Time trends

De ‘Oorlog tegen Tabak’ wordt al gevoerd sinds 1964. Er werden gouden bergen beloofd, maar is er een effect te merken na veertig jaar? Dertig jaar is de maximale tijd die nodig is om te bepalen of er effect van een ingreep op dit gebied te meten is.

Een gepensioneerd Amerikaans farmaceutisch chemicus had alle tijd om de effecten, aan de hand van onafhankelijke gegevensbronnen, op een rijtje te zetten en ontdekte dat al die overheidsdollars voor niets zijn geweest.

ABSTRACT: Most scientists believe a change in prevalence of a health risk behavior in a population will manifest itself between ten and thirty years after the change takes place. Since 1964, the War on Tobacco has caused one of the largest changes in health behavior of a population ever known, over a relatively short period of time. It is now past the time when we have a right to expect profound changes in the health profile of Americans due to the War on Tobacco. This article examines the smoking behavior of various American birth groups, identifies the years when smoking related disease should occur based on the age of these birth groups, and concludes no significant health profile changes have occurred that can be credited to the War on Tobacco. Further, it is concluded  that no cost savings treating tobacco related disease have been realized.  Comparisons are made to the health profiles of Europeans who have had no War on Tobacco which confirm our War on Tobacco is worthless. Literature is cited that predict our young generations of Americans can also expect no health bonanza from the War on Tobacco.

The author is a retired pharmaceutical chemist who noticed no improvement in health of Americans while accessing the medical literature during his career. His personal health profile does not include any smoking related diseases, and he believes anti tobacco  activists and public health officials have diminished the quality of his life more than any tobacco company.


by David Kuneman

I have been compiling smoking rates and smoking related disease rates gathered over the last thirty years by government agencies.  Time trends of most medical conditions blamed on smoking are stationary or rising even as the prevalence of smoking declines.  The data I am about to present should be of interest to anyone concerned  with claims smoking rates are a major health cost to society. I have compiled as much of this data as possible from Statistical Abstracts of the United States (SAUS) and the National Cancer Institute’s Surveillance, Epidemiology, and End Results ( SEER) Program.  The SEER Program is a scientifically selected population-based registry of 9 geographic regions within the United States used to identify and quantify our overall health trends. These are widely regarded as the best data available. The tables below present these data. The smoking rate data in some tables are inserted by me for the convenience of the reader, and not part of the original database cited. Birth group refers to birth year plus or minus five years.  Thus, for example, birth group 1905 refers to those born from 1900 to 1910. 

Identification of years when most smoking related disease should strike.

I have obtained historical (cigarette only) smoking data from J. Harris,  J. Nat. Cancer Inst. , Vol. 71, #3, Sept, 1983, p473 of the birth groups of Americans old enough to be in the smoking related disease risk group between 1973 and 1998.Current smoker at age 60, birth year 1925, data were extrapolated by me because these are 1985 data and Harris published during 1983. Harris concluded male smokers born 1905 +5 and females born 1925 +5smoked the heaviest lifetime dose.  By birth group, male smokers peak cigarette use was actually males born 1925 during 1953, but they began quitting much earlier in their lives than prior birth groups and smoked almost exclusively cigarettes; their lifetime consumption was less than 1905 males.  Harris reported the mixing of cigarettes with other tobacco products was more common among earlier male birth groups. I also conclude the 1905 male birth group smoked the heaviest lifetime dose because males born 1905 inhaled more tobacco combustion products: first, they also smoked more cigars, ( Table I ) second, they smoked more unfiltered cigarettes which were most popular before the 1950s when these males were already in middle age. The plain fact is when a filter cigarette is smoked, one can see tar build up in the filter. This is tar that is never ingested by the smoker. Studies consistently claim unfiltered cigarettes are more dangerous than filtered. Third, they arrived at the age of smoking inception during the 1920’s when smoking was very popular and acceptable, and fourth, they were already average age 59 when the first Surgeon General’s report was published in 1964, supposedly too late to benefit much from the rash of quitting that followed.  At the time of Harris’ publication, cigars were thought to be less risky than cigarettes which was the official explanation why males who lived during most of the 1800’s never suffered from lung cancer. When cigar smoking again gained popularity during the 1990’s, reports that cigars are indeed as hazardous as cigarettes began appearing. If true, then males born 1905 ( smoking both) ought to be most at risk from smoking related disease. However, these recent cigar reports never gave a satisfactory explanation why lung cancer was extremely rare before 1930.

Table I ,Domestic Cigar Consumption, 5-year intervals, in Millions

Per Capita Male/year over age 21* 
 5-Year Interval Large Cigars/yr Small Cigars/yr Large Cigars Small Cigars  
 1920-1924 7,018.6 594.5 223 19
 1930-1934 4886.1 286.5 130.5 7.5
 1940-1944 5449.6 132.7 125 3.0

*Calculated from data  from “Historical Statistics of the U.S”. House document #93-78 part1 and the Production of Tobacco, W.W. Garner, U.S. Dept of Agriculture, 1947.

Males, average birth years 1895 and 1905, reached the age of smoking inception when males smoked an average of 223 cigars per capita annually. Latter birth groups reached the age of smoking inception after cigar consumption began declining.

Another reason to believe male smokers born 1905 were the heaviest is due to their low quitting/ decade-after-age-30-numbers.  These data in Table II are from Harris.  We have been bombarded with claims substantial reduction of smoking related disease occurs when people quit before the disease strikes, and even quitting after the disease strikes can have benefits.  Birth group 1905 males ought to have more smoking related disease because they smoked the most at age 50, and almost as much as 1895 males at age 60.   The cigar issue doesn’t apply to female birth groups. Female birth groups 1915 and 1925 actually had somewhat similar smoking histories, both smoking 37% at age 50, but those born 1925 took up the habit earlier in their lives while the 1915 birth group smoked the most at age 60.  Smoking at age 50 ought to be the most important parameter because most smoking related cancers occur after age 60 and claims are common that ten years after quitting, risk of cancer declines dramatically.

Based on age of these birth groups,  smoking related disease incidences during 1973 occurred  mostly among those born 1905 +5,  during 1985 among those born 1915, and during 1996-1998 among those born 1925.  The slightly longer time span, I have presented, between birth and expected incidences at later intervals is due to overall life expectancy increases occurring from 1970 to 1998.

( refer to J. Amer. Med Assoc., Feb9, 1994, p435, T5. Most smoking related cancers are diagnosed age 60 to 80, with peak age 70. Birth year overlap is about 25% over ten-year intervals )

Table II, Harris Cigarette-only Smoking Prevalence by Birth Group

 Ave Birth Year + 5 Ever Smoked Smoker Age 50 Smoker  Age 60 After Age 30
 MALES 1895 50% 50% 50% 0%
 1905 62% 60% 49% 5%
 1915 68% 59% 44% 9%
 1925 69% 50% 33% 17%
 FEMALES 1895 12% 12% 11% 0%
 1905 24% 24% 23% 0%
 1915 38% 37% 31% ~0%
 1925 42% 37% 25% 7%

From SAUS1992, T19&39 and SAUS 2000, T13, I obtained data from the U.S. Census Bureau, Current Population Reports, and have calculated percents of each birth group still living among those over age 60, of the same sex, during various decades. m= male, f= female  NA= not yet age 60.

Table III, Percent of all Persons Over Age 60, by Birth Group

Birth Group:1895190519151925
during1970’s31.3%m 33.4%f57.7%m 52.2%fNANA
during1980’s11.3%m 16.8%f31.4%m 33.3%f57.3%m 49.8%fNA
during1990’s 12.6%m 19.5%f33%m 34%f54.3%m 46.3%f
during1999  13.8%m 18.6%f35.9%m 35.7%f

Among those over age 60, the highest portion of persons are under age 70 during all years studied. Therefore one should expect smoking related disease patterns during any decade to be most dependant on smoking behavior of those born 60 to 70 years earlier.  The reason vertical data total more than 100% is because our population  over age 60 increased during the later years. The reason the 1999 horizontal data don’t total 100% is the 1935 birth group is then over age 60 and not included in this table.  Male and female 1935 birth groups both smoked less than those birth groups shown.

