Methods: We use the National Longitudinal Study of Adolescent and Adult Health (ADD Health) data. The analysis sample include 10,784 participants who participated in three waves of data from adolescence to young adulthood (average age 28). The dependent variable is depressive symptoms measured by the 10-item Center for Epidemiologic Studies-Depression Scale. The independent variable is ACEs. Nine binary variables are used to capture the following adversities respectively: physical abuse, physical neglect, emotional abuse, emotional neglect, sexual abuse, or growing up in the household with problems of divorce/separation, incarceration, mental illness, or suicide attempt. The mediating variable is self-esteem measured by Rosenberg’s short form scale. Demographic and socioeconomic characteristics are included as covariates (age, gender, race and ethnicity, education, marital status, public assistance, homeless experience). We performed a Latent Class Analysis (LCA) to identify the subgroups of ACES and a path analysis to simultaneously estimate the direct and indirect effects of ACEs on depressive symptoms through self-esteem. All analyses are performed using sampling weights to account for the complex survey design and non-response bias.
Results: LCA classifies three subgroups of ACEs, child maltreatment (19%), household dysfunction (15%), and low adversity (66%). The two groups of child maltreatment and household dysfunction are more likely socioeconomically disadvantaged than low adversity group. For instance, those with college or higher degree are fewer in household dysfunction (50.16 %) and child maltreatment (62.50%) groups than low adversity group (68%). Proportions of public assistance recipients are higher in child maltreatment (29.47%) and family dysfunction (38.66 %) groups than low adversity group (18.37%). Path analysis presents that ACEs is associated with depression and that self-esteem significantly mediates the associations. Most striking findings are that child maltreatment group reports significantly lower levels of self-esteem and higher levels of depression than household dysfunction and even low adversity groups.
Conclusion and Implications: This study provides important evidence that the latent construct of ACEs is comprised of three subgroups and that the detrimental pathway from ACEs to depressive symptoms occurs by decreasing self-esteem. It is important to note that child maltreatment appears to result in worse effects on self-concept and mental health. We discuss research and practice implications that should be addressed to prevent enduring effects of early adversities.