Deplorably Bad Study on Mobile Phones And Saliva Published

August 1st, 2013
submit to reddit Share

I have been taking some time off.   I was not planning on posting again until mid next week, but this recent news story is so hideous, it was impossible for me to ignore it.   Recently a study was published which claims to have found marked differences in the saliva of heavy cell phone users.   This would be significant, if it were true, because it could show a direct biological effect on the saliva glands and, by extension, the possibility that this could lead to cancer or another condition.

It has been making the rounds in the mainstream media, as one might expect.  These kind of studies are almost guaranteed to generate a lot of press.

Via Science Daily:

Heavy Cell Phone Use Linked to Oxidative Stress
July 29, 2013 — Scientists have long been worried about the possible harmful effects of regular cellular phone use, but studies so far have been largely inconclusive. Currently, radiofrequency electromagnetic fields, such as those produced by cell phones, are classified as possibly carcinogenic to humans (Group 2B) by the International Agency for Research on Cancer (IARC). A new Tel Aviv University study, though, may bring bad news.

To further explore the relationship between cancer rates and cell phone use, Dr. Yaniv Hamzany of Tel Aviv University’s Sackler Faculty of Medicine and the Otolaryngology Head and Neck Surgery Department at the Rabin Medical Center, looked for clues in the saliva of cell phone users. Since the cell phone is placed close to the salivary gland when in use, he and his fellow researchers, including departmental colleagues Profs. Raphael Feinmesser, Thomas Shpitzer and Dr. Gideon Bahar and Prof. Rafi Nagler and Dr. Moshe Gavish of the Technion in Haifa, hypothesized that salivary content could reveal whether there was a connection to developing cancer.

Comparing heavy mobile phone users to non-users, they found that the saliva of heavy users showed indications of higher oxidative stress — a process that damages all aspects of a human cell, including DNA — through the development of toxic peroxide and free radicals. More importantly, it is considered a major risk factor for cancer.

The findings have been reported in the journal Antioxidants and Redox Signaling.

For the study, the researchers examined the saliva content of 20 heavy-user patients, defined as speaking on their phones for a minimum of eight hours a month. Most participants speak much more, Dr. Hamzany says, as much as 30 to 40 hours a month. Their salivary content was compared to that of a control group, which consisted of deaf patients who either do not use a cell phone, or use the device exclusively for sending text messages and other non-verbal functions.

Compared to the control group, the heavy cell phone users had a significant increase in all salivary oxidative stress measurements studied.

“This suggests that there is considerable oxidative stress on the tissue and glands which are close to the cell phone when in use,” he says. The damage caused by oxidative stress is linked to cellular and genetic mutations which cause the development of tumors.

The fact that this was even published leads me to believe that the journal in question must have extremely poor standards for peer review.

The number of study subjects is pretty small, and that itself would call into question any findings.   However, I will not bother critiquing the statistical distribution or significance of the study, because none of that actually matters, and doing so would dignify the validity of the study’s methods.   In fact, regardless of how dramatic and significant the findings of such a study are, they are irrelevant to the debate on mobile phones and health because of a massive foundational flaw in the study.

When conducting a study of this type, it’s important to have a good demographic match between the study group and the control group.   The two groups should be as close as possible in terms of age, gender, ethnicity, background and so on.   Otherwise, the subjects should be randomly selected.   This assures that the only thing different between the two groups is the variable being studied.  In this case, that would be mobile phone usage.

Therefore, to conduct a study on mobile phone usage, you would want two groups that are very similar but where one group consists of heavy phone users and the other of those who rarely or never use a mobile phone.   That’s getting increasingly difficult in this day and age, but it should still be possible to find those who do not use mobile phones at all, and therefore would have to simply compare very heavy users to those who only use their phone for talking occasionally.  This, of course, could be verified through billing data.

The authors of this study instead chose to use deaf individuals for their control group.  Although this does offer a population sample that would not be using mobile phones, it is such a poor demographic match, it undermines any conclusion of the study.   In effect, the study simply shows that there is an apparent difference in salivary oxidative stress between deaf and non-deaf individuals.

This is hardly a groundbreaking discovery.   The demographics between deaf and non-deaf members of the population are pretty obvious.  Deaf people are not likely to be involved in certain jobs, such as those which include a lot of direct contact with the public.   They would be more likely to work in closed environments, where direct communications with the general population is not a requirement.   Deaf individuals are likely to have attended special schools and socialize in more closed groups.  They certainly would not be frequent patrons of concerts or other events where hearing is important.   Also, since deafness can be caused by any number of medical, environmental or congenital conditions, it would be expected that a history of certain medical conditions would be especially high in the deaf community.

But above all else, deaf persons generally do not talk.   Of course, some deaf people do have the ability to speak, having been taught to produce speech sounds, even without the ability to hear them. However, such individuals still are not likely to speak anywhere near as much as those who are not deaf and would have learned how to speak much later in life.   On the other hand, those who can hear, speak all the time without much thought.  This is hugely important, because the act of talking stimulates saliva production.   It also forces a person to breathe through their mouth, resulting in more moisture evaporating and the potential for a dryer mouth.

