尿液中鄰苯二甲酸酯代謝物、雙酚A葡萄糖苷酸、1-羥基芘葡萄糖苷酸、全氟碳化合物與白三烯素E4含量，與女性荷爾蒙、嬰兒出生健康指標、過敏症狀之相關性

Abstract

鄰苯二甲酸酯 (phthalates)、雙酚A (bisphenol A)、全氟碳化合物(perfluoroalkyl chemicals, PFACs)以及多環芳香烴(polycyclic aromatic hydrocarbons, PAHs)為普遍存在於環境中之污染物；前三類為工業原料，常被應用於塑膠、食品包材、餐具及個人保健用品等，多環芳香烴則是經由不完全燃燒產生，交通排放為最主要的污染源。過往毒理學以及流行病學研究顯示，這些污染物皆具有生殖及發展毒性，有可能會影響女性荷爾蒙之分泌，也可能會對母體內之胎兒發育以及幼童健康造成不良影響。因此，本研究探討鄰苯二甲酸酯、雙酚A、全氟碳化合物以及多環芳香烴之暴露與女性荷爾蒙分泌之關聯、其產前暴露對嬰兒出生健康指標之影響、以及對孩童發生過敏之關聯。本研究對象分為兩部分，第一部分為60名液晶顯示幕工廠女工，第二部分則以Taiwan Birth Panel Study中138位產婦、196位兩歲兒童與74位五歲兒童為對象。所收集之尿液樣本以極致液相層析串聯式質譜儀分析16種待測物之濃度，包含4種鄰苯二甲酸酯代謝物 (monoethyl phthalate, MEP; mono-n-butyl phthalate MnBP; monobenzyl phthalate, MBzP, 和mono-(2-ethyl-hydroxyhexyl) phthalate, 5OH-MEHP)、雙酚A葡萄糖苷酸 (bisphenol A glucuronide, BPA-G)、1-羥基芘葡萄糖苷酸 (1-hydroxylpyrene glucuronide, 1-OHP-G)、9種全氟碳化合物 (perfluorobutanoic acid, perfluorohexanoic acid, perfluorooctanoic acid, perfluorononanoic acid, perfluorodecanoic acid, perfluoroundecanoic acid, perfluorododecanoic acid, perfluorooctane sulfonate,和perfluorohexane sulfonate)與白三烯素E4 (leukotriene E4, LTE4)。嬰兒出生健康資料、產婦及兒童之飲食、居住環境、過敏診斷與個人資料等由問卷取得，並使用多變項線性迴歸分析與邏輯迴歸進行統計分析。 待測物中以鄰苯二甲酸酯代謝物的檢出率最高，為30-73%；其中又以MEP檢出較高濃度(幾何平均數 = 15.2-25.5 μg/g creatinine)，MnBP居次(4.16-15.1 μg/g creatinine)，5OH-MEHP (3.74-8.01 μg/g creatinine)和MBzP (0.38-0.64 μg/g creatinine)則較低。全氟碳化合物之檢出率最低，僅見於極少數樣本。MEP (濃度範圍< LOD-8,995 μg/g creatinine)、MnBP (< LOD-3,213 μg/g creatinine)、BPA-G (< LOD-24,416 μg/g creatinine)與1-OHP-G (< LOD-1,124 μg/g creatinine)濃度變動較大，有些研究對象具有極高暴露。液晶顯示器工廠女工之荷爾蒙研究顯示，MBzP與BPA-G在濾泡期、排卵期或黃體期皆對促濾泡刺激素(follicle-stimulating hormone, FSH)有負向相關，MEP與BPA-G則分別在各期對PdG (pregnanediol glucuronide，為尿液中黃體素代謝物)有正向或負向相關，BPA-G對各期E1C濃度亦有正向相關，但皆未達統計顯著，亦無劑量效應關係，故認為鄰苯二甲酸酯與雙酚A的暴露對女性荷爾蒙分泌並無顯著影響。尿液中待測物濃度和嬰兒出生健康指標如體重、身長、頭圍或妊娠年齡的關係則依待測物和健康指標之不同而有所差異。嬰兒出生體重和MnBP、1-OHP-G有負向相關，和5OH-MEHP則有正向關係； MEP和MBzP對出生身長有負相關，高暴露組的效應大於低暴露組，且MEP的極高暴露組(前百分之十)對極低暴露組(小於偵測極限)有顯著負相關，5OH-MEHP則有正相關但無劑量效應關係；苯二甲酸酯或雙酚A對頭圍與妊娠年齡的增減則無明確關係。過敏孩童與健康孩童尿液中待測物濃度差異絕大部分未達統計顯著，但發現MBzP暴露越高可能增加發生過敏的風險，此外，在五歲孩童中發現MnBP之極高暴露組(前百分之二十五)發生過敏之風險為極低暴露組(小於偵測極限)的六倍 (β = 1.84, OR = 6.29, p < 0.05)。其餘之鄰苯二甲酸酯、雙酚A或多環芳香烴則未發現與過敏情形有關。有無使用益生菌之兩歲孩童尿液中待測物差異未達統計顯著，但使用益生菌之孩童尿液中BPA-G較高，且有無使用益生菌的族群其極高暴露組(前百分之十)的尿中BPA-G濃度有顯著差異，推測使用益生菌可能會提高BPA攝取。 本研究中嬰兒出生健康指標之研究對象主要來自臺北市及其鄰近地區，足以代表本區域族群之鄰苯二甲酸酯、雙酚A或多環芳香烴的暴露情形。而先前相似之台灣研究，其族群則是以台灣南部為主，且DBP暴露較DEP暴露高，和本研究之觀察相反。 本研究發現尿液中MEP、MBzP與BPA-G濃度與女性荷爾蒙含量有所關聯，但影響並未達統計上之顯著；鄰苯二甲酸酯(DEP、DBP與BBzP)與多環芳香烴之產前暴露則對嬰兒出生體重或身長有負向影響；DBP與BBzP暴露為孩童發生過敏之可能風險因子。其餘尿液中待測物則和所關心之健康效應無明顯統計相關性。為達到更佳統計效力，建議未來研究尚需要納入其他未考慮之影響因子以及增加研究之個案數目。在女性荷爾蒙研究方面，須將其他可能影響荷爾蒙分泌之因素如壓力、工作輪班、其他化學物質暴露納入考量；嬰兒出生健康指標亦可能受其他化學物質影響。此外，增加樣本數及提升待測物檢出率來增加統計解釋力亦為改善之道；其他可能影響過敏發生之物質如塵蟎、花粉等也應該在探討孩童過敏時納入考慮。 本研究找出可能造成健康效應之危險因子並建議可改善之處，以供未來進行大型研究之參採。

