Lustberg, M and E Silbergeld. 2002. Blood lead levels and mortality. Archives of Internal Medicine 162:2443-2449.

Lead's toll on public health reaches far beyond the impact on childhood cognitive development and aggression. In this article, Lustberg and Silbergeld examine blood lead levels measured during the period 1976-1980 and calculate the effect of elevated levels on mortality rates over the subsequent decade. They document significant increases in mortality for people with blood lead levels between 20 and 29 µg/dL. The elevated mortality was observed for all mortality causes pooled together and also specifically for deaths related to circulatory illnesses and cancer.

How big a deal is this? Fifteen percent of the US population during the years 1976-1980 had lead levels within this range. Those still living are likely to continue to bear the additional health burden of lead poisoning from that era. And today, an estimated 1.7 million Americans in the general public have blood lead levels of at least 20 µg/dL. Moreover, the concentration of lead sufficient to raise mortality rates is well beneath the current federal standard of 50 µg/dL for workers.

Putting lead in paint and in gasoline, allowing lead smelters to foul their local environment and lead mining to pollute watersheds will surely rank as among the great failures of public health protection in the 20th century.

read about lead in Detroit neigbhorhoods

What did they do? Lustberg and Silbergeld used newly available data on the health status of people who had participated in the second National Health and Nutrition Examination Survey (NHANES II) during the years 1976 to 1980. In the original survey, detailed medical information was obtained from over 20,000 participants, including laboratory examinations, assessment of social and demographic factors (e.g., education and smoking), etc. Blood lead levels areavailable for 10,049 of the participants. The health of participants was tracked through 1992. By December 1992, 2145 of the participants with complete information had died, or roughly 23%.

Lustberg and Silbergeld focused their work on individuals whose lead levels were lower than 30 µg/dL between the ages of 30 and 74 years. This upper lead level was used as a cutoff to exclude lead exposed workers. Applying these selection criteria left them with a pool of 4190 individuals.

For their analysis, Lustberg and Silbergeld used a technique called survival analysis. This allowed them to separate out the effects of lead from other well-established factors that typically are associated with lead and would confuse the analysis. Among the many confounding variables they included in their analsysis were age, sex, race, education, income, smoking, exercise, and body mass index.

What did they find? The average lead level in Americans whose blood was sampled during the study (1976-1980) was 13 µg/dL. By 1992, 929 died (22.2%). Of those, 424 (10.1%) died of a circulatory disease and 240 (5.7%) of cancer.

In their statistical analyses, Lustberg and Silbergeld first looked at what factors were associated with higher blood lead levels. They found that being male, being older, living in an urban area, getting exercise, smoking and higher body mass index all were statistically associated with higher lead levels. Conversely, being white, having a college education and a greater income were associated with lower lead levels.

They then performed an analysis of the factors associated with higher mortality rates, without considering blood lead levels. Here, being male, being older, smoking more, exercising less and having a lower income were all linked to greater mortality. None of this is surprising.

They then performed analyses with lead, using multivariate statistical techniques to adjust for the confounding effects of confounding variables. The consistent finding of this work was that blood lead levels above 20 µg/dL were linked to greater overall mortality, higher mortality due to circulatory diseases, and higher mortality from cancer (table below). These causes of death were highly important in the study population, as 71% of deaths were attributable to either circulatory diseases or cancer.

Lustberg and Silbergeld also report that a detailed examination of their data indicate that African-Americans and other non-white participants were more sensitive to lead than whites, experiencing "increased mortality at lower blood levels." In African-Americans, blood lead levels between 10 and 19 µg/dL were associated with a 2.63-fold increase in mortality to all causes, compared to a 7% increase in whites (1.07-fold increase).

What does it mean? These remarkable results provide compelling new reasons for public and private measures to reduce lead exposures. Lead levels found in many Americans during the study period increased mortality rates by almost 50%. The sample size of the study is very large; hence this result is highly significant statistically (p = 0.003).

Elimination of lead from gasoline and measures reducing exposure to lead from other sources (e.g., lead paint) have brought lead levels down significantly in the average American. Results announced by the US Centers for Disease Control in January 2003 indicate that average lead levels today are approximately one-tenth of what they were in 1976. Selected groups of people, however, continue to be exposed to lead at dangerous levels.

For example, the current Occupational Safety and Health Administration "action level" for workers exposed to lead on the job is 50 µg/dL, far above the exposures found by Lustberg and Silbergeld to be associated with an increase in mortality. Children in housing with lead paint, and children living in areas with lead soil contamination lingering on after industrial activities (e.g., Detroit) are also at risk.