The Comorbidity of Conduct Problems and Depression in Childhood and Adolescence

Clinical Child and Family Psychology Review, Vol. 9, Nos.
3/4, December 2006 (Ó 2006)

DOI: 10.1007/s10567-006-0011-3
The Comorbidity of Conduct Problems and Depression

in Childhood and Adolescence
Jennifer C. Wolff,1 and Thomas H. Ollendick1,2
An extensive body of research documents the high prevalence of comorbidity among child and
adolescent disorders in general and between conduct problems and depression in particular.
These problems co-occur at significantly higher rates than would be expected by chance and
their comorbidity may have significant implications for nosology, treatment, and prognosis.
Four main hypotheses have been put forth to account for these high rates of comorbidity.
First, comorbidity may be a result of shortcomings associated with referral or informant
biases. Second, comorbidity may be an artifact of overlapping definitional criteria. Third, one
disorder may cause the other disorder by influencing the developmental trajectory and placing
an individual at increased risk for further difficulties. Finally, comorbidity between two dis-
orders may be explained by shared underlying causal or risk factors. The purpose of this
review is to explore these possibilities, concentrating primarily on the common risk factors of
parent psychopathology, emotion regulation, and cognitive biases that may underlie the
co-occurrence of these two disorders. Based on our review, we propose a model for the
development of comorbidity between these two disorders.
KEY WORDS: aggression; child; comorbidity; conduct problems; depression
On the surface, the association between conduct sadness as well as loss of energy, apathy, and sleep
problems (including both oppositional defiant disor- problems. Given these symptomatic disparities, it is
der and conduct disorder) and major depressive dis- perplexing that comorbid conduct problems and
order is unexpected given the dissimilarities in depressive disorders occur at significantly higher rates
symptoms characterizing each of the disorders. than would be expected by chance (Capaldi, 1991).
According to the Diagnostic and Statistical Manual The paradox of the apparent distinctiveness of these
of Mental Disorders (DSM-IV; American Psychiatric disorders coupled with their high rates of comorbid-
Association, 2000), symptoms associated with oppo- ity raises the question of how such seemingly distinct
sitional defiant disorder include argumentativeness, disorders come to co-exist in the same individual.
loss of temper, and disregard for authority and The purpose of this review is threefold. First,
symptoms of conduct disorder consist of physical epidemiological data related to rates of overlap
fighting, property destruction, and other delinquent between conduct problems and depression are
acts. Symptoms of depression, on the other hand, are examined. Second, four possible explanations that
thought to relate more to internalized feelings of might help to inform this co-occurrence are discussed
and critically reviewed. Third, based on these find-
ings, a model for the development of comorbidity is
1
presented. The model integrates common causal
Department of Psychology, Virginia Polytechnic Institute and
factors in the development of conduct problems and
State University, Blacksburg, VA, USA.
2
To whom correspondence should be addressed; e-mail: tho@ depression, recognizing their reciprocal relationship.
vt.edu A developmental psychopathology approach serves
201
1096-4037/06/1200-0201/0 Ó 2006 Springer ScienceþBusiness Media, Inc.
202 Wolff and Ollendick
as the framework for this model and various trans- 1.8% of 11-year-olds met criteria for depression. The
actional processes that contribute to comorbidity are expected rate of comorbidity in this sample would be
discussed. In all, the present review provides a syn- the product of 9.2 and 1.8% or, 0.2%. However,
thesis of the comorbidity between these two disorders overall rates of comorbidity observed in this study
and suggests plausible pathways for its occurrence. were between 1.5 and 2.9% (95% confidence inter-
vals). Based on these findings, the observed comor-
BACKGROUND AND SIGNIFICANCE bidity of these presumably distinct child psychiatric
disorders far exceeds that expected by chance alone.
Epidemiology There exist a host of challenges associated with
the identification and comparison of prevalence rates
Studies of comorbidity between depression and across studies, including differences among studies in
conduct problems suggest that each of these disorders terms of diagnostic precision and methodology,
is associated with increased risk for the other. For assessment techniques, the ascertainment method
example, in a meta-analysis completed on child and (e.g., community-based or clinic-referred), and dis-
adolescent community samples, it was reported that tinct characteristics of the samples. The age of the
5.8–14.7% met criteria for oppositional defiant dis- children included in the samples makes comparisons
order or conduct disorder and 1.8–8.0% met criteria particularly difficult as prevalence rates tend to vary
for depression (Angold and Costello, 1993). In ref- with age. Similarly, gender differences complicate the
erence to comorbidity, however, 22.7–83.3% of those comparison of these findings as rates are oftentimes
with depression also met criteria for oppositional found to be different for boys and girls at certain
defiant disorder or conduct disorder, whereas 8.5– points in their development. Although such limita-
45.4% of those with oppositional defiant disorder or tions exist, the ever-expanding database on this topic
conduct disorder also met criteria for depression has made significant advances in examining rates of
(Angold and Costello, 1993). Thus, those who have comorbidity while largely accounting for such
one disorder appear to be at increased risk for the factors.
other.
Among studies of clinic-referred youth, rates of Age of Onset
comorbidity between these two disorders are even
higher. In a recent study conducted by Greene et al. An examination of age of onset of each of these
(2002), more than 30% of clinically referred children disorders, alone and in combination, serves as an
diagnosed with severe major depression also met introduction to the possible relationship between
criteria for conduct problems. Conversely, of those conduct problems and depression. Although
with conduct problems, more than 50% met criteria researchers differ in opinion on the sequence of onset
for major depression. In examining rates of comor- between conduct problems and depression, both
bidity as they relate specifically to conduct disorder conduct disorder and depression tend to increase in
and oppositional defiant disorder, comorbid depres- prevalence during adolescence (Cohen et al., 1993).
sion appears to be more prevalent in those with As will be discussed shortly, some researchers suggest
conduct disorder compared to those with opposi- that conduct problems emerge first and the ensuing
tional defiant disorder. For example, in a study of social failures experienced by these children and
inpatient adolescents, Arredondo and Butler (1994) adolescents contribute to the development of
found that 27% of the adolescents with oppositional depression. Others, however, suggest that depression
defiant disorder (compared to 76% with CD) met the may emerge first followed by the development of
diagnosis of a mood disorder. conduct problems as a means of acting out or
It is clear that the observed rates of comorbidity compensating for some of the internalized feelings
in epidemiological surveys and clinical samples associated with depression.
exceed those expected by chance alone. As described Many clinical and community-based studies
by Caron and Rutter (1991), the expected rate is have found that conduct problems precede rather
obtained by multiplying the base rate of each of the than follow the development of depression. For
separate conditions. Using the epidemiological survey example, Biederman et al. (1995) reported that the
completed by Anderson et al. (1987) as but one age at onset for oppositional defiant disorder and
example, 9.2% of 11-year-olds were diagnosed with conduct disorder was 6–7 years of age, compared to
oppositional defiant disorder or conduct disorder and about 8 years of age for depression. Among
Comorbidity of Conduct Problems and Depression 203
hospitalized girls in another study, the first symptom unlike the ‘‘masked depression’’ phenomenon put
of conduct disorder was reported to occur at 8.2- forth some years ago.
years-old and that of depression at 13.5 years
(Zoccolillo and Rogers, 1991). By contrast, Kovacs Sex
et al. (1988) found that depression was more likely
to precede conduct problems. In their study of 8- to Several studies have found sex differences in
13-year-old children, more depressed children with rates of comorbid conduct problems and depression.
comorbid conduct disorder developed depression In terms of individual disorders, the rate of conduct
prior to conduct disorder (56%) rather than conduct disorder is considerably higher for adolescent males
disorder prior to depression (25%). Thus, it appears than for adolescent females (Robins and Price, 1991),
that conduct problems may precede depression in whereas the prevalence of depression is higher in
some cases but that depression may precede conduct adolescent females (Lewinsohn et al., 1993). In a
problems in other cases. Overall, studies suggest that review of child and adolescent population studies,
the prevalence of comorbidity reaches a peak Zoccolillo (1992) found that the co-occurrence of
around middle adolescence (Beyers and Loeber, these disorders is most likely to occur in boys in
2003; Loeber et al., 1994). This phenomenon may be preadolescence, diminishing in late adolescence and
due to the fact that conduct problems tend to into adulthood, but most likely in girls during mid-
decrease as children move out of adolescence and adolescence into adulthood. Overall, such studies
into adulthood while depression tends to increase as have demonstrated that co-occurring conduct prob-
children move into adolescence and then on into lems are more likely to occur in males than in females
adulthood. although some have found varying rates depending
The timing of the overlap in these problems on the source of the information (Keiley et al., 2003).
tends to show qualitative differences in the types of Interestingly, a greater proportion of girls than
behaviors displayed (Loeber et al., 1994). Specifically, boys tend to exhibit an onset of conduct problems
when depression occurs in early adolescence, boys during adolescence. Although in the majority of cases
with comorbid conduct problems tend to show overt adolescent-onset conduct problems tend to be less
(e.g., bullying, fighting) and covert aggression (e.g., severe, girls displaying symptoms of adolescent-onset
stealing, vandalism, lying), whereas if the overlap conduct problems seem to show a more severe type of
emerges in middle- to late-adolescence behaviors are disturbance that is similar in many respects to boys
characterized as showing more conflict with authority with the childhood-onset pattern (Frick, 1998). Thus,
(e.g., stubbornness, defiance, truancy). Thus, quali- girls with adolescent-onset conduct disorders tend to
tative differences are associated with age of onset of have high rates of neuropsychological dysfunction
comorbidity. and are at high risk for having significantly impaired
Interestingly, there is some evidence that the age adult functioning (Moffit, 1993). Some have specu-
of onset of either comorbid disorder may be later lated that this finding may be linked to the concurrent
than is typical in non-comorbid individuals. For rise of depression in adolescent girls and the delete-
example, Kovacs et al. (1988) found that the average rious effects of comorbid conduct problems and
age of onset for major depression in boys having depression. Along these lines, Loeber and Keenan
comorbid conduct problems was 12.2 years, com- (1994) hypothesized a paradoxical effect of the child’s
pared to 10.8 years for those with no comorbid gender, suggesting that on average the severity of
condition. This finding may suggest that the presence conduct problems in girls is greater than that for boys
of one disorder serves a protective function against when the probability of comorbid depression is taken
the development of the second disorder. For our into account. Hence, somewhat surprisingly, the
purposes here, it may be the case that children with implications of comorbidity may be more profound
conduct problems are able to act out their feelings of in girls than boys.
anger, irritability, or sadness, thus delaying the onset
of full-blown depression. An alternate explanation Implications of Comorbidity
for these findings may be that one disorder masks or
conceals the presence of the other; that is, as children The high prevalence rates of symptom overlap
receive clinical services for one disorder (in this case among children and adolescents with conduct prob-
conduct problems), the other disorder may become lems and depression underscore the importance of
more apparent. This phenomenon would not be examining these comorbid conditions. Quite
obviously, the issue of comorbidity raises questions to receive treatment, had poorer global functioning,
concerning the nosology of disorders. Whether con- and were more likely to attempt suicide. Thus, com-
duct problems with and without comorbid depression orbidity signaled a poorer clinical profile.
(or vice versa) should be considered to be two distinct Additionally, comorbidity may have implica-
