Estimating lead-attributable mortality burden by socioeconomic status in the USA

Chengzhe Tao,Zhi Li,Yun Fan,Yuna Huang,Tingya Wan,Mingxue Shu,Shuwen Han,Hong Qian,Wenkai Yan,Qiaoqiao Xu,Yankai Xia,Chuncheng Lu,You Li

DOI: https://doi.org/10.1093/ije/dyae089

2024-06-12

Abstract:Background: This study aimed to estimate population-level and state-level lead-attributable mortality burdens stratified by socioeconomic status (SES) class in the USA. Methods: Based on the National Health and Nutrition Examination Survey (NHANES), we constructed individual-level SES scores from income, employment, education and insurance data. We assessed the association between the blood lead levels (BLL) and all-cause mortality by Cox regression in the NHANES cohort (n = 31 311, 4467 deaths). With estimated hazard ratios (HR) and prevalences of medium (2-5 μg/dL) and high (≥ 5 μg/dL) BLL, we computed SES-stratified population-attributable fractions (PAFs) of all-cause mortality from lead exposure across 1999-2019. We additionally conducted a systematic review to estimate the lead-attributable mortality burden at state-level. Results: The HR for every 2-fold increase in the BLL decreased from 1.23 (1.10-1.38) for the lowest SES class to 1.05 (0.90-1.23) for the highest SES class. Across all SES quintiles, medium BLL exhibited a greater mortality burden. Individuals with lower SES had higher lead-attributable burdens, and such disparities haver persisted over the past two decades. In 2017-19, annually 67 000 (32 000-112 000) deaths in the USA were attributable to lead exposure, with 18 000 (2000-41 000) of these deaths occurring in the lowest SES class. Substantial disparities in the state-level mortality burden attributable to lead exposure were also highlighted. Conclusions: These findings suggested that disparities in lead-attributable mortality burden persisted within US adults, due to heterogeneities in the effect sizes of lead exposure as well as in the BLL among different SES classes.