Preventing depression in high-income countries—A systematic review of studies evaluating change in social determinants

We conducted a systematic review to examine whether changes in social determinants can contribute to the prevention of depression, in order to provide input for policy development and to highlight research gaps. Social determinants were defined as the structural conditions in which people live that shape their health and were categorized according to whether they pertained to societal arrangements, material resources distributed through these arrangements, or social resources that follow from interactions between people. To capture all relevant evidence we included studies that measured depressive disorders, depressive symptoms, psychological distress, mental health and prescription rates of antidepressants. We searched three databases (Medline, Embase and Psychinfo) from their inception till December 2022 and supplemented our search by reference and citation searching of the included studies. Studies were synthesized qualitatively and we used the Validity Assessment tool for econometric studies to assess study quality. Prospero submission number CRD42021236132 A total of 31,103 titles were identified, 135 studies met our inclusion criteria. The majority of studies were conducted in the United States (n = 45) or the United Kingdom (n = 39). Studies used longitudinal data (n = 61); repeated cross-sectional data (n = 20); or evaluated an intervention study (n = 7). Study designs included natural experiments (n = 19), while some used propensity score matching to construct a quasi-experiment (n = 11). Analysis methods included difference-in-difference approaches (n = 30) or regression analysis in varying forms. We found evidence that strategies that promote paid employment and parental leave policies can reduce risk of depression whereas reduced entitlements to social welfare (particularly when accompanied by obligations to enter employment), loss of income, instability of housing and collective insecurity increase depression risk. A number of studies examined moderation by gender, age category or ethnicity and of these gender was the most commonly observed moderator. Few studies tested underlying causal mechanisms with formal mediation analyses. These studies provide important indications of how intervening on social determinants of health can shape risk for depression. However, the included studies do not fully capture the complexity of the relationships between determinants and the mechanisms driving them. Future studies could take this into account, for instance by using systems approaches.