Introduction Social integration could constitute a pathway and/or a buffer in the association between stressful life events in childhood (SLEC) and depression in adulthood and this may differ according to the adults' socioeconomic status and neighbourhood environment. Our objective was to (1) Investigate the association between SLEC and depression in adulthood, (2) Test the mediating and/or moderating effect of social integration in the association between stressful life-events in childhood and depression in adulthood.

Methods Data. SIRS cohort study, a longitudinal epidemiological survey of 4560 adults of the Paris metropolitan area conducted since 2005. Data collection in 2005, 2007 and 2010. Outcome: New cases of depression identified by the Mini-Diag in 2010. Individual variables: SLEC, indicators of social integration (social support and social capital), socioeconomic status. Neighbourhood characteristics: socioeconomic type of the neighbourhood of residence. Statistical analysis: multilevel logistic regression and structural equations models.

Results Stressful life events (social or psychological conditions) in childhood were associated with the occurrence of a depression in 2010 (RMSEA=0.079; NFI=0.94). After including these events in a fully adjusted model, they were associated with a lack of social support in 2005, which in turn was associated with depression 5 years later, but with a different magnitude according to individuals and/or neighbourhood SES.

Conclusion Identification of pathways and buffers between stressful life events in childhood and depression in adulthood contributes to the knowledge for a comprehensive model of the intergenerational transmission of social inequalities in mental health and could guide the mental health public policies in specific sub-groups of population.