Trajectories and predictors of child abuse in Chinese children aged 4–7 years: A growth mixture model analysis

Trajectories of child abuse and associated predictors have not been fully elucidated to date, particularly under the Chinese cultural context. The current study attempted to explore trajectories of child abuse and predictors in a Chinese community sample based on a longitudinal study. Three waves of data were collected from 521 caregivers who had children aged 4–7 years with caregiver-reported questionnaires. An unconditional growth mixture model analysis was used to examine developmental trajectories, and a conditional growth mixture model analysis was conducted to examine predictors for trajectories. The results showed that changes in child abuse, physical abuse, and emotional abuse over time were heterogeneous, including in the None to Low class (93.8 %, 93.6 %, and 92.4 %) and the Sustained Abuse class (6.2 %, 6.4 %, and 7.6 %). The children’s sex, caregivers’ education levels, family income, and caregivers’ childhood trauma predicted the classification of child abuse, physical abuse, and emotional abuse (β = −0.72 to 1.32, S.E. = 0.18–0.39, p <.01), e.g., girls and caregivers’ low childhood trauma predicted individuals belonging to the sustained abuse class of physical abuse, and these factors also predicted child abuse within this class (β = −1.28 to 2.85, p <.05). These findings suggest that different subtypes of child abuse may have different trajectories, and prevention strategies should be implemented based on specific subtypes. Additionally, the findings highlight that many more longitudinal studies should be conducted on child abuse and examine changes in abuse over time.