Recreating the Doll Doctors Study Graph

I said a couple of days ago that I would have another go at creating the graph which showed mortality in relation to heavy smokers, moderate smokers, light smokers and non smokers.

What drew me actually redrawing that graph was that I noticed that the two axes, the x and the y, were not to the same scale. I’ll show you what I mean:


By the way, that graph is of survivors. I had to change it round to convert it into deaths.

If you look carefully, you can see that the ten percent divisions on the vertical axis correspond with about five year periods on the horizontal axis. [There is nothing wrong with that, but it can give false impressions] Also, you will note that the horizontal axis starts at age 35. [Again, nothing wrong with that in itself] Even so, since the human life-time is maximised at about 100 years, it fits nicely into a ‘square’ graph with equal axes of ages and percentages.

My first effort was fine, but I made a couple of (rather insignificant) errors. One was to chose my reference points in a rather higgledy-piggledy manner. Another was to clutter the graph with text. There isn’t much difference between the two graphs, except that the new one one is ‘smoother’.

The original graph with which I was working was actually a little smaller than the one above and there were no grid-lines (the ones in the pic above are mine), and so you can see the real difficulties of picking out reference points. Her is the revised graph:


What you can see immediately when you compare the two graphs is that, what seems to be a gentle, gradual pegging out of heavy smokers years before non smokers is nothing of the sort. Nothing much happens at all until around the age of fifty, then, fairly suddenly, things start to happen. At 50, 2% of non smokers had died and 8% of heavy smokers with moderates and lights somewhere between the two. At 60, 8% of non smokers had died and 18% of heavy smokers,  but light smokers were only 12%. At 70, non smokers were 18% dead and light smokers were 31%.

You may say that the graph shows that smoking is pretty awful, and you may be right, but we must remember that no other factors have been taken into account. What about alcohol consumption? What about air pollution? What about locality of both work and home? What about war experiences? (I am considering only external facts here and not stuff like genetic susceptibility) But what intrigues me most is the parallel nature of the lines once people start to peg out. Do you see the significance? It means that heavy smokers, moderate smokers, light smokers and non smoker were all dying at the same rate, even though non smokers survived for longer before they started to peg out at the same rate. What ought to have happened, if smoking was so bad, is that the lines should have continued to diverge as they did between the ages of 50 and 65.  They should have fanned out. What that suggests to me is that non smoking, at best, merely postpones the inevitable.

But there is another point worth observing, which is that Doll’s definition of light smoking was between 1 and 14 cigs per day. In today’s climate, I think that light smoking would be confined to less than 10 per day. I think that nowadays, there would have been at least four categories: 1 – 9, 10 -19, 20 – 29, 30 +. It could even extend to five by going 30 – 39, 40 +.

Also, there is another peculiarity which is the statement below Doll’s graph. It is hard to read. Here is what it says:

Overall survival but with cigarette smokers subdivided by the amount they were smoking at the time their last questionnaire was returned”

I suppose that, if heavy smokers were going down like ninepins, Doll had no option, but who is to know when the last questionnaire was issued and returned?

There is another observation worth noting which is that, at the age of 80, there were still 20% of the original heavy smokers still alive. Why? How did they come to survive into very old age? What was it about them which defied the consensus?

Also, we have Garyk’s calculations in the previous post a couple of days ago which indicate that there is only a fractional difference between smokers who die from ‘smoking related diseases’ and non smokers who die from the same diseases. It may be true that smokers die earlier from these diseases, but non smokers still die from the same diseases.

Further, in the longer view, McMillan estimates that more and more people will be living with cancer as time progresses and more and more people live longer and longer. It has been estimated by McMillan that some 40% of aged people with be living with cancer in a few years time.


Whenever I think about these things, I cannot help but return to the McTear Case. Here was a guy who contracted squamous cell carcinoma of the bronchi and died at the age of 53 or thereabouts. His ‘CV’ indicates that he was a right character – couldn’t hold down a job, thrown out of the army, constantly in trouble with the police, set fire to his own house a couple of times, had a family history of LC, but smoked a lot and was chosen by ASH ET AL to be the test case in the the McTear Case. ASH ET AL must have been struggling to find a single person who qualified to be the test case if they had to chose McTear. It looks to me a though they had no candidates who were living quiet lives and being very good but still got LC and smoked a lot. If they had, they would have chosen such a person, wouldn’t they?


What is becoming more and more obvious by the day is that any minority, whether they be smokers, fatties, or mental patients, cannot expect to receive the support and protection of politicians; indeed, condemnation and persecution is what they can expect. Will there be any consolation? Well – no – until a modern day Wilberforce stands up and condemns the slavery and persecution.


16 Responses to “Recreating the Doll Doctors Study Graph”

  1. Tony Says:

    Just a trivial correction. From the McTear judgement:
    [1.2] Alfred McTear died, aged 48, on 23 March 1993.

    • junican Says:

      Thanks for that, Tony. I searched and searched for that info but could not find it. How did you discover it?

      • Tony Says:

        I was actually looking for something else at the time. I was sure I had read somewhere that ASH had a further 1,100 cases lined up ready to go after the McTear case which they were expecting to win. I think, in your summary, you gave a figure of 300. Anyway, I couldn’t find the 1,100 figure but happened to notice Mr McTear’s age while I was looking.

      • junican Says:

        They did have several hundred to chose from but the great majority were ‘out of time’. My memory of the numbers involved is hazy. I think that they finished up with about 20 – 30 which may have been viable and they chose McTear. But I’m not sure.

  2. Frank J Says:

    Well done, Junican, excellent. Simple, straight forward and understandable.

