Feb 27 2014

Warning Labels on Cell Phones

Hawaii Senate Bill 2571, which is making its way through the legislature, would require that a large non-removable warning label be attached to the back of every cell phone. Originally the warning label was to read, “This device emits electromagnetic radiation, exposure to which may cause brain cancer. Users, especially children and pregnant women, should keep this device away from the head and body.” A revised version of the bill, however, changed the warning to, “To reduce exposure to radiation that may be hazardous to your health, please follow the enclosed product safety guidelines.”

This seems like an example of clear nanny-state overreach, but worse it is not based on science. According to reports every expert consulted by the relevant committees argued against the measure, but the legislators passed it anyway. The measure has one more committee to get through, and then it would go to the House for a vote.

As I have written before (see also here and here) there is no clear link between cell phone use and brain cancer. The plausibility of a link is low but not zero. Non-ionizing radiation is not energetic enough to break chemical bonds, and therefore should not cause DNA damage that could lead to cancer. However, an alternate physical mechanism cannot be ruled out, and biology is complicated, so I don’t think we can rule out a possible connection on theoretical grounds alone. We can just say it’s unlikely.

The National Cancer Institute summarizes the situation thusly:

It is generally accepted that damage to DNA is necessary for cancer to develop. However, radiofrequency energy, unlike ionizing radiation, does not cause DNA damage in cells, and it has not been found to cause cancer in animals or to enhance the cancer-causing effects of known chemical carcinogens in animals.

Epidemiological studies have been a little mixed, as is generally the case with this kind of evidence, but at this point the evidence is fairly solid that there is no link between cell phone use and brain cancer for exposure less than 10 years. The data for >10 years is less certain, but this is mostly because it takes more time to gather this data and we are just getting to the point that there is epidemiological evidence available for long exposures.

The NCI fact sheet also contains an excellent summary of all the relevant studies and the official opinions of various regulatory and health organizations. The bottom line is that there is no clear link. Epidemiological data cannot rule out a small connection, but it is reassuring when no clear signal is evident. This means, at worst, there is a small effect.

As is also often the case there are outliers. In this research I find one research group, Hardell and Carlberg, consistently find an association, at apparent odds with the rest of the scientific community. Some of their studies are also a little odd, like this one looking at brain cancer survival and cell phone use.  The results are all over the place, and their choice of statistical analysis seems a bit arbitrary. In other words, the data gives off a whiff a p-hacking. Without consistent independent replication, I usually do not put much weight on such outliers.

One other type of evidence that is relevant is the incidence of new cases of brain cancer. If cell phone use is a significant cause of brain cancer, then we would expect the incidence to be going up over the last 20 years as cell phone use has skyrocketed. However, the incidence has been stable to slightly decreasing. In the last 10 years the incidence of brain and other nervous system cancers has decreased on average by 0.2% per year. This is pretty strong evidence against any significant cell phone effect.

Other types of cancers have also been investigated. A recent study from Denmark showed no association between cell phone use and skin cancer. A possible association between carrying cell phone in one’s bra and breast cancer made the rounds in the media, but this is based on a case series of four patients. This is the most preliminary type of evidence, and should only be used as an indication for further research.


The scientific evidence does not point to any cancer or other health risk from cell phone use (unless you text while driving). The International Agency for Research on Cancer, American Cancer Society, National Institute for Environmental Health Sciences, the FDA, CDC, and FCC have all concluded that there is no clear evidence of any risk, although the IARC considers non-ionizing radiation a “possible” carcinogen.

Despite this Hawaiian politicians feel they know better and that they need to warn the public with a giant sticker on the back of every cell phone. If anything the sticker should warn about the risk of texting or using a cell phone while driving. That is a proven risk.

The evidence does not seem to rise to the level that an intrusive warning is indicated. The unintended negative consequence of such labels is warning fatigue. If the public is constantly bombarded with warnings of marginal or just possible risk, they will tend not to take such warnings seriously. They become part of the background noise. Then, when a warning would be actually beneficial, it is lost in this noise.

In this case, politicians should listen to scientists. If only.

17 responses so far

17 thoughts on “Warning Labels on Cell Phones”

  1. zorrobandito says:

    Warning fatigue is a real risk, as we in California can testify. Prop 65 (“there’s a substance somewhere in here which might be harmful”) has chiefly benefited sign-painters. Every business establishment has such a sign. (If you eat the wallpaper paste in here you could die!)

