台灣地區慢性肺部疾病盛行分析暨風險因子探討

Prevalence and Risks for Chronic Pulmonary Diseases in Taiwan

王玉純(Yu-Chun Wang)

指導教授 ： 宋鴻樟

共同指導教授 ： 林嘉明(Jia-Ming Lin)

國立臺灣大學/公共衛生學院/環境衛生研究所/博士(2007年)

https://doi.org/10.6342/NTU.2007.01303

摘要

慢性肺病(Chronic pulmonary disease)是全球近10年來盛行率逐漸升高的肺部疾病，其成因複雜，與宿主本質及外在環境因子有關，有需要就國人的現況進行探討。本研究之重點即利用現有全民健保資料庫、空氣污染和氣象監測資料以及疾病管制局所提供之病毒合約實驗室檢測資料，進行外在環境因子對本國成人慢性呼吸道疾病之近年影響研究，以期能獲知外在環境因子、肺炎、流行性感冒以及流行性感冒疫苗接種對疾病盛行影響之程度。
研究進行分為四個方向。首先，利用全民健保資料分析國內慢性呼吸道疾病最近7年之就醫盛行率，並探討疾病嚴重度與個案本質之相關性；第二，建立氣象因子（尤其是極端氣象）和空氣污染物等外在環境因子，以波以松回歸模式(Poisson regression model)估計對疾病就醫之相對危險性；第三，分析國內肺炎及流感地區別就醫之盛行率，並針對氣象因子、空氣污染物及流感病毒活性之間進行一相關性分析；第四，針對慢性肺部疾病患者，探討其肺炎及流感就醫發生對其後續慢性肺病發生之存活分析，計算其危險比。
慢性肺部疾病盛行率有逐年下降之趨勢，盛行率以台北最低，高雄、花東地區最高，年長者為主要危險人口，並以肺炎與流感為慢性肺病住院個案最顯著的疾病合併症，勝算比為2.16 (95% CI: 2.00-2.34, p < 0.0001)。肺炎及流感盛行率升高至1998-1999年之後便逐年下降，地區盛行率台中、高雄地區相近，台北最低，年齡15歲以下族群具有較高的風險，並以慢性肺部疾病為最相關的疾病合併症。
氣候因子相關分析結果顯示，平均溫度(average temperature)在5百分位以下時對台灣三大都會區的慢性肺病就醫風險最高，RR=1.18 (95% CI: 1.13-1.23, p < 0.0001)，其次是最高溫度(maximum temperature)在5百分位以下時，RR=1.11 (95% CI: 1.07-1.15, p < 0.0001)，最低溫度(minimum temperature)與就醫危險並未達顯著相關性。分地區比較，台北及台中地區均在極低溫時(<5%)對慢性肺部疾病有較高的激發風險，但是在高雄卻是偏高溫時影響較大。分疾病比較，支氣管/肺氣腫/氣喘(bronchitis/emphysema/asthma)對大氣環境反應較顯著，慢性阻塞性肺病(chronic obstructive pulmonary disease)則無。