I’ve written several times over the years about the overblown claims of harm attributed, largely—but not exclusively—by cranks, to cell phone radiation. It’s been claimed that radiation from cell phones can cause brain tumors (there’s no convincing evidence that this is true), breast cancer (the evidence for these claims is so incredibly flimsy—and featured by Dr. Oz, to boot!—that this is not a credible claim), and a wide variety of other health issues. Indeed, if you believe the cranks, the mobile phone companies are the equivalent of tobacco companies denying that their products cause massive harm.
Never mind that, based on pure physics alone, it is incredibly unlikely that radio frequency radiation can cause cancer. I’ve described it many times, but it’s always worth describing again. This particular form of radiation is just too low energy to break chemical bonds in DNA, a prerequisite for most mutations. On the other hand, a lot of the debunkings of the claimed cell phone-cancer link often rely on what I like to refer to as a “Cancer Biology 101” understanding of how cancer develops in which DNA strand breakage is an absolute requirement for the development of cancer, and I’ve gotten into trouble for being more open to the possibility that there might actually be a plausible biological mechanism by which cell phone radiation might cause cancer. (Usually, it’s physicists, hence the “Cancer Biology 101” understanding.) As I like to say whenever this topic comes up, it is highly implausible based on basic science that cell phone radiation could cause cancer. It’s not homeopathy level-implausible, but it’s pretty implausible. Nor is it impossible, as has been claimed, because there may be biological mechanisms behind cancer that we do not yet understand, and it’s almost always physicists with little knowledge of epigenetics and other mechanisms of cancer development who make such dogmatic claims. Absent compelling evidence of a link between cell phones and cancer, then, it is not unreasonable to rely on the basic science and consider the possibility of such a link to be remote.
Although I’ve dealt with the issue of cell phones and cancer time and time again, it occurs to me that I haven’t dealt with another claim frequently made about adverse health events due to cell phone radiation, namely that it’s deadly to our vital essence sperm. The other day, I saw a ridiculously fear mongering article in The Telegraph entitled Mobile phones are ‘cooking’ men’s sperm:
Fertility experts are warning man that using a mobile for as little as an hour a day is “cooking sperm” and lowering level significantly.
The new study shows that having a mobile phone close to the testicles – or within a foot or two of the body – can lower sperm levels so much that conceiving could be difficult.The findings have led to a leading British fertility expert to advise men to stop being addicted to mobile phones.
The study – by highly respected specialists – found that sperm levels of men who kept their phones in their pocket during the day were seriously affected in 47 per cent of cases compare to just 11 per cent in the general population.
Professor Martha Dirnfeld, of the Technion University in Haifa, said: “We analysed the amount of active swimming sperm and the quality and found that it had been reduced.
“We think this is being caused by a heating of the sperm from the phone and by electromagnetic activity.”
What utter rot!
This is irresponsible journalism (and science communication, to boot—I mean you, Prof. Dirnfeld!) at its worst. Basically, the message is: Hey, you men! Stop being addicted to your mobile phones! They’re sapping your precious bodily fluids frying your sperm! Stop being so into your cell phone, you idiots! Of course, that might not be such bad advice, but why it might be reasonable advice has absolutely nothing to do with sperm counts, sperm motility, or infertility. If it’s good advice, the reason has more to do with being less obsessed with a handheld device and more interested in what’s going on around oneself.
Whenever I see an article like this, I always have to go straight to the study. That proved to be harder than usual, which lead me to curse journalists for not providing direct links, DOI numbers, or even formal references to studies that they report on. Eventually, I did find the paper. It’s an article in Reproductive BioMedicine Online by Zilberlicht et al, entitled Habits of cell phone usage and sperm quality – does it warrant attention? At this point, I was tempted to invoke Betteridge’s Law of Headlines and leave it at that, just answering no, but I’m a glutton for a study, even a bad one. It’s one of my blogging specialties.
