I quit for the longest of time...(started at 16), all of the sudden I found myself in Colombia, and all the restrictions I've gotten used to (no smoking in public places, etc) where non-existent. Further more, the packing of the cigarettes was still all branded, almost no warnings, and all kinds of 'clicky' flavored types. Don't want a whole pack? no problem, we got you! Buy singles on every corner...
Needless to say I started smoking again... I then quit again for a few years... went to northern Africa, and guess what? People literally smoking at the airport, £1 packs of camel, etc...
I haven't had one in a bit now... People think the restrictions don't work, but they absolutely do.
> People think the restrictions don't work, but they absolutely do
It's wild to me that people make assertions like this. We can (and should) argue about the effectiveness of smoking restrictions on decreasing cancer and heart disease, but to make the argument that they have no effect is just silliness.
You hear the same arguments from the right whenever gun control, environmental protections or really any sort of government regulation comes up, and from the left (though to a much lesser degree) on the topic of tax rates. We should all be able to agree that any sort of regulation will have a marginal effect, and be able to have a discussion about the level of that effect.
Counter-dote: I started smoking when deployed. It took me a few years to quit and that was a smoking -> chew -> patch/gum -> nothing.
No restrictions on sale other than 18+ (I was 24 when I started), and inside public places were only restricted a few years after I came back so I smoked a fair bit in local bars.
Smoking and similar habits are such a wild spectrum you can't prescribe any one thing and restrictions might help some who are prone but aren't really a cure.
Now that we are no longer used to smell the smoke, even getting close to someone smoking a cigarette makes me nauseous. Does not work with rolling tobacco, though.
> If you think that’s the stupidest thing you’ve ever heard, you can take it up with Sir Ronald Fisher, who actually made this argument with apparent sincerity in a 1957 letter to the British Medical Journal. I mention this in case you didn’t already know what an ass he was.
Not only did Fisher state this, and not only was he an ass, but it is almost embarrassing —it turns out he was partly right. You can read all about it in Pearl’s gem The Book of Why.
In essence Fisher argued that there could be a confound between the neurochemical and behavioral propensity to smoke, and the risk of developing lung cancer.
And strange as it seems to me, the DNA variant in CHRNA5 that is most highly associated with higher risk of nicotine addiction also modulates (independently) lung cancer risk. His principle was kinda-sorta defensible for an avid smoker.
Mendelian randomization of people by their CHRNA5 gene status can indeed demonstrate that Fisher was wrong. The majority causal risk is in fact smoking.
Interesting! Within almost every truth there lurks a valid but ultimately very minor counter-point.
Sorta a tangent here, but it's annoying that this is the very thing that causes people biased to not believe to latch onto as evidence they're right. Once introduced, further reasoning or curiosity about the subject tends to stop before weighing matters of degree, which leads to missing the big picture.
To generalize a lot, this is kinda where we are in the current meta of truth fracturing. Our biases drive us to discover nuances we grasp onto to avoid believing what we don't want to.
The idea that there's some confounding factor isn't in itself implausible, but a factor of 16 increase in risk is huge. Of course, to really be sure you should also look at how lung cancer risk and smoking correlates between countries, and within a country over time. And you can segment the population by sex and ethnicity. Oh and also look for a dose-dependent effect, very important. Heavier smokers should be at a higher risk.
Of course, the link between cancer is smoking is even stronger if some of the "non-smokers" were exposed to an increased level of cigarette smoke.
Both my parents are lifelong smokers from their early teens and were a year or two apart in age.
My mom died of lung cancer at 54.
My dad is still going strong at 77 and smokes Camel unfiltered.
Humans are weird.
Regardless, you don't want to die from lung cancer. It is pretty horrible. Your body attacks itself, fluid collects in the lungs, it is painful, and you basically drown from within. Your only option is high doses of morphine and one day you just don't wake up. I wouldn't wish that on anyone.
People always talked about how long George Burns survived as a cigar smoker, but luck, genetics and how much you inhale all alter your risk profile. Cigar smokers tend not to inhale, but they still expose their throats and stomachs to all that smoke. Throat cancer sounds terrible.
Camel unfiltered is some sort of statistical anomaly. Hope for those genes.