    With the result that no one pays the slightest attention to these signs.

  2. carbonUnit says:

    Please don’t use the term ‘nanny state’, it sounds so… conservative.

  3. Gallenod says:

    I’m not sure why the term “nanny state” should be considered derogatory. Nannies are people we employ to care for and protect persons (generally young children) who are innocent about and incapbable of dealing on their own with the risks and responsibilities in their environments. A true nanny state that followed this role model would only provide appropriate care and regulation and let go when the charges in its care demostrated that they were capable of taking care of themselves.

    I suppose, however, most adults feel they are competent to assess and make their own decisions about risk and responsibility and resist any attempt to protect them from their own folly through societal regulation.

    While that may be the case, significant fractions of the populace frequently demonstrate an inability to both understand reality and act rationally. Belief in psychics, astrology, psuedoscience, and other quackery is a distressingly large part of humanity, not counting delusional belief in their own abilities and “common sense.”

    “Yeah, other people might crash if they text while driving, but not me!”

    But the Hawaii cell phone labeling proposal isn’t something a nanny, who I’d generally consider a reponsible role model, would do. Good nannies base their decisions on reality or they don’t stay hired very long. Labeling cell phones based on unsubstantiated speculation is more of a “fear the boogeyman” tactic. That, to me, is less a nanny state than a “puppet state” with no mind of its own dancing on strings pulled by others.

    But perphaps that’s what we get when electibility trumps capability in our electoral process?

  4. TheFlyingPig says:


    I think the primary objection to the “nanny state” is that the relationship between an individual and the government is not analogous to the relationship of a child and a nanny. Aside from anarchists, most people are OK with the government protecting people from external sources of harm (eg, stopping theft). But nannies must go further and protect children from internal sources of harm and control their activities to a large degree – preventing them from making bad decisions, controlling where they have to be and when they have to be there, forcing them to do chores or other unwanted activities, etc. It’s these latter activities that people who want limited government object to, and it’s not based on the assumption that adults will make good decisions.

    It’s based on the ideology that whereas nannies have the right and responsibility to protect children under their care from themselves and control their behavior to some extent, nobody and no entity has the right to do that to/for adults (extreme circumstances excepted). IOW, adults have the right to full autonomy, even from government; children do not have the right to full autonomy, not even from their nanny. Furthermore, practical considerations aren’t the point of this argument, so even if a society is filled with delusional and harmful beliefs/practices (eg, Christianity, anti-vaxxers), the government does not have the right sanction (in either sense) those practices.

    But yeah, even if you don’t buy my above arguments, this Hawaii cell-phone labeling is a stupid and unscientific version of the nanny state that we can all oppose. It’s even worse than the whole GMO labeling hysteria.


  5. kwillcox says:

    The experiment, led by Yaniv Hamzany, an ear, nose and throat specialist with Tel Aviv University’s Sackler Faculty of Medicine, compared the saliva of 20 heavy cellphone users (those who spoke on their cellphones an average of 30 hours per month) with the saliva of 20 individuals, most of whom are deaf and either don’t use a cellphone or use it for non-verbal functions like texting. The results: The heavy cellphone users showed more signs of oxidative stress.

    What does that term mean? Essentially, cell damage. More specifically, it’s a molecular imbalance between antioxidants and pro-oxidants, or free radicals. Free radicals are a natural byproduct of metabolism but are also linked with aging and, when they outnumber the antioxidants, cell damage occurs.

    Hamzany concludes that exposure to cellphones causes oxidative stress, which harms human cells and DNA and contributes to a “carcinogenic effect.” He adds that his study corroborates the results of so many others, which find correlation, not causation, between cellphones and cancer risk. “It does give us an idea of the harmful effect of the telephone,” he says, advocating caution rather than discontinuing cellphone use. “Maybe use it less often, maybe not expose children to it,” he tells U.S. News. “Maybe there is a kind of additive effect, and in the long run it could cause tumors.”