disorders or as different manifestations of the same tions for the course and treatment of psychopa-
condition has important implications for better thology. To date, few studies have examined the
understanding etiology, disease progression, and treatment of youth with comorbid conduct prob-
treatment response (Biederman et al., 1995). The lems and depression (see Rapp and Wodarski,
identification of two distinct but co-existing condi- 1997; Rohde et al., 2004). Results of the few
tions is important so that each disorder can in turn be medication trials that have examined efficacy in
treated. On the other hand, different treatment comorbid samples indicate that antidepressant
approaches may not be necessary if the symptoms medications may be efficacious for adolescents with
represent the same underlying processes (Biederman comorbid depression. Still, those depressed adoles-
et al., 1995). cents with comorbid conduct problems tend to be
Moreover, this co-occurrence carries important less responsive to medication than those having
ramifications in terms of the degree of impairment other co-occurring disorders (Hughes et al., 1990).
and the course of psychopathology for these children In another study investigating the efficacy of the
and adolescents. Specifically, those with comorbid Adolescent Coping with Depression Course (CWD-
conditions are typically more severely impaired than A), a cognitive behavioral group intervention for
children with either disorder alone (Nottelmann and adolescent depression, those with comorbid conduct
Jensen, 1994). In combination, these problems may disorder had significantly higher rates of depression
magnify or exacerbate the deleterious effects of either recurrence over a 2-year follow-up period (Rohde
disorder alone and contribute to high levels of psy- et al., 2001). In sum, the few studies that have
chosocial impairment in various areas of functioning. accounted for comorbidity of conduct problems
For example, research demonstrates that risk for and depression in treatment efficacy studies dem-
suicidality increases in depressed adolescents as a onstrate poorer outcomes in those adolescents dis-
function of pre-existing conduct problems (Capaldi, playing co-occurring conduct problems and
1991, 1992). Indeed, a number of studies have dem- depression.
onstrated that comorbid emotional symptoms and Overall, these findings illustrate the potential
conduct problems are common among children and implications of comorbidity in terms of nosology,
adolescents who commit suicide. Moreover, this risk degree of impairment, prognosis, and treatment.
appears to be higher for girls than boys (Loeber and While the implications of any single disorder can be
Keenan, 1994). Longitudinally, youths with both profound, the co-occurrence of a second disorder
conduct problems and depression are also prone to appears to compound and exacerbate the deleterious
increased long-term problems in functioning effects of the individual disorder.
(including adult criminality), compared with de-
pressed-only youth (Capaldi, 1992; Harrington et al., OVERVIEW OF PATHWAYS AND
1991; Kovacs et al., 1988). CONDITIONS LEADING TO COMORBIDITY
In one study, Lewinsohn et al. (1995) examined
the consequences of comorbidity in a community Despite the importance of research in this area,
sample of 1,507 adolescents. Participants with ‘‘pure’’ studies of comorbidity have largely failed to consider
and comorbid disorders were compared on six clinical pathways to the development of co-occurring disor-
outcome measures. When comparing those diagnosed ders and the conditions under which comorbidity
with major depressive disorder only to those with a occurs or fails to occur. Moreover, these studies
major depressive disorder and comorbid disruptive rarely consider the reasons these dual diagnoses occur
behavior, those with the comorbid disorder were with such high frequencies. Thus, questions remain as
significantly more likely to receive treatment, had to the origins of comorbidity and the mechanisms
poorer global functioning, and had more academic that underlie this phenomenon. While at present
problems. Similarly, when examining those with a there is not sufficient evidence to fully answer these
disruptive behavior disorder alone compared to those questions, the extant literature lends itself to specu-
with a behavior disorder and comorbid depression, lation on the onset and causes of developing comor-
those with the comorbid condition were more likely bid psychopathologies.
Certainly, without due consideration of particu- likelihood of referral for either disorder separately
lar mechanisms in the development of comorbid (Caron and Rutter, 1991).
diagnoses, little can be done to effectively prevent the In addition, comorbidity diagnosed in clinic
development of further psychopathology. With this settings may simply reflect more extensive psycho-
problem in mind, the literature suggests four possible pathology in those referred to such settings as com-
explanations for comorbidity (Angold et al., 1999; pared to those in the general population (Nottelmann
Caron and Rutter, 1991; Seligman and Ollendick, and Jensen, 1995). For example, clinic-referred indi-
1998). First, it may be that comorbidity is a meth- viduals may have more severe symptomatology, be
odological artifact based on shortcomings associated more impaired, and come from families that generally
with referral biases and the use of multiple infor- feel more overwhelmed by their children’s problems,
mants. Second, although conduct problems and as compared to those who do not seek treatment for
depression are believed to be two distinct constructs, psychological symptoms (Angold et al., 1999).
the extent of overlapping definitional criteria may Another potential source of methodological
produce inflated rates of comorbidity. Third, one error is clinician bias. Previous work has related rates
disorder may cause or put an individual at increased of comorbidity to clinical interviewers who might be
risk for the second disorder. Fourth, two disorders more attuned to the possibility of multiple disorders
can be comorbid because they share underlying risk (Angold and Costello, 1993). Moreover, it was sug-
factors. It is important to note that the first two gested that parents, teachers, and children may be
explanations suggest that comorbidity is the result of similarly sensitized in that reporting one problem
methodological flaws, whereas the second two may make them more sensitive to other problems and
explanations suggest that comorbidity is due to the hence more likely to report another disorder. How-
nature of the two disorders. Each of these explana- ever, this possibility has received little support;
tions will be explored briefly as they relate to the especially with the widespread use of structured
relationship between conduct problems and depres- interviews. In fact, it is hypothesized that the opposite
sion in children and adolescents. While an exhaustive effect may be true – that comorbidity may actually be
review of this vast literature was not undertaken, the underreported in instances that clinical interviewers
present review is illustrative of the findings. are not open to the possibility of alternate explana-
tions for emerging symptoms (e.g., clinicians assume
METHODOLOGICAL ARTIFACT? that internalizing and externalizing disorders cannot
co-exist).
Some have suggested that comorbidity is a The increased use of multiple informants has
methodological artifact rather than a real psycho- also raised questions regarding the legitimacy of
pathological phenomenon (see Angold et al., 1999). comorbidity. In particular, reliance on multiple
In particular, those who question high rates of com- informants frequently results in varying opinions and
orbidity have suggested that the effects of referral interpretations of problem behaviors (Youngstrom
bias, clinician bias, and the use of multiple infor- et al., 2003). Combining information from various
mants, may inflate rates of co-occurrence among sources often requires employing some rule for
clinical disorders. making decisions. Typically, this means using the ‘‘or
As was noted earlier, past studies of comorbidity rule’’ that if either the parent or child endorses the
were often based on clinic samples. Since they were disorder, it is considered present. However, this rule
not population-based surveys, these results were raises the possibility that different informants may
likely to be subject to referral biases and Berkson’s have different interpretations of the same behavior or
bias, producing higher comorbidity rates in the clinic may view different behaviors as being typical of the
population than those in the general population child based on the environment where they observe
(Angold et al., 1999). According to Berkson’s bias, the child (e.g., home, school, with peers). For exam-
whenever less than all subjects with a particular dis- ple, a parent might view the behavior as oppositional,
order are referred, selected samples will contain a whereas the child might indicate feeling depressed,
disproportionately large proportion of patients leading to disparity in diagnostic labels (and
showing multiple problems and probable comorbid- increased likelihood of comorbidity). In one recent
ity. As illustrated earlier, this bias is presumably study, different decision rules resulted in the classifi-
because the referral likelihood for subjects with more cation of anywhere from 5 to 74% of the participants
than one disorder will be a function of the combined classified as having comorbid internalizing and
externalizing symptoms (Youngstrom et al., 2003). system (Angold et al., 1999). In other words, these
Such difficulties with reconciling discrepancies disorders are not actually comorbid at greater than
between reporters may lead to inflated rates of chance levels, but they appear to be so because of
comorbidity (Jensen et al., 1999). lack of specificity in the diagnostic nomenclature.
However, Angold et al. (1999) proposed several To examine this possibility, it is first necessary to
reasons that the use of multiple informants alone examine DSM-IV criteria for different disorders and
does not explain the high rates of comorbidity. First, compare the specific symptoms associated with each
Angold and colleagues highlight the finding that disorder. In comparing conduct disorder and
adult studies that frequently rely on single reporters depression, the criteria for these disorders do not
(primarily self-report) also find high rates of comor- seem to overlap, at least on the surface. On the other
bidity. Second, those child studies using only self- hand, it is known that irritability is common in
report have reported similar rates of comorbidity as childhood depressive disorders and that several
those using multiple informants. Third, in many symptoms of oppositional defiant disorder appear to
studies comparing differences in the number of be related to irritability (e.g., often loses temper).
comorbid diagnoses reported by parents, children, or Irritability, frustration, and angry outbursts are ei-
both informants, no significant differences were ther primary or secondary features of ODD (APA,
found (e.g., Jensen et al., 1995, 1999), which stand in 2000). Furthermore, a tendency to display intense
contrast to the findings of Youngstrom and col- and negative emotions is an important aspect of the
leagues (2003). Together, these findings discredit the ‘‘difficult temperament’’ that has been linked to the
argument that comorbidity is the result of the use of development of both conduct problems and depres-
multiple informants. In fact, estimates of co-occur- sion. In effect, the presence of one disorder partially
rence may be more accurate when based on multiple fulfills the criteria for a second disorder, and thus, in
sources of information, because studies have sug- operational terms, implies the existence of the second
gested that a single reporter may be biased and that disorder. If such an effect was the sole cause of the
multiple informants yield a more complete ‘‘picture’’ apparent comorbidity between the two disorders,
of the child (Ollendick and Hersen, 1993). then the coexistence of the two disorders would be
In summary, it appears that comorbidity is not present only in those manifesting the shared symp-
the result of some methodological flaw. Increased use toms and not in those whose disorders did not
of community-based samples and multiple infor- involve such symptoms (Angold and Costello, 1993).
mants has primarily ruled out the possibility that Along these lines, one study controlled for these
systematic errors in studies of comorbidity are overlapping symptoms (using subtraction and pro-
responsible for the high prevalence rates observed in portion methods) in 341 clinically referred children
numerous studies. Such findings suggest that com- and then reevaluated each child’s diagnosis (Bieder-
orbidity is a ‘‘real’’ phenomenon and an area in need man et al., 1995). Eliminating these overlapping
of further research. symptoms (‘‘angry mood’’ for depression) failed to
reduce elevated rates of comorbidity between these
FLAWS IN CURRENT NOSOLOGIES? problems. Such findings suggest that comorbidity is
not just an artifact of overlapping diagnostic criteria.
It is possible that the co-occurrence of these Although some minor overlap may be perceived, the
problems is so common that the distinctiveness of the vast majority of the criteria offer an acceptable level
purportedly separate diagnostic entities should be of specificity to distinguish the disorders. The diag-
