The opponents of e-cigarettes have determined that ‘ultrafine particles’ or ‘particulates’ are an issue they can work with. But this campaigning gambit, it turns out, involves a crude scientific sleight of hand. We’ll explore this by looking at some examples and the appropriate responses.
Note: Are you looking at this and thinking ‘that looks a bit long’? If so, Carl Phillips has provided: What is wrong with ecig particulate claims – the simple version. The short version is that e-cig aerosols are have completely different physics (they are liquid droplets) and completely different chemistry (they have hardly any of the toxic properties) to the particulates generated by combustion or pollution processes. So why would anyone expect them to have the same impact on health? But I’d be delighted if you press on…so let’s start with…
The University of California at San Francisco – Professor Stanton Glantz
Professor Stanton Glantz, perhaps the world’s top anti-vaping activist, is the leading advocate of the ‘ultrafine particles’ theory of e-cigarette health risk. In a discussion of e-cigarette science, he asserts for example:
Notice the conflation of cigarettes and e-cigarettes in this statement. This, as we shall see, is at the heart of the sleight of hand. Reference 4 and the link (below) are directed to two sources on secondhand tobacco smoke, which have nothing whatever to do with e-cigarettes, even though e-cigarettes is the subject he is actually writing about. Reference 4 is as follows:
 Institute of Medicine. Secondhand Smoke Exposure and Cardiovascular Effects: Making Sense of the Evidence. 2009. Also see Barnoya J, Glantz S. Cardiovascular effects of secondhand smoke: nearly as large as smoking. Circulation 2005 May 24;111(20):2684-98 and Grana RL, Benowitz N, Glantz SA. E-cigarettes: a scientific review. Circulation 2014 May 13;129(19):1972-86. doi: 10.1161/CIRCULATIONAHA.114.007667.
The third and final citation in this reference his own review of e-cigarettes in Circulation is discussed below. Professor Glantz is duly taken down by Dr Farsalinos in his critique of this assertion. In his posting: Abuse of evidence and argument: a response to Stanton Glantz’ criticisms of an expert letter to WHO on tobacco harm reduction, Farsalinos comments:
However, this is a direct mis-presentation of evidence, since there is not even a single study indicating that particles emitted from e-cigarettes represent a risk factor for cardiovascular disease. The review cited (co-authored by the writer of the critique) is again a mis-presentation of science, making an arbitrary conclusion that the evidence linking environmental pollution or cigarette smoke particles with cardiovascular disease can be applied to e-cigarettes. (original emphasis)
The important and obvious point for scientists and non-scientists alike is that it is not just the size of the particles that matters, but what they are made of. That is how reactive and toxic the surface chemistry is and whether they are liquid and likely to be easily absorbed or solid and likely to lodge in lung tissue. There is a big difference between the soot from diesel engines, emissions from power stations, photochemical smog or biomass burning (including the hot reactive organic particles in tobacco smoke) and the relatively benign components of an e-cigarette vapour droplet. In an act of heroic obfuscation, in their review of e-cigarettes published in Circulation as mentioned above, Glantz and colleagues manage to admit that e-cigarette vapour might be totally different to cigarette smoke and other environmental pollution, and by implication, that the conclusions that follow from assuming they are equivalent are worthless.
E-cigarettes deliver nicotine by creating an aerosol of ultrafine particles. Fine particles can be variable and chemically complex, and the specific components responsible for toxicity and the relative importance of particle size and particle composition are generally not known. Given these uncertainties, it is not clear whether the ultrafine particles delivered by e-cigarettes have health effects and toxicity similar to the ambient fine particles generated by conventional cigarette smoke or secondhand smoke. There is strong evidence, however, that frequent low or short-term levels of exposure to fine and ultrafine particles from tobacco smoke or air pollution can contribute to pulmonary and systemic inflammatory processes and increase the risk of cardiovascular and respiratory disease and death. (emphasis added) – Grana, Benowitz, Glantz, Circulation 2014.
Despite the tortuous formulation aimed at sounding is if it is obvious (but not yet proven) that e-cigarettes and cigarette particles would have the same effect, this paragraph actually acknowledges that the chemical composition of the particle is critical. Translation: e-cigarette vapour bears no relation to cigarette smoke or anything else generally regarded as hazardous in particulate form, but we’ll suggest it is like tobacco smoke all the same.
