Family Resilience of Chinese Rural Left-behind Adolescents: The Predictors and Outcomes from a Latent Profile Analysis

Background and Purpose: Family resilience is a significant factor in enhancing the quality of life for adolescents, particularly for Chinese rural left-behind adolescents who face multiple adversities challenging their quality of life. Nevertheless, few studies have examined whether distinct family resilience subgroups of rural left-behind adolescents exist, and which ecological factors differentiate these different profiles of family resilience. This study aims to investigate family resilience characteristics in a sample of Chinese rural left-behind adolescents, identify the observable family factors that could distinguish different profile of family resilience, and examine the relationships between these profiles and distal outcome, namely quality of life.

Methods: A cross-sectional study was conducted at a rural area in Sichuan province, China, in 2023. The sample included 790 left-behind adolescents aged between 10 to 16 years, recruited from three schools (47.8 % female, Mage = 13.49 ±1.18). Four models were constructed to examine family characteristics, including family structure (model 1), parent- adolescent communication frequency (model 2), parent- adolescent communication channels (model 3), and caregiver characteristics (model 4). Descriptive and correlation analyses were performed first. Subsequently, Latent Profile Analysis (LPA) was employed to identify distinct subgroups based on family resilience, and Regression mixtures analyses were applied to investigate the predictor and outcomes associated with different family resilience profiles. Data analyses were conducted using SPSS and Mplus.

Results: Three distinct subgroups of rural left-behind adolescents based on family resilience were identified: “high family resilience” (41%), “moderate family resilience” (50%), and “low family resilience” (9%). In model 1, we found that members of the three profiles differed in gender, sibling number, birth order, and parental marital status. Model 2 suggested that whether paternal-adolescent or maternal-adolescent communicate more than once a week can distinct high and moderate family resilience groups with low family resilience group. Model 3 indicated that gender, adolescents communicate with father online or through telephone are associated with higher probabilities of belonging to the high and moderate family resilience group. Model 4 highlighted differences among the profiles in the relationships between adolescents and their main caregiver. Furthermore, the second regression mixture model revealed that all three groups of adolescents exhibited significant levels on the general quality of life and four subdomain quality of life.

Conclusions and Implication: This study underscores the significance of identifying different levels of family resilience among rural left-behind adolescents according to observable family level predictors. These findings can facilitate social worker, teachers, and other service providers in identifying low family resilience groups and designing tailored parental and family intervention. By understanding the distinct profiles of family resilience, targeted support can be provided to enhance the quality of life of rural left-behind adolescents and their families.