Evidence For The Layman

Key terms

ETS is the acronym of Environmental Tobacco Smoke, or secondhand smoke.

What is epidemiology?

Epidemiology is the study of disease in the human population. Epidemiological studies use statistics to try to investigate links between diseases and factor which may cause those diseases. Diet, occupation, and family medical histories are examples of factors investigated.

What is a relative risk?

A relative risk is the statistical measure which expresses the findings of an epidemiological study. It measures whether consuming or being exposed to something (eg., drinking whole milk, living or working with a smoker) raises or lowers risk of contracting a particular disease.

What do leading scientists and scientific organizations say about relative risk numbers?

• A relative risk of 1.0 means no increased or decreased risk of disease.

• Relative risk below 1.0 is said to be evidence of a possible reduced risk of disease for consuming or being exposed to something. An example is eating a high fruit diet as determined in a study published in Nutrition and Cancer in 1988.

• A relative risk above 2.0 or 3.0 is said to be evidence of a possible increased risk of disease. An example is eating a diet very high in saturated fat as determined in a study published by the U.S. National Cancer Institute in 1993.

• A relative risk between 1.0 and 2.0 is generally referred to as a " weak association" from which no clear conclusions can be drawn. Examples include drinking 1 to 2 glasses of whole milk per day, eating one biscuit a day, drinking chlorinated water, eating pepper frequently and being exposed to second hand tobacco smoke.

How epidemiology applies to ETS

" In epidemiologic research, relative risks of less than 2 are considered small and are usually difficult to interpret. Such increases may be due to chance, statistical bias, or the effects of confounding factors that are sometimes not evident." ( Press release, U.S. National Cancer Institute, Oct. 1994)

" Small associations below 2.0 may be beyond the limits of reliable epidemiological inference." (J. Peto, 1992, IARC)

" In general, I think its a general opinion, and my opinion , that if a relative risk figure goes below about three, then the significance of that result becomes open to question. The relative risks we are talking about with ETS are well below three, and most of them well below two. If it was anything else but ETS I think people wouldn't bother with it at all. They wouldn't discuss it. The relative risks are far too low." ( Proffessor John Robert Ashford, Managing Director of Exeter Health Information Services, comments on OSHA proposed rule on Indoor Air Quality - 1995)

" Differences in risk of 50% ( Relative risk of 1.5) are small in epidemiological terms and severely challenges our ability to distinguish whether it reflects cause and effects or whether it simpl;y reflects bias."(Lynn Rosenberg, Boston University School of Medicine quoted in Press Release U.S. National Cancer Institute Oct 26,1994)

" Epidemiological studies, in general are probably not able , realistically, to identify with any confidence any relative risks lower than 1.3 ( that is a 30% increase in risk) in that context, the 1.5 [reported relative risk of developing breast cancer after abortion] is a modest elevation compared to some other risk factors that we know cause disease." ( Dr. Eugenia Calle, Director of Analytic Epidemiology, American Cancer Society, Washington Post - Oct 27,1994)

Of approximately 40 studies on second hand smoke and lung cancer, about eighty percent did not report a statistical significant overall increase in risk among non smokers married to or living with smokers. The remaining studies report, at most, weak associations.

"The statistical evidence does not appear to support a conclusion that there are substantial health effects of passive smoking." (Dr. Jane Gravelle, et al, Senior Specialist in Economic Policy, Congressional Research Service, in Testimony before the Subcommittee on Clean Air and Nuclear Regulation Committee on Environment and Public Works U.S. Senate -- May 11, 1994)

"The totality of data on ETS and lung cancer does not support the claim made in the draft EPA report that ETS is responsible for an increased incidence of lung cancer in the United States. There is no scientifically valid basis for conducting a risk assessment on ETS or classifying ETS as a known carcinogen or even probable human cargcinogen." (Dr, W. Gary Flamm, Science Regulatory Services International, Former Director of the Office of Toxicological Sciences, The U.S. Food and Drug Administration -- FDA)

"Even when overall risk is considered, it is a very small risk (1.19) and it is not statistically significant at a conventilonal 95 per cent level." (Congressional Research Service, Report on Environmental Tobacco Smoke and Lung Cancer Risk, Nov. 14, 1995)

"...It seems that the epidemiological database as it stands today does not support an association between ETS exposure and lung cancer. The workplace data reports no elevation in risk and the spousal esposure data reports either no increase in risk or an increase that is so small that it can be entirely accounted for by bias and confounding factors that have not been properly addressed in the studies." (Dr. Hitoshi Kasaga, Tokai University School of Medicine, Japan, in "Environmental Tobacco Smoke," H. Kasuga, (Ed.), Springer - Verlag, p. 84, 1993)

About estimating second hand smoke exposure through the use of questionaires:

It is clear that misclassification and recall bias plague ETS epidemiology studies. It is also clear from the simulations that modest, possible misclassification and recall bias rates can change the measured relative risk results, possibly in dramatic ways. (Enironmental Tobacco Smoke and Lung Cancer Risk, Congressional Research Service Report, p. 45, Nov. 14, 1995)

"The results of the study suggest that questionaires lead to a large amount of exposure misclassification which must be taken into account when assessing the effect of ETS exposure on pregnancy and other health outcomes and that an objective method to measure exposure which is sensitive, accurate, and reliable is needed to validate them." ("Measurement of Exposure To Environmental Tobacco Smoke in Pregnant Women Using Questionaire, Personal Monitor and Urine Cotinine: A Problem in Exposure Modelling", T.Z. O'Connor, B.P. Leaderer, T.Holford, and M.B. Bracken (Yale University), Proceedings of Indoor Air, 1993, vol. 3, 1993)

Most of the studies on seconhand smoke and lung cancer do not sufficiently take into account potential confounding factors:

"The world-wide evidence supporting the existence of differences in lifestyle, etc. between families containing smokers and thos without smokers, for all nationality studied, is now considerable. It is important, therefore, that the studies of ETS and lung cancer collect information on the exposure of cases and controls to all factors that cause lung cancer and to which the may have been exposed and that these confounding factors are included in the analysis jointly with ETS exposure to judge where that true association lies" (Dr. Anthony Springall, Fellow of The Royal Statistical Society, testimony to the US Occupational Safety and Health Administration, 1994)

"Exposure to seconhand smoke includes many factors. Tha colture of the home where there is a smoker includes not only smoking, but also other lifestyle habits. Many of these, such as low vitamin intake and high consumption of alcohol, also increase the risk of cancer." (Genevieve Matanosky, MD. Professor of Epidemiology, at The Johns Hopkins School of Public Health in "Characterisics of nonsmoking women in NHANES I and NHANES II Epidemiologic foolow-up studies with exposure to spouses who smoke" - published in the American Journal of Epidemiology - July 1995)

"The major sources of uncertainty for interpreting the epi results are counfounders - factors other than ETS which could explain the measured risk values, and misclassification." (Congressional Research Service Report on Environmental Tobacco Smoke and Lung Cancer - November 14, 1995)

"Some studies make considerable efforts to control for other factors and to verify the classification of subjetc into the proper categories; other do little in that regard. Even the best of studies, however, face practical limitations on their abilities to verify and control." (Congresinal Research Service Report on Environmental Tobacco Smoke and Luyng Cancer Risk - November 14, 1995)

Comments - It is hard to conceive that all these scientists are wrong. The antismokers are quick to label any study, or scientist speaking against the on-going hysteria as corrupted by the Tobacco Companies. In reality, however, it is very improbable that all these scientist (and many more, not reported here) have been hired - or influenced - by pro-smoking forces. A scientist gains his/her reputation through a lifetime of research and hard work. Only twisted minds would conceive a mass-sellout of the discipline of science.

A BALANCED PERSPECTIVE

The conclusions of science

This sounds worrysome. But let us look ot the EPA findings a little closer.

The EPA said that based on a compilation os 11 U.S. epidemiologial studies, nonsmokers married to smokers had a "relative risk" of 1.19 for lung cancer -- in other words, a 19% increased risk of lung cancer -- compared to nonsmokers with nonsmoking spouses. A figure like 1.19 (or 19%) may sound meaningful. But epidemiological studies have attributed even higher relative risk number for lung cancer to drinking 1-2 glasses of whole milk per day (relative risk = 1.62) [1], for heart disease to eating one biscuit a day (relative risk = 1.49) [2], and for rectal cancer to drinking clorinated water (relative risk = 1.38)[3]

Relative risks in perspective

The reason for this caution is due to the methodology used by epidemiologists to identify potential health risks - that is to say, comparson of statistics and not clinical research of people affected by a specific disease. They compare the incidence of disease in individual exposed to a possible suspected cause to the incidence of disease in non-exposed individuals. One can then, in principle, determine if there is an association between the exposure and the disease.
Let's take a real example:
Research finds that of 100,000 people who do not drink whole milk, 10 of them die of lung cancer. Then it is found that of 100,000 people who drink 1-2 glasses of whole milk per day, 16 die of lung cancer. Therefore the relative risk for lung cancer from drinking 1-2 glasses of whole milk is 16 over 10, or 1.6.

This means that if we have available a database large enough, we can find a correlation between all kind of activities and disease. We could say (and this is not a real example) that people who take more than one shower every day may show a relative risk of contracting rectal cancer greater than people who take only one shower a day, and so on. That is the reason why scientists do not consider significant a relative risk factor smaller that 2.0 - 3.0. The hysteria was launched by the media, picked up by opportunistic politicians and administrators, and supported by the antismoking industry. This created public opinion pressure confirmed by opinion polls. The local and central government responded with repressive regulation. But reality stands: a relative risk factor of 1.19 is INSIGNIFICANT.