My old man is over 70 now, lifelong smoker. One of his first jobs in his mid-teens was breaking up asbestos lining in steel moulds/pipes, and he sometimes tells me of the slew of now-banned chemicals his old man would use around the home, e.g. IIRC, some kind of pest detterent powder he would sprinkle handfuls of in the pantry, which was some variety of DDT. It's crazy to me that he's still going; I still expect a call some day, "well, turns out it's got me".
Yeah my mother has been smoking since Kennedy was president, and she's still going. I do think that even without cancer or emphysema, the quality of her life is poor. My wife plays sports with ladies older than my mom and it's a bummer to see mine unable to do much.
My mom died from lung cancer at 50. Horrible horrible way to go. My dad is 79 and (thankfully) going reasonably well. Bit started smoking in early early teens ;-(
Tangential, but if you haven't had your house tested for radon, I strongly recommend it. Radon is a radioactive decay product of uranium in rock and soil, naturally accumulates in houses, and is the second leading cause of lung cancer in the US after smoking.
I got a RadonEye meter to check the levels in my old place. It was around 4 pCi/l, which is apparently equivalent to smoking a pack a day in terms of relative cancer risk. After adding some ventilation, it dropped to lower than 1. Great.
The next place I moved into had a radon level of 14+ (!!) pCi/l. I couldn't believe the meter reading, so double-checked with another meter, which produced the same result. Got a radon mitigation system installed.
Seriously, check it out, especially if your place sits below grade or you live in a risk area. It could save your life.
Identical situation. In the hills of LA, bought a house, tested at 14+. Down below 2 most of the time after mitigation. Found out the hills of LA are one of the radon hotspots in California. I talked to many real estate agents, contractors, and neighbors and no one is aware of it, no one took me seriously, and generally no one wants to hear about it.
Definitely recommend. We were at around 14 (pCi/l) as well. Peaks up to 20. Got an active mitigation system installed last year for around $1k. Hasn’t exceeded 2 since.
Unforeseen benefit of no longer having to run my dehumidifier this summer. So it’ll end up paying for itself in electricity cost after maybe a decade hah.
The author is a PhD in CS from Berkeley and teaches data science at a (decent I think?) college. And yet, he writes completely incorrect statements about statistics. Why?
> "(1) that the supposed effect is really the cause, or in this case that incipient cancer, or a pre-cancerous condition with chronic inflammation, is a factor in inducing the smoking of cigarettes, or (2) that cigarette smoking and lung cancer, though not mutually causative, are both influenced by a common cause, in this
case the individual genotype."
> If you think that’s the stupidest thing you’ve ever heard, you can take it up with Sir Ronald Fisher, who actually made this argument with apparent sincerity in a 1957 letter to the British Medical Journal. I mention this in case you didn’t already know what an ass he was.
Of course, it's very possible that there's a common cause. For example, people who are very depressed may choose to smoke, and depression may also cause cancer. I'm not saying it does, I'm just saying it's clearly a possibility that needs experiments to disprove.
The author's claim is so wrong, and yet he seems to genuinely believe it.
It also seems unprofessional to call Sir Fisher an "ass", which is called an ad-hominem attack (attack on the person is obviously not a valid logical argument), and one of a dead person at that.
Fact is: Fisher made fundamental contributions and the poster himself uses these on a daily basis: "Student" [Gossett]'s t-distribution, Fisher's linear discriminant, a precursor of linear discriminant analysis, ANOVA, Fisher information, Fisher's scoring, Minimum Fisher information, the F-distribution, Fisher–Tippett–Gnedenko theorem, the Fisher-Yates shuffle algorithm, the von Mises–Fisher distribution, inverse probability, Fisher's permutation test Fisher's inequality, sufficiency of a statistic, and more. His contribution to biology are even more impactful.
It also happens that Fisher was a proponent of the "Frequentist" interpretation of probabilities, whereas the poster belongs to the recently more popular "Bayesian" camp of ideology (yes, ideology, if you think statistics as a part of mathematics is free from that, welcome to real life!).
