Would You Believe 40 Percent?

Remember that unpublished study claiming that a smoking ban in Helena, Montana, cut heart attacks in half? It has now been published—in the British Medical Journal, no less. Along the way, the authors' confident assertions have become much more cautious.

Last year Stanton Glantz, the California anti-smoking activist who helped prepare the study, said it showed that "banning secondhand smoke…immediately starts saving lives." Richard Sargent, a local physician and smoking ban proponent who was the lead author, said Helena's ordinance "led to an immediate and dramatic decline in the number of heart attacks."

By contrast, the published study concludes that "laws to enforce smoke-free workplaces and public places may be associated with an effect on morbidity from heart disease." Sargent et al. cite "Helena's small size" (and the correspondingly tiny numbers of heart attacks seen there each month) as "an important limitation of the study." They note that a before-and-after study such as this one cannot prove causation, since "there is always the chance that the change we observed was due to some unobserved confounding variable or systematic bias." And they concede that "we did not make any direct observations to measure how much the exposure to secondhand smoke was reduced during the months when the law was enforced."

Another striking change: The 60 percent drop that Sargent et al. attributed to the ban last year has become a 40 percent drop—a number that is still preposterously large, even if you accept the anti-smoking movement's claims about the effects of secondhand smoke. As I noted last year, the American Heart Association blames secondhand smoke for about 5 percent of heart disease deaths.

"The effect associated with the smoke-free law may seem large," Sargent et al. admit. But they claim it is consistent with epidemiological research indicating that exposure to secondhand smoke is associated with a 30 percent increase in heart attack risk (a number that is itself implausibly high, since smoking, which involves much higher levels of exposure, is associated with a heart disease risk increase of about 70 percent). "If all this effect were to occur immediately," Sargent et al. write, "we would expect a fall of –.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 –32.23 to –0.8 admissions)."

Here Sargent et al. make a calculation error that someone at the BMJ should have caught. If secondhand smoke increased the incidence of heart attacks by 30 percent, and if this effect could be reversed immediately, and if Helena's ban completely eliminated secondhand smoke (which it didn't, since it did not apply to private residences), the drop would be 23 percent, not 30 percent, meaning nine fewer heart attacks rather than the 12 that Sargent calculate or the 16 that they report. Notice, too, how wide the confidence interval is, ranging from less than one to more than 32, which reflects the small sample size.

Hedging their bets, Sargent et al. note that a smoking ban not only reduces exposure to secondhand smoke; it also encourages smokers to quit or cut back. "While both of these effects are probably occurring," they say, "we do not have a large enough sample size to estimate their relative contribution to our results." In other words, the reduction in heart attacks could have occurred partly or entirely among people who stopped smoking because of the ban, in which case it might have nothing to do with secondhand smoke—a pretty important concession given the way anti-tobacco activists (including Glantz's friends at Americans for Nonsmokers' Rights) are using the study to push for more smoking bans.

But even if Helena's ordinance caused every smoker in town to quit (which it assuredly didn't), it still could not possibly be responsible for a 40 percent drop in heart attacks, unless smoking's contribution to heart disease is twice what the CDC says it is. An immediate 40 percent drop in heart attacks across the whole population also implies a much more dramatic effect than is seen in studies of people who quit smoking.

Sargent et al.'s final defense is that the numbers are unimportant. Co-author Robert Shepard, another local physician and smoking ban advocate, told the Helena Independent Record: "I've said all along that it doesn't matter whether the drop is 60 or 80 or 40 percent. The important thing is, the drop isn't zero." Sargent added: "What's the right answer? I don't know. I can tell you with confidence that it's between 24 and 87 percent."

It's a mystery where those numbers came from. But does Sargent really think it's possible that secondhand smoke, smoking, or both together account for something like nine out of 10 heart attacks? "It will take four or five studies to figure it out," he said, "and they will have to be done in communities bigger than Helena." They will also have to be done by researchers who are interested in finding the truth, as opposed to ammunition for their cause.