Do we underestimate risk of cardiovascular mortality due to lead exposure?

Studies using data from the National Health and Nutrition Examination Survey-III (NHANES-III) have demonstrated significant prospective associations between blood lead levels and increased mortality. Bone lead represents cumulative lead burden and thus is a better biomarker for assessing chronic impacts, but its in vivo assessment requires special K-x-ray fluorescence (KXRF) instrumentation. Our team recently developed an algorithm predicting bone lead levels from a combination of blood lead levels, age and other socioeconomic and behavioral variables. We examined the associations of our algorithm-estimated bone lead levels and mortality in NHANES-III. We included 11,628 adults followed up to December 31, 2019. Estimated tibia lead and patella lead levels were calculated using our prediction algorithms. We used survey-weighted Cox proportional hazards models to compute hazard ratios (HRs) and 95 % confidence intervals (CIs). During the median follow-up of 26.8 years, 4900 participants died (mortality rate = 1398 per 100,000 adults/year). Geometric means (95 % CIs) of blood lead, predicted tibia lead, and predicted patella lead were 2.69 μg/dL (2.54, 2.84), 6.73 μg/g (6.22, 7.25), and 16.3 μg/g (15.9, 16.8), respectively. The associations for all-cause mortality were similar between blood lead and bone lead. However, the associations for cardiovascular mortality were much greater with predicted bone lead markers compared to blood lead: for comparing participants at the 90th vs. 10th percentiles of exposure, HR = 3.32 (95 % CI: 1.93–5.73) for tibia lead, 2.42 (1.56–3.76) for patella lead, 1.63 (1.25–2.14) for blood lead. The population attributable fractions for cardiovascular disease mortality if everyone's lead concentrations were declined to the 10th percentiles were 45.8 % (95 % CI: 28.1–59.4) for tibia lead, 33.1 % (18.1–45.8) for patella lead, and 22.8 % (10.4–33.8) for blood lead. These findings suggest that risk assessment for cardiovascular mortality based on blood lead levels may underestimate the true mortality risk of lead exposure.