Phthalate esters, bisphenol A, and perfluoroalkyl chemicals (PFACs) are emerging contaminants, and are often used in consumer materials such as plastics, food containers, and personal care products. Polycyclic aromatic hydrocarbons (PAHs) are pollutants generated from incomplete combustion. These contaminants possess developmental and reproductive toxicity, induce allergic symptoms and affect excretion of female hormones. Therefore, our study investigated the associations between the exposure to phthalate esters, bisphenol A, PFACs, and PAHs and on the indices of female hormones, birth outcomes, and allergy. This study determined urinary concentrations of four phthalate metabolites (monoethyl phthalate, mono-n-butyl phthalate, monobenzyl phthalate, and mono-(2-ethyl-hydroxyhexyl) phthalate), bisphenol A glucuronide (BPA-G), nine perfluoroalkyl chemicals (perfluorobutanoic acid, perfluorohexanoic acid, perfluorooctanoic acid, perfluorononanoic acid, perfluorodecanoic acid, perfluoroundecanoic acid, perfluorododecanoic acid, perfluorooctane sulfonate, and perfluorohexane sulfonate), leukotriene E4 (LTE4) and 1-hydroxypyrene glucuronide (1-OHP-G) with ultra-high performance liquid chromatography/tandem mass spectrometry, then associated the concentrations with the female hormones levels of female workers who were employed in a LCD factory. Birth outcomes of newborns and the incidence of allergy among children at age of 2 and 5 from the subjects of the Taiwan Birth Panel Study (TBPS) were also examined with the urinary concentrations of the above analytes in the subjects. There were 60 female worker samples, 138 maternal urine samples, 196 and 74 samples of 2 and 5 years old children, respectively, were included in our study. The data of birth outcomes, demography, diet and living surroundings, and allergy diagnoses were collected by a structured questionnaire. Multiple linear regression analysis was conducted between urinary analyte concentrations and female hormones (follicle-stimulating hormone, FSH; E1C, a metabolite of estradiol; and pregnanediol glucuronide, PdG) or birth outcomes (birth weight, birth length, head circumference, and gestational age). Logistic regression analysis was applied to evaluate the associations between urinary analyte concentrations and incidence of allergy. Phthalate metabolites have higher positive rates among the 16 analytes (approximately 30-73%); MEP (GM = 15.2-25.5 μg/g creatinine) had relatively higher levels among the four phthalate metabolites, followed by MnBP (4.16-15.1 μg/g creatinine), 5OH-MEHP (3.74-8.01 μg/g creatinine) and MBzP (0.38-0.64 μg/g creatinine). PFACs were not detected in most urine samples. The concentrations of MEP (range = < LOD-8,995 μg/g creatinine), MBP (< LOD-3,213 μg/g creatinine), BPA-G (< LOD-24,416 μg/g creatinine), and 1-OHP-G (< LOD-1,124 μg/g creatinine) varied larger than those of other analytes, and some subjects would have been exposed to higher amount of these chemicals. Regarding the effects on female hormones, urinary concentrations of MBzP and BPA-G showed inverse trends to levels of FSH, that of BPA-G showed positive trend to E1C, and those of MEP and BPA-G showed positive and negative trends to levels of PdG in follicular phase, ovulation phase and luteal phase, respectively; however, the above relationships did not reach statistical significance and were not dose-dependent. The