Pathways between Child Maltreatment, Psychological Symptoms, and Life Satisfaction: A Network Analysis in Adolescent Inpatients

Abstract

Child maltreatment is a risk factor for mental disorders and decreased life satisfaction during adolescence. We investigated whether child maltreatment would link to life satisfaction both directly and through psychological symptoms, whether these relations would change from admission to discharge after treatment, and which types of maltreatment, symptoms and facets of life satisfaction would be most influential in adolescent inpatients with internalizing mental disorders. N = 896 adolescent receiving inpatient psychotherapeutic treatment completed questionnaires on child maltreatment experiences, current psychopathology and subjective life satisfaction at admission and discharge (n = 765). Main diagnoses were affective (n = 322), eating (n = 447), obsessive-compulsive (n = 70) and anxiety disorders (n = 57). Network models of child maltreatment, psychopathology and life satisfaction nodes were estimated at admission and discharge and compared using network comparison tests. Potential causal shortest pathways were investigated using directed acyclic graphs.Network models were stable with no significant differences between admission and discharge. Strongest nodes of each cluster were "emotional abuse" (child maltreatment), "worthlessness", "thinking about dying" and "feeling lonely" (psychopathology) and "satisfied with life" (life satisfaction) at both admission and discharge. Emotional neglect showed direct connections to life satisfaction, indicating its relevance for therapeutic interventions. At both admission and discharge, "sexual abuse" indirectly predicted lower life satisfaction through psychological symptoms. In conclusion, child maltreatment is directly and indirectly connected to life satisfaction in adolescents with mental disorders. Emotional abuse and neglect were especially important in linking child maltreatment to life satisfaction and psychopathology.

Keywords: Adolescents; Childhood trauma; Depression; Eating disorder; Life satisfaction.

Fig. 1

Estimated networks. (A) admission (n = 896) and (B) discharge (n = 765) networks. Nodes (circles) represent symptoms, maltreatment types and facets of life satisfaction. Width and saturation of an edge (lines connecting nodes) represent the strength of the correlation, whereas color indicates positive (blue) or negative (red) correlations. Node color represents belonging to either symptom, child maltreatment or life satisfaction communities. In the variability network (C), edges represent the standard deviation of edges between two specific nodes across admission and discharge networks. See Table 1 for full node descriptions and communities (i.e., scales). A high-resolution version of the figure can be found here: https://osf.io/pyke9/

Fig. 2

Standardized node strength (sum of all edges of a given node to all other nodes) and expected influence centrality (summed weight of a node’s edges shared with remaining nodes) for admission and discharge networks. A high-resolution version of the figure can be found here: https://osf.io/pyke9/

Fig. 3

Shortest directed pathways from child maltreatment to life satisfaction nodes (circles) at (A) admission (n = 896 patients) and (B) discharge (n = 765 patients) based on directed acyclic graphs (DAGs) of N = 10,000 bootstrapped samples. Node color represents belonging to either symptom, child maltreatment or life satisfaction communities. Edges (arrows connecting nodes) represent potentially causal pathways between nodes exceeding the empirically determined threshold were retained; direction of edges was included if present in at least 51% of the bootstrapped samples. Edge thickness represents confidence in prediction. DAGs are combined for all forms of child maltreatment (pathways from emoabu–emoneg to SWLS are represented by dashed lines, paths associated with sexual abuse are shown with solid lines). A high-resolution version of the figure can be found here: https://osf.io/pyke9/