本研究整合地理資訊系統、遙感探測技術以及統計模型，以探討環境綠蔽度與雙相 情緒障礙症之關聯。方法：本研究經由健保資料庫承保抽樣歸人檔以獲取2003年至2012年台北 都會區（含台北市及新北市）各鄉鎮雙相情緒障礙症之季發生率資訊；並透過MODIS之全球常 態化差異植生指標（Normalized Difference Vegetation Index, NDVI）資料庫，以量化環境綠蔽 度；最後使用廣義加乘混合模型，在校正社會經濟、人口學、環境、共病症等可取得之風險因 子與時空自相關情況下，探討環境綠蔽程度與雙相情緒障礙症發生率之關聯性。結果：研究結 果發現，NDVI與雙相情緒障礙症發生率之關聯達到統計上之極顯著水準（p<0.01），係數值 為-0.51（95%CI: -0.68, -0.48），取自然對數後得相對風險值為0.60（95%CI: 0.51, 0.71），代表 當鄉鎮之NDVI每上升一個四分位距（IQR=0.33）時，雙相情緒障礙症之發生率會降至原來的 0.60倍。結論：由此可知，環境綠蔽度的增加與雙相情緒障礙症發生率的減緩具有正向的關聯 性。（台灣衛誌 2018；37(5)：514-526）

Objectives: This study investigated the association between surrounding greenness and bipolar disorder in the Taipei metropolitan area (Taipei City and New Taipei City) using a geographic information system and remote sensing technologies. Methods: The seasonal incidence of bipolar disorder from 2003 to 2012 for each township in the Taipei metropolitan area was calculated based on medical records extracted from the Longitudinal Health Insurance Database 2000. Surrounding greenness information was collected as the normalized difference vegetation index (NDVI) provided by NASA’s Moderate Resolution Imaging Spectroradiometer. The generalized additive mixed model was applied to examine the relationship between surrounding greenness and bipolar incidence after adjusting for socioeconomic status, demographic characteristics, environmental factors, and comorbidity. Results: Accounting for potential confounding variables, we found a statistically significant negative relationship between NDVI and bipolar incidence (regression coefficient: −0.51; 95% confidence interval [CI]: −0.68, −0.48; p<0.01). The adjusted relative risk for bipolar incidence was 0.60 (95% CI: 0.51–0.71) per interquartile range increment (0.33) on NDVI. Conclusions: Surrounding greenness is statistically associated with bipolar incidence in the Taipei metropolitan area. (Taiwan J Public Health. 2018;37(5):514-526)