Fear-mongering – unhelpful in the fight against cancer

Because it is hard to assess the cancer risk of, say, e-cigarettes or fracking, we are often content to trust what the media, lobby groups or scientific papers say. But while we may not be masters of sophisticated risk assessment tools, a little common-sense understanding goes a long way to debunking many scary claims.

Originally published in Daily Maverick on 27 January 2015

According to new research, the cancer risk of using e-cigarettes is ‘15 times greater than smoking cigarettes’. Oh boy, did I ever get that wrong. Also, fracking ‘could be sowing the seeds for a cancer epidemic in the Karoo’. Dear me. If true, I’d look a right fool.

What are we to do with all these alarming headlines? I get them sent to me routinely, by people who, one assumes, are trying to prove that something I wrote turned out to be terribly wrong.

Usually, when I appear to be wrong, reality has erred. Still, it is possible for the science on a subject to change. New information comes to light. New research is conducted. That’s what science does: it makes discoveries, it explains the unexplained, and it corrects its own past errors.

CancercellDid it do so in these cases? Or did nothing change at all, and was the alarmism of the headlines merely unthinking media sensationalism, or well-calculated activist repetition?

I wrote recently about e-cigarettes, and noted that cigarette smoke is likely an order of magnitude, and possibly up to 450 times, more toxic than the vapour of e-cigarettes. I based this claim on a study of a dozen e-cigarettes published in the peer-reviewed BMJ Tobacco Control journal.

My pen was hardly cool when Agence France-Presse distributed a tract that got published under the alarming headline that e-cigarette cancer risk is ’15 times greater than smoking cigarettes’. (Note the blame marks around that phrase. Headline writers use it to disclaim responsibility for a statement that may be unsupportable.)

For a start, let’s go to a less hysterical source. Here is the 550-word letter to the editor in the New England Journal of Medicine that led to this startling headline. According to its authors, if you use a 5V vaporiser in your e-cigarette, you risk inhaling high levels of formaldehyde compounds. At 3.3V, by contrast, none of this stuff is produced.

The study’s lead author told his local newspaper about these formaldehyde compounds: “We don’t know the health effects of this. It could be benign.”

Yet, they assume it is bad for you, because formaldehyde itself is a carcinogen, as are so many substances, natural and synthetic. But let’s grant them that, and stipulate that the hazard is significant.

To assess risk, we need to figure out how likely it is that this hazard actually materialises. It’s hard to tell, because the good scientists who penned the brief missive were not good enough to disclose the resistance of the vaporiser coils they used, or the wattage it drew. Without at least one of these numbers, it is impossible to determine whether 5V was a reasonable voltage to use in their e-cigarette test.

We can, however, make an educated guess, since a standard commercial device comes with a coil in the 2.4Ω to 2.8Ω resistance range, and a 3.3V to 3.7V battery. E-cigarette forums dedicated to building or modifying vaporisers are great sources for engineering knowledge about these machines. This is where one can find tables that list battery voltage against coil resistance, to produce a good vapour without overheating the coil. One discussion compares variable voltage with variable wattage. Another covers standard coil resistance and battery voltage, and the effect that these might have on the vapour.

These pages show (and anyone with high-school physics should know) that voltage is not the sole determinant of how hot the coil gets. There’s nothing wrong with a 5V battery, provided that you use a coil with at least 3Ω, and ideally 4Ω, of resistance. The latest new craze is sub-ohm vaping, in which the devices use batteries nowhere near 5V, but present a far greater risk of overheating.

These pages also show that if the letter authors used a ‘commercially available’ system with a coil in the 2.4Ω to 2.8Ω range they were operating it well outside recommended limits. They might argue that the device allowed them to set the voltage that high, but that doesn’t wash. It also allows you to change the coil, and sellers of variable voltage rigs don’t recommend them to first-time buyers.

Still, any idiot could buy one. Would your average vaper know to avoid the high setting? The egg-heads had a machine drag on the e-cigarette. If they’d bothered to taste the stuff, or ask some vapers to do so for them, they’d have known what happens when the coil gets too hot. The vapour becomes acrid and metallic. If you’re human, you’d stop sucking, because you’re not a lab robot. If the taste does not get better after a cool-down period, you curse and replace the coil.

What happens if you’re a complete dunce, and burn your coil every day all day? It turns out that it’s pretty much the same as overheating your frying oil, or burning your food: you might suffer some ‘incremental lifetime cancer risk’. That term means you’ll be slightly more likely to die of cancer if you continue to be exposed to the stuff for the rest of your life.

The models used to calculate the incremental lifetime cancer risk are usually developed at high doses, because low-dose responses are hard to measure. This produces a slope relating human body weight to exposure per day. Whether the relationship continues to hold at low doses is determined by whether or not the offending molecule is ‘genotoxic’. That is, is it merely toxic, or does it cause damage to DNA which could cause uncontrolled cell division? If the latter, it is supposed that any exposure at all could trigger cancer.

The actual incremental lifetime cancer risk from breathing formaldehyde for the rest of your life is given as 4.2×10-3, or 0.0042. That is almost five times as high as the risk they cite for cigarettes.

But that risk only materialises if you are exposed in the first place (which you won’t be very often, because overheated vapour taste terrible), and you continue that daily exposure consistently for the rest of your life (which you won’t do, because you’d prefer better-tasting vapour).

