REFERENCE -- Full Text of Edmund Contoski letter to Rep. Meslow on the merits of secondhand smoke scientific studies

Nov. 21, 2004


I’m a retired environmental consultant. I was Director of Planning for an internationally renowned environmental consulting firm doing business in more than forty countries. It’s understandable that you, as well as the general public, would regard environmental tobacco smoke (ETS—secondhand smoke) as a health hazard. After all, that’s what you’ve been told over and over. The Star Tribune and Pioneer Press won’t print any scientific evidence challenging their editorial positions on this. I’ve written nine or ten letters to them on this, citing scientific facts and studies, but they wouldn’t print one word of them. So I took out a full-page ad in a local newspaper to bring some of these truths to light (see enclosed).

The newspapers and anti-smoking zealots claim that there is “overwhelming evidence” that ETS is a dangerous carcinogen. But the independent health research firm Littlewood & Fennel surveyed all available studies of ETS. In their report to the National Toxicology Program’s Board of Scientific Counselors on Carcinogens, they reported that the overwhelming majority (over 75 percent) showed no association between ETS and lung cancer. They said the remaining studies (23%) showed “weak” statistical associations and included “substantial problems with bias, misrepresentation of relative risks and unacceptable epidemiological methodology.” Significantly, L&F state: “We have no affiliation with either industry or government. Indeed, this was a labor of love for sound science and intelligent public policy. Specifically: We own no stock or stock options or other financial instruments of [affected] companies…have not been paid consultants to either industry or government in this matter…have received no fee or honorarium for these comments, nor for conducting research related to this material.”

The pressure for smoking bans appears to reflect what L&F call the effort of “avowed anti-smoking advocates determined to somehow prove that ETS is a human carcinogen in the face of irrefutable evidence to the contrary.” To such people, the truth doesn’t matter. They are determined to get their way, even if it means deliberately promoting false or misleading information in order to provide cover for anti-smoking legislation. Unfortunately, they have been fairly successful. If you are not one of those willfully deceiving the public on this issue, then you are a victim of those who are—or else you have some special scientific expertise that qualifies you to contradict the findings of distinguished scientific health researchers and to disregard long established scientific standards.

When EPA declared ETS a carcinogen, it violated its own Risk Assessment Guidelines. It also violated the standards of the National Cancer Institute and the Paris-based International Agency for Research on Cancer. It violated accepted statistical standards (see Breslow, Day Statistical Methods in Cancer Research, Vol. 1, p. 38.) It also failed to meet any of the nine Bradford-Hill criteria for assessing biological and epidemiological data. Zero for nine! What special scientific expertise do you possess that qualifies you to disregard these established scientific standards and persist in treating ETS as a carcinogen? Have you even heard of the Bradford-Hill criteria? Have you even seen EPA’s Risk Assessment Guidelines or know that EPA violated them? If ETS is as dangerous as we have all been told, wouldn’t that be proved by meeting these scientific standards rather than violating them?

Dr. Alvan Feinstein, who was on the faculty of the Yale University Medical School for 30 years, has been called the “Father of Clinical Epidemiology.” Professor Melvin W. First is Professor Emeritus of Environmental Health at Harvard University. Dr. Gary Becker is a Nobel Prize winner. What scientific credentials do you possess that qualify you to disregard and contradict the research and conclusions of these distinguished scholars that ETS is not a carcinogen?

Then there’s the General Accounting Office. It has no medical expertise, but it is extremely well qualified in statistical studies. And it has been severely critical of the statistical manipulations that EPA engaged in to try to make ETS a carcinogen. Are you better qualified in statistical matters than the GAO?