During 1980, the males born 1905 , still alive, were average 75 years old, and females born 1925, 55 years old. Considering average life expectancy, ( 72 for males, and 79 for females),  most of the heaviest male smokers were no longer alive to influence the incidence trend of smoking related diseases during  and after 1980.  This would leave only males born 1915&1925 who smoked 1% to 10% less at age 50 and 5% to 16% less at age 65 at risk for these diseases after 1980. After 1990, males and females born 1915 have less influence on smoking related disease incidences because they comprise only 33-34 % of all persons over age 60.  Females, still alive, had steady smoking rates at age 50 and 60. Therefore we should expect peak smoking related disease incidences among males during the 1970’s and after 1980, declining incidences. Among females, rising incidences during the 1970’s and steady smoking related disease incidences after 1980.

There is, however, a reason to expect smoking related disease declines even earlier than 1980.   IF claims are true smoking reduces life expectancy an average of 14 years, then all of the projections made above should have occurred 14 years earlier.  Perhaps the average life expectancy numbers for males is composed of 70% never smoking males and males who quit before age 50, living to an average age of  76 and 30% at risk smokers living to an average age of 62.  The overall life expectancy would still be age 72, but the smoking component would have died at average age 62, or  average 1967 among those born 1905 and average 1977 among those born 1915.  Those persons, still living, in the birth groups 1905-1915 during, 1980, and 1990 would then be mostly those who never smoked or quit early in their lives.  This would leave only males, birth group 1925 and latter, who smoked less than birth groups 1905 and 1915 at age 50 and much less at age 60 alive to suffer smoking related diseases after 1980. Similar calculations for females, average life expectancy 79, yield  average death age 69 for the smoking component, and by 1985,  only the portion of the 1915 birth group that didn’t smoke is still alive instead of almost all the 1915 birth group and this would lead to declining smoking related disease among still-living women because their smoking exposure thereafter would be declining. 

I have addressed an alternate approach to determine if smokers live an average 14 years less than nonsmokers. From SAUS, 1992,T199, I have obtained data from U.S. Centers for Disease Control, Office of Smoking and Health surveys of those aged 25-44 years-old during 1965, and aged 45-64 years-old, during 1985:

* during 1965, of those aged 25-44, 49.5% smoked and their quit ratio was 23.6%

*during 1985, of those aged 45-64, 31.6% smoked, and their quit ratio was 49.7%.

The definition of quit ratio is “percent of persons who ever smoked who are former smokers”.  From this data, it is possible to calculate percent ever smokers of both surveys.  Subtracting 100 from the quit ratio yields percent ever smokers who are current smokers. Dividing percent current smokers by percent ever smokers who are current smokers = percent ever smokers. I obtained:  *for those aged 25-44 during 1965, 64.8% ever smoked.

*for those aged 45-64 during 1985, 62.8% ever smoked.

But, these are the same people, simply aged 20 years. IF many more of the ever smokers and current smokers than never smokers had died off before 1985, they would not have been able to participate in the 1985 survey and the 1985 survey results would have yielded much lower ever smoker results.  As a consequence, I conclude the difference between 64.8% ever smokers during 1965 and 62.8% ever smokers of the same group, during 1985 proves smokers don’t die off 14 years earlier than nonsmokers. Perhaps, the two percent difference is due to smoking, but, as will be discussed later, this 2% could easily be explained by other risks more prevalent among smokers, such as blue collar or military employment.

Yet, according to the 1990 Surgeon General’s Report, “The Health Benefits of Smoking Cessation”, after ten years of abstinence, the risk of lung cancer is 30-50% of the risk of continuing smoking. Cessation also reduces the risk of cancers of the larynx, oral cavity, esophagus, pancreas, and bladder. Heart disease death is reduced 50% after only one year of abstinence.  After quitting, stroke risk returns to ordinary levels after  5 to 15 years.  With sustained abstinence,  chronic obstructive pulmonary disease risk returns to normal, and peripheral vascular disease risk is substantially reduced. He also states death risk due to influenza and pneumonia is substantially reduced. Overall, those who quit before age 50, have one-half the risk of dying in the next 15 years than those who continue to smoke.  Below, we will determine if that actually happened.

From the Tables I and III, it is possible to calculate composite smoking data of all those over age 60 during decades 1970,1980, and 1990 to prepare the following table which compares it to health outcomes that actually occurred during those decades. Bear in mind these composite figures are only accurate if smokers live as long as nonsmokers; otherwise smoking data would be lower.


MALES:  ( born1905 are heaviest smokers)     
composite ever smoking among those now over 6051.5%58.4% 59.9% 
composite smoking @ age 50 among those now>6050.4%52.7% 46.3% 
composite smoking @ age 60 among those now>6044.0%40.6% 32.4% 
life expectancy gains of preceding decade 1.9 1.41.2
CANCERS, AGE ADJUSTED/100,000  (all)372 439 549
Lung73.8 83.8 69.8
Colon and Rectum53.2 63.0 51.7
Pancreas12.9 11.3 9.9
Oral Cavity and Pharynx17.5 17.1 13.6
Urinary Bladder25.6 26.9 29.0
Esophagus5.5 6.3 6.7
Larynx8.4 8.7 5.6
FEMALES ( born 1925 are heaviest smokers)     
composite ever smoking among those now over 6016.5%26.9% 32.4% 
composite smoking @ age 50 among those now>6016.5%26.4% 29.7% 
composite smoking @ age 60 among those now>6015.7%23.1% 22.1% 
life expectancy gains of preceding decade 1.9 0.60.4
CANCERS, AGE ADJUSTED/100,000  (all)292 339 416
Lung18.3 35.3 43.4
Colon and Rectum41.6 45.4 38.2
Pancreas7.8 8.2 7.7
Oral Cavity and Pharynx6.2 7.0 5.6
Urinary Bladder6.3 7.9 7.4
Esophagus1.8 1.9 1.6
Larynx1.3 1.8 1.4
CHRONIC COND, both sexes/1000     
Male/Female smoker, 45-64 years old , %41.2/33.439.3/30.733.4/29.2 27.1/24.0
Heart Conditions 75.482.6 78.9
Hypertension 114.3125.1 107.8
Varicose Veins of Lower Extremities 27.230.6 28.2
Chronic Bronchitis 36.149.7 53.7
Asthma 31.236.8 55.6
Chronic Sinusitis 145.0139.0 126.2
Cataracts 22.6n/a 26.7
OTHER DISEASES, both sexes/100,000     
Cerebrovascular disease hosp. discharge rates254 384  
Chronic Obstructive Pulmonary Disease   deathsincreasing
DAYS OF DISABILITY, both sexes     
Overall Smoking Trend, age 26 and over39.1%36.9%32.8%29.8% 
Work Loss Days5.4  5.6 
Bed Disability Days6.1  6.5 
Restricted Activity Days14.6  15.2 

Life expectancy gains, source:SAUS1992,T104; SAUS2000, T117
SEER cancer incidence data, source: http://seer.cancer.gov/csr/1973_1999/overview.pdf
Chronic conditions prevalence data, source: SAUS,1988,T172; SAUS,1992,T195;SAUS,2000,T220
Days of disability, source:SAUS,1992,T188
Male/Female smoker 45-64 year old data, source, “Health, United States”,2001, table 60
Overall Smoking Trend, age 26 and over source: SAUS,1992,T197


Selected Life Table Values

YearAve Life Remaining @ age 50
197023.128.8 years

Source:SAUS1992,T104; SAUS2000, T117

Males gained the most life expectancy /decade (1.9years) from 1970 to 1980 while most of the heaviest smoking component born 1905 were still alive and most at risk from smoking related diseases. Females also gained 1.9 years/decade life expectancy 1970-1980 but most of their heaviest smokers were not yet at the age where smoking related disease would strike and influence life expectancy.  If past smoking history were an important determinant, males would have experienced the lowest life expectancy gains during the 1970’s. If quitting by age 50 cut the risk of dying in half over the next 15 years, males would have experienced increased life expectancy gains during the 1990’s because their smoking at age 50 had declined from 52.7% to 46.3%. The data do fit the predictions made by the Surgeon General for females, as their smoking at age 50 and 60 increases, their life expectancy gains diminish in the subsequent decade, but when combined with male data, life expectancy gains do not relate to quitting before age 50 or 60. Males and females are realizing  declining life expectancy gains per decade despite ever fewer of them continuing to smoke. According to SAUS, 2001,T191, only 14.6% of males, and 11.5% of females over age 65 smoked during 1990. Yet during the 90’s, we witnessed the lowest life expectancy gains of all.

