A follow-up study of 304 cases of suspected pathology caused by benzene seen in 1950-71

• Published in: Medicina del lavoro Vol. 80; no. 5; p. 397
• Main Authors: Vai, T, Radice, L, Catenacci, G, Biscaldi, G P, Guercilena, S, Pesatori, A C, Bertazzi, P A
• Format: Journal Article
• Language: Italian
• Published: Italy 01-09-1989
• Subjects: Benzene - adverse effects; Cause of Death; Female; Follow-Up Studies; Humans; Italy; Leukemia - chemically induced; Leukemia - mortality; Male; Neoplasms - chemically induced; Neoplasms - mortality; Occupational Diseases - chemically induced; Occupational Diseases - mortality; Time Factors; Italy
• ISSN: 0025-7818

A group of 304 subjects admitted to the Institutes of Occupational Health of Milan and Pavia (Italy) between 1951 and 1970 for suspected benzene intoxication were retrieved from hospital records and their mortality experience was examined up to 31 December, 1986. The aims of the study were: 1) to estimate, in quantitative terms, the risk associated with benzene exposure in that area in the time period considered (these risks had already been reported, but merely as case reports); 2) to investigate the possibility of an increased frequency of neoplasms other than leukaemia; and 3) to further investigate the exposure history of these subjects. In the absence of data on the population at risk, mortality was analysed via the mortality odds ratio (MOR) method. A local population mortality experience was used as reference. Twenty-eight malignant neoplasms were observed (MOR = 2.2; 95% confidence interval = 1.3-3.7), 15 of which were haematologic neoplasms (MOR = 13.3; 95% confidence interval = 8.0-22.2). No odds ratio increases were observed for any of the other tumour sites or types. Eleven observed blood diseases represented a large numerical increase in the odds ratio. For all the haematologic neoplasms there had been an estimated exposure to airborne benzene concentrations above 20 ppm. However, no conclusions on dose-response relationships can be drawn from these data because of the approximate evaluation of individual exposure and, especially, because the study group most probably consisted of a highly selected sample of the exposed population.