He further held strong views about races, perhaps because he was a professor of eugenics (which we today know as incorrect unscientific nonsense, but obviously that was not the state at the time), and there is am ongoing controversy regarding whether he was a racist or not. What I read about him suggests to me I may not have liked him, but to just stamp him as "ass" wihtout elaborating is too cheap for a statistcs professional in my view. But then this is a personal blog, where the poster can write whatever he wants under US law.
"Sometimes interpreting data is easy. For example, one of the reasons we know that smoking causes lung cancer is that when only -6dB of the population smoked, +6dB of people with lung cancer were smokers. If you are a doctor who treats patients with lung cancer, it does not take long to notice numbers like that."
What was the change? +12dB or 10^1.2, which is 16x.
This ties into the odds form of Bayes' theorem given in the post; this representation of probability is
After that, a number of well-designed experiments on rodents demonstrated that smoking causes (strong causation) cancer (in rodents) beyond a reasonable doubt (as in, if you disagree, you'd better bring some impeccable data).
"...the pre-cancerous condition with chronic inflammation, is a factor in inducing the smoking of cigarettes,"
The author mocks this possibility, but the causes of chronic inflammation are not entirely understood, and it is linked with depression, cancer, obesity, stress, and a lot of other human conditions. We can't rule out that a stressed out person might want to smoke, be more likely to get addicted, and also have chronic inflammation which opens the door to the development of uncontrolled malignancies.
"Finally, looking at that figure you might wonder why the relative risk of smoking has increased so much. Based on my first pass through the literature, it seems like no one knows. There are at least three possibilities:
- Over this period, cigarettes have been reformulated in ways that might make them more dangerous.
- As the prevalence of smoking has decreased, it’s possible that the number of casual smokers has decreased more quickly, leaving a higher percentage of heavy smokers.
- Or maybe the denominator of the ratio — the risk for non-smokers — has decreased."
My first thought was "second hand smoke". My logic is, when smoking was more normalized, non-smokers were exposed to some of the same risk as smokers due to smoking indoors, etc, resulting in higher rates of lung cancer among non-smokers.
I think it has to be said, correlation is not causation. While I believe that in this case it's almost certainly true, it's still not scientific to claim that it is indeed true until you can prove causation.
When you do the math, it turns out that 100% of people who confuse correlation with causation end up dead. Dangerous stuff.
I always found the de-alcoholization of mouth wash a bit interesting. Seemed to happen a little after big tobacco lawsuits. Almost like they know something and worry about it.
As with coca-cola decreasing bottle sizes (before that was the cool thing to do). We went from 591mL to 500mL. And new smaller cans in addition to the 355mL cans (222mL). Huge promotions on Coke Zero. Decreased the sugar content in the regular product. https://www.cbc.ca/news/health/why-coke-is-lowering-its-suga...
[+] [-] pelagicAustral|2 years ago|reply
Needless to say I started smoking again... I then quit again for a few years... went to northern Africa, and guess what? People literally smoking at the airport, £1 packs of camel, etc...
I haven't had one in a bit now... People think the restrictions don't work, but they absolutely do.
[+] [-] jkubicek|2 years ago|reply
It's wild to me that people make assertions like this. We can (and should) argue about the effectiveness of smoking restrictions on decreasing cancer and heart disease, but to make the argument that they have no effect is just silliness.
You hear the same arguments from the right whenever gun control, environmental protections or really any sort of government regulation comes up, and from the left (though to a much lesser degree) on the topic of tax rates. We should all be able to agree that any sort of regulation will have a marginal effect, and be able to have a discussion about the level of that effect.
[+] [-] hankchinaski|2 years ago|reply
[+] [-] ohthatsnotright|2 years ago|reply
No restrictions on sale other than 18+ (I was 24 when I started), and inside public places were only restricted a few years after I came back so I smoked a fair bit in local bars.
Smoking and similar habits are such a wild spectrum you can't prescribe any one thing and restrictions might help some who are prone but aren't really a cure.
[+] [-] thefz|2 years ago|reply
[+] [-] robwwilliams|2 years ago|reply
Not only did Fisher state this, and not only was he an ass, but it is almost embarrassing —it turns out he was partly right. You can read all about it in Pearl’s gem The Book of Why.
