Objective Asbestos is a known human carcinogen, with evidence for malignant mesothelioma (MM), cancers of lung, ovary, larynx and possibly other organs. MM rates are predicted to increase with a power of time since first exposure (TSFE), but the possible long-term attenuation of the trend is debated. The asbestos ban enforced in Italy in 1992 gives an opportunity to measure long-term cancer risk in formerly exposed workers.

Methods Pool of 43 previously studied Italian asbestos cohorts (asbestos cement, rolling stock, shipbuilding), with mortality follow-up updated to 2010. SMRs were computed for the 1970–2010 period, for the major causes, with consideration of duration and TSFE, using reference rates by age, sex, region and calendar period.

Results The study included 51 801 subjects (5741 women): 55.9% alive, 42.6% died (cause known for 95%) and 1.5% lost to follow-up. Mortality was significantly increased for all deaths (SMR: men: 1.05, 95% CI 1.03 to 1.06; women: 1.17, 95% CI to 1.12 to 1.22), all malignancies combined (SMR: men: 1.17, 95% CI to 1.14 to 1.20; women: 1.33, 95% CI 1.24 to 1.43), pleural and peritoneal malignancies (SMR: men: 13.28 and 4.77, 95% CI 12.24 to 14.37 and 4.00 to 5.64; women: 28.44 and 6.75, 95% CI 23.83 to 33.69 and 4.70 to 9.39), lung (SMR: men: 1.26, 95% CI 1.21 to 1.31; women: 1.43, 95% CI 1.13 to 1.78) and ovarian cancer (SMR=1.38, 95% CI 1.00 to 1.87) and asbestosis (SMR: men: 300.7, 95% CI 270.7 to 333.2; women: 389.6, 95% CI 290.1 to 512.3). Pleural cancer rate increased during the first 40 years of TSFE and reached a plateau after.

Discussion The study confirmed the increased risk for cancer of the lung, ovary, pleura and peritoneum but not of the larynx and the digestive tract. Pleural cancer mortality reached a plateau at long TSFE, coherently with recent reports.