Green space structures and schizophrenia incidence in Taiwan: Is there an association?

A growing body of research indicates overall greenness offers potential psychological benefits. However, few studies have explored green space structures and their potential association with mental disorders. The aim of this study was to determine the existence of such an association. From two million randomly sampled people in Taiwan's National Health Insurance Research Database, we selected 3 823 patients that received a first-time diagnosis of schizophrenia from 2005–2016. Moreover, we used a geographic information system and a landscape index to quantify three characteristics of green space structures including area and edge, shape, and proximity. Additionally, we collected the normalized difference vegetation index and enhanced vegetation index data to reconfirm the association between overall greenness and schizophrenia incidence. We used the indices to determine individuals' exposure according to their residential township. Spearman's correlation analysis was conducted to select variables by considering their collinearity. Cox proportional-hazards models were applied to assess the relationship between green space exposure and schizophrenia incidence following adjustment for potential confounders, such as air pollution (NO2, SO2, ozone, and PM10), temperature, precipitation, and socioeconomic status, which are risk factors. We found a negative association between most green space structures indices and schizophrenia incidence. Our findings suggest that for green spaces, a larger mean patch area and edge density, higher complex (higher perimeter–area ratio), and greater proximity (higher contiguity index, aggregation index, and contagion index), may reduce the risk of schizophrenia. A sensitivity test and subgroup analysis revealed similar results.