The path of least resistance
It is one of the old saws that lightning follows the path of least resistance, which is wrong, as it follows the path of highest electric field strength. In the youth of your bending author you could still see, against the London skyline, ball lightning conductors. Since rod conductors had been invented by one of those colonial rebels, Benjamin Franklin, they were not approved in Georgian society. The skyline has gone, but no doubt some of the Georgian balls remain.
In human affairs, following the path of least resistance is a more reliable rule. Policemen harass motorists, as it is an easier way of getting brownie points than catching burglars or muggers. Youngsters are encouraged to avoid difficult but rewarding things, such as physics, marriage or mathematics. Researchers in universities go with the flow and produce the results that the purse-holding bureaucrats require of them, rather than take the difficult path of scientific truth for its own sake. Journalists reproduce ready-made stories from university public relations offices, as it is much less tiresome than doing some real investigation. People accept naked PC propaganda from the media, about such things as global warming, genetic modification and passive smoking, as it is much easier than actually having to think about it. Meanwhile, true scholarship has taken refuge in The Web, among the pornographers and trivialists.
By Old Ned
The Wild Pink
The summer is a wonderful time of year for nature lovers. I was woken at the crack of dawn by the song thrush in the chestnut tree behind the house, and what a noise ’e made later, when the old sparrow hawk carried him off, all a screeching an a yowling. ’Ow I did larf! But, mind you, your Mother Nature’s a wonderful thing. Have you ever thought what an amazing stroke of luck it is that a cat has two holes in its fur just where its eyes are? The old dog fox has been a visiting regular this year. ’E soon cottoned on to the hunting ban. ’E got into old George’s hen house the other day. It don’t take ’im long to bite the ’eads off a couple a dozen chickens. You should a seen old George’s face. ’Ow we did larf! Old Reynard does like ’is bit of fun. I bin a spraying of my horganic crops with that pyrethrum. The ol’ doc says the ee pee hay declared it a probable human cars in again. Being ahejimicated ’e uses a lot of long words. Still if it don’t work I’ll ’ave to use some of the private stuff from the back of the shed. We calls it the midnight service – as the old vicar says “let us spray.” Since we went horganic all the money worries are behind us. It all goes for three times the price down at the market, but them posh housewives don’t like holes in their produce, so you ’as to do the necessary if necessary. From my window I can see the delicate bloom of a Wild Pink peeping out shyly from a crack in the footpath, but a bit o’ Roundup will soon see ’im off. Summer is also the time for occasional little visitors. Last week we spotted our first protestors. It’s the warm weather what brings ’em out. They said they was Green Peas and they had come to trample our potatoes, on the grounds that our area was a suspected GM trial site. I don’t know what they are on about, as most people round here drive Japanese cars. Anyway old George managed to convince them that the potatoes were really cabbages, and they marched off singing “we shall not be overcome.” Well, I ’opes you have enjoyed this little bit of comooning with nature, specially if you are one of them townees what don’t come across it much. I ’opes to be with you again around muck-spreadin’ time.
Why, I can smile, and murder while I smile,
And cry, 'Content,' to that which grieves my heart,
And wet my cheeks with artificial tears,
And frame my face to all occasions.
I'll drown more sailors than the mermaid shall;
I'll slay more gazers than the basilisk;
I'll play the orator as well as Nestor,
Deceive more slily than Ulysses could,
And, like a Sinon, take another Troy.
I can add colours to the chameleon,
Change shapes with Proteus for advantages,
And set the murd'rous Machiavel to school.
Can I do this, and cannot get a crown?
Tut! were it further off, I'll pluck it down.
All the Winston Smiths in the Orwellian Ministry of Truth must be working overtime as the history of the great foot and mouth epidemic is being rewritten. Guess what! It turns out that the Great Leader was in favour of vaccination after all, but he was too shy and retiring to say so. How lucky we are to have a leader so full of modest stillness and humility. The staging of his reluctant admission conjures up the image of the Bard’s Richard III standing between two clergymen, denying his ambition. It was all those nasty farmers and food industry moguls that forced his hand. This is leadership?
The Times of Saturday June 29th devotes a whole page to the revisionist process. Why ministers fought shy of mass vaccination goes the headline. As we observed 16 months ago, the whole disaster was down to the ministry (MAFF, now DEFRA; the names have been changed to protect the guilty), which had been out of control for years. They were, as the cliché has it, a disaster waiting to happen.
