Psychological Symptom Covariation Amongst Young Adolescents: The Role of Maltreatment in the Structural Characteristics of Psychopathology

Background: A prominent feature of psychopathology is covariation between different symptoms – commonly termed, from a categorical perspective, comorbidity. In both community and clinical samples, strong symptom covariation is evidenced. Comorbid symptomatology, as opposed to single syndrome, is generally associated with greater functional impairment and poorer treatment outcomes. As most manualized treatments have been developed with respect to a single diagnosis, application of treatment manuals to these cases is not straightforward. The issue also has important implications for psychiatric nosology. Understanding the sources of covariation has important implications for understanding etiology, the course of a syndrome, prevention and treatment. Research has attempted to delineate the general causes of comorbidity, but few efforts have been made to determine if the structural characteristics of psychopathology are shared across groups with different patterns of risk. This study examines whether maltreated and non-maltreated high-risk adolescents evidence the same relations between different symptom types.

Methods: A gender-balanced, ethnically diverse urban sample of 303 adolescents with a documented history of maltreatment and a comparison group of 151 non-maltreated adolescents with similar demographic profiles was recruited. The mean age was 10.9 years. Measurement models were built for three internalizing constructs including depression, posttraumatic anxiety, generalized anxiety, and three externalizing constructs including aggression, delinquency and impulsivity. Utilizing the measurement models, a series of confirmatory factor models with theoretical precedent were fit to characterize the relations among these variables in both groups. The six models represent increasingly parsimonious – more highly covarying models – of psychopathology, culminating in a model where a single factor optimally explains the observed correlations between all six symptom types. Selection of the best fitting model was informed by a range of fit indices but determined by the Akaike Information Criterion.

Results: The measurement models functioned equivalently across maltreatment and comparison groups. The best-fitting model for both groups was a model where correlated internalizing and externalizing factors were each hierarchically related to their respective three first-order factors. A multiple-group analysis with this model further examined the equivalence of the parameters between groups. Maltreated and comparison groups showed very similar patterns except that the correlation between internalizing and externalizing factors was approximately twice as strong in the maltreated group.

Conclusions and Implications: The findings suggest that two higher-order factors optimally explain the relationships between depression, posttraumatic anxiety, generalized anxiety, aggression, delinquency and impulsivity in a large group of young adolescents. Both maltreated and comparison groups exhibited a similar preference for the two-factor hierarchical model, as opposed to a single higher-order factor, or models featuring only first-order factors. The findings represent the model identified by Achenbach. It suggests that symptom covariation or comorbidity is a natural consequence of the existence of higher order processes that unify apparently discrete symptoms clusters. Attention may be profitably directed towards investigations of higher-order processes rather than discrete diagnoses. Maltreated youth may feature a more highly covarying form of distress. Future investigations could examine if maltreatment experience catalyzes a stronger relationship between internalizing and externalizing behavior, possibly warranting special attention on treating higher-order processes.