30 Minutes

In 2010 media reports incorrectly claimed half an hour of mobile phone use per day increased the risk of brain cancer, but this was a misinterpretation of the results of the 13 country INTERPHONE project. The false claim is based on a subset of results from a very small group of cancer patients in the study who reported they used their mobile phone for more than 12 hours a day (not 30 minutes a day) - which is highly unlikely.

Follow-up INTERPHONE validation studies found evidence that people diagnosed with a brain tumour over-reported their past mobile phone use and that this 'recall bias' was more likely if subjects perceive that mobile phone use is associated with brain tumours, as has been widely speculated in the media.

Because of these potential biases the INTERPHONE researchers specifically warned against focusing on extreme values: "Rather than focus on the most extreme values, the interpretation should rest on the overall balance of evidence."

The overall results of the study - which included the 'heavy user' subset - found there was no link between the two most common types of brain cancers and mobile phone use: "Overall, no increase in risk of glioma or meningioma was observed with use of mobile phones."

The 30 minute figure is a misinterpretation of the amount of use reported by INTERPHONE's heaviest users who all reported more than 1640 hours lifetime use spread out over 10 years - this corresponds to about a half-hour per day. But this is just a cut off point for 'heavy users' in the study and, as pointed out earlier, the patients in this group reported much higher than 1640 lifetime hours and this over-reporting (i.e. an improbable more than 12 hours day of actual usage) is likely to have caused the false link which did not show up in the overall combined results.

Unfortunately, a study of this size and significance that found no overall link should have helped to dispel some people's concerns but the early misreporting of some selective and biased results may have increased concerns.