The popular press has recently reported claims that harmful levels of lead and other metals can be found in vapor, attributing this to them coming off the coil. This is the result of a recent article out of Johns Hopkins Bloomberg School of Public Health that appeared in Environmental Health Perspectives. (EHP is the go-to journal for publishing alarmist junk papers about chemical exposures, equivalent to the role played by the journals Tobacco Control and Pediatrics for anti-tobacco junk science.)
The authors recruited vapers to bring their open-system hardware and e-liquid to a lab, and then they measured metal concentrations in the liquid and in aerosol captured after machine puffing. They found low concentrations of metals — aluminum, copper, zinc and others — in the unused e-liquid (in its store-bought container), higher concentrations in vapor captured after it was produced by the e-cigarettes, and higher still in the liquid remaining in the tank. However, “higher still” was still far from high enough to pose any worry. This did not stop the authors from claiming, incorrectly, that vapers are exposed to hazardous levels of metals.
The press articles about it emphasized the one metal that lay readers worry most about, lead. This was because the alarmist, inaccurate and inappropriate press release sent out by Johns Hopkins emphasized that metal. Obviously this was designed to sow fear, not to inform.
Every serious analysis of this study has pointed out that the calculation of exposure levels was hugely inaccurate (e.g., this one by Jim McDonald, which also discussed the folly of vapers volunteering to participate in anti-vaping studies like this). The paper’s authors effectively assumed — hiding it in technical calculations, of course — that every breath a vaper takes is a puff on an e-cigarette. They then noted that the total exposure from this fictitious level of exposure (barely) reached certain thresholds for exposure. Obviously this is wrong by a few orders of magnitude. One need only read the methods to see the error, but the authors knew that health reporters do not bother to read the papers they are reporting about.
The concentrations of lead reported were close to the “action level” for lead in drinking water, the concentration is not supposed to be exceeded. But this is based on drinking liters of water per day (and being exposed to aerosols from showers and cooking), not the few milliliters of e-liquid someone would consume. Moreover, those limits are designed to protect young children, who are particularly vulnerable to the brain development effects of lead.
This observation also means that the erroneous claims about total exposure cannot be attributed to an innocent calculation mistake. The authors report the quantity of metals per quantity of liquid as their main results. They could not help but have noticed that the levels were borderline acceptable for drinking water, and thus that total daily e-liquid consumption could not possibly be at dangerous levels. Authors do sometimes make calculation errors innocently (and the journal review process, contrary to popular misconceptions, almost never corrects them), but not when there is such an obvious crosscheck on their conclusions.
But there are also other major problems with this study.
The Daily Vaper asked Drexel University Associate Professor Igor Burstyn — an exposure scientist, environmental and occupational epidemiologist, and author of a landmark paper on vapor chemistry — about the article and how such a study could be done correctly.
Burstyn immediately noticed that the calculation of total daily exposure was wrong and noted that he had demonstrated the correct calculation in his paper. He added the observation, which also appears in his paper but has been largely ignored in both research and criticism, that atoms of a metal is not a useful measure for assessing health effects. Some molecules that include a particular metal are quite harmful while others are harmless. Attempts to alarm people about metal exposures almost always invoke the risk from the more harmful molecules even when there is no evidence those are present. Burstyn also noted that even ignoring the major error in calculating exposure, the threshold levels the authors compared their estimates to (which they barely reached) are not levels that are considered risky. Rather they are conservative thresholds for saying “at or below this level, there is clearly so little risk that we should not even bother to think about it.” In short, the detected exposure levels are far below any cause for concern.
Burstyn then observed that this research was apparently not really designed to assess whether metals came off of the coils, as is implied in the article. If that were the goal, the researchers would have used a purified liquid with no detectable quantities of the metals in it. It is a clear mistake to complicate the assessment by testing whatever e-liquid someone happened to be using.
There are several explanations for the odd result that the liquid in the tank had a higher concentrations of metals than the captured aerosol, which is generated by running that same liquid through the coil, the supposed source of the metal. The coil could be leaking metal, some of which diffuses back up the wick to the tank when the device is not in use, but the aerosol should have higher concentrations. Theoretically metal could have been lost to the gas phase and so not captured in the aerosol, though Burstyn observes that we would predict the opposite: The gas should have little metal in it, and so the captured droplets should have a higher metal concentration. Perhaps the wicking and aerosolization process is somehow effectively distilling the liquid, preferentially taking up liquid with less metal in it, and leaving an increased concentration in the tank. Using a simulated e-liquid with no detectable levels of the metals in it could have helped explain this odd result and distinguish among these hypotheses.
In addition, Burstyn observes that proper research of this kind must include a measure of metal depletion. In this case, that would mean at least weighing the coil before and after extended use to see if the loss of mass could explain the increases in the liquid that were observed. The conclusion that the coils are depositing metal into the liquid is nothing more than an assumption. It would be interesting to know more about sources of metals in e-liquids, even though there does not appear to be any health risk, but we are not going to learn it from studies like this.
The previous science lesson article explained how epidemiology researchers usually just report whatever result their calculations spit out, without ever digging deeper to understand what is really going on. The same is typically true of public health toxicology researchers. Instead of acting like scientists and doing the work to figure out what their results really meant before publishing them, these authors just asserted a story without checking it.
Burstyn concluded that these problems mean there are probably more subtle errors in the study too. He likened the experience of reading it to grading papers by mediocre first-year public health students. But at least his students, if writing about this topic, would presumably have read his paper and not made the simple mistakes. The Johns Hopkins authors clearly never read it.
A subtle effect of junk science like this, if it is taken seriously by regulators and lawmakers, is to benefit major manufacturers of cigalikes and other closed systems at the expense of open systems. It is relatively easy for a closed-system manufacturer to test all their device configurations, using all their liquids along with a clean test liquid, and demonstrate that harmful levels of metals are not present. Open system hardware and liquid manufacturers will not be able to do that if it is demanded. Thus, even apart from the public panic created by the dishonest press release, this approach will play into FDA’s goals of “simplifying” the market down to a few major manufacturers that they can better control.