台灣尿中砷暴露之潛在來源探討：2019與2023年全國代表性人體生物監測調查

Abstract

緒論: 砷 (Arsenic) 是一種廣泛存在於自然環境中的類金屬，長期暴露可能增加癌症、心血管疾病及神經發育異常等健康風險。近年世界衛生組織 (WHO) 指出，無機砷為全球食源性化學危害中造成疾病負擔最高的化學物質。雖然台灣飲用水砷暴露已大幅改善，人體生物監測結果顯示尿中砷暴露仍普遍存在，因此有必要重新評估現代台灣一般族群主要的砷暴露來源。 方法: 使用2019年及2023年台灣人體生物監測調查 (Taiwan Human Biomonitoring Survey, THBM) 資料，為具全國代表性之橫斷性研究，共納入2,119名參與者。尿液檢體共分析五種砷物種，包括亞砷酸鹽 (arsenite, AsIII)、砷酸鹽 (arsenate, AsV)、單甲基砷酸 (monomethylarsonic acid, MMA)、二甲基砷酸 (dimethylarsinic acid, DMA) 以及砷甜菜鹼 (arsenobetaine, AsB) 並將AsIII、AsV、MMA與DMA之總和定義為InAs4。飲食資訊使用國民營養健康狀況變遷調查 (Nutrition and Health Survey in Taiwan, NAHSIT) 設計之飲食頻率問卷，環境暴露資訊則透過THBM問卷蒐集。研究首先進行單變項分析並採用假發現率 (false discovery rate, FDR) 校正多重比較，再建立多變項線性迴歸模型，調整年齡、性別、居住地區及調查年度等共變項。此外，建立同時納入飲食與環境因子之 combined models，並以教育程度、家庭收入及吸菸狀態進行敏感度分析。所有分析皆採調查加權分析方法以考量複雜抽樣設計。 結果： 族群尿液中AsB與InAs4之加權中位數分別為16.98及13.37 μg/g creatinine。AsB與InAs4飲食與環境合併模型分別納入2,037名及562名具有完整分析資料的參與者。在AsB combined model中，每日魚類攝取、其他海鮮、家禽及畜肉／紅肉攝取均與較高的尿中AsB濃度相關。在InAs4 combined model中，每日魚類、麵包及堅果種子類攝取與較高的尿中InAs4濃度相關，而每日深綠色蔬菜攝取則呈負相關；米飯攝取的信賴區間包含虛無值。居家附近化學工廠、其他工廠及焚香習慣在環境多變項模型中與較高的InAs4濃度相關，但在同時納入飲食與環境因素後，評估的環境因素均未維持統計顯著性。額外調整社經指標及吸菸狀態後，主要結果方向大致一致。 結論: 本研究結果顯示，飲食可能是當代台灣一般族群尿中砷暴露的重要潛在來源，而環境因素的關聯則較不明確。未來仍需結合食物及環境樣本之砷物種測量，進一步確認不同暴露來源及其相對貢獻。本研究凸顯砷物種分析在辨識不同暴露來源的重要性，研究結果可作為未來人體生物監測及砷暴露風險管理之參考。

Introduction: Arsenic is a metalloid widely present in the natural environment, and long-term exposure may increase the risks of cancer, cardiovascular diseases, and adverse neurodevelopmental outcomes. In recent years, the World Health Organization (WHO) has identified inorganic arsenic as the foodborne chemical hazard contributing the greatest disease burden worldwide. Although arsenic exposure from drinking water has substantially declined in Taiwan, human biomonitoring data indicate that urinary arsenic exposure remains prevalent. Therefore, it is necessary to reassess the potential sources of urinary arsenic exposure in the contemporary general population of Taiwan. Materials and Methods: This nationally representative cross-sectional study used data from 2,119 participants in the 2019 and 2023 Taiwan Human Biomonitoring (THBM) surveys. Urine samples were analyzed for five arsenic species, including arsenite (AsIII), arsenate (AsV), monomethylarsonic acid (MMA), dimethylarsinic acid (DMA), and arsenobetaine (AsB). InAs4 was defined as the sum of AsIII, AsV, MMA, and DMA. Dietary information was collected using a food-frequency questionnaire developed for the Nutrition and Health Survey in Taiwan (NAHSIT), whereas environmental exposure information was obtained through the THBM questionnaires. Individual exposure factors were first evaluated in screening analyses, with false discovery rate (FDR) correction applied for multiple comparisons. Multivariable linear regression models were subsequently constructed with adjustment for age, sex, residential region, and survey year. Combined models simultaneously incorporating dietary and environmental factors were also established. Sensitivity analyses were conducted with additional adjustment for educational level, household income, and smoking status. All analyses were survey-weighted to account for the complex sampling design. Results: The survey-weighted median urinary concentrations of AsB and InAs4 were 16.98 and 13.37 μg/g creatinine, respectively. The combined models for AsB and InAs4 included 2,037 and 562 participants with complete analytical data, respectively. In the AsB combined model, daily fish consumption and the consumption of other seafood, poultry, and livestock/red meat were associated with higher urinary AsB concentrations. In the InAs4 combined model, daily consumption of fish, bread, and nuts and seeds was associated with higher urinary InAs4 concentrations, whereas daily consumption of dark green vegetables was inversely associated with InAs4 concentrations. The confidence interval for rice consumption included the null value. Residential proximity to chemical factories or other factories and incense burning habit were associated with higher InAs4 concentrations in the environmental multivariable model. However, none of the environmental factors evaluated remained statistically significant after dietary and environmental factors were included simultaneously. The directions of the main findings were generally consistent after additional adjustment for socioeconomic indicators and smoking status. Conclusions: The findings suggest that diet may be an important potential source of urinary arsenic exposure in the contemporary general population of Taiwan, whereas associations with environmental factors were less evident. Future studies incorporating arsenic-species measurements in food and environmental samples are needed to further identify different exposure sources and their relative contributions. This study highlights the importance of arsenic speciation in identifying different exposure sources and may provide a reference for future human biomonitoring and arsenic exposure risk management.