贝叶斯核机器回归模型在环境健康研究中的应用及R语言实现

doi:10.12183/j.scjpm.2025.0148

华南预防医学 ›› 2025, Vol. 51 ›› Issue (2): 148-153.doi: 10.12183/j.scjpm.2025.0148

贝叶斯核机器回归模型在环境健康研究中的应用及R语言实现

欧可儿^1,3, 郑雯媛^2,3, 李杏³, 曾韦霖³, 容祖华³, 龚德鑫³, 肖建鹏³

1.广东药科大学公共卫生学院,广东广州 510310;
2.暨南大学基础医学与公共卫生学院;
3.广东省疾病预防控制中心广东省公共卫生研究院

收稿日期:2024-05-16 发布日期:2025-03-18
通讯作者: 肖建鹏,E-mail：jpengx@163.com
作者简介:欧可儿（2000—）,女,在读硕士研究生,主要研究方向：环境与健康
基金资助:
国家自然科学基金（82373636）; 国家重点研发计划（2018YFA0606200）; 广东省医学科研基金（A2023036、A2023038）

Application of Bayesian kernel machine regression model in environmental health research and R implementation

OU Keer^1,3, ZHENG Wenyuan^2,3, LI Xing³, ZENG Weilin³, RONG Zuhua³, GONG Dexin³, XIAO Jianpeng³

1. School of Public Health, Guangdong Pharmaceutical University, Guangzhou, Guangdong 510310, China;
2. School of Medicine, Jinan University;
3. Guangdong Provincial Institute of Public Health, Guangdong Provincial Center for Disease Control and Prevention

Received:2024-05-16 Published:2025-03-18

摘要/Abstract

摘要： 目的贝叶斯核机器回归（Bayesian kernel machine regression,BKMR）模型是近年快速发展的统计分析方法,本研究旨在介绍BKMR模型在多污染物暴露的健康效应研究中的应用及其在R语言的实现。方法以美国国家环境健康科学研究所（National Institute of Environmental Health Sciences,NIEHS）公开的数据集为例,运用BKMR模型分析多污染物暴露对健康的影响,并基于R语言介绍其分析步骤。结果依托R语言的“bkmr”程序包,BKMR模型可以探究暴露与健康的单暴露-反应关系,多种暴露之间的交互作用,估算单暴露的效应和多暴露物的联合效应。结论 BKMR模型可同时分析多污染物暴露与健康结局的暴露反应关系和健康效应,是研究复合环境暴露健康影响的一种新分析方法。

关键词: 贝叶斯核机器回归, 多污染物暴露, 健康效应, R语言

Abstract: Objective The Bayesian kernel machine regression (BKMR) model is a new method developed rapidly in recent years. This study aims to introduce the application of the BKMR model in the study of the health effects of multi-pollutant exposure and R implementation. Methods Taking the dataset published by the National Institute of Environmental Health Sciences (NIEHS) as an example, we used BKMR model to analyze the impact of multi-pollutant on health, and introduced its implementation steps based on R language. Results Using on the "bkmr" package of R language, the BKMR model could analyze the univariate exposure-response relationship between the exposure and the health outcomes, the interactions among multiple exposures, and estimate the effect of single exposure and the joint effect of multi-pollutant exposure. Conclusion The BKMR model can analyze the exposure-response relationship and health effects of multi-pollutant exposure on health outcomes, which is a new analytical method for studying the health effect of combined environmental exposures.

Key words: Bayesian kernel machine regression (BKMR), Multi-pollutant exposure, Health effect, R language

中图分类号:

R195.1

欧可儿, 郑雯媛, 李杏, 曾韦霖, 容祖华, 龚德鑫, 肖建鹏. 贝叶斯核机器回归模型在环境健康研究中的应用及R语言实现[J]. 华南预防医学, 2025, 51(2): 148-153.

OU Keer, ZHENG Wenyuan, LI Xing, ZENG Weilin, RONG Zuhua, GONG Dexin, XIAO Jianpeng. Application of Bayesian kernel machine regression model in environmental health research and R implementation[J]. South China Journal of Preventive Medicine, 2025, 51(2): 148-153.

