Is Epidemiology ‘Science’?

If you said that Epidemiology is A science, then you could possibly get away with it, provided that you use the word ‘science’ loosely. But it is hard to see how you could describe epidemiology as ‘science’.

Let us take an example.

Dr John Snow, in 1854, observed that an outbreak of cholera was localised to a specific, though not precise, area of London. By mapping the incidents, he was able to pinpoint the source – a specific water pump on Broad St. The people who got cholera used that water pump. But Snow already believed that cholera was not caused by a ‘miasma’ in the air, but was caused by contamination of the water supply by sewage.

Did Snow perform ‘science’? Well, not exactly. He found the source of the cholera outbreak, and he found the ’cause’ in the sense of sewage contaminating the water supply at that pump, but he had not found the cause of cholera. It took some 30 years before the actual cause of the disease known as cholera, which caused the deaths, was discovered:

In 1883 a German physician, Robert Koch, took the search for the cause of cholera a step further when he isolated the bacterium Vibrio cholerae, the “poison” Snow contended caused cholera. Dr. Koch determined that cholera is not contagious from person to person, but is spread only through unsanitary water or food supply sources, a major victory for Snow’s theory. The cholera epidemics in Europe and the United States in the 19th century ended after cities finally improved water supply sanitation.”

It was Koch who performed the science. But it would be unfair to Snow to say that he did not point the way to the actual scientific discovery of Koch. His surmise that it was the sewage which was causing the problem was not actually correct. It was the bacterium, which just happened to be in the sewage.

But where do you draw the line? You could also say that Koch himself only showed that Vibrio cholerae was, in some way, involved. Only if the ‘pathways’ of how the bacterium did the damage would ‘scientifically prove’ that the bacterium actually caused the deaths of the victims.


The Doctors Study showed a strong numerical link between smoking and lung cancer in about one in seven smoking doctors as compared with non-smoking doctors where the numbers indicated only a remote incidence of lung cancer. You could compare that finding with Dr Snow’s identification of the cluster of cases around the Broad St water pump. The Doctors Study, no matter how meticulous, was not science. The science starts ONLY after the Doctors Study. As we saw in the McTear Case in 2005, the ‘science’ is nowhere near complete. In fact, it has not even started. What we have been told, over and over again, is that there are 4000 (and increasing) chemicals in tobacco smoke and that there are 60 carcinogens. However, it has not been shown how any of these substances can cause problems in the quantities inhaled. For example, I doubt that a cholera outbreak would have occurred if the incidence of Vibrio cholerae had been tiny. I suspect that ‘the dose creates the poison’ applies also to bacterial infection, or that bodily defences, weakened by old age or for some other reason, cannot cope.


There is some sense in the Doctors study because the incidence of LC was so much larger in heavy smokers as compared with non-smokers – like 20 (?) times larger. But it is still only numbers. You could reasonably describe Doll’s statement that smoking causes LC in the same way that Snow said that sewage causes Cholera. It’s all right saying that there are 4000 chemicals and 60 carcinogens, but which of the 4000 chemicals or 60 carcinogens actually causes the LC? The McTear Case showed that the ‘Experts’ do not know. Nor can they identify ANY LC  incident which is caused by tobacco smoke. They cannot do it. So, it is possible to infer that the ‘Experts’ are at the stage where Dr Snow blamed ‘sewage’. The anti-smoker Zealots blame ‘sewage’ in the form of tobacco smoke, but cannot say what the ‘bacterium’ is, or, indeed, if there is a ‘bacterium’.

But what is most important is that the Zealots have gradually, and deliberately, distorted the value of ‘Relative Risk’. In the Doctors Study, the RR was 20:1 as regards heavy smokers getting LC as compared with non-smokers. Even that RR was not ‘proof’ of anything. We are now seeing RRs of less that 2:1 being claimed as ‘proof’ of whatever. EG. “People who eat lots of fat are 25% more likely to get X disease”. In that example, the RR is 1.25:1. Any epidemiologist worthy of the name would tear his beard off since the ‘confounders’ are almost infinite.

And yet tobacco control (and especially ecig control) is getting away with such blatant distortion.

Where are the defenders of proper epidemiology? And proper statistics?

