Abstract
Multiple forms of abuse may co-occur, resulting in specific abuse typologies. A stratified random probability survey was conducted in Denmark with 4,718 participants, aged 24, from the 1984 birth cohort. A total of 2,980 interviews were successfully conducted. Latent class analysis was implemented using 20 categorical abuse experience items across four domains of childhood maltreatment. Logistic regression was conducted to ascertain whether abuse typologies could be differentiated by child protection status and gender. Four distinct abuse typologies were revealed: a non-abused group, a psychologically maltreated group, a sexually abused group, and a group experiencing multiple abuse types. Child protection status and female gender were predictive of group membership in certain abused groups compared to a non-abused group.
Acknowledgments
We thank the Danish National Research Foundation for generous research support.
Notes
a Variables combined due to similarity in question wording and low-level endorsement. See Christoffersen et al. (in press) for individual items.
Note. AIC = Akaike information criterion; BIC = Bayesian information criterion; SSABIC = sample-size-adjusted BIC; LRT = Lo-Mendell-Rubin adjusted likelihood ratio test.