    Dr. Lennart Hardell and his colleagues in Sweden just published the third in a series of papers on the use of wireless phones, including cell phones and cordless phones, and the risk of malignant and non-malignant brain tumors. The latest paper describes a new case-control study that examines the association between mobile phone use and brain cancer risk. In these studies, the cases were diagnosed with brain tumors between 2007 and 2009. (1)

    The study updates earlier research from case-control studies conducted by the Hardell group and extends the prior research by examining the effects of wireless phone use, i.e., cell phone and cordless phone use, on brain tumor risk for people who have used these phones for up to 25 or more years.

    Overall, the research found that people who used wireless phones for more than a year were at 70% greater risk of brain cancer as compared to those who used wireless phones for a year or less. Those who used wireless phones for more than 25 years were at greatest risk—300% greater risk of brain cancer than those who used wireless phones for a year or less.

    The total number of hours of wireless phone use was as important as the number of years of use. A fourth of the sample used wireless phones for 2,376 or more hours in their lifetime which corresponds to about 40 minutes a day over ten years. These heavier users had 250% greater risk of brain tumors as compared to those who never used wireless phones or used them for less than 39 hours in their lifetime.

    A similar analysis reported in the 13-nation Interphone study funded partly by the World Health Organization found a 182% greater risk of brain cancer among those who used cell phones for 1,640 or more hours in their lifetime.

    In the current study, for all types of wireless phone use, brain cancer risk was found to be greater in the part of the brain where the exposure to wireless phone radiation was highest—in the temporal or overlapping lobes of the brain on the side of the head were people predominantly used their phones.

    Given consistent results from multiple case-control studies that long-term use of mobile phones (i.e., ten or more years) is associated with brain cancer especially near where the phone is predominantly used, the International Agency for Research on Cancer should strengthen its 2011 assessment of radiofrequency energy from “possibly carcinogenic” to “probably carcinogenic” to humans.

    More importantly, governments around the world should heed the results of these studies. The public must be educated about the need to take simple precautions whenever using wireless devices. Governments must strengthen regulatory standards for wireless radiation and must fund research independent of industry to develop safer technologies.

    The paper was published online in the peer-reviewed journal, International Journal of Oncology. The abstract and a link to this paper appears below along with the abstracts for the Hardell group’s two prior papers from this study. All three papers are open access. (1-3)


    (1) Lennart Hardell, Michael Carlberg, Fredrik Söderqvist, Kjell Hansson Mild. Case-control study of the association between malignant brain tumours diagnosed between 2007 and 2009 and mobile and cordless phone use. International Journal of Oncology. Published online September 24, 2013.

  6. ConspicuousCarl says:

    “has chiefly benefited sign-painters.”

    Using paint which, no doubt, contains substances known by the State of California to cause cancer.

  7. kwillcox says:

    I am not surprised that results are all over the place. There is so much money to be made in wireless radiation based products, that studies that show no damage are to be expected. That said, this study stands alone. almost all studies show no effect until someone is calling half an hour for a decade.

  8. Hoss says:

    Hardell has done a number of studies that, so far, couldn’t be replicated.

  9. steve12 says:


    “There is so much money to be made in wireless radiation based products, that studies that show no damage are to be expected. ”

    OK, you must have delved into this issue. Can you give me a run down of the results of these studies broken down whether they were industry vs. non-industry funded?

  10. chrisj says:

    Great article. It is amazing how little respect people have for expert opinion when it goes against what they believe. I just want to add important aside. Steve wrote, “The scientific evidence does not point to any cancer or other health risk from cell phone use (unless you text while driving).” It’s not just texting while driving. Just talking on the phone while driving is probably a health risk too. I think this is important to point out because while it is obvious to everyone and common sensical that texting while driving is dangerous, it might surprise some people that it is dangerous just to talk on a phone and drive (even if it is a hands free headset). http://unews.utah.edu/old/p/062206-1.html

  11. BillyJoe7 says:


    Do you understand why scientists look at ALL the evidence when coming to a CONSENSUS about any scientific question?
    Do you understand that what you are doing is CHERRY PICKING and why this is not science?

  12. kwillcox – I specifically mentioned that Hardell and Carlberg are outliers and their results have not been replicated, so referencing their latest study is odd. If their results were anything close to the truth, there would be a huge signal in the data that everyone would see, and brain cancer rates would be clearly rising, which they are not.

