On 2016 Aug 05, Zvi Herzig commented:

Adding to Bates' comments below, here are some specific misrepresentations of the primary evidence by the Pisinger and Døssing review, relating to almost every major issue in e-cigarette toxicology (not an exhaustive list):

• “Some studies found high maximum concentrations of total TSNA”, citing studies showing TSNAs 100-200 times below cigarette smoke levels [1-3].
• “Exposure to formaldehyde was comparable with smoking", referring to a study calculating formaldehyde levels nine times below that from tobacco smoke [1].
• “Propylene glycol has been found to exacerbate and/or induce multiple allergic symptoms in children”, citing a study stating that "apparently… outcomes were not driven by propylene glycol” [4].
• “Values below the threshold limit don't necessarily protect against the health effect of 200–300 daily inhalations over decades”, referring to safety limits calculated for 8 hours exposures “day after day, over a working lifetime” [5].
• “These metals appear on the U.S. Food and Drug Administration's 'Harmful and Potentially Harmful Chemicals' list”, referring to metals detected below levels acceptable to the FDA for chronic inhalation [6-8].

[1] Goniewicz ML, Knysak J, Gawron M et al. Levels of selected carcinogens and toxicants in vapour from electronic cigarettes. Tob Control 2014

[2] Kim HJ, Shin HS. Determination of tobacco-specific nitrosamines in replacement liquids of electronic cigarettes by liquid chromatography-tandem mass spectrometry. J Chromatogr A 2013

[3] Farsalinos KE, Romagna G, Voudris V. Authors miss the opportunity to discuss important public health implications. J Chromatogr A 2013

[4] Choi H, Schmidbauer N, Spengler J et al. Sources of propylene glycol and glycol ethers in air at home. Int J Environ Res Public Health 2010

[5] ACGIH. Chemical Substances Introduction. acgih.org 2016. http://www.acgih.org/tlv-bei-guidelines/tlv-chemical-substances-introduction

[6] Siegel M. Metals in Electronic Cigarette Vapor are Below USP Standards for Metals in Inhalation Medications. The Rest of the Story: Tobacco News Analysis and Commentary. 2013.http://tobaccoanalysis.blogspot.com/2013/04/metals-in-electronic-cigarette-vapor.html .

[7] Farsalinos K. Metals and nanoparticles in e-cigarettes. 2013.http://www.ecigarette-research.com/web/index.php/2013-04-07-09-50-07/2013/89-metals-and-nanoparticles-i n-e-cigarettes-commentary.

[8] Farsalinos K, Voudris V, Poulas K. Are Metals Emitted from Electronic Cigarettes a Reason for Health Concern? A Risk-Assessment Analysis of Currently Available Literature. International Journal of Environmental Research and Public Health 2015

On 2016 Aug 05, Zvi Herzig commented:

None

On 2016 Aug 16, Clive Bates commented:

Note: The focus on conflicts of interests (COI) received criticism both for exaggeration and for the tacit assumption that a COI declaration somehow invalidates the scientific merit of the work. See Kosmider L, 2016 Ideology versus evidence: Investigating the claim that the literature on e-cigarettes is undermined by material conflict of interest and the reply from the authors Pisinger C, 2016 Reading the conflict of interest statement is as important as reading the result section: Response to the letter by Dr. Kosmider.

COI declaration is not intended as a form of self-invalidation, but for transparency. It is absurd to assert that it is as important as the substance of the work itself. COI tends to focus on conflicts arising from commercial interests, but many researchers in this field have undisclosed conflicts relating to funders, regulators, employers' prior policy positions, and their long-held beliefs. If these authors want to investigate COI (not the ostensible purpose of this work), then they should do it rigorously, and with a broader view of COI.

On 2016 Aug 05, Clive Bates commented:

Despite the title, the review does not address the "health effects of electronic cigarettes". Or perhaps to the extent that it does, it does not find anything negative to report. This may explain the diversionary conclusion, which resorts to an argument about competing interests rather than a critique of actual scientific findings [see note appended below].