Since speaking has such direct and obvious effects on a person’s mouth and saliva glands, it is absolutely ridiculous to consider a group of deaf and non-deaf persons to be an even remotely good demographic match for something like saliva tests.


This entry was posted on Thursday, August 1st, 2013 at 10:49 am and is filed under Bad Science, inverse square, Not Even Wrong, Obfuscation. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
View blog reactions

9 Responses to “Deplorably Bad Study on Mobile Phones And Saliva Published”

  1. 1
    DV82XL Says:

    I was going to e-mail you about this one because is so bad but resisted because you were taking a break. I’m glad you posted because this is breathtakingly bad science. I was particularly disappointed to see it posted in Science Daily which usually has more sense.


    Quote Comment
  2. 2
    John Says:

    The actual study in question came out almost a year ago, so I am not sure why it hit the presses just now, but I have seen it reported in a bunch of places online. I guess one published it and others followed.

    If you want the whole study you can find links on this page http://www.ncbi.nlm.nih.gov/pubmed/22894683

    The full study text is not available for free. You need a journal service or pay for a copy.

    Maybe someone can track down a place it was posted in its entirety.

    But I agree that without even knowing the distribution of the results, the control is so bad it invalidates any conclusion.


    Quote Comment
  3. 3
    DV82XL Says:

    Can’t find a free copy to read the whole study just yet, but I see that it has been pointed out by several reviewers that the conclusion that the condition of deafness lowers salivary oxidative stress levels would be the move valid one given the design of this study.


    Quote Comment
  4. 4
    Blubba Says:

    Rather than say the study is bad it might be more appropriate to say the study was overhyped by the media. I am not a biostatistician nor do I perform studies for a living but my understanding is that correlational studies like this are limited by their very nature because they don’t allow cause and effect to be determined directly, such as with the correlation between smoking and lung cancer. The study abstract clearly states their observations lead them to a hypothesis that the use of mobile phones may cause oxidative stress and modify salivary function. Hypothesis, not conclusion. Subsequent studies may well determine that heavy phone users tend to have higher stress jobs, like lawyers and realtors, while deaf people are more likely to work as librarians and florists. And yet other studies may determine (if they haven’t already) that people with higher stress jobs show it in their spit chemistry. Or subsequent studies may show that the differences persist even after discounting for such affects. The point is I have no reason to believe the study was poorly done…as far as it went. The bad guys are the kooks and lazy reporters who are giving it more press than it deserves.


    Quote Comment
  5. 5
    DV82XL Says:

            Blubba said:

    Rather than say the study is bad it might be more appropriate to say the study was overhyped by the media.

    While it is a valid point that this study has gotten more coverage in the general media than it deserves, the fact is that it was very poorly done. Only twenty subjects in test group and a clearly inappropriate control group are the first indications that something is very wrong and this is followed by frankly unsupportable conclusions. Regardless, this study doesn’t pass the basic smell test for good science and can be rejected on this factor alone.


    Quote Comment
  6. 6
    Blubba Says:

            DV82XL said:

    While it is a valid point that this study has gotten more coverage in the general media than it deserves, the fact is that it was very poorly done. Only twenty subjects in test group and a clearly inappropriate control group are the first indications that something is very wrong and this is followed by frankly unsupportable conclusions. Regardless, this study doesn’t pass the basic smell test for good science and can be rejected on this factor alone.

    Wouldn’t the sample size needed to support the statement that there was a significant increase in salivary oxidative stress indices in mobile phone users depend on how large the increase was? If they consistently found that the levels were, say, an order of magnitude larger than the controls twenty subjects may have been more than enough to be suggestive for all I know (which is all it appears the researchers were claiming) whereas an average difference of 10% might be inconclusive given the sample size. Without seeing the data and methodology I don’t see how anyone can make a definitive statement one way or the other.


    Quote Comment
  7. 7
    drbuzz0 Says:

            Blubba said:

    Wouldn’t the sample size needed to support the statement that there was a significant increase in salivary oxidative stress indices in mobile phone users depend on how large the increase was? If they consistently found that the levels were, say, an order of magnitude larger than the controls twenty subjects may have been more than enough to be suggestive for all I know (which is all it appears the researchers were claiming) whereas an average difference of 10% might be inconclusive given the sample size. Without seeing the data and methodology I don’t see how anyone can make a definitive statement one way or the other.

    I would say it does not matter how large the difference is. If they found that every single study group member had a much higher level than even a single member of the “control” group, that only proves that deaf people have a lower level than those who are not deaf.


    Quote Comment
  8. 8
    DV82XL Says:

            Blubba said:

    Wouldn’t the sample size needed to support the statement that there was a significant increase in salivary oxidative stress indices in mobile phone users depend on how large the increase was?

    I agree with Steve the experimental design is so deeply flawed it is impossible to say the results support any meaningful conclusion.


    Quote Comment
  9. 9
    Chris Says:

    “On the other hand, those who can hear, speak all the time without much thought.”

    There is so much rich material in that one little sentence……

    I fully agree with your argument – apples and oranges.


    Quote Comment

Leave a Reply

Current month ye@r day *

Please copy the string FuGsBk to the field below:

*

Protected by WP Anti Spam