Such was the disgust and horror that this saga aroused in your bending author that Number Watch has made thirteen comments since the very first that was entitled Mad, mad, mad in February last year. In those far off, innocent times the total of animals slaughtered was 15,000, which was the number of the month. Who would have thought that it would end up 400 times bigger?
The scapegoats have been identified and punished. Nick Brown was the political scapegoat. Incompetent he might have been but, as noted in our first report, he bravely stood in front of the serried ranks of empty New Labour benches and did his best to defend the indefensible. The lay scapegoat was a rather inadequate pig farmer Bobby Waugh, who is under house arrest for cruelty to animals. The judge made a point of saying that he was not before the court to act as a scapegoat for the outbreak of the disease. Oh yeah?
The really sick part of all this is that it was all swept under the carpet for the purposes of the general election. If Britain had not had a uniquely incompetent opposition, they would never have got away with it. While the Great Leader was touring the country glad-handing the populace and beaming his famous smile, the project of which he had taken personal charge was killing thousands of innocent and healthy animals every day. It continued after the election, but the capricious media had moved on.
Farmers who lost animals were compensated, but the world is not told about all the others. A neighbour of your bending author, for example, has just lost family home . He was a pig farmer, whose pigs were perfectly healthy, but he was not allowed to move them and in the end they were valueless, though he received no compensation. Apart from the Thatcher experiments in monetarism, there have never been so many lives and livelihoods ruined by Government action.
On The Times page there is a reasoned commentary by Magnus Linklater, which must carry the most understated headline of the year – Prevention was always preferable to mass killing.
Missing from the whole page was any mention of the epidemiologist whose computers orchestrated the crescendo of killing, Professor Roy Anderson; so in his honour here is a verse from T S Eliot’s Old Possum’s Book of Practical Cats:
Macavity, Macavity, there's no one like
For he's a fiend in feline shape, a monster of depravity.
You may meet him in a by-street, you may see him in the square
But when a crime's discovered, then Macavity’s not there!
On the seventh floor of the World Health Organisation (WHO) headquarters in Geneva last Thursday, an air of excitement and apprehension was building. Excitement, because for three days 25 top scientists from across the globe had been closeted in a small conference room, their heads bent in urgent discussion over a scientific discovery of global significance. Apprehension, because the nature of that discovery was one that would raise serious concerns over the most basic aspect of our daily lives -our eating habits.
The reason why these men and women had flown to Geneva was a chemical called acrylamide.
Yes folks, it’s the old acrylamide scam again. <material removed> the horror of thousands of parts per billion of the stuff in all the nation’s favourite snacks. They go on breathlessly:
There was no escaping the fact that a huge range of everyday foods contained potentially dangerous levels of a cancer-causing toxin.
What is strangely absent is any proof that the stuff is carcinogenic at all. Number Watch has not seen such evidence, but would not mind betting that it is those damned special tumour-prone rodents who have been fed the stuff at ludicrous concentrations. In other words the Concentration Fallacy.
No wonder the epidemiologists were excited. Instead of being faced with begging on the streets they will now be able to put organic bread on the table for their families for years to come on the strength of this one.
Much of the debate centres on the European limits for this terrible substance, which is 10 parts per billion. Now, if you consider the most toxic substance known to man, botulin (you know, the stuff that toxiphobiac Californian women inject into their faces) an order of magnitude calculation suggests that at this concentration one crisp (chip in the US) would kill a child. For a run of the mill poison, such as sodium cyanide, however, you would need about 50,000 times as much; say well over a thousand standard pub packets, all in one go. Now, while admittedly carcinogenicity is not the same as toxicity, a little evidence that acrylamide is more dangerous than the very stuff of detective fiction would not go amiss.
Help! We are drowning in Red Tape
Where is the wisdom we have lost in knowledge?
Where is the knowledge we have lost in information?
T S Eliot, Choruses from The Rock
(click for full size)
108 is the age of the great great grandmother who starved her self to death because she had been moved from her nursing home, which had been closed down in the face of onerous new regulations.
800 is the number of nursing homes being shut every year because of the cost of complying with the regulations.
94 is the number of pages in the book outlining the new regulations.
