Objective To identify modifiable risk factors in the development and persistence of depression. Design Prospective observational cohort study. Setting Healthy midlife women from well-women clinics in a Singapore tertiary hospital enrolled from September 2014 to October 2016. Population Enrolled participants at baseline (n=1201) who were followed-up 6.6 years later. Methods Sociodemographic characteristics and medical history were collected using a questionnaire. Conditions such as urinary incontinence and sleep quality were ascertained using validated questionnaires, and physical performance was objectively-measured. Changes in health conditions and physical performance measures were calculated by subtracting baseline from follow-up values. Modified Poisson regression analyses examined associations between risk factors and outcomes. Main outcome measures Depressive symptoms were defined using the Center for Epidemiological Studies-Depression (CES-D) scale score ≥16 and/or use of anti-depression medication. Incident cases were defined as new cases that emerged during follow-up, while persistent cases were present at both timepoints. Results After 6.6 years, depression developed in 120 out of 741 (16.2%) participants, while persistent depression was present in 34 out of 146 (23.3%) with baseline depression. Urinary incontinence (adjusted relative risks, aRR: 1.51, 95% confidence interval, CI: 1.06-2.13), poorer perceived health (1.78, 1.25-2.52), lower education attainment (1.72, 1.02-2.90), and alcohol consumption (1.81, 1.11-2.96) at baseline were independently associated with incident depression. Improvements in one-leg stand duration over 6.6 years reduced the risk of incident depression (0.98, 0.96-0.99). Baseline poor sleep quality was a risk factor for both incident and persistent depression. Conclusions Poor sleep, urinary incontinence and weak physical performance are potentially modifiable risk factors for incident and persistent depression.