SPECIAL SECTION

Symptomatic Improvement in Children With ADHD

Treated With Long-Term Methylphenidate and
Multimodal Psychosocial Treatment
HOWARD ABIKOFF, PH.D., LILY HECHTMAN, M.D., RACHEL G. KLEIN, PH.D., GABRIELLE WEISS, M.D.,
KAREN FLEISS, PSY.D., JOY ETCOVITCH, M.A., LORNE COUSINS, PH.D., BRIAN GREENFIELD, M.D.,
DIANE MARTIN, M.A., AND SIMCHA POLLACK, PH.D.

ABSTRACT
Objective: To test the hypotheses that in children with attention-deficit/hyperactivity disorder (ADHD) (1) symptoms of
ADHD, oppositional defiant disorder, and overall functioning are significantly improved by methylphenidate combined
with intensive multimodal psychosocial treatment compared with methylphenidate alone and with methylphenidate plus
attention control and (2) more children receiving combined treatment can be taken off methylphenidate. Method: One
hundred three children with ADHD (ages 79), free of conduct and learning disorders, who responded to short-term
methylphenidate were randomized for 2 years to (1) methylphenidate alone; (2) methylphenidate plus psychosocial
treatment that included parent training and counseling, social skills training, psychotherapy, and academic assistance,
or (3) methylphenidate plus attention psychosocial control treatment. Assessments included parent, teacher, and psy-
chiatrist ratings, and observations in academic and gym classes. Results: Combination treatment did not lead to
superior functioning and did not facilitate methylphenidate discontinuation. Significant improvement occurred across all
treatments and continued over 2 years. Conclusions: In stimulant-responsive children with ADHD, there is no support
for adding ambitious long-term psychosocial intervention to improve ADHD and oppositional defiant disorder symptoms.
Significant benefits from methylphenidate were stable over 2 years. J. Am. Acad. Child Adolesc. Psychiatry,
2004;43(7):802811. Key Words: attention-deficit/hyperactivity disorder, oppositional defiant disorder, school obser-
vations, long-term stimulant treatment, psychosocial treatment.

Accepted January 30, 2004.
Drs. Abikoff, Klein, and Fleiss are with the NYU Child Study Center, New
York University School of Medicine, New York; Drs. Hechtman and Greenfield
are with the Department of Psychiatry, McGill University and Montreal
Childrens Hospital, Montreal, Quebec, Canada; Ms. Etcovitch is with
Montreal Childrens Hospital, Dr. Weiss is with the University of British
Columbia and British Columbia Childrens and Womens Hospital, Vancouver,
British Columbia,Canada; Dr. Cousins is with McGill University and the
Summit School, Montreal, Quebec, Canada, Ms. Martin is with Nassau Com-
munity College, Garden City, NY; and Dr. Pollack is with the Department of
Computer Information Systems and Decision Science, St. Johns University,
Queens, NY.
The study was supported NIMH grants RO1 MH44848 (H.A.) and RO1
MH44842 (L.H.).
Correspondence to Dr. Abikoff, NYU Child Study Center, 215 Lexington
Avenue, 13th Floor, New York, NY 10016; e-mail: abikoh01@med.nyu.edu.
0890-8567/04/430708022004 by the American Academy of Child
and Adolescent Psychiatry.
DOI: 10.1097/01.chi.0000128791.10014.ac
The merits of stimulant medication in the treatment of
attention-deficit/hyperactivity disorder (ADHD) have
long been established. In addition to improving cardi-
nal symptoms, short-term stimulants also enhance aca-
demic productivity and accuracy (Carlson et al., 1992;
Douglas et al., 1986) and teacher, parent, and peer
interactions (Granger et al., 1996) as well as improving
antisocial behavior (Hinshaw, 1991; Hinshaw et al.,
1992; Klein et al., 1997). However, improvement is
not maintained when medication is discontinued, and
the lack of long-term efficacy has been a concern (Abi-
koff and Gittelman, 1985a; Gittelman-Klein et al.,
1976). Although there may be long-term benefits of
early stimulant treatment (Paternite et al., 1999), dif-
ficulties often continue through adolescence and early