Put this all together, and the claim that e-cigarettes pose a cancer risk ‘15 times greater than smoking cigarettes’ is outright fear-mongering. And frankly, it shouldn’t take a journalist to explain that to a professor, four PhDs and the editor of the New England Journal of Medicine.

Note the process we followed here. We characterised the hazard: an overheating vaporiser coil produces formaldehyde compounds that likely release formaldehyde which can cause cancer in humans.

Then we established two independent facts, usually described as the exposure assessment and the toxicity assessment. (In other risk settings, different criteria might apply, but the principle is the same.)

The first determines how a person might become exposed to the hazard. If there is no pathway for people to become exposed to a hazardous substance, then the substance poses no risk. If there is a pathway, we need to know how likely it is that exposure will occur.

The second works out the so-called ‘dose-response’ of the human body to the toxin in question. The human body can tolerate small doses of all but the most lethal toxins. Possible exceptions are ‘genotoxic’ substances that, like ionising radiation, can damage DNA. I say possible, because we are exposed to small doses of genotoxic substances naturally, just like we absorb a modest amount of ionising radiation just from our surroundings.

Finally, we estimate the risk, which is the magnitude of the hazard times the probability that it might actually happen. In this case, even if the hazard is assumed to be significant, the risk is negligible.

If the risk were not negligible, we’d have to go a few steps further, in order to manage the risk, and communicate it to those affected. A constructive response would mitigate the risk without compromising the perceived or actual benefits of that which caused the hazard in the first place.

Either way, fear-mongering headlines are not justified. Given that the authors of the letter might have anticipated their worst-case scenario (‘15 times’) to be used by the media, it seems reasonable to suspect that their intent was lobbying or self-promotion.

A similar case of mis-assessing risk can be found in a recent shale gas story.

The government will likely soon issue shale gas exploration permits (though it would have been smarter to do so while the oil price was still over $100). In anticipation, the Cancer Association of South Africa (CANSA) got behind an effort by the Treasure the Karoo Action Group (TKAG) to conduct baseline studies of water resources in the Karoo.

In principle, that’s a great idea. I’ve proposed such studies myself. If an industrial or mining process might pose risks, it would be good to be able to monitor it, and to prove (or disprove) later pollution claims. The more we know, the less there is to fear.

In contrast to CANSA’s fairly measured statement, however, Carl Albrecht, its acting head of research, told the media that shale gas drilling was “sowing the seeds of a cancer epidemic in the Karoo”.

That language is unscientific, inflammatory, manipulative and untrue. It is designed only to whip up public fear, rather than to mitigate any risks to public health. It relies on anecdotal evidence instead of systematic, long-term studies of pollution related to shale gas drilling.

The article reports that CANSA has taken a stand in opposition to shale gas drilling, alongside the TKAG. This kind of lobbying appears to me far outside CANSA’s remit. A reputable group like CANSA ought to do better than react in knee-jerk fashion to the mere naming of a chemical. It should be aware of how actual risk is assessed.

In the newspaper story, Albrecht frets over formaldehyde and benzene, without even an acknowledgement to how sparsely populated the Karoo is. Elevated levels of these chemicals have never been found more than a few hundred metres from active drilling sites, nor have they persisted long-term.

While ad-hoc measurements have occasionally raised cause for concern, numerous more systematic studies and regulatory reports have found nothing alarming. Even a hardcore anti-fracking activist group concedes that it could find no causal connection between its air measurements and shale gas operations, nor could it find any chemical concentrations exceeding legal standards.

More importantly, gas, when used to generate electricity, produces less than half the carbon dioxide of coal, 80% less carbon monoxide and nitrogen oxides, and virtually eliminates smog-causing sulphur dioxide and particulates. Minor and transient pollution at drilling sites should be viewed in this context.

Even if you’re working on the drill site, the chance of exposure to benzene or formaldehyde in high enough dose and duration to cause cancer is low. If you’re anywhere else, your risk is, to a rounding error, nil.

Carcinogens are all around us, both naturally and otherwise. In most cases, our exposure to such hazards is very limited, or the dose is low. Either way, the consequent risk is also low. That’s why most carcinogens never go on to cause cancer.

In the unlucky event that they do, scientists are producing great advancements in treatment and preventative care. In developed economies, cancer rates have plummeted, and a recent paper argues that by 2050, most cancers in those under 80 will have become preventable. (Cue more media sensationalism.)

Scientific progress sometimes poses threats to human health. More often, however, they increase prosperity and contribute to winning the battle. Fear-mongering certainly doesn’t help the fight against cancer.


  1. About the Formaldehype.

    I agree. It’s just a smoke screen to scare people. While I don’t doubt the raw data of the study, the conclusions drawn and the headlines generated are just bovine excrement hitting the media fans.

    This wasn’t the first wave of scaremongering with Formaldehyde, only the most excessive. So far. And there are even more variables than just the total power (voltage, current, resistance). It also depends on the design of the liquid flow to the coil, how much an atomizer can handle. So, even subohm vaping with 30W or more is not a problem, as long as the atomizer and the battery are designed to handle it.

    My own “review” of the first study Freaking Formaldehyde also contains links concerning the other waves.

  2. Pingback: Freaking Formaldehyde | Norbert Zillatron

Comments are closed