A new article in the journal Tobacco Control details the difficulties in using machines to predict the levels of toxins and carcinogens to which a given person will be exposed when he smokes a particular brand of cigarette. Smoking machines are used to produce the misleadingly precise tar and nicotine "yields" on cigarette packages, which may have little or no relationship to the health hazards associated with a given brand. The main problem (as I've mentioned before) is that people, unlike machines, engage in "compensatory behavior" to achieve the dose of nicotine to which they're accustomed. They adjust the number of puffs and the degree of inhalation to compensate for changes in the nicotine content of the smoke. They also tend to cover the filter ventilation holes that help cigarette manufacturers achieve lower tar and nicotine ratings. The authors of the Tobacco Control article note that the standard method of generating these numbers, which uses puffing parameters that "systematically underestimate smoking behaviour in humans," is "widely recognised to be inadequate." They consider four alternative standards, all of which involve more-intense smoking, and conclude that they would not be significantly more reliable in predicting actual human exposure.
Two other approaches discussed in the article look more promising. One is to measure the amount of a given toxin or carcinogen per milligram of nicotine. Unfortunately, this ratio is not constant for any given brand; it varies with the intensity of smoking, which in turn varies across smokers, depending on the level of nicotine they like. Complicating things further, there are many potentially dangerous chemicals in cigarette smoke, and their levels may move in different directions when cigarette design or smoking behavior changes. Still, for some of these chemicals, there are large differences in levels per milligram of nicotine across brands that probably do translate into differences in actual exposure.
The other approach that tries to take compensatory behavior into account uses a machine protocol designed to achieve a particular dose of nicotine regardless of a cigarette's design. Differences in toxin yields across brands measured by this method presumably would signify differences in actual exposure, at least at a given level of nicotine intake. Such measurements would be far from perfect, but they would be a more reliable guide than the current numbers. Even if a better method of predicting exposure to particular smoke constituents can be found, there remains the question of how changes in levels of specific toxins affect disease risk.
Meanwhile, the uncertainty surrounding these numbers not only casts doubt on the tobacco companies' marketing of "low-yield" cigarettes as a supposedly safer alternative to regular cigarettes. As Michael Siegel notes on his tobacco policy blog, it also highlights the silliness of complaining about a slight increase in the machine-measured nicotine yields of major cigarette brands that may or may not have occurred in recent years. More important, it reveals the irrationality of existing and proposed government regulations. It is the federal government, after all, that requires cigarette companies to advertise the misleading yield numbers. And as Siegel points out, a bill that would give the FDA regulatory authority over tobacco, backed by Philip Morris as well as leading anti-smoking groups, would extend the use of questionable yield numbers and add to their perceived credibility. Under the bill, the FDA would have the authority to set maximum levels for specific components of tobacco smoke, based on the same sort of machine-generated numbers that public health officials and anti-smoking activists roundly condemn as not only inaccurate but fraudulent.
"This whole thing is a huge hoax in the making," Siegel writes. "It has the potential to institutionalize the fraud that the tobacco companies have committed, but to put it into the hands of our own government." He also argues (as I have) that the regulatory regime created by the bill would make it virtually impossible to introduce genuinely safer tobacco products. "Ultimately," he says, "there's only one way that I think even has the potential to be successful in developing safer cigarettes or other tobacco products. And that's to allow the free market system to work. Free market competition could, possibly, result in a race to see which company could come up with safer products." To make that possible, the government would have to stop requiring misleading product comparisons, permit the introduction of new products without pharmaceutical-style regulatory review, and allow companies to make truthful statements about the potential health advantages of those products, holding them liable for fraud but not for the voluntary choices of informed consumers.