802 J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 43:7, JULY 2004


adulthood (Barkley, 1990; Gittelman et al., 1985;
Mannuzza et al., 1993; Weiss and Hechtman, 1993).
Symptomatic persistence adversely affects multiple
functions including academic and occupational attain-
ment and interpersonal relationships. Notably, the per-
sistence of ADHD is a significant risk factor for later
antisocial and substance use disorders (Gittelman et al.,
1985; Mannuzza et al., 1993).
In the hope of reversing these patterns, psychosocial
treatments have been developed to have an effect on
dysfunctions not regularly normalized by stimulants
(Hinshaw et al., 2002). The current study provides
systematic information of the incremental benefit of
combining stimulant and psychosocial treatments for
an extended period of time. It evaluates the adjunctive
efficacy of intensive multimodal psychosocial treatment
(MPT) in children with ADHD treated with methyl-
phenidate. The study tests whether 1 year of combined
methylphenidate and MPT confers significantly better
function in social, behavioral, emotional, and academic
domains and better parental functioning compared
with treatment with methylphenidate alone in children
with ADHD and whether gains are maintained over a
second year.
This paper reports treatment effects on ADHD and
oppositional defiant disorder (ODD) symptoms and
childrens overall functioning. Other treatment out-
comes, i.e., childrens social and academic perfor-
mance, are communicated in Abikoff et al. (2004) and
Hechtman et al. (2004a,b).
The following hypotheses are addressed: Over a
1-year period (year I), (1) there is significant advantage
to adding a multimodal psychosocial intervention to
methylphenidate treatment and (2) after year I, stimu-
lant treatment can be withdrawn more successfully in
the combination treatment than methylphenidate
alone group. Additionally, we predicted that the supe-
riority of the combination over methylphenidate alone
would result from the specific content of the psycho-
social treatment and not from its nonspecific features.
Therefore, it was hypothesized that the combination of
methylphenidate and MPT would be superior to meth-
ylphenidate plus attention control psychosocial treat-
ment (ACT).
Finally, it was hypothesized that relative advantages
associated with 1 year of combined treatment would
persist. Hence, we predicted superiority of combined
treatment during a second year (year II) of mainte-
J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 43:7, JULY 2004
TREATMENT EFFECTS ON ADHD SYMPTOMS
nance treatment. We hypothesized that treatment
groups would demonstrate different patterns of func-
tion over time. Specifically, significant incremental im-
provement during year II was predicted with combined
treatment, relative to methylphenidate alone, and
methylphenidate plus an attention control. In the latter
two groups, a flattening of treatment effects was pre-
dicted.
METHOD
Details of the design and its rationale are presented in Klein et al.
(2004). Briefly, the study was conducted at two large medical cen-
ters (New York and Montreal) between 1990 and 1995. Medica-
tion-free boys and girls, 7.0 to 9.9 years of age (mean 8.2 0.8),
93% male, mostly white, without a diagnosis of conduct or learning
disorder, met diagnostic and severity criteria for ADHD (n = 103).
Because treatment included 2 years of methylphenidate, children
had to exhibit meaningful improvement in a 5-week open trial of
methylphenidate.
Treatments
Children were randomly assigned for 2 years to (1) methylphe-
nidate alone (M) (n = 34), (2) methylphenidate plus MPT (M +
MPT) (n = 34), or (3) methylphenidate plus ACT (M + ACT) (n =
35).
Measures
Parent Ratings. Parents completed the Conners Parent Rating
Scale (Goyette et al., 1978), whose Hyperkinesis Index was an
outcome measure; and the Home Situations Questionnaire (Bark-
ley, 1990). It yields the number of problematic situations and their
severity.
Teacher Ratings. Teachers completed the Hyperactivity and Con-
duct Problem Factors of the Conners Teacher Rating Scale (Goy-
ette et al., 1978) and the School Situations Questionnaire (Barkley,
1990). Like the Home Situations Questionnaire, the School Situ-
ations Questionnaire yields the number of problematic situations
and a severity score. Both scales have adequate norms, reliability,
and validity and demonstrate sensitivity to treatment effects (Bar-
kley, 1990).
Psychiatric Ratings. Based on clinical interviews, child psychia-
trists completed a DSM-III R checklist for ADHD, ODD, and
conduct disorder (CD) symptoms, and a Childrens Global Assess-
ment Scale (C-GAS), a measure of functional competence with
good interrater and testretest reliability (Shaffer et al., 1985).
School Observations. Because the treatment blind could not be
protected, objective assessments of classroom behavior provided
independent assessment of treatment effects. Children were ob-
served twice at each assessment in academic and gym classes. The
Classroom Observation Code (Abikoff et al., 1980), which differ-
entiates children with ADHD from normal children and is sensitive
to treatment effects (Abikoff and Gittelman, 1985b), was used
during academic classes. Categories of interference, off task (mea-
sures of impulsivity and sustained inattention), and gross motorall
were combined into an ADHD composite used previously as an
outcome measure (Klein and Abikoff, 1997).
803
ABIKOFF ET AL.