Also, U.S. District Judge William Osteen found that “EPA cannot show a statistically significant association between ETS and lung cancer.” He continued: “In conducting the ETS Risk Assessment, EPA disregarded information and made findings on selective information, did not disseminate significant epidemiological information; deviated from its Risk Assessment Guidelines; [and] failed to disclose important findings and reasoning….” That’s exactly what those favoring smoking bans are doing! Have you bothered to read Judge Osteen’s ruling? Before you pass a smoking ban, you ought to do so. It does not matter that Judge Osteen is not a scientific expert, because he hires verifiably independent scientific experts to advise him. Maybe you legislators should do the same if you do not have the scientific expertise that’s available to the U.S. Fifth District Court and choose to contradict that court’s findings.*

“EPA disregarded information and made findings on selective information,” said the court. Here’s an example. There were fourteen studies available on ETS in the workplace. Thirteen of these showed no association between lung cancer and ETS. EPA disregarded those thirteen and based its decision on workplace danger on a single study, the Fontham study, which showed a risk ratio of 1.29—a number still far too small to be a risk by accepted standards. But that being all EPA could come up with, EPA claimed that RR to be significant. Now, however, it turns out that the Fontham study was flawed in its interpretation of the data, and when subsequent researchers made the necessary correction, the result was a risk ratio of 1.0. Precisely no effect.

How do we know whether a health risk is high or low? There must be some standard of measurement. What is it? The scientifically accepted standard for quantifying a health danger in epidemiological studies is risk ratio (RR), also known as relative risk. But the propaganda about the “danger” of ETS is strangely silent about RRs. Why? Because the RRs for ETS are so low they do not qualify for serious consideration according to accepted standards. Despite using its invalid statistical manipulations, EPA could come up with an RR of only 1.19 for ETS. That’s far below the standards of the National Cancer Institute or the International Agency for Research on Cancer.

Dr. Eugenia Calle, director of analytic epidemiology for the American Cancer Society, says that RRs below 1.3 cannot even be reliably identified. When a study showed an RR of 1.5 between abortion and breast cancer, Dr. Calle was quick to state that a RR of 1.5 is too low to call abortion a risk factor for cancer. So how can RRs below that be called significant for ETS? Now you know why all the proponents of the smoking ban refuse to discuss risk ratios and simply claim ETS is “dangerous.”

They also try to talk about the percentage increase in RR without mentioning the RR, in order to make it sound scarier. For example, they may say that there is a 50% increase in the risk, without mentioning that this means a risk ratio of 1.5, which is still far below the level of meaningful risk, as Dr. Calle noted. Proponents of smoking bans talk about “hundreds” of studies—but they don’t reveal what the RRs are for those studies or that those numbers show the opposite of what they claim. The RRs for those “hundreds” of studies show the absence of danger from ETS, as does EPA’s RR of 1.19 for ETS.

RR of 1.0 is neutral—no effect—while smaller numbers indicate a negative association, that is, a health benefit. Surprisingly, a large number of studies indicate a health benefit from ETS, particularly for children. EPA deliberately excluded any research studies that resulted in RRs below 1.0. They did so for two reasons: (1) including such studies would further lower their 1.19 RR for ETS, which was already unacceptably low by scientific standards, and (2) it would undermine EPA’s claim that there was no safe level, no “threshold,” for ETS.

Obviously, if they recognized any benefit from low levels of ETS, their claim of danger from even the smallest amount of ETS would go out the window. Thus EPA refused, for example, to include the Wu-Williams/Blot study—which was a very large study—because it showed a very significant negative association between lung cancer and ETS. By setting aside this and many other studies whose results EPA didn’t like, it was left with basing its conclusions on only eleven studies, the majority of which consisted of 40 or fewer people.

To attempt to beef up its numbers, EPA then engaged in a “meta-analysis,” a method of combining separate studies in order to constitute a larger sample. However, EPA did not follow the accepted statistical procedures here. It did not use all the available data (leaving out studies it didn’t like), and it combined studies structured in different ways, resulting in a statistically invalid combination of “apples and oranges.” With such a contrived statistical procedure, it managed to reach an RR of 1.19 for its meta-analysis, but it could not reach the required 95% confidence level (CL) for this.

This recognized statistical standard means that the likelihood of the results occurring by mere chance is no more than five percent. So EPA lowered the CL standard to 90% instead of 95%, doubling the chance of being wrong. EPA never did this before or since. So why should anyone consider this degradation of statistical standards valid for ETS? Significantly, in the seven years following EPA’s meta-analysis, five valid meta-analyses were performed by others. These show no association between lung cancer and workplace exposure to ETS (combined RRs from 0.98 to 1.04, with an average of 1.01 and a 95% confidence level.)

True scientists concede their hypotheses are wrong when contrary to data. Political activist/advocates dismiss the data in order to hang onto their scientifically unsupportable political agenda based on a preordained “conclusion.” They start with a “conclusion” and then work backward attempting to justify it. When that can’t be done honestly, they substitute misrepresentation and willful manipulation for the tenets of sound science. Politicized science is no science at all. Claiming ETS is a carcinogen is now “politically correct,” but it is clearly politicized science.

Unhappily, the once respected American Cancer Society has now come down on the side of politicized science. It now supports smoking bans even though it has sponsored at least two studies in the U.S. showing no association between lung cancer and ETS. One of these may well be the largest study ever done on the subject, comprising 375,000 people. (And that’s not even counting the Enstrom/Kabat study, which the American Cancer Society helped set up, supplied data for, and financed—until preliminary results indicated the opposite of what the ACS wanted. The ACS then discontinued its funding and denounced the study, which covered 100,000 Californians over 38 years. Other funding was found, and the study concluded: “The results do not support a causal relation between environmental tobacco smoke and tobacco related mortality.”) I can find no ACS study with a positive association for ETS and cancer.

You probably saw ads in the Twin Cities newspapers, sponsored in part by the ACS, showing a bar waitress holding a tray and the message “69 carcinogens.” That may be true, but it is deliberately misleading with the intent to scare people. It is misleading because a single cup of coffee contains 100 carcinogens. And drinking three cups of coffee per week exposes you to greater danger of cancer than pack-a-day secondhand smoke for 40 years!

EPA has claimed that ETS accounts for 3,000 deaths annually. This is a computer-generated figure that has been thoroughly discredited. I would be glad to explain at some future time the faulty assumptions behind this absurd figure. For now, let’s just note that both Hennepin and Ramsey County Boards chose to use this discredited figure rather than a more accurate and up-to-date estimate from the Congressional Research Service. The CRS is a branch of the Library of Congress. It is non-partisan, highly respected, and is not beholden to the tobacco industry. The CRS—looking at the same data as the EPA, in fact, in reviewing EPA’s study (at the request of Congress)—concluded: “It is possible that very few or even no deaths can be attributed to ETS.” (see page 55 of CRS report.) But the two counties wanted political cover for a smoking ban rather than the truth. So they used the discredited figure of 3,000 deaths. The resolutions passed by both counties also claim that ETS is a Group A Carcinogen—without telling people that a federal judge 6 years ago rescinded that classification for ETS.

I was at the public hearing in Bloomington. There was not a single bit of scientific health evidence presented by anyone who testified in favor of the smoking ban. A number of doctors testified, and their testimony consisted of saying, in essence, “I’m a doctor, and I’m in favor of the smoking ban.” That is not proof of anything. That is opinion, not evidence. No doctor cited a single epidemiological study or a single risk ratio. No doctor could state he had treated even a single patient from ETS. No one presented or even claimed to have seen a single death certificate saying anyone had died from a disease caused by ETS. The so-called Bloomington Advisory Board of Health Report was a farce. It contained not a single RR and instead consisted mainly of such things as studies about economic impacts on bars and restaurants and public opinion polls of citizens of Bloomington showing popular support for the ban. Those subjects are emphatically NOT health studies.

I hope you will be more honest on this issue than Hennepin and Ramsey Counties and the cities of Bloomington, Minneapolis and St. Paul. This is an issue in which neither they nor you nor other units of government should be involved. It’s a bogus argument that health issues trump property rights, because this is not a health issue. Phony health issues do not—or at least should not—trump individual rights, specifically property rights. That is the real issue here. Bar and restaurant owners should have the right to establish the conditions for business on their property, and customers should then be free to patronize them or not. There is no need for government intervention here at all. Government is supposed to protect people’s rights, not take them away under phony pretexts, such as health “dangers” that are blatantly counter to abundant scientific evidence.

Yours truly,
Edmund Contoski

*The U.S. Fifth District Court decision referred to in the letter was overturned on appeal. However, this action was purely a procedural matter that said nothing about the merits of the issues in the case. The appeals court ruled that EPA's decision about ETS being a carcinogen was not a reviewable action and therefore the Fifth District Court had no jurisdiction. That's all the court said; it declined to comment on the science or the merits of EPA's action.