questioned. In the past, some have suggested that noses for each of the disorders require multiple
comorbidity arises because these conditions reflect a symptoms, making this explanation an unlikely can-
general syndrome reflecting the individual’s general- didate for explaining the co-occurrence of disorders
ized vulnerability to psychiatric disorder that is (Zoccolillo, 1992). Moreover, if poorly delineated
manifest in comorbidity between different types of criteria were an explanation, then comorbidity should
the same disorder (Zoccolillo, 1992). Conversely, decrease as the number of conduct disorder symp-
several authors have pointed out that comorbidity toms increases, in fact, the opposite occurs (Robins
could be due to the fact that individual ‘‘non-specific’’ and Price, 1991). That is, the more symptoms of
symptoms may be shared by disparate diagnoses conduct disorder, the less likely the occurrence of
(e.g., Caron and Rutter, 1991), with the result that a depression. In fact, the opposite pattern is obtained.
certain amount of overlap is built into the diagnostic This finding suggests that comorbidity is a real
phenomenon rather than a methodological or It has been theorized that the progression of
definitional flaw. conduct problems into depression is related to the
In sum, definitional or criteria-based flaws are chain reaction of developmental failures experienced
not a viable explanation for the co-existence of by youth with conduct problems (Capaldi 1991, 1992;
conduct problems and depression. In general, the Capaldi and Stoolmiller, 1999; Patterson et al., 1992).
literature suggests that conduct problems and Specifically, the combination of lack of competence
depression are distinct entities. Moreover, similarities and negative reactions from others may result in
in diagnostic criteria do not appear to account for pervasive failures in adaptation that subsequently
high rates of comorbidity. Put simply, empirical evi- contributes to a child’s vulnerability to the onset of
dence demonstrates that symptom overlap and depressive symptoms (Capaldi and Stoolmiller,
developmental explanations do not account for high 1999). For example, higher levels of conduct prob-
rates of comorbidity. lems are associated with peer rejection, problems in
the parent–child relationship, and failure to develop
DOES THE COURSE OF ONE DISORDER academic skills (e.g., Capaldi, 1991). Such conse-
AFFECT THE OCCURRENCE OF THE OTHER quences often affect social development and con-
DISORDER? tribute to associations with deviant peers, negative
attitudes, and may in turn contribute to depressed
A third possible explanation for the overlap of mood.
conduct problems and depression is that comorbidity In a longitudinal study, Capaldi (1992) found
occurs because one disorder causes or puts an indi- that boys with high conduct problems at Grade 6
vidual at risk for the other. Three such types of reported significantly higher depressed mood
association may occur between these disorders (Fer- symptoms at Grade 8. Some have hypothesized that
gusson et al., 1996). First, conduct problems may be a one cause of this phenomenon may be that the
direct cause of affective disorder (e.g., Capaldi 1991, social consequences of antisocial behavior may in-
1992; Capaldi and Stoolmiller, 1999). Second, crease with age. Social consequences in adolescence
depression may be a direct cause of conduct disorder (e.g., poor family relations, academic under-
(Kovacs et al., 1988; Puig-Antich, 1982). Third, achievement) may become more costly for the
conduct problems and depression may be reciprocally individual as they enter a stage when peer rela-
related so that they influence one another in a tionships become increasingly important. However,
simultaneous and transactive fashion. In other words, controlling for prior depressed mood, conduct
while conduct problems increase the risk for depres- problems at Grade 8 did not predict higher levels of
sion, depression in turn, leads to symptoms of con- depressed mood at Grade 12 (Capaldi and
duct problems. As suggested by Seligman and Stoolmiller, 1999). Beyers and Loeber (2003) sug-
Ollendick (1998), for any of these direct causal gest that this pattern of results may indicate that
assumptions to be true, it must be shown that one the effect of conduct problems on later depressed
disorder temporally precedes the other and puts an mood may be limited to early adolescence or select
individual at elevated risk for the development of the transition periods.
second disorder.
Depression Preceding Conduct Problems
Conduct Problems Preceding Depression
However, a subset of studies has found the
As mentioned previously, the age of onset for reverse to be true. At the forefront of this alternate
conduct problems typically precedes the age of onset position is the work of Kovacs et al. (1988) who
for depression (Biederman et al., 1995; Nock et al., found that in the majority of cases involving
2006; Zoccolillo and Rogers, 1991). In fact, the comorbid depression and conduct disorder, depres-
National Comorbidity Survey Replication study sion was diagnosed first. This early study followed
found that conduct disorder precedes depression in depressed children ages 8–13 over several years. Over
72% of cases (Nock et al., 2006). Moreover, those the course of the study, 23% developed conduct
with both active and remitted conduct disorder had disorder. Of these comorbid cases, 56% of partici-
significantly higher risk for developing depression pants developed depression prior to conduct disorder
later in life suggesting that conduct problems may be while 25% developed conduct disorder prior to
a marker of risk for subsequent psychopathology. depression. While these results demonstrate that, at
times, depression precedes conduct problems, the depression to conduct problems as the literature in
sample used for the study was small and does not this area is limited.
permit ready generalizations. Moreover, the sample
was selected based on a diagnosis of depression, Suggesting Common Ground
which may have increased the likelihood of reporting
a subsequent diagnosis of conduct problems. If the More recent studies have proposed a dynamic
reverse had been true (i.e., the sample was selected link between conduct problems and depression that
based on diagnosis of conduct problems) they may occurs concurrently and prospectively (Lahey et al.,
have been more likely to find subsequent diagnoses of 2002). In examining 7- to 12-year old boys (at Wave
depression. Hence, limitations of this study render the 1) over a 7-year period, wave-to-wave changes were
results difficult to interpret. paralleled by correlated changes in the number of
Another way to examine the progression from symptoms of depression. In other words, higher levels
one disorder to another is to look at the effects of of CD in Wave 1 were accompanied by higher aver-
treatment for one disorder on the other disorder. For age levels of depression symptoms in Wave 1.
example, in one study, Puig-Antich (1982) found that Moreover, these symptoms remained relatively stable
clinic-referred children who met criteria for major across waves two to seven. Thus, the findings suggest
depression often exhibited comorbid conduct prob- that when depression increases from one wave to the
lems. Moreover, symptoms of conduct problems im- next, it is accompanied by concurrent increases in
proved when depression remitted. The study initially conduct problems and vice versa.
sought to examine the effectiveness of imipramine in In a similar longitudinal investigation, Beyers
pre-pubertal major depressive disorder. Clinical and Loeber (2003) examined relations between
observations during the 5-week imipramine trial and depressed mood and delinquency while controlling
during follow-up thereafter indicated that successful for common risk factors. Results indicated that
outcome of participants’ mood disorder was followed depressed mood predicted concurrent delinquent
by alleviation of conduct disorder behaviors in the acts, while at the same time delinquent acts predicted
majority of the participants. However, these results depressed mood. However, the relationship impli-
should also be interpreted cautiously since no pure cated was not a symmetrical one. In fact, depressed
conduct problem control group was included, and mood had a more robust effect on delinquency than
because contrary results have been reported else- delinquency had on depression. Hence, these findings
where, albeit in somewhat older samples (Harrington suggest that while one disorder may predict the other,
et al., 1991). depression may have a stronger effect on conduct
It should also be noted that Patterson et al.’s problems such that delinquent acts increase at a
(1992) model further predicts that as children progress higher than would be expected rate. While such
to later stages, the depressed mood resulting from studies illustrate the relationship between conduct
earlier antisocial behavior and its negative conse- problems and depression in boys, there remains a
quences is expected to precede more delinquent lack of research on this issue in girls. Thus, little is
behaviors. For instance, it has been found that known about the dynamics of these disorders as they
depressed mood in boys is associated with escalation to relate to females.
more serious and varied kinds of delinquent behavior Finally, the work of Fergusson et al. (1996)
(Beyers and Loeber, 2003; Loeber et al., 1994). Basi- offers an alternate explanation for the relationship
cally, a transactional model is hypothesized in which between these dimensions. Specifically, he posits that
conduct problems give rise to depression which, in the relationship between these disorders relates to
turn, contributes to more delinquent acts. risk factors common to both disorders. Structural
Alternately, it has been suggested that youth equation modeling suggested that a substantial
with depression may develop conduct problems as the component of the comorbidity between conduct and
irritability associated with depression becomes more affective disorders arose because the risk factors
severe. Difficulties regulating the irritability and associated with the development of conduct problems
negative affect associated with depression may con- overlapped with the risk factors for adolescent
tribute to increased conflict with others, opposition- affective disorders. In particular, of the shared vari-
ality, and subsequent acting out behaviors. Such an ance between conduct disorder and affective disor-
explanation warrants further study as do additional ders, more than two thirds was accounted for by
mechanisms that may explain the progress from common risk factors. Accordingly, results suggested
that the link between the two disorders was not a relationship between conduct problems and depres-
direct relationship, but an indirect one accounted for sion may be explained by their common risk factors,
by other factors. as will be discussed next. Research on the mecha-
It should be noted that research on the temporal nisms that may explain the path from one disorder to
ordering of these problems is limited and several another is limited and warrants further investigation.
factors have been neglected. First, research is hin-
dered by the lack of longitudinal investigations of this COMMON RISKS?
topic. Even those longitudinal studies that have
explored this topic have tended to cover a relatively The final explanation for high rates of comor-
brief period of time. Second, differences in the bidity between conduct problems and depression
sequalae of disorders for males and females have not involves the presence of common risk factors. Ques-
been adequately examined. Third, several studies tions of vulnerabilities, risk factors, and familial
have focussed on the onset of disorders without (genetic and environmental) transmission that in-
accounting for the emergence of early symptomotol- volve adolescent characteristics and the familial and
ogy. For example, a depressed child may begin to extra-familial context in which development takes
display temper outbursts, but may not meet full cri- place remain at the forefront of research on comor-
teria for oppositional defiant disorder. While the bidity. As Caron and Rutter (1991) noted some years
onset of a disorder serves as one marker of comor- ago, antecedent causal factors may lead to comor-
bidity, it may be useful for clinical purposes to con- bidity in two ways. First, the risk factors for one
sider co-occurring symptoms that may subsequently disorder may be the same as the risk factors for the
develop into a second disorder. Finally, research in other disorder. Second, even though two disorders
this area has failed to identify various mechanisms for may have apparently different risk factors, these risk
risk as well as moderators in this relationship. For factors may correlate with each other. In both in-
example, severity of the initial disorder may be a stances, the comorbidity between disorders arises
primary predictor of which individuals subsequently because the risk factors and life pathways that lead to
develop a second disorder. Along these lines, Nock one disorder overlap or correlate with the risk factors
et al. (2006) found that conduct-disordered individ- and life pathways that lead to the other disorder.
uals with increased symptom number and severity In general, research has shown support for the
were more likely to develop a second disorder. Fur- inter-correlation and overlap of risk factors between
ther elucidation of such factors is needed. conduct problems and depression (e.g., Fergusson
An alternate explanation for these findings may et al., 1996). Such findings suggest that the risk
be that because conduct problems have extremely factors and life pathways that lead to increased risk
deleterious effects on interactions between children of conduct disorder have much in common with the
and their adult caretakers, parents and teachers may risk factors and life pathways that contribute to
be more likely to recognize symptoms of conduct depression. There are many levels of risk factors,
problems as opposed to depression. Thus, children including child, family, school, peer, neighborhood,
with conduct problems may be more likely to present and culture, which may contribute to comorbidity.
to clinic settings for this reason, and symptoms of Moreover, a number of common genetic, biological,
co-occurring depression may be overlooked by parent physiological, psychological, environmental, and so-
or teacher raters. Accordingly, it may be that cial factors may be implicated in the development of
depression precedes conduct problems in more cases comorbidity. While a plethora of possibilities exists,
than the previously reviewed studies suggest. the present review selectively examined the common
In brief, however, findings demonstrate that in risk factors of parental psychopathology, emotion
the majority of cases depressive symptoms tend to regulation, and emerging cognitive biases and sought
emerge after the onset of conduct problems. In a to find common and unique aspects of this risk; as
smaller proportion of cases, depression seems to such, it was not intended to be an exhaustive review
occur earlier than the onset of conduct problems, of all of the aforementioned possibilities. These risk
although these findings may be tainted by referral factors were chosen because they illustrate both
biases. Moreover, studies suggest a reciprocal rela- shared and unique sources of overlap in the symp-
tionship, as a rise in one disorder is often associated toms associated with conduct problems and depres-
with an increase in the other. Still, based on the work sion. However, it is also important to keep in mind
of Ferguson et al. (1996), it appears that much of the that despite the presence of various risk factors, many
individuals demonstrate great resilience and do not depression is related to their mother’s experiencing
go on to display comorbid psychopathology. more depression. In addition to this risk factor for
depression, other studies have implicated the role of
Parental Psychopathology parental alcoholism (e.g., Puig-Antich, 1989). Thus,
children of depressed or substance abusing parents
Previous research has indicated that children of are at increased risk for depression.
parents with various forms of psychopathology are at A few studies have examined the relationship
increased risk for many psychological problems between comorbid conduct disorder and depression
including conduct problems and depression. Since a in youth and psychopathology among family mem-
complete discussion is beyond the scope of this paper, bers. For instance, Puig-Antich (1989) studied the
only the primary conclusions that can be drawn from family history of 95 prepubertal probands (aged 6–
this literature will be presented. 12 years). Results showed that depression in the
For parents of children with conduct problems, probands correlated with high family prevalence of
high rates of parental depression, antisocial person- depression and alcoholism. However, results further
ality disorder, and substance abuse have consistently indicated that when family members of depressed
been reported (e.g., Frick et al., 1992). Depression, children had low rates of depression, the children
particularly in mothers, has been linked to child were more likely to exhibit comorbid conduct disor-
conduct problems in several studies (e.g., Phares and der (in comparison to those with high family rates of
Compas, 1992). In general, depressed mothers are depression). This finding suggests that when family
more negative in their interactions with children than members have not experienced depression, these
non-depressed mothers. They appear to use more children may be more likely to act out their inter-
physical punishment, are more verbally aversive, nalized feelings. In addition, Williamson et al. (1995)
monitor and supervise their children’s activities less found that the rate of antisocial personality disorder
effectively, engage in fewer affectionate interactions among relatives of youth with depression and con-
with their children, and respond to their children with duct disorder was elevated compared with the rate
less warmth (see McMahon and Wells, 1998). Such among relatives of youth with depression only. Thus,
parenting practices may in turn contribute to coercive comorbid children are more likely to have family
interactions between the parent and child and result members with antisocial personality disorder.
in more opportunities for the child to engage in more Although family factors are often considered to
conduct problems behaviors. High rates of parental be primarily psychosocial in nature, such variables
antisocial personality disorder have also been contain both genetic and environmental influences.
reported for parents of children referred for conduct For example, O’Connor et al. (1998) demonstrated
problems (e.g., Lahey et al., 1988). Although some that genetic factors accounted for approximately half
studies have found an association between maternal of the variance in the overlap of depression and
antisocial behavior and child conduct problems, antisocial symptoms. In terms of environmental
research has consistently shown that the association influences, research has also linked parent depression
is stronger between father and child (e.g., Frick et al., to lower levels of nurturance and affection as well as
1992). Additional studies have also pointed to greater control, hostility, and conflict (see Jewell and
increased risk of conduct problems associated with Stark, 2003 for review). Although space consider-
parental alcoholism and other substance use (e.g., ations preclude a thorough presentation of such
Loeber et al., 1995). mechanisms, the role of such processes should be
Similarly, estimates of the incidence of depres- recognized.
sion in the children of depressed parents range from Unfortunately, the majority of studies have ten-
14% to almost 50% (see Trad, 1987). In fact, several ded to examine the role of parent psychopathology in
studies have found that depression in one member of isolation. This decision presents several problems.
a family magnifies the risk of depression twofold First, when family characteristics are examined, they
(e.g., Williamson et al., 1995). As reviewed by Stark usually focus on single childhood conditions, either by
et al. (2000), the offspring of depressed parents are excluding potential participants who have other forms
more likely to develop depression if the parent’s of psychopathology or by ignoring potential
depression had an early onset, has been recurrent, or co-occurring conditions (Marmorstein and Iancono,
if the parent was repeatedly hospitalized. Hammen 2004). Thus, little is known about the role of parental
(1991) has also noted that the severity of a child’s psychopathology as it relates to co-occurring disorders
in their offspring. Second, looking at these factors in high, moderately low), Eisenberg et al. (2000)
isolation does not allow for the examination of how hypothesized that the interaction of these variables
risk factors operate together in placing a child at risk would result in differences in adjustment and quality
for psychopathology. Third, because family risk fac- of social behavior. Results indicated that negative
tors are often interconnected with one another, their emotionality was a general risk factor related to both
independent examination is probably confounded by internalizing and externalizing disorders. In addition,
these overlapping variables (Frick et al., 1992). In regulatory undercontrol was a predictor of external-
other words, although factors such as parental psy- izing behavior problems, whereas regulatory over-
chopathology may explain part of the variance in the control (e.g., behavioral inhibition) predicted
development of child psychopathology, they do little internalizing problems.
to explain the mechanisms of change underlying their In another study, Eisenberg et al. (2001) exam-
development (e.g., dynamics in the parent–child ined emotion regulation in children with internalizing
relationship, parenting techniques). and externalizing problem behaviors. Based on parent
Overall, those studies that have looked at com- and teacher reports as well as observations during
orbidity as it relates to parental psychopathology several behavioral tasks, various similarities and dif-
point to the common risks of parental depression ferences emerged between groups. In general, children
(particularly in the mother) and substance abuse with externalizing problems, compared with children
(primarily in the father) and the distinct risk factor of with internalizing problems and non-disordered chil-
paternal antisocial personality disorder for conduct dren were more prone to anger, impulsivity, and low
problems. Risks for comorbidity appear to be similar regulation. Children with internalizing symptoms
to those with conduct problems with at least one were prone to sadness, low impulsivity, and were more
distinction. Specifically, research has demonstrated regulated on measures of attentional and inhibitory
that when a ‘‘mismatch’’ occurs between parent and control. Those children that were comorbid in the
child where the child demonstrates depression, but study showed a similar presentation as those with
the parent does not, the child may be at increased risk externalizing behaviors. Specifically, comorbid chil-
for the development of comorbid conduct problems dren were higher on anger and impulsivity, but lower
and depression. Thus, it may be that families showing on attentional regulation and inhibitory control.
goodness-of-fit between parent and child psychopa- It is important to recognize that emotion regu-
thology are less prone to having a child with a co- lation involves both physiological and behavioral
morbid presentation. Such a possibility remains to be components, both of which are expected to influence
empirically verified however. pathways to child psychopathology. It has been
argued, for example (see Eisenberg et al., 2001), that
Emotion Regulation the Behavioral Inhibition System (BIS) and the
Behavioral Activation System (BAS), with connec-
Emotion regulation deficits also have ties to both tions to the limbic system, may underlie high and low
conduct problems and depression. In discussing this levels of regulatory control. For example, BIS has
risk factor, it is necessary to first define emotion been implicated in behaviors related to overcontrol
regulation. For purposes of this review, emotion (e.g., social withdrawal), whereas low BIS and/or
regulation is conceptualized as the internal and high BAS are associated with more externalizing
external processes involved in initiating, maintaining, behaviors including impulsivity (Gray, 1987). Others
and modulating the occurrence, intensity, and have pointed to regulatory differences related to right
expression of emotions to accomplish one’s goals (see and/or left frontal asymmetry, although findings are
Eisenberg et al., 1997; Grolnick et al., 1996). Based somewhat contradictory at this time (e.g., Forbes
on the work of Block and Block (1980), it has been et al., 2006).
argued that children with externalizing problems are Along these lines, Cole et al. (1996) examined
undercontrolled in their expression of affect, whereas expressive and physiological aspects of emotion reg-
those with internalizing problems are overly con- ulation during a negative mood induction in pre-
trolled or constrained in their expression of emotions. schoolers with varying degrees of behavior problems.
For example, in contrasting those with various Findings indicated that at follow-up, inexpressive
styles of emotion regulation (i.e., highly inhibited, children had higher rates of both externalizing and
optimal regulation, undercontrolled) to those with internalizing disorders while expressive children had
varying levels of negative emotionality (i.e., moderately higher rates of externalizing disorders. Moreover,
inexpressive children had the highest heart rate, and depression. Specifically, the role of social infor-
lowest vagal tone, and smallest autonomic nervous mation processing and schemata in aggressive and
system (ANS) change during negative mood induc- depressive youth represents a general common risk
tion. Highly expressive children had the slowest heart factor between these two groups. While there are
rate, highest vagal tone, and largest ANS change. many similarities in the type of distortions and defi-
Such physiological differences in the two groups may cits displayed by these children, there are also distinct
suggest different emotion regulation strategies differences in the processing patterns and basic
between the groups. For example, inexpressive chil- schemata of those with each type of psychopathol-
dren may not show external reactivity (e.g., show ogy.
facial response), but they may focus internally on Research exploring various knowledge struc-
distress. Based on these results, the authors suggest tures, or schemata, tend to differentiate these two
that inexpressive children with disruptive behavior groups of children. In general, aggressive children are
may be at increased risk for subsequently developing thought to possess inflated self-concepts, underesti-