After gym observations, observers completed the CTRS Hyper- TABLE 1

kinesis Index and IOWA CTRS (Loney and Milich, 1982), both Conners Parent Rating Scale and Home Situations Questionnaire
with demonstrated sensitivity to treatment effects (Abikoff and Git-
Treatment Group
telman, 1984) (behaviors coded in gym are presented elsewhere
[Abikoff et al., 2004]). M M + MPT M + ACT
Trained observers rated two children: the study child and an
average classmate of unremarkable comportment identified by Measure Mean SD Mean SD Mean SD
teachers. Whenever study children changed classes, another peer CPRS
was observed. (Because comparison children were anonymous and
Hyperkinesis Indexa,c
were unaware of being observed, parental consent was not required
by the institutional review board or schools.) Pretreatment 1.9 0.5 1.9 0.5 1.9 0.5
Children were evaluated twice before experimental treatment: 6 mo 1.2 0.5 1.2 0.6 1.1 0.5
once at pretreatment and at the end of the 5-week open methyl- 12 mo 1.1 0.6 1.2 0.6 1.0 0.6
phenidate trial (medication baseline, when only classroom observa- 18 mo 0.9 0.6 1.0 0.5 1.0 0.6
tions were repeated). Assessments were obtained after 6, 12, 18, and 24 mo 1.0 0.6 0.9 0.5 0.8 0.4
24 study months to identify the timing of hypothesized treatment HSQ
differences. Situations (N )a,b,d
Pretreatment 12.9 3.5 13.1 3.3 12.6 3.4
Data Analyses 6 mo 11.9 3.6 12.1 2.6 11.0 3.3
There were no significant group site or group site time 12 mo 9.9 4.4 11.3 3.7 11.1 4.5
interactions. Repeated measures over time for dependent variables 18 mo 8.8 4.7 9.9 4.0 10.8 4.5
were modeled as a mixed-model analysis of covariance implemented 24 mo 9.4 4.4 9.9 4.1 9.1 4.7
in Proc Mixed (SAS v8.1, Cary, NC), controlling for socioeco- Severitya,e
nomic status. Empirical data exploration indicated that an unstruc- Pretreatment 3.6 1.6 3.8 1.9 3.7 1.7
tured covariance model best fit the data. Model parameter estimates 6 mo 2.7 1.5 2.4 1.1 2.4 1.3
and their standard errors were generated through maximum likeli- 12 mo 2.3 1.7 2.4 1.4 2.4 1.5
hood functions.
18 mo 1.8 1.5 2.1 1.4 2.4 1.7
Differential treatment effects in year I compared status at pre-
treatment and at medication baseline with status at 6 and 12 24 mo 1.7 1.5 1.9 1.3 1.6 1.2
months. For hypothesized differential maintenance effects, Proc Note: CPRS = Conners Parent Rating Scale (range 03); M =
Mixed analyses compared the 12-, 18-, and 24-month data for
methylphenidate; MPT = multimodal psychosocial treatment;
differential patterns of change. The above tests yield main effects for
group and time and group time interaction effects. The latter are ACT = attention control psychosocial treatment; HSQ = Home
the main interest of the study. Situations Questionnaire (situations, range 116; severity, range
To control for multiple tests, was set at p < .01, two tailed; 09). Group time interactions: none significant.
a
p values between .05 and .01 are noted as trends in the tables. Full Time effects in year I: pretreatment versus 6 and 12 months,
tables with F values are available from the authors. p < .001.
b
Time effects in year I: 6 versus 12 months, p < .01.
c
Time effects in year II: 12 versus 18 months, p < .03; 12 versus
RESULTS 24 months, p < .01.
d
Time effects in year II: 12 versus 18 months, p < .04.
Children were in the clinical range on all measures e
Time effects in year II: 12 versus 24 months, p < .02.
obtained from parents, teachers, classroom observers,
and clinicians (Tables 16). For example, on the
teacher hyperactivity factor, children obtained a mean Parent and Teacher Ratings. No significant group
score of 2.4 (range 03). In school, children were rated differences or group time interactions were obtained
as problematic in nine of 12 situations and on average for any parent or teacher ratings (Tables 1 and 2).
had 11 of 13 ADHD symptoms. The C-GAS mean Psychiatric Ratings. No differential treatment effect
(<55) indicates impaired overall function. Observed was obtained for ADHD or ODD symptoms at home
rates of disruptive and inattentive behavior were mark- or school (Table 3) or for rates of diagnosed ADHD,
edly elevated. ODD, and CD (Table 4). Overall, only 12.1% con-
tinued to meet DSM-III-R criteria for ADHD by the
Year I Treatment Effects From Pretreatment end of year I.
These analyses test the hypothesis that, over year I, Before treatment, 53% of the children received a
M + MPT led to superior function compared with M diagnosis of ODD. As shown in Table 4, the rate of
alone and M + ACT, relative to function before treat- ODD decreased significantly in all groups without dif-
ment. ferential treatment effects.

804 J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 43:7, JULY 2004

 TREATMENT EFFECTS ON ADHD SYMPTOMS

TABLE 2 TABLE 3
Conners Teacher Rating Scale and School Mean Number of ADHD and ODD Symptoms
Situations Questionnaire Treatment Group
Treatment Group
M M + MPT M + ACT
M M + MPT M + ACT
Symptoms Mean SD Mean SD Mean SD
Measure Mean SD Mean SD Mean SD
ADHD
CTRS factor Schoola,b
Hyperactivitya Pretreatment 11.7 1.8 11.8 1.6 11.6 2.0
Pretreatment 2.4 0.6 2.5 0.5 2.3 0.6 6 mo 5.5 3.4 5.1 3.8 4.6 3.2
6 mo 0.9 0.6 1.1 0.7 1.0 0.8 12 mo 4.1 3.3 3.7 2.9 3.3 3.0
12 mo 1.2 0.9 0.9 0.8 0.9 0.7 18 mo 4.9 3.4 3.5 3.0 4.3 3.8
18 mo 0.9 0.5 0.9 0.7 0.9 0.8 24 mo 4.2 3.5 3.6 2.2 2.9 3.2
24 mo 1.1 0.8 1.0 0.7 0.7 0.4 Homea,b
Conduct problemsb Pretreatment 11.2 2.0 11.7 1.7 11.3 2.2
Pretreatment 1.2 0.6 1.2 0.6 1.1 0.6 6 mo 5.3 3.3 5.0 3.7 4.5 3.3
6 mo 0.9 0.5 0.9 0.5 0.7 0.4 12 mo 3.9 3.2 3.6 2.9 3.2 3.0
12 mo 0.9 0.7 0.9 0.6 0.7 0.5 18 mo 4.8 3.4 3.4 2.9 4.0 3.6
18 mo 0.7 0.6 0.7 0.5 0.7 0.6 24 mo 4.0 3.4 3.5 2.1 2.7 3.1
24 mo 0.7 0.6 0.8 0.5 0.6 0.4 ODD
SSQa Schoola,b,c
Situations Pretreatment 4.5 2.2 4.3 2.6 4.7 2.5
Pretreatment 9.2 2.9 9.5 2.7 10.1 1.7 6 mo 2.9 2.1 2.5 2.4 2.4 2.1
6 mo 5.1 4.0 5.7 4.1 6.4 4.0 12 mo 2.6 2.5 2.2 2.0 1.3 1.7
12 mo 4.6 4.0 6.1 4.5 5.5 4.3 18 mo 3.0 2.5 1.6 1.8 2.1 2.3
18 mo 5.0 4.2 5.8 3.4 5.2 4.6 24 mo 2.4 2.6 1.9 2.3 1.9 2.5
24 mo 6.2 3.8 5.1 4.2 5.3 3.8 Homea,b,c
Severitya Pretreatment 4.4 2.2 4.2 2.5 4.4 2.4
Pretreatment 5.5 2.1 5.5 2.2 5.7 1.9 6 mo 2.9 2.1 2.5 2.4 2.4 2.1
6 mo 1.9 1.9 1.8 1.9 2.3 2.2 12 mo 2.5 2.5 2.0 1.9 1.3 1.7
12 mo 1.7 2.1 2.2 2.1 1.7 1.7 18 mo 3.0 2.5 1.6 1.8 2.0 2.2
18 mo 1.7 1.8 1.8 1.2 2.5 2.6 24 mo 2.4 2.6 1.7 2.2 1.8 2.3
24 mo 2.0 1.7 2.2 2.0 1.6 1.7
Note: ADHD = attention-deficit/hyperactivity disorder; ODD,
Note: CTRS = Conners Teacher Rating Scale (range 03); SSQ = oppositional defiant disorder. Group time interactions: none sig-
School Situations Questionnaire (situations, range 112; severity, nificant.
a
range 09). Group time interactions: none significant. Pretreatment versus 6 and 12 months, p < .001.
a b
Pretreatment versus 6 and 12 months, p < .001. 6 versus 12 months, p < .01.001.
b c
Pretreatment versus 6 and 12 months, p < .000. Trend in group time effect in year II: M + MPT versus M at
18 months, p < .01.

