環境因素如溫度和空氣污染物等,已被認定為影響人類健康的重要因素,其中某些疾病的死亡率和發病率,可能與氣候改變或空氣污染物的濃度有關。過去學者在進行此一議題的研究前,多數聚焦於都會區民眾的死亡率與氣候之關係,並以卜瓦松迴歸模式來進行分析。若要以疾病發生率而非死亡率做為影響健康之指標,且針對不同地區民眾進行研究,則必須考慮到不同地區之間的異質性。 各地區之異質性除了與各地區的氣候特性、空氣品質有關外,還跟各地區的人口特質有關。例如,不同地區之民眾對於氣候的耐受力可能不同,不同地區之民眾的健康習慣也可能不同;否則,跨地區的比較可能不夠精確。因此,本研究認為,分析的模式應適度納入此變異的考量,允許各地區有其特質,以便進行後續分析。我們根據每日最低氣溫的第5百分位和每日最高氣溫的第95百分位,對各個區域定義了不同的極端溫度閾值。此外,我們採用臺灣本島的七個空氣品質監測區範圍定義,以控制空氣污染濃度的差異,並利用廣義線性模型,及納入同一地區不同日數之相關性,來評估極端氣溫相關因子對健康之影響;並以腦血管疾病和缺血性心臟病新發生病例數為例。模式採用負二項(negative binomial)模式來納入多餘變異(over dispersion),並利用擬概似函數訊息(quasi-likelihood information criterion, QIC)選擇了最恰當之相關性矩陣。結果顯示,在考慮了各地區一般情況地氣溫條件及定義相對情況下的高低溫之後,有些地區的相對極端溫度顯著地影響腦血管疾病和缺血性心臟病的新發生病例數。如相對極端溫度對高屏和花東空品區有不良影響,其它地區則只受相對極端高溫或相對極端低溫之影響。為了證明模式的配適表現,我們也與其他目前研究常使用的統計模式進行比較,從模式的配適表現來看,因為我們的模式能較為公正地為每一區域民眾說明相對極端氣候的影響,優於其他分析模式,能提供更好的估計。
The environmental factors like temperature and air pollutants have been recognized as important factors for human health, where mortality and morbidity of certain diseases may be related to the abrupt climate change or the air pollutant concentration. In order to clarify such effects and to account for the heterogeneity between different regions, we used different threshold values of extreme temperature, based on the 5th percentile of daily minimum temperature and the 95th percentile of daily maximum temperature, to define relative extreme temperature for each region. In addition, we adopted the definition of seven air quality regions in Taiwan to accommodate the differences in air pollution concentrations. We then applied a generalized linear model to evaluate the disease-associated factors on new cases of cerebrovascular disease and ischemic heart disease. Results showed that the local climate factors significantly influenced the occurrence of new cases of cerebrovascular disease and ischemic heart disease in certain regions, but not all. For instance, the high and low temperature extremes are associated with more cases of cardiovascular diseases in Kao-Ping, and Hua-Tung regions. However, in North, Yun-Chia-Nan, and Kao-Ping regions only high temperature extreme are associated with more new cases of ischemic heart disease; while only low temperature extreme correlates more new cases of ischemic heart disease in Yilan air quality region. To demonstrate the performance of our model, we also compared with other current statistical models. Our model provides better estimates and outperforms than other analyses.