They are cowering in the trenches of their jobs, that’s where they are. Having said that, there must be thousands of retired epidemiologists and statisticians who read all this junk and sigh. But what can they do? They are armed with bows and arrows and spears, whereas the opposition is armed with fighter and bomber jets, tanks, nicotine patches and bottles of purified water.


So there are two scenarios which seem to make sense:

1. Let tobacco control destroy itself under its own inherent contradictions (“It is good for smokers to stop to save NHS costs, and it is all right for those costs to be increased by smokers stopping and thus living longer and being old and decrepit longer”).

And at the same time:

2. Find work-arounds just as people found ‘work-arounds’ concerning rationing in WW2. Many a dead rabbit found itself in the pot without passing through some government rationing scheme. There were many ‘rabbit sausages’ going the rounds of butchers during WW2. My Dad, and everyone else with a garden, bred chickens during WW2. It was the patriotic thing to do. Did anyone worry about salmonella? Did they heck! Were there copious regulations? Were there heck! It is quite possible that regulations beget problems as a result of their tendency to embed errors.


Some years ago, when I was worried to death about the consequences of taking herself on the cruise where our granddaughter was going to get married (because of the known and unknown uncertainties), I had a think. This think is not unusual and is often recommended, but rarely followed. The think is: “WHAT IS THE WORST SCENARIO?”

As regards actually going on holiday, the worst scenario is that you do not go. Is that so awful? The worst scenario is that you stay at home. You may lose some money, but that is no more than an inconvenience.

But there is a worse scenario, which is that you get ill on holiday. It could happen to anyone. Just as TC avoids proof of smoking harm, holiday insurance companies avoid the simple matter of illness. Everything is blurred. That, in itself, is also a contamination of epidemiology. Next time that I take herself to Spain, if it ever happens, I shall ensure that no stone is unturned as regards travel insurance. The main thing will be: “What additional cover is there which is better than EU Health common benefits?”

Accept the maxims: a) TC will destroy itself eventually, and, b) that smokers will get smart and stop paying tobacco taxes.

It is really very obvious.


We should all be thankful that we did not face that question in the trenches of WW1.




8 Responses to “Is Epidemiology ‘Science’?”

  1. Frank J Says:

    “so much larger in heavy smokers as compared with non-smokers – like 20 (?) times larger”

    The ‘wisdom of the crowd’ saw several problems with Doll when this came out. First was the difference being much smaller, insignificant in fact, in rural areas which was pointed out by Bessie Braddock in Parliament and, to date, has no response.

    Notwithstanding that you have to die of something, anyway, they also saw that his statistics did not point to a ‘general’ cause. It was only a problem with ‘some’ people. e.g. 0.1 being the non smokers risk and – 20 times larger in heavy smokers which I also remember as being regarded as 30+ per day – rises to 2%. i.e. 98% of smokers didn’t develop it. Even if we take 8% (80% larger risk) it’s still remains that 92% don’t develop it. Surely, if health was the problem Doll et al would then have analysed why it occurred in only a few people and not anywhere near the main body. He didn’t.

    Those were the points raised in those days and, of course, were not answered. They remain unanswered today. That is why people were suspicious and dismissive of both Doll and the use of statistics.. Some of us, still are.

    • junican Says:

      Doll was aware of the problem of the small number of smokers who got LC, but preferred to highlight the high RR. As you would expect, the figures for heart problems as cause of death were much closer together, but Doll somehow managed to claim that an RR of 2 for smokers in that dept was trustworthy enough for him to claim that smoking caused 50% of smoker deaths. That is where the ‘Smoking kills 50% of smokers’ came from.
      Your comment about the urban/rural divide is quite correct. There was also a study in South Africa by Kitty Little which showed that ;smoking related deaths’ were more common in inland cities which suffered smogs than the windy, coastal cities. There was another one which showed that people who live near the top of high-rise blocks suffer less illhealth than those who live near the bottom. Air quality, anyone?
      Interesting,is it not?, that studies of that nature no longer take place? Instead, money is spent in vast amounts on finding slight variations on the ‘tobacco harm’ meme.

