Heart Attacks Rise After Smoking Bans As Often As They Fall

Jacob Sullum | May 1, 2009

A recent national study of the relationship between smoking bans and trends in hospital admissions and mortality should (but probably won't) put a stop to claims that such laws lead to immediate, dramatic reductions in heart attacks. The study, published by the National Bureau of Economic Research (NBER), is by far the largest to be done so far. It covers an eight-year period and includes more than 200,000 admissions for acute myocardial infarction (AMI), plus more than 2 million heart attack deaths in 468 counties across the United States. By contrast, the U.S. studies finding big declines in heart attacks following smoking bans have covered short periods (six months to a year), focused on single locations (such as Helena, Montana, and Pueblo, Colorado), and included hundreds of hospital admissions. Here is what the NBER study's authors, led by the RAND Corporation's Kanaka Shetty, found:

Workplace smoking restrictions are unrelated to changes in all‐cause mortality or mortality due to other AMI in all age groups. Restrictions on smoking of any sort are associated with reduced all‐cause mortality among the elderly (‐1.4%, 95% CI: ‐3.0 to 0.2%) but the result is only significant at the 10% level (p=0.06) (see Table 2). We find no statistically significant reduction in admissions due to AMI among working‐age adults (‐4.2%, 95% CI: ‐10.2 to 1.7%, p=0.165) or among the elderly (2.0%, 20 95% CI: ‐3.7 to 7.7%, p =0.48) following the enactment of a workplace smoking restriction (see Table 3). We similarly find no evidence of reduction in admissions for other diseases in any age group...

In contrast with smaller regional studies, we find that workplace bans are not associated with statistically significant short-term declines in mortality or hospital admissions for myocardial infarction or other diseases. An analysis simulating smaller studies using subsamples reveals that large short-term increases in myocardial infarction incidence following a workplace ban are as common as the large decreases reported in the published literature.

I italicized that last part because this is what I've been saying for years: In a big country with lots of jurisdictions, there are bound to be places where heart attacks drop substantially after smoking bans are passed. There will also be places where they remain flat or go up. Unless that first category is bigger, there isn't even prima facie evidence that smoking bans reduce heart attacks. But if you focus only on the places where heart attacks do happen to drop sharply, you can trick people into believing such outcomes are typical. Which is what anti-smoking activists like Stanton Glantz have done, and they've done it pretty successfully, judging by their credulous reception in the press. Shetty et al. put it more politely:

Comparisons of small samples...might have led to atypical findings. It is also possible that comparisons showing increases in cardiovascular events after a smoking ban were not submitted for publication because the results were considered implausible. Hence, the true distribution from single regions would include both increases and decreases in events and a mean close to zero, while the published record would show only decreases in events. Thus, publication bias could plausibly explain why dramatic short‐term public health improvements were seen in prior studies of smoking bans.

Shetty et al. also note that the sharp drops in heart attacks highlighted by the likes of Glantz were never a biologically plausible result of reducing exposure to secondhand smoke:

The mechanism for these tremendous declines in AMI rates reported in the small‐scale studies is unclear...The estimates of risk due to ETS exposure due to public smoking from these small‐scale studies are similar in magnitude to those from studies of intensive household exposure to secondhand smoke...The similarity implies exposure to secondhand smoke presents large heath risks at low levels and no additional health risks at higher levels, which seems unlikely.

I would add that the 30 percent risk increase associated with household exposure in those studies seems implausibly large when compared to the heart disease risk associated with smoking, which delivers much higher levels of toxins than breathing air mixed with secondhand smoke. (Smokers are about 100 percent more likely than nonsmokers to develop heart disease and about 70 percent more likely to die from it.) But even if exposure to secondhand smoke does not increase the risk of heart disease, or if  smoking bans do not affect exposure much, you would expect them to cause a long-term drop in heart disease to the extent that they encourage smokers to cut back or quit. It just wouldn't happen right away, and it wouldn't be dramatic enough to grab people's attention in the way Glantz likes to do.

More on smoking bans and heart attacks here.

[via The Rest of the Story]

Comments

Nikola T|5.1.09 @ 12:52PM|#

Keeping with the mentality of his base, I propose he draw from the following:

Judges,
Joe Brown
Larry Seidlin
Judy Sheindlin
Greg Mathis
Glenda Hatchett

Nikola T|5.1.09 @ 12:53PM|#

Oops wrong thread... How'd that happen?

Damn, cross ME off the list!

|5.1.09 @ 1:20PM|#

"Youse guys like a smoke?"