700,000 is the number of National Health Service beds blocked by elderly people who are well enough to leave but cannot find a place in a home.
25 Billion is the number of pounds that the Chancer of the Exchequer has confiscated from pension funds
It was the Conservatives who began the attack on the elderly by removing tax relief on the cost of their care and cutting benefits for people in care homes, but that was nothing compared with the effects of New Labour's love affair with bureaucracy and regulation.
At the other end of life, the Great Leader had refused to intervene in the disastrous new examination fiasco. This generation of children is the most examined ever. Tales of stressed pupils are rife in all communities and no doubt there will be the usual stories of teenage suicides. Many independent schools and the select state school attended by the Great Leader's son have opted out of the system. However, nothing is allowed to block New Labour's obsession with targets, tests, league tables and the general overweening bureaucracy. The penultimate year of school, the lower sixth form, was for your bending author, as for many others, the most rewarding and educational experience of life, precisely because it was unexamined. New Labour, however, cannot understand people educating themselves for the sheer enjoyment of it. They must be coerced, formulated, impaled upon the pin of bureaucratic assessment. That is the socialist way.
Spot the seasonal variationA letter from our indefatigable correspondent in Puerto Rico
I'm messing around with some numbers (bad) that I came around during some investigation of mine, shown in the graphs below. I also include a bar version. Seems that there was a more or less log increase with time which peaked on July of the second year, but mercifully there now seems to be a decreasing trend up to the present (this past June). This means that the danger is decreasing, thank the Creator! Maybe you might be able to find a good fit for this curve, as I'm no good at that. Hope you can glean some insight out of it. After all, the badnumberists get plenty out of much less!
By the way, this is a graph of your numbers of the month.
Food for thought
Naturally, smoking adversely affects human fertility, as it does everything else in the world of epidemiology, but now, thanks to valuable research at King’s College London, added to it are such things as beer, soya and vegetables.
Researchers at the Metropolitan University of Nether Wallop have been aware of this for some time and its Contraceptive Advice Department has been recommending bathing the genitalia in soy sauce or draught real ale as a normal part of foreplay, though there is a strong faction who favour bottled Guinness, and the academic dispute rages into the night.
Things can get out of hand, however. Celia Spume of Neasden relates that she named her youngest Cider "on account of I reached for the wrong bottle in the moment of passion." Mrs Glenda Pustule of West Grinstead reports that she and her husband started out on ordinary mild soy sauce, but soon graduated to the strong stuff. She adds “Then, before we knew where we were, it was the full king prawn chop suey with a side dish of water chestnuts and bamboo shoots. I don’t know where it will all end.”
Dr Gertrude Narwhale, Reader in Contraceptive Gastronomy at the University, says “There is no need for feelings of guilt about such practices. Variety is the spice of life and a little seasoning is quite harmless. In fact, I always recommend taking a large pinch of salt with the products of such research.”
Publish and be biased
It seemed too good to be true when the New York Times decided to publish an article on the importance of negative results , but of course, they missed the main point. They seem to think that the main reason for publishing negative results is to save others from repeating the experiments. In fact, it is the almost universal non-publishing of negative results that sustains most of the epidemiological fictions that arise from apparent positive results.
Let us demonstrate with some simple modelling. Assume that there is a politically correct surmise that passive drinking causes toe-nail cancer. Assume also that there is actually no such correlation and that the probability of any member of the populace contracting the disease in a year is 0.0001. Eager to get on the bandwagon, no fewer than one hundred professors world wide each detail a research student to do a study. Each research student is especially diligent and finds 100,000 people who have been exposed to passive drinking. After a year they look to see how many have contracted the disease. It is obvious from the figures that we would expect about ten. The 100 results, however, are scattered about this figure and we can model this process by, for example, the MathCad expression:
The results of the 100 studies are then as in the graph below.
This is where the publication bias comes in. Only 46 of the 100 results are positive in that they produce more than the expected number. The rest are discarded and the professors go on to some other topic. So the distribution of published results is then as in the second graph. The average of the whole 100 studies is 10.11, about what we expected, but the average of the published studies is now 12.975.
At this stage along comes a nameless US Government agency to do a meta-study, combining all the results. The next you hear is that all the media are carrying headlines bellowing that Passive drinking causes a thirty percent increase in toenail cancer. This all shows how easy it is to get a Relative Risk of 1.3 on the basis of no effect at all.