Update: for a more complete discussion of tobacco smoke physics and chemistry, see Chapter 3 of the Surgeon General 2010 report: How Tobacco Smoke Causes Disease: The Biology and Behavioral Basis for Smoking-Attributable Disease. The SG states the obvious – the chemical composition is at the heart of the toxicology:
Cigarette smoke is a complex mixture of chemical compounds that are bound to aerosol particles or are free in the gas phase. Chemical compounds in tobacco can be distilled into smoke or can react to form other constituents that are then distilled to smoke. Researchers have estimated that cigarette smoke has 7,357 chemical compounds from many different classes (Rodgman and Perfetti 2009). In assessing the nature of tobacco smoke, scientists must consider chemical composition, concentrations of components, particle size, and particle charge (Dube and Green 1982). These characteristics vary with the cigarette design and the chemical nature of the product.
The SG goes further and identifies leading culprits in chemicals present in tobacco smoke…
Fowles and Dybing (2003) suggested an approach to identify the chemical components in tobacco smoke with the greatest potential for toxic effects. They considered the risk for cancer, cardiovascular disease, and heart disease. Using this approach, these investigators found that 1,3-butadiene presented by far the most significant cancer risk; acrolein and acetaldehyde had the greatest potential to be respiratory irritants; and cyanide, arsenic, and the cresols were the primary sources of cardiovascular risk. Other chemical classes of concern include other metals, N-nitrosamines, and polycyclic aromatic hydrocarbons (PAHs). This evaluation, along with the Hoffmann list of biologically active chemicals (Hoffmann and Hoffmann 1998), was used to select the chemicals reviewed in this chapter. Other chemical components with potential for harm will be identified as analysis of tobacco smoke becomes more complete and cigarette design and additives change.
So if and when these compounds, or others with equivalent toxic effects, show up in material quantities in vapour products operated under realistic conditions then we can start to be concerned about whether droplet size makes it worse. But there is no point in starting with droplet size before establishing a toxic risk from the composition.
London School of Hygiene and Tropical Medicine – Professor Martin McKee
Professor Martin McKee, an anti-e-cigarette activist and media commentator, has had a go at bringing Professor Glantz’ arguments into Britain, via the New Scientist. McKee asserts:
… there is increasing evidence that the ultrafine particles released in e-cigarettes may have adverse consequences for the cardiovascular system. For these reasons, a wide range of medical and public health bodies, including the World Health Organization, have expressed serious concern about their promotion.
The problem is that the hyperlink to ‘evidence’ in this in this statement actually provides no evidence. It certainly does not show that ultrafine particles released from e-cigarettes have adverse effects on the cardiovascular system. The authors of the cited article were eventually forced to acknowledge this in subsequent exchanges. The study cited merely discovers the obvious fact that the e-cigarette aerosol contains, well, an aerosol, and that this is counted as particulate matter in measurements. The study in question (Schober et al 2014) does not attempt to make the case that the particulates are harmful, simply that they exist. Also note that in Professor McKee’s world, if enough organisations express concern about something it sort of becomes true, but they are all basically repeating the same flawed argument.
Once again, Dr Farsalinos delivers the punch in a letter to the journal criticising Schober et al. In a write up on his must-read blog Farsalinos summarises the exchange.
For the issue of PM, they [Schober et al] acknowledged that the composition is different, and added that “it cannot be deduced from this, that long-term inhalation of 1,2-propanediol nano-droplets must be completely safe.” Well, in a similar way we can support that there is currently no evidence showing that inhalation of 1,2-propanediol (PG) nano-droplets is unsafe.