In essence Fisher argued that there could be a confound between the neurochemical and behavioral propensity to smoke, and the risk of developing lung cancer.
And strange as it seems to me, the DNA variant in CHRNA5 that is most highly associated with higher risk of nicotine addiction also modulates (independently) lung cancer risk. His principle was kinda-sorta defensible for an avid smoker.
Mendelian randomization of people by their CHRNA5 gene status can indeed demonstrate that Fisher was wrong. The majority causal risk is in fact smoking.
[+] [-] npunt|2 years ago|reply
Sorta a tangent here, but it's annoying that this is the very thing that causes people biased to not believe to latch onto as evidence they're right. Once introduced, further reasoning or curiosity about the subject tends to stop before weighing matters of degree, which leads to missing the big picture.
To generalize a lot, this is kinda where we are in the current meta of truth fracturing. Our biases drive us to discover nuances we grasp onto to avoid believing what we don't want to.
[+] [-] gweinberg|2 years ago|reply
Of course, the link between cancer is smoking is even stronger if some of the "non-smokers" were exposed to an increased level of cigarette smoke.
[+] [-] latchkey|2 years ago|reply
My mom died of lung cancer at 54.
My dad is still going strong at 77 and smokes Camel unfiltered.
Humans are weird.
Regardless, you don't want to die from lung cancer. It is pretty horrible. Your body attacks itself, fluid collects in the lungs, it is painful, and you basically drown from within. Your only option is high doses of morphine and one day you just don't wake up. I wouldn't wish that on anyone.
[+] [-] hinkley|2 years ago|reply
Camel unfiltered is some sort of statistical anomaly. Hope for those genes.
[+] [-] 000ooo000|2 years ago|reply
[+] [-] gedy|2 years ago|reply
[+] [-] mongol|2 years ago|reply
[+] [-] NikolaNovak|2 years ago|reply
My mom died from lung cancer at 50. Horrible horrible way to go. My dad is 79 and (thankfully) going reasonably well. Bit started smoking in early early teens ;-(
[+] [-] jejeyyy77|2 years ago|reply
[+] [-] ternus|2 years ago|reply
I got a RadonEye meter to check the levels in my old place. It was around 4 pCi/l, which is apparently equivalent to smoking a pack a day in terms of relative cancer risk. After adding some ventilation, it dropped to lower than 1. Great.
The next place I moved into had a radon level of 14+ (!!) pCi/l. I couldn't believe the meter reading, so double-checked with another meter, which produced the same result. Got a radon mitigation system installed.
Seriously, check it out, especially if your place sits below grade or you live in a risk area. It could save your life.
[+] [-] colordrops|2 years ago|reply
[+] [-] evanreichard|2 years ago|reply
Unforeseen benefit of no longer having to run my dehumidifier this summer. So it’ll end up paying for itself in electricity cost after maybe a decade hah.
[+] [-] 000ooo000|2 years ago|reply
[+] [-] _cs2017_|2 years ago|reply
> "(1) that the supposed effect is really the cause, or in this case that incipient cancer, or a pre-cancerous condition with chronic inflammation, is a factor in inducing the smoking of cigarettes, or (2) that cigarette smoking and lung cancer, though not mutually causative, are both influenced by a common cause, in this case the individual genotype."
> If you think that’s the stupidest thing you’ve ever heard, you can take it up with Sir Ronald Fisher, who actually made this argument with apparent sincerity in a 1957 letter to the British Medical Journal. I mention this in case you didn’t already know what an ass he was.
Of course, it's very possible that there's a common cause. For example, people who are very depressed may choose to smoke, and depression may also cause cancer. I'm not saying it does, I'm just saying it's clearly a possibility that needs experiments to disprove.
The author's claim is so wrong, and yet he seems to genuinely believe it.
[+] [-] jll29|2 years ago|reply
Fact is: Fisher made fundamental contributions and the poster himself uses these on a daily basis: "Student" [Gossett]'s t-distribution, Fisher's linear discriminant, a precursor of linear discriminant analysis, ANOVA, Fisher information, Fisher's scoring, Minimum Fisher information, the F-distribution, Fisher–Tippett–Gnedenko theorem, the Fisher-Yates shuffle algorithm, the von Mises–Fisher distribution, inverse probability, Fisher's permutation test Fisher's inequality, sufficiency of a statistic, and more. His contribution to biology are even more impactful.
