of treatment with stimulant medication. Several indices of academic achievement indicated that never-medicated or minimally-medicated participants with childhood ADHD, but not those who received consistent treatment for at least one year with stimulant medication, performed significantly worse than controls with no prior history of ADHD. Notably, this finding was found on standardized academic achievement measures (i.e., subtests of the WIAT-II), as well as classroom performance, as assessed by high school GPA. On each of these measures, those with childhood ADHD who received stimulant medication performed midway between the unmedicated ADHD and control groups. These treatment-related findings emerged even more clearly in secondary analyses comparing the two ADHD subgroups; after controlling for severity of childhood ADHD, those receiving at least one year of stimulant treatment had significantly higher scores on all three measures of academic achievement and a significantly higher high school GPA than those not receiving such treatment. These findings are consistent with studies demonstrating that stimulants improve short-term academic performance (Tannock et al. 1989; Evans and Pelham 1991; Wilens and Biederman 1992; MTA Cooperative Group 1999) and extend them to suggest possible, albeit modest, long-term benefits of stimulant treatment.

Although children treated with stimulant medication as compared to those without such treatment had improved academic performance, academic achievement was not normalized in the medicated ADHD youth. This may be due in part to the fact that a significant minority of children with ADHD also have comorbid learning disorders (Halperin et al. 1984; August and Garfinkel 1990), which are not corrected by stimulant medication (Gittelman et al. 1983). Further, while stimulant medication has been consistently shown to result in behavioral and cognitive improvement in children with ADHD, normalization in these domains is only achieved for a portion of successfully treated children (Swanson et al. 2001). Thus medication likely mitigates behavioral and cognitive difficulties, which in turn, impacts academic progress and performance. However, there is little reason to believe that there should be complete normalization in academic performance.

Nevertheless, it is notable that we have been able to detect clear, yet limited, gains in long-term academic outcome associated with stimulant treatment whereas other studies have failed to do so. This may be due to several features of this study as compared to others. First, we had in our sample a substantial number of adolescents who never received any treatment with medication (n = 27), and the mean duration of treatment for our treated group was substantially longer than most other studies. Importantly, all of the studies reviewed by Barkley and Cunningham (1978) and most by Schachar and Tannock (1993) had treatment durations of less than one year. As such, those participants would have been considered unmedicated in this study. With current knowledge about the chronic nature of ADHD, there is no reason to believe that treatment for a few weeks or months should yield lasting or long-term benefit with respect to outcome.

Another important difference between this study and others in the field is that our sample was not comprised primarily of middle class Caucasian youth. Rather, our sample contained an ethnically diverse sample of individuals with many being of low SES. As such, many of the children, irrespective of whether they did or did not receive stimulant treatment, likely had less access to resources commonly employed both within and outside of the home by middle and upper class families to promote academic success. Although we cannot be certain, and many other factors are at play (e.g., use of different tests), it is quite possible that this lower SES contributed to the fact that academic achievement standard scores went down over time irrespective of treatment. Unfortunately, we do not have childhood scores in our control group, who were of similar SES, to determine whether their relative performance also declined with age. It is notable however, that the decline in reading scores from childhood to adolescence was substantially less in the medicated group, although not quite reaching statistical significance.

To our knowledge, these data are the first to demonstrate across several measures a positive effect of extended stimulant medication treatment on long-term academic outcome in clinically-referred children with ADHD. Nevertheless, the magnitude of the effect is modest and these data are limited due to several factors. First, we were only able to re-evaluate 90 out of our 169 original participants (53%). This is likely due largely to the fact that the research was not originally designed as a longitudinal study. This limitation is somewhat mitigated by the fact that those participants who were seen for follow-up did not differ on baseline measures from those that were not seen for follow-up. A second limitation is the naturalistic design of this follow-up study and the impossibility of tight experimental control due to ethical constraints (i.e., prospective assessment following adhered to random assignment to medicated and unmedicated groups). Therefore factors other than medication may account for the findings. For example, it is possible that probands who received medication had greater parental involvement or access to health care, factors that may have contributed to their greater academic progress (Larzelere et al. 2004). However, this possibility is at least somewhat mitigated by the fact that the groups did not differ with regard to SES.

