1 Department of Health Sciences, Université du Québec à Rimouski, Rimouski, QC, Canada. william_gilbert@uqar.ca.
2 Substantive-Methodological Synergy Research Laboratory, Department of Psychology, Concordia University, Montreal, QC, Canada.
3 Optentia Research Unit, North-West University, Vanderbijlpark, South Africa.
4 Department of Education, University of Helsinki, Helsinki, Finland.

Numerous studies have sought to determine whether low self-esteem acts as a risk factor for depressive symptoms (i.e., a vulnerability model) or whether depressive symptoms lead to a decrease in self-esteem (i.e., a scar model). Although both models have received some support, very little research has: (a) addressed this question across critical life transitions likely to modify this pattern of associations, such as the transition to adulthood; (b) sought to identify the psychological mechanisms (i.e., mediators) underpinning these associations. The present study was designed to address these two limitations, focusing on the directionality of the associations between depressive symptoms and self-esteem from mid-adolescence to early adulthood while considering the role of motivational factors, namely mastery (intrinsic/extrinsic) and performance (approach/avoidance) goals as conceptualized in achievement goal theory. A sample of 707 Finnish adolescents aged 15-16 (52.1% boys) was surveyed six times up to the age of 25. Results from a cross-lagged panel model (CLPM) revealed that depressed individuals were more likely to have low self-esteem, although self-esteem protected against depressive symptoms between ages 16-17 to 20-21. Moreover, while self-esteem promoted mastery-extrinsic goals which in turn reinforced self-esteem, depressive symptoms promoted performance-avoidance goals which led to more depressive symptoms and lower self-esteem. Overall, these findings highlight (1) the long-lasting negative consequences of depressive symptoms on self-esteem and (2) the crucial role played by academic motivation in explaining the development of depressive symptoms and self-esteem over time. In turn, these results help refine the vulnerability and scar models, and suggest that motivational factors should be considered in prevention and intervention efforts among young populations.