Comorbid Depression and ADHD in Children and Adolescents

Attention deficit hyperactivity disorder (ADHD), the most common diagnosis in child psychiatry, appears to be more challenging to diagnose and treat when there is a comorbid depressive disorder. In community samples, estimates of the prevalence of depression among patients with ADHD range from 13% to 27%, while clinical sample reports have run as high as 60%.^1-3 Conversely, among children and adolescents with depression, various studies have reported widely varying rates of ADHD (from less than 5% to more than 50%); a recent study in very young children reported a rate of 42%.^4,5 Nevertheless, ADHD and depressive disorders are often difficult to separate in clinical practice.

Depressed children often show more irritability and inattention than sadness, which may lead to a misdiagnosis of ADHD.⁶ Various algorithms and evidence-based treatment recommendations for comorbid ADHD and depression point in somewhat differing directions.^3,7,8 In clinical practice, moderating factors that include age and sex as well as additional comorbid conditions add complexity.

In this article, we review key issues in the diagnosis, treatment, and outcome of comorbid ADHD and depression. We also highlight the importance of moderating variables and consider the validity of the distinction between internalizing and externalizing disorders. Our review of academic studies is complemented with new naturalistic findings from 3419 patients who were routinely assessed at baseline and 4-month follow-up in private practice.

Diagnostic controversies

Some conditions, such as thyroid disease, drug abuse, sleep disorders, learning disabilities, bipolar disorder, or attachment disorders may easily be misdiagnosed as ADHD if not screened for.^6,9 Source bias in the diagnosis of ADHD remains a concern, parent and teacher ratings of schoolchildren are frequently inconsistent, and the validity of the most widely used instrument to diagnose ADHD—the Conners’ Rating Scale–Revised—has been seriously questioned.^3,10 DSM-IV split the ADHD diagnosis presented in DSM-IIIR into inattentive and hyperactive/impulsive subtypes, and more recent research suggests other ways of subtyping ADHD^11-13:

• With and without antisocial disorders.
• ADHD alone, ADHD with oppositional defiant disorder/conduct disorder (ODD/CD), ADHD with anxiety, and ADHD with ODD/CD and anxiety.
• Persistent versus nonpersistent ADHD.

Turning conventional wisdom on its head, research findings suggest that from a familial perspective, the assessment of ADHD may be more valid in adults than in children.¹⁴ Such findings can lead to polarized conclusions, including hypotheses that ADHD is genetically heterogeneous or that ADHD is not a disease but rather a group of symptoms that represents a common behavioral pathway for a range of emotional, psychological, and learning problems.^3,13

Similarly, the diagnosis of depression in children is not without concerns. Historically, clinicians maintained that young children have difficulty naming affect and considered the diagnosis of mood disorders in children controversial.^3,15 However, recent studies show more typical presentations of major depressive disorder (MDD) in children.⁵ In addition, there have been widely discrepant reports of the prevalence of depression among children with ADHD. The Multimodal Treatment Study of Children With ADHD (MTA) found mood disorders in only 4% of study participants, which is far lower than other studies and suggests that instruments used in that study may have been insensitive to depressive symptoms.¹⁶ Carefully controlled academic studies (mostly with boys) suggest that depression and ADHD are indeed distinct entities and not simply the product of overlapping symptoms, rater bias, or demoralization.⁹

Internalizing and externalizing disorders

Although ADHD is often considered an externalizing (overtly disruptive) disorder (as are ODD and CD) and depression is often considered a (nondisruptive) internalizing disorder (as is anxiety), the distinction between these 2 “super categories” is less clear than the names suggest. Externalizing disorders are related to a range of anxiety disorders. Findings from studies of patients with ADHD alone, with ADHD and dysthymia, and with dysthymia alone, indicate that the rates of externalizing symptoms or ODD are not significantly different.^17,18 Other research shows that the association of depression with ODD/CD (one an internalizing and one an externalizing disorder) was almost as strong as that of depression and anxiety (2 internalizing disorders).⁴

Complementing these academic studies, we collected data from our clinic at Alabama Psychiatric Services (APS)—a private practice group with 11 offices, 38 psychiatrists, 45 nurses, and 34 therapists—that routinely integrates baseline and follow-up assessments into its inpatient, outpatient, and partial hospital services to adults, children, and adolescents. Our findings shed light on the relationship between the internalizing and externalizing disorders.¹⁹ We administered the Child and Adolescent Symptom Checklist (CAS) to parents of 3419 outpatients. The CAS is a fully automated, psychometrically sound rating scale (E. J. F., unpublished data, 2005), with 5 subscales that assess for ADHD:

• ADHD-inattentive (ADHD-I).
• ADHD-hyperactive/impulsive (ADHD-H).
• Generalized anxiety disorder.
• ODD.
• MDD.

The items on each subscale parallel the criteria for DSM-IV Category A of the corresponding diagnosis, and responses are on a 4-point scale of frequency. A factor analysis (principal components with varimax rotation) of CAS data reveals that 2 factors account for 77% of the variance.

The correlation between the factor and the underlying variables—factor loading—is shown in the Table. The 5 common childhood disorders may be represented by 2 factors that do not neatly follow the internal/external dichotomy. On factor 1 (a distress factor) anxiety and depression load heavily, and ODD loads moderately; on factor 2 (an attentional factor) ADHD-I and ADHD-H load heavily, and ODD loads moderately. In short, anger and acting out are components of both the distress and attentional factors. These findings align with clinical reality and indicate that irritability is often present in anxiety, depression, and ADHD.

Underlying factors

The relationship between depression and ADHD may vary as a function of moderating variables (sex, age, comorbid ODD), ADHD subtype, and data source (doctor, child, parent, teacher). For example, when we look at our entire sample, depression scores have a significantly higher correlation (P < .001) with ADHD-I than with ADHD-H. However, if we look at these same correlations within age groups (younger than 10 years, 10 to 14 years, older than 14 years) the highly significant differences in the correlations between depression and ADHD subtypes remain only for the youngest group (P < .0001) and there is merely a trend in the 2 older groups. Clinically, this highlights the importance of considering behaviors that are normed by age.

As another example of the power of moderating variables, our data showed that by controlling for ODD symptoms, a clear distinction emerged between depression and ADHD-H symptoms, while the relationship between depression and ADHD-I symptoms was significantly attenuated. This underscores the relative independence of depressive and ADHD syndromes when angry/irritable features are not present, suggested by the factor analysis in the Table. More generally, these results indicate that failing to sufficiently account for moderating variables and ADHD subtype may help explain why different investigators have reached different conclusions about whether the comorbidity between depression and ADHD is an epiphenomenon (accounted for by a third variable).^4,20

The relationship between ADHD and depression symptoms can be refined even further by looking at the results by sex. On the CAS, the relationship between ADHD-I and depression scales was significantly stronger (P < .001) for boys than for girls (whether controlling for ODD or not). When controlling for ODD, the correlation for girls (r = .06) was clinically meaningless, while still statistically significant in our large sample. This finding suggests that when ODD features are absent, depression and ADHD-I are less closely related in girls, and thus more clearly distinguishable than in boys, in whom significant overlap remains and differential diagnosis may be more difficult.

Genetic research suggests that MDD with ADHD is an etiologically distinct subtype in females but not in males.²¹ Biederman and colleagues²² found that girls with ADHD were less likely than boys with ADHD to have MDD. This is an interesting finding because the prevalence of depression in the general population is higher in girls than in boys. These lines of evidence converge to indicate a clearer distinction between ADHD and depression in girls than in boys.