In other words, they made no assessment of risk of the e-cigarette aerosol, and so cannot conclude whether there is a risk. So they resorted to the scientific scoundrel’s defensive attack of ‘can’t prove completely safe‘, which is a scientific non sequitur. Note also the use of the chemical term 1,2 propanediol instead of the common name ‘propylene glycol’. Why did Schober et al do that, given the likely wide interest in these findings? Here’s a thought: which is more unfamiliar and dangerous sounding? In fact PG is ‘Generally Accepted as Safe’ for use in foods by the FDA, widely used in cosmetics and in pharmaceuticals as an excipient (see Dow safety information). Schober et al was also subject to a robust critique by Carl Phillips and Igor Burstyn, the latter an expert in occupational hygiene and exposure assessment. That is well worth reading in full, but here is one of the more damning statements from it:
The device the authors used to detect “particles” does not distinguish between droplets and solid particles; to assess any health-relevant particles the authors should have used gravimetric techniques that determine the mass of solid particles emitted into the air. As such, the authors’ work suffers from inadequate testing of their major conclusion and confirmation bias: they assumed health-relevant particles would be present in the aerosol, performed a test that was incapable of ruling that out, and then interpreted their results as confirmation.
129 “public health and medical authorities”
In a letter to WHO Director General Margaret Chan, 129 public health academics and activists signed up to the following statement drafted by Professor Glantz:
There is already good evidence that ENDS emissions release several toxic substances into the environment that cause harm to health. These substances include ultrafine particles […]
Again references to Schober et al are used to justify this statement or they simply note the presence of ‘particulates’ as if that is sufficient to justify the claim that these are harmful to health. In response, scientists who work in this field replied with a strongly critical letter to Dr Chan, calling for “dispassionate interpretation and presentation of evidence”. The fascination with particle size, rather than chemical composition, was briskly dismissed:
Reference is also made to ‘ultrafine particles’ as though it is particle size that governs risk to health. In fact the lowest size particles are water vapor, which carries no health risk. What matters most is the chemical composition of the particulate matter – and this differs completely between tobacco smoke and e-cigarette vapor.
World Health Organisation – Tobacco Free Initiative
Regrettably WHO out-sourced its science to UCSF and has taken its scientific cue from Professor Glantz. It has therefore ended up promoting the ‘particulates theory’ at a global scale, while attempting to draw on WHO’s standing in air quality science in an effort to legitimise its tobacco policy. In its September 2014 paper on ‘Electronic Nicotine Delivery Systems (ENDS)’ for COP-6 WHO suggested that evidence from environmental air quality could be applied to e-cigarette aerosol. The sleight of hand in applying science and evidence that has been developed for one thing (atmospheric air pollution) to something completely different (e-cigarette vapour) is not acknowledged, and its unclear if the authors were even aware that they were doing that. Here is the relevant passage:
15(b). However, epidemiological evidence from environmental studies shows adverse effects of particulate matter from any source following both short-term and long-term exposures. The low end of the range of concentrations at which adverse health effects has been demonstrated is not greatly above the background concentration, which for particles smaller than 2.5 μm has been estimated to be 3–5 μg/m3 and increases with dose, which means that there is no threshold for harm and that public health measures should aim at achieving the lowest concentrations possible .
The source  given for this statement is WHO air quality guidelines for particulate matter, ozone, nitrogen dioxide and sulfur dioxide: summary of risk assessment. Geneva: World Health Organization; 2006.
However , these guidelines do not in fact apply to any source and this is incorrectly asserted in the WHO ENDS paper. The guidelines are, as any reasonable person would expect, entirely focussed on the type of particulate matter that one might expect to form hazardous air pollutants – diesel exhaust, power stations emissions, biomass combustion and fine particle of dust or degrading road surfaces etc. They give no basis whatsoever to claim that the these conclusions can be generalised to e-cigarette aerosol. WHO’s ENDS paper (not its air quality guidelines) has simply assumed that the evidence from the particles routinely studied for air pollution science can be generalised to e-cigarettes, but it cannot. The WHO air quality guidelines confirm their focus:
[PM10] is primarily produced by mechanical processes such as construction activities, road dust re-suspension and wind, whereas [PM2.5] originates primarily from combustion sources. In most urban environments, both coarse and fine mode particles are present, but the proportion of particles in these two size ranges is likely to vary substantially between cities around the world, depending on local geography, meteorology and specific PM sources. In some areas, the combustion of wood and other biomass fuels can be an important source of particulate air pollution, the resulting combustion particles being largely in the fine (PM2.5) mode.