It is also quite possible that only those participants who had a positive clinical response to medication remained in treatment for longer than a year. As such, academic improvement may be more closely linked to clinical response rather than stimulant treatment per se. Unfortunately, limitations in our naturalistic follow-up design preclude us from systematically evaluating the relationship between symptom change and academic progress. Not only were different measures of symptom severity used during childhood and adolescence, but the critical time to have evaluated the clinical improvement would have been while the participants were still in high school; at the time of this study, many were not.

Although our acquisition of school records and the objective nature of the test measures provides confidence in our academic outcome assessment, history of medication treatment, as reported by the parent and adolescent, may be subject to recall inaccuracies associated with retrospective reports. This is most likely true for the dose and duration of medication treatment as well as the precise timing of medication onset and discontinuation. It was not uncommon, for instance, for parents and youth to have no recall about medication dosages. As a result, amount of medication, which might be a critical variable, could not be assessed in this study. With regard to duration, most families were able to provide what appeared to be a reasonable estimate of duration (e.g., “He took the medicine from 2^nd to 4^th grade,” or “He tried it for a week or so and we didn't like it and stopped”), but undoubtedly there was some error in these retrospective estimates. On that basis, we did not conduct primary hypothesis-testing analyses of duration of treatment. Rather, we attempted to focus on predictor variables for which we had the most confidence, such as whether or not the child ever took medication, and if so, whether or not the medicine was taken regularly for an extended period of time. Although we operationally defined consistent treatment as being for at least one year, the mean duration of medication treatment was considerably longer than that (on average several years duration). Nevertheless, we secondarily examined correlations between duration of medication treatment and academic outcomes among medicated (ADHD-M) participants, and found that no significant correlations emerged.

While it is possible that other psychosocial or educational interventions impacted upon outcome, these were often vaguely described by parents (e.g., “He received something in school” or “He received on and off tutoring from a classmate”), and thus difficult to quantify. Thus, we could not identify methods for integrating non-pharmacological interventions into our analyses. Limited statistical power precluded our ability to conduct analyses among medicated participants to examine whether treatment during some ages had a greater impact on academic achievement than at other ages; also, the restricted number of female participants precluded analyses of gender differences. Finally, this study is limited by the fact that the never-ADHD controls were recruited as adolescents, and as such, could not be incorporated into any longitudinal analyses.

This study provides a unique perspective in that participants were not recruited for a specific treatment study during childhood, and for this reason, a substantial proportion never received stimulant medication. Therefore the results of this study may represent a more realistic comparison among treatment groups as treatment is provided in the community, as opposed to more rigorous and well-controlled treatments typically provided by research studies. Further, the sample of this study is distinctive in that a lower SES and a substantial number of racial and ethnic minority groups are represented among medicated and unmedicated ADHD probands, as well as control participants.

Due to ethical constraints limiting investigators from directly manipulating the impact of medication on long-term outcome, or the use of placebo conditions, patterns of results emerging from multiple less-than-perfect naturalistic studies such as this one will need to be examined. Newer studies that follow children more frequently (i.e. annually or biannually) are likely to obtain higher-quality data regarding medication status with more precise quantification regarding dose, duration, and age of treatment, as well as a lower attrition rate of participants.

Conclusions

Overall, we found that children with ADHD who were treated with stimulant medication showed academic gains on several measures relative to children with ADHD who did not receive such treatment. Nevertheless, academic achievement was not normalized. Despite several limitations, many of which are inherent to our naturalistic follow-up design, this study represents a first step toward understanding the possible long-term benefits of stimulant medication for the improvement of academic performance in children with ADHD.

Footnotes

Acknowledgments

The authors acknowledge Tobey Busch and Dana Barowsky for their help and support throughout this project.

Disclosures

Dr. Newcorn is a recipient of grants for research support from Eli Lilly, McNeil, Novartis and Shire; an advisor/consultant for Eli Lilly, Novartis, McNeil, Shire, Cephalon, Cortex, Pfizer, Lupin; and a Speaker for Eli Lilly, McNeil, Novartis, and Shire. Drs. Powers, Marks, Miller, and Halperin have no financial ties or conflicts of interest to disclose.

This research was funded by grant # R01 MH060698 from the National Institute of Mental Health.

References

Achenbach

. Manual for the Child Behavior Checklist 4/18, 1991 Profile. Burlington VT: University of Vermont Department of Psychiatry, 1991.

