Szmigielski S

This paper examines the relationship between exposure to RF radiation and cancer occurrence in military personnel in Poland during the years 1971 to 1985. Exposure was based on military personnel records of job description, but individual exposure measurements were not available. The EM fields were pulse modulated (150-3500 MHz), and about 80% of the individuals were exposed to less than 2 W/m2. The annual incidence of various cancers was calculated using the total "non-exposed" personnel of approximately 124,000 per year, and the "exposed" population of approximately 3,700 per year. Increased incidences in the exposed versus non-exposed groups were found for several cancers, including the oesophagus and stomach (3.24 times), colorectal (3.19), brain (1.91), and blood and lymphatic systems (6.31).

There are considerable weaknesses in this study. The exposure to EM fields is based on job description, but no details of actual exposure are available, and no information is provided about the duration of exposure. Since information was extracted from medical records on cancer cases, it is possible that more information was available on RF and other exposures on these cancer cases than for healthy personnel. This would introduce a bias that might artificially increase the estimated incidences in the "exposed" personnel. Only annual incidences per 100,000 are given. Actual numbers of cases are not provided. The author states, however, that "one case of malignancy in the group of 3700 exposed subjects during 15 years accounts for the incidence of 1.8 cases per 100,000 annually." This would mean that there would be only 24 cases of malignancies of the blood and lymphatic system in the 15 years of the study. This is a very small number, and a considerable difference to the calculated incidences would arise if exposure was wrongly assigned in even a few cases. No information is given about confounding factors, although the author emphasises that there is potential for electric workers to be exposed to "leukaemiogenic factors". There is no evidence of age adjustment, which could make a considerable difference to the calculated outcomes.

Overall, the paper is very weak in its methodology, and the results must be treated with extreme caution.