comorbid depression. However, because the study did mate their social rejection, and show a hostile attri-
not follow these children past first grade, this possi- bution bias (see Rudolph and Clark, 2001 for review).
bility was not fully explored. Conversely, the negatively biased schemata of
It is also necessary to consider the means by depressed children contribute to negative views of the
which emotion regulation might lead to emotional or self, current circumstances, and the future (Beck,
behavioral problems. In some cases, problems in 1967). In particular, according to learned helplessness
emotion regulation (e.g., over/undercontrol) may theories of depression (Abramson et al., 1978),
directly influence the child’s propensity to display depressed individuals tend to attribute the causes of
conduct problems or depression (Frick and Morris, negative events to stable, global, and internal factors.
2004). For example, a child with low regulatory In comparing the conceptions of self and peers in
abilities may be quick to show anger and act youth with aggression, depression, or a combination
aggressively or a child who is overcontrolled may be of the two, Rudolph and Clark (2001) demonstrated
quick to internalize feelings of sadness and to with- that children in the depressed and depressed-aggres-
draw. However, emotion regulation may also affect sive groups demonstrated more negative conceptions
the development of these behaviors through indirect of both self and peers than did non-symptomatic and
means (Frick and Morris, 2004). Mechanisms such as aggressive children. Aggressive children, on the other
parenting style, social cognitive skills, and peer rela- hand, demonstrated a self-enhancement bias. Thus,
tions may mediate the relationship between emotion depressed and comorbid individuals appear to hold
regulation and the development of these disorders. negative views of the self, whereas aggressive indi-
Finally, it should be recognized the majority of viduals show more inflated ideas of the self.
research in this area discusses internalizing/external- In addressing the role of social information
izing disorders rather than conduct problems and processing, Dodge’s model of the steps children
depression in particular. Therefore, the present dis- engage in before enacting social behaviors provides a
cussion is somewhat speculative as it is not known if framework for discussion. As reviewed by Dodge
these descriptions apply to the primary disorders in (1993), this dynamic model of steps includes encoding
question. Although those disorders that comprise and interpreting social stimuli as well as generating
internalizing disorders (anxiety and depression) and and evaluating responses before enacting a behavior.
externalizing disorders (attention-deficit/hyperactivity Deficits and distortions at any one of these steps may
disorder (ADHD), oppositional defiant disorder, and independently contribute to the development of
conduct disorder) bare many relations to one another, conduct problems and/or depression (e.g., Quiggle
there are clear distinctions in the presentation of each. et al., 1992). Indeed, as reviewed by Dodge (1993),
Thus, although these findings are assumed to relate to studies of social information processing in children
conduct problems and depression, their generaliz- with conduct problems and depression indicate
ability to distinct disorders remains unknown. general and specific difficulties at each stage of
processing.
Cognitive Deficits and Distortions In the first step, encoding social cues, individuals
must pay attention to as many relevant cues as pos-
Researchers have also identified cognitive deficits sible in an unbiased manner. For aggressive children,
and distortions as a link between conduct problems research demonstrates both deficits and distortions
related to the coding of social stimuli. Specifically, most positively evaluated response for enactment. A
aggressive boys have been found to encode fewer cues number of factors are involved in children’s evalua-
than non-aggressive boys before making attributions tions of responses including the moral (‘‘good’’ versus
about another’s intent (Dodge and Newman, 1981) ‘‘bad’’) acceptability of a response and the degree of
and their encoding tends to be systematically biased confidence children have in their ability to enact each
toward hostile cues (Crick and Dodge, 1996). Con- response (i.e., self-efficacy). Aggressive children judge
versely, those with depression tend to attend to neg- aggression to be less morally ‘‘bad’’ than other chil-
ative cues about the self particularly in response to dren (Deluty, 1983). Aggressive children are also
situations involving affiliative loss or failure (Beck, more confident in their ability to aggress than their
1967). Thus, both children with conduct problems non-aggressive counterparts (Quiggle et al., 1992).
and depression demonstrate biases in coding. How- Depressed children, on the other hand, tend to expect
ever, those with conduct problems show biases more positive outcomes for withdrawal (Quiggle
toward encoding hostile acts, whereas depressed et al., 1992).
children are primed to attend to failure, loss, and In one of the few studies to evaluate information
negative self-reference (Dodge, 1993). processing in comorbid aggressive and depressed
The second step of processing involves the children, Quiggle and colleagues (1992) found that
mental representation and interpretation of the those having both conduct problems and depression
encoded cues. Research shows that children with demonstrated processing characteristics similar to
conduct problems show a hostile attribution bias as both subgroups. In other words, comorbid children
they tend to selectively attend to or misperceive showed the cognitive patterns of both the aggressive
hostile cues in social interactions (Crick and Dodge, and depressive subgroups in an additive manner. The
1996; Dodge and Frame, 1982). Again at this stage, only seemingly unique difference in the comorbid
depressed children tend to make internal, stable, and group emerged on response generation when
global attributions for negative events (e.g., Abram- comorbid children tended to generate more purely
son et al., 1978). Using the Children’s Attributional affective responses to negative situations. The authors
Style Questionnaire, several studies have replicated suggest that comorbid children become more frus-
this finding (e.g., Abela, 2001). Additional studies trated than non-comorbid individuals when they do
have shown that depressed children also demonstrate not obtain their desired goal. In turn, they feel more
hostile attribution bias. In comparing those with angry and helpless, similar to their depressed peers.
aggression and depression, one study (Quiggle et al., However, they may focus more on the anger and
1992) showed that aggressive children who showed a frustration rather than the helplessness. Thus, they
hostile attributional bias were more likely to report may be less likely than those with depression alone to
that they would engage in aggressive behavior and give up or withdraw. Instead, they might continue to
indicated that aggression would be easy for them. be frustrated and aggressive. Such an account, how-
However, depressed children were more likely to ever, is only speculative.
attribute negative situations to internal, stable, and
global causes. Summary of Common Risk Factors
In the third step, children access one or more
possible behavioral responses from long-term mem- While this review has indicated several factors
ory or construct a novel response. Both the quantity common to both conduct problems and depression, it
and the quality of responses generated at this stage is necessary to point out that while such studies have
have been studied. Aggressive boys tend to generate a shown that these factors are correlates of each dis-
fewer number of responses than popular boys order, sufficient evidence has not been offered to
(Richard and Dodge, 1982). Moreover, they tend to show that they are causal risk factors antecedent to
show proportionately more aggressive responses the onset of the disorders. Even when studies do take
(Dodge, 1980). Comparatively less research has a longitudinal approach, they typically follow these
examined response generation in depressed children. children for only a short period of time. In addition,
However, one of the few studies to examine this issue these studies often follow children at quite young
found that depressed children generated more irrele- ages—before they have even reached the period of
vant responses to problems (Quiggle et al., 1992). greatest risk for comorbidity (see Seligman and
In the fourth step, children evaluate the previ- Ollendick, 1998). Since knowledge of causal risk
ously accessed or constructed responses and select the factors relies heavily on the results of experimental
trials as opposed to observational or even longitudi- one another and influence relations between risk fac-
nal research the present findings fall short of dem- tors and problem behaviors. Unfortunately, the
onstrating a causal link. majority of studies on risk factors stop at the point of
Moreover, a clear paucity of research fails to identifying these common variables. While a number
explain how such common factors relate to comorbid of factors have been identified, far less is known about
individuals. While ample findings have demonstrated the mechanisms by which they operate (Rutter, 1997).
correlates of each disorder alone, the current research Studies continue to provide retrospective, correla-
lends itself largely only to speculation as to how these tional results that are not able to explain how these
issues might present in comorbid individuals. factors influence the development of psychopathology.
While identifying these factors is an appropriate
INITIAL CONCLUSIONS first step, little can be done in terms of interventions
without knowing how these factors interrelate and
The above discussion explored various aspects of how they operate upon one another. Indeed, knowing
comorbidity, focussing on potential explanations for what characteristics or events precede the onset of
the overlap of conduct problems and depression. comorbid disorders does not adequately explain the
Various methodological flaws, including Berkson’s mechanisms by which these risk factors lead to the
bias and the use of multiple informants, were explored development of a specific disorder. To account for
as possible explanatory factors. However, the fact that this limitation, investigators are increasingly incor-
similar rates of comorbidity have been found in non- porating a developmental psychopathology frame-
clinical samples primarily ruled out this possibility. work to address the full array of features associated
The previous discussion also entertained the idea that with these disorders (Toth and Cicchetti, 1999).
the presence of shared diagnostic criteria may explain
the covariance of these disorders. However, exami- Developmental Psychopathology
nation of the symptom criteria for conduct problems
and depression showed that criterion overlap is min- The goal of developmental psychopathology is
imal and this factor alone does not account for to ‘‘integrate often disparate fields of study into an
comorbidity. Next, the potential causal roles between interdisciplinary perspective that informs efforts to
conduct problems and depression were discussed. prevent and ameliorate maladaptation and psycho-
Although only tentative conclusions may be drawn, pathology across the lifespan,’’ (Toth and Cicchetti,
the literature suggests that in the vast majority of 1999, p. 16). Under this framework, the organization
individuals, conduct problems develop prior to of specific problem behaviors is considered in the
depression, although in other individuals the reverse context of a series of age- and stage-relevant tasks,
may be true. Thus, it appears that conduct problems which are important to adaptation over time. The
exert a main effect on depression and vice versa. role of isolated risk factors and causal pathways in
Finally, various common risk factors for the two the development of child and adolescent psychopa-
disorders were explored. The review of the literature thology are considered within a transactional context
demonstrated that conduct problems and depression involving the growing child and his/her environment.
have common (e.g., maternal depression) and unique Based on this framework, the individual child and
risk factors (e.g., parental antisocial personality their environment are intricately intertwined in the
disorder). Thus, the co-occurrence of these disorders development of child psychopathology.
may be explained, in part, by their overlapping risk Thus, the following discussion considers the
factors. Still, there remains a lack of research on the dynamic development of comorbidity as it emerges
presentation of these risk factors in comorbid indi- over time and integrates the aforementioned risk fac-
viduals. Thus, a model by which these factors might tors into this theoretical framework for understanding
develop in comorbid individuals is next proposed. the etiology of conduct disorder and depression. The
model is, of course, tentative since firm empirical
MODEL FOR THE DEVELOPMENT support for the various connections is not yet available.
OF COMORBIDITY
Overview of Model
Many of the factors reviewed above are interre-
lated and serve to compound one another though a The proposed model (see Figure 1) builds upon
complex network of factors. These variables influence the work of Fergusson et al. (1996) and Weiss et al.
In general, our model suggests common and

unique risk factors that might account for the initial
onset and interaction between disorders as well as
risk factors which may lead to comorbidity. The
model accounts for the development and mainte-
nance of conduct problems and depression (alone and
in combination), the interconnectedness of various
common and unique risk factors, and the direct and
indirect relations between these disorders over time.
In this vein, the model (adapted from Fergusson
et al., 1996 and Weiss et al., 1998) assumes that
1. The child’s level of conduct problems at Time 1
is a function of the unique and common risk fac-
Fig. 1. Model for the development of comorbid conduct prob- tors that contribute to the development of this
lems and depression (adapted from Fergusson et al., 1996).
disorder.
2. The child’s level of depression at Time 1 is a
function of the unique and common risk factors
(1998) to explain the development of comorbid con- that contribute to the development of this disor-
duct problems and depression. As noted earlier, der.
Fergusson and colleagues sought to explain how 3. Conduct problems and depression are recipro-
common risk factors account for relations between cally related so that one disorder increases the
conduct problems and affective disorders over time. child’s risk for the second disorder (at Times 1
Basically, common risk factors were expected to and 2) and vice versa.
account for the relationship between these disorders. 4. The child’s level of conduct problems at Time 2
However, this model did not explain the exegesis of is a function of his or her preexisting level of
the initial disorder or account for unique risk factors. conduct problems and depression at Time 1 and
Moreover, when tested, the common risks in this known unique and common risk factors for con-
model did not fully account for comorbidity as duct problems.
hypothesized. 5. The child’s level of depression at Time 2 is a
It is suggested that the work of Weiss et al. function of his or her preexisting level of depres-
(1998) may serve as a practical addition to this model sion and conduct problems at Time 1 and the
with a focus on common and unique features of unique and common risk factors for depression.
psychological disorders. These authors argued that it 6. Common risk factors may influence the develop-
might be useful to conceptualize childhood disorders ment of conduct problems and/or depression at
in terms of generality and specificity. Weiss and col- Times 1 and 2.
leagues combined these factors into a conceptual and In sum, the model allows for several potential
data-analytic model for characterizing different fea- pathways to comorbidity. To be considered a com-
tures of childhood psychopathology. Specifically, the mon risk factor, a variable must be shown to signif-
model pooled these factors in an additive manor icantly relate to both conduct problems and
resulting in the following equation: Expression of depression. To be defined as a unique risk factor, a
Syndrome = Common Features + Broadband- variable must discriminate between conduct problems
Specific Features + Error [Individual Differences]. and depression, showing significant correlation with
While this model may account for the expression and only one of the disorders, but not the other.
differentiation of a single disorder, it does not ac- Accordingly, common risk factors are related to the
count for the expression of comorbidity and the development of conduct problems, depression, or
interaction between disorders. In other words, the their comorbid condition, while unique risk factors
model shows how disorders may be distinct from one serve to differentiate conduct problems and depres-
another, but it does not explain how such distinct sion. The existence of these disorders (alone or in
characteristics may coexist in the same individual. combination) is considered a byproduct of these
For these reasons, a modified model is needed to various risk factors as they influence the child’s
account for comorbidity. development over time.
Risk Factors one in which there is somewhat of a cyclical pattern

