17. Pediatric Psychopharmacological and Combined Interventions

Stimulant medications can be delivered in a variety of ways, including immediate release, intermediate, extended release, and transdermal patches (Connor, 2006; Wilens, 2001). The main differences in the preparation of stimulants have to do with how quickly the medication is absorbed by the central nervous system, the rate of absorption, or peak plasma or brain concentration of the stimulant (Connor, 2006). Immediate release preparations include: methylphenidate (brand name, Ritalin, Methylin, Metadate), dextromethylphenidate (brand name Focalin), dextroamphetamine (brand name Dextrostat, Dexedrine), and mixed amphetamine salts (brand name Adderall). Intermediate release preparations include: methylphenidate (brand name, Ritalin SR, Methylin ER, Metadate ER), and dextromethylphenidate (brand name Dexedrine spansule). Methylphenidate can also be delivered in quicker acting, longer lasting intermediate formulas (brand name Metadate CD, Ritalin LA). Extended release formulas include methylphenidate (brand name Concerta) and mixed amphetamine salts (brand name Adderall XR). See Connor (2006) for additional information about onset of action, peak clinical effect, serum half-life, duration of behavioral effects and required number of daily doses. Transdermal patches have also been developed that deliver methylphenidate for up to 12-hour periods.

Standard dosing practices have changed over the years, from BID (morning and noon administration) to TID which includes a three-times-per-day dosing with a late afternoon dose (Corkum, Panton, Ironside, MacPherson, & Williams, 2008). Long acting or sustained release formulas also treat symptoms after school. Research to date shows that the long acting delivery systems appear to be effective (Biederman, Faraone, Monuteaux, & Grossbard, 2004).

The transdermal methylphenidate system also seems effective for improving core symptoms of ADHD, academic performance, and deportment based on teacher and parent ratings (McGough et al., 2006). The patch was well tolerated and, when present, adverse effects were mild to moderate in intensity. The adverse effects were similar to those reported in oral delivery systems, and abated after several weeks.

The potential benefits of stimulant medications are well documented and include enhanced performance on the major symptoms of ADHD, including impulse control, motor coordination, and vigilance (Connor, 2006; Pliszka, 2003); improved cognitive functioning (Barkley, DuPaul, & McMurray, 1991); increased academic productivity and accuracy (Balthazor, Wagner, & Pelham, 1991); decreased off-task behaviors (Barkley & Cunningham, 1979); decreased aggression (Hinshaw, Henker, Whalen, Erhardt, & Dunnington, 1989); improved peer relations (Hinshaw, 1991); fewer negative commands from teachers (Barkley, 2006), and improved interactions with parents (Barkley, 2006).

In general stimulant medications produce improvement in compliance, impulsive aggression, social interactions and academic performance (Wilens & Spencer, 2000). Many of the improvements in interpersonal domains occur not only because the child’s behaviors improve, the behaviors of the adults supervising the child are indirectly affected (e.g., fewer negative interactions, less hostility) (Connor, 2006). Social skills, per se, do not necessarily improve without specific skill development and contingencies to support these new skills.

Despite positive results, individual response remains highly variable (DuPaul, Barkley, & McMurphy, 1991), and children should be carefully monitored for adverse side effects (Connor, 2006; Wilens, 2001). Also, a number of environmental factors may affect a child’s positive response to stimulant medication. For example, Barkley and Cunningham (1980) found that the better the mother-child relationship, the greater the positive response rate in the child.

Multimodal treatment generally includes parent and child education about ADHD, stimulant medication, behavioral therapy, and educational interventions for psychosocial and academic difficulties. The Multimodal Treatment Study of Children with ADHD (MTA Cooperative Group, 2004a, b) investigated the efficacy of medication alone (MedMgt), behavioral and psychosocial treatments alone (Beh), combined medication with behavioral interventions (Comb), and a control group who received treatment in a community setting (CC) or treatment as usual. The study is complex in design and is ongoing, following study participants from childhood into early adulthood.

