Predictors of admission and readmission to hospital for major depression: A community cohort study of 52,990 individuals

•We identified predictors of first time hospitalisation for major depressive disorder (MDD); and MDD readmission thereafter.•A broad range of social, health and lifestyle factors were associated with de novo admission for MDD.•Only older age and a prior non-depression related psychiatric hospital episode were associated with readmission for MDD.

BackgroundOur current knowledge about predictors of admission and re-admission to hospital as a result of major depressive disorder (MDD) is limited. Here we present a descriptive analysis of factors which are associated with MDD hospitalisations within a large population cohort.MethodsWe linked participants of the Scottish Health Survey (SHS) to historical and prospective hospital admission data. We combined information from the SHS baseline interview and historical hospitalisations to define a range of exposure variables. The main outcomes of interest were: (1) first time admission for MDD occurring after the SHS interview; and (2) readmission for MDD. We used Cox regression to determine the association between each predictor and each outcome, after adjusting for age, gender and deprivation quintile.Results52,990 adult SHS participants were included. During a median follow-up of 4.5 years per participant, we observed 530 first-time admissions for MDD. A relatively wide range of factors - encompassing social, individual health status, and lifestyle-related exposures - were associated with this outcome (p<0.05). Among the 530 participants exhibiting a de novo admission for MDD during follow-up, 118 were later re-admitted. Only older age (over 70) and a prior non-depression related psychiatric admission were associated with readmission for MDD.LimtationsMDD was defined using records of International Classification of Disease hospital discharge codes rather than formal diagnostic assessments.ConclusionThese findings have implications for mental health service organisation and delivery and should stimulate future research on predictive factors for admission and readmission in MDD