Health effects in biomedical research laboratory personnel in Sweden : cancer occurrence and reproductive outcomes

Abstract: The main purposes of this thesis were to study mortality and cancer incidence among individuals working in biological and medical (biomedical) research laboratories as well as to investigate the possible adverse effects on reproduction in female laboratory personnel. This work is based on two cohorts. The study base of the cancer studies included employees who had worked at least one year and a day, 50% or more of full time, at laboratory departments (n=5,035) and at non-laboratory departments (n=2,923) at Karolinska Institutet and at the universities of Lund, Gothenburg and Linköping. 70 laboratory departments and 34 non-laboratory departments participated in the study, where the latter were included as an internal reference group. The number of laboratory and non-laboratory research groups participating in the questionnaire-based study of cancer and specific laboratory exposures was 714, encompassing 5288 individuals (3277 laboratory personnel with 1476 men and 1801 women, 2011 non-laboratory personnel with 1290 men and 721 women). The cohort of the reproduction studies comprised 1052 women born in 1945 or later, who had worked for one year or more during the period I January 1990 through 31 December 1994 in a laboratory or non-laboratory department at Karolinska Institutet or at the universities of Gothenburg, Linköping, Lund, Stockholm, Umeå or Uppsala and had given birth to at least one child during that period. The standardized mortality ratio (SMR) both for laboratory and non-laboratory personnel indicated lower mortality than for general population. Individual cancer types were studied by calculating standardized incidence ratios (SIR). There were an increase of SIRs for brain tumors among male laboratory scientists, 3.11 (95% confidence interval 0.85-7.96) after 10 years of work, and for breast cancer among female scientists, 1.62 (CI 0.78-2.98). For women who worked with organic solvents and with carcinogens of class 2B, elevated SIRs of 2.73 (CI 1. 10-5.63) and 3.15 (CI 1.16-6.85), respectively, were obtained for malignant melanoma. Work with DNA and RNA gave SIRs of 1.75 (CI 0.574.07) and of 1.78 (CI 0.584.14) for female breast cancer. For male personnel, prostate cancer and use of recombinant-DNA techniques gave a SIR of 1.79 (CI 0.49-4.59). However, the number of cases for all cancer types in research laboratories was small. The studies of pregnancy outcomes in connection to work with chloroform showed an odds ratio (OR) of 2.3 and a 95% confidence interval of 0.9-5.9 for spontaneous abortions. The OR for 'Large for gestational age' (LGA) infants in association with the mother's laboratory work was 1.9 (CI 0.7-5.2) and for male infants in connection to mother's work with radioactive isotopes, 3.4 (CI 0.8-14.2). The number of malformations was low in this cohort. For preterm births work with bacteria gave an OR of 2.7 (CI 1.2-6.5). Time-to-pregnancy was used to estimate the fecundability, i.e. per cycle probability to conceive a clinically detectable pregnancy. The fecundability ratio (FR) between exposed and unexposed cycles was calculated; a value below unity indicates subfertility. Work with organic solvents in general gave a decreased adjusted fecundability ratio (FR) of 0.76, 95% CI of 0.64-0.89. Moreover, work with cell techniques and use of viruses also showed decreased FRs, 0.83 (CI 0.70-0.99) and 0.73 (CI 0.56-0.95), respectively, In conclusion, the total mortality as well as the incidence of all cancers together was lower in both the laboratory and the non-laboratory groups than in the general population, but slightly increased risks were seen for male brain tumors and female breast cancer among laboratory personnel. Work with solvents showed an elevated SIR of malignant melanoma in female laboratory personnel. Concerning reproductive health, no major risks were obtained for most outcomes. However, mother's work in laboratory showed an increase of LGA infants and an association was seen between reduced fecundability and use of solvents, cell techniques or viruses.