Did you notice, by the way, that one of the studies produced a RR of greater than 2? Which will perhaps help to answer one of the the most frequent questions posed to Number Watch - why are real scientists not prepared to accept risk ratios of less than 2?
In case you think that there is any exaggeration in the above, the effect is actually understated. The random numbers would normally be divided by another set of random numbers, the estimates for the non-exposed population, which further increases the scatter. Furthermore, epidemiologists vary in their squeamishness about publishing low RRs, which moves the average even higher and nicely rounds the distribution, so that it looks more genuine. Also we did a relatively large number of studies in order to be able to produce histograms, but the argument applies however many or few studies are involved.
It does not matter much which sort of study is done. All that matters is the expected number (10 in this case, quite a common sort of number). You might, for example, have done a retrospective study. If you have a thousand toenail cancer patients and the probability of any individual being exposed to passive drinking is one percent, you again have an expected number of 10.
According to friend Poisson, the resulting scatter (standard deviation) as a fraction of the mean for an expected value of N is the reciprocal of the square root of N, so for N = 10 it is about 30%, which is what we got.
Funnily enough, about 1.3 is very common RR in published epidemiological studies. Some of the really tacky ones, such as the EPA study on ETS, even go down to less than 1.2.
Its a funny old world.
Dangerous and destructive nonsense!
Just when you thought that the numbers in the modelling exercise for Publish and be biased (above) might be atypical or unfair, this happened .
(click for full size)
Yes, a bunch of quacks in America have done a "statistical survey" of the effects of Hormone Replacement Therapy. The media uniformly and dutifully reported the whole ragbag. The Times sub-headings read:
US study halted after health fears rise
Patients suffer 41% increase in stroke risk
22% increase in risk of heart disease
Regular readers will note Relative Risks of less than 2, which are not generally accepted by real science. Having a computer model to hand, or our one hundred imaginary professors if you like, it seemed too good an opportunity to miss. There was a slight modification to include the fact that the number of controls was the same as the number of victims.
Unlike climate models there is nothing secret about it and any one with MathCad or similar can reproduce it with ease.
Taking one set of figures (the most drastic) from The Times, for strokes, ignoring the usual odd fraction of women as the numbers never quite add up, we can deduce that there were 8,304 women who took the therapy and 8,304 in the control group. The annual probability of getting a stroke was 0.00195, so the annual number of strokes in the control group was about 16 and the number in the victim group was about 23. This is a RR of 1.44, but what is 0.03 between friends?
On the assumption that there is no real effect, so the control and victim groups are statistically identical, two sets of 100 random binomial numbers were generated (using for the cognoscenti rbinom(100,8305,0.00195)) and one set was divided by the other. The results are then Relative Risks for the case where there is no causal relationship.
The results in the form of a histogram, were as follows:
The remarkable thing is that our 100 imaginary professors, working on no real effect at all, have actually done as well on average as the real quacks. For, when we include the publication bias effect, as in the second histogram, the imaginary professors got an RR of 1.43.
Furthermore, this was not an isolated case. Running the model ten times produced RRs of 1.44, 1.51, 1.49, 1.53, 1.38, 1.65, 1.41, 1.63, 1.40 and 1.53.
The question arises as to why these results are more dramatic than in the previous modelling exercise above. The reason is that this time we have included the effect of the scatter of the control group, which is the denominator in the RR. Thus, while both groups produce the expected relative scatter (standard deviation/mean) of 0.25 (i.e. the reciprocal of the square root of the expected number, 16), when we divide one by the other the relative scatter is increased by more than half.
So all this panic is caused by experimental results that are just about exactly what you would expect if there were no causal effect at all .
Overcome by the "significance" of their findings, the "researchers" called a premature halt to the study and there are now proposals that this (presumably) valuable therapy will be withdrawn. The shares of the company producing the drug fell to a two year low.
Another aspect of this scare is a new variation of publication bias. The Times chose to ignore two apparently beneficial (though equally fatuous) effects of this data dredge, as good news is no news.
Footnote for mathematicians: obviously an analytical solution to this problem would be preferable. I have tried and failed. It was beyond the mathematics of a simple engineer. I am aware that in the case of normal variates the quotient follows the rather beastly Cauchy density function, but I cannot work out what it should be for these asymmetric distributions. Any advice would be appreciated.