研究期間(1996-2002)就醫次數介於7-28次的族群（疾病嚴重度2）對大氣環境也較為敏感。空氣污染物對慢性肺病的影響，在台北、高雄地區以NOx最具相關，台中地區則和PM10及SO2的相關最大。此外，春秋季的低濕度、夏秋季的低風速及夏秋季高日照率，均是除了平均溫度之外顯著增加慢性肺部疾病就醫風險的氣象相關因子。除了慢性阻塞性肺病，當週肺炎及流感就醫對當週慢性肺部疾病就醫風險均達顯著相關。
低溫和台北地區的肺炎及流感就醫相關最大，當週相對危險1.31 (95% CI: 1.20-1.42, p<0.0001)，在高雄地區，相對危險為1.19 (95% CI: 1.09-1.31, p<0.0001)，相較之下，台中約延遲四週才會看到低溫的影響。低相對濕度對肺炎及流感就醫亦有顯著的風險存在。空氣污染物以NOx增加就醫的相對危險最具一致性，每上升1 ppb NOx，整合三都會區的相對危險為1.005 (95% CI: 1.003-1.007, p<0.0001)，台北地區1.004 (95% CI: 1.001-1.007, p= 0.0175)，台中地區1.005 (95% CI: 1.002-1.008, p= 0.0027)以及高雄地區1.011 (95% CI: 1.007-1.015, p< 0.0001)。此外，台北地區PM10濃度亦是會增加就醫危險之空氣污染，相對危險1.003 (95% CI: 1.002-1.004, p<0.0001)。
觀察肺炎及流感暴露對慢性肺病就醫的存活分析，冷季(11 - 4月)30天累積存活率低於熱季(5 - 10月)的。流行性感冒疫苗接種實施後高於疫苗接種計畫實施前，男性低於女性。在接種前，高雄地區的30天累積存活率最低為0.42，接種後以花東地區累積存活率最低，為0.31。同樣的，疫苗接種前以0-14歲及50-69歲累積存活率最低，但在接種後，60-69歲族群累積存活率上升最多，由0.42至0.54。細分不同的慢性肺病，肺炎及流感對慢性阻塞性肺病就醫的危險比最高，達11.7 (95% CI: 8.64-15.7, p <0.0001)，支氣管/肺氣腫/氣喘次之，危險比4.82 (95% CI: 4.60-5.05, p <0.0001)；對兩種疾病均有被診斷的族群，影響最小，危險比2.59 (95% CI: 2.42-2.78, p <0.0001)。校正年齡、性別、地區、疫苗接種、個案疾病分類、時間（年）及季節的影響後，肺炎及流感暴露對慢性肺病就醫之危險比為3.91 (95% CI: 3.76-4.07, p <0.0001)。
本研究總結高雄地區為三都會區慢性肺部疾病及肺炎及流感就醫盛行率較高地區，大氣環境的低溫對台北及台中地區的慢性肺部疾病就醫風險貢獻較大，但空氣污染對高雄地區貢獻較大。相較之下，肺炎及流感的暴露危險性一致，並遠高於大氣環境的貢獻，然而，肺炎及流感亦受地區性氣象及空氣污染情形所影響。

