Department of Environmental Health, Harvard T.H. Chan School of Public Health, 677 Huntington Ave, Boston, MA 02115, USA.
Nicholas School of the Environment, Duke University, 9 Circuit Dr, Durham, NC 27710, USA.
Chemosphere. 2023 Feb;315:137705. doi: 10.1016/j.chemosphere.2022.137705. Epub 2022 Dec 30.
Humans are exposed to increasingly complex mixtures of hormone-disrupting chemicals from a variety of sources, yet, traditional research methods only evaluate a small number of chemicals at a time. We aimed to advance novel methods to investigate exposures to complex chemical mixtures. Silicone wristbands were worn by 243 office workers in the USA, UK, China, and India during four work shifts. We analyzed extracts of the wristbands for: 1) 99 known (targeted) chemicals; 2) 1000+ unknown chemical features, tentatively identified through suspect screening; and 3) total hormonal activities towards estrogen (ER), androgen (AR), and thyroid hormone (TR) receptors in human cell assays. We evaluated associations of chemicals with hormonal activities using Bayesian kernel machine regression models, separately for targeted versus suspect chemicals (with detection ≥50%). Every wristband exhibited hormonal activity towards at least one receptor: 99% antagonized TR, 96% antagonized AR, and 58% agonized ER. Compared to men, women were exposed to mixtures that were more estrogenic (180% higher, adjusted for country, age, and skin oil abundance in wristband), anti-androgenic (110% higher), and complex (median 836 detected chemical features versus 780). Adjusted models showed strong associations of jointly increasing chemical concentrations with higher hormonal activities. Several targeted and suspect chemicals were important co-drivers of overall mixture effects, including chemicals used as plasticizers, fragrance, sunscreen, pesticides, and from other or unknown sources. This study highlights the role of personal care products and building microenvironments in hormone-disrupting exposures, and the substantial contribution of chemicals not often identifiable or well-understood to those exposures.
人类会从各种来源接触到越来越复杂的混合激素扰乱化学物质,但传统的研究方法一次只能评估少数几种化学物质。我们旨在开发新的方法来研究复杂化学混合物的暴露情况。243 名美、英、中、印的办公室工作人员在四个工作班次期间佩戴了硅胶手环。我们分析了手环提取物中的:1)99 种已知(靶向)化学物质;2)1000 多种通过可疑筛选初步确定的未知化学特征;3)通过人体细胞测定对雌激素(ER)、雄激素(AR)和甲状腺激素(TR)受体的总激素活性。我们使用贝叶斯核机器回归模型评估了化学物质与激素活性之间的关联,分别针对靶向化学物质和可疑化学物质(检测≥50%)进行评估。每个手环都对至少一种受体表现出激素活性:99%拮抗 TR,96%拮抗 AR,58%激动 ER。与男性相比,女性接触的混合物更具雌激素性(调整国家、年龄和手腕带中皮肤油脂丰富度后,高出 180%)、抗雄激素性(高出 110%)和更复杂(中位数为 836 种检测到的化学特征,而 780 种)。调整后的模型显示,共同增加化学浓度与更高的激素活性之间存在强烈关联。一些靶向和可疑化学物质是整体混合物效应的重要共同驱动因素,包括用作增塑剂、香料、防晒霜、杀虫剂的化学物质,以及其他或未知来源的化学物质。这项研究强调了个人护理产品和建筑微环境在激素扰乱暴露中的作用,以及通常无法识别或了解的化学物质对这些暴露的大量贡献。