Regular correspondent Matt DiMeo writes
I'm not sure I follow your analysis in your "Dangerous and destructive nonsense" article.
Based only on the information in the Telegraph article to which you linked, I don't see how publication bias applies. This seems to be the results of one monolithic study, rather than a "meta-study" aggregating results from multiple studies.
Am I missing something obvious in either your analysis or the Telegraph article?
Your previous "Publish and be biased" article did an excellent job of explaining the publication bias effect, but I don't see how it should apply to this particular case.
The numbers were obtained from the coverage in The Times. This link was not given, as non UK readers have difficulty of access once the item is archived. Chopping off the bottom half of the distribution might be thought of as a device to isolate the supposed malign results. If you leave the distribution intact you always get an average RR of about 1. Equally you can chop off the top of the distribution to estimate the likely benign RRs and show that they are equally insignificant. Having said that, most of the UK coverage ignored the supposed benefits in favour of the scare, therefore exercising publication bias.
Before someone else raises it, another potential criticism is that the calculation only dealt with the annual probability, as the risks were quoted on a per annum (and per 10,000) basis, whereas the original study went on for five years. Because of the diminishing returns in the way Poisson works this only gives you and advantage of root five, or a little over two. Running the model on this basis still produces results of 1.4 or greater, though admittedly below the one in twenty level of chances of being wrong by accident adhered to by epidemiologists. Sceptics do not generally accept this 5% criterion; particularly in a case like this, where there is a data dredge. We know from the media coverage that at least five possible effects were included and there might have been more.
Here are the actual annual probabilities calculated by reader John S Baltutis from the write up in the Nando Times.
Yet another try. It appears that the answer to Matt lacked perspicuity. If you want to determine the average deviation from the mean (as opposed to the standard deviation, which is the RMS deviation from the mean) you have to start from the mean and integrate outwards, otherwise your result will always be around zero. The chopping off talked about is just a convenient way of doing this separately for the upper and lower halves of the distribution (which is asymmetric). The model for publishing bias happened to do this conveniently, but using it seems to have fogged the issue by implying that it was inherent to the study under discussion, which it is not. The average RR if you use the whole distribution will always be approximately 1.0, which conveys no information at all. The reason for doing 100 imaginary studies is to obtain a distribution shape in the absence of the ability to do it analytically. It does not imply that there is more than one study under discussion.
All that being said, large parts of the media reported only the bad news, which reintroduces a form of publication bias. The basic point is that the incremental risks here are very small (far too small to worry the life out of millions of women) and the RRs are not of an acceptable magnitude. As for the figure of 0.05, here is how it was discussed in Sorry, wrong number!, note particularly item 3 (and while you are at it, the cartoon on the front of the book):
Here is a disturbing quotation from Matthews’ pamphlet:So just what was the profound logic that led Fisher to choose that talismanic figure of 0.05, on which so much scientific research has stood or fallen? Incredibly, as Fisher himself admitted, there weren’t any. He simply decided on 0.05 because it was mathematically convenient.
It is now quite clear that the choice was a very bad one. It means that many supposedly significant scientific results are entirely spurious. This number has become an icon throughout science. It is the determinant of whether research is submitted or accepted for publication. Throughout the world scientists sieve their numbers, looking for the magic breakthrough. If P is greater than 0.05 they abandon the line of research and go on to something else. If it is less, they throw their hats in the air and start writing. In this I have been as guilty as the next man. There are important consequences in all this:
1. One in twenty of all scientific tests on non-existent effects and using accepted statistical significance is wrong.
2. There is a massive tendency to publication bias. Not only was the one wrong positive result published, but the nineteen negative results were not published.
3. As soon as the researchers get their 0.05 they stop taking measurements to avoid reducing the publishability of their results, for nowadays researchers are measured by their publications.
4. Researchers plug their results into a statistical package, turn the electronic handle and see whether they have won the prize on the statistical fruit machine. They do this without attempting to understand the theory behind the statistics they are using.There is a powerful motivation towards cheating, even sub-consciously. This is exacerbated by the publish-or-perish attitude arising from modern auditing.
Have you ever wished you had not started something?
Last word on Dangerous and destructive nonsense!
