Introduction Assessment of occupational pesticide exposure in epidemiological studies of chronic diseases is particularly challenging. Biomonitoring of current pesticide levels might not correlate with past exposure relevant to disease etiology, and indirect methods often rely on workers’ imperfect recall of exposures, or workers’ job titles.

Within the IMPRESS project (www impress-project.org) we investigated how exposure assessment method (EAM) type for assessing occupational pesticide exposure influenced risk estimates for some chronic diseases.

Methods In three meta-analyses the influence of type of EAM on the pooled risk of prostate cancer (25 articles), Non-Hodgkin’s lymphoma (NHL) (29 articles), and Parkinson’s disease (PD) (34 articles) was investigated using sub-group analyses by type of EAM. Data were mainly obtained from a previous systematic review conducted by the authors. Categories of EAM types used were: group-level assessments (e.g. job titles), self-reported exposures, expert-level assessments (e.g. job-exposure matrices), and biomonitoring. Further sub-group analyses were made by study design and publication year.

Results EAM types were not associated with statistically significantly different pooled risk estimates regarding any health outcome. However, for all health outcomes, case-control studies showed consistently higher risk estimates when expert-level assessments were used compared with self-reports. Overall, case-control designs showed higher risk estimates than cohort designs. Cancer studies showed higher risk estimates in later publications, whereas PD studies showed higher risk estimates in earlier publications.

Conclusion Overall, EAM type in studies of occupational pesticide exposure appears not to affect risk estimates of prostate cancer, NHL, and PD. Nevertheless, in case-control settings self-reported exposures might yield lower risk estimates than expert-based methods, possibly resulting from a higher degree of exposure misclassification due to workers imperfect recall of exposures. In systematic reviews of health effects of occupational exposure to pesticides, study design, year of publication, and exposure assessment method(s) should be taken into account.