So what about the study? Let’s just say that the abstract is not promising:
Male infertility constitutes 30–40% of all infertility cases. Some studies have shown a continuous decline in semen quality since the beginning of the 20th century. One postulated contributing factor is radio frequency electromagnetic radiation emitted from cell phones. This study investigates an association between characteristics of cell phone usage and semen quality. Questionnaires accessing demographic data and characteristics of cell phone usage were completed by 106 men referred for semen analysis. Results were analysed according to WHO 2010 criteria. Talking for ≥1 h/day and during device charging were associated with higher rates of abnormal semen concentration (60.9% versus 35.7%, P < 0.04 and 66.7% versus 35.6%, P < 0.02, respectively). Among men who reported holding their phones ≤50 cm from the groin, a non-significantly higher rate of abnormal sperm concentration was found (47.1% versus 11.1%). Multivariate analysis revealed that talking while charging the device and smoking were risk factors for abnormal sperm concentration (OR = 4.13 [95% CI 1.28–13.3], P < 0.018 and OR = 3.04 [95% CI 1.14–8.13], P < 0.027, respectively). Our findings suggest that certain aspects of cell phone usage may bear adverse effects on sperm concentration. Investigation using large-scale studies is thus needed.
Wow! Sounds damning, doesn’t it? Well, on the surface it does. Let’s dive deeper, as I am wont to do with studies like this.
First of all, let’s look at the central hypothesis, namely that cell phone radiation causes decreased sperm count and motility, both of which are associated with male factor infertility, for obvious reasons. There’s only one plausible biological mechanism (and, even then, it’s not so plausible) to explain how cell phone radiation might decrease sperm count and motility. After all, it’s well known that increased temperature is associated with decreased sperm count and quality. It’s the very reason that fertility doctors recommend that men being evaluated for infertility wear boxers instead of briefs.
So what about this study?
One thing I noticed is that the study in question was published in September 2015. This lead to me wonder: Why is there news coverage of this paper now, given that it was published nearly six months ago? I can only speculate that some editor somewhere was looking for a story and heard about this. (I mean, come on. Who can resist a headline like Great balls of fire! Mobile phones are ‘cooking’ men’s sperm?) I don’t know about Martha Dirnfeld, but if someone called me about a paper I published in February, I’d wonder what was going on, although I’d probably appreciate the attention and cooperate the way that Prof. Dirnfeld did.
The study itself is fairly unremarkable. It is an unrandomized study of men presenting to Carmel Medical Centre for a first time semen analysis as part of an infertility workup. So right there, we know that this is not a representative sample of men. There is no “normal” control group because all of these men were being evaluated for male factor infertility. The study is a relatively small one, only 106 men, of which 26 men were excluded because they didn’t meet the study criteria, which means only 80 men were analyzed. Here are the things they were asked about:
This included questions regarding their demographic background, i.e. age, place of living, number of children, occupation, ethnicity and educational status. There were also questions on their general medical history and fertility-related conditions (i.e. varicocele, orchitis), as well as lifestyle habits such as smoking and consumption of alcohol. Further questions accessed information about daily habits of cell phone usage such as the number of devices used and the duration of daily use (talking). The latter was classified by four categories: less than 30 min, 30–60 min, 60–120 min and over 120 min. The usual location of the device while talking, carrying and charging was assessed separately.
And:
Data regarding the use of accessories such as hands-free devices and earphones were collected as well. Other variables included the number of years that an individual owned a cell phone, talking while the device is being charged (as a categorical yes/no question) and talking in low reception areas (defined as: elevators and underground floors). Information on cell phone types, models and frequencies was not collected.
So right away, I see a number of problems. First, not collecting information on cell phone types is a big issue, as different models emit different amounts of radiofrequency radiation. A more problematic issue is a major assumption behind the study, specifically that time spent talking on one’s mobile phone correlates with exposure of a man’s genitals to the evil magic radio waves. First, even though the data for this study were collected in 2011 and 2012 (one wonders why it took so long to analyze the data, given that this was a prospective study), by then, time spent speaking arguably no longer represented a good measure of cell phone use by then. After all, the iPhone 4 was released in 2011. A better question would be how long one is on one’s mobile phone, either speaking or surfing the Internet. I don’t know about you, but I hardly ever use my iPhone for telephone conversations any more; at least 90% of its use is to access the Internet.
There’s another problematic issue with this study. I don’t know about most men, but when I’m using my phone, be it to have a conversation, to text, or surf the Internet, it’s nowhere near my family jewels, if you know what I mean. Even so, these results are far from convincing. For example, there are a lot of unimpressive p-values here. The finding that talking duration of more than one hour per day was associated with a higher rate of abnormal sperm concentration than talking less than one hour per day resulted in a p<0.04, which probably means it was between the p-value between 0.03 and 0.04, because otherwise the authors would have reported p<0.03.
