- And that Danish study
Posted by walter
on May 22, 2020, 1:44 pm, in reply to "For clarity, the SA article was an opinion piece by Moskowitz himself"
|
Just picking up after the Interphone study, where the writer says: "The dose-response curve from this undertaking is telling, because it clearly does not betray any obvious signs of correlation." That's because the methodology was chosen to dilute the effects. Just like a haystack is 'telling', because it clearly doesn't show any obvious signs it contains a needle. It's a great way to hide a needle. Let's hide another needle. He continues in his merry way, with another study with suspiciously poor, industry-serving methodology: "A similar Danish cohort study also did not reveal any obvious link between phone usage and tumor rates." This study was published in 2011, but the data ran only up to 2007. Thirteen years old data? Two major methodological problems: 1) Potential misclassification of exposure. The study was based on phone records, so didn't know who was actually using the mobile phone. Mobile phones were not as common as they are now. So for example, a sharing couple, or parent either allowing offspring to use the mobile (or giving it to them after a period of their own use), might appear from the data to be one heavy user. This is mentioned at the end under study limitations - which in the world of dodgy mobile phone safety research is where researchers purge themselves of any sin, leaving the media (and especially the 'science' blogosphere) to run the story for them. "Because we excluded corporate subscriptions, mobile phone users who do not have a subscription in their own name will have been misclassified as unexposed. Also, as data on mobile phone subscriptions were available only until 1995, individuals with a subscription in 1996 or later were classified as non-users." The authors did a further analysis which purported to show the misclassification did not affect the conclusion, but were obliged to admit that "this enormously reduced the number of cases". In fact "the incidence rate ratio in men for central nervous system tumours was 1.07 (95% confidence interval 0.86 to 1.34; n=83), and for glioma it was 1.08 (0.77 to 1.51; n=36)." A slight increase in effect but not statistically significant - having lumped opposites together, you can't recover from it by studying them separately later because you have reduced your 'data power'. 2) Another er, slight limitation is that the study didn't measure anything. They didn't have the data on actual use, and to 'measure' exposures, went by the length of time the mobile contract was held! "Another limitation of our study is that the dose- response analyses are based on years since first subscription and we did not have information on the actual amount of mobile phone use. Therefore, we could not examine the risk restricted to the subgroup of heaviest users" https://www.bmj.com/content/343/bmj.d6387 Despite these obvious biases, in the conclusion the study noted that for users of 10 years or more (ie, at least 10 years of subscription) gliomas in both men and women were elevated slightly. This should give cause for concern. The writer is holding up as quality a study where the authors themselves said they could not examine a high dose. The problems with these two studies are obvious. In my view these limitations were guaranteed to mute any genuine effect out of existence. What was telling is that despite large built-in biases, both studies still found something relating the mobile use to gliomas. I'll just finish by noting that the author does nothing to support his claims that they are 'quality'; indeed they have been widely criticized on many counts - possibly they are the two most criticized studies in the field. |