Treatment Resistance in Youths With ADHD and Comorbid Conditions

Since its introduction in DSM-III in 1980, attention-deficit/ hyperactivity disorder (ADHD) has proved to be a developmental disorder with many causes and complex behavioral, cognitive, and emotional manifestations that can impair academic functioning, occupational achievement, social relationships, and self-esteem.^1-3 The immediate and long-term impairments of ADHD are of great concern, particularly as more and more longitudinal studies document lifelong patterns of underachievement, sexual-reproductive and driving risks, alcohol(Drug information on alcohol) and substance abuse, mood disorders, and reduced executive functions and coping skills.^4-8

Recently, comorbid psychiatric conditions associated with ADHD have been identified as major obstacles to successful treatment and functional outcomes. The high prevalence of these conditions (Table 1) along with their negative impact on treatment outcomes has been documented in numerous studies of patients across their life spans.^2,9-12 It is evident that comorbid conditions impose heavier burdens on patients with ADHD (as well as their families), and that they make it more difficult for clinicians to choose interventions that are likely to succeed with a high degree of certainty. Whereas the evidence base for multimodal treatment of ADHD "simplex" in children and adolescents is well established,¹³ such is not the case for patients with complex forms of the disorder.

The following case vignette illustrates a common occurrence in modern psychiatric practice: the presence of severe comorbidity in ADHD patients leading to a less than optimal treatment outcome.

Tommy is a 13-year-old boy with severe combined-type ADHD and dyslexia who is in your office with his family for a follow-up visit to discuss how his stimulant medication is working.

His parents report that his teachers see a modest improvement in his ability to focus in class and get his class work done, along with less calling out, tapping on the table, and fidgeting in his seat.

By contrast, at home, Tommy is very argumentative about doing his homework and has refused to do it altogether. He is easily angered, has been having difficulty with sleeping, and has been spending less time with friends and family members, preferring to stay in his bedroom playing computer games.

In your office, Tommy confides that he is worried about his ability to handle all the demands of his seventh-grade teachers, many of whom he dislikes. He expresses a great deal of negativity about school and feels convinced that he is "too stupid to make it." Tommy requests that he no longer be required to take medication because he thinks it is not helping him and instead makes him feel "weird and dull."

This article reviews the causes of treatment resistance in young patients with ADHD and describes strategies for integrating pharmacotherapy and psychotherapy to avoid resistance and promote optimal functioning.

The NIMH Multimodal Treatment Study of Children With Attention-Deficit/ Hyperactivity Disorder (MTA study) is the largest and most comprehensive study of the differential effects of 4 treatment strategies after 14 months of intervention on the core symptoms as well as key functional measures of childhood ADHD. The sample of 579 patients (80% males) enrolled in the study included 34% with comorbid anxiety disorder, 40% with oppositional defiant disorder, 14% with conduct disorder, and 11% with a tic disorder. The socioeconomic status (SES) of the study sample was 19% low income, 41% lower-middle income, and 37% upper-middle income. Roughly 30% were living in single parent families, over 90% had parents who graduated from high school, and 84% of fathers and 72% of mothers were employed.

In addition to comparing the efficacy of different treatment strategies (stimulant medication treatment, behavior management, combined treatment, and community care), a major goal of the MTA study was to identify moderators (factors intrinsic to the sample) and mediators (factors pertaining to the intervention process) of treatment outcome.

Owens and colleagues¹⁴ and Hinshaw¹⁵ reviewed the extensive data collected from the study and found significant differences in treatment response based on the presence of several moderators (Table 2). For example, chil-dren with comorbid anxiety disorders showed a stronger response to behavioral intervention (based on parental reports of ADHD and internalizing symptoms) than those without anxiety. Children from low SES families had a better response to combined treatment on teachers' ratings of social skills. Children with more severe initial ADHD symptoms showed a relatively poorer response to both medication management and combined treatment compared with those with less severe symptoms.

