Patterns and Dimension of Child Maltreatment: A Latent Class Analysis

Background: Many studies have reported on the negative effects of child maltreatment on adolescent and adult functioning regarding offending behaviors, mental health, and substance use problems. In this field of research, some researchers have focused on specific types of child maltreatment that affect the developmental outcome of interest (Grogan-Kaylor & Otis, 2003; Kim & Cichetti, 2010; Yun, Ball, & Lim, 2011). Other researchers, such as Robert F. Anda and Vincent J. Fellitti, emphasize the cumulative effect of multiple child maltreatment. Thus, in their Adverse Child Experiences (ACEs) study, Anda and Fellitti measure child maltreatment as the aggregate sum of multiple maltreatment experiences. This study aims to explore the underlying patterns and dimensions of child maltreatment. By identifying latent variables that better capture the child maltreatment experience of individuals, this project addresses not only specific types of maltreatment but also co-occurring maltreatment.

Methods: Data come from the Seattle Social Development Project (SSDP), a longitudinal study that looks at the development of positive and problem behaviors among adolescents and young adults. This study started during the 1980’s with 18 elementary schools in the Seattle Public School District, over-sampled in high-crime neighborhoods. The survey on child maltreatment was based on the Childhood Trauma Questionnaire short form (CTQ-SF, Bernstein, et al. 2003), a 28-item self-report, retrospective inventory of child maltreatment history that assesses for physical abuse, emotional abuse, sexual abuse, emotional neglect, and physical neglect. This survey was administered when participants were 24 years old. Using a 5-point likert-type scale ranging from “never true” to “always true”, the CTQ-SF scales demonstrate good reliability across samples (Bernstein et al., 2003). In order to run Latent Class Analysis, all items above were dichotomized.

Results: Based on the model with the above 8 items, 6-latent classes were identified. Chi-square test of model fit was used to check the goodness-of-fit. Individuals in class 5 were identified as those without child maltreatment history. Class 1 individuals were identified as victims of sexual abuse only. Class 2 individuals indicated experiences of multiple types of maltreatment. With the highest probability of physical neglect due to parental drug use, it is likely that these individuals became victims of other maltreatments as a result of parental drug use (i.e. drug dealers or parent’s partners as abuse perpetrators while parents neglected). Class 3 individuals were identified as victims of co-occurring physical abuse and emotional abuse. Class 4 individuals were mainly victims of neglect, with low levels of other types of maltreatment. Finally, class 6 individuals were victims of long-term co-occurring physical, sexual, and emotional abuse.

Conclusion: Different types as well as co-occurrence of maltreatment categorize individuals into different classes. Therefore, child maltreatment can neither be understood as an aggregate sum of occurrences nor separated into specific types of maltreatment. It is important to look at both aspects. This has important implications for researchers connecting child maltreatment experience to various developmental outcomes as well as practitioners working in child welfare settings.