Ars Technica on that lousy rat cell phone study:
Ars Technica on that lousy rat cell phone study:
The study, which was not properly peer reviewed—despite what some outlets have reported—is chock full of red flags: small sample sizes, partially reported results, control oddities, statistical stretches, and a slim conclusion. In short, “there is nothing in this report that can be regarded to be statistically significant," Donald Berry, a biostatistics professor at the University of Texas MD Anderson Cancer Center, told Ars. "The authors should have used the 'black box warning.'"
If cellphone radiation really does cause cancer, this study wouldn’t have proven it. And the mountains of preexisting data on the topic all point to mobile devices as posing zero to very low risks. This includes a recent Australian study that found no significant increases in brain cancer since the introduction of mobile phones.
So how did this study grab headlines? First, the authors are researchers at the National Toxicology Program (NTP), which years back received millions of dollars from the government to set up this experiment, the largest animal study to date on the subject. In carefully designed and expensive setups, researchers exposed more than 2,000 rats and mice to wireless frequencies using two common signal modulations: Code Division Multiple Access (CDMA) and Global System for Mobile Communications (GSM). In batches of 90, rats got full-body exposure to 900MHz frequencies of both types of signal modulations. Mice received the same treatment, but at 1900MHz frequencies. Exposure came in 10 minute stretches over 18 hours, with a total of nine hours a day, seven days a week. In humans, that would equate to a lot of phone time.
After collecting two years' worth of data, the researchers are now reporting the results from the rats. In short, the researchers report “low incidence” of brain and heart cancers (malignant gliomas in the brain and schwannomas in the heart) in male rats compared to controls. Female rats showed no increased incidence of cancers at all. The mouse data has yet to be released.
While the sex differences in the rats are unexplained, the larger concern lies with the control rats, which were not exposed to any cellphone radiation. Historically, the types of rats used in the study, a breed called Sprague Dawley rats, have on average about a one or two percent chance of getting either of the two cancers. In some studies, cancer rates in Sprague Dawley controls ran as high as six to eight percent. But in the cellphone study, none of the control rats had either type of cancer at all. And these cancer-free controls died early.
Why the rats exposed to cellphone radiation would live longer than control rats is another unexplained quirk of the study. But it’s potentially a big problem since the cancers that were found in the exposed rats tend to develop later in the rats’ lives. So, if the control rats had lived just a bit longer, they perhaps would have developed one of the cancer types. And just one control rat turning up with cancer would eliminate the statistical significance of the cancer link in male rats.
“The results might just be random false positives,” Christopher Schmid, a biostatistician and founder of the Center for Evidence Based Medicine in Public Health at Brown University, told Ars. “I would need more data to be convinced and more consistent scientific rationale for the findings.”
Yet, with the potential that the low incidence of cancer is real and with the huge swath of the human population that could be affected, the authors felt compelled to publish the preliminary finding. However, they didn’t take the traditional route of submitting their rat findings to a peer-reviewed journal. This would involve having anonymous scientific peers look over the data and analyses and an editorial board decide whether the study is worthy of being published in their august pages. Instead, the NTP researchers selected their own reviewers to comment on the study and then published the study and the reviewers’ comments on a pre-publication website called bioRxiv.
Interestingly, the three selected reviewers were largely critical of the study. One reviewer, Michael Lauer of the National Institutes of Health’s Office of Extramural Research, succinctly concluded, “I am unable to accept the authors’ conclusions,” after questioning the authors’ methods, statistics, and reporting of partial results.
“Why aren’t we being told, at least at a high level, of the results of other experiments... ” Lauer questioned. “In the absence of knowing other findings, we must worry about selective reporting bias.”
http://arstechnica.com/science/2016/05/study-that-found-cell-phones-cause-cancer-in-rats-is-riddled-with-red-flags/
The study, which was not properly peer reviewed—despite what some outlets have reported—is chock full of red flags: small sample sizes, partially reported results, control oddities, statistical stretches, and a slim conclusion. In short, “there is nothing in this report that can be regarded to be statistically significant," Donald Berry, a biostatistics professor at the University of Texas MD Anderson Cancer Center, told Ars. "The authors should have used the 'black box warning.'"
If cellphone radiation really does cause cancer, this study wouldn’t have proven it. And the mountains of preexisting data on the topic all point to mobile devices as posing zero to very low risks. This includes a recent Australian study that found no significant increases in brain cancer since the introduction of mobile phones.
So how did this study grab headlines? First, the authors are researchers at the National Toxicology Program (NTP), which years back received millions of dollars from the government to set up this experiment, the largest animal study to date on the subject. In carefully designed and expensive setups, researchers exposed more than 2,000 rats and mice to wireless frequencies using two common signal modulations: Code Division Multiple Access (CDMA) and Global System for Mobile Communications (GSM). In batches of 90, rats got full-body exposure to 900MHz frequencies of both types of signal modulations. Mice received the same treatment, but at 1900MHz frequencies. Exposure came in 10 minute stretches over 18 hours, with a total of nine hours a day, seven days a week. In humans, that would equate to a lot of phone time.
After collecting two years' worth of data, the researchers are now reporting the results from the rats. In short, the researchers report “low incidence” of brain and heart cancers (malignant gliomas in the brain and schwannomas in the heart) in male rats compared to controls. Female rats showed no increased incidence of cancers at all. The mouse data has yet to be released.
While the sex differences in the rats are unexplained, the larger concern lies with the control rats, which were not exposed to any cellphone radiation. Historically, the types of rats used in the study, a breed called Sprague Dawley rats, have on average about a one or two percent chance of getting either of the two cancers. In some studies, cancer rates in Sprague Dawley controls ran as high as six to eight percent. But in the cellphone study, none of the control rats had either type of cancer at all. And these cancer-free controls died early.
Why the rats exposed to cellphone radiation would live longer than control rats is another unexplained quirk of the study. But it’s potentially a big problem since the cancers that were found in the exposed rats tend to develop later in the rats’ lives. So, if the control rats had lived just a bit longer, they perhaps would have developed one of the cancer types. And just one control rat turning up with cancer would eliminate the statistical significance of the cancer link in male rats.
“The results might just be random false positives,” Christopher Schmid, a biostatistician and founder of the Center for Evidence Based Medicine in Public Health at Brown University, told Ars. “I would need more data to be convinced and more consistent scientific rationale for the findings.”
Yet, with the potential that the low incidence of cancer is real and with the huge swath of the human population that could be affected, the authors felt compelled to publish the preliminary finding. However, they didn’t take the traditional route of submitting their rat findings to a peer-reviewed journal. This would involve having anonymous scientific peers look over the data and analyses and an editorial board decide whether the study is worthy of being published in their august pages. Instead, the NTP researchers selected their own reviewers to comment on the study and then published the study and the reviewers’ comments on a pre-publication website called bioRxiv.
Interestingly, the three selected reviewers were largely critical of the study. One reviewer, Michael Lauer of the National Institutes of Health’s Office of Extramural Research, succinctly concluded, “I am unable to accept the authors’ conclusions,” after questioning the authors’ methods, statistics, and reporting of partial results.
“Why aren’t we being told, at least at a high level, of the results of other experiments... ” Lauer questioned. “In the absence of knowing other findings, we must worry about selective reporting bias.”
http://arstechnica.com/science/2016/05/study-that-found-cell-phones-cause-cancer-in-rats-is-riddled-with-red-flags/
Bullshit, IOW.
ReplyDeleteI'M IMMORTAL!!!!
ReplyDelete