    I assume these are all deaths and not just LC. To me, it shows the paucity of their arguments. It is significant not for what it specifically tries to say but what it doesn’t say and the mass of relevance it leaves out. We can see why Sir Ronald Fisher condemned it as only being worthy of further investigation and we can also see why the great Sir Richard himself failed to submit it as evidence in the McTear case.

    If the whole study was supposedly an investigation into causes of ‘deaths’ it is amateur and wholly inconclusive, a shocking attempt at an impartial study, at best, a starting point. But it is obvious that it wasn’t meant to be impartial. It was a policy driven study to condemn smoking, searching for any reason at all in support. To any rational being it doesn’t even achieve that successfully. For most of my life I’ve accepted an increased ‘risk’ because I smoked. Having seen these figures and their lack of substance, no longer will I believe in any increased risk due, solely, to smoking. Overall, it’s rubbish.

    It would be torn apart, legally. Politicians, however, are a different breed, Easily influenced and manipulated.

    • junican Says:

      It isn’t just lung cancer – it is all mortality.
      I agree that it was motivated because a couple of other big studies were commenced about the same time – around 1950. Doll said that there was no intention to keep the study going for so long; I think that twenty years was the expected extent. You can see from that that there was a political purpose involved, probably connected to the WHO. I suspect that there was a deliberate intent to build up ‘the body of evidence’ and that the health depts of the UK and the USA were especially involved.

  3. Linda Says:

    I was fighting against a cancer stage 4.I think it is very important that family support to win, because i was very weak;really helped me participate in one group of affected by cancer, so my mood improved, also helped me a lot a medical adviser in (they are doctors),this is important .I recomended not surrender, because sometimes the first treatment does not work as me, and sometimes change doctors it is necessary.Read positive thinking books gave me more energy.During my cancer,i changed my diet,now i eat vegetarian organic food(now i not eat meat).I think is a set of things that help.

  4. garyk30 Says:

    I may be reading this chart poorly; but, it seems to me that the avg age of death for the n/s was about 82 and the avg age of death for the moderate smokers(m/s) was about 76.

    If 50% of the m/s died between 76 and 100, 24 years, the average rate was 2% per year.

    About 10% of the m/s would have died between 76 and 82 and 40% would have lived beyond 82.
    (2%/year x 5 years = 10% and 50% – 10% = 40%)

    Thus, 40% of the m/s lived beyond that age compared to 50% of the n/s.

    40% of m/s vs 50% of n/s lived beyond the median age of death for the n/s.

    60% of m/s vs 50% of n/s died pre-mature deaths.

    A. m/s are only 20% more likely to die pre-maturely than n/s.

    B. That is a relative risk of 1.2

    C. A relative risk of 1.2 is not statistically significant

    Both m/s and n/s have the same probability of pre-mature death.

    Altho smoking is claimed to be the biggest ’cause’ of pre-mature death, m/s and n/s have the same chance of dying a pre-mature death!!

    • garyk30 Says:

      A pre-mature death being one that occurs below the average age of death of the n/s.

      • garyk30 Says:

        Anti-smokers seem to believe that you can, somehow, have less than 50% of deaths be pre-mature. LOL

      • junican Says:

        The Zealots deliberately claim ‘premature’ death by comparing smoker deaths with non smoker deaths. They take no account of the possibility that the later deaths of non smokers may be out of step. Perhaps non smoker deaths should not be taken as the ‘ideal’ or the ‘norm’. Perhaps the lifestyles of non smokers are not that pleasant. Or rather, perhaps their lifestyles were not that pleasant during the period of this study, especially in its early years. Sure they lived longer, but were they happier?

  5. Frank J Says:

    I feel all these figures are a straw man. There is no such thing as a ‘normal’ being or life. Heavy smokers e.g. supposedly die earlier. But why are they ‘heavy’ smokers? are they more susceptible to stress and its effects? or are they just what used to be called nervous? Are non smokers such because they are cautious and take fewer, if any, risks going about their lives? Do they pace themselves in a slower fashion? What are their relative adrenalin levels? What sort of lives do they, respectively, lead? What sort of wear and tear do they give themselves? What is their definition of ‘quality’? And, importantly, what effect do any these things have on them?

    Notwithstanding many other physical factors and feasible combinations, these questions are not addressed in the above ‘survey’

    It’s a straw man and complete bollox!


    • junican Says:

      At the age of 60, 25% of heavy smokers had died. Why had those particular individuals died and not the other 75%? At the age of 70, 45% of heavy smokers had died. 55% had not. Again, we ask why had those who had died died, and why had the survivors not? Even at the age of 80, 20% of heavy smokers had still not succumbed. Why not?
      So there are two really important questions:
      1. Why did a smallish number of heavy smokers die comparatively early, and,
      2. Why did a smallish number of heavy smokers survive for so long?

      There MUST have been other factors involved.

      • garyk30 Says:

        God loved the ones that died and wanted them in heaven with Him?

  6. Frank J Says:

    These are the simple arguments that have been around for 50 years and have failed to be answered. No attempt at any form of ‘cohort’ is any response as no ‘cohort’ whatsoever can take all into account. Peer review is no answer, either, as they can only pass opinion based on the parameters of the exercise. That is why the original surveys, both Hospital and Doctor’s, were treated with such disdain over that period. After 50 years of being swatted away, the antis have ‘overcome’ them by, simply, not allowing them to be heard.

    It’s quite simple, really. Doll’s survey is as much bollocks now as it was then and only SHS has carried the day by changing the argument, what we used to call the ‘Assumptive Close’. It’s a distraction. If SHS is dealt with, it all falls. Dead in the water. It should still be the main aim, irrespective of Politician’s views.

    Thought required as to how to keep throwing it at them, again and again. Lets get straight to the meat.

Comments are closed.

%d bloggers like this: