POSSIBLE EXPLANATIONS FOR CO-OCCURRING DEPRESSION AND CONDUCT PROBLEMS
Although there are a variety of reasons that conduct problems and depression may cooccur, I will focus on three potential explanations for the co-occurrence of conduct problems and depression that have garnered support based on research to date: (1) depression symptoms confer risk for conduct problems, (2) conduct problems confer risk for depression, and (3) shared risk factors account for their co-occurrence (Angold et al., 1999; Caron & Rutter, 1991; Klein & Riso, 1993; Simonoff, 2000). Evidence and potential models for each of these explanations are presented next
Explanation 1: Depression Symptoms Confer Risk for Conduct Problems
The first explanation for co-occurring depression and conduct problems is that depression is a risk factor for conduct problems. One possible model for this relation is that depressed children are irritable and withdrawn at home, which leads to parent-child conflict (Sheeber, Hyman, & Davis, 2001). Such conflict and subsequent increased levels of depression may lead to problematic peer relationships (e.g., rejection, victimization) and withdrawn behavior at school (Rudolph & Clark, 2001). As part of their depression, children may have difficulty
concentrating, which leads to poor academic performance. Over time, parents’ hostile behaviors may be increasingly replaced with withdrawal from the child, as parents recognize that their behaviors do not serve to modify the child’s behaviors. As these children age, peer difficulties and parental detachment persist, increasing the likelihood of associating with deviant peer groups (McGrath & Repetti, 2002).
Some research involving clinic-referred depressed children suggests that the onset of depression precedes the onset of conduct problems (Geller et al., 1985; Kovacs et al., 1988; Puig-Antich, 1982). Collectively, however, these studies are descriptive and do not suggest mechanisms through which depression may lead to conduct problems. Furthermore, these studies do not consider ODD symptoms, which are a common precursor to and may represent a prodromal form of CD (Lahey, Loeber, Quay, Frick, & Grimm, 1992; Moffitt, 1993a). Thus, the inability to determine whether early depressive symptoms (e.g., difficulty concentrating, social withdrawal), unmeasured ODD, or other mechanisms influenced the development of conduct problems in these studies limits the conclusions that can be drawn regarding how conduct problems may have developed in addition to depression over time (Loeber & Keenan, 1994; Rohde et al., 1991). Moreover, because support for this explanation has been limited primarily to clinical samples, it is not clear how useful this explanation will be for understanding conduct problems and depression in community-based samples in which comorbidity rates and symptom severity are likely to be lower.
Explanation 2: Conduct Problems Confer Risk for Depressive Symptoms
A second explanation is that conduct problems are a risk factor for depression. This ordering is consistent with the stage model proposed by Patterson and colleagues (1989), which indicates that children with conduct problems are more likely to fail in interpersonal relationships and academic settings (Capaldi, 1991, 1992; Patterson, DeBaryshe, & Ramsey, 1989). In this type of “failure” model, conduct problems interfere with an acquisition of social skills. These social skills deficits are exacerbated by conflictual interpersonal relationships,
which lead to associating with deviant peers and depressed mood. Consistent with this explanation, several prospective studies using community-based samples of adolescents have suggested that conduct problems precede depression significantly more often than depression precedes conduct problems when they co-occur (Capaldi, 1992; Fleming, Boyle, & Offord, 1993; Sack et al., 1993). Furthermore, among clinic-referred boys, CD measured in childhood predicted later depression even when initial levels of depression were controlled, but
depression did not predict later CD (Lahey et al., 2002). Thus, the findings of these studies lend support to the explanation that conduct problems precede depression, though this pattern may be evidenced earlier in clinic-based (Lahey et al., 2002) than community-based (Capaldi, 1992; Fleming et al., 1993) samples. Nevertheless, a general reliance on adolescent samples limits our understanding of how facets of this comorbid condition unfold during childhood, when the mechanisms and symptoms may be more amenable to intervention. In addition, the
ability of these studies to account for the comorbidity of conduct problems and depression is limited by a failure to test the shared risk factors explanation for comorbidity (Explanation 3, presented next) concurrently with the explanations that one disorder is a risk factor for the other (Explanations 1 and 2).