[Counterfactual note: PM10 means particles less than 1o microns – one hundredth of a millimetre – in diameter. PM2.5 means smaller than 2.5 microns – the particle size affects penetration and pathways into the body]
Unsurprisingly, there is no reference to any epidemiological evidence that relates to e-cigarettes aerosol or to anything remotely similar chemically or physically. Even the epidemiology for different types of products of combustion is limited. However, WHO’s air quality experts are content to conclude that all of these sources of products of combustion can be treated in a similar manner. This is a reasonable judgement in my view, but not one that can be reasonably extended to e-cigarette aerosol particulates, given that are quite different in chemical and physical composition to solid phase products of combustion. To the extent that these guidelines say it is not possible to isolate the impacts of specific pollutants, this is the key paragraph:
Although few epidemiological studies have compared the relative toxicity of the products of fossil fuel and biomass combustion, similar effect estimates are found for a wide range of cities in both developed and developing countries. It is, therefore, reasonable to assume that the health effects of PM2.5 from both of these sources are broadly the same. By the same token, the WHO AQG [air quality guideline] for PM can also be applied to the indoor environment, specifically in the developing world, where large populations are exposed to high levels of combustion particles derived from indoor stoves and fires.
This is all focussed on products of combustion. I respectfully submit that it is an unreasonable leap to conclude that this discussion embraces e-cigarette vapour, which is not a product of combustion at all… and therefore WHO mislead the Conference of the Parties of the FCTC with its support for the ‘particulates theory’.
E-cigarette Summit presentation by particulates specialists – WHO air quality science
The subject of particulate matter from e-cigarettes was raised at the E-cigarette Summit in London (13 November 2014) in the form of a presentation (Chamber study on second hand e-cigarette “smoke” – preliminary results) by two scientists, Professor Gordon McFiggans and Professor Roy Harrison OBE. The presentation started with an account of the health impacts of particulates – showing a very large annual premature death toll of 200,000 for the UK (more than for smoking) and 340,000 years of lost life. This is certainly a troubling health burden (though the figures are strongly contested)…. but is it in any way applicable to e-cigarettes? Nothing in the report suggests it is.
The presentation mentioned that there are four broad categories of particulates:
- Primary mechanical: airborne particles like dust from deserts, degrading road surfaces, pollen etc
- Primary combustion: products of combustion like diesel smoke, power station emissions, and biomass burning – this category would include tobacco smoke
- Secondary: pollutants that form through atmospheric reactions like photochemical smog
- Volatile primary / secondary – things like clouds, spray perfumes and air fresheners – this category would include e-cigarette vapour
They then went on to describe measurements of e-cigarette aerosol in a chamber study, showing that e-cigarettes produced measurable amounts of this 4th class of particulates. However, they did not show that the 4th class of particulate are actually harmful, or that e-cigarettes might be expected to be harmful and contribute to the death toll. And that is rather important… so I put the obvious question, in so many words: isn’t it composition that matters most…? Cigarette smoke particles are solid and lodge in the lungs and have highly reactive surface chemistry with potentially thousands of products of combustion, many of which will be carcinogenic and reactive. This is quite different to vapour aerosol. As Carl Phillips and Igor Burstyn put it in their critique of Schober et al
While droplets are particulates in the broadest sense of the term, in the context of environmental pollution that term generally refers to fine solid particles that can lodge in or be absorbed through the lungs intact. A liquid, of course, just dilutes into the bloodstream or other bodily liquids, regardless of particle size and deposition location. Thus, the extensive discussion of particulate size, let alone the explicit claims about health implications, is highly misleading.
Professor McFiggans was generous enough to give a confident and honest answer, which can be summarised (I paraphrase): of course the composition and toxicity matters. Professor Harrison, however, was undeterred and in his reply referred the audience to the latest “REVIHAAP report” of WHO. He argued that it showed that the epidemiology did not allow us to distinguish between different types of particulate. I thought this was an odd formulation at the time: having insufficient evidence to distinguish between types of air pollutant is not the same as having evidence that they should be treated the same. It turns out the report itself is rather more informative than this.
I decided to look the report up: and discovered this is the Review of evidence on health aspects of air pollution – REVIHAAP project: final technical report. Professor Harrison was an expert reviewer. The key section for our purposes is A2 page 10-31 (PDF extract). A review of this shows there is absolutely no basis for assuming equivalence between e-cigarette aerosol and the environmental particulates in general. A few observations from the report now follow.
