Hill and Doll Doctors Study and Deaths from Lung Cancer since 1980

First, can I say that this concerns only the United Kingdom. World-wide ramifications are just too complex to consider.

I have been pondering the Hill and Doll Doctors Study for some time. I went to some trouble to précis the reports which Doll published in the British Medical Journal and the results of the précis are in the sidebar.

I believe, with some reluctance, that we have to accept that the original data upon which the Doctors Study was based were genuine and not fiddled. It is a terrible thing to even suggest that Doll fiddled the figures, but he was a zealot of the most zealous kind. But there were other people involved, and it is unlikely that these people would have stood for blatant fiddling, and so we must discount that possibility. That means that we must accept the figures as accurate.

Which means that we must accept Doll’s contention that fifty times more heavy smokers caught lung cancer than did non-smokers. That is, for every one non-smoker who died from lung cancer, fifty heavy smokers died from lung cancer. In fact, he claimed that there was a progression – light smokers fared worse that non-smokers; moderate smokers fared worse that light smokers; heavy smokers fared worse that moderate smokers.

Doll therefore concluded that smoking itself caused lung cancer.

——————-

Despite the fact that we must accept the figures and data in Doll’s reports, we can be disappointed. In fact, epidemiologists ought to have been very disappointed by the reports. Think for a moment. What was the one absolutely certain fact which Doll had? Was not the ONLY absolutely certain fact at the beginning of the study the dates of birth of the doctors? I am really surprised, therefore, that Doll broke his groups of doctors into ten-year age groups. Considering that the ages were the only certain facts, a much narrower selection ought to have been chosen – say, four year groups.

Also, it is easy to forget that Doll must have known the geographical distribution of the doctors since he wrote to them individually at home. We know this because he himself said that there were occasions where relatives answered the questionnaire when the doctor himself was ill. Therefore, there was an excellent opportunity to see whether or not there were ‘hotspots’ of lung cancers here and there – in heavily industrialised areas as compared with country areas, for example. That opportunity was missed.

A query also arises about Doll’s division of smokers into only three groups – 1 – 14 cigs per day, 15 – 24, and 25 +. It must surely be obvious that a person who smokes 14 cigs per day must be 14 times at greater risk than a person who smokes 1 cig per day, and yet Doll grouped them together and called them ‘light smokers’. Superficially, this may seem unimportant, but it is only unimportant if the quantity smoked is unimportant! But Doll said that the quantity smoked is VERY important! This is one of several ‘contradictions’ within Doll’s reports.

Another point, which Doll dismissed, was that, out of the 60,000 doctors approached, 20,000 ignored the survey. And yet the survey asked only a few simple questions. Why did so many doctors ignore it? Doll said that he did not follow up on his first circular. Some people might say that this could be important since it brings into the study an element of ‘self-selection’.

————————

In 1962, the Chief Medical Officer of the UK produced a report called ‘Smoking and Health’, which was published in 1963. Recently, the Holy Zealots of the Tobacco Control Inquisition have been celebrating fifty years since the publication of that report.

We must think about time-scales here. Doll’s ten year report was only published in 1964, but the CMO was in cahoots with Doll and must have known the contents before he published the 1962 report. It is a curious fact that Hill produced his well-known dissertation on what is required for an epidemiological study to be reasonably assumed to indicate ’cause and effect’ in 1965, after Hill had ‘retired’ at the age of 64. Is it not fairly obvious that there was some sort of power struggle? Was there not a difference of opinion between Hill, Doll and the CMO about the conclusions being drawn as to ’cause and effect’ from the Doctors Study and other studies about that time? I think so. It is obvious to anyone who is prepared to study Doll’s reports that smoking is only POSSIBLY a cause of lung cancer and almost certainly one factor out of many.

———————

It is reasonable to assume that the prevalence of smoking would be most intense during the war years 1939-45. In that period, vast numbers of Briton’s were ‘living on their nerves’ and taking solace from the enjoyment of tobacco. Who can blame them? But even then, there were eugenicists hanging around and frowning. It was not long after the war that American children’s comics started demonising tobacco – ‘the evil Nic O’Tine’ was painted in much the same terms as Jews in Nazi Germany – skinny, sneaky, with bent noses and evil eyes. In fact, it is not unlikely that the anti-tobacco crusade began very shortly after the end of Prohibition.

——————

The question of ’cause and effect’ can only be certified after the passage of a sufficient period of time when the proposal of ’cause and effect’ is dependent upon a sufficient passage of time. Erm… that makes sense to me because I wrote it! Put it this way. Doll said that his Doctors Study showed that it takes decades for the damaging effects of smoking to be revealed. It makes sense, therefore, to assume that it takes decades of NOT smoking for the effects of NOT SMOKING to be revealed.