Clinical assessments of overall functioning on the
C-GAS also failed to differ across treatments
(Table 4).
Classroom Observations. Interobserver agreement was
conducted in approximately 15% of observations. Each
sites trainer functioned as the standard. Phi coeffi-
cients of interval scores for categories ranged from 0.83
to 0.92.
Impulsive, inattentive, and hyperactive behaviors
(interference, off task, and gross motor) showed no
significant advantage of M + MPT. The composite
measure of ADHD also failed to reveal significant
treatment differences (Table 5).
Table 6 presents observer ratings on the IOWA
CTRS for subjects and normal peers. Inattention and
oppositional behavior in structured classes and gym
showed significant improvement in all groups (p = .01)
without advantage for combination treatment.
Year I Treatment Effects From Medication Baseline.
Classroom observations, which were the only measures
of ADHD repeated after the 5-week clinical methyl-
phenidate trial, showed no differential treatment effect
compared with functioning after brief methylphenidate
treatment. As shown in Tables 5 and 6, scales com-
pleted by observers showed no relative advantage for
children on the combination treatment.

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 43:7, JULY 2004 805

ABIKOFF ET AL.

TABLE 4 trend in favor of M + MPT over M treatment was

Psychiatric Ratings of ADHD, ODD, CD, and Overall Function found (p < .04). No other difference occurred.
Treatment Group
M M + MPT M + ACT Year II Treatment Effects
a a
N % N % Na % Year I findings fail to support the superiority of the
ADHDb combination of M + MPT over M alone. In light of
Pretreatment 34/34 100 34/34 100 35/35 100 these results, there is little relevance to testing the hy-
6 mo 9/32 28.1 6/31 19.4 5/31 16.1 pothesis that year I advantages persist during a second
12 mo 5/30 16.7 4/30 13.3 2/31 6.5
18 mo 6/25 24.0 3/29 10.3 4/38 14.3
year of maintenance treatment. Nonetheless, year II
24 mo 5/25 20.0 1/29 3.4 3/27 11.1 outcomes inform on the hypothesis that M + MPT is
ODDb significantly superior to other treatments during main-
Pretreatment 16/34 47.1 18/30 58.1 19/35 54.3 tenance treatment.
6 mo 8/32 25.0 5/31 16.1 5/31 16.1 Parent and Teacher Ratings. None of the parent or
12 mo 7/30 23.3 4/30 13.3 2/31 6.5
teacher questionnaires yielded significant group or in-
18 mo 6/24 20.8 3/27 11.1 3/28 10.7
24 mo 5/26 17.2 4/29 13.8 4/27 14.8 teraction effects, failing to support a differential clinical
CDb trajectory across the three treatments during year II
Pretreatment 0/34 0 0 0/35 0 (Tables 1 and 2).
6 mo 0/32 0 2/31 6.5 1/31 3.2 Psychiatric Ratings. Psychiatrists ratings of ADHD
12 mo 0/30 0 0/30 0 1/31 3.2 symptoms in school and at home did not differentiate
18 mo 1/24 4.2 0/27 0 0/22 0
24 mo 1/24 4.0 1/29 3.4 0/27 0
the treatment groups (Table 3). With regard to ODD
symptoms, significantly fewer occurred at 18 months
C-GASb Mean SD Mean SD Mean SD in the M + MPT group than the M alone group (p <
Pretreatment 53.0 7.1 54.9 6.6 55.1 6.8 .01) (Table 3) but not at other time points. Rates of
6 mo 67.1 7.2 68.5 8.5 66.9 7.7 diagnosed ADHD, ODD, and CD did not differ
12 mo 69.4 9.3 69.6 9.7 68.6 7.9 across treatments (overall, 11.1%, 15.8%, and 2.5%,
18 mo 70.6 9.8 69.6 8.5 66.3 7.2
24 mo 72.4 11.6 69.3 11.1 72.3 10.5
respectively). Similarly, treatment groups did not differ
significantly in global functioning during year II.
Note: ADHD = attention-deficit/hyperactivity disorder; ODD = Classroom Observations. Classroom behaviors during
oppositional defiant disorder; CD = conduct disorder; C-GAS =
year II yielded no significant group or interaction ef-
Childrens Global Assessment Scale.
a
Number diagnosed/total number in group. Missing data and fects (Tables 5 and 6).
dropouts account for varying numbers.
b
No contrast significant. Time Effects
Main effects for time, which reflect change over time
Placebo Substitution at the End of Year I regardless of treatment condition, were examined to
Relapse was defined as fulfilling diagnostic criteria provide heuristic information regarding the possible at-
for ADHD, being rated worse by two of three raters, tenuation of effects with long-term methylphenidate
and a 25% increase on the CTRS hyperactivity factor. treatment. Time effects during year I, relative to pre-
We tested for differential survival to single-blind pla- treatment, and after brief methylphenidate treatment
cebo substitution as well as time maintained on pla- are summarized as well as time effects during year II.
cebo. Year I Time Effects Relative to Pretreatment. Signifi-
All children (100%) relapsed when switched to pla- cant time effects were obtained on all measures, show-
cebo and were placed back on methylphenidate regard- ing improvement at 6 and 12 months. Moreover,
less of the treatment received during the previous year. children continued to improve between 6 and 12
The mean number of days to reinstitution of methyl- months of treatment on parent ratings of problematic
phenidate for each group was as follows: M alone, 8.6 situations (Table 1) and on symptoms of ADHD and
5.4 (range 122); M + MPT, 17.1 16.2 (range ODD at home and school (Table 2) but not on teacher
562); and M + ACT, 11.7 12.8 (range 169). A ratings (Table 3).

