A growing body of literature has elaborated the life-course and cohort patterns in the relationships between social factors and depression in Western societies. Nonetheless, far less research has focused on whether inequalities in social status have caused the inequality in misery over the life course in Eastern societies such as Taiwan, which is a collectivist society that has undergone tremendous social change. This research examines the life-course depression trajectories, with taking cohort variations into consideration, and assesses the multidimensional effects of education on depression in a network perspective. This study is based on the nationally representative samples from the repeated cross-sectional Taiwan Social Change Survey and from the longitudinal Survey of Health and Living Status of the Middle Aged and Elderly in Taiwan.

Results reveal a U-shaped aging trajectory in depression: depression declines in early adulthood, bottoms out in middle age, and then rises again in late life. This trajectory is the composite outcome established by factors associated with historical trends in education, differential survivals, life stages, health decline, and maturity. Moreover, the direction of the trajectory depends on education. For the well-educated Taiwanese, depression decreases from early adulthood to middle life and maintains relatively stable in old age. For the less educated, depression increases steeply over the life course. Taken together, the education-based disparity increases with age and the pattern even strengthens across more recent cohorts, consistent with respectively the cumulative advantage theory and the rising importance theory. Although late-life convergence is found in cross-sectional analyses, aging vector analyses with FIML estimation and Gompertz survival analysis suggest that selective mortality is the plausible reason.

Meanwhile, education is not the only root cause of psychological well-being in Taiwan. Social relationships factors—such as children's education, co-residence, social support, and familial negative interaction—also demonstrate substantial influence on depression, but mediate educational effects slightly. However, in the aging vector analyses, education is the resource that consistently displays negative coefficients with respect to the slope of depression. Consistent with the resource substitution theory, educational effects are greater for those in disadvantageous statuses. Therefore, increased education is the most specific resource that suppresses the progression of depression over the life course and under difficult times.