Arvo building, seminar room B342, address: Arvo Ylpön katu 34
Doctoral defence of M.Sc. Sie Sie Siew
The field of science of the dissertation is Occupational Health.
The opponent is Associate Professor Beate Pesch (Ruhr University Bochum, Germany). Professor Clas-Håkan Nygård acts as the custos.
The language of the dissertation defence is English.
Occupational exposures (to wood dust, iron and welding fumes) and risk in specific respiratory cancers
The aim of the study was to explore associations between selected occupational exposures and the targeted cancer sites in the entire national workforce. Occupational exposures were estimated from occupational titles obtained from the censuses. Quantitative job exposure estimates were estimated according to the Finnish national job exposure matrix (FINJEM) or the Nordic Occupational Cancer Study job exposure matrix (NOCCAJEM) that cover more than 300 specific job titles and exposures in four periods: 1945–59, 1960–74, 1975–84, to 1985–94. The Nordic study cohort consists of all male workers who had participated in the censuses 1970–1990 in Finland, 1960–1980 in Norway, 1960–1990 in Sweden, and 1981 in Iceland. Three other papers specifically included the Finnish part of cohort i.e. the cohort in Finland only. The range of population covered was between 1.2 million and 14.9 million in four Nordic countries.
Cancer cases in this thesis were diagnosed in 1961–2005. Altogether 44,492 incidence cases of primary lung cancer were included, based on the data in the Finnish Cancer Registry. The numbers of men diagnosed with primary nasal cancer and nasopharyngeal cancer were 292 and 149 in Finland, and 2839 and 1747 in four Nordic countries.
For each occupation with exposure to ‘iron and welding fumes’, standardized incidence ratios (SIRs) of lung cancer were calculated (i.e. ratio of observed to expected number of cases). SIRs for nasal cancer, nasopharyngeal cancer and lung cancer were calculated in occupations with exposure to wood dust or formaldehyde. Relative risks (RRs) of each cancer were calculated by comparing three categories of cumulative exposures (CEs) with the unexposed category using Poisson regression models and with an adjustment for confounding factors.
Hazard ratios (HRs) were calculated for individual-level data derived from occupational history recorded in all available census data in conditional logistic regression models. Work-related co-exposures (e.g. asbestos and silica) and confounding factors (socio-economic status, SES, and smoking) were adjusted whenever available.
The results demonstrated that in comparison to the unexposed, workers most heavily exposed to iron/welding fumes had 40% more lung cancer risk, when asbestos and silica were controlled, without adjustment for SES and smoking. When these two non-occupational risk factors were also controlled, the RR decreased but remained >1 in all CE categories. One novel finding was the detection of an association between one histological subtype of lung cancer, squamous-cell lung carcinoma, and exposure to iron fume and dust (2-fold risk) as well as to welding fumes (55% excess).
Independent roles between CE to iron and welding fumes could not be properly distinguished due to high correlation of both exposures in same individuals. Co-exposure to asbestos contributed up to 40% greater lung cancer risk at the heaviest asbestos exposure level (>13.23 fibers/cm3-years). Risk attributable to co-exposure to silica was small. In general, Finnish workers exposed to the greatest level of wood dust had 60% more risk to develop nasal cancer as compared to non-exposed workers. Elevated risks at all CE levels were above RR 2.0 in the study extended to workers in four Nordic countries. Drastic excess from adenocarcinoma histological subtype appears to dominate the risk. The most heavily exposed workers (CE ≥ 28.82 mg/m3-years) has a 29-fold risk of nasal adenocarcinoma. Experiencing light wood dust exposure (≤ 6.70 mg/m3-years) led to a three-fold risk even after formaldehyde exposure was controlled in the model.
In this study, the findings with wood dust exposure confirmed the risks in nasal cancer, which is in line with the increased risks detected by the previous studies, but no association with nasopharyngeal cancer. Weak but persistent lung cancer risk was observed in all levels of exposure to iron and welding fumes. Detection of risks were noticeable already at the lowest CE level when regulatory limits were applied, which rises the concern relating to adequacy of existing occupational exposure standards or recommended guidelines.
This study provides reasonable confidence in results due to the strengths and quality in size, reliability of nationwide registered cancer data, and at least partial adjustment for exposures contributed from both work-related and non-work-related confounding factors. Assigning numerical exposure estimates to workers in the absence of personal monitoring data is challenging. Large-scale record linkage is a considerably efficient solution as it can generate a tremendous volume of nationwide and even multinational data on decade-long retrospective work history.
The dissertation is published in the publication series of Acta Universitatis Tamperensis; 2376, Tampere University Press, Tampere 2018. The dissertation is also published in the e-series Acta Electronica Universitatis Tamperensis; 1882, Tampere University Press 2018.