以多層次分析探討空氣污染對學童氣喘與肺功能之相關性研究

The association of air pollution with asthma and lung function of children in multilevel analysis

曾暐婷(Wei-Ting Tseng)

指導教授 ： 李建宏

高雄醫學大學/健康科學院/公共衛生學研究所/碩士(2008年)

摘要

氣喘及肺功能可能會受社區、個體等多層次因素所影響，而過去單一層次的研究對整個族群的了解有限。本研究使以多層次模式分析學童個體層次因素及學校社區層次因素對肺功能的影響，了解肺功能在學童間和學校間的變異。
本研究分析民國84年高屏區所進行的學童呼吸系統健康檢查資料，僅納入距空氣品質監測站3公里以內的國中，共56所88,902位學童接受International Study of Asthma and Allergy in Childhood（ISAAC）錄影帶問卷篩檢氣喘，並調查學童個體等危險因子。除氣喘之外，16,287位有進行肺功能檢測。
結果發現，學童FVC、FEV1、FEV1/FVC、PEF、PEF%等肺功能和氣喘盛行狀況在一定的程度上受其所在的學校不同而不同。在分別納入不同個體層次的變數後，不管是社區層次或個體層次的變異都有部分被解釋，且在兩個層次裡仍有變異存在。此外，發現學童肺功能的變異仍約有2.56~9.95%來自於社區層次，氣喘盛行的變異仍有2.81%，即不同學校間的學童這些肺功能和氣喘仍有群聚的效應存在。
考慮社區層次空氣污染後，臭氧確實能降低兩個層次的FEV1/FVC變異，總懸浮微粒和落塵量對於PEF%也有同樣的結果。而氣喘的部份，則是二氧化氮、臭氧、碳氫化合物、落塵量能部份解釋社區層次的變異情形。
BMI超過23.1 kg/m2 對FVC和FEV1的影響會隨學校不同而不同，即學校的脈絡現象會修飾個體層次BMI超過23.1 kg/m2 這變項和FVC和FEV1之間的關係。
運用隨機效應的多層次分析模式較傳統固定效應的分析模式合理及正確，在考量個體層次因素的影響後，社區層次的脈絡因素亦對學童的FVC與FEV1有顯著的影響性。

關鍵字

氣喘 ； 空氣污染 ； ISAAC問卷 ； 肺功能 ； 多層次分析

並列摘要

Asthma and lung function may be affected by both school (community) and schoolchildren (individual) level factors. In the single level analysis we can not completely understand whole group health disparity. We used multilevel analysis to understand the association between different level factors with asthma and lung function. And we estimated both between schoolchildren and between school variance in asthma and lung function
By using the video edition of International Study of Asthma and Allergy in Children (ISAAC), we studied the prevalence of asthma from 88,902 schoolchildren in 56 junior high schools. And each school is within 3 kilometers from air-monitoring station. Besides, we also measured the lung function of the schoolchildren.
The study found that distribution of FVC, FEV1, FEV1/FVC, PEF, PEF%, and prevalence of asthma differed from each school. After adjusted schoolchildren level factors, variance between schoolchildren and between school were partly explained but still existed. About 2.56~9.95% and 2.81% of the schoolchildren residual differences in lung function and asthma were related to school level. The potential clustering effect of asthma and lung function was found within schools.
After taking into account the outdoor air pollutant, ozone reduced two levels variance of FEV1/FVC. Total suspended particulate and airborne dust particles had the same effect on PEF%. And a portion of school level asthma variance was explained by NO2, O3, Carbon hydrogen compound, and airborne dust particles.
The effect of BMI above 23.1 kg/m2 on FVC and FEV1 was based on different school. The school context modifies the schoolchildren association between BMI above 23.1 kg/m2 with FVC and FEV1 these two lung function.
The random effect of multilevel analysis is more rational and correct than fixed effect of single level analysis. Schoolchildren asthma and lung function differences were partly conditioned by the schoolchildren composition of the schools and school contextual factors.

並列關鍵字

Asthma ； Air pollution ； ISAAC ； Lung function ； Multilevel analysis

參考文獻

1. Bateman, E.D., et al., Global strategy for asthma management and prevention: GINA executive summary. Eur Respir J, 2008. 31(1): p. 143-78.

2. Yeh, K.W., W. Fang, and J.L. Huang, Increasing the hospitalization of asthma in children not in adults - from a national survey in Taiwan 1996-2002. Pediatr Allergy Immunol, 2008. 19(1): p. 13-9.

3. Ho, W.C., et al., Air pollution, weather, and associated risk factors related to asthma prevalence and attack rate. Environ Res, 2007. 104(3): p. 402-9.

4. Tsai, S.S., et al., Air pollution and hospital admissions for asthma in a tropical city: Kaohsiung, Taiwan. Inhal Toxicol, 2006. 18(8): p. 549-54.

5. 行政院衛生署, 氣喘診療指引. 2000, 台北市: 衛生署.