Thanks to reader Dennis A. Falgout, the original paper has been tracked down in all its tacky glory. It was even worse than your cynical author imagined. Here is the results section. The figures to take note of are those emphasised in red. Remember a figure close to unity means that there is a one in twenty chance that there is no increased (or decreased) risk at all, even if you take it all at face value. The chutzpah of these clowns (and there are a lot of them) is simply unbelievable. Millions of women around the world have been put through the mincer just so that they could have a tatty publication to their names.
Results On May 31, 2002, after a mean of 5.2 years of follow-up, the data and safety monitoring board recommended stopping the trial of estrogen plus progestin vs placebo because the test statistic for invasive breast cancer exceeded the stopping boundary for this adverse effect and the global index statistic supported risks exceeding benefits. This report includes data on the major clinical outcomes through April 30, 2002. Estimated hazard ratios (HRs) (nominal 95% confidence intervals [CIs]) were as follows: CHD, 1.29 (1.02-1.63) with 286 cases; breast cancer, 1.26 (1.00-1.59) with 290 cases; stroke, 1.41 (1.07-1.85) with 212 cases; PE, 2.13 (1.39-3.25) with 101 cases; colorectal cancer, 0.63 (0.43-0.92) with 112 cases; endometrial cancer, 0.83 (0.47-1.47) with 47 cases; hip fracture, 0.66 (0.45-0.98) with 106 cases; and death due to other causes, 0.92 (0.74-1.14) with 331 cases. Corresponding HRs (nominal 95% CIs) for composite outcomes were 1.22 (1.09-1.36) for total cardiovascular disease (arterial and venous disease), 1.03 (0.90-1.17) for total cancer, 0.76 (0.69-0.85) for combined fractures, 0.98 (0.82-1.18) for total mortality, and 1.15 (1.03-1.28) for the global index. Absolute excess risks per 10 000 person-years attributable to estrogen plus progestin were 7 more CHD events, 8 more strokes, 8 more PEs, and 8 more invasive breast cancers, while absolute risk reductions per 10 000 person-years were 6 fewer colorectal cancers and 5 fewer hip fractures. The absolute excess risk of events included in the global index was 19 per 10 000 person-years.
Yet another footnote: Number Watch was guyed (and why not?) on FreeRepublic.com for describing the Cauchy distribution as "beastly". Now you might well prefer an over-used anal Americanism to express disagreeableness; that is a matter of taste. Giving the formula for the density function did not add much to the total of human knowledge. However, to quote the monumental Methods of Mathematical Statistics by Cramér, on the subject "No moment of positive order, not even the mean, is finite." That , however you prefer to express it, is somewhat inconvenient.
Meanwhile, back at the great antioxidant scam
How wonderful! How fragrant! Let Joy be unconfined!
Thanks to a tip from reader David Delany, your bending author has had the opportunity to read the account of a properly conducted randomised trial with a commonsense interpretation or the results. Furthermore, it is a pleasure to be contradicted in the assertion above that negative results are never published. As the great R A Fisher observed “without randomisation there is no significance”. Unfortunately, as with so many prophets, Fisher was doomed to have his writings selectively interpreted by his followers and had the dubious responsibility of foisting the P<0.05 criterion upon the world.
When you spend much of your time with your snout rooting in the muckyard of junk science, it is easy to forget that you only have to raise your head to rediscover the purity of real science.
This paper, produced by the Heart Protection Study Group, examined the proposition that antioxidant vitamin supplements would reduce the incidence of vascular disease, cancer and other adverse outcomes. The trial involved 20,536 UK adults with coronary disease. It was properly designed and the procedure is carefully detailed in the paper. Here are the results:
There were no significant differences in all-cause mortality (1446 [14·1%] vitamin-allocated vs 1389 [13·5%] placebo-allocated), or in deaths due to vascular (878 [8·6%] vs 840 [8·2%]) or non-vascular (568 [5·5%] vs 549 [5·3%]) causes. Nor were there any significant differences in the numbers of participants having non-fatal myocardial infarction or coronary death (1063 [10·4%] vs 1047 [10·2%]), non-fatal or fatal stroke (511 [5·0%] vs 518 [5·0%]), or coronary or non-coronary revascularisation (1058 [10·3%] vs 1086 [10·6%]). For the first occurrence of any of these "major vascular events", there were no material differences either overall (2306 [22·5%] vs 2312 [22·5%]; event rate ratio 1·00 [95% CI 0·94-1·06]) or in any of the various subcategories considered. There were no significant effects on cancer incidence or on hospitalisation for any other non-vascular cause.
