台灣老年人長期空氣污染暴露與認知功能缺損之關聯性：6年的世代研究

Abstract

研究背景：隨著老年人口快速的上升，罹患認知功能障礙及失智症的長者也隨之增加。過去流行病學的研究顯示，長期暴露於空氣污染物中會影響認知功能，然而此發現於並不一致。此外，過去研究主要探討單一或特定空氣污染物所造成的影響，少有研究同時探討多種污染物與認知功能之間的關係。 研究方法：本研究為臺灣老年人流行病學研究（Taiwan Initiative for Geriatric Epidemiological Research，TIGER）的一部分，於基線時（2011-2013年）共收入605位(65歲以上之社區老年人)參加臺大醫院老人健康檢查，本研究平均每兩年追蹤一次，至2019年已完成3次的追蹤，排除有以下情形之受試者 ：蒙特利爾認知功能評估台灣版（Montreal Cognitive Assessment-Taiwan version，MoCA-T）分數小於21分、使用阿茲海默症藥物的患者、中風病史、腦部腫瘤大於等於3公分及腦部曾經受過損傷失去意識大於30分鐘、於兩年後的追蹤MoCA-T變化大於等於12分、居住在非暴露評估範圍、於基線前五年暴露濃度有缺失值、無法再次確認居住地點者，排除後共有499位受試者納入本次的分析。六種常見空氣污染：二氧化硫（sulphur dioxides, SO2）、二氧化氮（nitrogen dioxides, NO2）、粒徑小於等於2.5微米（um）之細懸浮微粒（particulate matter in diameter ≤ 2.5 µm, PM2.5）、粒徑介於2.5-10 um之粗懸浮微粒（particulate matter in diameter 2.5-10 µm, PMcoarse）、一氧化碳（carbon monoxide, CO）及臭氧（ozone, O3），取自臺灣行政院環保署的開放資料庫，納入台灣北部共26座測站自1993年開始的監測資料。將受試者歷年的居住位置及測站資料利用貝氏最大熵法之時空分析（Space-Time analysis Rending with Bayesian Maximum Entropy, STAR-BME）進行時空推估，模擬每位受試者的暴露資料，並計算每位受試者長期的平均暴露濃度。接著利用廣義線性混合模型（generalized linear mixed model，GLMM）加上隨機截距（random intercept），調整認知功能相關的重要變數後，探討17至23年空氣污染暴露與認知功能間於6年追蹤間的關係。認知功能的評估包含整體認知功能（用Montreal Cognitive Assessment –Taiwanese version, MoCA-T)評估及四項認知功能領域 （邏輯記憶力、專注能力、執行能力及語言能力）。 研究結果：除了PM2.5之外，其他污染物的長期平均暴露濃度低於環保署的標準（SO2＝4.5 ± 0.1 ppb；NO2＝27.2 ± 0.2 ppb；PM2.5＝19.1 ± 0.1 µm/m3；PMcoarse＝28.7 ± 0.2 µm/m3；CO＝1.4 ± 0.02 ppm；O3＝40.9 ± 0.2 ppb）。隨著長期PM2.5平均暴露濃度每上升一個四分位距單位(1.2 μg/m3)，整體認知功能（MoCA-T）會變差（β＝-0.12，95%信賴區間：-0.24至-0.002），並且於語言流暢度也有類似的發現（β＝-0.06，95%信賴區間：-0.10至-0.01）；然而隨著PMcoarse的暴露濃度每上升四分位距單位（1.8 μg/m3)，對邏輯記憶功能會呈現保護作用（β＝0.07至0.10）。而執行功能、專注功能與六項空氣污染物之間並未發現有關聯。 討論：長期暴露於空氣污染物中可能會造成老年人認知功能的危害，亦有可能會造成未來失智症的風險。因此，除了積極改善空氣品質外，期望未來於修定空氣品質標準時，能將認知功能相關的研究納入考量。

Background: As population aging increased, the prevalence of cognitive impairment and dementia increases rapidly. Previous epidemiologic studies showed that exposure to ambient air pollutants has been related to cognitive impairment, but their findings remain inconsistent. Most previous studies focused on single or few air pollutants. Studies simultaneously exploring multiple air pollutants and air pollutants are limited. Method: This study is part of the ongoing Taiwan Initiative for Geriatric Epidemiological Research (TIGER). A total of 605 community-dwelling elders aged 65 or older were recruited from the elderly health checkup program of National Taiwan University hospital at baseline (2011-2013) with three biennial follow-up. After exclusion the score of Montreal Cognitive Assessment-Taiwan version (MoCA-T) of baseline ≤ 21, use of medication for treating Alzheimer’s disease, stroke, brain tumor ≥ 3 cm, head trauma with loss of consciousness for ≥ 30 mins, living outside of the area for exposure assessment, lack of the exposure data 5 years before the baseline, and incomplete residential address information. A total of 499 participants were included for analyses. The data of six air pollutants were obtained from 26 monitor stations of Taiwan Environmental Protection Administration (EPA) between 1993 and 2010 in northern Taiwan. These pollutants included sulphur dioxide (SO2), nitrogen dioxide (NO2), particulate matter in diameter ≤ 2.5 µm (PM2.5), particulate matter in diameter 2.5-10 µm (PMcoarse), carbon monoxide (CO) and ozone (O3). Spatial-Time Analysis Rendering with Bayesian Maximum Entropy (STAR-BME) was utilized to establish the spatial-temporal model and then estimated the individual exposure of each air pollutant based on participants’ residual address to calculate the long-term average exposure concentration. Generalized linear mixed model (GLMM) with random intercept was utilized to assess the association of 17- to 23-year exposure to air pollutants with global or domain-specific cognition over 6 years follow-up adjusting for important covariates. Cognitive function include global cognition (assessed by Montreal Cognitive Assessment-Taiwanese version, MoCA-T) and domain- specific cognition (logical memory, attention, verbal fluency, and executive function).Result: Except PM2.5, the long-term exposure to other pollutants were below the EPA air quality standards (SO2: 4.5 ± 0.1 ppb, NO2: 27.2 ± 0.2 ppb, PMcoarse: 19.1 ± 0.1 µm/m3, PM2.5: 28.7 ± 0.2 µm/m3, CO: 1.4 ± 0.02 ppm, O3: 40.9 ± 0.2 ppb). An interquartile range (IQR) increase in PM2.5 was associated with poor global cognition [β＝-0.12, 95% confidence interval (CI) = -0.24 to -0.002] and verbal fluency [β＝-0.06, 95% CI = -0.10 to -0.01]. In contrast, an interquartile range (IQR) increase in PMcoarse was associated with better logical memory [β＝0.07 to 0.10]. No significant association was observed between other pollutants and the performance of executive function or attention. Conclusion: Long-term exposure to air pollutants may have detrimental effect on cognitive function of elderly and increase the risk of dementia in the future. Therefore, in addition to improve air quality, it is important to consider cognition as one of the health outcomes for revising air quality standards.