Scientific claims in the FDA deeming regulation (part 3 of ???)

by Carl V Phillips

Continuing this series with page 21 of the draft regulations.

I will skip the paragraph that is about little cigars basically being the same as cigarettes, and thus perhaps subject to their flavor bans, but will use it to anchor one observation:  Most of the content analyzed in the previous post, as well as much of what follows, conflates observations about very different products.  FDA is cogent when making points about little cigars being basically no different from cigarettes.  But they present an utter muddle when they try to talk about, say, flavors in general, as if they have similar implications for e-cigarettes, hookahs, and cigars.

Moreover, efforts to improve public health by reducing the prevalence of cigarette smoking may be undermined by tobacco users switching to other tobacco products.

This is the transition from talking about little cigars to e-cigarettes.  It further speaks to the inappropriate conflation.  Because this series is looking purely at the scientific claims, I will not delve into the ethical implications of, in effect, complaining “what consumers prefer to do is at odds with what we are trying to force them to do.”

The scientific evidence remains as yet unclear what the public health impact will be from products such as e-cigarettes.

Contrast this assertion with the claims analyzed in the previous post in this series.  In the material analyzed there, numerous affirmative assertions about responses to flavors and such were made based on either no evidence or very thin and tenuous evidence.  Here they are making a negative assertion about limited knowledge in a context where there is far more known than there is about the flavor assertions.

Of course, if the statement is interpreted as “we do not know, down to three significant figures, what the impact will be” then it is obviously true, as it is about everything.  But to take the natural interpretation, rather than the vacuous extreme interpretation, they seem to be asserting “we cannot predict whether e-cigarettes will be quite positive, quite negative, or somewhere in between.”  This is simply false.  The number of Americans who have quit smoking using e-cigarettes is already approaching the point that it would be arithmetically impossible for the net impact to be negative.  Meanwhile, there is no basis for claiming there are any negatives at all.

More youth who report they would never have used a tobacco product are experimenting with e-cigarettes (Ref. 4, 18);

This is the type of sloppy sentence that is common for creating innuendo in propaganda.  It is entirely inappropriate for a regulatory document.  More than what?  Presumably the intended claim is “more than in the past”, because this is what Reference 4 (the notorious over-concluding CDC report) found.  But, of course, an increase in experimenting with something that is brand new is obviously inevitable.  Note also that the CDC, to my knowledge, did not attempt to assess whether an individual would have never used a tobacco product (how could they?).  Needless to say, Reference 18 — a 1981 document about cigarette use (which I now have updated the analysis of in the previous post) — has even less information about experimenting with e-cigarettes.  (Its inclusion here is obviously a substantive typo, which illustrates a different sort of question about the validity of the document.)

The innuendo in this statement is that the number of experimenters who were not already tobacco users is large, which is false.  Moreover, given the previous sentence, the innuendo is that this has a substantial negative public health impact.  However, there is no reason to believe that even if every single teenager were forced to experiment, that there would be a public health impact, let alone that it would be negative if there were one.

the number of cigarette smokers who actually quit tobacco product use with e-cigarettes is low (Ref. 19);

This is pure word games.  A casual reader would interpret this as “the number of cigarette smokers who quit smoking…”.  Of course, the role of e-cigarettes for the vast majority of users is an ongoing substitute for smoking.  Assuming e-cigarettes are being considered tobacco products — which, of course, is FDA’s position — this means that most successful cases of smoking cessation would not be counted.  Moreover, “low” is a misleading claim in itself.  Low compared to what?

Reference 19 is about a randomized trial of e-cigarettes in a cessation clinic setting (Bullen et al.), which has nothing to do with the number of smokers who have actually quit.

current cigarette users experimenting with e-cigarettes have become dual users (id.)–with unknown health impacts.

The first part of this is obviously true, and indeed understated (as phrased, all it says is that there are two or more such individuals).  Many e-cigarette users transition off of smoking gradually, and given the trends, the number in that transition phase is probably increasing.  Some of them might never complete the transition.  However, there is overwhelming evidence that smoking less is better than smoking more, and there is no reason to believe that there is any synergistic harm from smoking less while also using e-cigarettes.  Thus the health impacts are, in fact, known to be positive.

Again, the referenced clinical cessation experiment tells us nothing about what is happening in the world, and thus is irrelevant to the claim.  What I believe we are witnessing there is the freshman term-paper phenomenon of authors relying on the aside introductory assertions of other authors that are not supported by the reported research in question.  Instead of vetting the assertions by trying to trace them back to real evidence — which, if found, should be what is cited, not the secondary summary statement in an irrelevant research report — naive authors often act as if anything that someone wrote down must be true.

Although the health consequences of e-cigarettes are not well understood because of their relatively new entrance into the market, the health concerns and addictive properties of other tobacco products have been widely recognized in Surgeon General Reports and scientific literature.

It is difficult to tell exactly what they are claiming here, due to a complete lack of citations.  While random[*] citations are sprinkled throughout most of the text, this statement that explicitly refers to previous literature includes none. [*That adjective is perhaps hyperbole, but not too far from being literally true.]

The first part of the sentence is another example of conflating lack of exact knowledge with lack of sufficient knowledge at the relevant level.  No, we do not know the health consequences of e-cigarette use within a factor of two or four.  But we do have sufficient evidence to conclude that the risk is in the range of two orders of magnitude less than that from smoking.  Needless to say, reference to “other tobacco products” (which includes the aforementioned little cigars, which are basically the same as cigarettes) is irrelevant.  There is a complete failure to recognize the known differences among products.

My reference to little cigars is not out of context.  The next statement in the document is:

When similar products are taxed or regulated differently, substitutions across products occur. For example, industry documents indicate that tobacco firms have been aware of disparities in the legal treatment of cigarettes and cigars and have made efforts to develop small cigars that cigarette smokers would smoke (Refs. 20 and 21). Sales of small cigars quadrupled in  the early 1970s, when cigars were taxed at a much lower rate than cigarettes and cigarette (but not small cigar) advertisements were banned from television and radio (Ref. 21).

This discussion of non-cigarette combustibles continues for the remainder of p.22.  While I am not analyzing FDA’s understanding of combustibles, the following is telling about their epistemic approach:

While researchers posited that this change in prevalence rates is likely due to the lower taxes (and ultimately lower cost to the consumer) ([Ref 22] at 566), the lack of regulation over certain tobacco products may be a contributing factor. Without a common regulatory framework, tobacco firms can exploit differences in regulatory requirements to drive consumers to different product markets.

Even though FDA acknowledges that there is general agreement that the substitution of products is driven primarily by different tax rates, they assert, without support, that their regulation (which does not change tax rates) would matter.  Reading between the lines, what they seem to be saying is “we intended to ban the flavors that some people prefer in these other combustible products, and then there will be less desire to use them rather than cigarettes.” Or perhaps what they have in mind is advertising.  What is important is that FDA is making hand-waving assertions about regulation per se having some particular impact (which is obviously not true) without offering any support for why particular regulation would have an impact.  They are clearly avoiding communicating what regulation they intend that they think would have an impact.  In other words, they are reporting their unsubstantiated guesses about the impact of a hidden agenda.