Females with increasing smoking exposure, gained 3.0 years of life expectancy from 1970 to 1998, males, with declining smoking exposure, gained 4.5. Overall, the influence of smoking patterns after age 50 is obviously of little importance on remaining life expectancy. With this dramatic change, males ought to have fared much better than females.  Further, females have always had a higher life expectancy and this is a disadvantage when comparing net gain with males. Finally, most of these males at age 50 were employed, and tend to have more hazardous occupations than females. The odds a male dies due to an accident are twice those of a female. SAUS 2000, T705 indicates fatal accidents at work declined from 21/100,000/yr  during 1960 to 4/100,000/yr  during 1998.   For reference, lung cancer kills about 50/100,000 Americans annually. The extra 1.5 years of life males gained past age 50 vs. females could easily be explained by improved workplace safety alone. We should expect males to attain proportionately  more life expectancy based on this accident data.  All deaths due to accidents and violence declined from 101.9/100,000 white males and 42.4 white females during 1970 to 77.3 white males and 33.2 white females during 1998.  According to a study published in the Journal of Drug Issues, 1993, 62% of prison inmates smoke. Because criminals are much more likely to die of violence than the rest of us, it is highly plausible some of the excess mortality attributed  to smoking is actually due to the lifestyle of criminals. The Chicago Tribune reported Mar. 21, 1997 that smokers who attend church regularly have 4 to 7 times lower risk of high blood pressure than smokers who do not. It also reported church-going smokers have the same risk of high blood pressure as non-churchgoing nonsmokers. It is obvious the lifestyle of criminals is a health risk, and that they smoke more than the general population causes some of this risk behavior to be falsely attributed to smoking. Together, accidents and violence represented 5% of all deaths in 1970 ( about the same as lung cancer) and therefore strongly influence overall life expectancy; particularly since accidents and violence are not age dependant. Accidental and violent deaths are mostly male and had they not declined, the male life expectancy increase between 1970 and 1998 ( 4.5 years) would have probably been similar to females ( 3.0 years). This similarity would rule out the possibility the female data do fit predictions made by the Surgeon General. The different male and female life expectancy gains between 1970 and 1998 are explained by factors other than smoking.


 According to http://seer.cancer.gov/csr/1973_1998/overview.pdf , Table I-3 , the percent change of lung and bronchus cancer incidence between 1950-1998 was 248 % / for colon and rectum, 0.6% / pancreas, 13.7% / oral cavity and pharynx, -40% / esophagus, -2.7% /  larynx, 20% / and urinary bladder, 53.8%.  During 1950, the average birth year of those at risk was 1885. According to Harris, only 33% of these males and 3% of these females ever smoked during their lives and by 1950, only 25% of these males, and about 2% of these females still smoked. The heavier smoking 1895 birth group ( table II) was only average 55 years old, still too young to influence this data much. As already discussed, those most at risk during 1998 had much higher exposure to smoking. Considering these dramatic differences in smoking behavior, any cancer due to smoking ought to be much more prevalent during 1998. Only cancers of the lung/bronchus , oral cavity and pharynx, pancreas, urinary bladder, and larynx fit this pattern. We can immediately eliminate cancers of the colon and rectum, oral cavity and pharynx, and esophagus as being dependant on smoking behavior of birth groups. Below, data are presented in more detail for each of the 1905, 1915, and 1925 birth groups for the years when their cancer would be most likely to occur.

If smoking contributes substantially to the incidence of these cancers, one  would expect males peak incidence during the 1973 period when Harris’ heaviest lifetime male smokers were still alive, and of age where these cancer most often strike.   If the smoking component of Harris’ heaviest smoking males lost an average of 14 years of life expectancy, then these peak incidences would have occurred during the 1960s. Male lung and laryngeal cancers peak incidence rates occurred between ten and twenty years after they should have ( 1985) if past smoking history is any measure of vulnerability .  Another way to examine this data is to compare the different smoking histories of males diagnosed with lung and laryngeal cancers during 1973 and 1996-98.  Males born 1905 and diagnosed in 1973 only quit 5%/decade after age 30. Males born 1925 and diagnosed in 1996-98 quit 17%/decade after age 30.  Yet both birth groups had the same lung, and colon-rectal cancer rates. Cancers of the esophagus and bladder are still increasing while lifetime exposure to smoking continues to decline. This eliminates bladder cancer, which had been previously included from the 1950-1998 data. Only cancers of the pancreas and larynx survive all comparisons but lung cancer will be considered still included because of the strong correlation with the 1950-1998 data and a proximate correlation  with the 1973-1996-98 data. For females, steady incidences of these cancers would be expected after 1980 while half of the 1915 and almost all of the 1925 birth group were still alive, and this assumes smoking females’ life expectancy is 69 due to 14 years diminished life expectancy. If female smokers live as long as nonsmokers,  then female incidence trends should be peaking now.  However, conclusions male pancreatic and laryngeal cancer are related to smoking history, are contradicted by declining female pancreatic and laryngeal cancer trends from 1985 to 1998.   ( They should have been steady.) Despite smoking much less, females born 1905 had slightly more pancreatic and the same laryngeal cancer rates during 1970-73 as females born 1925 during 1996-98. It can be concluded birth group 1925 females who were the first to have measurable quitting after age 30 did not benefit relative to 1915 females. ( Table II). The more detailed analysis presented here, continues to support the conclusion from the 1950-1998 data that cancers of the colon-rectum, oral cavity-pharynx, and esophagus are not related to smoking behavior of birth groups. The parallel trends of male and female incidences of most of these cancers, including the “all” cancer data, are striking, even though their historical smoking patterns are quite different. Some early evidence suggests female lung cancer has peaked, but this can only be correlated with their past smoking behavior if we assume no diminished life expectancy and then we can’t explain why their laryngeal cancer rates declined after 1985. More years have to pass before we can reach any firm conclusions concerning females, but it is clear substantial quitting between male birth groups 1905 and 1925 did not benefit the latter and save society costs of treating any of their cancers, except possibly, of the lung, larynx and pancreas. 

Davis, et. al. J. Amer. Med. Assoc. Feb. 9, 1994, p 434 took a broader approach and found the same trends, using a log-linear Poisson model. ” Successive birth cohorts of women have experienced dramatic increases in smoking-related cancer, with women born in the 1920s and 1930s developing six times more cancer than women in the baseline birth cohort 1888 through 1897. Although the explosive rise in smoking-related cancer appears to have peaked, the relative risk for the most recent cohort in Fig. 4 is still about 5 (RRc, 4.93; 95%CI, 3.83 to  6.37). Among men, the incidence of smoking-related cancer increased with birth cohort to the point where rates for men born 1923 through 1932 were about 40% greater (RRc, 1.39, 95%CI,1.00 to 1.94). than for men born 1888 through 1897. In the more recent birth cohorts of men, however, these rates have decreased.  For instance, men born in the 1940s had roughly the same risk of developing smoking-related cancer as men born around the turn of the century.”

From Harris, females born 1895 ever smoked 12% ,smoked 11% at age 60 and 10% at age 70. Females born 1925 smoked three times as much and, according to Davis, had six times as much smoking related cancer. Males born 1895 ever smoked 50%, smoked 50% at age 60, and 37% at age 70. Males born 1925 ever smoked 69% , smoked 33% at age 60 and much less at age 70.  Although smoking more when they were young, males born 1925 quit to levels below the 1895 birth group by age 50 and still suffered 40% more cancer.  Among these male and female birth groups, smoking related cancer is obviously out of sync with smoking history.  Then how can we correlate their smoking histories with the costs of treating these cancers?


Statistical Abstracts of the U.S. publishes data on selected chronic conditions, but unfortunately the NHIS warns data collected prior to 1979 cannot be compared to latter data because the data collection procedures changed in 1978.  We cannot compare data when the heaviest male smokers were still alive (1970s) with data obtained while the lifetime male smoking exposure was declining. We can do this for females because their peak exposure birth groups were mostly still alive during 1980.  Hence the combined data should reflect progressively declining male exposure to cigarette smoking, and steady exposure among females. IF smoking shortens life expectancy 14 years, then all data below should be declining.  I inserted age 45-64 smoking rates because alternately, these data can be compared with that prevalence data to assess how current smoking influences these conditions instead of the past smoking histories of Harris’ heaviest smokers during the years leading up to the prevalence data.  Claims are common that quitting smoking results in a decline of 50% for heart disease after only one year. Then, a decline of 4% current smoking between 1973 and 1985 should have resulted in less heart conditions and hypertension by 1985, and a decline of 9% smoking between 1973 and 1996 should have resulted in even lower prevalence by 1996. Examining current smoking behavior of each gender has little or no effect on the prevalence of conditions at the same year, or latter years.  A decline of these conditions due to most of Harris’ heaviest smoking males dying before 1980 is also not indicated. 

