With the “gateway effect” dead in the water formaldehyde has become the new cause du jour of the anti-vaping nuts. The latest assault has been launched through the letters page of a major US medical journal. Let’s look at the facts.
So the New England Journal of Medicine has joined in the latest crusade against electronic cigarettes. Yesterday they published a letter to the editor about a new study claiming that e-cigs produce more formaldehyde – a carcinogen – than actual burning tobacco-filled cigarettes do. I already blogged about this somewhere else, but there’s a lot more that needs to be said. Basically this study was carried out by researchers who were either utterly clueless or intent on pushing their own agenda at the cost of scientific integrity, and for reasons I’ll explain I strongly suspect it’s the latter.
What the study claims is that electronic cigarettes, when run at high voltages, generate enough formaldehyde to make them between five and fifteen times as much of a cancer risk as real cigarettes. One of the study authors, James Pankow, admitted that this claim could have done with some more context but “we just wanted to get it out.” The extra context of course is that formaldehyde isn’t the only carcinogen – or even one of the significant carcinogens – in cigarette smoke. In fact while it’s classed as a known carcinogen there’s (at most) vague knowledge on what cancers it’s actually associated with and how strongly. Complicating all of this is the fact that formaldehyde is ubiquitous. It’s everywhere; all living things produce it and it’s naturally present in every fruit, vegetable or piece of meat you eat. Every breath you exhale contains formaldehyde from your lungs, whether you vape or not. As context for the study this is all highly relevant, and the fact it was omitted doesn’t look encouraging.
Of course lots of other essential details were omitted too. The letter to the NEJMdidn’t mention what type of e-cig was used in the tests, just that it created huge amounts of formaldehyde at high voltages. Luckily a quick-thinking vaper emailed one of the responsible scientists and asked; much to their amazement they actually got a reply (this doesn’t happen much when dealing with anti-vaping loons):
“The results published in this letter come from samples vaped using a Sailebao CE4 cartomizer with a 2.1 ohm coil and an Innokin iTaste VV V3.0 battery.”
Hmm, okay. Is the Sailebao CE4 clearomiser (it’s not a cartomiser) a suitable device for testing high-power vaping? I would say not. CE4 clearos are usually supplied with starter kits that include an eGo battery, and these batteries output a fixed 3.7 volts. They’re entry level technology, and old entry level technology at that – three years old, at least. This also looks like a particularly cheap example, but the CE4 is a simple enough design. It’s also a top coil design; the heating element is at the top of the tank, just below the mouthpiece, and two long silica wicks hang down into the liquid. This means liquid transport to the coil depends on capillary action, which is not a fast process. If the coil is vaporising liquid too quickly it will outrun the wicks’ ability to keep it supplied and the liquid that remains won’t be evaporated; it will be burned, producing a foul-tasting phenomenon known as a dry hit. That, of course, is why nobody uses CE4s on a high-powered device. When Dr Konstantinos Farsalinos did his own research he tried using a CE4 at a power output of 9 watts, but had to replace it with a bottom-coil Evod because it was consistently dry-hitting. The power output of the 2.1 ohm CE4 used in the new study, when run at 5 volts, is 12 watts.
A CE4 works just fine on a lower voltage battery, of course, and it doesn’t produce any formaldehyde either – when Pankow and his associate David Peyton tested their equipment at 3.3 volts they found no formaldehyde at all. That’s only to be expected; the wicks can keep up at that sort of power output so no liquid would be burned. Unfortunately they either didn’t test at the iTaste VV’s full range of voltages or didn’t release the results; all they’ve revealed is the figures for 3.3V and 5V. It would be nice to know how much formaldehyde was created at 3.7V – i.e. what the clearo is supposed to be used at – but I can guess anyway: None. I suspect that if they’d produced a proper voltage/output curve, like they should have done, we’d see no formaldehyde at all until pretty close to maximum power, at which point the level would rise steeply.
But wait. Are we actually talking about formaldehyde at all? I’d naturally assumed that this, at least, was correct – that formaldehyde had been released in the vapour at high voltages. Then Dr Farsalinos intervened to point out that no, it wasn’t. What Pankow and Peyton were measuring was actually a class of chemicals called formaldehyde hemiacetals. These are a combination of formaldehyde and alcohols; they’re produced when propylene glycol breaks down at high temperatures to produce formaldehyde, which then reacts with more propylene glycol (an alcohol) to form the hemiacetals. Pankow and Peyton classed formaldehyde hemiacetals as “Formaldehyde Releasing Agents” and backed this up with a link to another paper, showing how FRAs can cause health risks.
Well, there’s a serious problem with this. The paper they linked to wasn’t talking about formaldehyde hemiacetals at all; it was discussing several completely different classes of chemicals. In fact there’s no evidence that formaldehyde hemiacetals even release formaldehyde, or that they pose any risk of cancer or any other health problems. To do what this study did, and assume they pose the same cancer risk as free formaldehyde, is utterly dishonest. In fact it’s actually possible that hemiacetals could have a protective effect against formaldehyde!
So this study was using equipment inappropriately, in a way that it never would be used in real life, then misrepresenting the results to make completely inaccurate claims about a danger that does not in fact exist. Why? What possesses scientists to do this? Well, let’s follow the money. The NEJM, like most journals, asks authors to fill out a conflict of interest form. Here it is.
If you look at the list of donors declared by each scientist there are two names at the top that are quite significant; these are Michael J Dowd and Patrick J Coughlin, who contributed “philanthropy to support research”. So who are Dowd and Coughlin? Philanthropists? Well, not exactly. They appear to work for this law firm, and have previously been involved in class action suits against tobacco companies. Lawyers don’t generally pay for research unless they have an ulterior motive, commonly involving another giant class action suit and another few million in fees. So does anyone believe that this pair funded their pet scientists out of a healthy spirit of open-minded inquiry? No, me neither.
This post was originally published on 22 January 2015