Then there’s the issue of univariate versus multivariate analysis. Whenever you examine multiple variables for their potential correlations with various outcomes, you’ll run into the problem of multiple comparisons, where, the more variables you examine the greater the chance of seemingly “positive” results by random chance alone. That’s why Student’s t-test is not appropriate for multiple comparisons, for example, and it’s why univariate analysis is fine for an initial exploration of multiple variables but has to be followed by an analysis that takes multiple comparisons into account. And, as you see in the abstract above, there were only two things that came out as statistically significant: talking while charging the device and smoking, both of which were associated with abnormal sperm concentration.
Then there was another finding that is basically meaningless. Basically, among men who reported holding their phones ≤50 cm from the groin, the “rate of abnormal sperm concentration showed a non-significant trend towards a higher value among participants who reported generally keeping their cell phones at a distance ≤50 cm from the groin compared with those who kept it at a distance >50 cm from the groin (47.1% versus 11.1%). No association was found between any of the factors investigated and between semen volume and progressive motility (data not shown).”
In other words, the authors were really, really, really stretching to eke out a seeming correlation relevant to a link between cell phones radiation and decreased sperm count and male factor infertility. Heck, there wasn’t even a correlation between keeping the phone near one’s testicles and…anything examined! Basically, this study is nonsense, and the authors do a serious song and dance to try to make statistically insignificant findings sound important:
The participants in this study, who reported talking on their phones while the device was being charged, were more likely to have abnormal semen concentration. To our knowledge, this aspect of cell phone use has not been previously addressed. During charging of cell phones, two changes occur: (i) the external power source by itself emits energy; and (ii) due to the continuous supply of energy from the external source, the device transmits at a higher power, without the need for energy saving, in contrast to the usual talking mode.
None of which was measured. And:
Participants who constantly carry the device at a distance ≤50 cm from the groin were found to have a higher rate of abnormal sperm concentration. Although the association did not reach statistical significance, it appears that sperm parameters may be affected, even during a stand-by mode (when RF-EMR is emitted from the device for short durations).
Huh? No it doesn’t. Your data didn’t reach statistical significance, and we have no idea what you mean by “approaching” statistical significance! How close was your p-value to 0.05? If you’re serious about arguing that there is an effect but that your sample size was just too small to detect it, you really should give the actual p-value, rather than just “NS” for “not significant.” Such an argument tends to be a lot more convincing if p=0.06 than it is if p=0.25. I’m by no means a p-value Nazi (if I were, I wouldn’t be so much into science-based medicine rather than evidence-based medicine), but I also realize that “approaching statistical significance” usually means, “no matter how much we tortured the data we just couldn’t make the p-value fall below 0.05.”
Basically, if you look critically at the data, one thing that becomes very clear is that the studies purporting to show a link between cell phone radiation and male infertility are uniformly horribly designed and unconvincing. This one is no exception. The Telegraph published a rebuttal article that’s not bad, but still leaves something to be desired. First, it quotes experts correctly pointing out just how bad these studies are. Second, it quotes Allan Pacey, Professor of Andrology at the University of Sheffield:
“What is more of a problem is men – and couples – waiting until they are older before they are trying for a baby and then trying to blame their infertility on things like mobile phones when it’s their combined age which is probably a bigger problem.”
Yes. This is very likely a factor. Unfortunately The Telegraph couldn’t resist going here:
Pacey’s opinion – that mobile phone radiation is being seized upon by infertile men as a convenient scapegoat for their condition – is supported by Stephen Harbottle, a Consultant Embryologist at Cambridge IVF.
“There are a number of environmental factors that may be contributing to the general decline we have seen in male fertility over the last 15 years,” says Harbottle, who believes mobile phones are much further down the list of causes than recent research would have us believe.
“Probably most significant factor is oestrogen-mimicking compounds in the food chain. These can enter the food chain via pesticides and some plastic food wrappings.
You knew it wouldn’t be long before they brought pesticides and BPA into it. There really isn’t any good evidence to support this assertion.
The bottom line is that cell phone usage as measured by a study like the Technion study is almost certainly a confounder, a surrogate for some other factor that is known to be related to infertility that the study doesn’t control for. By reporting a small, crappy study like this as evidence that cell phones are “cooking men’s sperm” journalists exercise the very worst tendencies in science reporting, tendencies that I’ve railed at since the beginning, particularly when it comes to vaccine reporting.
For shame, anyone who credulously swallowed this story and ran with it.