It also happens that Fisher was a proponent of the "Frequentist" interpretation of probabilities, whereas the poster belongs to the recently more popular "Bayesian" camp of ideology (yes, ideology, if you think statistics as a part of mathematics is free from that, welcome to real life!).
He further held strong views about races, perhaps because he was a professor of eugenics (which we today know as incorrect unscientific nonsense, but obviously that was not the state at the time), and there is am ongoing controversy regarding whether he was a racist or not. What I read about him suggests to me I may not have liked him, but to just stamp him as "ass" wihtout elaborating is too cheap for a statistcs professional in my view. But then this is a personal blog, where the poster can write whatever he wants under US law.
[+] [-] YokoZar|2 years ago|reply
We have mountains of direct evidence with underlining theory for the straightforward explanation. Sometimes we need to believe that evidence.
[+] [-] crdrost|2 years ago|reply
"Sometimes interpreting data is easy. For example, one of the reasons we know that smoking causes lung cancer is that when only -6dB of the population smoked, +6dB of people with lung cancer were smokers. If you are a doctor who treats patients with lung cancer, it does not take long to notice numbers like that."
What was the change? +12dB or 10^1.2, which is 16x.
This ties into the odds form of Bayes' theorem given in the post; this representation of probability is
so roughly 0dB = 50%, 5dB = 75%, 10dB = 90%, 20dB = 99%, 30dB = 99.9%...[+] [-] aldarion|2 years ago|reply
[+] [-] dekhn|2 years ago|reply
https://pubmed.ncbi.nlm.nih.gov/2000852/
After that, a number of well-designed experiments on rodents demonstrated that smoking causes (strong causation) cancer (in rodents) beyond a reasonable doubt (as in, if you disagree, you'd better bring some impeccable data).
[+] [-] bentt|2 years ago|reply
The author mocks this possibility, but the causes of chronic inflammation are not entirely understood, and it is linked with depression, cancer, obesity, stress, and a lot of other human conditions. We can't rule out that a stressed out person might want to smoke, be more likely to get addicted, and also have chronic inflammation which opens the door to the development of uncontrolled malignancies.
Still, don't smoke.
[+] [-] yread|2 years ago|reply
https://www.pnas.org/doi/10.1073/pnas.1718185115
Smoking actually kills more people by causing heart disease than by causing cancer. But the risk isn't that much increased.
[+] [-] neonate|2 years ago|reply
[+] [-] seccess|2 years ago|reply
- Over this period, cigarettes have been reformulated in ways that might make them more dangerous.
- As the prevalence of smoking has decreased, it’s possible that the number of casual smokers has decreased more quickly, leaving a higher percentage of heavy smokers.
- Or maybe the denominator of the ratio — the risk for non-smokers — has decreased."
My first thought was "second hand smoke". My logic is, when smoking was more normalized, non-smokers were exposed to some of the same risk as smokers due to smoking indoors, etc, resulting in higher rates of lung cancer among non-smokers.
[+] [-] andrewinardeer|2 years ago|reply
[+] [-] bottled_poe|2 years ago|reply
[+] [-] Zaskoda|2 years ago|reply
[+] [-] Zaskoda|2 years ago|reply
When you do the math, it turns out that 100% of people who confuse correlation with causation end up dead. Dangerous stuff.
[+] [-] WolfeReader|2 years ago|reply
https://www.cancerresearchuk.org/about-cancer/causes-of-canc...
[+] [-] Scoundreller|2 years ago|reply
As with coca-cola decreasing bottle sizes (before that was the cool thing to do). We went from 591mL to 500mL. And new smaller cans in addition to the 355mL cans (222mL). Huge promotions on Coke Zero. Decreased the sugar content in the regular product. https://www.cbc.ca/news/health/why-coke-is-lowering-its-suga...
[+] [-] hn72774|2 years ago|reply
[+] [-] ekianjo|2 years ago|reply
[+] [-] tensor|2 years ago|reply