Hardell L, Nasman A, Pahlson A, Hallquist A, Mild KH.
This study examined 209 people with brain tumour, identified from records in Swedish cancer centres. The patients were diagnosed during 1994-96. Each subject was matched on sex, age and study region with two controls. In all, 425 controls were included. Exposure to cellular telephones, and to other radiofrequencies and chemicals, was ascertained by postal questionnaire, which was supplemented in some cases by telephone interview.
Seventy eight cases (37.3%) and 161 controls (37.9%) reported use of cellular telephone. The mean number of hours of phone use for cases was 511 (median 136) and for controls 428 (median 136).
The study found no association between the occurrence of brain tumours and the amount of cellular phone use. No association was seen when the number of years of exposure was considered separately. A "somewhat increased risk" that was not statistically significant was found for tumours in the temporal or occipital lobe of the brain on the same side the cellular phone was used. This was found only for users of the analogue NMT system. The risk was 2.56 times greater for the right side and 2.10 for the left side, but was based on only 7 cases on the right and 5 cases on the left. The authors state that "the results were based on low numbers and must be interpreted with caution."
The main weakness of a case-control study such as this is that it might be subject to recall bias i.e. it is more likely that people with a brain tumour remember past cell phone use compared with controls. The authors state that "validation of exposure to cellular phones from telephone companies would have been desirable. However, it turned out to be impossible to get such data."