Other considerations.

Other sources can be used to check and illustrate the it is alcohol consumption that is the major source of lung cancer. Also, it is possible to calculate that smoking can be no more than a very small source of lung cancer.

A chinese investigation.

This survey (http://carcin.oxfordjournals.org/cgi/content/full/25/11/2177) of lung cancer patients in the Xuan Wei province in China has established the correct risk of lung cancer from smoking. It is a workplace related survey, in which the investigators are attempting to speculate that coal smoke is a source of lung cancer. In the section “Results” it is evident that for males the risk increase for lung cancer from smoking is correctly established to be 1.69. i.e. a risk increase of approximately 70 %.

Note also that one of the researchers in the team is from the US federal agency EPA, an agency that claims that exposure to ETS is a source of lung cancer among never smokers.

The survey was published in july of 2004.

British workplace report.

From this British workplace report ( http://www.hazards.org/cancer/hsecriticism.htm ) it is evident that smoking cannot be a significant cause of lung cancer. In section 18 the workplace investigators have established that the risk of lung cancer from smoking is 71 %. It is therefore known in wider circles that smoking involves a very modest risk of lung cancer, in the order of 70 %.

The report was published in june 2007.

A crude estimate of lung cancer from smoking.

American authorities apply a computer model named SAMMEC to create propaganda about smoking. The Lung Cancer Alliance quotes numbers from SAMMEC ( http://www.lungcanceralliance.org/facing/risks.html ). From the section “But what if I quit smoking” can be derived the following table and corresponding estimate of lung cancer risk from smoking, labelled Crude RR.

Table 1. Crude estimate of the risk of lung cancer from smoking, from SAMMEC.

The calculations presume that 'all other factors are equal'. Of course all other factors are not equal. Alcoholics are far less likely to give up smoking than non-drinkers, so the estimate is skewed. But it is clearly evident that the risk of lung cancer from smoking cannot be 1950 % as claimed in the Doctors Study 1994 report, table III.

It comes as no surprise that the crude estimate of RR for smoking and lung cancer is the same for males as females, here at 2.6, corresponding to an increase risk of 160 %.

The negative dose response of smoking on mortality.

It is known in closed circles that there is a peculiar effect in the smoking surveys among those that smoke many cigarettes, more than 30 – 40 cigarettes per day. In cohorts with high cigarette consumption the risk of lung cancer and mortality drops. Investigators that are intent on deamonizing smoking mask this by making an upper bound of say more than 25 cigarettes per day. Thus they can hide this peculiar inverse effect at high rates of cigarette consumption.

This effect is due to the fact that alcoholics and heavy drinkers have a favorite consumption of cigarettes, typically with a mximum of around 30 cigarettes per day. This is something that is typically not visible in most smoking surveys. But there are exceptions to this rule. Every now and then this peculiar inverse effect appears. An example of this is from a Norwegian survey that was published in the magazine called Tobacco Control in 2005 ( http://tobaccocontrol.bmj.com/cgi/content/full/14/5/315?ijkey=1dbe89423f819c1e7205860d195c753ae06aa8a4 ).

From Table 2 in the survey it is possible to study the inverse effect on the female smokers.

Figure 1. Excerpt from table 2 of the Norwegian survey. Mortality rate is per 100.000. The phantom effect from alcohol decreases at higher levels of cigarette consumption. The female drinkers in this Norwegian survey prefer smoking between 20 and 24 cigarettes per day. Or perhaps it would be more precise to say that the density of drinkers among smokers is the highest in the cohort of those smoking between 20 and 24 cigarettes per day.

The observations of a physician.

Physicians seem to have a great deal of opinions about what one should or should not do. One Dr. De Winter has his own website. He has made certain observations on smoking and drinking. From the section on alcohol ( http://www.dew-health.org/html/alcohol___cancer.html ) there is this quote:

“The rates of lung cancer and laryngeal cancer in smoking drinkers, however, differ dramatically from those seen in non-drinking smokers; alcohol alone, therefore, may play the more significant role at this particular site.”

It is very easy to establish whether alcohol or smoking plays the most significant role. Simple inspection of the correlation between alcohol consumption and lung cancer demonstrates that alcohol is the dominant factor.

The inhalation mystery.

The first survey that Austin Bradford Hill and Richard Doll completed on smoking and lung cancer was a socalled case control study. It was published in 1950 in the British Medical Journal, BMJ, 1950, Volume 221, page 739-748. From it comes one observation that has been the subject of several speculations and debates. It is the following quote, from page 746:

“In fact, whether the patient inhaled [tobacco smoke] or not did not seem to make any difference.”

The explanation is that alcohol consumption is the dominant factor for lung cancer. It therefore plays no practical role whether one inhales tobacco smoke or not.

Explanation of the MRFIT survey.

There are a few surveys that point to the fact that quitting smoking does not result in any measureable health gains. One such survey is the MRFIT survey ( http://jama.ama-assn.org/cgi/content/abstract/248/12/1465?ijkey=774236c019225b9c4ae342eb40b03f34f4788ed6&keytype2=tf_ipsecsha ), that was published in 1982. The authors of the report are clearly puzzled at the results.

The survey followed two cohorts of mid-aged males. The first group was the intervention group, the other was the control group. The test group was systemtically exposed to smoking cessation counselling and and dietary advice. After seven years, the survey showed no effects. There was a small variation in the causes of death among the two groups. The test group actually had a higher mortality rate than the control group.

The simple explanation is that smoking is too modest a health risk to be measureable. The smoking surveys measure the effects of drinking. Alcohol was not a part of the MRFIT survey, so no result came out of it.

If one were to conduct a survey corresponding to the MRFIT survery, but only concentrating on the effort to convince the test group members of alcohol abstinence, the result would be overwhelming. Presumably just five years of testing would be sufficiant to document a very strong effect of alcohol abstinence.

Conclusion.

Other sources than the smoking surveys document that smoking cannot be a major source of lung cancer or excess mortality. Peculiarities from smoking surveys are easily explained with alcohol consumption.