Initial findings after 14 months of treatment showed that all four groups improved (MTA Cooperative Group, 2008a). However, there were some significant findings that are summarized: (1) MedMgt produced larger benefits than behavior therapy; (2) Comb treatment did not significantly increase the overall benefits of MedMgt alone; (3) participants in the Comb treatment group had 20 a percent lower dose of medication with similar results as those children in the MedMgt group who had higher doses; (4) Comb treatment was superior to other treatments for children in families on public assistance; (5) Comb treatment was superior to other treatments for children with ADHD and comorbid anxiety, and (6) families with higher rates of attendance at monthly clinic visits had better treatment outcomes. Medication as prescribed in the MTA study produced clinical improvement in the core symptoms of ADHD -- hyperactivity, inattention and impulsivity. Comb treatment (medication and behavioral therapy) was relatively superior to Beh therapy alone and to CC. Even though some children in CC care received medication (38%), they fared less well compared to those in the MedMgt group. These differences may be the result of lower medication doses prescribed in CC compared to those receiving the MTA MedMgt algorithm. The Beh therapy was also found to be more effective than CC – treatment as usual.

Youth receiving medication in CC group had better outcomes than the non-medicated CC youth, and outcomes were similar to those in the Beh group. However, children in the MedMgt had better outcomes than those receiving treatment as usual in the CC group. It is likely that the manner in which medications are administered and monitored in the community affects long-term outcomes.

The MTA did a 24-month follow-up to measure the long-term outcome of treatment (MTA Cooperative Group, 2008a). The major outcomes include the following. (1) There was a persistent relative superiority of MedMgt and Comb groups over the Beh and CC treatment groups, although effect size was lower than 14-month analyses. (2) While participants taking medication at 24 months had better outcomes than those who were not medicated, there was a partial loss of the relative benefits of medication compared to the data from the 14-month interval. (3) Youths who were continually medicated had slower growth gains compared to non-medicated youths (1 cm/year reduction; 1.2 kg/year in weight gain), and stimulant growth suppression may continue when medication treatment is maintained. (4) The compliance with assigned medication doses dropped at the 24-month follow-up, while the percentage of children in the Beh group increased their use of medication. This may have affected the results between the treatment groups (see #1 above). (5) Children who stopped medication at the 24-month follow-up had greater overall deterioration in outcomes, while those children starting medication after the initial study phase improved during the 10-month interval.

The 36-month follow-up of the MTA study revealed interesting and sometimes confusing results at the 2nd follow-up (MTA Cooperative Group, 2008a, 2008b). The relative superiority of Comb and MedMgt over the Beh and CC was completely lost at 36 months. The relative superiority of the MTA medication algorithm was lost, and continued medication was a marker of deterioration in some children. There were three different outcome trajectories that are important. Two groups (66% of sample) showed large initial improvement on medication. Class 1 (52%) had a large initial improvement that was maintained over time, while another Class 3 (14%) had large initial improvement that was not maintained at 36 months. In fact, Class 2 showed deterioration at the 36-month follow-up. Family demographics showed that this group had higher rates of adversity than the other groups. Further, Class 2 (34% of sample), had initial modest improvement on medication that gradually increased over time, and was significantly better than those who were not medicated. The first two groups (66%) of the sample had a large and significant improvement with medication over time, and for Class 1 the magnitude of improvement was even greater at 36 months.

The MTA Cooperative Group (2008b) also investigated moderator variables to determine how comorbidity factors impact outcomes. Children with comorbid anxiety had improved outcomes when given Beh therapy, and with outcomes similar to both the MedMgt and the Comb therapies. The Beh, Comb, and MedMgt were superior to treatment as usual (CC therapy) for children with anxiety. Children with disruptive disorder, ADHD and anxiety also fared better with Comb treatment compared to the other therapies. Further, Comb, but not MedMgt, therapies reduced the persistence of oppositional defiant disorder (ODD) and mood disorders.

Family demographics appeared to impact the effectiveness of various treatment modalities. Children from families with income challenges (receiving public assistance versus those who were not) had better outcomes in the Comb therapy compared to other treatments (MTA Cooperative Group, 2008b). Significant reductions in negative/ineffective discipline practices only occurred in the Comb treatment group. In general, families with socioeconomic disadvantage responded most favorably to Beh treatment. In addition, children receiving Comb and Beh therapies had some protection against substance experimentation and delinquency compared to MedMgt and CC groups that were not protected.