Explanation 3: Shared Risk Factors Account for Co-Occurring Conduct
Problems and Depression
A third possibility involves the shared risk factors explanation for comorbidity, which suggests that depression and conduct problems are associated with overlapping and unique factors and that comorbidity stems from shared risk factors (Angold et al., 1999; Caron & Rutter, 1991; Klein & Riso, 1993). At the psychological level of analysis, shared risk factors for conduct problems and depression involve at least seven domains: (1) child emotion dysregulation and emotional lability (Beauchaine, 2001; Snyder, Schrepferman, & St. Peter, 1997); (2) child attention problems (Bird, Gould, & Staghezza, 1993; Drabick, Beauchaine, Gadow, Carlson, & Bromet, 2006; Treuting & Hinshaw, 2001); (3) low levels of parent-child communication and parental detachment (Drabick, Gadow, et al., 2006; Fergusson, Lynskey, & Horwood, 1996; Ge, Best, Conger, & Simons, 1996); (4) ineffective and coercive discipline (Compton, Snyder, Schrepferman, Bank, & Shortt, 2003; Loeber, Farrington,
Stouthamer-Loeber, & Van Kammen, 1998; Patterson et al., 1989); (5) parental psychological problems, including maternal depression and marital discord (Fendrich, Warner, & Weissman, 1990; Loeber et al., 1998); (6) peer relationship problems (Capaldi, 1992; Fergusson et al., 1996; Keiley, Lofthouse, Bates, Dodge, & Pettit, 2003; Rudolph & Clark, 2001); and (7) poor academic performance (Capaldi, 1992; Loeber et al., 1998; Velez,
Johnson, & Cohen, 1989).
Unfortunately, the usefulness of these purported shared risk factors for understanding cooccurring depression and conduct problems is limited. First, these risk factors are not specific to co-occurring depression and conduct problems, which limits their utility for etiological and intervention models for this comorbid condition (Steinberg & Avenevoli, 2000). Second, presuming these are “shared” risk factors is arbitrary as these factors have not been tested effectively in other viable roles, such as mediators or moderators, for co-occurring depression and conduct problems. Indeed, many of these same risk factors are included in Explanations 1 and 2 above as the mechanisms that facilitate development of the comorbid condition (e.g., conduct problems lead to academic and interpersonal difficulties, which may lead to depression). Thus, it is plausible that these factors each play a role in the development of the comorbid condition, but their effects differ depending on when they are experienced. Third, these risk factors are likely inter-related, which restricts the ability to predict risk from these factors individually (Deater-Deckard, Dodge, Bates, & Pettit, 1998; Greenberg, Speltz, DeKlyen, & Jones, 2001).
It is furthermore likely that there are reciprocal and transactional influences among these risk factors. Using a developmental psychopathology perspective, a model for the development of depression and conduct problems that ties these various domains together could be posited. For instance, the ability to regulate emotions, attend to demands, and inhibit impulsive responding affects how well children tolerate frustration and negotiate interpersonal interactions (Bagwell, Molina, Pelham, & Hoza, 2001; Shaw, Owens, Giovannelli, & Winslow, 2001). Thus, a child who is emotionally dysregulated (Domain 1) and/or exhibits attention problems (Domain 2) may be less responsive to or compliant with parental demands, which may lead parents to spend increased amounts of time addressing discipline issues. Over time, these interactions may lead to parent-child relationships that are characterized by parental detachment and poor parent-child communication (Domain 3).
Concurrently, because of the potentially demanding and time-intensive nature of parenting an emotionally dysregulated and inattentive/impulsive child, parents may be more likely to engage in inconsistent and coercive discipline (Domain 4). In response to such discipline, children may use aggressive and depressive behaviors to deflect family members’ behaviors and facilitate their removal from aversive interactions (Patterson, 1982). Coercive parentchild interchanges are particularly likely if parents are experiencing their own psychological problems, parenting disagreements, or marital difficulties (Domain 5). In school, an emotionally dysregulated and/or inattentive child may have problematic peer relationships and experience peer rejection and/or victimization (Domain 7), which ultimately may lead to associating with deviant peers. Last, with increasing academic demands as children age, problems with inattention and emotion dysregulation are likely to
lead to poor academic performance (Domain 8). Thus, earlier difficulties may be precursors for later risk factors (e.g., childhood attention problems may be related to preadolescent academic problems; preadolescent peer rejection may be related to adolescent deviant peer involvement). Furthermore, the severity and accumulation of risk over time will impact when children exhibit conduct problems and depression, and suggest that concurrent consideration of multiple domains may better capture the processes that place children at risk for the development of comorbid depression and conduct problems (Deater-Deckard et al., 1998).