Among the high-risk individuals that were studied, these antioxidant vitamins appeared to be safe. But, although this regimen increased blood vitamin concentrations substantially, it did not produce any significant reductions in the 5-year mortality from, or incidence of, any type of vascular disease, cancer, or other major outcome
For years now we have all been subjected to a deluge of advice from health gurus and substance-pushers that we should take this or that (tea , for instance) and be rewarded by greatly reduced health risks. Large sums of money were made by university research groups for epidemiological tests on these. Here is what Sorry, wrong number! had to say:
Another potent example is the great free radical scare. These entities are charged combinations of atoms that normally exist only for a short time. They are essential to a number of chemical processes, including those in living cells. We are told that free radicals damage DNA presumably as a result of high dosage experiments. The result of extrapolating these results down to very small concentrations has led to a whole chain of missing links that end up with the antioxidant “breakthrough”. The reasoning is that, if free radicals are bad, then anything that neutralises them must be good. The most remarkable turn round was with red wine. After years of attack by the alcohol SIFs, it was suddenly the flavour of the month, because it contains flavonoids, which are anti-oxidants. University groups were actually paid large sums of money by grocery firms to establish which varieties of wine were best for your health. By large sums I mean, for example, the £5 million paid by Safeways to Glasgow University to test 65 different wines. As a result Chilean Cabernet Sauvignons were held up as the best wines to stave off heart disease and cancer. Eventually someone did some more science and found that all alcohol was beneficial (in moderate doses). Another group claimed that tea was more effective. They were generously financed by a well known tea merchant.
The bad news is that the establishment and the media could not possibly let the great unwashed know about this, so we will have to give it the stamp.
If you want an example of good research with a positive outcome , here is one from the same people. Note particularly the p values, which are two orders of magnitude lower than those in the HRT study. There is a difficulty of explanation here. The relative risk of not taking statin is about 1.21, which is a level we have often derided. You have to take two factors into account. First, this is not an epidemiological observational study; it is a properly designed randomised trial. Second, the significance levels are one hundredth of those normally quoted in epidemiology (0.05) and the smaller the better. This is all neatly summarised by considering the number of adverse events. It was of the order of 8 per 10,000 people in the HRT study, while in the statin study below it is of the order of 800 per 10,000 people. In other words, if we take everything at face value, for the vulnerable, taking statin is 100 times more likely to save your life than giving up HRT.
It is all getting rather complicated; perhaps we should move onto something else.
MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20 536 high-risk individuals: a randomised placebo-controlled trial
Background Throughout the usual LDL cholesterol range in Western populations, lower blood concentrations are associated with lower cardiovascular disease risk. In such populations, therefore, reducing LDL cholesterol may reduce the development of vascular disease, largely irrespective of initial cholesterol concentrations.
Methods 20 536 UK adults (aged 40-80 years) with coronary disease, other occlusive arterial disease, or diabetes were randomly allocated to receive 40 mg simvastatin daily (average compliance: 85%) or matching placebo (average non-study statin use: 17%). Analyses are of the first occurrence of particular events, and compare all simvastatin-allocated versus all placebo-allocated participants. These "intention-to-treat" comparisons assess the effects of about two-thirds (85% minus 17%) taking a statin during the scheduled 5-year treatment period, which yielded an average difference in LDL cholesterol of 1·0 mmol/L (about two-thirds of the effect of actual use of 40 mg simvastatin daily). Primary outcomes were mortality (for overall analyses) and fatal or non-fatal vascular events (for subcategory analyses), with subsidiary assessments of cancer and of other major morbidity.