with one giving rise to the other and vice versa.
Although at first glance, and as evidenced by our Accordingly, each disorder places an individual at
review, there appears to be a great deal of overlap in risk for the development of the other disorder.
factors related to the onset of conduct problems and However, the interaction between these two disorders
depression. However, our review also illustrates how does not necessarily have to be symmetrical in that
these can be broken down into common and unique one disorder affects the other to the same degree as
factors. For example, while both groups have diffi- the other (Loeber and Keenan, 1994). Based on the
culty regulating emotion, the review of the literature reviewed literature, it seems quite probable that
demonstrates that those with externalizing problems conduct problems increase the risk of depression
are undercontrolled while those with internalizing more than the reverse direction. Although the exact
problems are overcontrolled (e.g., Eisenberg et al., mechanisms by which these processes take place are
2001). Moreover, the previous discussion of risk not known, several theories have been presented.
factors serves as an example of how these factors may In particular, children may progress from
interrelate in a transactional manner. For example, depression to conduct problems as a means of acting
parent psychopathology may give rise to negative out their internalized feelings. Related to the idea of
self-concepts, which may, in turn lead to depression ‘‘masked depression,’’ depressed children may
in the child. Furthermore, the depression in the child become aggressive as a means of externalizing some
may lead to more depression in the parent as the of the feelings, particularly of sadness, that they
parent may, for example, question their parenting experience. Others have suggested that depression
abilities if the child develops behavior problems. may lead to conduct problems through cognitive
Thus, a cycle in which parent and child influence one debilitation in the midst of frustrating situations
another’s psychopathology is suggested. Overall, (Greene and Doyle, 1999). Children experiencing the
according to the model, common risk factors give rise irritability and distractibility associated with depres-
to either disorder, while unique factors differentiate sion may have increased difficulty thinking through
between these conditions. Various combinations of various situations and responding appropriately. For
these unique and common risk factors may contrib- example, depression might in some cases impair an
ute to comorbidity. individuals’ concern about the adverse consequences
Following the onset of conduct problems or of their actions, thereby increasing risk for conduct
depression, the model further posits that conduct problems (Lilienfeld, 2003). Moreover, the negative
problems may exert a main effect on depression (or affect that is associated with depression may by
vice versa) or one disorder might exert an indirect annoying or irritating to others and, in turn, may
effect on the other through common risk factors. The have negative effects on relationships with significant
direct link between conduct problems and depression others in the child’s life which may contribute to
in the present model is maintained for two reasons. greater conflict with others (Oland and Shaw, 2005).
First, given the number of risk factors believed to be Such pathways from depression to conduct problems
related to these disorders, it seems that it is not likely are worthy of consideration in the development of
that the correlation between the disorders will be fully comorbidity.
explained by common risk factors. If for example, In terms of the path from conduct problems to
any one of these factors is not appropriately included depression, it is believed that as hypothesized by
in the model then the mediation of this relationship Patterson and colleagues (1992), the adverse conse-
will not be complete. While Fergusson et al. (1996) quences and life circumstances associated with con-
attempted to explain this relationship through indi- duct problems may lead to more depressed thoughts
rect means they were only successful in explaining such as negative views of the self, current circum-
part of the variance in the co-occurrence of disorders. stances, and the future. As mentioned previously,
Second, and more importantly, this direct relation- specific social and academic problems that are often
ship is maintained in the current model since several seen as byproducts of conduct problems include
studies have shown that one disorder may place an academic underachievement, peer rejection, problems
individual at increased risk for comorbidity (e.g., in parent–child relationship, and/or legal difficulties
Capaldi, 1991). (Capaldi, 1991). The accumulation of negative reac-
The nature of the relationship between conduct tions from others coupled with failures in various
problems and depression appears to be a reciprocal realms of functioning may subsequently increase a
child’s vulnerability to depression. Thus, it appears practice are warranted. First, in terms of research, it
that there are of number of possibilities to explain the is important for studies to use techniques that reduce
path from conduct problems to depression. artifactual comorbidity. As has been reviewed, vari-
ous methodological techniques (e.g., use of clinical
samples or multiple informants) may over inflate
Summary
rates of comorbidity. To address some of these issues,
it is recommended that studies make use of general
In sum, the proposed model incorporates com-
population samples. In addition, to reduce error rates
mon and unique factors related to the development of
based on the use of multiple informants, it is sug-
comorbidity over time. Both direct and indirect
gested that discrepancies between reporters be
relationships between conduct problems and depres-
resolved by a trained clinician rather than the
sion have been reviewed and incorporated into this
so-called ‘‘OR’’ rule (Jensen, 2003). Since some
framework.
informants may be more reliable or accurate than
A significant drawback of this model is its ne-
others, the resolution of this issue by a third, ‘‘expert’’
glect of risk factors that might be unique to comorbid
party seems to be a better option. An alternate
individuals. While the model allows for risk factors of
solution to the use of multiple informants involves
conduct problems and depression to be compiled in
the application of confirmatory factor analyses to
an additive manner, it does not account for variations
data from two informants (e.g., Keiley et al., 2003). It
in these factors which may be unique to children with
is believed that in employing such techniques,
comorbid disorders. Still, those studies that have
researchers may be better able to concentrate on
examined risk factors in comorbid individuals seem
‘‘true’’ cases of comorbidity.
to suggest the risk factors for comorbidity consist of a
Next, to address the temporal ordering of
compilation of factors related to both conduct
disorders and the possibility that one may give rise to
problems and depression, and that separate factors
the other, a greater emphasis on longitudinal inves-
may not exist.
tigations is needed. While conduct problems tend to
It should be reiterated that despite the presence
precede depression in the majority of cases, further
of various risk factors, many individuals demonstrate
research is needed to elucidate the nature of these
great resilience and do not go on to display psycho-
relationships. Investigations of these disorders over
pathology. This resilience may reside in the presence
time may clarify the developmental course of these
of a number of factors in the child’s life that may
disorders as they tend to wax and wane over time
counteract these risk factors. For example, an easy
(Caron and Rutter, 1991; Loeber and Keenan, 1994).
temperament, high level of intelligence, or the pres-
Such studies allow for better elucidation of various
ence of positive social support may mitigate the
antecedent risks and outcomes of the disorders as
relationship between the aforementioned risk factors
well as mechanisms which may explain the path from
and subsequent psychopathology within the pro-
conduct problems into depression or vice versa. To
posed model.
date, research on such mechanisms and predictors of
Finally, while the proposed model is believed to
comorbidity have been limited.
elucidate relationships between conduct problems
Along these lines, it is hoped that the presently
and depression, it is recognized that these processes
proposed model may lend itself to longitudinal
remain speculative at this time and await empirical
research. It is believed that this model lends itself to
verification. However, in as much as the proposed
the integration of various risk factors and examina-
model helps to elucidate the interconnectedness of
tion of the processes by which comorbidity develops.
various factors, the heuristic utility of the model is
Accordingly, various mediating and moderating
that it brings a developmental perspective on psy-
variables in the development of comorbidity must be
chopathology and provides a framework for inte-
thoroughly delineated. In particular, the risk factors
grating various mechanisms of risk.
reviewed in this study merit further investigation as
parental psychopathology, emotion regulation, and
IMPLICATIONS FOR CLINICAL RESEARCH maladaptive cognitions place individuals at risk for
AND PRACTICE the development of both disorders. Specifically,
greater delineation of the common and unique as-
Based on the review of the literature and pro- pects of these risk factors that may contribute to one
posed model, suggestions for future research and or both disorders is needed.
In terms of clinical practice implications, the delineated (Caron and Rutter, 1991). Moreover,
high prevalence of comorbid conduct problems and although studies of comorbidity have failed to show
depression highlight the need for interventions spe- convincingly which risk factors are common or
cific to this population. It is also hoped that preven- unique to conduct problems and depression and how
tion research in this area may be benefitted by the these factors interrelate in the development of com-
proposed model as specific common and unique risk orbidity, important advances have been made. While
factors for these disorders are identified and targeted. identifying these factors is an appropriate first step,
Identifying these early risk factors may help clinicians models of risk, such as the one presented, are needed
to identify individuals who are at risk for comorbidity to explain the mechanisms by which these risk factors
and to develop more appropriate interventions which ultimately lead to the development of comorbidity
include prevention efforts focussing on one or both and its effective treatment.
disorders. For example, as early differences in emo-
tion regulation are identified, prevention programs
REFERENCES
may be designed to target these children before the
onset of one or both disorders. Moreover, because Abela, J. (2001). The hopelessness theory of depression: A test of
comorbid children and adolescents demonstrate poor the diathesis—stress and causal mediation components in third
treatment outcomes (e.g., Hughes et al., 1990; Rohde and seventh grade children. Journal of Abnormal Child
Psychology 29(3): 241–254.
et al., 2001) more research must identify effective Abramson, L., Seligman, M. E., and Teasdale, J. D. (1978).
treatment modalities for these individuals. Indeed, Learned helplessness in humans: Critique and reformulation.
while research has provided us with some empirically Journal of Abnormal Psychology 87(1): 49–74.
American Psychiatric Association. (2000). The Diagnostic and
based guidelines for treating these disorders alone, Statistical Manual of Mental Disorders (4th ed., text revised),
much less is known about effective means of treating Washington, DC, Author.
those having symptoms of both disorders. In fact, Anderson, J. C., Williams, S. M., McGee, R., and Silva, P. A.
(1987). DSM-III disorders in preadolescent children: Preva-
presently no empirically supported treatments have lence in a large sample from the general population. Archives
been developed for youth who present with these co- of General Psychiatry 44(1): 69–76.
occurring conditions (Ollendick and King, 2004). Angold, A., and Costello, E. (1993). Depressive comorbidity in
children and adolescents: Empirical, theoretical, and method-
Given the high rates of comorbidity between these ological issues. American Journal of Psychiatry 150(12): 1779–
disorders, it is all the more important to account for 1791.
treatment variables specific to these individuals in Angold, A., Costello, E., and Erkanli, A. (1999). Comorbidity.
Journal of Child Psychology and Psychiatry and Allied Disci-
order to determine what treatments work best for plines 40(1): 57–87.
which individuals under which circumstances. Arredondo, D. E., and Butler, S. F. (1994). Affective comorbidity
In the mean time, because the literature suggests in psychiatrically hospitalized adolescents with conduct disor-
der or oppositional defiant disorder: Should conduct disorder
that co-varying conduct problems and depression be treated with mood stabilizers?. Journal of Child and
often present more similar to pure conduct problem Adolescent Psychopharmacology 4(3): 151–158.
behaviors than to pure depression, it is hypothesized Beck, A.T. (1967). Depression, Harper and Row, New York.
Beyers, J., and Loeber, R. (2003). Untangling developmental
that treatments targeting the risks associated with relations between depressed mood and delinquency in male
co-occurrence may most productively apply treat- adolescents. Journal of Abnormal Child Psychology 31(1): 247–
ments found to be effective for conduct problems. 266.
Biederman, J., Faraone, S., Mick, E., and Lelon, E. (1995).
Future research should explore not only the efficacy Psychiatric comorbidity among referred juveniles with major
of such treatments for children with co-occurring depression: Fact or artifact?. Journal of the American Academy
conduct problems and depression, but also the of Child and Adolescent Psychiatry 34(5): 579–590.
Block, J., and Block, J. (1980). The role of ego-control and ego-
efficacy of treatments tailored to the distinct charac- resiliency in the organization of behavior. In Collins, W. S.
teristics of these comorbid individuals. (Ed.), Minnesota Symposia on Child Psychology: Vol. 13.
Development of Cognition, Affect, and Social Relations (pp. 39–
101). Hillsdale, NJ: Erlbaum.
CONCLUSIONS Capaldi, D., and Stoolmiller, M. (1999). Co-occurrence of conduct
problems and depressive symptoms in early adolescent boys:
Clearly, the study of comorbidity and its devel- III. Prediction to young–adult adjustment. Development and
Psychopathology 11(1): 59–84.
opment remain in their ‘‘embryonic state’’ (Loeber Capaldi, D. (1991). Co-occurrence of conduct problems and
et al., 2000, p. 1475). Although comorbidity has been depressive symptoms in early adolescent boys: I. Familial
documented in many epidemiological and longitudi- factors and general adjustment at Grade 6. Development and
Psychopathology 3(3): 277–300.
nal studies, the risks and mechanisms related to its Capaldi, D. (1992). Co-occurrence of conduct problems and