参考文献

[1] 韦兰成,徐怡璐,吴彬,等. 重金属复合暴露健康效应的研究进展[J]. 毒理学杂志,2021,35(3):251-254,261.
[2] Zhu J, Chen L, Liao H.Multi-pollutant air pollution and associated health risks in China from 2014 to 2020[J]. Atmospheric Environment, 2022, 268: 118829.
[3] Smarr MM, Mirzaei Salehabadi S, Boyd Barr D, et al.A multi-pollutant assessment of preconception persistent endocrine disrupting chemicals and incident pregnancy loss[J]. Environ Int, 2021, 157: 106788.
[4] MacLehose RF, Dunson DB, Herring AH, et al. Bayesian methods for highly correlated exposure data[J]. Epidemiology, 2007, 18(2): 199-207.
[5] Cox LA Jr.Implications of nonlinearity, confounding, and interactions for estimating exposure concentration-response functions in quantitative risk analysis[J]. Environ Res, 2020, 187: 109638.
[6] Bobb JF, Valeri L, Claus Henn B, et al.Bayesian kernel machine regression for estimating the health effects of multi-pollutant mixtures[J]. Biostatistics, 2015, 16(3): 493-508.
[7] 申炜垌,张佳晨,郭建勇,等. 重金属汞、铅、镉低剂量暴露与原发性高血压的关联性研究[J]. 实用预防医学,2023,30(4):385-389.
[8] Xie T, Wang M, Jiang R, et al.Comparative study on anthropogenic impacts on soil PAHs: Accumulation and source apportionment in tourist and industrial cities in Hebei Province, China[J]. Sci Total Environ, 2024, 912: 168435.
[9] Li H, Deng W, Small R, et al.Health effects of air pollutant mixtures on overall mortality among the elderly population using bayesian kernel machine regression (BKMR)[J]. Chemosphere, 2022, 286(Pt 1): 131566.
[10] Bobb JF, Claus Henn B, Valeri L, et al.Statistical software for analyzing the health effects of multiple concurrent exposures via Bayesian kernel machine regression[J]. Environ Health, 2018, 17(1): 67.
[11] Taylor KW, Joubert BR, Braun JM, et al.Statistical approaches for assessing health effects of environmental chemical mixtures in epidemiology: Lessons from an Innovative Workshop[J]. Environ Health Perspect, 2016, 124(12): A227-A229.
[12] Zhao N, Smargiassi A, Hatzopoulou M, et al.Long-term exposure to a mixture of industrial SO₂, NO₂, and PM_2.5 and anti-citrullinated protein antibody positivity[J]. Environ Health, 2020, 19(1): 86.
[13] Goodrich AJ, Kleeman MJ, Tancredi DJ, et al.Ultrafine particulate matter exposure during second year of life, but not before, associated with increased risk of autism spectrum disorder in BKMR mixtures model of multiple air pollutants[J]. Environ Res, 2024, 242: 117624.
[14] Preston EV, Webster TF, Claus Henn B, et al.Prenatal exposure to per- and polyfluoroalkyl substances and maternal and neonatal thyroid function in the Project Viva Cohort: A mixtures approach[J]. Environ Int, 2020, 139: 105728.
[15] Wang X, Gao D, Zhang G, et al.Exposure to multiple metals in early pregnancy and gestational diabetes mellitus: A prospective cohort study[J]. Environ Int, 2020, 135: 105370.
[16] Carrico C, Gennings C, Wheeler DC, et al.Characterization of weighted quantile sum regression for highly correlated data in a risk analysis setting[J]. J Agric Biol Environ Stat, 2014, 20(1): 100-120.
[17] Keil AP, Buckley JP, O'Brien KM, et al. A quantile-based g-computation approach to addressing the effects of exposure mixtures[J]. Environ Health Perspect, 2020, 128(4): 47004.
[18] Chen L, Sun Q, Peng S, et al.Associations of blood and urinary heavy metals with rheumatoid arthritis risk among adults in NHANES, 1999-2018[J]. Chemosphere, 2022, 289: 133147.
[19] Guo X, Wu B, Hu W, et al.Associations of perchlorate, nitrate, and thiocyanate with metabolic syndrome and its components among US adults: A cross-sectional study from NHANES[J]. Sci Total Environ, 2023, 879: 163083.
[20] Oluwayiose OA, Houle E, Wu H, et al.Urinary phthalate metabolites and their mixtures are associated with advanced sperm epigenetic aging in a general population[J]. Environ Res, 2022, 214(Pt 4): 114115.

贝叶斯核机器回归模型在环境健康研究中的应用及R语言实现

Application of Bayesian kernel machine regression model in environmental health research and R implementation

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 4

Metrics

本文评价

推荐阅读 0

[1]	李杏, 刘涛, 肖建鹏, 吴君乐, 马文军. R语言在公共卫生领域的应用：Meta分析[J]. 华南预防医学, 2020, 46(5): 568-570.
[2]	曾韦霖, 刘涛, 肖建鹏, 马文军. 广义线性模型在时间序列研究中的应用及R语言实现[J]. 华南预防医学, 2020, 46(3): 321-323.
[3]	胡建雄, 陈思齐, 刘涛, 肖建鹏, 曾韦霖, 马文军. R语言在公共卫生领域的应用：基本数据处理[J]. 华南预防医学, 2020, 46(2): 183-184.
[4]	陈思齐, 胡建雄, 刘涛, 曾韦霖, 马文军. R语言在公共卫生领域的应用:基础入门篇[J]. 华南预防医学, 2020, 46(1): 98-100.