806 J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 43:7, JULY 2004

 TREATMENT EFFECTS ON ADHD SYMPTOMS

TABLE 5
Mean Rates of Observed ADHD Behaviors in Academic Classes
Treatment Group
Normal M M + MPT M + ACT
Observed Behavior Mean SD Mean SD Mean SD Mean SD
a
Interference
Pretreatment 8.0 5.9 16.3 11.1 15.8 8.1 17.4 7.8
Medication baseline 8.1 5.7 9.2 5.9 9.0 6.0
6 mo 5.9 4.6 9.0 7.7 7.1 6.6 6.7 5.8
12 mo 6.1 5.1 7.4 5.6 8.2 8.2 6.3 6.1
18 mo 5.6 5.3 6.8 7.1 6.7 6.9 4.3 4.3
24 mo 4.5 4.3 6.6 7.5 7.6 7.4 5.1 5.4
Off taska
Pretreatment 2.2 2.9 6.5 6.2 4.8 3.8 6.0 6.3
Medication baseline 3.2 5.2 1.6 3.1 2.9 4.6
6 mo 0.9 1.6 2.9 4.4 1.5 2.6 1.4 1.9
12 mo 1.8 3.7 1.9 3.7 2.1 2.6 2.1 4.4
18 mo 1.5 3.0 2.9 3.7 1.6 2.4 1.3 1.7
24 mo 1.6 4.1 1.9 2.6 3.6 3.4 2.8 4.3
Gross motora,b
Pretreatment 1.3 1.3 3.1 2.5 2.9 2.1 3.7 2.8
Medication baseline 2.0 3.2 1.9 1.3 1.9 1.5
6 mo 1.8 2.4 2.7 2.5 1.1 1.1 2.3 2.1
12 mo 1.6 1.6 1.7 1.8 1.8 2.1 1.4 1.3
18 mo 1.4 1.6 1.5 1.4 1.3 0.9 1.4 1.3
24 mo 1.2 1.5 1.2 1.4 1.5 1.5 1.1 1.2
ADHD Compositea,c,d
Pretreatment 11.5 7.8 25.9 15.9 23.9 10.8 27.1 12.6
Medication baseline 13.3 8.3 12.7 7.6 13.7 9.2
6 mo 8.6 5.8 14.6 11.5 9.7 7.0 10.5 7.5
12 mo 9.5 7.3 11.0 8.4 12.1 10.6 9.8 8.1
18 mo 8.6 7.9 11.1 7.9 9.6 8.1 7.1 5.9
24 mo 7.2 7.1 9.7 10.1 12.6 10.6 9.0 8.2

Note: ADHD = attention-deficit/hyperactivity. Group time interactions: none significant.

a
Time effects in year I: pretreatment versus 6 and 12 months, p < .001.
b
Trend in group time interaction: M + MPT versus M + ACT at 6 months, p < .02; M + MPT versus M at 6 months, p < .003.
c
Trend for time effects in year II: 12 versus 18 months, p < .05.
d
Time effects in normals: Pretreatment to 12 months, p < .02; 12 to 24 months, p = .12.

Year I Time Effects Relative to Medication Baseline.
Relative to medication baseline, school observations
did not reflect reduced rates of ADHD behaviors
(Table 5). However, all groups improved on the
inattention/overactivity ratings in academic (p < .01)
and gym classes (p < .001) (Table 6).
Year II Time Effects. All groups maintained treat-
ment gains during year II, with some further improve-
ment over time. Parents reported lower scores on the
Hyperkinesis Index (p < .01) and trends on Home
Situations Questionnaire problematic situations (p <
.04) and severity (p < .02) (Table 1).
Teacher ratings, psychiatrist ratings of ADHD and
ODD symptoms, and observer ratings did not show
significant time effects between 12 and 24 months.
DISCUSSION
This dual-site study represents an effort to provide a
broad psychosocial treatment program aimed at opti-
mizing multiple aspects of function in children with
ADHD treated with methylphenidate. Measures of
ADHD and related behavior problems did not indicate
any meaningful advantage for the combination of

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 43:7, JULY 2004 807

ABIKOFF ET AL.

