Talk:Smoker's paradox

Sources review
Firstly, I acknowledge the work that has been put in this article; a "thank you" to contributors. However, I must mention some issues I may have:
 * "Smoker's paradox" is only mentioned as a term in relation to cardiovascular disease, COVID-19, and general in-hospital mortality. It is not mentioned nor described as a concept in most of the other sources. Instead, a reverse relationship is mentioned, without being described as contrary to what expected based on the mechanisms of the disease. In other words, the article is written simply on the possible beneficial effects of smoking in various diseases, and in my opinion it editorializes by collectively naming them "smoker's paradox". There is very limited literature on smoker's paradox as a general concept (see bullet 3).
 * Specifically on cardiovascular disease, smoker's paradox has not been definitely proved, may have been influenced by various confounding factors in research, and "has lost momentum in recent years". While there is an important disclaimer against any though of smoking as non-toxic, the article goes on to list a series of diseases that "may" have inverse relationship with smoking, without mentioning the multiple cofounding factors.
 * The only sources that cross the boundaries of specific diseases and cover the likelihood of beneficial smoking effects in more than one systems are: Baron 1996, which does not mention the term "smoker's paradox"; Wang 2020, which is a conference abstract and has serious sources of bias (not mentioning confounding factors like age and performance status), and is actually a primary source; and Usman 2020 . The latter is the most interesting one, as it is recent, it concentrates on COVID, it specifically mentions the term, and devotes a paragraph-long review on it. Other than that, the actual term is only mentioned in citations on cardiovascular disease (NB: I have no full access to some of the articles). Is Usman 2020 enough for an encyclopedia article on the subject? (Of note, Usman 2020 draws the conclusion on smoker's paradox in COVID-19 that "the data supporting smoker’s paradox claims are limited and questionable").
 * Multiple sources are primary ones: Weinblatt 1968, Harles 1982, Anderson 2008 (also a preclinical study on tissues), Brunet 2008, Cramer 1986, Galo 2018 (though I admit a meta-analysis could be cited instead), van Duijn 1991, Loftus 1996, Hjern 2001, Wang 2020, Cohen 2001, Desai 2009. I also note how old many citations are. Does smoker's paradox merit an actual article on it, or could it be covered with passing references on smoking effects in the articles on specific diseases, and in Health effects of tobacco, section "benefits"?

NikosGouliaros (talk) 22:59, 30 March 2021 (UTC)

Statistical anomaly
Do some sources discuss this as a statistical anomaly? It would not be surprising viewed from this perspective. — Paleo Neonate  – 00:54, 11 April 2021 (UTC)