Those decades have now passed. Smoking prevalence started to fall, among men, shortly after the war, which is to be expected, since the ‘pressure’ was off. Also, despite the end of the war, much deprivation was still in evidence as a result of the war, and so the cost of smoking was a relevant factor in the budget of ordinary people. Starting before 1950, smoking prevalence began to diminish. The diminishing of smoking gathered pace so that, between 1960 and 2000, smoking prevalence among men fell from some 80% to 30%. Over that period of time, the cost of tobacco rose and rose, and, at the same time, more and more goods were coming onto the market and more and more advertising was taking place, especially on the TV. It is reasonable to assume that lots of people stopped smoking in order to buy goods. I certainly did. The only reason that I could afford my first car was because I stopped smoking for twelve months. It was nothing to do with health – it was everything to do with cost.

—————-

It is important for us to consider ‘the delayed effect’ of smoking in connection with lung cancer. According to Doll and others, it takes decades for smoking to affect people. But that is not what Doll’s Doctors Study shows. According to the study, most doctors started smoking about the age 19. But the study showed that the effects began about the age of 35. That is only 16 years – it is not ‘decades’. Even so, why did the effect, even after ‘decades’ affect only a few people?

—————-

The ’delayed effect’ has never been adequately quantified. How could it be, since it is an artificial construct which depends upon a circular argument? (“We know that heavy smokers get LC 50 times more often than non-smokers after 10, 20, 30 years, give or take a decade or two. We know this because we know that it takes a decade or two for the effect to ensue. We know this because smokers do not get LC for a decade, or two or three or not at all”).

——————-

There are so many confounding factors, known or unknown, that the only way to verify whether or not smoking ’causes’ lung cancer is to WAIT AND SEE what happens, as regards the incidence of LC deaths, after an appropriate length of time during which smoking prevalence diminishes. That has been happening for some fifty years now. So what has happened to lung cancer in men? Has it diminished pro rate with the diminishing of smoking prevalence, even after decades of elapsed time? I think that we shall find that the diminishing in LC in no way reflects the diminishing of smoking prevalence.  But I have seen some fancy statistics which purport to show large falls in LC deaths since 1950-ish, but I some of them are not really relevant IF there is a decades-long delay in the effects of smoking. For example, if smoking in men only started to decline significantly around 1970, the fall in LC deaths should only start to be revealed  around 1990 or so. Falls in LC deaths in the 50s, 60s, 70s and 80s ought not to effected by cessation during those decades, but they could show a downward trend for other reasons.

Another very important factor is the improvements in LC detection and treatment  since about 1960. I would suggest that people read this paper:

http://www.wtflungcancer.com/lung-cancer-surgery-demystified-guest-post-by-dr-david-tom-cooke/

It is called ‘Lung Cancer Surgery Demystified’. It is very simple and straightforward, and reveals some of the techniques now available, such as PET scans, radiation and directed chemotherapy. We might therefore expect to see a lowering in LC deaths, especially in otherwise healthy younger people, even if those deaths are merely postponed.

The death stats have been quite difficult to dig out for years prior 2006 because they were not put on-line automatically. What was put on-line was the original data sheets, and I had to extract the data which I wanted and then use excel to refine them further. But I got them in the end. But they are very simple figures, and I do not see any reason that they should be otherwise. I reason that, if smoking causes LC, and if people stop smoking more and more over a period of time, or, better still, never start smoking, then LC deaths should diminish and diminish. Whether or not they get LC later in life is not relevant – they should not get LC at all if LC is caused by smoking. So let us look at some figures.

———————–

There have been enormous changes in the demographic of and size of the population of England and Wales since 1950, as we know. To take account of these changes would be impossible for me. There has also been great change in pollution levels of all sorts and medical techniques have improved immeasurably. I reasoned that the simplest way to count LC deaths was as a proportion of all deaths and, possibly, as a proportion of all cancer deaths. I went back to 1980 (being around the time that smoking cessation and the fall in starting smoking ought to start to show some effect. Although I have female figures as well, I intend to concentrate only on men, since men were by far the biggest smoking group.

TABLE A.

MALE LUNG CANCER DEATHS.

Year………LC deaths…….% of total deaths……% of all cancer deaths.

1980……….25250……………8.65…………………………..36.33.

1990……….22900……………8.26………………………….30.31.

2000……..17,700……………6.93……………………………25.39.

2006……..16,700……………6.93……………………………23.00.

2011………16,700……………7.11…………………………….22.18.

