Bill King, Coral Gartner and Ron Borland
In April 2019, the first case appeared of what soon became an epidemic of highly unusual lung damage cases in the United States, with the patients first presenting for medical treatment with a form of lipoid pneumonia. The US Centers for Disease Control and Prevention (CDC), the Food and Drug Administration (FDA) and numerous state departments of health began investigating in July. By the end of the second week of September, the epidemic numbered 380 probable cases and there had been six deaths, mostly among young people. Infection was soon ruled out as the cause and an unknown chemical agent or agents was deemed the probable cause. It quickly became apparent that the cases were all strongly linked to use of electronic ‘vaping’ devices.
Most cases have so far been confined to the United States (with reports emerging of at least one case in Canada). Given that vaping is widespread internationally and has been popular for around a decade, this appears to be a new hazard associated with vaping. The investigations are ongoing and have not decisively identified the cause, beyond being a chemical agent delivered via ‘vaping’. However, exactly what constitutes vaping is not straightforward, and the technological, human and societal elements of vaping are constantly evolving. Electronic vaporisers are used both as an alternative to smoking tobacco, as well as for recreational purposes. More recently, these devices have been used or modified to enable vaping of other liquids, including cannabis-derived products, such as tetrahydrocannabinol (THC). In the case of cannabis products, the carrier liquid is typically an oil. This is likely to be more problematic than the carrier liquids used to vape nicotine, which are water soluble. Vaping cannabis oils, known as dabbing, is an increasing phenomenon that may be driven by the legalisation of cannabis in many jurisdictions.
Currently, the key focus of the outbreak investigation conducted by the New York State Department of Health is unlicensed (or ‘bootleg’) ‘dabs’ containing vitamin E acetate, which is commonly used to dilute the THC while thickening the liquid to hide the dilution. Among lung damage cases residing in New York State who submitted vape products they had recently used for testing, all submitted at least one cannabis product containing vitamin E acetate. Other states have also identified clusters of cases in which all reported purchasing cannabis vape cartridges from black market vendors. The picture emerging of this outbreak is similar to recent clusters of otherwise healthy people suddenly becoming seriously ill or dying as a result of methanol poisoning associated with consumption of tainted alcoholic beverages in Indonesia, Costa Rica and Australia.
Insofar as vitamin E acetate, and potentially other similar contaminants, is accepted as the primary agent of concern for having caused the vaping-lung damage cases, one might expect that public health advice would stress the exceptionally high risk of vaping bootleg cannabis products, and that consumers should cease using these. That was the approach taken by the FDA in a consumer advisory issued on September 7th:
“Because consumers cannot be sure whether any THC vaping products may contain vitamin E acetate, consumers are urged to avoid buying any vaping products on the street, and to refrain from using THC oil or modifying/ adding any substances to products purchased in stores.”
However, the CDC has been giving the public very different advice. In a public statement on its website on September 16th, the CDC advised concerned members of the public:
“While this investigation is ongoing, if individuals are concerned about these specific health risks, they should consider refraining from using e-cigarette or vaping products.”
In contrast, CDC statements to the media carried fewer qualifications:
“While this investigation is ongoing, people should not use e-cigarette products,” Dana Meaney-Delman of the CDC said in a call Friday. That broad recommendation is because “there is a diversity of products” related to e-cigarettes, some containing THC, or tetrahydrocannabinol, marijuana’s primary psychoactive component, and some containing nicotine, she said.
Layden et al summarised CDC’s rationale for advising abstinence form ALL forms of vaping, including vaping for nicotine. They wrote:
“E-cigarette liquids and aerosols have been shown to contain a variety of chemical constituents that may have adverse health effects. Major declared constituents in nicotine-based e-cigarettes include propylene glycol and glycerine, in addition to nicotine. Identified contaminants include polycyclic aromatic hydrocarbons, nitrosamines, volatile organic chemicals, and inorganic chemicals such as toxic metals. Endo-toxins and flavouring compounds such as diacetyl and 2,3-pentanedione [NB: two naturally occurring constituents of butter that also naturally occur in tobacco and have been used as additives in both cigarettes and e-liquids] have also been discovered.”
Is an urgent health warning that extends to all vaping products including commercially produced nicotine products justified? No-one has argued credibly for any serious acute health risk from vaping as a result of any of the chemicals listed above, which occur in vapour at much lower levels than the levels in cigarette smoke. Collectively, the presence of these chemicals provides credible evidence that long term vaping will carry a health risk, but it does not seem plausible that exposures at lower levels than tobacco smoke are responsible for the serious acute respiratory syndrome seen in the current outbreak.
The public health priority, as the US FDA correctly understood, should be to discourage the use of bootleg products, especially those related to vaping cannabis products. In our view, the aim of preventing further lung damage cases is much more likely to be achieved with a targeted approach that excludes improbable causes, than with the blanket ‘don’t vape’ approach of CDC. Including people who are unlikely to be at acute risk is both morally questionable and likely to undermine the credibility that is critical for agencies like CDC to be most effective.
While CDC’s overreaction might be understandable, as all the cases have occurred within its jurisdiction, the response of authorities in other countries is harder to understand. For example, as far as we know, there have been no reported cases of lipoid pneumonia linked to vaping in Australia. However, the Australian Chief Medical Officers issued a joint statement about the outbreak on September 13th that contained a general warning against vaping in all forms, stating that “Anyone using e-cigarette products or who is exposed to e-cigarette emissions and/or e liquids is potentially at risk”.
The logic of starting warnings broad and narrowing them down later is flawless from a certain standpoint: it means that the early warnings are bound to encompass a more specific cause of harm that emerges later. However, it also creates the risk of over-reach and reputational damage to trusted institutions. The earlier warnings might later be interpreted by some as false alarms, bringing into question the authority of the agencies issuing them. For example, advising everyone to immediately stop drinking alcoholic beverages because of methanol deaths would be considered inappropriate and likely to be ineffective, although it would arguably be effective at preventing methanol poisoning if everyone followed this advice.
We end by noting a specific danger about premature advice for ex-smokers to cease vaping. While not smoking or vaping is the safest option, many smokers who have switched to vaping to reduce their health risk have done so because they found abstinence from smoking difficult to achieve and maintain. If they heed warnings to stop vaping, those with strong nicotine dependence may rapidly relapse to smoking. That is guaranteed to cause harm.
Bill King is a Research Fellow at the Behaviour Change Unit, School of Psychological Sciences at the University of Melbourne. Coral Gartner is an Associate Professor of Public Health at the School of Public Health, University of Queensland. Ron Borland is Professor of Psychology – Health Behaviour at the School of Psychological Sciences, University of Melbourne.