It does, however, attempt to summarise various proxies for health effects but does this very poorly. Throughout the substance of the article, it fails to internalise the most basic concept in toxicology:

"the dose makes the poison" (attributed to Paracelsus, 1538).

The assessment of e-cigarette science by the Royal College of Physicians (London), Nicotine without smoke: tobacco harm reduction, 28 April 2016 included this review but passed over it in the following terms:

Recent reviews of the health effects of toxins inhaled during normal use of e-cigarettes have expressed concerns over potential adverse effects based on the presence of these contaminants, but not their levels, which are generally the more important determinant of toxicity.

The danger of a review that approaches toxicology in this way is that plays into public fears inspired by chemical names and may deter smokers, who are at well-established serious risk from high exposures to harmful agents, from switching to the much safer vapour products. In doing so, the careless framing of scientific findings can cause actual harm.

In addition, such work provides a citable source for irresponsible activism, and so can entrench misunderstanding in regulatory bureaucracies like the WHO, FDA and European Commission - further amplifying the harm through ill-conceived regulation.

For any future work, these authors or others should carefully distinguish between the following:

• The detection of the presence of a hazardous agent in vapour, recognising a de minimis level for which there is no material concern - see Burstyn I, 2014 for an approach that adopts occupational exposures as a frame of reference.

• Human exposure to a hazardous agent that may or may not present a risk to health recognising that magnitude of exposure and materiality matter. For an insight into how to approach this, see Farsalinos KE, 2014. These authors provide a good example in Table 3, showing e-cigarette nitrosamine exposure to be about 1,000 times lower than in cigarette smoke and comparable to the residual levels found in regulated medicinal NRT.

• Observed damage to cells found in vitro following exposure to vapour that may have no bearing on human exposure or risk. See criticism of the extreme misinterpretation of one cell study finding in PubMed Commons on Yu V, 2016. This study provides a good example of what not to do.

• Observed impacts in animal studies, which are rarely a reliable proxy for human experience - see discussion by the Laura and John Arnold Foundation of the weakness of animal studies: Why journalists should stop publishing mouse studies

• The illusion of risks induced by operating vaping equipment in a way that no human user will ever experience other than momentarily. The Jensen RP, 2015 study on 'hidden formaldehyde' is a case in point and has been severely and appropriately criticised (Bates CD, 2015) for running equipment too hot and in 'dry puff' conditions, and then sensationalising the findings as if they showed a cancer risk, and suggesting these may be greater than for smoking.

• Observed changes to the human body that are not necessarily an indication of harm to health. Nicotine and coffee both create changes to the cardiovascular system, but neither is a direct cause of material harm - see PubMed Commons review of Carnevale R, 2016.

• Signs of immediate health gains when e-cigarettes are used as an alternative to smoking - for examples see Polosa R, 2015 and Polosa R, 2014 for examples of 'harm reversal' in chronic respiratory patients.

Finally, given that almost all e-cigarette users are former or current smokers, then any attempt to provide useful information on e-cigarette risk should always use cigarette smoking and related risk as a point of reference, not only the risks compared to complete abstinence. To avoid this comparison is to miss the value of harm reduction as a strategy in tobacco control. There are many examples of this failure - to take just four:

• Allen JG, 2016 raised the alarm about an additive, diacetyl, but failed to point out that, as one critic put it, "average diacetyl exposure from vaping is 750 times lower than from smoking".

• Yu V, 2016 despite concluding that "e-cigarette vapor, both with and without nicotine, is cytotoxic to epithelial cell lines and is a DNA strand break-inducing agent", the study showed that cells exposed to the e-cigarette vapour medium were still alive after 8 weeks but all were dead in the tobacco smoke extract within 24 hours. This important comparison was buried in the paper, not mentioned in the abstract, and not included in the authors' publicity statements about the study.