TABLE 6
IOWA CTRS Ratings in Academic and Gym Classes
Treatment Group
Normal M M + MPT M + ACT
Setting and Factor Mean SD Mean SD Mean SD Mean SD
Classroom
Inattention/overactivitya,b,c
Pretreatment 0.3 0.3 1.1 0.6 0.9 0.6 1.2 0.6
Medication baseline 0.4 0.4 0.6 0.5 0.6 0.5 0.5 0.5
6 mo 0.2 0.2 0.6 0.6 0.3 0.3 0.4 0.4
12 mo 0.3 0.3 0.4 0.4 0.4 0.4 0.3 0.1
18 mo 0.2 0.3 0.3 0.3 0.2 0.2 0.3 0.3
24 mo 0.2 0.3 0.4 0.5 0.5 0.6 0.3 0.5
Oppositional/defianta,d
Pretreatment 0.03 0.1 0.2 0.3 0.2 0.4 0.4 0.5
Medication baseline 0.05 0.1 0.2 0.4 0.1 0.2 0.1 0.2
6 mo 0.02 0.1 0.3 0.6 0.0 0.1 0.1 0.2
12 mo 0.03 0.1 0.1 0.1 0.1 0.1 0.0 0.1
18 mo 0.1 0.2 0.0 0.04 0.0 0.1 0.1 0.2
24 mo 0.02 0.1 0.1 0.2 0.2 0.3 0.1 0.1
Gym
Inattention/overactivitya,b,e
Pretreament 0.2 0.3 0.7 0.5 0.6 0.4 0.7 0.5
Medication baseline 0.2 0.2 0.4 0.4 0.3 0.3 0.3 0.3
6 mo 0.2 0.2 0.4 0.6 0.3 0.3 0.2 0.3
12 mo 0.2 0.3 0.1 0.2 0.2 0.3 0.2 0.3
18 mo 0.2 0.3 0.2 0.2 0.2 0.3 0.1 0.1
24 mo 0.1 0.2 0.2 0.5 0.2 0.2 0.2 0.3
Oppositional/defianta
Pretreatment 0.1 0.1 0.2 0.2 0.2 0.2 0.3 0.5
Medication baseline 0.1 0.2 0.2 0.4 0.2 0.3 0.2 0.3
6 mo 0.04 0.1 0.2 0.5 0.1 0.1 0.1 0.2
12 mo 0.1 0.3 0.1 0.2 0.1 0.2 0.1 0.3
18 mo 0.1 0.2 0.1 0.2 0.1 0.2 0.0 0.1
24 mo 0.1 0.1 0.1 0.3 0.0 0.1 0.2 0.2
Note: IOWA CTRS = Conners Teacher Rating Scale (range 03).
a
Time effects in year I: pretreatment versus 6 and 12 months, p < .02.001.
b
Medication baseline versus 12 months, p < .01.001.
c
Significant time effects in year II: 18 versus 24 months, p < .012.
d
Trends in group time interaction: M + MPT versus M at 6 months and M + ACT versus M at 6 months, p < .02; M + ACT versus
M at 12 months, p = .01.
e
M + ACT < M at 6 months, p < .03.

methylphenidate and MPT over methylphenidate
alone, and methylphenidate could not be discontinued
more successfully in children who received the combi-
nation treatment.
Negative findings were obtained in spite of efforts to
disprove the null hypothesis, i.e., to find treatment
differences. Thus, multiple sources of information were
obtained. The statistical analytic strategy reflects
greater concern for type II errors (, power) compared
with the conventional, conservative strategy of concern
for type I errors (, significance). We conducted mul-
tiple analytic methods including repeated measures,
heuristic random regression, and mixed models analysis
of variance. In addition, the facts that parents, teachers,
and psychiatrists were not blind to treatment and that
parents were involved in treatment delivery should
have biased the results in favor of psychosocial treat-
ment. The MTA Cooperative Group study is the only