    • Some French bloke Says:

      Doll et al always deliberately disregarded the urban-rural divide, which should have been a consideration, even as regards the London Hospitals Study whose subjects and controls came from Europe’s largest Metropolitan Area. Over an area which is half the size of Corsica (or over 7 times that of the Isle of Man), marked variations in population densities and degrees of urbanization (particularly back in 1950), are bound to have been reflected in a significant gradient for LC prevalence, correlated further with the phenomenon of heavy smoking being more common in the central area than in the outskirts. Hence, even a slightly higher cigarette consumption per capita translated to a considerably higher risk of LC, and was eagerly presented as the cause thereof!
      Add to that that their classification of social classes in Table II of the London Hospitals Study didn’t permit identification of (very) possible occupational exposures, and top it all off with a mention of their inept handling of the inhaling paradox, and you get an idea of how a potentially useful but auxiliary science has been turned into a train wreck at the hands of pretentious and reckless professionals.

      While quite a lot of junk science can be pretty innocuous (even proving more entertaining than scary to the enlightened reader), their piece of “research”, in view of its multiple and invariably destructive consequences over the decades, belongs with the toxic waste of junk science.

      • junican Says:

        Indeed! I should imagine that finding out where people lived and what work they did would have been easy to discover when questioning them. I wonder if Doll did that in the Hospital Study but ignored the results? He was certainly aware of the Doctors’ home addresses in the Doctors’ Study.
        And he dropped the inhaling question!

  2. raven921 Says:

    Is it science?
    Well you tell me,try doing the first part of this class and tell me if you can get past the obvious bias in part two.
    I couldn’t.
    The course is free and I did learn quite a bit about epidemiology and how it worked in the past where it had a real ability to help track down what was killing the poor worldwide.
    Oh it didn’t cure anything but then it moved on,it morphed.
    It decided like all the religious nutters that there was so much more work to be done….

  3. garyk30 Says:

    ” By 10 September, 500 people had died and the mortality rate was 12.8 percent in some parts of the city.”

    Solved by removing the handle from a water pump. There were no ‘Draconian’ restrictions on peoples’ lifestyles.

    Smoking ‘related’ mortality is (USA) 450,000 out of a population of 315,000,000 or 1/10th of 1%.
    The anti’s want 20% of the adults to alter their lifestyles and severe rules on how businesses can be operated.

    Lung Cancer death is a very rare event.
    In the USA, there are 160,000 yearly deaths in a population of 315 million or 5/100ths of 1%.

    Increased risk of a very rare event is still a very rare event.

    If one compares the chances of ‘Not Dying’ from Lung Cancer, in any given year, we find two numbers:
    1. Doll’s Doc Study: current smokers had 99.8% of never-smokers chances of ‘not’ dying from lung cancer.

    2. USA data shows a 99.95% factor.

    Since I am not scared by elusive shadows, I refuse to be ‘panicked’ by the fear of a very slight risk of lung cancer ‘related’ to my smoking.

    Here is a Relative Risk to consider.
    In Doll’s Doc study, compared to smokers, never-smokers were only 1.07 times more likely to not die from a smoking ’caused’ disease.

    As was pointed-out the other day, it is very strange that folks that have smoked for 50 years are diagnosed with lung cancer at the same age as those that have never-smoked.

    • junican Says:

      I think that a big problem revolves around the fact that, despite all the research and masses of jargon involved, no one really knows what causes a cell to become a ‘zombie’. (What I mean by ‘zombie’ is that it refuses to die when it should, but serves no useful purpose except to reproduce).
      If I was thinking very broadly, I would surmise that it would be possible to discover the cause of some specific cancer, and that it might be possible for a ‘cure’ to be effected. However, only comparatively young people would have the physical ‘robustness’ for the treatment to work.
      I am struggling to explain what I mean.
      When we get old, infirmities take longer to sort themselves out. Can we call that ‘metabolism’? I am 75. It is noticeable to me that cuts take a lot longer to heal, the stiffness from physical activity takes much longer to relax, my joints ache more easily in the cold. My ‘metabolism’ is slowing down. If a cancer ‘cure’ depended upon a strong metabolism, then that ‘cure’ would not work for old people. In other words, death is still inevitable, if only because the cures that are effective in younger people are not effective in older people.
      But none of that will happen if research funds continue to be directed into prohibitions.

Comments are closed.

%d bloggers like this: