產前金屬與農藥暴露對新生兒臍帶血去氧核醣核酸甲基化之相關性與新生兒出生結果之關聯探討

Abstract

背景與目的：根據目前研究已知母親懷孕期間暴露到環境中微量金屬與農藥會影響孩童之出生結果與神經發展行為，亦有研究指出此影響會影響孩童往後成長期間之疾病發生。然而，以目前已知文獻而言，環境中微量金屬與農藥暴露對於基因甲基化之關聯探討仍為少數，且環境暴露對於孩童神經行為發展與出生結果一直尚未有明確的解釋機轉。故本研究目的為透過結構性問卷與基因甲基化，了解環境暴露與出生結果及2歲孩童神經行為發展之間的關聯。 方法：本研究採用2004年至2005年間所建立的Taiwan Birth Panel Study世代研究族群，經篩選後共440對母親與孩童。本研究採用母親於生產後所填答之結構性問卷，與胎兒臍帶血樣本，進行基因甲基化的測量，並結合先前已測量完之臍帶血金屬與農藥濃度使用廣義線性迴歸模式進行關聯性研究，以及對於基因甲基化與嬰兒出生結果、2歲孩童神經行為發展使用廣義線性迴歸模式進行關聯性研究。 結果：母親懷孕期間微量金屬暴露對於LINE-1甲基化及Alu甲基化有顯著負向趨勢(p<0.05)；LINE-1低度甲基化與妊娠週期、Apgar評分有顯著相關(β= -0.78, p=0.0055; β= -0.23, p= 0.0012)；Alu甲基化與妊娠週期、新生兒頭圍、Apgar評分有顯著負向趨勢(p<0.05)；LINE-1甲基化在華語嬰幼兒溝通發展量表中的總分、辭彙表達、語法複雜度及語言使用有顯著負向趨勢(p<0.05)，Alu低度甲基化與總分、辭彙表達、語法複雜度及語言使用有顯著相關(β=-144.62, p=0.0108; β=-131.33, p=0.0137; β=-11.66, p=0.0112; β=-3.66, p=0.0204)，igf2甲基化與與總分、辭彙表達有顯著負向趨勢(p<0.05)，與語法複雜度及語言使用有顯著相關(β= -31.40, p=0.0020; β=-14.47, p<0.0001)；LINE-1甲基化在1歲半至5歲兒童行為檢核表中，與總分、內化量尺、外化量尺呈現U型分布；在兒童綜合發展測驗中，LINE-1低度甲基化與總分、語言商數、社會商數及自助商數呈現相關(β=-9.98, p=0.0257; β=-8.46, p=0.0220; β=-11.90, p<0.0448; β=-13.06, p=0.0209)，Alu甲基化與總分、語言商數、社會商數及自助商數亦有顯著相關。 結論：本研究結果顯示，除了環境微量金屬暴露與低度全基因體甲基化有負面的影響，低度甲基化對於新生兒結果亦有負面影響；本研究結果亦顯示低度甲基化與孩童2歲時神經智能行為發展的負面影響有關。

Background: The effects of metal and pesticide exposures can cause adverse neonatal birth outcomes and children neurobehavioral developments. However, association studies among exposures and DNA methylations are quite few, and there is no study focus on associations between DNA methylations and neurobehavioral developments. Objectives: 1) To evaluate associations between prenatal exposures to metal and pesticide and DNA methylation levels in umbilical cord blood; 2) to assess relationships among DNA methylation levels in umbilical cord blood and infants birth outcomes; 3) to assess relationships among DNA methylation levels in umbilical cord blood and children neurobehavioral developments measured in 2 year-old. Methods: A total of 440 mother-infant pairs from Taiwan Birth Panel Study (TBPS) after exclusion. 18 metals and 2 pesticides were measured in umbilical cord blood. DNA methylation levels of LINE-1, Alu and igf2 were also measured in leukocytes from umbilical cord blood. We estimated associations between exposure levels and DNA methylation levels, as well as associations between DNA methylation levels and children neurobehavioral developments in 2 year-old via generalized linear regression models. Results: Our data shows trend effects in low levels of metals exposure and LINE-1 methylations as well as Alu methylations (p<0.05). LINE-1 methylation levels are associated with gestational age and Apgar score (β= -0.78, p=0.0055; β= -0.23, p= 0.0012). Trend effects exist between Alu methylation levels, gestational age, infant head circumference, and Apgar score (p<0.05). In MCDI measurement, trend effects are in LINE-1 methylations, total score, words production, syntax complexity, and language use (p<0.05). Alu methylations are associated with total score, words production, syntax complexity, and language use (β=-144.62, p=0.0108; β=-131.33, p=0.0137; β=-11.66, p=0.0112; β=-3.66, p=0.0204). Trend effects exist in igf2 methylations, total score, and words production (p<0.05); igf2 also associated with syntax complexity and language use (β= -31.40, p=0.0020; β=-14.47, p<0.0001). In CBCL/1 frac12;-5 measurement, we found U shape between total score, internalizing score, and externalizing score. In CDIIT measurement, LINE-1 methylations are associated with total score, language DQ, social DQ, and self-help DQ (β=-9.98, p=0.0257; β=-8.46, p=0.0220; β=-11.90, p<0.0448; β=-13.06, p=0.0209). Alu methylations are also associated with total score, language DQ, social DQ, and self-help DQ. Conclusion: Our study indicated that environmental exposures are associated with global hypomethylations. Besides, our study showed that hypomethylations are related to adverse birth outcomes. Furthermore, our study revealed that hypomethylaitons are associated with children neurobehavioral developments in 2 years old.