The authors after a review come to the conclusion that studies of electromagnetic effects can be divided in two groups; those that are funded by telecommunication industry and those that are not, with the first having numerous procedure fault to arrive to the conclusion of no effect of telephony, while others financied by public funds are more correctly designed and have other conclusion.
" Uncertainty about the association between health risks and exposure to radiofrequency radiation emitted by cellular and cordless mobile phones can be addressed by a critical analysis of the methodology used in studies assessing this relationship. Studies funded by cellphone companies give reassuring conclusions but are affected by biases and flaws, whereas public-funded studies are without these errors and show acute and chronic effects, including head tumors, findings supported by biological evidence. "
For example (taken from table 2) here are some of the methodology errors they detect in the Interphone negative studies, based on a “non-blind" protocol; and the reliability of Hardell positive studies, based on a "double-blind” protocol.
|▪ inadequacy of "regular use of cell phones" defined as "at least 1 phone call/week, for at least 6 months": 2-5 min/day, often for <5 years;||▪ MP use is significant: from over 16 to over 32 min/day, for ≥10 years;|
|▪ inadequate exposure or latency time in relation to time required for diagnosing the tumors concerned: fewer than 5% of cases have latency time ≥10 years;||▪ 18% of cases were exposed for or from ≥10-15 years;|
|▪ fails to include cordless users, subjects younger than 30, and people living in rural areas, even though these groups have high exposure;||▪ includes them;|
|▪ fails to distinguish tumor laterality in relation to laterality of MP-use;||▪ tumor laterality is always considered in relation to MP-use laterality;|
|▪ fails to consider other types of malignant and benign head tumor, except for astrocytomas, neuromas, meningiomas and salivary gland tumors (1 study);||▪ other types of head tumor are considered separately;|
|▪ participation of controls is reduced to 60%, at times <40%, with prevalence of the exposed;||▪ exposed and non-exposed controls participate in equal proportion and at high percentage (nearly 90%);|
|▪ the patient, interviewed face-to-face when in a confused state during the post-operatory period, may report the recent laterality of use which, as a result of the disturbances brought about by the tumor, may not be the side habitually used before tumor development;||▪ the data are collected through a questionnaire sent to cases on dismissal from hospital, when they are recovering;|
|▪ the negative findings are publicized as fully reassuring even though these at times include positive data indicative of increased carcinogenic risk, eg. for only ipsilateral tumors, or only in the subgroup exposed for ≥10 years, or only in residents in rural areas.||▪ the currently positive findings are correctly examined.|
They review cellphone-company funding of studies in this very interesting reading.