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 3rdRevisedWashington, DC: American Psychiatric Association, 1987.

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th. Washington, DC: American Psychiatric Association, 1994.

August

, Garfinkel

. Comorbidity of ADHD and reading disability among clinic-referred children. J Abnorm Child Psychol, 18:29–45. 1990.

August

, Holmes

. Behavior and academic achievement in hyperactive subgroups and learning-disabled boys. AJDC, 138:1025–1029. 1984.

Barbaresi

, Katusic

, Colligan

, Weaver

, Jacobsen

. Modifiers of long term school outcomes for children with Attention-deficit/Hyperactivity Disorder: Does treatment with stimulant medication make a difference? Results from a population-based study. J Develop Behav Pediatr, 28:274–287. 2007.

Barkley

. Behavioral inhibition, sustained attention, and executive functions: Constructing a unifying theory of ADHD. Psychological Bull, 121:65–94. 1997.

Barkley

, Cunningham

. Do stimulant drugs improve the academic performance of hyperkinetic children? Clin Pediatr, 17:85–92. 1978.

Barkley

, Fischer

, Edelbrock

, Smallish

. The adolescent outcome of hyperactive children diagnosed by research criteria: I. An 8-year prospective follow-up study. J Am Acad Child Adolesc Psychiatry, 29:546–557. 1990.

10.

Barkley

, Fischer

, Smallish

, Fletcher

. The persistence of attention-deficit/hyperactivity disorder into young adulthood as a function of reporting source and definition of disorder. J Abnorm Psychol, 111:279–289. 2002.

11.

Biederman

, Faraone

. Attention-deficit hyperactivity disorder. Lancet, 366:237–248. 2005.

12.

Chacko

, Pelham

, Gnagy

, Greiner

, Vallano

, Bukstein

, Rancurello

. Stimulant medication effects in a summer treatment program among young children with attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry, 44:249–257. 2005.

13.

Dykman

, Ackerman

. Attention deficit disorder and specific reading disability: Separate but often overlapping disorders. J Learning Disabilities, 24:96–103. 1991.

14.

Evans

, Pelham

. Psychostimulant effects on academic and behavioral measures for ADHD junior high school students in a lecture format classroom. J Abnorm Child Psychol, 19:537–552. 1991.

15.

Evans

, Pelham

, Smith

, Bukstein

, Gnagy

, Greiner

, Altenderfer

, Baron-Myak

. Dose-response effects of methylphenidate on ecologically valid measures of academic performance and classroom behavior in adolescents with ADHD. Exper Clin Psychopharmacol, 9:163–175. 2001.

16.

Faraone

, Biederman

, Lehman

, Spencer

, Norman

, Seidman

, Kraus

, Perrin

, Chen

, Tsuang

. Intellectual performance and school failure in children with attention deficit hyperactivity disorder and in their siblings. J Abnorm Psychol, 102:616–623. 1993.

17.

Fischer

, Barkley

, Edelbrock

, Smallish

. The Adolescent outcome of hyperactive children diagnosed by research criteria: II. Academic, attentional, and neuropsychological status. J.Consult Clin Psychol, 58:580–588. 1990.

18.

Frick

, Kamphaus

, Lahey

, Loeber

, Christ

MAG

, Hart

, Tannenbaum

. Academic underachievement and the disruptive behavior disorders. Journal Consult Clin Psychol, 59:289–294. 1991.

19.

Gittelman

, Klein

, Feingold

. Children with reading disorders–-II. Effects of methylphenidate in combination with reading disabilities. J Child Psychol Psychiatry, 24:193–212. 1983.

20.

Greenhill

, Findling

, Swanson

. A Double-blind, placebo-controlled study of modified-release methylphenidate in children with attention deficit/hyperactivity disorder. Pediatrics, 109:39–46. 2002.

21.

Halperin

, Gittleman

, Klein

, Rudel

. Reading-disabled hyperactive children: A distinct subgroup of attention deficit disorder with hyperactivity? J Abnorm Child Psychol, 12:1–14. 1984.

22.

Hechtman

, Abikoff

, Klein

, Weiss

, Respitz

, Kouri

, Blum

, Greenfield

, Etcovitch

, Fleiss

, Pollack

. Academic achievement and emotional status of children with ADHD treated with long-term methylphenidate and multimodal psychosocial treatment. J Am Acad Child Adolesc Psychiatry, 43:812–819. 2004.