• Schweitzer KS, 2015 attempted to show that nicotine causes "dose-dependent loss of lung endothelial barrier function, which is associated with oxidative stress and brisk inflammation". But this study has been criticised for using nicotine exposures 500-50,000 times higher than nicotine concentrations in the blood of typical smokers or vapers: Cell studies on e-cigarettes: don’t waste your time reading (at least most of) them

• Paley GL, 2016 concluded that e-cigarette battery fires and explosions were a "significant public health risk", based on six reported incidents. The authors asserted this without noting that in 2011 there were 90,000 smoking-material fires in the U.S. resulting in an estimated 540 civilian deaths, 1,640 civilian injuries and $621 million in direct property damage. To the extent vaping replaces smoking it will reduce this toll and result in a significant public health win.

On 2016 Aug 05, Clive Bates commented:

Despite the title, the review does not address the "health effects of electronic cigarettes". Or perhaps to the extent that it does, it does not find anything negative to report. This may explain the diversionary conclusion, which resorts to an argument about competing interests rather than a critique of actual scientific findings [see note appended below].

It does, however, attempt to summarise various proxies for health effects but does this very poorly. Throughout the substance of the article, it fails to internalise the most basic concept in toxicology:

"the dose makes the poison" (attributed to Paracelsus, 1538).

The assessment of e-cigarette science by the Royal College of Physicians (London), Nicotine without smoke: tobacco harm reduction, 28 April 2016 included this review but passed over it in the following terms:

Recent reviews of the health effects of toxins inhaled during normal use of e-cigarettes have expressed concerns over potential adverse effects based on the presence of these contaminants, but not their levels, which are generally the more important determinant of toxicity.

The danger of a review that approaches toxicology in this way is that plays into public fears inspired by chemical names and may deter smokers, who are at well-established serious risk from high exposures to harmful agents, from switching to the much safer vapour products. In doing so, the careless framing of scientific findings can cause actual harm.

In addition, such work provides a citable source for irresponsible activism, and so can entrench misunderstanding in regulatory bureaucracies like the WHO, FDA and European Commission - further amplifying the harm through ill-conceived regulation.

For any future work, these authors or others should carefully distinguish between the following:

• The detection of the presence of a hazardous agent in vapour, recognising a de minimis level for which there is no material concern - see Burstyn I, 2014 for an approach that adopts occupational exposures as a frame of reference.

• Human exposure to a hazardous agent that may or may not present a risk to health recognising that magnitude of exposure and materiality matter. For an insight into how to approach this, see Farsalinos KE, 2014. These authors provide a good example in Table 3, showing e-cigarette nitrosamine exposure to be about 1,000 times lower than in cigarette smoke and comparable to the residual levels found in regulated medicinal NRT.

• Observed damage to cells found in vitro following exposure to vapour that may have no bearing on human exposure or risk. See criticism of the extreme misinterpretation of one cell study finding in PubMed Commons on Yu V, 2016. This study provides a good example of what not to do.

• Observed impacts in animal studies, which are rarely a reliable proxy for human experience - see discussion by the Laura and John Arnold Foundation of the weakness of animal studies: Why journalists should stop publishing mouse studies

• The illusion of risks induced by operating vaping equipment in a way that no human user will ever experience other than momentarily. The Jensen RP, 2015 study on 'hidden formaldehyde' is a case in point and has been severely and appropriately criticised (Bates CD, 2015) for running equipment too hot and in 'dry puff' conditions, and then sensationalising the findings as if they showed a cancer risk, and suggesting these may be greater than for smoking.

• Observed changes to the human body that are not necessarily an indication of harm to health. Nicotine and coffee both create changes to the cardiovascular system, but neither is a direct cause of material harm - see PubMed Commons review of Carnevale R, 2016.

• Signs of immediate health gains when e-cigarettes are used as an alternative to smoking - for examples see Polosa R, 2015 and Polosa R, 2014 for examples of 'harm reversal' in chronic respiratory patients.

Finally, given that almost all e-cigarette users are former or current smokers, then any attempt to provide useful information on e-cigarette risk should always use cigarette smoking and related risk as a point of reference, not only the risks compared to complete abstinence. To avoid this comparison is to miss the value of harm reduction as a strategy in tobacco control. There are many examples of this failure - to take just four:

• Allen JG, 2016 raised the alarm about an additive, diacetyl, but failed to point out that, as one critic put it, "average diacetyl exposure from vaping is 750 times lower than from smoking".