It is noteworthy that despite 45-64 year old males smoking 12% less and 45-64 year old females smoking 7% less during 1998 than 1985, the prevalence of heart disease is higher in 1996. Hypertension is down slightly, but better drugs exist now to fight this condition. Chronic sinusitis is improving and it seems reasonable smoking can aggravate this condition. Asthma and chronic bronchitis clearly bear no past or present historical relationship to smoking. Varicose veins of the lower extremities has not changed. Overall, the Surgeon General’s promises don’t seem to be holding up for these chronic conditions and no cost savings have been realized treating fewer of these chronic conditions.


 Some other diseases are blamed at least in part, on smoking, but incidence or prevalence data are not reported in the references cited above.  I have identified other sources the interested reader might wish to review.

Cerebrovascular disease hospital discharge rates are rising, from 254/100,000 persons in 1970 to 384/100,000 in 1985.   (http://www.cdc.gov/mmwr/preview/mmwrhtml/00001370.htm) The authors report it’s prevalence is difficult to ascertain. Again, with fewer of Harris’ heaviest male smokers still living in the 1985 population, and less current smoking among those 45-64 years old, one should expect a decline of this disease. This decline could be partially offset by increased smoking of females, at age 50 and age 60 (10%)in the birth groups at risk in 1985, but not to the extent that an overall rise of 130 discharges would result.  But that offset would have been modified somewhat by a decline of 4% among males over age 60 in 1980.  With a 6% overall average smoking exposure increase at age 50 and 60 among those now over age 60 during 1980, the 1985 data show a 51% increase in CVD relative to 1973, or 45% more of an increase than we would expect if CVD rates were smoking-dependant.  Modern claims are that CVD risk declines rapidly after quitting smoking. Smoking had declined from 41.2% to 39.3% for 45-64 year-old males and from 33.4% to 30.7% for 45-64 year-old females between 1973 and 1980. Although ever-smoking status of at risk females was increasing, current risk as measured by current smoking was declining for both males and females. Allowing 5-years for this supposed lower risk to be realized, (1985) one can simply not explain why actual CVD occurrence was 51% higher.

 Chronic Obstructive Pulmonary Disease death rates and hospitalizations are increasing, but mild to moderate COPD decreased from 1971-1975 to 1988-1994 among those over 55 years of age. Among other groups, rates have been  stable. (www.cdc.gov/mmwr/preview/mmwrhtml/ss5106a1.htm)

COPD is believed to be the cumulative result of long-term smoking although some improvement is alleged to occur after smoking cessation. Mild to moderate COPD is declining which is what would be expected based on smoking reductions among those aged 45-64 years and the dying off of most of Harris’ heaviest smoking males by 1980.  What of the deaths and hospitalizations?  From  SAUS1992, T113, the age adjusted death rate for COPD was 11.6/100,000 in 1970, and 15.9 in 1980.  From SAUS2001, T 106 the death rate was 19.7/100,000 in 1990 and 21.3 in 1998.  The only long-term conclusion is that serious COPD, resulting in death, doubled over the last thirty years, during a time when lifetime smoking exposure of those over age 60 is now in rapid decline due to increasing number of those over age 60 now belonging to the lesser-smoking 1935 birth group in 1998. Society did not save the costs of treating fatal cases of COPD in 1998 despite lower smoking at age 50 and age 60 among those over age 60 in 1990.


Antismoking activists and other health groups claim one major cost of smoking is days lost from work  Yet from the disability-days data, Table IV, it is clear no employer cost savings were realized as a result of the overall smoking decline among those over age 26 from 40% to 30% during the 1970-1989 time period. 

According to Healthy People, 2000 Review,1992, Table 3, during 1987, 29% of all persons over age 20 smoked; 36% of all blue collar workers smoked; and 42 % of all military personnel smoked.  Smoking was more prevalent in high risk occupations.  These people tend to work outside in inclement weather, work with dangerous substances and mechanical equipment, and work odd hours. It is probably because they have higher risk occupations, they are more likely to take a sick day, or become disabled; not because they smoke.


I used incidence or prevalence data wherever possible because their time trends are more directly influenced by cause, and less influenced by improvements in medical technology over the reported periods.  Death rates from most of these conditions are declining because of improved treatment, not because these conditions are occurring with decreased frequency.   The purpose of the War on Tobacco was to prevent these diseases from occurring in the first place, and in this regard, it has already proved itself a failure; despite overwhelming cooperation from the public, as evidenced in the first table.

Using the same analysis for the prevalence of selected chronic conditions data and SEER cancer incidence data presented above, it is clear we should have expected national declines among these diseases after 1985, but only if smoking rates bear a strong relationship to public health. .  If smoking rates actually have little influence on these disease trends,  even when the  people suffering from them are smokers, then we would expect time trends similar to what we are actually recording.  Even while smoking declines, the true causes remain and many appear to be increasing.

While it is too early to be certain, it appears today’s younger Americans will suffer ever increasing rates of smoking related cancers in the future, despite never smoking as much as the generations of the past.  From the J. Amer. Med. Assoc. article, (Davis) previously cited, the following table is constructed:  M= male, F= female


(SAUS 1996, T222) SEER
Age       Period         smoking rate at
 specified age
smoking rel. cancer
20-241973-197743.1%M  34.1%F45M    35F
20-241978-198237.5M    33.8F26M    24F
20-241983-198732.6M    30.4F43M    31F
25-341973-197750.5M    38.8F202M  141F
25-341978-198243.9M    33.7F195M  166F
25-341983-198738.2M    32.0F288M  176F
35-441973-197751.0M    39.8F1057M  636F
35-441978-198241.8M    37.0F1127M  735F
35-441983-198737.9M    31.5F1297M  774F

Between 1973-1977 and 1983-1987, smoking declined about 20% among all these younger age groups , but smoking related cancer increased about 20%.  Even without smoking as much during the beginning of their lives, the incidence of smoking related cancer is increasing with each new age group.  The War on Tobacco has also been a complete failure for the health of our younger citizens.  Society is not saving costs of treating smoking related cancer incidences among our youth many more of whom never took up smoking in the first place.

Brown and Kessler, J. Nat Cancer Inst., vol 80, #1, Mar. 1988 performed a computer analysis of smoking and lung cancer trends and project if smoking behavior remains steady after 1990, lung cancer rates during 2030 will be 43 male and 39 female/100,000. If the NCI objectives of 15% smoking prevalence by year 2000 had been achieved, 29 male and 23 female/100,000 would have been expected by 2030.  This is rather surprising. By 2030, the average birth year of lung cancer victims will be 1955.  These future victims are now age 45. They currently smoke at the rate of 29% males and 25% females, only half the rate of their grandparents when they were age  45.  Yet the computer projections are that these 1955 birth group males will suffer as much lung cancer as the male birth group 1895 who, according to Harris, smoked  cigarettes 48% at age 35 and  also smoked cigars. Female 1955 birth group will suffer the same lung cancer rate as the 1925 female birth group. Yet the 1925 birth group smoked at 42% age 45 and the 1955 birth group now smokes 25% at age 45.  Both these1955 birth groups contain twice the never smokers as the 1905 male and 1925 female birth groups. Brown and Kessler’s lung cancer projections for year 2000 are close to what has actually occurred during 2000 which helps validate their computer model.  Even for the future, there will be too much lung cancer to be explained by our diminished smoking rate.

These trends have not gone unnoticed by health scientists and public health officials, and they have published reasons why we should not expect public health improvements until even more smokers quit.

