David W. Kuneman Speaks On The Missouri BanCurrently the claim is secondhand smoke kills about 65,000 persons/ year, but recent studies of nicotine in nonsmokers have claimed to show a 90% reduction in secondhand smoke exposure relative to the past.
If all this is true, then secondhand smoke must have been killing about ten time as many persons, or about 650,000 persons / year in the 1970s. In 1970, for example, there were 1,921,000 total deaths reported in the USA, from all causes. This means secondhand smoke must have been responsible for about 35% of all deaths in 1970. However, since about 40% of us smoked, then those 670,000 deaths must have occurred among the 60% of us who did not smoke. This is a very rough calculation, but this means approximately, if 60% of those 1,921,000 deaths were nonsmokers, then approximately 1,152,000 total nonsmokers died from all causes, and 650,000 of those were caused by secondhand smoke. This is more than one out of two deaths among nonsmokers. This is of course, patently ridiculous, and it would have been impossible for our public health authorities to not notice. It is also mathematically impossible if you audit the death tables, subtracting all the deaths due to causes not possibly related to smoke exposure,( accidents for example) and tallying the number of reported deaths that could have been caused by smoke among nonsmokers. Also, if the current claim that first-hand smoking kills 450,000 persons/year is true, then first-hand smoking must have been killing 900,000 persons/year in 1970. Totaling this with the 650,000 gives 1.5 million deaths due to smoking and secondhand smoking, or 75% of all deaths in 1970.
However, toxicologists generally find the risk of disease rises with the square of the exposure to a substance. If secondhand smoke claims are true, then secondhand smoke was killing 6.7 million nonsmokers/year in the 1970s. But, with only 1.9 million recorded deaths from all causes/ year, this possibility does not exist.
Please take note, that epidemiology, which is the science used to derive statistical relationships between exposure to a risk factor and disease is generally not practiced today as it was in the past. Researchers who make claims like the ones above are supposed to check those claims against all available other evidence. I have an old epidemiology text book, Cancer Epidemiology, Johns Hopkins Press, 1967 and it says statistical relationships, such as those which claim secondhand smoke causes lung cancer, heart disease, many other disorders, and now even teenage obesity cannot be considered anything but hypothesis unless all other evidence supports the hypothesis.
Chapter 5 : A statistical relationship by itself is a hypothesis
* could be an artifact
*could be an indirect association
*could be a causal associationdetermining if the effect is real:
1.population studies - etiological hypothesis- Morbidity studies- Mortality studies-time trends
This is why I presented data in earlier speeches that we are no healthier today, than in the 1970s, and states with smoking bans are no healthier than states without smoking bans.2. biological plausibility I emailed all of you a recent report from the Journal of the National Cancer Institute that genetic sequencing of nonsmoker's lung cancer showed the absence of the same genetic damage found in smoker's lung cancers.
3. additional epidemiological evidence
*strength of the association: First-hand smoking and lung cancer: risk is 10-times higher than nonsmokers. This is considered proof.
* specificity of the association: An association can be weak, but be a fact if it is specific. For example Vit D deficiency and rickets. Researchers can be fairly certain if they find a 10% dietary deficiency in Vit D causes a 10% increase in rickets cases.
*dose-response relationship: Again the Vit D example. A 100% deficiency of Vit D causes 100% occurrence of rickets.
*consistency of the association: All Vit D experiments conclude the same facts.Today, many studies have been conducted which claim secondhand smoke causes disease, but many more have been conducted which find secondhand smoke does not cause disease. My own review of these studies has found that those that claim the risks are real neglect to take into account the other evidence which "Cancer Epidemiology" teaches us must also be considered before concluding a statistical relationship is real.
For example, Enstrom and Kabat in the UCLA study which I previously presented to the Justice and Health Committee, measured population time trends with respect to exposure to smoke and heart disease and lung cancer in nonsmokers, in addition to nonsmokers' risk from spousal smoking. They found smoke does not cause cancer or heart disease in nonsmokers. In contrast, the EPA report which did not measure time trends found that smoke does cause lung cancer and heart disease in nonsmokers.
