Timing Matters: Examining Linkages Between Age of Maltreatment and Onset of Offending

Background and Purpose: A large body of evidence has shown that exposure to childhood maltreatment is associated with poor behavioral outcomes, including juvenile delinquency and adult crime. Less attention has been paid to the impacts of adolescent maltreatment. Likewise, no known studies have examined whether the developmental timing of maltreatment is linked to the age that offending behaviors emerge (i.e. onset) from adolescence through early adulthood. Addressing these gaps in the literature, this study considers two related questions: (1) Does the onset of offending among maltreated children and adolescents, respectively, differ from non-maltreated participants? (2) Are maltreated children and adolescents more likely to initiate offending at particular developmental stages (early adolescence; late adolescence; early adulthood) than non-maltreated participants?

Methods: This investigation includes 1,411 participants from the Chicago Longitudinal Study, a prospective panel investigation of predominantly African American participants who were born in 1979 or 1980. Administrative child maltreatment records were used to create dichotomous measures indicating if a subject experienced: (a) childhood maltreatment (< age 12; n=133), and/or (b) adolescent maltreatment (age 12-17; n=78). Juvenile and adult offending histories were gathered through age 27 from county, state, and federal databases. These records were used to create a continuous outcome representing a participant's age of first arrest. This variable also served as the basis for constructing a polytomous outcome differentiating four groups: (1) non-offenders, (2) early-onset offenders (first juvenile petition, < age 15), (3) late-onset offenders (first juvenile petition, age 15-17), and (4) adult-onset offenders with an arrest conviction but no juvenile arrest record. Analyses predicting age of first arrest were performed using Cox proportional hazards regression. Models analyzing the polytomous measure of offending were conducted using multinomial logistic regression with non-offenders serving as the reference group. All analyses controlled for an array of characteristics associated with maltreatment and/or crime (e.g., family poverty).

Results: Adjusted hazard ratios from Cox regression models indicated that childhood maltreatment, Wald (df=14, p=.002) and adolescent maltreatment, Wald (df=14, p<.001) were both associated with an earlier onset of offending. Results from multinomial logistic regressions revealed that childhood (OR=3.34, p<.001) and adolescent (OR=5.58, p<.001) maltreatment increased the likelihood of early-onset juvenile offending. Childhood (OR=2.12, p<.01) and adolescent (OR=4.40, p<.001) maltreatment also increased the likelihood of late-onset juvenile offending. However, neither childhood nor adolescent maltreatment was significantly associated with adult-onset offending. Secondary analyses demonstrated that initial estimates were robust to alternative model specifications, such as covarying maltreatment frequency (i.e. number of reports) and analyzing different ways of defining childhood and adolescent maltreatment.

Implications: Results from this study suggest that both childhood and adolescent maltreatment precipitated an earlier onset of offending behaviors. Given the known associations between early-onset delinquency and the frequency and persistence of later offending, these findings indicate that experiencing maltreatment at any age may elevate the risk of adopting more serious patterns of maladaptive behavior. This evidence has implications for future research examining offending trajectories among maltreated children and adolescents as well as designing interventions to prevent their engagement in delinquent behaviors.