23.

Hechtman

, Weiss

. Long-term outcome of hyperactive children. Am J Orthopsychiatry, 53:532–541. 1983.

24.

Hinshaw

. Externalizing behavior problems and academic underachievement in childhood and adolescence: Causal relationships and underlying mechanisms. Psychological Bull, 111:127–155. 1992.

25.

Jastak

, Wilkinson

. The Wide Range Achievement Test–Revised (WRAT-R) Wilmington, DE: Jastak Associates, 1984.

26.

Lambert

. Adolescent outcomes for hyperactive children. Am Psychologist, 43:786–799. 1988.

27.

Larzelere

, Kuhn

, Johnson

. The Intervention selection bias: An under recognized confound in intervention research. Psychological Bull, 130:289–303. 2004.

28.

Loney

, Kramer

, Milich

. The hyperactive child grows up: Predictors of symptoms, delinquency, achievement at follow-up. Psychosocial Aspects of Drug Treatment for Hyperactivity. KD

Gadow

, J

Loney

. Boulder, CO: Westview Press, 1981; 381–415.

29.

Loney

, Milich

. Hyperactivity, inattention, aggression in clinical practice. Advances in Developmental and Behavioral Pediatrics. M

Wolraich

, D

Routh

. Greenwich, CT: JAE, 1982; 113–147.

30.

Lyon

. Learning disabilities. The Future of Children, 6:54–76. 1996.

31.

Mannuzza

, Klein

, Bessler

, Malloy

, Hynes

. Educational and occupational outcome of hyperactive boys grown up. J Am Acad Child Adolesc Psychiatry, 36:1222–1227. 1997.

32.

Mannuzza

, Klein

, Bessler

, Malloy

, LaPadula

. Adult outcome of hyperactive boys. Educational achievement, occupational rank, and psychiatric status. Arch Gen Psychiatry, 50:565–576. 1993.

33.

MTA Cooperative Group. A 14-month randomized clinical trial of treatment strategies for attention-deficit/hyperactivity disorder. Arch Gen Psychiatry, 56:1073–1086. 1999.

34.

MTA Cooperative Group. National Institute of Mental Health Multimodal Treatment Study of ADHD follow-up: 24-Month outcomes of treatment strategies for attention-deficit/hyper-activity disorder. Pediatrics, 113:754–761. 2004.

35.

MTA Cooperative Group. 3-year follow-up of the NIMH MTA Study. J Am Acad Child Adolesct Psychiatry, 46:988–1001. 2007.

36.

Murphy

, Barkley

. Attention deficit hyperactivity disorder adults: Comorbidities and adaptive impairments. Comp Psychiatry, 37:393–401. 1996.

37.

Nakao

, Treas

. Updating occupational prestige and socioeconomic scores: How the new measures measure up. Sociological Methodology, 24:1–72. 1994.

38.

Paternite

, Loney

, Salisbury

, Whaley

. Childhood inattention overactivity, aggression, and stimulant medication history as predictors of young adult outcomes. J Child Adolesc Psychopharmacology, 9:169–184. 1999.

39.

Pelham

, Carlson

, Sams

, Vallano

, Dixon

, Hoza

. Separate and combined effects of methylphenidate and behavior modification on boys with attention deficit-hyperactivity disorder in the classroom. Journal of Consulting and Clinical Psychology, 61:506–515. 1993.

40.

Rasmussen

, Gillberg

. Natural outcome of ADHD with developmental coordination disorder at age 22 years: A controlled longitudinal, community-based study. Journal of the American Academy of Child & Adolescent Psychiatry, 39:1424–1431. 2000.

41.

Satterfield

, Satterfield

, Cantwell

. Three-year multi-modal treatment study of 100 hyperactive boys. Behavioral Pediatrics, 98:650–655. 1981.

42.

Schachar

, Tannock

. Childhood hyperactivity and psychostimulants: A review of extended treatment studies. Journal of Child and Adolescent Psychopharmacol, 3:81–97. 1993.

43.