While the MTA study has been the most comprehensive of its kind, there are complexities to the data and analyses that have created controversies and confusion. The study is ongoing, and eight-year follow-up data were presented at the MTA Research Symposium (2007), 10-year data analyses have been completed, and 12-year follow-up is in progress (MTA Cooperative Group, 2008b). To date the study has shown long-term benefits of stimulants over a one-year period. These findings are consistent with those reported by Abikoff et al. (2004) which documented benefits of stimulant medications over a two-year period when medications are properly monitored and adjusted. At the 24-month follow-up, the MTA data found that the relative benefits of stimulants were gradually reduced when children return to community care. In the three-year follow-up, treatment benefits dissipated completely for some children, while others continued to improve. In the future, the MTA study will address important questions about quality care for ADHD, the ultimate effects on height and weight for children who show stimulant-related growth suppression, and long-term functional outcomes of treatment (MTA Cooperative Group, 2008b).

A non-stimulant medication, atomoxetine hydrochloride (brand name Strattera) was recently approved and has shown to be effective for reducing the core symptoms of ADHD (Kelsey et al., 2004). Comparison studies are somewhat inconsistent, with some reports showing greater improvements on Adderall (extended release mixed amphetamine salts) compared to Strattera (Faraone, Wigal, & Hodgins, 2007), while others show that Strattera is comparable to that of methylphenidate (Kratochvil et al., 2002). Barkley, Anderson, and Kruesi (2007) also found that Strattera improved self-ratings of ADHD symptoms in adults with ADHD and self-evaluations in a driving simulator, although these improvements were not noted for examiner-rated driving performance. Bohnstedt et al. (2005) also found that parent and teacher ratings of ADHD showed significant improvement for children on atomoxetine compared to placebo. Parent ratings appeared to detect “a larger effect and accounted for more unique variance in the prediction of treatment type, independent of teacher-based ratings” (p. 158).

Other medications used for children who are considered non-responders to stimulant medications include imipramine (Connor, 2006), and MAOI (Zametkin, Rapoport, Murphy, Linnoila, & Ismond, 1985). Desipramine selectively blocks the uptake of NE, and imipramine may block the uptake of serotonin. While antidepressants may increase NE availability at the synapse (Hunt et al., 1991), these are not the frontline medications of choice for most children with ADHD.

Wood, Crager, Delap, and Heiskell (2007) reviewed non-stimulant medications that have been used to treat ADHD. Tricyclic antidepressants have been used to treat individuals with ADHD, particularly those with comorbid disorders including mood, anxiety, oppositional and tic disorders. Selective serotonin reuptake inhibitors (SSRIs) have a more cautioned tale. In 2004, the FDA issued a “black box” warning about the risk for suicide in youths taking SSRIs. “The implementation of an SSRI in treating ADHD should only be considered when there is a dual diagnosis of ADHD with depression or anxiety” (Wood et al., 2007, p. 344). Other side effects for SSRIs (increased mania, agitation, insomnia, akathisia, and sleep difficulties) should be carefully monitored. Other antidepressants (i.e., bupropion, venlafaxine) block the reuptake of serotonin and norepinephrine, but children on these medications require regular electroencephalogram (EEG) monitoring to assess the risk for seizure activity in a very small percentage of individuals. Studies of antihypertensive medications show that clonidine (CLN) reduces ADHD symptoms (Wilens & Spencer, 1999). Clonidine appears to successfully reduce aggression, sleep disturbances, and tics [see (Wood et al., 2007) for a review]. While improvement in conduct-related problems has been shown when CLN is used in combination with stimulants, serious side effects have been reported. In rare instances, sudden death in children has been reported when stimulants and CLN are combined, but evidence for the linkage is not well established (Taylor et al., 2004). Although promising, less research has been conducted testing the efficacy of guanfacine (GFN).

A number of atypical antipsychotic drugs have been used for children experiencing high levels of behavioral problems, including aggression and disruptiveness. Risperidone (RPD) and aripiprazole (ARP) seem to improve severe behaviors, but are less helpful for core ADHD symptoms (hyperactivity and cognition) (Wood et al., 2007). Serious side effects reduce the widespread use of these medications, except for children with extreme behavioral problems. Carbamazepine (CBZ), a seizure medication, appears to be safe and effective for children, but serious side effects (i.e., dizziness, headaches, drowsiness, ataxia, blurred vision, nausea, vomiting, rash, heptic abnormalities) limit their widespread use as an alternative medication for ADHD (Silva, Munoz, Alpert, 1996; Pellock, 1987). Wood et al. (2007) conclude “Despite their variability as alternatives, each type of medication has certain limitations, side effects, and varying amounts of research available to substantiate its use for the treatment of ADHD. These medications offer several potentially successful options to those who fail to respond to stimulants or those who find the side effects of stimulants bothersome” (p. 341).

