Blood lead levels and lung cancer mortality: An updated analysis of NHANES II and III. (7th May 2021)

• Record Type: Journal Article
• Authors: Rhee, Jongeun; Graubard, Barry I.; Purdue, Mark P.
• Is Part Of: Cancer medicine. Volume 10:Number 12(2021)
• Page Start: 4066
• Page End: 4074
• Publication Date: 2021-05-07
• Subjects: blood lead -- lead -- lung cancer -- lung cancer mortality -- NHANES; 616.994005
• DOI: 10.1002/cam4.3943
• Languages: English
• ISSNs: 2045-7634

Abstract

Abstract: Previous analyses within the National Health and Nutrition Examination Survey (NHANES) II and III cycles suggested an association between blood lead levels (BLLs) and lung cancer mortality, although the evidence was limited by small case numbers. To clarify this relationship, we conducted updated analyses of 4, 182 and 15, 629 participants in NHANES II and III, respectively, (extending follow‐up 20 and 8 years) aged ≥20 with BLL measurements and mortality follow‐up through 2014. We fit multivariable Cox models to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) relating BLLs and lung cancer with adjustment for smoking and other factors. We did not observe an overall association between BLLs and lung cancer after adjustment for smoking (both surveys) and serum cotinine and environmental tobacco smoke exposure (NHANES III), although suggestive associations were observed among women (NHANES II: HR 2.7, 95% CI 0.7, 10.0 for ≥20.0 µg/dl vs. <10.0 µg/dl, P trend = 0.07; NHANES III: HR 11.2, 95% CI 2.1, 59.4 for ≥10.0 µg/dl vs. <2.5 µg/dl, P trend = 0.04). After stratifying on smoking status, an association with elevated BLLs was observed in NHANES II only among former smokers (HR 3.2, 95% CI 1.3, 8.0 for ≥15 vs. <15 µg/dl) and in NHANES III only among current smokers (HR 1.7, 95% CI 1.1, 2.8 for ≥5 vs. <5 µg/dl). In summary, we found elevated BLLs to be associated with lung cancer mortality among women in both NHANES II and III. Given the absence of an association among non‐smokers, we cannot rule out residual confounding as an explanation for our findings. Abstract :