Professor Harrison was, I think, basing his response on this conclusion from the report:
The integrated science assessment used evidence from both epidemiological and experimental studies to conclude that “there are many components contributing to the health effects of PM2.5, but not sufficient evidence to differentiate those constituents (or sources) that are more closely related to specific health outcomes” (EPA, 2009). Despite the increased number of studies (especially epidemiological) after 2009, the general conclusion remains the same. (page 12)
It should be obvious, as I hinted above, that this statement does not justify treating e-cigarette aerosol as though it is as harmful as diesel smoke, degrading road surfaces, power station emissions, smog etc. It simply means they don’t have the science to tell the impacts of the different particulate types apart, not that there is evidence that they are uniformly harmful. In fact there is much in the report to reassure e-cigarette users and to justify calling on the various opponents of e-cigarettes to withdraw their claims about e-cigarette ‘particulates’.
In a statement of the obvious, the first section acknowledges that chemical composition is critically important:
Epidemiological and toxicological studies have shown PM mass (PM2.5 and PM10) comprises fractions with varying types and degrees of health effects, suggesting a role for both the chemical composition (such as transition metals and combustion-derived primary and secondary organic particles) and physical properties (size, particle number and surface area) (page 10 – emphasis added)
This makes perfect sense to me: you would be concerned both about a toxic agent and the pathways through which it could expose a human, which may depend on its physical characteristics. But the agent has to be toxic in some way or the manner of exposure isn’t very important. Professor Robert West spoke from the floor and emphasised the point that you would expect there to be something toxic or carcinogenic about the particles before the size starts to become a concern. It is this insight that leads to the reductio ad absurdum case of water droplets (which also count as particulates) to challenge the notion that all particulates are harmful.
The report goes on to discuss particulate source types for which there is evidence of harmful effects (page 11):
B. Source types
A variety of air pollution sources have been associated with different types of health effects. Most of the evidence accumulated so far is for an adverse effect on health of carbonaceous material from traffic (see also Question C1). A more limited number of studies suggest that traffic-generated dust, including road, brake and tyre wear, also contribute to the adverse effects on health.
1. Coal combustion results in sulfate-contaminated particles, for which epidemiological studies show strong evidence of adverse effects on health.
2. Sources of PM emission relevant to health also include shipping (oil combustion) power generation (oil and coal combustion) and the metal industry (such as nickel).
3. Exposure to particles from biomass combustion – most notably residential wood combustion – may be associated not only with respiratory, but also with cardiovascular health.
4. Desert dust episodes have been linked with cardiovascular hospital admissions and mortality in a number of recent epidemiological studies.
E-cigarette aerosol is of course not mentioned, but nor is anything remotely resembling it in chemical or physical form. Note that it is implicit in this section that not all source types are assumed to have health effects simply by virtue of being particulates – why would they try to differentiate between source types otherwise? Most of the discussion in the paper of course relates to the actual particulates that are recognised air pollutants (from traffic etc), and it is these whose effect cannot be isolated from each other (page 12):
The particle constituents most often included in the studies have been sulfate and black carbon […]
In epidemiological studies, the effects of combustion-derived organic carbon are difficult to separate from those of black carbon and/or elemental carbon because of a high correlation due to the common source: combustion processes (WHO Regional Office for Europe, 2012). Elemental carbon is most strongly associated with primary combustion particles and primary organic carbon, whereas secondary organic aerosol formation is delayed with respect the primary emissions, because secondary organic carbon is formed during longer range transport in the atmosphere. Secondary organic carbon also has a significant biological component, but this part of PM has hardly been studied in relation to health effects.
So the discussion here is about evidence that would allow a distinction between several categories that are not relevant to e-cigarette aerosol. The report goes on to express a fair amount of concern about ‘ultrafine particle’ (less than 1/1oth micron = 100nm), but once again this refers to the primary source of these, namely products of combustion (page 17):
There is a general consensus that ultrafine particles are defined as particles smaller than 100 nm in mobility diameter and mostly stem from combustion processes in urban settings (Peters, Rückerl & Cyrys, 2011)
So again we are really talking about products of combustion in traffic, power stations, cooking and heating, and biomass burning. It is simply not legitimate to take research findings about the ultrafine soot and smoke particles and cut and paste those to e-cigarette aerosol, which has completely different chemistry and physical form.
