A mortality study of beryllium workers.

We aimed at investigating mortality among beryllium-exposed workers, according to solubility of beryllium and beryllium compounds. We conducted an historical cohort study of 16,115 workers employed during 1925-2008 in 15 facilities, including eight entailing exposure to insoluble beryllium and seven entailing exposure to soluble/mixed beryllium compounds, who were followed up for mortality until 2011. Data were analyzed using indirect standardization and Cox regression modeling. Lung cancer standardized mortality ratio (SMR, national reference rates) was 1.02 (95% confidence interval [CI]: 0.94-1.10) in the whole cohort, 0.88 (95% CI: 0.75-1.03) in the insoluble beryllium subcohort, and 1.09 (95% CI: 0.99-1.09) in the soluble/mixed beryllium subcohort. For lung cancer, there was an association with period of hire in soluble/mixed beryllium plants but not in insoluble plants, and, conversely, employment in soluble/mixed plants was associated with increased mortality only among workers hired before 1955. There was no trend with duration of employment. Mortality from chronic beryllium disease increased, in particular, among workers hired before 1955 in soluble/mixed beryllium facilities. There was no increase in lung cancer mortality in the entire cohort and lung cancer mortality was not increased among beryllium workers hired in 1955 or later in soluble/mixed beryllium facilities, or at any time among those employed in insoluble beryllium facilities.