posted by Elaine Keller (with input from Carl V Phillips)
Mr. Smith was delivering a speech across town in front of 800 people when the murder occurred. He had no blood on his clothing. Police could find no trace evidence at the scene that leads to Mr. Smith, and were unable to come up with a motive for the crime. Mr. Smith is a prime suspect.
The last sentence is a confounding conclusion. Not confounding in the epidemiologic sense – I will leave that topic to Carl – but in the sense of being baffling. It causes surprise or confusion because it acts against the reader’s expectations. When such conclusions appear in a murder mystery or puzzler, they can be entertaining. When they appear in scientific journal articles, they are perplexing. Or, in the spirit of this blog, they are lies.
Take the case of the paper published by German researchers on the subject of chemicals in electronic cigarette (e-cigarette) vapor.
Schripp T, Markewitz D, Uhde E, Salthammer T. Does e-cigarette consumption cause passive vaping? Indoor Air. 2012 Jun 2.
The final paragraph of the Conclusions section states,
“Overall, the e-cigarette is a new source of VOCs and ultra-fine/fine particles in the indoor environment. Therefore, the question of “passive vaping” can be answered in the affirmative. However, with regard to a health related evaluation of e-cigarette consumption, the impact of vapor inhalation into the human lung should be of primary concern.”
The reader is left with the impression that Schripp et al. found chemicals in e-cigarette vapor that would definitely endanger the health of users. The most likely way readers will interpret the phrase “passive vaping” is that the researchers found chemicals in the exhaled vapor that would be hazardous to the health of bystanders as well, given the fact that the CDC attributes 49,000 deaths each year to “passive smoking.”
The first rule of toxicology is “the dose makes the poison.” This means that it is important to know not only what chemicals are involved, but also the quantity of chemicals that are present. Almost any substance (even water) is toxic in large enough quantities and many “toxic” chemicals are harmless, or even helpful in some cases, in smaller quantities.
Fluoride is a good example of the first rule of toxicology. In concentrations of less than 0.5 percent in toothpaste, stannous fluoride and sodium fluoride helps strengthen teeth and prevent cavities. However, toothpaste overdose may cause stomach pain and possible intestinal blockage.
So what experiments did Schripp and his colleagues perform? There were two parts. The “large scale vaping/smoking experiment” was performed in an 8 cubic meter stainless-steel emission test chamber (about the size of the interior of American family SUV or minivan – with the windows up and the vents closed). A volunteer sat in the chamber and air quality was sampled after 20 minutes to establish a baseline. Then the volunteer was given an e-cigarette with one of three liquids: apple-flavored with no nicotine; apple-flavored with a nicotine concentration of 1.8%; and tobacco flavored with 1.8% nicotine. This was followed by a last trial, in which the volunteer smoked a cigarette.
For the large scale stainless steel chamber experiment, the researchers reported on the 20 compounds with the highest concentrations, comparing them to the concentrations at baseline. Fourteen of the 20 compounds that increased for the cigarette smoke showed no increase over baseline for the vapor of any of the three e-cigarette samples. The six compounds that did increase for vapor samples were 2-butanone (MEK), acetic acid, acetone, isoprene, formaldehyde, and acetaldehyde.
In their conclusions, the researchers failed to point out that many more compounds were found in smoke than in vapor, and they did not compare the quantities of compounds measured in vapor to those measured in smoke. The quantities measured in vapor ranged from 1/10th to 1/40th those generated by cigarette smoke.
Just how hazardous are the compounds in vapor? The Occupational Safety and Health Administration publishes permissible exposure limits (PELs) for hundreds of chemicals that might be present in the air at workplaces. Five of the six compounds were present in quantities that are less than 1% of the PEL. The sixth compound, formaldehyde, is produced naturally by the human body, and it was present at 2.4% of the PEL. If the researchers had provided this comparison in their data, it would have been obvious that their conclusions did not fit the facts.
Apparently the researchers were surprised at what they did not find. “Although 1,2-propanediol [propylene glycol] was detected in traces only within the 8 m³ chamber during the consumption of e-cigarettes, this compound must be released due to the visible fume in the exhaled breath. To determine the VOC composition in the breath gas directly, an e-cigarette smoker exhaled into a 10 L glass chamber.” (Interestingly, this could be interpreted as them saying, “we changed our methodology on the fly because we did not like the results we were getting.”)
Perhaps they also were surprised that nicotine did not show up in their list of the 20 compounds with the highest concentration in smoke. Analysis of the immediately captured breath in the glass chamber resulted in a different list of chemicals than the stainless steel chamber experiment. The abstract states, “Prominent components in the gas phase are 1,2-propanediol, 1,2,3-propanetriol, diacetine, flavourings and traces of nicotine. As a consequence, ‘passive vaping’ must be expected from the consumption of e-cigarettes.”
The sequence of these sentences would lead the reader to believe that the chemicals specified in the first sentence lead to a condition they call “passive vaping” implying that it is similar in risk to “passive smoking.”
The extremely low quantities in the stainless-steel chamber experiment indicate that most of the chemicals found in concentrated captured exhalation disappear in the ambient air. For example, although nicotine was present (at 1.4% of the exposure limit) in the glass chamber experiment, no nicotine at all was detected in the stainless-steel chamber experiment. A bystander would need to lock lips with an exhaling e-cigarette user to be exposed to all the “prominent components of the gas phase” measured in the glass container experiment.
Even with the lip-lock, the highest level of chemical exposure in the second experiment (glycerin) is only 9.5% of the PEL. Two of the chemicals are not considered harmful at all. Not surprisingly, the highest concentration was for 1,2-propanediol, aka the non-toxic carrier, propylene glycol. If passive vaping is supposed to mean that bystanders are exposed to harmful levels of chemicals, then neither experiment in the study proved the existence of passive vaping.
Nothing in the perplexing conclusion to this article informs the reader about the extremely low level of danger represented by the quantities of chemicals detected. An accurate conclusion might have stated, “The consumption of e-cigarettes causes emissions of aerosols and VOCs, such as 1,2-propanediol, flavoring substances and nicotine, into indoor air; however the quantities of these substances are so low that they do not present a health hazard to bystanders or to the users themselves.”