Quantitative exposure assessment in community-based studies

Abstract

Occupational epidemiology focuses on the associations between exposures at the workplace and disease outcomes, essentially concerned with the prevention of disease. Basically two types of studies can be distinguished in occupational epidemiology: industry-based studies which study a population at one or more workplaces, and community-based (also referred to as population-based) studies which study a selection of the general population working at a multitude of workplaces. Community-based studies typically have larger statistical power than industry based-studies when investigating risk factors for rare diseases and diseases with long latency, such as cancer and neurodegenerative diseases. They usually include many cases and offer the opportunity to collect complete work histories and detailed information on potential confounders. Therefore, community-based studies may provide an important complement to industry-based studies. The main limitation of community-based studies, however, is the assessment of occupational exposure. Generally, there is no detailed exposure information available. Exposure is therefore often assessed qualitatively (exposed versus non-exposed) or semi-quantitatively (low/medium/high), while exploring exposure-response relations and expressing (resultant) risks by unit of exposure requires quantitative exposure assessment. Up until now, the latter comes almost exclusively from industry-based studies. In this thesis the quantitative exposure assessment process is described for respirabe crystalline silica (RCS), a known lung carcinogen. Based on quantitative estimates derived from statistical modelling of available historical measurement data of RCS, a quantitative JEM (i.e. SYN-JEM) has been elaborated. The quantitative job-, region/country-, and time-specific SYN-JEM was subsequently applied in a community-based pooled case-control study on lung cancer (SYNERGY, including 17,705 cases and 21,813 controls). With the successful derivation of an exposure-response relation for RCS and lung cancer, we demonstrated the feasibility of assessing occupational exposure quantitatively in a community-based study. Based on this exposure-response relation, we estimated the risk of lung cancer to be 1.034 (95% CI 1.027-1.041) for each two-fold increase in cumulative RCS exposure. This estimate is consistent with previous industry-based studies. However, the large size of the study population from the pooled international community-based studies offers the possibility to perform more in-depth epidemiologic analyses. We can now, for the first time, explore possible interactions with smoking and other occupational exposures and study subtypes of lung cancer (i.e. squamous cell carcinoma, small cell carcinoma, or adenocarcinoma) or exposed sub-groups of interest (e.g. women or never smokers). One of the main requirements to perform exposure assessment in a quantitative way is the availability of sufficient (historical) measurement data. However, owing to the growing amount of exposure measurements digitized in databases, the inclusion of measurement data in exposure estimation in community-based studies has become feasible. In conclusion, the use of quantitative JEMs for community-based studies is an important methodological way forward in the derivation of exposure-response relations of occupational exposures and health effects.