One such reason was addressed in ” Twenty Five Years of Progress”, Surgeon General’s Report, 1989. On page 140, he stated that those smokers most at risk in 1985 were also smokers in 1965.  He went on to say  most male smokers in 1965 had also smoked pipes and cigars, but in 1985 almost exclusively cigarettes. Females made up 40% of current smokers in 1965, but nearly half in 1985. Female smokers who had started as teenagers were now approaching the age where smoking related diseases would strike.  He then pointed out that the American Cancer Society’s studies indicate remaining smokers had a higher percentage of heavy smoking.  Finally, during 1965, only ¼ of males who ever smoked had quit, while in 1985, ½ had quit.

Harris, 1983 already addressed the cigar and pipe smoking issue.  Many  males born 1881 to 1900 were still alive during 1965, but dying off rapidly.  Many were deceased before the 1973 incidence data.  Almost all were deceased by 1980,  and practically none influenced the incidence data after 1985. Therefore most lung cancer during 1973 should be correlated with cigarette smoking.  By 1985, the 1915 birth group was most at risk.  They had quit more than their predecessors. The Harris data  further indicate females born 1915 smoked 25% by 1935 when they were 20 years old. Many had started as teenagers.  Instead of approaching the age where smoking related diseases would strike, females starting as teenagers were already at the age (70) where smoking related diseases would strike in 1985.  

To address the Surgeon General’s issue that remaining smokers have a higher smoking rate,  I was able to obtain cigarette consumption rates per capita for 1970 and 1990 using US Dept. of Agriculture, Economic Research Service cigarette consumption/capita data published in SAUS, 1992, T1262, and current smoking rates, SAUS 1992, T 199 and SAUS 1996, T 222.  I divided the consumption/capita data by smokers/capita and obtained consumption/smoker data.   In 1970, the average smoker smoked 10638 cigarettes/year, and in 1990, 10980 cigs/year. The claim remaining smokers were heavier smokers in 1985, is obviously false.

It is clear the impact of smoking rates on the historical prevalence of these diseases is so small it cannot be detected because so many other underlying causes are of much greater importance. The smoking behavior of these birth groups is not a major determinant of prevalence of these diseases as these birth groups age. Therefore the smoking component of the costs of treating these diseases is also small. If smoking were public health enemy #1, the impact of our societies’ change in smoking behavior would be so large, other underlying causes could not mask the impact of smoking declines. The problem with statistical methods used to isolate the smoking contribution to the prevalence of these diseases is that there is no ideal population, risk-free of all other causes,  to compare smoker vs. nonsmoker.  Another problem is our health care community’s lack of access to healthy smokers. Smokers tend to be of lower socioeconomic status and less concerned with health care in general. As a consequence they are less likely to have health insurance and seek medical attention when they are healthy.  As a further consequence, healthy smokers are less likely to volunteer for studies attempting to assess the impact of smoking on disease. Our health care community has no alternative but to compare sick smokers to our more affluent overall population. This is why, sooner or later, smoking becomes associated with everything.

Nevertheless, our public health officials love these smoking studies. In our most recent Federal budget, the Dept. of Health and Human Services will consume more of our budget than any other  single Federal department. This includes Defense, and Homeland Security. Charged with improving the overall health of America, they draw big fat government salaries, pensions, and fancy titles. They can ask for ever bigger budgets and larger departments because an ever bigger war against tobacco must be fought. If they can’t promise us a healthier America, how else could they justify such a big budget? They then endow their buddies in academia and health care with bigger grants to produce more studies justifying widening this war. No matter how many smokers quit, they always claim more quitting is necessary before public health improves.  As long as they have tobacco to blame for lack of progress, they don’t have to do their job. Tobacco becomes the dog who ate their homework.

Current media claims are that smoking causes 440,000 deaths per year. According to SAUS, 2001, T107, 2,222700 deaths occurred among those over age 34 during 1998.  Smoking never killed anyone under age 34. Therefore, if the claim of 440,000 deaths is true, smoking is responsible for 19.7% of all deaths over age 34. During 1998, 22% of all adults smoked. Therefore, if you believe this claim, almost everyone who smokes, dies of smoking related causes. This is impossible, because smokers are also at risk of all causes of death that kill nonsmokers.  How can only 10% of smokers die of the same causes that kill 1,782,700 nonsmokers each year?

Current media claims are that smoking causes 1/3 of all cancer.  From SEER sources previously cited, our total male cancer rate (all sites) was 372/100,000 males during 1973 and 549/100,000 1996-98.  IF smoking caused 1/3 of all cancers during 1973, then smoking caused 124 male cancers/100,000 and this was due to Harris’ heaviest male smokers smoking 60% when they were 50 years old.  During the 1996-98 period, the smoking related cancer component would be related to Harris’ 1925 birth group smoking 50% at age 50.  This would be 103/100,000 cases based on the ratio.  Subtracting 103 from 550/100,000 means 447/100,000 cancers during the 1996-98 period were not smoking related.  This 447/100,000 incidences exceeds the total cancer, including smoking related, during 1973. Therefore even if all smoking had been eliminated by 1980, males would still have more cancer today than during 1973.  Where are our public health officials?  From SAUS2001, T119, during 1973 we spent  $101 billion dollars on health care. During 1996, we spent $1.038 trillion dollars on health care. This year we’ll spend about $1.5 trillion. Are we getting our moneys’ worth?


I would like to focus the remainder of this article on lung cancer. Lung cancer and smoking have the strongest statistical correlation. One would think an excess relative risk of 10 to 20 fold would be strong enough to offset the statistical errors induced by other possible lung cancer causes.

Modern research has concluded wood fires produce the same combustion products as cigarettes. The EPA’s website contains references to articles with data on the composition of wood smoke. ( http://www.epa.gov/ttn/atw/burn/woodsmoke1993.pdf)  Benzo-A-pyrene, the chief carcinogen of tobacco smoke has been found in wood smoke according to the National Academy of Sciences as presented in Chemical Carcinogens, ACS monograph 173, chapter 7, Hoffmann and Wynder.  This makes sense, wood burns at approximately the same temperature as tobacco, contains the same minerals, carbohydrates, amino acids, and cellular material. Archeologists believe mankind has used wood fire for at least 50,000 years, and find many of the remnants of these ancient fires in caves where the smoke was confined and breathed by our ancestors. Recently, PBS produced a nova educational treatise on what is known of mummies studied to date. It reported the Ice Man’s lungs were found to be coated with campfire soot. Most of us cooked over wood fires until the middle of the 18th century. One would think we should have developed a natural resistance to these combustion products by now.

During the eighteenth century tobacco production flourished in this country.  The U.S. Department of Agriculture’s  the Production of Tobacco, by W.W.Garner detailed the magnitude of this industry.  In 1843, the French Tobacco Monopoly began manufacturing cigarettes. The Dictionary of Statistics, 1889 Michael Mulhall  reported 1883 annual consumption of 59 oz per capita of tobacco in the U.S. , the equivalent of about 200 packs of cigarettes/adult male/year. Being a nation of mostly farmers, I’m sure many people also raised tobacco for personal consumption which was not counted.  The first cigarette rolling machine was patented in 1881 and by 1900 there were over 160 commercial brands of cigarettes.

In 1912, Adler published a monograph on lung cancer and due to the rarity of the disease asked “Is it worthwhile to write a monograph on primary malignant tumors of the lung?”  Not until 1930 was lung cancer finally recognized as a separate disease.

The sensible scientist will always attempt to associate a recently induced  cause to explain any new disease that seemingly emerges  from nowhere.  Thus it was with Doll and Hill, two British scientists increasingly alarmed with the rapid rise in lung cancer during the 1940’s.  At first, they suspected industrial carcinogens , asphalt, insecticides, etc.  They investigated many modern substances recently introduced into our society, but found no relationship except with smoking.  They never however, measured the total combined impact of all modern pollution, on lung cancer.