Biological plausibility has been claimed to exist because one can hypothesize smokers risks would just be scaled down in nonsmokers exposed to smoke. For example the EPA did do this, but recently an article in the Journal of the National Cancer Institute proved biological plausibility does not exist.
These findings "support the hypothesis that at least two distinct molecular pathways are involved in the pathogenesis of lung adenocarcinomas, one involving EGFR TK domain mutations and the other involving KRAS gene mutations," the authors write. These results also "suggest that exposure to carcinogens in environmental tobacco smoke may not be the major pathogenic factor involved in the origin of lung cancers in never smokers but that an as-yet-unidentified carcinogen(s) plays an important role."Strength of the Association....... Generally, antismoking activists claim smoke increases the risk of lung cancer in nonsmokers 20%, and heart disease 30%. The American Cancer Society's guidelines are that any risk under a 100% excess cannot be considered evidence that the hypothesis is true. The U.S. Justice Dept. has adopted the same standard which is why secondhand smoke lawsuits have never successfully been brought by plaintiffs. " Cancer Epidemiology" states a strength of association such as first-hand smoking and lung cancer is considered proof of a fact, but weak associations are not. Indeed, weak associations such as Vit E is the sex vitamin, high doses of Vit E prevents heart disease and cancer, and now high doses of Vit. E causes heart disease and cancer have been proven false. Female hormone replacement therapy now has been found to increase the risk of cancer and heart disease while formally it was thought to prevent these. Then Dr. Atkins comes along and claims diets rich in saturated fats prevent obesity and heart disease, and a quick study by the American Heart Assn. agrees. It is not possible secondhand smoke causes teen obesity because first-hand smokers are generally found to have 7% less body fat than nonsmokers. Lastly, the American Lung Association claimed in the 1980s that smoking caused 75% of all cases of stomach ulcers. Now we know 90% are caused by the bacterium H.Pyolri, and the other 10% are adverse reactions to medication. ( Incidentally, now that we know what causes ulcers, we can cure them.) In summary, there is not one weak association discovered in the last 50 years, that turned out to be a fact, secondhand smoke related, or not.
Specificity of the Association: Indeed secondhand smoke is blamed for practically all diseases. This idea is ridiculous. Particularly since tobacco smoke is composed of the same substances that wood smoke is composed of, and we've been breathing that for over 50,000 years without all these diseases being associated with it. Again, teenage obesity???? There is a mountain of evidence smokers have a lower percentage of body fat on average than nonsmokers. How could smoke affect one group one way, and another group in the opposite way?
Dose-response relationship: Lawrence Garfinkle who was vice president of epidemiology of the American Cancer Society published in 1980 that nonsmoking spouses of two-pack-a-day smokers were less likely to get lung cancer than spouses of lighter smokers. For that reason, and after correcting for urban residency, he concluded smoke does not cause lung cancer in nonsmokers. Overall, dose response data are mixed. Some studies find a direct relationship, others an opposite relationship. The only ones the antismoking activists tell you about are the ones which find a direct relationship.
Consistency of the association: many secondhand smoke studies find a negative correlation, many a zero correlation and some a positive correlation.... Why do we only believe the positive ones???? What is actually going on is that the standard deviation of all the study results is greater than the average result of all the studies combined. The mathematical uncertainty is greater than the amount of risk claimed. No real scientist will rely on data like that.
For more, please go to http://www.junkscience.com/ The site is run by Dr. Steven Milloy, who works at the Cato Institute of Business and Government. Please click on the "short course" This will give you a better perspective of what I'm trying to tell you than I can give you in five minutes. Since a recent Gallup poll found 53% of Amercians don't believe secondhand smoke is a serious risk, you owe it to us to investigate every possibility that smoke may not cause disease and personally become as knowledgeable as possible about all sides of this issue before voting.
David W. KunemanAugust 24, 2005
I do not work for Big Tobacco, nor any other interest that pays me any compensation otherwise. I'm a retired pharmaceutical chemist, and became familiar with epidemiology as part of my work. I have had access to the medical literature for years, and have noticed many studies which find no harm from secondhand smoke that never get media attention. We should not legislate based on bad science!