Green space exposure predicts manic episodes in bipolar disorder: one year longitudinal analysis using smartphone-derived GPS data

Urban living is linked to increased psychiatric morbidity, while green spaces may protect mental health via stress-buffering effects. Yet, the role of green space exposure in the onset of affective episodes in bipolar disorder (BD) remains unclear. Using 12 months of GPS data from 29 participants with BD-I/II, we quantified daily exposure to green spaces (green area & NDVI; covariates: population density & imperviousness). Affective status (euthymia, depressive or (hypo)manic episode) was assessed via SCID-based biweekly clinician interviews (26 assessments per participant). Multilevel logistic regression examined moving averages and temporal variance of green space exposure during prodromal and episode weeks relative to euthymia. Reduced mean green space exposure and increased exposure variability robustly statistically predicted prodromal phases of (hypo)manic episodes, with variance effects persisting into acute mania. In contrast, depressive episodes showed limited prodromal predictability. These findings propose dynamic environmental signatures as early-warning signals and modifiable risk factors in BD.