Cumulative and Interaction Effects of Risk Factors On Depressed Mood Among Adolescents in Taiwan

Introduction: According to the ecological transactional model (Cicchetti & Toth, 1998), the transaction of multi-level risk factors is a contributor to depressed mood. Assessing the cumulative effects and interaction effects of risk factors at individual, family, and school levels allows researchers to gauge multiplicative and amplified impacts of a cluster of risk factors on adolescent psychosocial adjustment (Burchinal et al., 2000; Kraemer et al., 2001). Based on a nationally representative sample of adolescents in Taiwan, the study investigated the risk factors associated with adolescent depressed mood. The risk factors examined included family economic strain, parent-adolescent conflict, negative peer relationships, and academic expectation stress from teachers and parents. The study tested two hypotheses. First, a cumulative risk index, referring to the number of risk factors, was hypothesized to exert multiplicative effects on adolescent depressed mood. Second, two-way interactions of all risk factors were examined as to determine whether risk factors amplify each other.

Methods:Using a multi-stage, random cluster sampling method, this cross-sectional study collected self-report questionnaires from 1,306 students in Taiwan. Multiple logistic regressions were performed to examine relationships between risk factors, the cumulative risk index and adolescent depressed mood. To assess the two-way interaction effects of risk factors, hierarchical linear regression analysis and a simple slope analysis were conducted.

Results:Study participants’ mean age was 16.5 (SD = 0.90), and slightly more than half (53.2%) were females. Adolescent depressed mood was measured by the Center for Epidemiologic Study Scale (CES-D). Study participants overall did not exhibit a substantial extent of depressed mood (Mean = 19.55, SD = 10.77). However, around one fifth of the participants (20.9%) met the criteria of moderate-severe depressive symptoms (CES-D > 29). Multiple linear regression analyses suggested that all hypothesized risk factors significantly correlated with adolescent depressed mood. Among these factors, negative peer relationships accounted for the largest variance in adolescent depressed mood.

Multiple logistic analyses indicated that compared to adolescents who had no presenting risk factors, the odds of having moderate-severe depressed mood were 4.49 (95% CI: 2.61-7.73), 11.64 (95% CI: 6.78-20), 31.59 (95% CI: 17.38-57.40), and 175.6 (95% CI: 37.22-858.57) times greater in the adolescents having 1, 2, 3, 4 risk factors, respectively. The hierarchical regression model indicated that the interaction term of family economic strain x academic expectation stress was significant. The simple slope analysis revealed that the association between academic expectation stress and depressed mood increased when the family economic strain was high.

Conclusions and Implications: Study findings suggest that the ecological perspective is helpful in guiding researchers to comprehensively examine risk factors associated with adolescent depressed mood. The cumulative risk index can be used as a simple indicator to identify the high-risk group. Taiwanese adolescents who experience multiple stressors are in need of interventions which not only address risk factors across ecological levels, but in particular attend to adolescents’ family economic disadvantages. Further studies are warranted to investigate factors which can protect adolescents from cumulative and interaction effects of risk factors.