There are a number of factors which should be taken into account before drawing any firm conclusion, and alarming the public. For instance, such small increase as 1.19 from 1.0 could be due to one of the 20 or more recognized risk factors for lung cancer such as family history, diet, type of job and alcohol consumption. Epidemiological studies on second-hand smoke cannot take all these factors into account. In fact, they generally address very few if any of them. Also, epidemiologists use questionnaires to gather data. In the case of second-hand tobacco smoke, researchers question non-smookers who have contracted lung cancer and who liverd or worked with smokers to ascertain how much of their spouses or colleagues smoked in their presence. Sometimes they ask the patients to think back 30 years and remember how many cigarettes were consumed around them. And often researchers question not the non-smokers concerned but their surviving relatives. Unfortunately, human memory if far from perfect, particularly if you are trying to remember what someone else did many years ago.
This, and the possible influence of other risk factors, explains epidemiologists' reluctance to draw firm conclusions based on relative risks of 2.0 or less. Many have raised serious questions about the validity of the EPA's classification of second-hand smoke as a known carcirogen based on a relative risk of 1.19.

Criticism of EPA

Journalists have been taken to task for accepting the EPA's conclusions at face value while rejecting independent science. In Forbes Media Critic Magazine, Jacob Sullum wrote: "Despite serious questions about the report's assertion that ETS causes lung cancer and the process by which the EPA reached that conclusion, leading US newspapers have treated this assertion as a scientific fact. In so doing, not only have they exaggerated what is known about the effects of ETS, but they have missed an important story about the corruption of science by the political crusade against smoking" [7]

Some other pertinent comments on the EPA risk assessment and on epidemiology in the area of secondhand smoke:

• Dr. Jane Gravelle of the US Congressional Research Service testified to the US Senate in 1994: "The statistical evidence does not appear to support a conclusion that there is a substiatial health effect of passive smoking". [8]

• M.E. LeVois of Environmental Health Resources: "The ETS lung cancer epidemiological data provide no scientifica basis for government regulation of smoking in the workplace" [9]

• Kent Jeffreys of the Alexis de Tocqueville Institution: "The EPA has manipulated selected portions of the existing literature until it produced the desired result." [10]

• Dr. Gary L. Huber, professor of medicine at the University of Texas Health Center: "In this report on TES (second-hand smoke), the EPA did not comply with accepted principles of toxicology, chemistry and epidemiology, not with its own guidelines for undertaking cancer risk assessment. In fact, the conclusions drawn are not ebven supported by the EPA's own statements" [11]

Government

There is no question that the vast majority of health scientists and experts are truly devoted to improving public health (including this writer), but a rush to publicize relative risk statistics without due explanation of the limitations can lead to concern and unjustified regulation. Over time, a closer look at the evidence may produce conflicting and confusing information.

When will it end? Margarine and Apple under Investigation

• We were told several years ago to substitute margarine for butter, only to be told later that margarine's trans-fatty acids might be worse for our arteries.
• We were told that apple growers were causing ancer with a pesticide called Alar, which EPA promplty banned. Later it surfaced that thoudands of apples would have to be consumed daily for years on end to possibly result in cancer.
• Women were informed by a prestigious study that a diet heavy in fat would increase the risk of breast cancer. Now that has been contradicted by other studies.
• Again, women were informed that silicone leaking out of silicone breast implants would create breast cancer and rehumatoid artritis. This created a sweeping wave of hysteria and multi-million dollar law sutes in the U.S.A. and to a lesser extent, in Canada, and it is still going on today. Immediately after the information was spread by the media with the usual alarmistic tones, hundred of women came forward with stories attributing all forms of artritis and breast cancer to silicone breast implants. Recently, all reputable scientists have come forward stating that there is absolutely no relationship between breast cancer, rehumatoid artritis, and silinone implants. The logic of scapegoating continues.
• The May 1996 edition of the British Medical Journal published a study urging the government to cut the population's salt intake in the face of "overwhelming evidence" that excessive consumption was causing high blood pressure, heart desease and strokes. The May 1996 edition of the Journal of American Medical Association published another study concluding that "dietary salt intake has little effect on blood pressure in the population at large." [12]

How long can we go on trying to bend society to every study, and every hysteria arising from them? How long can we withstand the stress of adapting to every noisy interest group claiming to be affected by discomfort, real or perceived? The social and financial cost of "political correctness" cannot be withstood forever: the supercosts of all this enormous safety/security/state protection machine is well hidden in the books of our economy, but it is showing up in the end of taxation, and continuously reduced quality of life. Money once used for productive ends is now diverted to an enormous, non-producing machine designed to "protect us", and achieve an utopistically perfect society, resulting in moral and financial banckrupcy. When are we going to pull the plug?

Risk in perspective