It remains unclear whether a clinically easier differential diagnosis in females produced a cleaner proband sample that led to a familial distinction or, conversely, whether a true etiological distinction leads to an easier differential diagnosis in girls than in boys. Clinically, it appears that a more rigorous assessment procedure is indicated whenever features of ODD are present with ADHD and depression, and that differential diagnosis may be more complicated in boys than in girls.
A strong relationship was found between ADHD and depression that was modified by age and comorbidity.23 This relationship was intensified when ODD/CD was present in older, but not younger, children. In general, ADHD appears at a younger age and depression at an older age, and, as expected, the strength of the association between the 2 disorders increases from childhood through adolescence.23,24
The importance of considering moderating variables and ADHD subtype is underscored by our findings (Figure), which show levels of depression as a function of sex, age, and ADHD subtype. The Figure shows that:
- Children and adolescents with ADHD-combined (ADHD-C) subtype generally have higher levels of depression than do youths with ADHD-I.
- Depression increases with age, and the rate of increase is greater in girls than in boys.
- Girls older than 10 years generally have higher levels of depression than boys, whereas the reverse appears true in children younger than 10 years.
This distinction underlies the finding that symptoms of depression are more highly correlated with inattentive symptoms than with hyperactive/impulsive symptoms, yet a diagnosis of ADHD-C in children is more likely to be comorbid with depression than a diagnosis of ADHD-I.
The main limitation of this study is the anomalous finding indicating that girls who are older than 14 years with ADHD-C have a slightly lower rate of depression than boys. Although the data are based on a sample of 920 patients with ADHD, by taking sex, age, and ADHD subtype into account, the results for the oldest girls may have limited reliability because of a rather small sample size.
These considerations help clarify some variable findings in the literature. For example, Biederman and colleagues22 found that boys with ADHD have higher rates of depression than girls, which, in girls older than age 10, is at variance with our findings. However, without understanding the mix of ADHD subtypes and ages of their study samples, it is not possible to directly compare their results with ours. Moreover, because their study was well resourced and in an academic center, it maximized specificity by requiring that 3 elements be satisfied for inclusion in the study:
- The clinical diagnosis.
- Confirmed by a telephone questionnaire with the mother.
- An in-person structured interview.
Because the criteria required multiple screens there may be questions of generalizability. For example, while the ADHD CAS scores for our sample were comparable to those of patients in the MTA, we found higher ODD scores in our sample than those found in the MTA participants.16,25 Perhaps the demanding protocol required for inclusion in the MTA selected a specialized subsample in which the families were more compliant with requests and the children had lower levels of ODD. If so, the degree to which the results of the MTA study could be generalized remains an open question.
Outcomes
Although comorbidity in ADHD generally leads to worse outcome, the impact of comorbid depression remains unclear. While Rostain9 highlights parental depression as a predictor of poor outcome in ADHD in his report on the MTA study, he does not list the child’s depression as a moderator of outcome.
Findings from our clinic show that patients with ADHD-C and comorbid depressive disorders have higher scores on both ADHD-I and ADHD-H scales than patients without comorbid depression. We also found that those without depressive comorbidity at baseline had a higher probability of being relatively free of ADHD symptoms at the 4-month follow-up (P < .05). However, patients without depressive comorbidity had lower ADHD scores to begin with, so they had less “distance” to recovery.
When we controlled for the degree of initial severity of ADHD, there was not a true difference in the degree of relative improvement between patients with and patients without depressive comorbidity at either the 4- or 8-month follow-up. By contrast, high scores on ODD predicted significantly less relative improvement at both the 4- and 8-month follow-ups (P < .05 for each). Moreover, patients with ADHD-C who initially scored high on both ODD and depression scales did worse at 4 months (P = .08) and significantly worse at 8 months (P < .05). These findings suggest that one reason that comorbidity has affected outcome differently in various studies in child and adolescent psychiatry is the failure to account for the full range of comorbid conditions.26-29 In this case, differing results emerge depending on whether ODD was considered.
Treatment issues
Generally, it appears best to treat the depression first, because it is clinically the most limiting condition and depressed patients show worse cognitive impairment.30 Unfortunately, most clinicians treat the ADHD first, usually at the insistence of the parents (and because it is possible to get a quick response). However, the depression may persist, and then one is forced to add to the treatment regimen.
Looking at individual medications may help fine-tune clinical decisions. Bupropion, although not a stimulant per se, may improve both ADHD symptoms and comorbid depression; it is as effective as methylphenidate(Drug information on methylphenidate).31,32 Findling33 reported that patients with ADHD and depression treated with fluoxetine(Drug information on fluoxetine) or sertraline(Drug information on sertraline) monotherapy were less depressed but showed no improvement in ADHD symptoms. Additional treatment with a psychostimulant was necessary to effectively address chronic ADHD. Conversely, the psychostimulants may not provide observable antidepressant effects, for which additional serotonin reuptake inhibitor treatment is required. In a Lilly-funded study of adolescents with ADHD and MDD, atomoxetine(Drug information on atomoxetine) was an effective and safe treatment for ADHD but showed no efficacy in treating MDD.34 Consistent with these findings, the Texas Children’s Medication Algorithm Project protocol for treatment of comorbid ADHD and depression recommends treating the more severe disorder first, and, if the other disorder does not respond, treating it as well.8
In ADHD with bipolar depression, stimulants alone may destabilize bipolar disorder and should be avoided as monotherapy. Using a mood stabilizer, both bipolar and ADHD symptoms may improve, and the cautious addition of a stimulant may further benefit both conditions.35
On the other hand, Goodwin and Jamison36 hypothesize that “. . . many children with purported ADHD develop bipolar disorder in adulthood, raising the possibility that the ADHD-like presentation in childhood may have represented an early manifestation of bipolar illness.” They add, “We would urge caution in the diagnosis and treatment of adult ADHD, always giving preference to initially diagnosing and treating mood disorders until euthymia is achieved before making the ADHD diagnosis or seeking to treat it with stimulants.”
Conclusion
Comorbid depression in patients with ADHD suffers from an “attention deficit” by both researchers and clinicians, compared with other comorbidities (eg, ODD, anxiety). Based on academic studies and data from APS, depression in ADHD appears to be a distinct comorbidity, increasingly prevalent in children as they get older, with a higher rate of increase in girls than boys. By considering moderating variables, our data illustrate why findings in the field are often contradictory to those in academic studies.
Fully accounting for moderating variables is a formidable task; even with our large database, some study cohorts were underpowered. Externalizing and internalizing disorders, as traditionally conceptualized, appear to be overlapping rather than exclusive categories, with anger and acting out (ie, ODD features) cutting across both categories.
Depression alone does not seem to worsen the outcome of ADHD. Although this counterintuitive finding illustrates the controversial nature of the debate about diagnosis of and comorbidity in ADHD, our data also suggest that ODD may worsen the outcome of comorbid ADHD and depression. Should this suggestion be confirmed by further research, it would indicate the need for more aggressive intervention for this nonresponding subgroup.
Our findings highlight the necessity for a careful assessment of children and adolescents with ADHD and depression, with special attention to comorbid ODD as well as other moderating variables. Contradictory find-ings in treatment outcome may result from a failure to assess carefully.
