Tuesday 15 November 2011

Heart miracles are impossible

King-size cigarette,
pint-size intellect
It's good to see Dr. Carl V. Philips back and blogging over at Ep-ology. In his last two posts he has been discussing the North Carolina heart miracle 'study', which is as bad a piece of advocacy-driven junk science as you will ever see.

In particular, he makes a point which I have tried to made before, which is absolutely fundamental to all the heart miracle studies. The results they report—of heart attacks falling by 17%, 21%, 40% or whatever—are simply impossible.

Let's go along with the "consensus" view that long-term secondhand smoke exposure increases the lifetime risk of heart disease by around 20-30%. Nevermind whether that is a realistic estimate. For good or ill, it is the figure used by the Surgeon General and other authorities, and it is accepted by those who conduct the heart miracle studies.

That being the case, is it plausible that the elimination of secondhand smoke from restaurants, offices and bars could reduce the heart attack rate by 21% (as reported in North Carolina) or 40% (as reported in Helena, Montana)?

It is not.

For one thing, most restaurants, some bars and nearly all offices were non-smoking before the ban. In addition, many non-smokers avoided the few remaining smoky venues before the ban. The vast majority of heart attack cases are elderly and not the kind of people to be out partying in bars, nor indeed working in pubs or waiting tables in restaurants. Furthermore, the amount of secondhand smoke inhaled by this subsection of non-smokers before the ban is minimal compared to the long-term exposure that the 20-30% figure is based on.

As Carl explains...

How many people go from being exposed to restaurant/bar smoke to unexposed as a result of the ban? It is a bit fuzzy to define this since there will be a lot of people whose exposure is reduced, and a spectrum of how much it is reduced. But we can start with the observation that roughly half of everyone had approximately zero such exposure before the ban, never or almost never going out to eat and drink, or avoiding smoking-allowed venues when they did...

Thus, even if you believed that exposure at the level of visiting restaurants and bars causes somewhat more than 20% increase in risk, which is an absurd belief in itself, there is no possible way the effect of the smoking ban could be more than about half of the claimed 21%.

Even if we assume that secondhand smoke does cause heart attacks, smoking bans have so little effect on so few non-smokers (and have no effect at all on the smokers, unless it compels them to quit), that the kind of reductions in the heart attack rate reported by these studies defy both science and common sense. If there is an effect, it is too small to measure and would never show up in population-level statistics. Once that is understood, it is obvious that any studies which claim a dramatic effect on the heart attack rate must be flawed, cherry-picked or distorted. Sure enough, when such studies are examined, they prove to be flawed, cherry-picked and distorted.

We can figure that half of the population was not exposed in the first place, that easily a third of those exposed were smokers, that many of those exposed had very minor and occasional exposure, and that many others that were exposed had only a minor reduction in exposure since most of their exposure was elsewhere. So it seems unlikely that even one-fifth of the population experienced a substantial reduction in exposure, getting the effect down below 1% of the total.

If, to take North Carolina as an example, the smoking ban caused the heart attack rate to drop by 21%—which it unequivocally did not—it follows that smoking in bars, restaurants and offices must have been responsible for a fifth of all heart attacks before the ban.

It is quite possible that thirty years of induced panic about passive smoking has persuaded many people that such diluted tobacco smoke is capable of wreaking such havoc, but the empirical evidence shows that it cannot be so. If it were, the relative risk from secondhand smoke exposure would be far higher than 20-30%. Indeed, secondhand smoke would be responsible for more heart attacks than smoking. It would mean that passive smoking (at work and at home) was the single biggest risk factor for heart attacks. Even the most tobaccophobic hypochondriac surely cannot believe such a thing.

In the case of Stanton Glantz's bar-lowering Helena study (2004), the smoking ban effect was even greater—an astonishing 40%. Again, this implies that smoking in a subsection of private venues was responsible for two-fifths of all heart attacks before the ban—a manifestly risible idea.

Interestingly, Glantz must have known that his findings were inherently implausible because he addressed them in the text of the study itself. His comments tell you much about the man's mathematical illiteracy and, sadly, tell us much about the decline of the peer-review process (the study was published in the prestigious British Medical Journal). He wrote:

The effect associated with the smoke-free law may seem large but is consistent with the observed effects of secondhand smoke on cardiac disease. Secondhand smoke increases the risk of a myocardial infarction by about 30%; if all this effect were to occur immediately, we would expect a fall of - 0.30 x 40.5 = - 12.2 in admissions during the six months the law was in effect, which is within the 95% confidence interval for the estimate of the effect (a drop of - 32.2 to - 0.8 admissions).

His argument here is that secondhand smoke exposure increases the risk of heart disease by 30% and so, "if all this effect were to occur immediately", a smoking ban should reduce the heart attack rate by around 30%. 40% is, he concedes, a little higher than might be expected but it is within the margin of error.

This piece of reasoning is so patently flawed that I still cannot believe it was allowed to be published. Let's leave aside the fanciful idea that the effect of a lifetime's exposure would suddenly be nullified by a smoking ban in non-domestic settings. The key point is that Glantz ignores the fact that secondhand smoke is one of dozens, if not hundreds, of risk factors for heart attacks (or heart disease—he treats them as if they were the same thing). He seems not to comprehend the difference between relative risk and absolute risk. He does not acknowledge that a relative risk which affects a subsample of the nonsmoking population is not going to have a commensurate effect on the entire population. And he implicitly treats secondhand smoke as if it were the sole cause of heart disease. These are staggering schoolboy errors for a man with pretensions of being an epidemiologist (which just goes to show that a degree in mechanical engineering is not always the best grounding for a career in cardiology).