"No, eh? We want our lungs to be pink when they fry us. Hey, we told 'em we didn't want a lawyer. Chip here probably just kill him anyway. Lawyers are for sucks."

High Every Body|5.1.09 @ 1:21PM|#

Nice article. BRB after I have a couple of Winstons out in the rain.

ed|5.1.09 @ 1:28PM|#

publication bias could plausibly explain why dramatic short‐term public health improvements were seen in prior studies of smoking bans

Lying doesn't carry the social stigma it once did in our culture.

libertarian democrat|5.1.09 @ 1:39PM|#

No, publication bias is generally unrelated to lying.

|5.1.09 @ 1:51PM|#

Stupidity and self-righteousness, yes. Lying, not really.

Simon|5.1.09 @ 2:03PM|#

This is really interesting work, both for what it says about the flawed case on behalf of smoking bans and the process of scientific publication broadly. Thanks very much for sharing.

Guess My Bra Size|5.1.09 @ 2:10PM|#

I wonder just how much tobacco production and smoking impacts global warming?

High Every Body|5.1.09 @ 2:50PM|#

GMBS,

Zero.

No Name Guy|5.1.09 @ 2:55PM|#

Publication bias, small sample sizes, limited length of studies - who would have thunk that the case the anti-smoking zealots would have rested on a tripod of such obvious junk.

(shaking head at the stupidity of the control freak nannies who push this shit on the rest of us.)

Shit, the simple facts without resorting to bullshit (dose response curve, for example) should suffice to convince folks that smoking is probably not in one's long term benefit. Besides, I don't like stinky clothes, so would voluntary choose a place that didn't have smoking.

Oh yeah, and I say this in spite of the fact that I buried my father a few years ago (lung cancer, 2 pack a day for 40 years) and my mother is currently battling cancer (1 pack a day for about the same amount of time).

|10.20.09 @ 7:16PM|#

But the real question is... "Do you smoke" ???

|5.1.09 @ 2:56PM|#

I wonder just how much tobacco production and smoking impacts global warming?

0.0000001% as much as all the "Go Green" webpages.

Kreel Sarloo|5.1.09 @ 3:51PM|#

It strikes me that since the damage done by smoking is long term and cumulative that it should take some time before any meaningful statistics on health benefits due to some societal reduction in smoking would become apparent. Years, maybe even decades, I'm thinking.

And that assumes that smoking bans actually do result in significant reductions in smoking.

|10.20.09 @ 7:18PM|#

I agree with you, it will take time. But we will see the positive results...

EscapedWestOfTheBigMuddy|5.1.09 @ 3:59PM|#

Kreel Sarloo,
Advocates of smoking bans have spent the last N years loudly making media hash of reported studies that show huge (but non-statistical) and immediate (but un-sustained) health benefits from bans.

And they have poo-pooed any suggestion that this data might not be conclusive. And all but alleged that opponents of the bans are baby killers.

All without the slightest suggestion of a mechanism for the reported "effect" or any care for the lack of statistical reliability.

Those of us who know something about the interpretation of data have been waiting for this---and we knew it was coming---so we can beat the putzes over the head with reality.

Not that they'll notice.

Paul|5.1.09 @ 6:33PM|#

Clearly, the bans need to strengthened.

Michael J. McFadden|5.2.09 @ 1:38AM|#

Speaking of such things as publication bias, readers might like looking over a study complete over three years ago that arrived at basically the same conclusions with similar large-scale data as the current RAND study:

http://www.scribd.com/doc/9679507/bmjmanuscript

and an article with the American Council on Science and Health about the publication bias Dave Kuneman and I ran into while trying to get it published:

http://www.acsh.org/factsfears/newsID.990/news_detail.asp

The RAND study uses somewhat more sophisticated statistical procedures than we did, but arrived at the same general conclusion: Helena and its clones are either deliberately selected and designed studies to devote smoking bans or they are all amazingly fortuitous "accidents" in having found subpopulations with AMI declines after bans.

I'll refrain from offering my own opinion on that question.

Michael J. McFadden
Author of "Dissecting Antismokers' Brains"

adina |5.2.09 @ 4:48AM|#

None of the findings were statistically significant at the 5% level, which is the bare minimum for publication at even the most mediocre journal. These data don't really add much to our understanding, save for the fact that they prompt further study.

adina |5.2.09 @ 4:49AM|#

I should add that I oppose smoking bans, MIs or no MIs.