808 J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 43:7, JULY 2004


other systematic investigation that has implemented an
intensive psychosocial intervention (MTA Cooperative
Group, 1999). Unlike the MTA (multimodal treatment
of ADHD) study, we did not assess the role of psycho-
social treatment alone because more than three treat-
ment groups would have limited power, and our
previous findings indicated that psychosocial treatment
was not competitive with stimulant treatment (Klein
and Abikoff, 1997). Indeed, in the MTA study, psy-
chosocial treatment was significantly inferior to medi-
cation, even though a sizable proportion of children
randomized to psychosocial treatment were placed on
medication during the course of the study and even
though treatment evaluators (parents and teachers)
were actively involved in the delivery of psychosocial
treatment. Unlike the MTA study, we did not imple-
ment active intervention in the classroom or an inten-
sive therapeutic summer program. Finally, our study
lasted 24 months, in contrast to 14-month MTA study.
In spite of these methodological differences, results
from the two studies are strikingly similar. Both failed
to document that MPT provides benefits above those
obtained with methylphenidate alone for ADHD and
ODD dimensional ratings (MTA Cooperative Group,
1999). Moreover, an additional year of maintenance
treatment failed to produce differential benefits in chil-
dren who received combination treatment compared
with those treated solely with methylphenidate. How-
ever, treatment gains from year I were maintained dur-
ing year II.
We considered possible reasons for the lack of ad-
junctive efficacy of psychosocial treatment. Perhaps if
attendance at psychosocial treatment sessions were
poor, as has occurred in other studies (Firestone et al.,
1981; Schachar et al., 1997), many might have received
inadequate treatment. This explanation seems unlikely
because families attended at least 75% of all treatment
sessions.
Perhaps treatment fidelity was compromised and
psychosocial treatments were not delivered as intended.
Ongoing supervision and audiotape reviews throughout
the course of the study protected against this possibility.
Perhaps medication compliance differed among
treatment groups, thereby affecting treatment out-
come. However, results of methylphenidate checks do
not support this explanation (Klein et al., 2004).
We considered the possibility that multimodal treat-
ment would be especially effective in ODD children
J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 43:7, JULY 2004
TREATMENT EFFECTS ON ADHD SYMPTOMS
with ADHD and reanalyzed year I results with ODD
diagnosis entered as another factor. Children with and
without ODD responded similarly and positively in all
three treatment groups. Finally, a ceiling effect with
short-term methylphenidate might have been reached,
precluding accrued benefits from the combination
treatment over medication alone. This possibility ap-
pears unlikely because time effects were found, indicat-
ing continued improvement with methylphenidate.
The hope was that the combination treatment would
reduce the incidence of CD. This was not the case.
Relatively few children in the study (4.7%) developed
CD by ages 9 to 12, when such onsets are not unusual
in boys with ADHD and ODD (Lahey et al., 2000). As
with other time effects, it is not clear whether methyl-
phenidate treatment, the common denominator across
treatments, was instrumental in reducing the develop-
ment of full-fledged CD. However, what is clear is that
oppositional symptoms, significant risk factors for CD,
were markedly reduced.
There was a slight advantage for children who re-
ceived combination treatment insofar as they remained
on placebo 8.5 days longer than those on methylphe-
nidate alone after 1 year. However, this difference falls
short of meaningful clinical advantage.
The absence of an untreated control group precludes
unambiguous conclusions regarding improvements
over time because maturational effects cannot be ruled
out. In spite of their ambiguous implications, time ef-
fects in children with ADHD point to functions that
might continue to improve with methylphenidate
treatment. During year I, children had significant re-
ductions in ODD and ADHD symptoms, which were
maintained during year II. Furthermore, the severity of
ADHD continued to improve over year II. Such im-
provements over time are all relatively small, and their
clinical significance is limited. Importantly, findings do
not indicate a loss of treatment efficacy over 2 years of
methylphenidate treatment. This finding argues against
concerns that stimulant effects attenuate with long-
term use (Cantwell, 1975; Kupietz et al., 1988). To our
knowledge, this 2-year study is one of the longest to
demonstrate the long-term efficacy of stimulant treat-
ment in children with ADHD.
Limitations
As in all clinical trials, the generalizability of findings
is defined in part by inclusion and exclusion criteria.
809
ABIKOFF ET AL.
Results are relevant to young, mostly white children
with ADHD who demonstrate some meaningful ben-
efit with short-term treatment. However, it would
seem pointless, if not unethical, to administer long-
term medication to children who did not demonstrate
short-term benefit. Divergent results might occur in
other clinical subgroups, such as groups with low so-
cioeconomic status, greater disadvantage, and comor-
bidity.
Findings do not apply to children with ADHD who
also have CD, although 30% of the children had symp-
toms of CD and 53% had ODD (Klein et al., 2004).
Virtually identical results were obtained in the MTA
study, which had broader inclusion criteria. Because of
the ambitious goal of altering the long-term course of
ADHD, both the MTA study and the current study
were restricted to 7- to 9-year-old children. Whether
similarly negative findings would apply to younger
children is unknown. So far, studies of psychosocial
interventions in young children have not included
medication comparisons (Barkley et al., 2000; Sonuga-
Barke et al., 2001).
Our study obtained daily report cards from teachers
but did not implement a school-based treatment. To
the extent that such interventions are crucial, we may
have missed efficacy for psychosocial treatment. This
probability seems unlikely because the MTA study ob-
tained similarly negative results in spite of providing
school-based treatment.
Clinical Implications
Findings document that initial benefits of methyl-
phenidate do not diminish over time; rather, incremen-
tal gains appear likely with extended stimulant
treatment. The study does not support the expectation
that stimulant-treated children with ADHD should
routinely receive psychosocial treatment to reduce
ADHD, ODD, and CD symptoms.
Disclosure: Dr. Abikoff is a member of the ADHD Advisory Board and
a principal investigator in clinical trials, Shire Pharmaceutical Co.,
and a member of the Metadate CD Advisory Board of Celltech Phar-
maceuticals. He is a recipient of an investigator-initiated grant from
McNeil Consumer and Specialty Pharmaceuticals. Dr. Hechtman re-
ceived research funding from Eli Lilly, Janssen Ortho, Purdue, Shire
Pharmaceutical Co., and GlaxoSmithKline Beecham and is on the
speakers roster of Shire Pharmaceutical Co., Janssen Ortho, and Eli
Lilly. Dr Klein is a member of the ADHD Advisory Board of Shire
Pharmaceutical Co.
810
REFERENCES
Abikoff H, Gittelman R (1984), Does behavior therapy normalize the class-
room behavior of hyperactive children? Arch Gen Psychiatry 41:449454
Abikoff H, Gittelman R (1985a), Hyperactive children treated with stimulants.
Is cognitive training a useful adjunct? Arch Gen Psychiatry 42:953961
Abikoff H, Gittelman R (1985b), The normalizing effects of methylpheni-
date on the classroom behavior of ADDH children. J Abnorm Child
Psychol 13:3344
Abikoff H, Gittelman R, Klein DF (1980), A classroom observation code
for hyperactive children: a replication of validity. J Consult Clin Psychol
48:555565
Abikoff H, Hechtman L, Klein RG et al. (2004), Social functioning in
children with ADHD treated with long-term methylphenidate and mul-
timodal psychosocial treatment. J Am Acad Child Adolesc Psychiatry 43:
820829
Barkley RA (1990), Attention-Deficit Hyperactivity Disorder: A Handbook for
Diagnosis and Treatment. New York: Guilford
Barkley RA, Shelton TL, Crosswait C et al. (2000), Multi-method psycho-
educational intervention for preschool children with disruptive behav-
ior. J Child Psychol Psychiatry 41:319332
Cantwell DP (1975), A critical review of therapeutic modalities with hy-
peractive children. In: The Hyperactive Child, Cantwell DP, ed. New
York: Spectrum, p 180
Carlson CL, Pelham WE Jr, Milich R, Dixon J (1992), Single and com-
bined effects of methylphenidate and behavior therapy on the classroom
performance of children with attention-deficit hyperactivity disorder.
J Abnorm Child Psychol 20:213232
Douglas VI, Barr RG, ONeill ME, Britton BG (1986), Short-term effects
of methylphenidate on the cognitive, learning and academic perfor-
mance of children with attention deficit disorder in the laboratory and
the classroom. J Child Psychiatry 27:191211
Firestone P, Kelly MJ, Goodman JT, Davey J (1981), Differential effects of
parent training and stimulant medication with hyperactives. J Am Acad
Child Psychiatry 20:135147
Gittelman-Klein R, Klein DF, Abikoff H, Katz S, Gloisten AC, Kates W
(1976), Relative efficacy of methylphenidate and behavior modification in
hyperkinetic children: an interim report. J Abnorm Child Psychol 4:361379
Gittelman R, Mannuzza S, Shenker R, Bonagura N (1985), Hyperactive
boys almost grown up. Arch Gen Psychiatry 42:937947
Goyette CH, Conners CK, Ulrich RF (1978), Normative data on revised
Conners Parent and Teacher Rating Scales. J Abnorm Child Psychol 6:
221236
Granger DA, Whalen CK, Henker B, Cantwell C (1996), ADHD boys
behavior during structured classroom social activities. J Atten Disord 1:
1630
Hechtman L, Abikoff H, Klein RG et al. (2004a), Children with ADHD
treated with long-term methylphenidate and multimodal psychosocial
treatment: impact on parental practices. J Am Acad Child Adolesc Psy-
chiatry 43:830838
Hechtman L, Abikoff H, Klein RG et al. (2004b), Academic achievement
and emotional status of children with ADHD treated with long-term
methylphenidate and multimodal psychosocial treatment. J Am Acad
Child Adolesc Psychiatry 43:812819
Hinshaw SP (1991), Stimulant medication and the treatment of aggression
in children with attentional deficits. J Clin Child Psychol 20:301312
Hinshaw SP, Heller T, McHale JP (1992), Covert antisocial behavior in
boys with attention-deficit hyperactivity disorder: external validation
and effects of methylphenidate. J Consult Clin Psychol 60:274281
Hinshaw SP, Klein RG, Abikoff H (2002), Childhood attention-
deficit/hyperactivity disorder: nonpharmacologic treatments and their
combination with medication. In: A Guide to Treatments That Work,
2nd ed., Nathan PE, Gorman JM, eds. New York: Oxford
Klein RG, Abikoff H (1997), Behavior therapy and methylphenidate in the
treatment of children with ADHD. J Atten Disord 2:89114
Klein RG, Abikoff H, Hechtman L, Weiss G (2004), Design and rationale
of controlled study of long-term methylphenidate and multimodal psy-
chosocial treatment in children with ADHD. J Am Acad Child Adolesc
Psychiatry 43:792801
J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 43:7, JULY 2004

