Small population size (anecdotal, really); no controls (unlike in twin studies).
It is not just sample of 1, but it is also non-randomized sample of 1. Reasercher didn't flip a coin and assign one twin to smoking another to non-smoking group (of one). If they are so identical in all regards (genetic and environmental), why would one be smoking heavily (52 pack-years at age 52). That's a clear marker of some sort of additional hardship in life, some unfortunate exposure or stress, bad marriage or stress at work, chemicals at work or home,.. (e.g. she had early on a tonsillectomy that nonsmoking sister didn't, clearly during childhood she was more susceptible to inflammation, or had a bad luck with doctors). When someone smokes in this day and age, and that heavily for a woman, or even during the last 2-3 decades of antismoking pressures and propaganda, that is clear a marker for something else that is likely to affect her health negatively (including her diet, skin) e.g. she clearly cared about her health much less than her non-smoking sister, considering all the scare stories abourt smoking everyone hears about from every source, including those on each pack of cigarettes.
That lack care for her own health couldn't just stop at heavy smoking, but it must have permeated all other aspects of her life (maybe she is a risk taker, contrarian,..). And that attitude toward her health and life obviously took its toll. What other risks she took, or advice she rejected, that the other one didn't. For example the non-smoking sister is using moisturizer, while the smoking one isn't, further confirming different attitudes toward their own care and looks. That can make skin look different, too, you know.
We can obviously discount the unlikely possibility that the smoking sister went out and read all the science about tobacco smoke and concluded that antismoking is a scientific fraud and so she decided to smoke rationally as adult (as I actually did; I didn't smoke in my teens or most of college despite great peer pressures to take up smoking to look cool; I started smoking fully in grad school as result of a rational conclusion, just when many of my friends succumbed to the antismoking pressures and quit).
Interesting few bits from the table, though -- the smoking sister, despite all the other odds against her (risk taking, lack of care for her health, possible other hardships which caused her to smoke against all advice), didn't get arthritis, while the non-smoking sister did by age 52. She also didn't have her gallbladder removed. Of course, therapeutic & protective effects of smoking in arthritis (
osteo threefold for knee, and
rheumatoid) are well established even in animal experiments, so that's not surprise. Gallbladder difference may be due to other anti-inflammatory or detox effects o tobacco. The most striking difference on the face photos is droopy, dull eyes and face of the non-smoking sister vs the bright and bushy-tailed, lively look of the smoking sister. One could almost tell which one will get Alzheimers or other form of senility in twenty years. As to wrinkles, yes, the smoking one appears more wrinkly, although well within the effects of a moisturizer that non-smoking sister is using. The non-smoking sister is also more overweight (BMI 33 vs 30.1 for smoking sister), which can smooth out some wrinkinling.
In summary, a very weak and very unconvincing case, with ambiguous net results (of all effects), at best, for the antismoking mission. Frankly, if it weren't for the antismoking hysteria and its big bucks sponsors, this kind of "scientific research" would have been laughed out of the journal by the referees after the first page. Today, of course, any junk study that "proves" <smoking bad, pharma potions good>, goes.
In Japan and Korea, 60-70 percent of men smoke, yet they tend to look more youthful . . .
Confounding variable "race" introduced. Perhaps non-smoking Japanese and Koreans look even younger than their smoking counterparts.
Back in 1940s and 1950s, actors and other celebrities were largely smokers, they didn't have botox or face lifts, yet they didn't look particularly wrinkled, certainly not more than nonsmokers of that era . . .
This is sweeping and subjective.
They were casual observations in response to casual observations of the same kind. My real argument (the scientific backing for more serious readers) was given right ofter those lighter comments.
Some of the underlying biochemical reasons why smoker's skin (and every other marker of youthfulness) would come out younger in any apples to apples comparisons (not just the same genetics, but sun & other exposures, diet, stress, socioeconomic status,...):
Twin studies are as apple-to-apple comparisons as ethically possible (as, obviously, researchers cannot
force people to start or quit smoking). Also, differences in sun exposure (and other lifestyle) can certainly be taken into account (via multiple regression, as in this twin study).
There are lots of things not taken into account, many of them deliberately. I have yet to see an antismoking non-randomized stats study (which is just about all of them, except for handful of randomized intervention trials that backfired and are never talked about) that brings up
self-medication confounding which ought to be right on the top of the confounding variables, considering the spectrum and depth of well established (and well known, but never publicized) protective and therapeutic effects of tobacco smoke, especially the upregulation of detox & antioxidant enzymes, MAO B effects, numerous anti-inflammatory effects,... None of them exists for the antismoking epidemiology matrix, hence they are never mentioned, let alone acknowledged as relevant confounding variables that need to be measured and their effects on the results quantified.
Further, as noted in the comments on the twin sisters case, the antismoking indoctrination itself is quite a relevant confounding variable. Somene who smokes in the face of the present day gross abuses of smokers at all levels (social, economic, psychological), "knowing" all about alleged harm and yet doing it, says a lot about person and his or her general attitude toward health, society, or any other advice, risk taking, as well as about unmentioned stresses and hardships in their lives. All these major confounders can skew and drown statistically many beneficial biochemical effects of tobacco smoke. These confounders are uniformly and deliberately ignored by the antismoking researchers, which strongly drives the statistical correlations on non-randomized samples in the directions disfavoring smokers (or masking the beneficial effects of t.s.).
As an illustration, just take a look at the
rheumatiod arthritis experiment mentioned above. Statistically, as nothed in the intro of the paper, smoker have higher rates of r.h., and doctors will advice r.h. patients to quit smoking (as they did to my mother in law), yet the experiment demonstrates clearly the protective and therapeutic effects of nicotine alone, and the additional beneficial effects from the real tobbaco smoke. The results are also perfectly expected from previously well known potent anti-inflammatory & analgesic effects of t.s. (and of nicotine alone), as well as the known negative correlation betwene smoking and osteo-arthritis. Clearly many r.h. patients, or those in preclinical stages, would perceive smoking as beneficial in many ways, to alleviate the symptoms, to tolerate pain better, to delay the onset and progression of the cartilage destruction. Hence there is a more than obvious
self-medication confounding that needs to be quantified and taken into account. Yet, such confounding is never mentioned in antismoking epidemiology, let alone quantified and taken into account. Thus the non-randomized stats from antismoking epidemiology indicate the negative effects of t.s. on r.h. which is exactly the opposite from the real effects (strongly beneficial), which results from the known effects of t.s., and which confirmed also by direct experiments. In short, the antismoking epidemiology has long degenerated into the pure junk science, worthless to anyone but those paying for it.
The studies on the individual effects of smoking you cite look interesting. I'm not going to argue smoking/tobacco/nicotine cannot have any positive effects (including hormetic). However, based on the aforementioned twin studies (and their literature reviews), I would conclude that, at least for skin-aging, the harms greatly outweigh the benefits. (Many of the skin benefits of smoking can be gained more safely, e.g., by supplementing on CoQ10, as found in tobacco leaves.)
I suppose, as long as you limit the benefits to those with well identified mechanisms and isolated compounds, you will be fine with pharma substitutes. Considering though how little the biology and biochemistry really know about even the mere biochemical networks of a single cell (just few tips of the iceberg of the vast intelligent network within each live cell, capable of engineering new live cells at the molecular level from scratch), let alone about astronomocally more complex network of such cellular networks, the organism, I would guess, sticking to the understood substitutes you would be limiting yourself to few percent of the full benefits at best, if not harming yourself (with unbalanced, dissonant mishmash of supplements).