Editorial by Elisabeth Cardis and Siegal Sadetzki on indications of brain-tumor risk in mobile-phone studies
In their article, the two scientists from the Spanish „Centre for Research in Environmental Epidemiology“ (CREAL) and from the Israeli Tel-Aviv University draw their conclusions from the 2010 in most parts published overall results of the multicenter study INTERPHONE and comment on these results. In their comprehensive assessment they put the results in the context of interpretations by other groups (reaching from “mobile phones are safe” to “mobile phones cause tumors”) and in the context of other individual epidemiologic EMF-studies on brain-tumor risk. As study coordinator, Cardis had an important role in the INTERPHONE study. Sadetzki was the principal investigator of the Israeli study part. The authors deal in particular with the timing of the study, the exposure variables of relevance, and the influence of methodological limitations.
Their key messages are:
- While several studies have found suggestions of an increased risk of brain tumors in relatively long-term users, no such increase was seen in the INTERPHONE main analyses. However, in a subanalysis in Annex 2 of the publication, where short-term users instead of never users were used as the reference category, an increased risk was seen among long-term users, “with an indication of a trend for increasing risk with increasing time since start of use”.
- Analysis of the level of use (in terms of amount and duration) is essential in order to evaluate whether a risk exists. In the INTERPHONE study, the median cumulative call time over life corresponded to only about 27 minutes of phone use per day even in the highest decile of cumulative call time, which is a low use intensity compared with today. Observations in the highest-exposed group are, however, particularly important, as this could be a sentinel group, signaling effects that might be found in studies with higher and more relevant exposure patterns. In INTERPHONE, a 40% increase in risk was seen for glioma (i.e. a benign or malign tumor in brain tissue) in the highest exposed group (highest decile of cumulative call time).
- Analyses of brain tumors without taking into account the hemisphere of their location are likely to “dilute” the risk. The findings in several studies of an increased risk for glioma among the highest users on the side of the head where the phone was normally used and, in INTERPHONE, in the temporal lobe (where most of the energy from the phone gets absorbed) are therefore important. “These are the findings that would be expected if there was a risk, as these are the a priori relevant exposure variables”, the authors say.
- Erroneous recall of the side of the head where the phone was normally held, in particular due to expectations of tumor patients (“laterality recall bias”), and higher refusal rates among population controls than among cases (“selection bias”) are the main biases that affect the results of INTERPHONE and similar studies. There are arguments for and against an underestimation of risk due to these biasing factors. Taken together, the results are most likely a mixture of bias and true effect.
The authors come to the following conclusion:
“It is not possible to evaluate the magnitude and direction of the different possible biases on the study results and to estimate the net effect of mobile phones on the risk of brain tumors. The overall balance of the above-mentioned arguments, however, suggests the existence of a possible association. While more studies are needed to confirm or refute these results, indications of an increased risk in high- and long-term users from Interphone and other studies are of concern.” Until definitive scientific answers are available, Cardis and Sadetzki advise the public, and particularly young people, to adopt simple and low-cost precaution measures, such as the use of text messages, hands-free kits and/or the loudspeaker mode of the phone.
Bibliography: Cardis and Sadetzki, Occup Environ Med published online January 24, 2011, Study Abstract
Extract from the EMF Brief 34 of 3. February 2011