Shaffer

, Fisher

, Dulcan

, Davies

, Piacentini

, Schwab-Stone

, Lahey

, Lichtman

, Canino

, Rubio-Stipec

, Rae

. The NIMH Diagnostic Interview Schedule for Children version 2.3 (DISC-2.3): Description, acceptability, prevalence rates, and performance in the MECA study. Methods for the epidemiology of child and adolescent mental disorders study. J Am Acad Child Adolesc Psychiatry, 35:865–877. 1996.

44.

Shaffer

, Fisher

, Piacentini

, Schwab-Stone

, Wicks

. Diagnostic Interview Schedule for Children (DISC-2.1P)-Parent Version. New York: New York State Psychiatric Institute, 1989.

45.

Silver

. The Relationship between learning disabilities, hyperactivity, distractibility and behavioral problems: A clinical analysis. J Am Acad Child Psychiatry, 20:385–397. 1981.

46.

Swanson

. School Based Assessments for ADD Students. Irvine, CA: K.C. Publishing, 1992.

47.

Swanson

, Kraemer

, Hinshaw

, Arnold

, Conners

, Abikoff

, Clevenger

, Davies

, Elliot

, Greenhill

, Hechtman

, Hoza

, Jensen

, March

, Newcorn

, Owens

, Pelham

, Schiller

, Severe

, Simpson

, Vitiello

, Wells

, Wigal

, Wu

. Clinical relevance of the primary findings of the MTA: Success rates based on severity of ADHD and ODD symptoms at the end of treatment. J Am Acad Child Adolesc Psychiatry, 40:168–179. 2001.

48.

Swanson

, Wigal

, Greenhill

, Browne

, Waslick

, Lerner

, Williams

, Flynn

, Agler

, Crowley

, Fineberg

, Regino

, Baren

, Cantwell

. Objective and subjective measures of the pharmacodynamic effect of Adderall in the treatment of children with ADHD in a controlled laboratory classroom setting. Psychopharmacol Bulletin, 34:55–60. 1998.

49.

Swanson

, Wigal

, Sonuga-Barge

, Greenhill

, Biederman

, Kollins

, Nguyen

, DeCory

, Hirshe Dirksen

, Hatch

S J

. COMACS Study Group: A comparison of once-daily extended-release methylphenidate formulations in children with attention-deficit/hyperactivity disorder in the laboratory school (the Comacs Study) Pediatrics, 113:206–216. 2004.

50.

Tannock

, Schachar

, Carr

, Logan

. Dose-response effects of methylphenidate on academic performance and overt behavior in hyperactive children. Pediatrics, 84:648–657. 1989.

51.

Torgersen

, Gjervan

, Rasmussen

. ADHD in adults: A Study of clinical characteristics, impairment and comorbidity. Nordic J Psychiatry, 60:38–43. 2006.

52.

Vaidya

, Austin

, Kirkorian

, Ridlehuber

, Desmond

, Glover

, Gabrieli

. Selective effects of methylphenidate in attention deficit hyperactivity disorder: A functional magnetic resonance study. Proc Nat Acad Sci, 95:14494–14499. 1998.

53.

Wechsler

. Wechsler Intelligence Scale for Children -Revised (WISC-R) San Antonio, TX: The Psychological Corporation, 1974.

54.

Wechsler

. Wechsler Intelligence Scale for Children-Third Edition (WISC-III) San Antonio, TX: The Psychological Corporation, 1991.

55.

Wechsler

. Wechsler Adult Intelligence Scale - Third Edition (WAIS-III) San Antonio, TX: The Psychological Corporation, 1997.

56.

Wechsler

. Wechsler Individual Achievement Test (WIAT) San Antonio, TX: The Psychological Corporation, 1992.

57.

Wechsler

. Wechsler Individual Achievement Test - Second Edition (WIAT-II) San Antonio, TX: The Psychological Corporation, 2001.

58.

Weiss

, Hechtman

, Milroy

, Perlman

. Psychiatric status of hyperactives as adults: A controlled prospective 15-year follow-up of 63 hyperactive children. J Am Acad Child Adolesc Psychiatry, 24:211–220. 1985.

59.

Wilens

, Biederman

. The Stimulants. Psychiatric Clin North America, 15:191–222. 1992.

60.

, Anderson

, Castiello

. Neuropsychological evaluation of deficits in executive functioning for ADHD children with or without learning disabilities. Develop Neuropsychol, 22:501–531. 2002.