There are ways to measure this impact. In Chemical Carcinogens, chapter 7, figure 3, reports a study of urban/rural differences in lung cancer rates:

City Population50,000+10,000-50,000TownsRural areas
Nonsmoking male lung cancer    14.7/100,0009.3/100,000         4.7/100,0000/100,000
Smoking male lung cancer          85.2/100,00070.9/100,000       71.7/100,00065.2/100,000

According to the American Cancer Society, urban pollution was at least partially responsible for this urban/rural difference in lung cancer even after adjustments were made for higher cigarette consumption among city dwellers. According to “Geographic Patterns in the Risk of Dying and Associated Factors” , U.S. Dept. of Health and Human Services, series 3, #18, p25 females also have highest incidences of lung cancer in metropolitan areas, particularly in portions of the south, northeast, and west coast. Lewis Herber in Crisis in our Cities, 1965, found increased cancer risk of most cancers in our urban areas.  The highest differential was for cancers of the respiratory system with a rate of 20.3 for urban, and 9.3 for rural dwellers. Persons at High Risk of Cancer, p 345 reports the male urban/rural ratio for lung cancer is 1.89 for males, 1.64 for females. All references I have obtained insist urban dwellers smoked more than rural dwellers, but I doubt twice as much. The International Encyclopedia of the Social Sciences , Vol, 14, 1968, p 337 reported rural non-farm persons smoked at the same rate as urban dwellers, but rural farm workers smoked less. As a side note, then the increased likelihood nonsmoking spouses of smokers are city dwellers could easily explain the results of most secondhand smoke studies apart from living with a smoker. I have never seen a “secondhand science” ( not a typo) study than controls for the urban/rural ratio. What I find particularly noteworthy is the 0 instances/100,000 nonsmoking males in rural areas (some of whom must have had smoking wives) , not different from lung cancer patterns a century ago when almost all of us were rural dwellers, but males commonly smoked cigars. 

In sum, back-extrapolating lung cancer time trends from the 1990’s to 1930 leads to a zero intercept around 1900.  Records from 1900 confirm lung cancer was very rare in 1900 but male tobacco smoking was common during all of the 1800’s. Historical Statistics of the U.S. House Document # 93-78 part 1 reports 4.07% of our population was aged over 65 during 1900, 9% in 1970. Enough people did live long enough to develop lung cancer at rates approximately half of today’s if tobacco had caused lung cancer during the 1800s. Claims are common lung cancer was not diagnosed during the 1800’s because X-ray technology was not available. This is nonsense. Lung cancer is too painful in it’s advanced stages to be missed by physicians; even during the 1800’s. Certainly, during autopsy, lung cancer would have been detected in these patients leading 1800’s physicians to conclude it was prevalent. Other cancers were widely recognized during the 1800’s. Today, even nonsmokers get lung cancer in urban areas.  For all the world, it appears we can have smoking without lung cancer, but we can’t have modern urban pollution without lung cancer.

This urban pollution factor holds up internationally too. Cancer Epidemiology: Methods of Study, The John Hopkins Press,  reported a study by Dean that ” immigrants to South Africa from Great Britain had a mortality rate from lung cancer about two times higher than South African born white men of the same age group and similar smoking histories.”  Persons at High Risk of Cancer, The National Cancer Institute, reported in chapter 21 migrant studies conclude “lung cancer risk of migrants from countries with high and low risks are displaced to a level intermediate to home and host populations”.

I have examined lung cancer incidence in Europe where until recently, there has been no War on Tobacco, and the results are surprising.  They’re  doing better than we are. From SAUS, 1992, T1366 and SAUS 1999, T1353,  I obtained age-adjusted lung trachea and bronchus cancer death rates of most European nations during the 1988 and 1994 periods. From SAUS, 1992 T 1361, life expectancy for 1991.  From the British Medical Journal, April 22, 2000, p1102 Cavelaars, et al , I obtained ever smoking and current smoking data of males and females 45-74 years old for 1990.  From SAUS,1999 T226, I computed American current, 45-74 years old smoking rates. I obtained USA ever smoker data from Harris.

CountryLung Cancer rate/yearLung Cancer rate/yearLife Expectancy%Current Smoker  M/F% Ever Smoker  M/F
W. Germany36.2/199535.8/198975.834.8/18.574.9/34.1
G. Britain47.0/199553.8/198976.536.5/27.678.4/51.4

USA male ever smoking rates are a little lower than most of these European nations; USA female ever smoking rates are higher than some European nations, but lower than others. As a result of the War on Tobacco, USA current smoking data are much lower than most other nations.   Yet between 1988 and 1994 our lung cancer death rate  increased slightly while Denmark and the Netherlands , the heaviest smoking nations declined.  Respiratory cancer is declining as an average of all the European nations, and probably stabilizing in America. Considering these numbers represent 275 million Americans and   approximately 300 million Europeans, basing conclusions on these data is not subject to confounding errors common in smaller studies.  All of these countries have modern cancer registries. We share about 65% of our genetic heritage with these Europeans, and about as similar a lifestyle as can be found between any two large populations, except they have not had a War on Tobacco and fewer have quit. If the War on Tobacco has value, that value should be detectable in these comparisons particularly with such a strong statistical association as is generally recognized between smoking and lung cancer.

The above data also provide insight concerning the possibility U.S. lung cancer rates would have continued to rise had it not been for the War on Tobacco. U.S. male rates peaked during 1985. Female rates appear to be peaking now. Had it not been for the War on Tobacco, we could assume our lung cancer would still be rising, but, since Europe is experiencing  lung cancer declines we can conclude our stabilization of lung cancer rates is not a result of the War on Tobacco. We can also conclude our societal costs of treating lung cancer have not benefited as a result of the War on Tobacco. The Europeans are saving more money treating lung cancer than we are.

The War on Tobacco has also not benefited our life expectancy, relative to the Europeans. Most of these countries have higher current smoker and ever smoker rates than we do. Yet they also have higher life expectancy except Portugal which, amazingly, is the country with the lowest smoking rate.

I must confess, I’m astounded that at least some confirmation of the statistical relationship between smoking and lung cancer ought to be realized when comparing heaviest smoking Americans with lighter smoking Americans and lighter smoking Americans with heavier smoking Europeans.  Published scientific studies consistently conclude these relationships are real. Tobacco: A Major International Health Hazard, Oxford University Press, 1986 summarizes these studies .

First, a dose-response relationship exists. Petro reports lung cancer is correlated with average packs/day smoked and years of smoking.

Second, after cessation, annual risk of lung cancer drops and is inversely proportional to years since cessation.

Third, tobacco smoke is bioactive and concentrated tar does induce skin cancers on laboratory animals.  However, inhalation studies are not so consistent. Tumors of the upper respiratory tract have been induced in rats and hamsters, according to Hoffmann and Wynder, but they reported in Chemical Carcinogens ” Inhalation studies have not yet clearly demonstrated that the exposure of animals to tobacco smoke leads to carcinoma of the lung.”

This apparent disparity can possibly be explained if one considers how skin cancers are induced in laboratory animals. Tobacco smoke is collected, concentrated, and then typically dissolved in a solvent. The tar dissolved in solvent is then painted onto the skin of the animal. Various solvents have been used. Some are now known to be carcinogens even without tobacco tar dissolved in them.  Acetone is commonly used. Acetone is probably not carcinogenic, but can react with many of the classes of compounds present in tobacco smoke to produce compounds that may be carcinogenic and have never been individually tested. Organic Chemistry, by Frank C. Whitmore, reviews some of these possible reactions.  Acetone can undergo what is known as aldol condensation with aldehydes commonly present in tobacco smoke. Formaldehyde and secondary amines, present in tobacco smoke can combine with acetone forming what are known as Mannich reaction products.  Glycols, formed by the partial combustion of sugars can react with acetone forming cyclic acetals and dioxolanes. Acetone reacts with ammonia and derivatives of ammonia, these include amino acids and amines present in tobacco smoke. Acetone condenses with reactive aromatic compounds (including PAH’s) forming bridge compounds. Hydroxyacid amides react with acetone to produce oxazoles.  Acetone reacts with cotarnine, an alkaloid, to produce open-chain compounds. It may also react with nicotine.  Flavor and fragrance chemists know flavors and fragrances are also composed of aldehydes, ketones, hydrocarbons, amines and glycols. It is commonly agreed these are not stable on storage, particularly when subject to light, moisture, trace metals, and oxygen. According to The Essential Oils, Vol. 1, D.Van Nostrand Co. Inc., Guenther, ed. ,reactions like the ones I propose for stored tobacco smoke extracts are common in stored flavors and fragrances. This paragraph is not meant to state with any degree of certainty these reactions actually occur, but I have never read of any attempt by those who study tobacco smoke extracts to be certain they do not. This possibility may explain why induction of skin tumors via tobacco tar dissolved in acetone more often produces tumors than inhalation studies using unaltered tobacco smoke.

What most people don’t realize is the same observations made about tobacco smoke have also been made about urban pollution.