Look at it this way. If using a mobile phone while driving increases your risk of having an accident by 90%, what will be the effect on the number of car crashes in a country that bans the practice?

The answer is that we do not know. There are countless other risk factors for car crashes and so, even if using a mobile phone has a substantial effect on individual risk, the effect at the population level will be too small to measure.

By Glantz's logic, however, the effect of a mobile phone ban will be to reduce the number of car crashes by 90%—because he doesn't understand the basic difference between individual relative risk and absolute risk to the population. How can he be so ignorant? There are, as Carl says, only two possibilities.

Interestingly, it is not entirely clear whether he spouts junk because he has not acquired a modicum of understanding about the science in the field where he has worked for decades, or because he is a sociopath-level liar; I am not entirely sure which is the more charitable interpretation.

Do go read both of Carl's pieces about the North Carolina nonsense:

Unhealthful News 189 - Absurd claims about the effects of smoking place restrictions, North Carolina edition (Part 1)


Unhealthful News 190 - Absurd claims about the effects of smoking place restrictions, North Carolina edition (Part 2)


8 comments:

Anonymous said...

Chris we got another live one,this time the MAYO CLINIC!


Smoking ban cut heart attacks risk in half
Published: Nov. 14, 2011 at 7:37 PM



Read more: http://www.upi.com/Health_News/2011/11/14/Smoking-ban-cut-heart-attacks-risk-in-half/UPI-80441321317442/#ixzz1dkHgem60

Harley

Anonymous said...

Dearieme. Surely people didn't burst into your house smoking cigarettes. If you suffered discomfort at work, then I can sympathise. If you chose to spend your leisure time in other people's smoky houses, smoky pubs or cafes, then I have no sympathy. If a pub allows smoking, people will go there to smoke. You wouldn't say a club had lousy manners because its loud music damaged your hearing. There has never been anything to prevent non smoking businesses opening and many did. Some have been successful and some had to shut through lack of demand. You can't blame that on smokers. I've never understood why non smokers didn't demand more non smoking pubs. We would now all be content. It worked well with cafes. Smokers went in Caffe Nero. People who didn't like smoke went in any of the others. Note that most smokers feel more comfortable if they walk into a place knowing they are not going to upset anyone. There is nothing I would like better than to be able to visit a smoking club - not a place that anyone upset by smoke could accidentally stumble into. I appreciated the completely separate smoking sections in some restaurants in the USA back in the 1980s. I also remember looking to get tables next to those with cigarette packets on them.

Anonymous said...

A quick note on the figures:

If mobile phone use increased accident risk by 90% and if everyone used mobiles while driving then you would have 190 deaths per N population compared to 100 in a similar population that did not use them. Hence banning their use (if 100% enforced) would reduce accidents by a maximum of 90. A reduction of 90/190 = 47%.

Similarly if every single person in Helena was an active smoker (approx 60% increased risk of heart disease supposedly) and everyone quit smoking after the ban then the maximum reduction in heart attacks would be 60/160 = 38%. As you rightly say this also ignores other risk factors so the reduction would be less than that. What's more I've seen evidence that giving up smoking is associated with a 4 fold increase in risk of hypertension (at least for the first year or so) which implies that the heart attack rate would probably rise even in this extreme scenario.

Glad Carl is back. I'll have a proper read later.

Tony

Anonymous said...

Oops, above should read:
'190 crashes per N population'

Michael J. McFadden said...

Chris, you mean you don't believe Stanton Glantz? I am Shocked, simply SHOCKED! ::sigh:: Don't worry. I'll get over it.

A couple of notes: I'd like to add to your Carl reading recommendations with this gem I just discovered on confounding:

http://ep-ology.blogspot.com/2011/03/unhealthful-news-74-controlling-for.html#comment-form

And in terms of mobile phone risks I corresponded a while ago with a Dr. Wilson who had done a study on mobile phone risks. I took his figures, compared them to EPA death-by-lung-cancer risks, and brewed up a Relative Risk estimate for driving with a driver who was text messaging once per hour vs. one who was smoking a cigarette once per hour.

Riding with the texting driver was 13,000 times as dangerous. The risk was roughly 1,300,000% higher. The odd follow up to that was that I assumed Professor Wilson would jump on the chance to make a real impact on teen driving/texting (which was his basic concern) by using the comparison. Nope: he wouldn't even answer my emails once he saw what I'd done with his data! Evidently the kids getting killed by texting weren't that important compared to the War On Smoking. ::sigh::

So are the Antis simply grossly and criminally incompetent, or sociopathic liars? Dunno. I'd ask 'em, but I wouldn't be able to trust their answers!

- MJM

Anonymous said...

Tony,

You left out the never-smokers that would be not exposed to SHS/ETS.

Also, about 85% of heart attacks occur to people over the age of 65 and only about 8% of the current smokers are over the age of 65.

At least from what I have seen in 50 years of smoking, there are very few folks over the age of 65 hanging out in bars and restaurants.

Plus, these people are talking about ER admissions for AMI, they have NO way of knowing the number of AMI's that never made it to the ER or never even tried to go to the ER.

They might as well be talking about bans 'causing' a change in the number of angels that dance on the heads of pins.

Gary K.

Anonymous said...

Gary,

I think we're broadly in agreement here. I simply described an extreme example where even if every one was an active smoker and all of them suddenly gave up you still couldn't reach Glanz's figures.

And yes angels on pinheads is a good metaphor.

Tony

dearieme said...

"Note that most smokers feel more comfortable if they walk into a place knowing they are not going to upset anyone": most smokers behaved as if they didn't give a toss whether they upset anyone. Of course people smoked all over me at my workplace, also without giving a toss. I loathe all the lies about passive smoking, but as for the bans, you chaps didn't half bring them on yourselves.