關鍵字

慢性呼吸道疾病 ； 肺炎及流感 ； 氣象 ； 空氣污染 ； 風險

並列摘要

Chronic pulmonary diseases (CPDs) are slowly progressive heterogeneous disorders of the airways with associated etiological factors remain nebulous and debatable. This study aimed to analyze the prevalence and risks of the diseases for the three metros, and Hualien/Taitung area in Taiwan. We used universal health insurance claims, measurements of air pollutants and weather, and virological surveillance data to clarify the relative importance of potential risk factors.
We first used the reimbursement claims file of an established cohort from the National Health Insurance programe to estimate the chronological prevalence of CPDs and pneumonia and influenza (P&I) and comorbidity associated with severity of these diseases by multiple polytomous regression analyses between 1996 and 2003. The Poisson regression model was used to estimate the risk for these two diseases associated with weather conditions and air pollution, etc. We also conducted survival analyses for patients to compare the difference between patients with and without pneumonia and influenza.
The chronological analysis for the population-based cohort demonstrated a dramatic pattern of declining prevalence of CPDs and P&I for the population in Taiwan two to three years after the launch of the nationwide health insurance. CPDs were the most common in the elderly but P&I were most prevalent in population aged 14 years and less that both with the highest rate in Kaohsiung, Taitung and Hualien and the lowest in Taipei. Pneumonia and influenza (P&I) were founded as the most important co-morbidities associated with the hospitalization for CPDs with an odds ratio of 2.16 (95% confidence interval (CI): 2.00-2.34, p < 0.0001).
The risk for CPDs was the highest when ambient average temperature was at extreme cold (less than 5 percentile in 3 metros) with an overall relative risk (RR) of 1.18 (95% CI: 1.13-1.23, p < 0.0001). This association was significant in Taipei and Taichung, but not in Kaohsiung. This atmospheric environment association was significant for both bronchitis/emphysema/asthma (BEA) and severity of the disease (7-28 clinic vists), but not the chronic obstructive pulmonary disease (COPD). NOx had a greater health impacts in Taipei and Kaohsiung, but PM10 and SO2 in Taichung. Low wind speed and high sunshine rate in summer and autumn, and low relative humidity in spring and autumn were also significant weather conditions associated with the occurrence of CPDs. In all association models, P&I was consistently a significant risk factor for CPDs.
The risk for P&I associated with extreme cold was greater in Taipei with a RR of 1.31 (95% CI: 1.20-1.42, p<0.0001) than in Kaohsiung with a RR of 1.19 (95% CI: 1.09-1.31, p<0.0001). Low relative humidity was also an important risk factor for P&I. Nitrogen oxides (NOx) was a significant air pollutant associated with P&I morbidity, with a RR of 1.004 (95% CI: 1.001-1.007, p= 0.0175) for per 1 ppb NOx increase in Taipei, 1.005 (95% CI: 1.002-1.008) in Taichung, and 1.011 (95% CI: 1.007-1.015) in Kaohsiung. The risk for P&I morbidity associated with PM10 of 1 microgram per cubic meter increase was also significant (RR=1.003, 95% CI: 1.002-1.004) in Taipei.
In the survival analysis for the association between clinic visits of P&I and CPDs, the results showed that a lower cumulative survival rate occurred in cold season, male, and period before the influenza vaccination program. Lower cumulative survival rates for CPDs also appeared in the Kaohsiung population and those aged 0-14 and 50-69 years old before vaccination. But, there was a significant increase in survival in cases aged 60-69 years old, rose from 0.42 to 0.54. There was a greater impact from P&I for COPD with a hazard ratio (HR) of 11.7 (95% CI: 8.64-15.7, p <0.0001) than for BEA (HR= 4.82, 95% CI: 4.60-5.05, p <0.0001). The effect was reduced for those diagnosed with both BEA and COPD (HR= 2.59, 95% CI: 2.42-2.78, p <0.0001). Overall, the HR associated with pneumonia and influenza for CPDs was 3.91 (95% CI: 3.76-4.07, p <0.0001) after controlling for age, sex, area, vaccination, disease category and season.
In conclusion, the prevalence of reimbursement claims for the care of CPDs during 1996-2002 in Taiwan varied by area in association with P&I prevalence, regional weather conditions and air pollution. P&I had strongest association with the risk for CPDs in the 3 study metros. Among the 3 areas, the ambient temperature was also a significant risk factor for CPDs in Taipei and Taichung and air pollution in Kaohsiung.

並列關鍵字

chronic pulmonary disease ； pneumonia and influenza ； weather ； air pollution ； risk

參考文獻

154. Chen L, Lu H-S, Chang H-C. Utilization rates of preventive health services provided for children by the National Health Insurance Program, 1996-2001 Taiwan J Pub Health.2004;23:37-44.

162. Kuo LC, Yang PC, Kuo SH. Trends in the mortality of chronic obstructive pulmonary disease in Taiwan, 1981-2002. J Formos Med Assoc.2005;104:89-93.

163. Jan IS, Chou WH, Wang JD, Kuo SH. Prevalence of and major risk factors for adult bronchial asthma in Taipei City. J Formos Med Assoc.2004;103:259-263.

1. Murray CJL, Lopez AD. Alternative projections of mortality and disability by cause 1990-2020: Global burden of disease study. Lancet.1997;349:1498-1504.

2. Fletcher C, Peto R. The natural history of chronic airflow obstruction. Br Med J.1977;1:1645-1648.