by Carl V Phillips
Sorry for the blog silence. Busy. It will continue for a few days. In the meantime, I recommend reading this post by Erik (Atakan) Befrits about Sweden caving to WHO pressure to try to use warning labels to try to scare people away from THR. It contains some important perspective that many of my readers my not normally get. (Obvious disclaimer: I don’t necessarily agree with every word of the post.)
The crux:
Make no mistake about this: This new law on snus will have exactly ZERO effect here in Sweden where ZERO people die from using snus. It will however have devastating effects for the health of smokers and users of toxic smokeless formulations in THE 193 OTHER recognized countries in the world.
Tobacco harm reduction, anti-THR lies, and related topics 
Professor Phillips, I would want to improve on the post, or the next one. Please let me know what you don’t agree with on mail. Thx
Wow. Thanks for such an insightful and enlightening response. I had no inclination or suspicion that the whole TSNA concept with regards to Swedish snus vs American ST could have been the result of a marketing ploy to damage THR in the U.S..
“But the thing is, the levels of TSNAs in Swedish products used to be considerably higher than they are now (based on what data we have, and inference) and snus did not cause harm back then either.”
Now THAT is interesting. How much higher? If it’s not the TSNAs (speaking of old school American smokeless tobacco) that are potentially harmful, then what other constituents could have been responsible in the past?
I did a paper c2005 that reviewed the available info on historical TSNA levels and found that the then-current levels in US product were about the same as the levels in Swedish product in the 1980s (which, of course, gave us the lack of cancers in the epidemiology from the 1990s). Since then both have continued to come down quite a lot, and probably the trend was generally downward before then too, but we don’t have measurements.
It is really not all that clear that old school American ST was harmful. There is a single outlier study from the 1970s that found a notable elevated oral cancer rate. Other very old US studies were fatally flawed (e.g., not controlling for smoking). Newer US studies have not found the same result. So that is just one study and has some definite limitations, so might well be wrong. The subjects of the study used an odd product, in any case, Appalachian-style dry snuff, not merely an older version of Copenhagen, so that might be a genuine difference. But if we assume there the different measured effect was real, it could be any number of chemical properties — we don’t have any of the product around to test, after all. Or it could have been an interaction effect: It caused cancer when coupled with being a woman (the main users of this odd product) who was born c1920 and lived a particular impoverished rural lifestyle, but it would not cause cancer even if the same product were used by typical ST users now. Epidemiology is tricky like that.
Thanks Carl. This is certainly troubling news.
What part of what Atakan’s post do you not agree with? ..just curious..
So at first I left out the word “necessarily” from the standard “I do not necessarily agree” disclaimer that I put in. I have since fixed that. But I supposed I must pay for my sloppy editing by answering the question. Give me a few minutes and I am sure I can find *something* :-). BRB.
(Sigh. I did not expect to have to be grading papers today.)
Ok, so for what I owe everyone per the above comments, I found just one thing in the OP that I starkly disagree with, though it is something that Atakan quoted, not something he wrote himself. In the quote from Siddiqi at the end, is the observation: “At one end (Swedish snus) is more or less harmless and can be used as harm reduction substitute while on the other hand the products used in Asia and Africa have very high TSNA and are extremely harmful.”
Technically the statement does not say that the other products are harmful *because* of high TSNA, but it clearly implies that. This common claims that the TSNA levels matter a lot not actually established and is supported only by the thinnest evidence. It is clearly the wrong perspective on South Asian products, whose harm clearly comes from the primary ingredients and other chemicals that have nothing to do with there being tobacco in the mix. I have written about that on this page quite a few times. The story with Sudanese-style products is not quite so obvious. Frankly, it is not quite so obvious they cause harm, because the epidemiology is so poor. If they do, it might be from high TSNA levels, but it might not. It is not like that is the only variable.
Basically the TSNA claims come from assuming the conclusion and then finding evidence that is *not inconsistent* with the conclusion and saying that therefore the assumed conclusion must be right. It is textbook bad science — exactly the “reasoning” that produced myths throughout time (see, e.g., “balancing the humours”). The whole concept was basically concocted by American ANTZ and Swedish marketers to try to argue that the Swedish data does not apply to American products, and so the ANTZ could keep lying about American ST and the Swedish companies could gain some competitive advantage. But the thing is, the levels of TSNAs in Swedish products used to be considerably higher than they are now (based on what data we have, and inference) and snus did not cause harm back then either.
One other point, not a disagreement, but a discussion: Atakan discusses the different implications of the “may cause harm” and “causes harm” phrasing on the warnings (which I realize is a translation, so nuances might be lost). From a proper epidemiologic perspective, such phrases do not normally mean any different from each other. That is, at an epidemiologic level, to say that something *does* cause harm is to say that it causes harm in *some* people (pretty much never everyone), which then at an individual level translates into *may* cause harm for you personally. Of course I realize that *may cause* can be interpreted as “in theory, this could cause harm to someone, but maybe not” — in which case is an utterly vacuous statement, since that is true of everything. I also realize that the vibe of the communication as interpreted non-technically is different. I agree that both versions of it are an overstatement, since communicating that vacuous “in theory” statement is obviously inappropriate, and so given the lack of evidence that there is harm, neither should be said. It obviously should not be changed in the direction of making this unsupportable claim “feel” even more definitive.