etiology and development are yet to be fully depressive symptoms in early adolescent boys: II. A 2-year
follow-up at Grade 8. Development and Psychopathology 4(1): tions for assessment and treatment. Clinical Child and Family
125–144. Psychology Review 2: 129–148.
Caron, C., and Rutter, M. (1991). Comorbidity in child psycho- Greene, R. W., Biederman, J., Zerwas, S., Monuteaux, M., Goring,
pathology: Concepts, issues and research strategies. Journal of J. C., and Faraone, S. V. (2002). Psychiatric comorbidity,
Child Psychology and Psychiatry 32: 1063–1080. family dysfunction, and social impairment in referred youth
Cohen, P., Cohen, J., Kasen, S., Velez, C. N., Hartmark, C., with oppositional defiant disorder. American Journal of
Johnson, J., Rojas, M., Brook, J., and Streuning, E. L. (1993). Psychiatry 159(7): 1214–1224.
An epidemiological study of disorders in late childhood and Grolnick, W., Bridges, I. J., and Connell, J. P. (1996). Emotion
adolescence: I. Age- and gender-specific prevalence. Journal of regulation in two-year-olds: Strategies and emotional expres-
Child Psychology & Psychiatry 34(6): 851–867. sion in four contexts. Child Development 67: 928–941.
Cole, P., Zahn-Waxler, C., Fox, N., Usher, B., and Welsh, J. D. Hammen, C. (1991). Depression Runs in Families: The Social
(1996). Individual differences in emotion regulation and Context of Risk and Resilience in Children of Depressed
behavior problems in preschool children. Journal of Abnormal Mothers, Springer-Verlag, New York.
Psychology 105(4): 518–529. Harrington, R., Fudge, H., Rutter, M., and Pickles, A. (1991).
Crick, N. B., and Dodge, K. A. (1996). Social information- Adult outcomes of childhood and adolescent depression: II.
processing mechanisms in reactive and proactive aggression. Links with antisocial disorders. Journal of the American
Child Development 67: 993–1002. Academy of Child and Adolescent Psychiatry 30(3): 434–439.
Dodge, K. (1993). Social-cognitive mechanisms in the development Hughes, C., Preskorn, S. H., Weller, E., Weller, R., Hassanein, R.,
of conduct disorder and depression. Annual Review of and Tucker, S. (1990). The effect of concomitant disorders in
Psychology 44: 559–584. childhood depression on predicting treatment response. Psy-
Dodge, K., and Newman, J. P. (1981). Biased decision-making chopharmacology Bulletin 26(2): 235–238.
processes in aggressive boys. Journal of Abnormal Psychology Jensen, P. (2003). Comorbidity and Child Psychopathology:
90(4): 375–379. Recommendations for the next decade. Journal of Abnormal
Dodge, K. (1980). Social cognition and chldren’s aggressive Child Psychology 31(3): 293–300.
behavior. Child Development 51: 162–170. Jensen Watanabe, P. S. H. K., Richters, J., Cortes, R., and Sharon,
Dodge, K., and Frame, C. L. (1982). Social cognitive biases M. (1995). Prevalence of mental disorder in military children
and deficits in aggressive boys. Child Development 53: 620– and adolescents: Findings from a two-stage community
635. survey. Journal of the American Academy of Child & Adolescent
Eisenberg, N., Cumberland, A., Spinrad, T., Fabes, R., Shepard, Psychiatry 34(11): 1514–1524.
S., Reiser, M., Murphy, B., Losoya, S., and Guthrie, I. (2001). Jensen, P. S., Rubio-Stipec, M., Canino, G., Bird Dulcan, H. R. M.
The relations of regulation and emotionality to children’s K., Schwab-Stone, M. E., and Lahey, B. B. (1999). Parent and
externalizing and internalizing problem behavior. Child Devel- child contributions to diagnosis of mental disorder: Are both
opment 72(4): 1112–1134. informants always necessary?. Journal of the American Acad-
Eisenberg, N., Fabes, R. A., Guthrie, I. K., and Reiser, M. (2000). emy of Child and Adolescent Psychiatry 38: 1569–1579.
Dispositional emotionality and regulation: Their role in Jewell, J. D., and Stark, K. D. (2003). Comparing the family
predicting quality of social functioning. Journal of Personality environments of adolescents with conduct disorder or depres-
& Social Psychology 78(1): 136–157. sion. Journal of Child & Family Studies 12(1): 77–89.
Eisenberg, N., Fabes, R. A., Shepard, S. A., Murphy Guthrie, B. Keiley, M., Lofthouse, N., Bates, J. E., Dodge, K. A., and Pettit,
C. I. K., Jones, S., Friedman, J., Poulin, R., and Maszk, P. G. S. (2003). Differential risks of covarying and pure compo-
(1997). Contemporaneous and longitudinal prediction of nents in mother and teacher reports of externalizing and
children’s social functioning from regulation and emotionality. internalizing behavior across ages 5 to 14. Journal of Abnormal
Child Development 68: 642–664. Child Psychology 31(3): 267–283.
Deluty, R. H. (1983). Children’s evaluations of aggressive, asser- Kovacs, M., Paulauskas, S., Gatsonis, C., and Richards, C. (1988).
tive, and submissive responses. Journal of Clinical Child Depressive disorders in childhood: III. A longitudinal study of
Psychology 12(2): 124–129. comorbidity with and risk for conduct disorders. Journal of
Fergusson, D. M., Lynskey, M. T., and Horwood, L. J. (1996). Affective Disorders 15(3): 205–217.
Origins of comorbidity between conduct and affective disor- Lahey, B. B., Loeber, R., Burke, J., Rathouz, P. J., and McBurnett,
ders. Journal of the American Academy of Child and Adolescent K. (2002). Waxing and waning in concert: Dynamic comor-
Psychiatry 35(4): 451–460. bidity of conduct disorder with other disruptive and emotional
Forbes, E. E., Shaw, D. S., Fox, N. A., Cohn, J. F., Silk, J. S., and problems over 17 years among clinic-referred boys. Journal of
Kovacs, M. (2006). Frontal asymetry, mother-child affect Abnormal Psychology 111(4): 556–567.
expression, and maternal depression as factors in child Lahey, B. B., Piacentini, J. C., McBurnett, K., Stone, P.,
behavior problems. Journal of Child Psychology and Psychia- Hartdagen, S. E., and Hynd, G. (1988). Psychopathology
try. and antisocial behavior in the parents of children with conduct
Frick, P., and Morris, A. (2004). Temperament and developmental disorder and hyperactivity. Journal of the American Academy
pathways to conduct problems. Journal of Clinical Child and of Child and Adolescent Psychiatry 27: 163–170.
Adolescent Psychology 33(1): 54–68. Lewinsohn, P. M., Hops, H., Roberts, R. E., Seeley, J. R., and
Frick, P. J. (1998). Conduct disorders. In Ollendick, T. H., and Andrews, J. A. (1993). Adolescent psychopathology: 1. Prev-
Hersen, M. (Eds.), Handbook of Child Psychopathology (3rd alence and incidence of depression and other DSM-III-R
ed., pp. 213–237). New York: Plenum Press. disorders in high school students. Journal of Abnormal
Frick, P. J., Lahey, B. B., Loeber, R., Stouthamer-Loeber, M., Psychology 102: 133–144.
Christ, M., and Hanson, K. (1992). Familial risk factors to Lewinsohn, P. M., Rohde, P., and Seeley, J. R. (1995). Adolescent
oppositional defiant disorder and conduct disorder: Parental psychopathology: III. The clinical consequences of comorbid-
psychopathology and maternal parenting. Journal of Consult- ity. Journal of the American Academy of Child and Adolescent
ing and Clinical Psychology 60(1): 49–55. Psychiatry 34(4): 510–519.
Gray, J. A. (1987). The Psychology of Fear and Stress, Cambridge Lilienfeld, S. O. (2003). Comorbidity between and within childhood
University Press, Cambridge. externalizing and internalizing disorders: Reflections and
Greene, R. W., and Doyle, A. E. (1999). Toward a transactional directions. Journal of Abnormal Child Psychology 31(3): 285–
conceptualization of oppositional defiant disorder: Implica- 291.
Loeber, R., and Keenan, K. (1994). Interaction between conduct Rapp, L. A, and Wodarski, J. S (1997). The comorbidity of
disorder and its comorbid conditions: Effects of age and conduct disorder and depression in adolescents: A compre-
gender. Clinical Psychology Review 14(6): 497–523. hensive interpersonal treatment technology. Family Therapy
Loeber, R., Burke, J. D., Lahey, B. B., Winters, A., and Zera, M. 24(2): 81–100.
(2000). Oppositional defiant and conduct disorder: A review of Richard, B. A., and Dodge, K. A. (1982). Social maladjustment
the past 10 years: I. Journal of the American Academy of Child and problem solving in school-aged children. Journal of
and Adolescent Psychiatry 39: 1468–1484. Consulting & Clinical Psychology 50(2): 226–233.
Loeber, R., Green, S. M., and Keenan, K. (1995). Which boys will Robins, L. N., and Price, R. (1991). Adult disorders predicted by
fare worse? Early predictors of the onset of conduct disorder in childhood conduct problems: Results from the NIMH Epi-
a six year longitudinal study. Journal of the American Academy demiologic Catchment Area project. Psychiatry: Journal for
of Child and Adolescent Psychiatry 34(4): 499–509. the Study of Interpersonal Processes 54(2): 116–132.
Loeber, R., Russo, M. F., Stouthamer-Loeber, M., and Lahey, B. Rohde, P., Clarke, G. N., Lewinsohn, P. M., Seeley, J. R., and
B. (1994). Internalizing problems and their relation to the Kaufman, N. K. (2001). Impact of comorbidity on a cognitive-
development of disruptive behaviors in adolescence. Journal of behavioral group treatment for adolescent depression. Journal
Research on Adolescence 44(4): 615–637. of the American Academy of Child and Adolescent Psychiatry
Marmorstein, N., and Iacono, W. G. (2004). Major depression and 40(7): 795–802.
conduct disorder in youth: Associations with parental psycho- Rohde, P., Clarke, G. N., Mace, D. E., Jorgensen, J. S., and Seeley,
pathology and parent–child conflict. Journal of Child Psychol- J. R. (2004). An efficacy/effectiveness study of cognitive-
ogy & Psychiatry 45(2): 377–386. behavioral treatment for adolescents with comorbid major
McMahon, R. J., and Wells, K. C. (1998). Conduct problems. In depression and conduct disorder. Journal of the American
Mash, E. J., and Barkley, R. A. (Eds.), Treatment of Academy of Child and Adolescent Psychiatry 43(6): 660–668.
Childhood Disorders (2nd ed., pp. 111–207). New York, NY: Rudolph, K., and Clark, A. G. (2001). Conceptions of relation-
Guilford Press. ships in children with depressive and aggressive symptoms:
Moffit, T. (1993). Adolescence-limited and life-course-persistent Social-cognitive distortion or reality? Journal of Abnormal
antisocial behavior: A developmental taxonomy. Psychological Child Psychology 29(1): 41–56.
Review 100(4): 674–701. Rutter, M. (1997). Stress research: Accomplishments and tasks
Nock, M., Kazdin, A., Hiripi, E., and Kessler, R. (2006). ahead. In Haggerty, R. J., Sherrod, L. R., Garmezy, N., and
Prevalence, subtypes, and correlates of DSM-IV conduct Rutter, M. (Eds.), Stress, Risk, and Resilience in Children and
disorder in the National Comorbidity Survey Replication. Adolescents: Processes, Mechanisms, and Interventions New
Psychological Medicine 36: 699–710. York: Cambridge University Press.
Nottelmann, E. D., and Jensen, P. S. (1995). Comorbidity of Seligman, L., and Ollendick, T. (1998). Comorbidity of anxiety and
disorders in children and adolescents: Developmental perspec- depression in children and adolescents: An integrative review.
tives. Advances in Clinical Child Psychology 17: 109–155. Clinical Child and Family Psychology Review 1(2): 125–144.
Nottelmann, E., and Jensen, P. (1994). Comorbidity of disorders in Stark, K. D., Bronik, M. D., and Wong, S. (2000). Depressive
childhood and adolescence: Developmental Perspectives. In Disorders. In Hersen, M., and Ammerman, R. T. (Eds.),
Ollendick, T. H., and Prinz, R. J. (Eds.), Advances in Clinical Advanced Abnormal Child Psychology (2nd ed., pp. 291–326).
Child Psychology18 New York: Plenum Press. Mahwah, NJ, US: Lawrence Erlbaum Associates Publishers.
O’Connor, T. G., Neiderhiser, J. M., Reiss, D., Hetherington, E., Toth, S. L., and Cicchetti, D. (1999). Developmental psychopa-
and Plomin, R. (1998). Genetic contributions to continuity, thology and child psychotherapy. In Russ, S. W., and
change, and co-occurrence of antisocial and depressive symp- Ollendick, T. H. (Eds.), Handbook of Psychotherapies with
toms in adolescence. Journal of Child Psychology and Psychi- Children and Families. Issues in Clinical Child Psychology (pp.
atry 39(3): 323–336. 15–44). Dordrecht, Netherlands: Kluwer Academic Publish-
Oland, A. A., and Shaw, D. S. (2005). Pure versus co-occurring ers.
externalizing and internalizing symptoms in children: The Trad, P. V. (1987). Infant and Childhood Depression: Developmental
potential role of socio-developmental milestones. Clinical Factors, John Wiley & Sons, Oxford, England.
Child and Family Psychology Review 8(4): 247–270. Weiss, B., Susser, K., and Catron, T. (1998). Common and specific
Ollendick, T. H., and King, N. J. (2004). Empirically supported features of childhood psychopathology. Journal of Abnormal
treatments for children: Advances toward evidence-based Psychology 107(1): 118–127.
practice. In Barrett, P., and Ollendick, T. H. (Eds.), Williamson, D. E., Ryan, N. D., Birmaher, B., Dahl, R. E.,
Handbook of Interventions that Work with Children and Kaufman, J., Rao, U., and Puig-Antich, J. (1995). A case–
Adolescents: From Prevention to Treatment (pp. 3–26). Lon- control family history study of depression in adolescents.
don: John Wiley & Sons, Inc. Journal of the American Academy of Child and Adolescent
Ollendick, T. H., and Hersen, M. (1993). Handbook of Child and Psychiatry 34(12): 1596–1607.
Adolescent Assessment, Allyn and Bacon, Boston. Youngstrom, E. A., Findling, R. L., and Calabrese, J. R. (2003).
Patterson, G. R., Reid, J. B., and Dishion, T. J. (1992). Antisocial Who are the comorbid adolescents? Agreement between
Boys, Castalia Publishing, Patterson, OR. psychiatric diagnosis, youth, parent, and teacher report.
Phares, V., and Compas, B. E. (1992). The role of fathers in child Journal of Abnormal Child Psychology 31(3): 231–245.
and adolescent psychopathology: Make room for daddy. Zoccolillo, M., and Rogers, K. (1991). Characteristics and outcome
Psychological Bulletin 111(3): 387–412. of hospitalized adolescent girls with conduct disorder. Journal
Puig-Antich, J. (1982). Major depression and conduct disorder in of the American Academy of Child and Adolescent 30(6): 973–
prepuberty. Journal of the American Academy of Child 981.
Psychiatry 21(2): 118–128. Zoccolillo, M. (1992). Co-occurrence of conduct disorder and its
Puig-Antich, J. (1989). A controlled family history study of adult outcomes with depressive and anxiety disorders: A
prepubertal major depressive disorder. Archives of General review. Journal of the American Academy of Child and
Psychiatry 46(5): 406–418. Adolescent Psychiatry 31(3): 547–556.
Quiggle, N., Garber, J., Panak, W., and Dodge, K. A. (1992).
Social information processing in aggressive and depressed
children. Child Development 63: 1305–1320.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

The Comorbidity of Conduct Problems and Depression in Childhood and Adolescence

Uploaded by

Copyright:

Available Formats

The Comorbidity of Conduct Problems and Depression in Childhood and Adolescence

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Comorbidity of Conduct Problems and Depression in Childhood and Adolescence

Uploaded by

Copyright:

Available Formats

Clinical Child and Family Psychology Review, Vol. 9, Nos.

3/4, December 2006 (Ó 2006)

The Comorbidity of Conduct Problems and Depression

Jennifer C. Wolﬀ,1 and Thomas H. Ollendick1,2

KEY WORDS: aggression; child; comorbidity; conduct problems; depression

In general, our model suggests common and

Risk Factors one in which there is somewhat of a cyclical pattern

You might also like