First, a dose-response relationship also exists. The data presented on page 14 from Chemical Carcinogens, chapter 7 shows a clear dose relationship exists between size of urban center and lung cancer risk.

Second, cessation of exposure to an urban center reduces lung cancer risk. Dean’s immigrant study to South Africa demonstrates smokers and nonsmokers had more lung cancer if born in Britain before immigrating to South Africa, but less than those remaining in Britain.  In Persons at Risk for Cancer, the National Cancer Institute, chapter 21 reported migrants’ risk was displaced to a level intermediate to home and host populations.

Third, animal studies of urban pollution confirm urban pollution is bioactive. Hoffmann and Wynder reported in Chemical Carcinogens they induced skin tumors in mice from organic fractions of air pollution from Detroit and New York City. They reported others had had similar success using air from other cities. ” Several potential contributors to urban pollution, such as extracts from chimney soot, road dust, and gasoline and diesel exhaust tars were carcinogenic in the experimental animal”. They report carcinogens present from these sources are oxidized by sunlight and ozone to produce even more potent carcinogens, not present in tobacco smoke.  Iron oxides, commonly present in urban air facilitated the carcinogenicity  of active PAH ( polynuclear aromatic hydrocarbons) as carriers for the induction of bronchogenic carcinomas in hamsters. Many researchers believe the carcinogenic effects of PAH’s are dependent on the enzymatic action of  aryl hydrocarbon hydroxylase .PAH’s are present in tobacco smoke and urban air pollution, but oxidation by ozone and sunlight circumvents the need for this enzyme for PAH’s to promote carcinogenicy in the lung from air pollution.  Huber, in Seminars in Respiratory Medicine, 1989 summarizes ” Overall, however, these bioassays for tobacco smoke have been inadequate, or perhaps ineffectively and inappropriately utilized; they have not provided acceptable analogues or models of lung cancer observed in humans, when, in fact, inhalation studies with some of the same kinds of animal bioassays and with environmental toxins not related with tobacco smoke do result in lung cancer similar to that which occurs in man”.

Everything that can be said about tobacco smoke is also true of urban pollution.  Asbestos and radon are also widely known to cause lung cancer. Yet when it is time to blame someone for the costs of treating lung cancer, the smoker gets the entire bill in the form of increased taxes, passed on expenses of lawsuits against the Tobacco Industry,  and the Tobacco Settlement. 

Urban pollution became prevalent long before lung cancer did.  Steel and coal pollution was prevalent by the mid 1800’s in our urban canters.  Wood smoke and coal smoke belched from our homes in urban centers more than 70 years before lung cancer rates began to rise in the 1930’s. The only kind of urban pollution that can explain time trends of lung cancer is motor vehicle.

Lets consider, lung cancer began to rise about 30 years after the introduction of the automobile. Males were primarily the users of the first automobiles and the first lung cancer victims.  Most of the first female smokers were stay-at-home moms. Harris  did report females born 1905 smoked about 24% during their entire lives, yet data presented previously indicate their lung cancer rates were disproportionately  low. Harris also reported those females had the same life expectancy as nonsmokers and that other researchers agree with him.  After the 1930’s, females began to drive, and 30 years later, during the 1960’s, their lung cancer rates began to rise. One could explain  the lower European lung cancer rates I presented when one considers they drive less than we do. They tend to use public transportation more and drive smaller, more efficient  cars when they do drive.  Although data are not presented, it is common knowledge the Japanese smoke much more than we do and get less lung cancer. Their motor vehicle use approximates the Europeans. Finally, our lung cancer incidences may not be responding to our smoking declines the way they should because we continue to increase our motor vehicle use.

The impact of motor vehicle pollution on lung cancer rates is just beginning to be realized. Recent media releases now implicate diesel exhaust as a carcinogen.  Unfortunately, diesels are exempt from pollution controls imposed on cars. Niles and Tan, U.S. Dept of Energy, in Analytica Chimica Acta, v. 221, (1989), found PAH’s in diesel exhaust, but more importantly, found nitrated PAH’s which they say are some of the most potent mutagens yet identified. For example, 1-nitropyrene has mutated bacteria without enzymatic action  and caused tumors at the site of injection in rats.  I suspect many high compression gasoline engines , used in cars during the 1950’sand 1960’s , also produced these emissions.

Your favorite corner gas station pump now carries a warning “Gasoline vapors have produced cancer in laboratory animals”, but there is much more.

Malcolm Pike, in Persons at High Risk of Cancer calculated every 10 nanograms/cubic meter of benzo-alpha-pyrene in city air is the equivalent of one cigarette smoked per day. He found 61 ng/cuM in air over Altoona,PA, and as little as 0.5 in Salt Lake City UT.

Ethylene dibromide was added to gasoline quite commonly until the middle 1980’s to scavenge lead built up from leaded gasoline use.  The white substance observed in the tailpipes of cars built in the 1960’s was lead bromide. EDB is volatile and was breathed by those exposed to gasoline fumes filling their cars or servicing them. According to Cancer and Chemicals, it has been found cancerous in both mice and rats often developing cancer as early as ten weeks after the start of the experiment. It has also been known to cause skin and eye injuries, respiratory tract inflammation, and even fatal heart problems.

Cancer and Chemicals also reports benzene in gasoline as high as eight percent. Benzene is most often associated with leukemia, but is now recognized as a general carcinogen.

Recently tire particles have been found to cause asthma in urban dwellers. www.s-t,com/daily/11-95/11-08-95/1108allergy.ht. Researchers at Bucknell U. have also found tire particles alongside our roadways and in our waterways. http://www.collegenews.org/x1939.xml. According to Chemical Carcinogens, rubber workers are at increased risk of bladder cancer because 2-napthylamine is used as antioxidant in tires. Carbon black is also added to tires, and rubber workers are at increased cancer risk from it. Chimney sweep’s cancer was attributed to soots, of which carbon black is a class member.  Tires contain cadmium. Huber states “Cadmium oxide, even in low accumulative dosages, induces the synthesis of metallothioneine within the lung; metallothioneme is a potent chemotaxin thought to be important to the recruitment of polymorphonuclear leukocytes as mediators of an inflammatory process in the pathogenesis of emphysema and chronic obstructive lung disease”. Weisburger in Chemical Carcinogens, listed cadmium as a direct-acting carcinogen.

Asphalt is classified by RTECS as a tumorigen and mutagen. (www.cdc.gov/niosh/rtecs/ci970feo.html ) It is made from the high boiling fraction of crude oil by partial oxidation to harden it. Kipling in Chemical Carcinogens ,describes this high boiling fraction as the fraction with the most carcinogenic activity.  Persons at High risk of Cancer reports persons with the occupation of gashouse worker, stoker, producer of asphalt, coal tar, and pitch workers, miners, still cleaners, and chimney sweeps have 2 to 6 times the risk of cancers of the lung, larynx, skin, and urinary bladder.  When a road is freshly paved, it is black, but after wear the crushed limestone component becomes exposed, and the surface becomes trenched .  Where does this asphalt go? You breathe it.

No review of risk of lung carcinoma from motor vehicles is complete without mention of asbestos use in brake pads.  Recently class action lawsuits have been filed on behalf of auto mechanics exposed while repairing brake systems. Mechanics, however are not the only ones exposed. Anyone with rally-style wheels knows this brake dust blows out of the brake assembly onto the wheel necessitating regular cleaning for appearance purposes.   The largest particles dissipate the least, the smallest travel the furthest, and probably breathed by persons occupying other vehicles .Hoffmann and Wynder report urban particulates contain silicates, metal oxides and carbon. Asbestos is one of these silicates.

What is less commonly known about braking systems is the nickel risk. Drums and rotors are commonly alloyed with 20% nickel which toughens the metal and makes brakes safer. As with brake pads, these drums and rotors do wear out. Brake dust is composed of not only asbestos, but also iron and nickel. Hoffmann and Wynder devote several paragraphs to the respiratory carcinogenesis of nickel. Animal experiments are positive for the free metal and most compounds tested. Persons at High Risk for Cancer reports iron oxide dust is also a lung cancer risk. Iron ore miners, metal grinders, and polishers, and iron foundry workers are at 2 to 5 times the risk of lung cancer.

