January 31, 2007

"The Tyranny of Deception"

50,000 dead. Per year. In the U. S. alone. From second-hand smoke.

For a couple of years now, I've been hearing that number cited left and right as a justification for all manner of public smoking bans or other government limitations on individual freedoms. And every time I hear it I think to myself, "Where did that number come from?" It's way too "round" for one thing, and even at that it just seemed unrealistically high. I doubted it. Not enough to check it out. I just doubted it.

On average, 1000 people per state. Not "harmed", not short of breath, not diseased. Dead. Every year. From cigarettes smoked by other people. Not buyin' it.

So along comes one Gio Batta Gori, with an article in yesterday's Washington Post called The Bogus 'Science' of Secondhand Smoke. Mr. Gori would seem to be a well credentialed commentator on the subject. He is an epidemiologist and toxicologist, and a former deputy director of the National Cancer Institute's Division of Cancer Cause and Prevention. So he's got my attention.

Gori's piece is well worth reading in full, but mainly it serves to show how difficult, if not impossible it is to measure the effects of second-hand smoke, and also how few, and how flawed are the major studies on which the estimates of deaths from second-hand smoke are based. Excerpting liberally, if I may:

Lung cancer and cardiovascular diseases develop at advancing ages. Estimating the risk of those diseases posed by secondhand smoke requires knowing the sum of momentary secondhand smoke doses that nonsmokers have internalized over their lifetimes. Such lifetime summations of instant doses are obviously impossible, because concentrations of secondhand smoke in the air, individual rates of inhalation, and metabolic transformations vary from moment to moment, year after year, location to location.

Typically, the studies asked 60--70 year-old self-declared nonsmokers to recall how many cigarettes, cigars or pipes might have been smoked in their presence during their lifetimes, how thick the smoke might have been in the rooms, whether the windows were open, and similar vagaries. Obtained mostly during brief phone interviews, answers were then recorded as precise measures of lifetime individual exposures.

In reality, it is impossible to summarize accurately from momentary and vague recalls, and with an absurd expectation of precision, the total exposure to secondhand smoke over more than a half-century of a person's lifetime. No measure of cumulative lifetime secondhand smoke exposure was ever possible, so the epidemiologic studies estimated risk based not only on an improper marker of exposure, but also on exposure data that are illusory.

Adding confusion, people with lung cancer or cardiovascular disease are prone to amplify their recall of secondhand smoke exposure. Others will fib about being nonsmokers and will contaminate the results. More than two dozen causes of lung cancer are reported in the professional literature, and over 200 for cardiovascular diseases; their likely intrusions have never been credibly measured and controlled in secondhand smoke studies. Thus, the claimed risks are doubly deceptive because of interferences that could not be calculated and corrected.

In addition, results are not consistently reproducible. The majority of studies do not report a statistically significant change in risk from secondhand smoke exposure, some studies show an increase in risk, and astoundingly, some show a reduction of risk.

Gori says that the U.S. Surgeon General's report from July, 2006 fails to back up with solid data its scary claims that...

"there is no risk-free level of secondhand smoke exposure," that "breathing secondhand smoke for even a short time can damage cells and set the cancer process in motion," and that children exposed to secondhand smoke will "eventually . . . develop cardiovascular disease and cancers over time."

While I am unconvinced that the very first breath of second-hand smoke "set(s) the cancer process in motion", there are clearly health hazards caused by exposure, and for once, invoking "the children" is more than cynical exploitation of emotion for political gain. Kids can't control the home environments they live in, and the threat to them shouldn't be minimized. (For the record, in case anyone suspects I have a dog in this fight...I don't. I smoked my last cigarette when my now 28-year old son was one, and my wife quit when she got pregnant with him.)

One of the problems in the counting of dead people from second-hand smoke, as this article points out, is that we have a lot of numbers, but very few names. The state of Pennsylvania, for example, doesn't even have a category for the purpose of recording a death caused by second-hand smoke, and officials cannot cite one death certificate that mentions second-hand smoke as a primary or even a contributing cause of death.

There are also widely varying results coming from different organizations who have studied the issue. The World Health Organization for one, says "Our results indicate no association between childhood exposure to ETS [environmental tobacco smoke] and lung cancer risk." The WHO study also found there was only "weak evidence" for a risk of lung cancer from spousal or workplace exposure.

The National Cancer Institute says 3,000 people die annually from the effects of second-hand smoke. Being the Cancer Institute, they probably concern themselves only with lung cancer deaths. That would be backed up by the Centers for Disease Control, whose estimates range from 38,000 to 60,000 a year total (you know...give or take 22,000), but says only 3000 of those are deaths from lung cancer. But...say CDC officials

"...the cardiovascular effects are about 15 times larger. We estimate at a minimum 35,000 to as high as 50,000 to 60,000 people die each year from heart attacks and strokes after being exposed to secondhand smoke."