associations between urinary analyte concentrations and birth outcomes varied among the analytes. MnBP and 1-OHP-G were negatively associated with birth weight but 5OH-MEHP were positively; MEP and MBzP showed negative and dose-dependent associations with birth length; compared to the subjects with urine levels below the LODs, the subjects at upper 10% of MEP levels showed significantly negative association with birth length; 5OH-MEHP had positive but not dose-dependent relationships to birth length. The existing data suggested that prenatal exposure to phthalate esters, BPA and pyrene does not affect head circumference and gestation age of newborns. No significant differences in urinary concentrations of most analytes between children with and without allergy; nevertheless, there was a trend on higher urinary levels of MBzP toward increased incidence of allergy and the 5-year-old subjects with upper 25% of MnBP levels showed significantly higher risk of allergy than those below the LOD (β= 1.84, OR = 6.29, p < 0.05); exposure to DEHP, DEP, BPA and pyrene were not associated with allergic symptoms. Urinary levels of phthalate metabolites showed no significant difference in consuming probiotics, but the levels of BPA-G were higher in the subjects consumed probiotics than those who never use them. The subjects from the TBPS were recruited from Taipei and adjacent areas; consequently, it is able to well illustrate the exposure to phthalate esters, BPA and PAHs in Northern Taiwan, while the subjects of previous similar studies were from Southern Taiwan. There were differences in the background levels and exposure profiles between the two different regions; exposure to DBP was higher than that to DEP in Southern Taiwan, but the profile was opposite in Northern Taiwan. The urinary concentrations of MEP, MBzP and BPA-G associated with female hormone levels, although not in statistical differences; prenatal exposure to DEP, DBP, BBzP and pyrene resulted in adverse effects on birth weight and birth length. Exposure to DBP and BBzP may be risk factors of allergy in children; the rest analytes did not show significant associations with adverse health outcomes that were concerned in this study. For further studies, more considerations on potential confounders and greater statistical power are desired to better evaluate the associations. In terms of the female hormone study, pressure, work shift and other chemical exposure (such as VOCs) may affect hormone levels. Previous studies showed that prenatal exposure to PCBs or PBDEs affects birth outcomes. Risk factors of allergy such as mite and pollen also need to be considered when investigating the associations between the analytes and the incidence of allergy. An additional strategy is to enhance statistical power by increasing sample sizes and increasing positive rates of analyte detection in samples. This study provided potential risk compounds to three categories of health outcomes and discussed crucial issues that could improve further studies in a large scale.