I grew up in a small city and lived in a rural area the first ten years of my adult life. When I moved to a large city, I noticed this “road film” would build up on my windshield. I never had this problem previously. Glass cleaner does a poor job removing it, but pure organic solvents such as paint thinners do a great job. PAH’s don’t dissolve well in water-based glass cleaners, but are quite soluble in paint thinners. Given the findings that PAH’s , nitrated PAH’s and partially oxidized PAH’s are present in vehicular emissions and urban air, I suspect road film is composed of  these compounds. I noticed cleaning my windshield wiper blades keeps them from streaking when it rains. Cleaning them with paint thinner on a paper towel leaves “black gunk” on the towel.  This is probably tire and asphalt dust. New wiper blades don’t produce this “black gunk” when cleaned. When you change the engine intake air filter on your car, look at it. Over time, it will become clogged with this black gunk. Since fresh air intakes of automobiles are located proximate to the wiper blades, these same road films and black gunks are drawn into the passenger compartment and breathed by the occupants.

Hoffmann and Wynder reported aza-heterocyclic compounds, epoxides, paraffins, olefins, peroxides, and epoxides in urban pollution. These are not present in tobacco smoke or wood smoke. Our lungs have not had 50-thousand years to develop resistance to these compounds and the others mentioned previously. The recent appearance of lung cancer coincides with the introduction of these compounds into the air we breathe. A sensible person would expect at least some lung cancer should be attributed to urban pollution.

The time trend behavior of lung cancer strongly suggests a multivariant mechanism. Hoffmann and Wynder postulate when the physiochemical defense mechanism of the lung is impaired, such as exceptionally thick mucus or impaired cilia action, then the residence time of inhaled particles is sufficient to allow carcinogenic agents to diffuse from the particles into the underlying cells. Smoking certainly thickens mucus and impairs cilia action. Thus smoking probably enhances vulnerability of the lung to cancer from these urban pollutants.  Smoking is well known to be synergistic with asbestos probably by slowing the elimination of asbestos from the lung. Cigar smokers didn’t get lung cancer during the 1800’s because modern urban pollution from vehicles didn’t exist. Today, studies indicate cigar smokers do get lung cancer, probably because the essential ingredient, vehicle pollution is now available to complete the process.

I like my car.  I’m not trying to shift all the blame for lung cancer off smoking and onto motor vehicles, but manufacturers of vehicles and fuels , and owners /users of motor vehicles are not expected to pay any of the societal costs of lung cancer.  Our holier-than-thou civic leaders, health professionals, media fanatics, and anti smoking activists all use motor vehicles but don’t accept any of the blame for our lung cancer epidemic because then they would have to accept some of the responsibility of the costs of treating it.  For these reasons, public attention continues to be drawn away from our motor vehicle problem even as real science dictates we should do otherwise.

In sum, motor vehicle pollution, and country of residence have sufficiently confounded the strong statistical association between smoking and lung cancer such that no reliable relationship exists between our past or present smoking / lung cancer rates that can be used to predict what will happen if smoking declines in the future. Then how can the weaker statistical relationships between smoking and other diseases, widely acknowledged to have confounding variables, be used to make public promises of improved health if the War on Tobacco is widened and intensified?


If differential smoking histories between various American birth groups can’t be used to demonstrate the impact of smoking on public health, then no major impact exists. This conclusion is further enforced by observations foreign populations don’t correlate with health claims made by our public health officials concerning our smoking risks of lung cancer.  Observations made in this article are in direct conflict with established studies. Therefore the individuals studied in our medical literature must not be representative of all Americans. Measuring health trends of populations negates the concern that biased data are being used because these trends include all confounding variables in proportion to their actual impact . The possibility even lung cancer is subject to confounding variables is demonstrated, and explains why even lung cancer doesn’t follow smoking trends. In this article, a much larger database is used which results in more reliable data than smaller studies.  Perhaps  an equivalent number of negative studies have been performed that do not enforce the conclusions of mainstream medical literature. Then averaging such studies with those which have been published would reach conclusions similar to mine. Given the mentality of editorial boards of our major medical journals, getting negative studies published would be virtually impossible.  I’m not sure an aspiring medical researcher would want his/her name on such a study. After spending more than $100,000 on a medical education, publishing such a study would be committing career suicide. Who would fund more research of scientists publishing negative studies?  I believe the data provided in this article proves our medical literature is biased.

Among American males, most smoking related disease during 1973, occurred among those born 1895 and 1905.  In addition to smoking more unfiltered cigarettes than latter birth groups, they also smoked cigars and hardly any quit before age 60. Most smoking related disease during 1996-1998 occurred among those born 1915 and 1925. They smoked more filtered cigarettes, and fewer cigars. Some quit before age 50 and many quit prior to age 60.  Their laryngeal and pancreatic cancer rates are lower, but no differences were realized for cancers of the lung, colon, oral cavity, bladder and esophagus. No difference was observed for any chronic conditions despite more quitting at the age where these conditions usually are manifest. They did not gain as much life expectancy/decade, after age 50, as the heavier smokers born 1895 and 1905 did.  Their quitting in latter years did not save society any costs of treating their diseases.

Among American females, those born 1895 and 1905 smoked less than those born 1915 and 1925.  Most of the earlier birth groups had died off by 1985, leaving only females with steady smoking rates at risk of smoking related diseases. Their lung cancer rates continued to rise despite steady smoking exposure, thereafter. Other smoking related cancers followed the same trends as males after 1985 even though their lifetime smoking exposure was steady while males’ was declining.  Their chronic condition prevalence was similar between 1980 when the lighter smokers were still alive, and 1996 when only the heavier smokers were still alive. The lighter smokers gained more life expectancy/decade after age 50 than the latter heavier smokers, but the heavier smokers still lived longer.  Their heavier smoking may have cost society more to treat their lung cancer during 1998, but society spent less treating their other smoking related cancers than those of the earlier, lighter smokers.  Society spent the same to treat their chronic conditions as the earlier, lighter smokers.

Scientific theories should be tested to insure their accuracy. Laboratories involved in medical, forensic or environmental testing routinely test their methodology to insure they are producing reliable results. Control samples, or spiked samples are always included in tests to prove no confounding factors are present that would invalidate results. Unlike smoking studies, these laboratories are regulated. Other branches of the science of medicine are also regulated, but statistical studies are the scientific equivalent of the Wild, Wild, West.   In statistical studies of smoking, when data exist conflicting with accepted theory, they are ignored. It’s bad enough that accepted principles of science are abandoned; it’s bad enough that smokers are being financially penalized for scientific principles that don’t exist; but even worse, we don’t alleviate suffering from these smoking related diseases while we bear the hardship of the War on Tobacco.

In closing, Principles and Procedures of Statistics, 1960, Steel and Torrie lists the features of the scientific method:

  1. A review of facts, theories, and proposals    (this was done)
  2. Formulation of a logical hypothesis subject to experimental testing    ( this was done)
  3. Objective evaluation of the hypothesis on the basis of experimental results ( this was not done)

Download poster


  • "Es ist schwieriger, eine vorgefaßte Meinung zu zertrümmern als ein Atom."
    (Het is moeilijker een vooroordeel aan flarden te schieten dan een atoom.)
    Albert Einstein

  • "Als je alles zou laten dat slecht is voor je gezondheid, dan ging je kapot"
    Anonieme arts

  • "The effects of other people smoking in my presence is so small it doesn't worry me."
    Sir Richard Doll, 2001

  • "Een leugen wordt de waarheid als hij maar vaak genoeg wordt herhaald"
    Joseph Goebbels, Minister van Propaganda, Nazi Duitsland

  • "First they ignore you, then they laugh at you, then they fight you, then you win."
    Mahatma Gandhi

  • "There''s no such thing as perfect air. If there was, God wouldn''t have put bristles in our noses"
    Coun. Bill Clement

  • "Better a smoking freedom than a non-smoking tyranny"
    Antonio Martino, Italiaanse Minister van Defensie

  • "If smoking cigars is not permitted in heaven, I won't go."
    Mark Twain

  • I've alllllllways said that asking smokers "do you want to quit?" and reporting the results of that question, as is, is horribly misleading. It's a TWO part question. After asking if one wants to quit it must be followed up with "Why?" Ask why and the majority of the answers will be "because I'm supposed to" (victims of guilt and propaganda), not "because I want to."
    Audrey Silk, NYCCLASH