The report does discuss the state of knowledge on particulate source types, reviewing epidemiology, clinical studies and toxicology for the sources under the main headings below (page 21-29):
(b) The role of source types
- Coal and oil combustion
- Biomass combustion
- Desert dust
- Ocean and sea
- Hazardous waste sites
No sign of anything here that looks remotely like e-cigarette aerosol. The only one that comes close is oceans and sea. There has been concern about sea salt, though of course there is no sea salt in e-cigarette vapour. However, in the report, the sea salt case establishes a useful precedent for our puposes. These are the conclusions for sea salt (page 29 – please read the document for the full text):
Epidemiological studies. […] All in all, there is little epidemiological evidence of the harmfulness of sea salt.
Clinical studies: […] The study provided clear evidence that PM dominated by sea salt and/or sea spray is far less toxic than equal amounts of combustion-derived PM
Toxicological studies: In no study published since 2005 has the role of sea spray and/or sea salt been investigated, although sea salt is not classified as a hazardous compound and it is plausible that at current exposure levels no harmful effects will occur.
Why dwell on sea salt? Well it shows that not everything that is classed as a particulate is classed as inherently harmful simply by virtue of being a particulate. It shows that other evidence can and should be gathered and that what the particulate is composed of actually matters. And it is not as if we know nothing of e-cigarette vapour – it has been fairly extensively studied, and I make no apology for again citing the 2014 Burstyn review: Peering through the mist: systematic review of what the chemistry of contaminants in electronic cigarettes tells us about health risks:
Current state of knowledge about chemistry of liquids and aerosols associated with electronic cigarettes indicates that there is no evidence that vaping produces inhalable exposures to contaminants of the aerosol that would warrant health concerns by the standards that are used to ensure safety of workplaces. However, the aerosol generated during vaping as a whole (contaminants plus declared ingredients) creates personal exposures that would justify surveillance of health among exposed persons in conjunction with investigation of means to keep any adverse health effects as low as reasonably achievable. Exposures of bystanders are likely to be orders of magnitude less, and thus pose no apparent concern.
This advice is sound: it sees no reason for alarm and no reason for concern about bystanders, but recommends the cautious approach of surveillance to detect adverse health effects in users should they arise.
Is there nothing to worry about?
The advocates of this ‘particulates theory’ should stop simply assuming that e-cigarette vapour, though chemically and physically completely different to air pollution particulates, can be treated as though it poses similar risks to health to diesel exhaust or cigarette smoke. Why not start instead from the insight that e-cigarette aerosol consists mainly of propylene glycol, which is widely used in food and as an excipient in pharmaceutical products, combined with nicotine and flavours?
There are of course risks worth exploring – though these would be dominated by exposure to the user rather than to bystanders. It is possible that flavouring agents could be respiratory sensitisers or have other impacts. It is possible that thermal breakdown products may pose a risk and can be avoided with good design or experience of use. It is possible to manage the levels of contaminant in e-liquids. This is a respectable agenda for regulation and regulatory science – though it should always be conducted with relevant context. That means (a) assessment of risks relative to smoking, and; (b) comparison with acceptable occupational exposure, to assist with benchmarking relative risk. We do not need a research industry making mountains out of molehills and scaring people into continued smoking.
Update: a couple of useful papers.
- Schulte P.A etc al (2010) Occupational exposure limits for nanomaterials: state of the art, J Nanopart Res (2010) 12:1971–1987
- Health Council of the Netherlands (2012) Working with nanoparticles: Exposure registry and health monitoring.
Both suggest that the concerns with ultrafine particle should be carefully differentiated according the physical and chemical characteristics of the particle, not just the size. They state the obvious in other words.
The concern about particulates per se has been manufactured by scientific sleight of hand, through heroic analogy with air pollutants with completely different chemistry and physical form. By frightening people and exaggerating or inventing risks with e-cigarettes, the advocates of this theory encourage continued smoking, make unethical interventions adult informed choice, mislead regulators, protect the cigarette-based business model of the tobacco industry, and prolong the smoking epidemic.
If you work in public health, those are quite bad things to be doing: please cease and desist.