Does second-hand smoke really cause cancer?

SHARE Does second-hand smoke really cause cancer?

Dear Cecil: Having recently debunked the Super Bowl Sunday violence story, perhaps you could check into this secondhand smoke business. I seem to remember that after the initial study came out blaming secondhand cigarette smoke for every kind of ill, this study was found to be seriously flawed. Is this another case like the “LSD causes chromosome damage” study? Rick Remaley, Chicago


Illustration by Slug Signorino

Cecil replies:

Sure, what the hell, why not insinuate myself into yet another hot-button topic? Then I’ll be ready to take on gun control, abortion, and which are smarter, cats or dogs.

Let me begin by saying that I’m allergic to tobacco smoke, and laws against smoking in public places have personally benefited me. In principle I don’t have a problem with banning public smoking: it’s an annoyance to nonsmokers and a danger to vulnerable folk such as asthmatics, children, and the elderly. All that having been said, the claim that “environmental tobacco smoke” (ETS) seriously threatens the health of the general public, and in particular that it causes lung cancer, is unproven at best.

There have been scores of studies on the health effects of ETS, but the one you’re probably thinking of was a 1993 report by the U.S. Environmental Protection Agency, which labeled ETS a class A carcinogen that caused approximately 3,000 lung cancer deaths among adult nonsmokers per year.

The EPA report put secondhand smoke on the political front burner once and for all. Countless jurisdictions relied on it when they banned public smoking. Today the U.S. has probably the most stringent regulation of public tobacco use of any major nation.

The tobacco industry and its allies were quick to attack the EPA report as “junk science” and filed suit to have it vacated. They won an important victory in 1998 when a North Carolina federal judge ruled that the EPA had made serious procedural errors and, worse, had “cherry-picked” its data to reach a preordained conclusion. The EPA has denied this charge and is appealing the decision.

The controversy over ETS and the EPA report has been marked by accusations of conspiracy, bias, and cooked data, so one has to tread carefully. Nonetheless, a few tentative conclusions can be drawn. The first is that under the most charitable interpretation the EPA’s evidence that ETS is carcinogenic comes perilously close to noise level — you’re not sure if you’re seeing a real effect or just random spikes in the data. The EPA report was based not on original research but on a “meta-analysis” of 11 existing studies; the analysis purported to show that ETS caused a 19 percent increased risk of lung cancer. While this seems like a respectably large number, remember it comes from an epidemiological study, which attempts to infer causality based on associations in the data — circumstantial evidence rather than a smoking gun. Whatever song and dance you may get from the statisticians, skeptical observers prefer to see an increased risk of at least 100 percent before they consider a relationship to be established beyond reasonable doubt.

The tobacco industry claims the EPA had to fudge the numbers just to arrive at 19 percent. For example, in calculating the probabilities, the agency used a “confidence interval” of 90 percent rather than the more stringent (and in my observation more common) 95 percent. The lower the standard, the more statistically significant your results can be made to seem.

Tobacco defenders claim that of four major ETS studies completed since the EPA report was released, two found no evidence that ETS increased cancer risk, one found weak evidence, and only one found strong evidence. The EPA’s take on it is that all four studies support its position. Sounds like bluff to me, but read the agency’s response and decide for yourself at

Smoking opponents say there’s a scientific consensus in the U.S. that ETS is bad, citing an impressive list of articles and official pronouncements — for example, a 1998 review in the Journal of the American Medical Association of 100 studies, 63 of which found some evidence of harm from ETS. I agree ETS is harmful, broadly speaking; the question is whether it causes lung cancer and other significant health problems, as the EPA claims. For years the tobacco industry denied any link between active smoking and lung cancer in the face of overwhelming evidence to the contrary, so you have to wonder when they make the same claim now about passive smoke. Nonetheless, one can’t escape the suspicion that this time the weasels may be right.

Cecil Adams

Send questions to Cecil via