Article Text


278 Mortality patterns in a cohort of 70,000 workers with blood lead measurements
  1. K Steenland1,
  2. Chowdhury2
  1. 1Atlanta, United States of America
  2. 2Emory University, Atlanta, United States of America


Objectives To determine whether adult lead exposure is associated with cause-specific mortality in a cohort with measured blood leads

Methods We studied 70,000 US adults from 11 states with blood lead levels measured between 1980–2005. Most were exposed occupationally. One-third had a single blood lead; the remainder had a median of 3 blood leads. Subjects were divided into four groups by highest blood lead (0–5, 6–24, 25–39, 40 + mg/dl; 16%, 33%, 34%, and 17% respectively.

Results Analyses were restricted to males (96% of deaths). There was a strong healthy worker effect in this young cohort (SMR all causes 0.73, 3561 deaths, US referent). Most cause-specific SMRs were unremarkable. The male lung cancer SMR showed a trend of increase across lead categories (SMRs 0.49, 0.61, 0.80, 1.21, 394 deaths, test for trend p = 0.003). With the lowest lead category as referent in Poisson regression, lung RRs were 1.00, 1.27, 1.80, and 2.85 by increasing lead category (test for trend p = <0.0001). Other smoking-related causes of death did not show consistent trends (eg., male COPD male SRRs 1.00, 0.21, 0.44, 0.63 (n = 129), male esophageal cancer SRRs 1.00, 1.29, 0.98, 1.43 (n = 40). There was a marked excess of ALS in the cohort (SMR 2.25, 43 observed), which however had an inverse relation to blood lead level (SMRS 11.1, 3.2, 0.9. 0.8 by increasing category). Numbers were small for other causes of interest (stomach cancer n = 23, kidney cancer n = 28, brain cancer n = 31). SMRs for stroke were low (0.59, 0.57, 0.81. 0.87) but showed a borderline significant increasing trend in Poisson regression (p = -0.08).

Conclusion We studied a large cohort with documented lead exposure. Results are not conclusive but suggest an association between lung cancer and lead exposure. Data are limited by a lack of work history, a limited blood leads per person, and relatively short latency.

Statistics from

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.