We thank Dr. Hocking for his useful comments concerning our study on occupational risk factors and high grade and low grade glioma. We agree that there are some methodological limitations in this study, which was carried out in the late 80s. For example, assessment of exposure to different types of electromagnetic fields could not be investigated at that time in the same way that would be possible today. We also agree that misclassification is a potential problem. It might indeed be difficult for people suffering from severe brain tumours to answer detailed questions on specific exposures associated with employment. Although all study centres used well-trained interviewers, the possibility cannot be excluded that recall bias due to lack of memory may have influenced the information on past exposure to occupational agents. This potential problem is an issue for most other brain tumour studies.
In our present study, efforts were made to schedule interviews as soon as possible after diagnosis. For cases who were too ill to answer questions, or who had already died, we interviewed proxy subjects in order to minimize bias that would have occurred if very severe cases had been excluded. Although we are aware of the problematic nature of information given by proxies, we decided to include their information in the analyses because sensitivity tests, including and excluding the proxy interviews, did not change the estimation of the risks substantially. Because people may not be aware of all the substances to which they are exposed at work, our main emphasis concerning interpretation of the results is focussed more on occupational activities than on substance exposure. Because of these problems, we are very careful in the interpretation of our results, especially when odds ratios did not reach statistical significance or when results were not consistent between the study centers.
Numerous other epidemiological studies have investigated the role of occupational exposures in the development of brain tumours with inconsistent findings. Despite the limitations mentioned above, our results are based on one of the largest case-control-studies on incident brain tumours in adults, and the data do not suggest that occupational exposure is a major risk factor for glioma development.