NIMH funded a large scale, multi-site study of the treatment of depression in teens (TADS, 2004). Treatment conditions were medication (fluoxetine) alone; cognitive behavioral therapy (CBT) alone; combined treatment (fluoxetine plus CBT) and placebo. The study found that antidepressant medication was effective for treating youths with depression, but the combined treatment produced the greatest improvement. CBT alone was less effective than medication alone, although both treatments were better than the placebo condition.

The combined treatment significantly reduced depressive symptoms, and was superior to medication and CBT alone (TADS, 2004). Teens in the study were carefully monitored for adverse side effects including gastrointestinal track events, sedation, and insomnia. Medication doses were adjusted to reduce these effects.

While 30 percent of teens had suicidal ideation at the start of the study, suicidal ideation was reduced in all groups. Decreases in suicidal ideation were greatest for teens in the combined treatment condition, while medication alone was not as protective. Based on this study, CBT appears to be an effective treatment for depression and protects teens with SI with carefully monitored medication (prozac).

The National Institute of Mental Health (NIMH) funded the nation’s largest study investigating 18–75-year-olds with treatment-resistant depression (Rush et al., 2004; Weissman et al., 2006). STAR*D data show that about half of all individuals with difficult to treat depression reached remission with additional treatment, but the odds were reduced with each new trial of medication (Rush et al., 2006). Approximately one-third of participants reached remission at Level 1 treatment when given Celexa (citalopram), an antidepressant. CBT was used either as a switch or add-on treatment during Level 2 of the study, while at Levels 3 and 4 new medications were either switched or added on. Drop-out of the study also increased when additional treatments were not effective. It is important to note that patients who have not responded positively to two prior antidepressants and then switch to a different class of antidepressants have only a minimal chance at remission by taking the new medication (Rush et al., 2004, 2006). Additional research is being conducted on the STAR*D sample to determine the efficacy of either adding on or switching to CBT, and to determine who responds to what treatment sequence.

In another phase of the STAR*D study, Zisook et al. (2007) investigated the impact of early versus late onset of depression to determine medication response across five age groups: childhood onset (ages <12), adolescent onset (ages 12–17), early adult onset (ages 18–44), middle adult onset (ages 45–59), and late adult onset (ages ≥ 60). “No group clearly stood out as distinct from the others. Rather, the authors observed an apparent gradient, with earlier ages at onset associated with never being married, more impaired social and occupational function, poorer quality of life, greater medical and psychiatric comorbidity, a more negative view of life and the self, more lifetime depressive episodes and suicide attempts, and greater symptom severity and suicidal ideation in the index episode compared to those with later ages at onset of major depressive disorder” (Zisook et al., 2007, p. 1539). Thus, age of onset was not associated with a difference in treatment response to the initial trial of citalopram.

In a study of 12–18 year olds who were non-responders to one SSRI, Brent et al. (2008) investigated other antidepressants, with and without CBT. Teens in the TORDIA study were exposed to the following treatments: (1) switch to a 2nd SSRI (i.e., paroxetine, citalopram, or fluxotine); (2) switch to different antidepressant medication plus CBT; (3) switch to venlafaxine, and (4) switch to venlafaxine plus CBT. Those teens who had switch to 2nd SSRI and also received CBT showed the most improvement. Treatment with venlafaxine (compared to another SSRI) was just as efficacious with fewer adverse effects than the other SSRIs.

In summary, continued research on the safety and efficacy of all antidepressants in children and adolescents is needed. While research has shown that some antidepressants can effectively treat depression in teens, especially when combined with cognitive behavioral therapy, the findings for young children are more equivocal. Regardless of the study, clinicians are advised to carefully monitor the adverse events that accompany antidepressants, particularly suicidal ideation and/or behaviors.