Klein RG, Abikoff H, Klass E, Ganeles D, Seese LM, Pollack S (1997),
Clinical efficacy of methylphenidate in conduct disorder. Arch Gen
Psychiatry 54:10731080
Kupietz SS, Winsberg BG, Richardson E, Maitinsky S, Mendell S (1988),
Effects of methylphenidate dosage in hyperactive reading-disabled chil-
dren. J Am Acad Child Adolesc Psychiatry 27:7077
Lahey BB, McBurnett K, Loeber R (2000), Are attention-deficit/hyper-
activity disorder and oppositional defiant disorder developmental pre-
cursors to conduct disorder? In: Handbook of Developmental Psychopa-
thology, 2nd ed., Sameroff AJ, Lewis M, Miller SM, eds. New York:
Kluwer Academic/Plenum, pp 431446
Loney J, Milich R (1982), Hyperactivity, inattention, and aggression in
clinical practice. In: Advances in Developmental and Behavioral
Pediatrics, Wolraich M, Routh D, eds. Greenwich, CT: JAI Press, pp
113147
Mannuzza S, Klein RG, Bessler A, Malloy P, LaPadula M (1993), Adult
outcome of hyperactive boys. Arch Gen Psychiatry 50:565576
J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 43:7, JULY 2004
TREATMENT EFFECTS ON ADHD SYMPTOMS
MTA Cooperative Group (1999), A 14-month randomized clinical trial of
treatment strategies for attention-deficit/hyperactivity disorder. Arch
Gen Psychiatry 56:10731086
Paternite CE, Loney J, Salisbury H, Whaley MA (1999), Childhood inat-
tention-overactivity, aggression, and stimulant medication history as
predictors of young adult outcomes. J Child Adolesc Psychopharmacol 9:
169184
Schachar RJ, Tannock R, Cunningham Corkum PV (1997), Behavioral,
situational, and temporal effects of treatment of ADHD with methyl-
phenidate. J Am Acad Child Adolesc Psychiatry 36:754763
Shaffer D, Gould MS, Brasic J (1985), CGAS (Childrens Global Assess-
ment Scale). Psychopharmacol Bull 21:747748
Sonuga-Barke EJS, Daley D, Thompson M, Laver-Bradbury C, Weeks A
(2001), Parent-based therapies for preschool attention deficit/hyper-
activity disorder. J Am Acad Child Adolesc Psychiatry 40:402408
Weiss G, Hechtman L (1993), Hyperactive Children Grown Up: ADHD
Children, Adolescents, and Adults. New York: Guilford Press
811