Notice the shaky "post hoc, ergo propter hoc" argument there. They say these people died after being exposed to second-hand smoke, which of course does not mean they died as a result of being exposed to second-hand smoke. As Gori's article states, "more than two dozen causes of lung cancer are reported in the professional literature, and over 200 for cardiovascular diseases; their likely intrusions have never been credibly measured and controlled in secondhand smoke studies."

Another widely cited article on the flawed second-hand smoke studies is this one by Thomas A. Lambert, an associate professor of law at the University of Missouri School of Law, which also appeared in the WaPo.

In releasing his ETS study this past June, Surgeon General Richard Carmona proclaimed that "there is no risk-free level of exposure to secondhand smoke" and that "even brief exposure to secondhand smoke has immediate adverse effects on the cardiovascular system and increases risk for heart disease and lung cancer."

The problem is, that's not what his study actually concluded. The report was a "meta-analysis" that combined the results of a number of previous studies, and those studies considered only chronic, long-term ETS exposure. The raw data covered in the report thus doesn't address the risk of short-term exposure that is the subject of Carmona's hyperbolic statements -- statements that conflict with the toxicology conventional wisdom that "the dose makes the poison."

Moreover, the report itself, which covered only studies published through 2002, ignored perhaps the largest ETS study ever conducted -- a 2003 study that followed, from 1959 to 1998, the health histories of more than 35,000 never-smoking Californians who were married to smokers. The authors found no "causal relationship between exposure to (ETS) and tobacco-related mortality," though they acknowledged that "a small effect" cannot be ruled out.

One of the calculations to be found in the Surgeon General's report is this:

Exposure of nonsmokers to tobacco smoke increases their risk of heart disease and cancer by as much as 30 percent.

Blogger Craig Westover works with that statistic a little bit, starting with the study's definition of "exposure":

One can arrive at a 30 percent increase risk from two studies that tracked the incidence of lung cancer in non-smoking spouses of smoking spouses and non-smoking employees in smoking environments. What these two studies show is that the longer the duration of exposure to secondhand smoke, the greater the risk of lung cancer. But here’s the kicker for a policy maker -- in the case of souses, the required duration of exposure before there was a correlation greater than 1.0 (1.0 means no connection) was after 29 years of exposure. In the case of an employee in a smoking establishment, the required exposure to exceed a risk of 1.0 was 21 years.

At 29 and 21 years respectively, the 95 percent confidence interval for the increased risk ranged from 1.03 to 1.50 -- an increased risk of 3 percent to 50 percent. The 30 percent figure in the article is pretty close to the mean of that distribution. What the full confidence interval indicates is that for these studies, there is as much likelihood that the real increased risk of secondhand smoke is 3 percent as it is 50 percent as it is 30 percent. In other words, to state there is 30 percent increased risk implies preciseness not supported by the data.

Let’s take it the next step. Fifty-percent, even 30 percent sounds pretty scary. But what does that mean in terms of real numbers. The rate of lung cancer among non-smoking individuals is about 10 in 100,000 people. If you’re a non-smoker, that’s your odds of getting lung cancer. If your spouse smokes, and you’ve been married 29 years or more, and we use the government’s 30 percent figure, your odds are now about 13 in 100,000. If you’re a non-smoker and you’ve worked in smoking environment for 21 years, you odds of getting lung cancer are about 13 in 100,000.

Not exactly cause for circling the wagons just yet if you're the spouse or employee of a smoker. The point is that the use of the scary "30% increase" statistic is deceptive and, one has to assume, intentionally so.

Going back to the Gori article on the issue of using misleading or inflated data as an "end justifies the means" tactic...

Some prominent anti-smokers have been quietly forthcoming on what "the science" does and does not show. Asked to quantify secondhand smoke risks at a 2006 hearing at the UK House of Lords, Oxford epidemiologist Sir Richard Peto, a leader of the secondhand smoke crusade replied, "I am sorry not to be more helpful; you want numbers and I could give you numbers..., but what does one make of them? ...These hazards cannot be directly measured."

It has been fashionable to ignore the weakness of "the science" on secondhand smoke, perhaps in the belief that claiming "the science is settled" will lead to policies and public attitudes that will reduce the prevalence of smoking. But such a Faustian bargain is an ominous precedent in public health and political ethics. Consider how minimally such policies as smoking bans in bars and restaurants really reduce the prevalence of smoking, and yet how odious and socially unfair such prohibitions are.

By any sensible account, the anachronism of tobacco use should eventually vanish in an advancing civilization. Why must we promote this process under the tyranny of deception?

Nothing I can add to that. Thanks for hanging in this long.

Posted by dan at January 31, 2007 11:38 PM