  13. Bill Openthalt says:

    chrisj —

    it might surprise some people that it is dangerous just to talk on a phone and drive (even if it is a hands free headset).

    It is equally dangerous to talk to a passenger, or do anything that distracts from the road and traffic conditions. Ideally, the driver should do nothing else but drive. No radio, no GPS, no squabbling kids. Like that is ever going to happen.

    kwillcox —

    For crying out loud:

    Fewer smokers == less lung cancer.
    More cellphones != more brain tumors.

    This is why we know smoking causes cancer, and cellphones don’t cause brain tumors. End of story.

  14. TheFlyingPig says:

    @Bill Openthalt

    “It is equally dangerous to talk to a passenger…”

    My understanding is that the studies show that it is more dangerous to talk on a cell phone than to an adult passenger. The explanation I’ve read is that adults will modulate the conversation based on traffic conditions. After reading a couple of these studies a while ago, I’ve become very aware of this phenomenon… while cruising on the freeway conversation is normal; but when the driver is making complex maneuvers or has to react to some danger, conversation slows down or stops.

    Dammit, after writing the first paragraph, I did a google search and it seems there are all sorts of claims out there. Oh well, I have so much confirmation bias for the story I told that I’ll stick with it for now.

  15. Psychbot709 says:


    Without having a reference to the source for Hamzany’s statements, I would point out a couple of problems here for consideration. First, what confounding factors could account for the discrepancies found in Hamzany’s study populations? Were the two cohorts randomized to minimize confounders? Second, what biological factors were analyzed from saliva? Third, how is oxidative stress measured reliably, and on what basis could this account for carcinogenic effects? Finally, the nature of the study design is not clear- is it retrospective or prospective? As a minor observation, the introduction to Dr. Hamzany seems like an appeal to authority fallacy.

    As for arguments regarding the nefarious interests of the telecom industry, if worth entertaining, I would point to the example of the tobacco industry. For decades, the tobacco interests pursued a casuistic agenda to discredit the clear statistical correlation between smoking and carcinomas. I think a wiser public (marginally wiser) and a certainly wiser health community are primed to this manipulation. I think that is the overall point of this blog post- there is no clear “smoking gun” link like in the tobacco case ; arguments to the contrary do not fit the sum of the epidemiological evidence ; the biological mechanisms to explain the carcinogenic effect is tenuous ; but, to exercise an abundance of caution, a dose-dependent phenomenon should be followed over time- there is no epidemic of brain cancer to urge for anything else. As with any new technology, unintended and adverse consequences can be expected if not anticipated, but brain cancer does not seem to be one of them for wireless telephony, so far. No nefarious subplot is needed.


    Re: kwillcox comments

    The Hamzany study has a few shortcomings that make it unhelpful in addressing the question. Firstly, the control group is not well matched – speech patterns and saliva production is different between hearing and deaf persons. Secondly, the sample size is too small to have any statistical power.

    The Hardell studies depend on exposure data which is unreliable; a shortcoming shared with Interphone and all other epidemiological studies that rely on retrospective data that are self reported, rather than measured.

    For cellphone case-control studies, patients with brain cancer are identified. These patients are then mailed a questionnaire, and are asked to recall from memory how much they used their cell phones over their lifetimes. Their answers are compared with a control group that does not have brain cancer. This type of study is subject to a number of potential sources of error. The most serious is “recall bias”. Patients with brain cancer who have heard that cellphones might cause brain cancer may unconsciously overstate their past use. The cancer may also have affected their memory.

    So, you really have to ask yourself – how accurate can this data be? Furthermore, how can you know what the actual RF exposure was during cellphone use? Exposure can vary by device model, frequency band, distance from cell site, talk/listen ratio, use of accessories – like headsets, carrying cases, hands-free kits, computer tethering, etc. Cumulative use data are only as good as people’s memory. And, nobody – sick or healthy – can accurately recall their prior use of cellphones over a 25 years period. And, as already mentioned, usage data is a very poor proxy for actual RF exposure measured with appropriate dosimetry systems.

    By the way, Interphone did not report a 182% increased risk. In the 10th decile of recalled cumulative call time, ≥1640h, the OR was 1.40 (95% CI 1.03–1.89) for glioma, and 1.15 (95% CI 0.81–1.62) for meningioma.

    There’s no smoking gun here!

Leave a Reply