Children of parents who reported even mild depressive symptoms showed relatively poorer responses to these treatments, and those with intelligence quotient (IQ) scores of less than 100 who had severe ADHD symptoms also responded less favorably to these interventions. Quite surprisingly, the sex of the child, his or her previous experience with stimulant medication, and the presence of either oppositional defiant disorder or conduct disorder had no discernable impact on treatment response.

With respect to mediator variables, keeping appointments turned out to be a highly significant factor for patients in the medication-management group but not for patients in the behavior-treatment group. Children in the community care group who received medications fared better than those in the community group who were not taking medications, but the children in the community care group were not as improved as the children in the medical center-based medication or combined-treatment groups. Children in the combined-treatment group who showed the greatest improvement in classroom social skills and behavior were from families who displayed the most improvement in discipline methods.

In addition, Owens and colleagues¹⁴ found that only 62% of the patients in the best outcome groups (medication management or combined treatment) had an excellent response (defined as reaching normal or near-normal status) compared with 30% of those who received behavior management or community care. In other words, even in the groups with the most favorable outcomes, more than one third of patients did not achieve an excellent response. This sobering fact is important to keep in mind when considering the prognosis for those who have ADHD.

Among children in the optimal treatment groups, having parents with even mild depression dropped the excellent response rate down to 45%. The investigators identified 5 moderator-defined groups of children receiving either combined or medication-management treatment:

• Group A (low initial ADHD severity, low parental depression) had 73% excellent responders.
• Group B (low initial ADHD severity, high parental depression) had 59% excellent responders.
• Group C (high initial ADHD severity and high IQ, high parental depression) had 48% excellent responders.
• Group D (high initial ADHD severity, low parental depression) had 48% excellent responders.
• Group E (high initial ADHD severity and low IQ, high parental depression) had 10% excellent responders.

The clinical implications are clear: parental depression needs to be addressed when it is present, and when ADHD symptoms are severe, additional treatment approaches (as yet unidentified in published studies) may need to be used in order to maximize the chances for success.

The results of the MTA study raise 2 important questions for clinicians when developing treatment plans for patients with ADHD and comorbidity:

• Is the goal of intervention symptom improvement or remission?
• How can treatment be designed to have maximal effects on the child or adolescent's functioning?

In their meta-analysis, Steele and colleagues¹⁶ proposed that remission be defined as "minimal or no symptoms of ADHD" (using standard rating scales). This would imply that on ADHD-rating measures, the average rating after treatment would be no more than 1 on scales that range from 0 to 3, and a score of no more than 2 on the Clinical Global Impressions Severity scale. The significance of achieving symptom remission is that it is associated with greater improvements in functional outcomes such as academic functioning,^17-19 driving performance,²⁰ and rates of substance abuse.^8,21,22

The MTA study showed that methylphenidate(Drug information on methylphenidate) alone, when titrated to optimal dosages, resulted in a 58% remission rate; when it was given in conjunction with behavior management, 68% of patients achieved remission. Only 34% of children who had received intensive behavior therapy alone (including parent training, summer camp, and classroom aides) achieved symptom remission.

Recently, randomized controlled clin- ical trials of ADHD medications have shown similar outcomes. In a study of 32 patients who received incremental daily doses of OROS methylphenidate (18, 36, and 54 mg), 66% achieved remission at the highest dosage.²³ In a randomized open-label study of 143 patients, 44% achieved remission on OROS methylphenidate (average daily dose, 37.8 mg) versus 16% on immediate-release methylphenidate (average daily dose, 32.2 mg).²⁴ The lower remis- sion rate for OROS methylphenidate may reflect the fact that lower dosages were administered. In a long-term (6-month) open-label trial of mixed amphetamine salts with 138 adolescent patients, remission rates averaged 61% with a range of 52% to 71% depending on the daily dosage given (10 to 40 mg).²⁵ Two randomized controlled trials of atomoxetine(Drug information on atomoxetine) reported remission rates of 28.6% and 27.0%, respectively, indicating a less robust response with this medication than with methylphen- idate and amphetamine.^26,27