• Yu V, 2016 despite concluding that "e-cigarette vapor, both with and without nicotine, is cytotoxic to epithelial cell lines and is a DNA strand break-inducing agent", the study showed that cells exposed to the e-cigarette vapour medium were still alive after 8 weeks but all were dead in the tobacco smoke extract within 24 hours. This important comparison was buried in the paper, not mentioned in the abstract, and not included in the authors' publicity statements about the study.

• Schweitzer KS, 2015 attempted to show that nicotine causes "dose-dependent loss of lung endothelial barrier function, which is associated with oxidative stress and brisk inflammation". But this study has been criticised for using nicotine exposures 500-50,000 times higher than nicotine concentrations in the blood of typical smokers or vapers: Cell studies on e-cigarettes: don’t waste your time reading (at least most of) them

• Paley GL, 2016 concluded that e-cigarette battery fires and explosions were a "significant public health risk", based on six reported incidents. The authors asserted this without noting that in 2011 there were 90,000 smoking-material fires in the U.S. resulting in an estimated 540 civilian deaths, 1,640 civilian injuries and $621 million in direct property damage. To the extent vaping replaces smoking it will reduce this toll and result in a significant public health win.

On 2016 Aug 05, Zvi Herzig commented:

None

On 2016 Aug 05, Zvi Herzig commented:

Adding to Bates' comments below, here are some specific misrepresentations of the primary evidence by the Pisinger and Døssing review, relating to almost every major issue in e-cigarette toxicology (not an exhaustive list):

• “Some studies found high maximum concentrations of total TSNA”, citing studies showing TSNAs 100-200 times below cigarette smoke levels [1-3].
• “Exposure to formaldehyde was comparable with smoking", referring to a study calculating formaldehyde levels nine times below that from tobacco smoke [1].
• “Propylene glycol has been found to exacerbate and/or induce multiple allergic symptoms in children”, citing a study stating that "apparently… outcomes were not driven by propylene glycol” [4].
• “Values below the threshold limit don't necessarily protect against the health effect of 200–300 daily inhalations over decades”, referring to safety limits calculated for 8 hours exposures “day after day, over a working lifetime” [5].
• “These metals appear on the U.S. Food and Drug Administration's 'Harmful and Potentially Harmful Chemicals' list”, referring to metals detected below levels acceptable to the FDA for chronic inhalation [6-8].

[1] Goniewicz ML, Knysak J, Gawron M et al. Levels of selected carcinogens and toxicants in vapour from electronic cigarettes. Tob Control 2014

[2] Kim HJ, Shin HS. Determination of tobacco-specific nitrosamines in replacement liquids of electronic cigarettes by liquid chromatography-tandem mass spectrometry. J Chromatogr A 2013

[3] Farsalinos KE, Romagna G, Voudris V. Authors miss the opportunity to discuss important public health implications. J Chromatogr A 2013

[4] Choi H, Schmidbauer N, Spengler J et al. Sources of propylene glycol and glycol ethers in air at home. Int J Environ Res Public Health 2010

[5] ACGIH. Chemical Substances Introduction. acgih.org 2016. http://www.acgih.org/tlv-bei-guidelines/tlv-chemical-substances-introduction

[6] Siegel M. Metals in Electronic Cigarette Vapor are Below USP Standards for Metals in Inhalation Medications. The Rest of the Story: Tobacco News Analysis and Commentary. 2013.http://tobaccoanalysis.blogspot.com/2013/04/metals-in-electronic-cigarette-vapor.html .

[7] Farsalinos K. Metals and nanoparticles in e-cigarettes. 2013.http://www.ecigarette-research.com/web/index.php/2013-04-07-09-50-07/2013/89-metals-and-nanoparticles-i n-e-cigarettes-commentary.

[8] Farsalinos K, Voudris V, Poulas K. Are Metals Emitted from Electronic Cigarettes a Reason for Health Concern? A Risk-Assessment Analysis of Currently Available Literature. International Journal of Environmental Research and Public Health 2015