[Figures derived from Office of National Statistics]

NB. I have only included the ‘% of all cancer deaths’ as vaguely relevant to the matter in hand. In reality, the decline could easily be explained by a rise in other cancer deaths, rather than a fall in LC deaths.

 

We could put alongside those figures the percentage fall in smoking prevalence (roughly) FROM 20 YEARS BEFORE:

TABLE B.

PERCENTAGE FALL IN SMOKING PREVALENCE.

Year………(% male smokers)………..% reduction on 1960 base.

1960……………..75%.

1970……………..70%…………………………5%.

1980…………….55%………………………..20%.

1990…………….40%………………………..35%.

(1998…………..35%…………………………40%)

You could be forgiven for saying, “Ah, but LC deaths have fallen quite substantially” Well, Yes, but the total number of deaths since 1980 has fallen. In 1980, the total of deaths was 581,400. In 2011, the total of deaths was 484,000. As you can see from the middle column of TABLE A, the percentage of male lung cancer deaths as a proportion of total deaths has fallen by about 1.5% over the thirty year period, even though smoking prevalence fell by about 35% among males DURING A THIRTY YEAR PERIOD IN AN ERA TWENTY YEARS EARLIER; that, is, the ‘delayed effect’ has been allowed for, but has made little difference.

But what about the vastly improved medical techniques for diagnosis and treatment? Are we to assume that the reduction in smoking prevalence, and only the reduction in smoking prevalence, has caused the reduction of about 1,5% over a period of thirty years? Are we to assume that the medical improvements have made no difference at all to LC mortality? “The absurdity of this reasoning is demonstrated by simply stating it”

But there are other factors also. The Clean Air Acts came into being around 1964. Are we to say that they also made no difference to LC deaths and COPD deaths?

———————————

CONCLUSION.

Have readers ever heard of ‘Occam’s Razor”? Briefly, it states that ‘the simplest solution is probably the most likely to be correct’. The above stats are very simple, but they are based upon the simple, and reasonable, expectation that thirty year of smoking diminution would have produced, by now, a substantial reduction in LC deaths, as a proportion of all deaths. If smoking causes LC, then a big reduction in smoking (after allowing for the delayed effect) must be followed eventually by a big reduction in LC as a cause of death. Note that Doll did not permit old age to be a cause of LC – 0nly smoking is the cause of LC. The above figures, in their little way, show that Doll was wrong. Certainly, smoking heavily was damaging, but only in the context of other relevant factors, such as genetics, air pollution, radiation, stress, and so on. The more you look at it, the sillier it seems to believe that smoking, in itself, causes lung cancer. The mere fact that so few smokers get LC should have flashed big red warning signs from the beginning.

Can I remind readers that Tobacco Control FAILED AND REFUSED to bring evidence before the Court in the McTear versus Imperial Tobacco Company Case, which concluded only in 2005? The Judge in that Case, Lord Nimmo Smith, roundly castigated Tobacco Control for that failure. Can I remind readers that Doll gave evidence in that Case as an expert witness?

Finally, I have the stats about total deaths, deaths by age range for 1980, 1990, etc., LC deaths in age ranges, but I see no point in wasting time on them. The only way to stop Tobacco Control in its tracks is to de-fund it completely. When that happens, it will disappear. But it will take some time, even after de-funding, for the iniquitous effects to fade away. Certainly, the legislation which dim-witted politicians have enacted will stand for a long time, and continue to be damaging. There is just a remote possibility that an amendment to the smoking ban could allow smoking rooms in pubs, clubs, etc, at the discretion of the owners, but don’t hold you breath.

The EU is particularly painful. In theory, I like the EU, and I think that it is a good thing to come together in those matters in which there is a common interest in standardisation. For example, you should be able to buy a washing machine made in Poland and expect it to be reliable, but it may well be designed in a different way from the norm. It may well be a good thing for us all to go metric. I have no problem with that. I also have no problem with the idea of the Euro being a common currency as a reserve currency. By that, I mean a currency which is automatically acceptable everywhere in the EU, but is valued in different Sovereign States in terms of their local currency. Thus, in the UK, Euros should be accepted in shops, pubs, etc – with an accepted exchange rate. Stirling is very readily accepted in Spain on those terms.

But where the EU has gone wrong is in matters of Health and such. There is no reason that Health should be standardised. Sovereign States can decide for themselves, according to their circumstances, what is important in matters of Heath. Some may have problems with obesity and others may have problems with thinness.

Above all, I am amazed at Cameron, the Prime Minister. At first, he seemed to be content to let his Ministers make decisions and stand or fall be them. Now, he seems to have fallen into the propaganda trap. Somehow or other, he has been persuaded that he had to have an opinion upon everything. Does anyone else think that this is a sign of weakness?

 

 

About these ads