Occupation and cancer - follow-up of 15 million people in five Nordic countries.

We present up to 45 years of cancer incidence data by occupational category for the Nordic populations. The study covers the 15 million people aged 30-64 years in the 1960, 1970, 1980/1981 and/or 1990 censuses in Denmark, Finland, Iceland, Norway and Sweden, and the 2.8 million incident cancer cases diagnosed in these people in a follow-up until about 2005. The study was undertaken as a cohort study with linkage of individual records based on the personal identity codes used in all the Nordic countries. In the censuses, information on occupation for each person was provided through free text in self-administered questionnaires. The data were centrally coded and computerised in the statistical offices. For the present study, the original occupational codes were reclassified into 53 occupational categories and one group of economically inactive persons. All Nordic countries have a nation-wide registration of incident cancer cases during the entire study period. For the present study the incident cancer cases were classified into 49 primary diagnostic categories. Some categories have been further divided according to sub-site or morphological type. The observed number of cancer cases in each group of persons defined by country, sex, age, period and occupation was compared with the expected number calculated from the stratum specific person years and the incidence rates for the national population. The result was presented as a standardised incidence ratio, SIR, defined as the observed number of cases divided by the expected number. For all cancers combined (excluding non-melanoma skin cancer), the study showed a wide variation among men from an SIR of 0.79 (95% confidence interval 0.66-0.95) in domestic assistants to 1.48 (1.43-1.54) in waiters. The occupations with the highest SIRs also included workers producing beverage and tobacco, seamen and chimney sweeps. Among women, the SIRs varied from 0.58 (0.37-0.87) in seafarers to 1.27 (1.19-1.35) in tobacco workers. Low SIRs were found for farmers, gardeners and teachers. Our study was able to repeat most of the confirmed associations between occupations and cancers. It is known that almost all mesotheliomas are associated with asbestos exposure. Accordingly, plumbers, seamen and mechanics were the occupations with the highest risk in the present study. Mesothelioma was the cancer type showing the largest relative differences between the occupations. Outdoor workers such as fishermen, gardeners and farmers had the highest risk of lip cancer, while the lowest risk was found among indoor workers such as physicians and artistic workers. Studies of nasal cancer have shown increased risks associated with exposure to wood dust, both for those in furniture making and for those exposed exclusively to soft wood like the majority of Nordic woodworkers. We observed an SIR of 1.84 (1.66-2.04) in male and 1.88 (0.90-3.46) in female woodworkers. For nasal adenocarcinoma, the SIR in males was as high as 5.50 (4.60-6.56). Male waiters and tobacco workers had the highest risk of lung cancer, probably attributable to active and passive smoking. Miners and quarry workers also had a high risk, which might be related to their exposure to silica dust and radon daughters. Among women, tobacco workers and engine operators had a more than fourfold risk as compared with the lung cancer risk among farmers, gardeners and teachers. The occupational risk patterns were quite similar in all main histological subtypes of lung cancer. Bladder cancer is considered as one of the cancer types most likely to be related to occupational carcinogens. Waiters had the highest risk of bladder cancer in men and tobacco workers in women, and the low-risk categories were the same ones as for lung cancer. All this can be accounted for by smoking. The second-highest SIRs were among chimney sweeps and hairdressers. Chimney sweeps are exposed to carcinogens such as polycyclic aromatic hydrocarbons from the chimney soot, and hairdressers' work environment is also rich in chemical agents. Exposure to the known hepatocarcinogens, the Hepatitis B virus and aflatoxin, is rare in the Nordic countries, and a large proportion of primary liver cancers can therefore be attributed to alcohol consumption. The highest risks of liver cancer were seen in occupational categories with easy access to alcohol at the work place or with cultural traditions of high alcohol consumption, such as waiters, cooks, beverage workers, journalists and seamen. The risk of colon cancer has been related to sedentary work. The findings in the present study did not strongly indicate any protective role of physical activity. Colon cancer was one of the cancer types showing the smallest relative variation in incidence between occupational categories. The occupational variation in the risk of female breast cancer (the most common cancer type in the present series, 373 361 cases) was larger, and there was a tendency of physically demanding occupations to show SIRs below unity. Women in occupations which require a high level of education have, on average, a higher age at first child-birth and elevated breast cancer incidence. Women in occupational categories with the highest average number of children had markedly lower incidence. In male breast cancer (2 336 cases), which is not affected by the dominating reproductive factors, there was a suggestion of an increase in risk in occupations characterised by shift work. Night-shift work was recently classified as probably carcinogenic, with human evidence based on breast cancer research. The most common cancer among men in the present cohort was prostate cancer (339 973 cases). Despite the huge number of cases, we were unable to demonstrate any occupation-related risks. The observed small occupational variation could be easily explained by varying PSA test frequency. The Nordic countries are known for equity and free and equal access to health care for all citizens. The present study shows that the risk of cancer, even under these circumstances, is highly dependent on the person's position in the society. Direct occupational hazards seem to explain only a small percentage of the observed variation - but still a large number of cases - while indirect factors such as life style changes related to longer education and decreasing physical activity become more important. This publication is the first one from the extensive Nordic Occupational Cancer (NOCCA) project. Subsequent studies will focus on associations between specific work-related factors and cancer diseases with the aim to identify exposure-response patterns. In addition to the cancer data demonstrated in the present publication, the NOCCA project produced Nordic Job Exposure Matrix (described in separate articles in this issue of Acta Oncologica) that transforms information about occupational title histories to quantitative estimates of specific exposures. The third essential component is methodological development related to analysis and interpretation of results based on averaged information of exposures and co-factors in the occupational categories.