Developmental pathways model of depression heterogeneity

Depressive disorders are a leading cause of disability, their first onsets peaks between ages 14-24, and they are a major concern in adolescent and emerging adult mental health. Although the majority of people with depression quickly recover with few, if any, recurrences, many suffer from a protracted and highly impairing illness. These time-limited depression (TLD) and chronic-intermittent depression (CID) groups appear to differ in developmental etiology, outcome, and treatment response, yet are rarely distinguished in research, gravely confounding the literature. To make progress in reducing depression burden, it is critical to delineate the developmental pathways and adult outcomes of these groups. We propose to test a dual developmental pathways model that hypothesizes that TLD results from severe life stressors. In contrast, CID develops from a familial liability to low reward responsiveness and childhood adversity, which produce blunted reward system functioning. Reward functioning may deteriorate further as the illness progresses. No study to date has examined whether exposure to severe life events and reward responsiveness prior to the first-onset of depression differentiate between these subtypes, which is essential to evaluating their role in etiology. Moreover, research has not traced the trajectory of reward function after CID onset to determine if it continues to worsen. We will build on an ongoing longitudinal study that has followed 550 young women (the high-risk demographic subgroup) from age 14 to 21 years, with > 85% retention. We propose to follow the cohort through age 26 (2 years beyond the highest-risk period) to accurately distinguish CID, TLD, and no depression groups. This will allow us to test the dual developmental pathways model. We also will trace trajectories of reward responsiveness assessed using self-report, electrophysiology, and a novel magnetic resonance imaging measure of dopamine signaling. Moreover, we will assess young adult outcomes (occupational and relationship functioning, suicidality, and mental health service utilization) to evaluate the public health significance of these subtypes. In addition, we will test the feasibility of developing a predictive algorithm for CID using a comprehensive battery of premorbid risk factors. Most research on depression to date conflates two markedly different populations, CID and TLD. We will test a model of their development and etiology by closely following a large sample from early adolescence to young adulthood, thus starting before and ending after the highest-risk period for depression onset. No previous study that begun prior to the risk period for depression has charted the course of depressive disorders at the level of resolution needed to distinguish CID from TLD. Delineation of these groups can greatly reduce the diagnostic heterogeneity that plagues depression research and treatment, which will increase the statistical power and interpretability of studies, help guide treatment development and selection, and direct services to the subgroup that most needs them.