Findings All-cause mortality was significantly reduced (1328 [12·9%] deaths among 10 269 allocated simvastatin versus 1507 [14·7%] among 10 267 allocated placebo; p=0·0003), due to a highly significant 18% (SE 5) proportional reduction in the coronary death rate (587 [5·7%] vs 707 [6·9%]; p=0·0005), a marginally significant reduction in other vascular deaths (194 [1·9%] vs 230 [2·2%]; p=0·07), and a non-significant reduction in non-vascular deaths (547 [5·3%] vs 570 [5·6%]; p=0·4). There were highly significant reductions of about one-quarter in the first event rate for non-fatal myocardial infarction or coronary death (898 [8·7%] vs 1212 [11·8%]; p<0·0001), for non-fatal or fatal stroke (444 [4·3%] vs 585 [5·7%]; p<0·0001), and for coronary or non-coronary revascularisation (939 [9·1%] vs 1205 [11·7%]; p<0·0001). For the first occurrence of any of these major vascular events, there was a definite 24% (SE 3; 95% CI 19-28) reduction in the event rate (2033 [19·8%] vs 2585 [25·2%] affected individuals; p<0·0001). During the first year the reduction in major vascular events was not significant, but subsequently it was highly significant during each separate year. The proportional reduction in the event rate was similar (and significant) in each subcategory of participant studied, including: those without diagnosed coronary disease who had cerebrovascular disease, or had peripheral artery disease, or had diabetes; men and, separately, women; those aged either under or over 70 years at entry; and--most notably--even those who presented with LDL cholesterol below 3·0 mmol/L (116 mg/dL), or total cholesterol below 5·0 mmol/L (193 mg/dL). The benefits of simvastatin were additional to those of other cardioprotective treatments. The annual excess risk of myopathy with this regimen was about 0·01%. There were no significant adverse effects on cancer incidence or on hospitalisation for any other non-vascular cause.
Interpretation Adding simvastatin to existing treatments safely produces substantial additional benefits for a wide range of high-risk patients, irrespective of their initial cholesterol concentrations. Allocation to 40 mg simvastatin daily reduced the rates of myocardial infarction, of stroke, and of revascularisation by about one-quarter. After making allowance for non-compliance, actual use of this regimen would probably reduce these rates by about one-third. Hence, among the many types of high-risk individual studied, 5 years of simvastatin would prevent about 70-100 people per 1000 from suffering at least one of these major vascular events (and longer treatment should produce further benefit). The size of the 5-year benefit depends chiefly on such individuals' overall risk of major vascular events, rather than on their blood lipid concentrations alone.
Lancet 2002; 360: 7-22
100% (as a proportion) is a difficult number to achieve. However, Number Watch, seems to have done this. According to a learned critic in Australia who has reviewed Number Watch:
What a ****ing woeful site. Everything on that is utter rubbish.
On July 19th the Daily Mail had one of those editions that seem to sum up the state of the nation.
Summer of Air Chaos yelled the front page. There has been a quiet take over the major trade unions by left wing militants. They have turned out to be no different from other fat cats in granting themselves huge salary increases, but now they are flexing their muscles. Blair’s union crisis says the inside page headline over a story including biographies of the hard left campaigners who have taken over at five of the biggest trade unions.
Taxes to go up? Brown still refuses to answer. The huge increases in state spending announced by the Chancer of the Exchequer have occurred at a time when the world economy is in an uncertain state and Britain is still reeling under the cost of new stealth taxes. The Institute for Fiscal Studies has identified a seven billion black hole in Brown’s calculations, but he simply ignores questions on the matter.
Pension fear for a whole generation Over 6,700 company schemes have been scrapped in the last year. Not only has Brown raided the pension funds directly for five billion a year, but most of his other stealth taxes fall on the industries in which pension funds invest. The news came just one day after the announcement that MPs had awarded themselves a 25% pension increase half funded by the taxpayer (they tried to get away with the whole lot, but this time the Treasury was on the taxpayer’s side), which brings us to:
Why I am so embarrassed to be an MP. George Osborne, Conservative MP for Tatton tells what is really going on at Westminster. Not only have MPs awarded themselves a huge pension increase on top of a huge salary rise, but their everyday conversation centres on such important matters as what summer travel junkets they have managed to wangle (they are now, of course off on their annual three-month holiday).
The secret diary of Greg ‘cut the crap’ Dyke aged 55 ¼ .Yes the unlovely oik who controls New Labour BBC has also joined the fat cat revolution. He gets a £97,000 bonus on top of his £357,000 salary. Having famously cut the perks (and jobs) of those lower down the pecking order he has spent £30,000 of the ordinary punter’s money on lavish junkets for the hierarchy.
Let the Bulldozer’s loose Countryside in fear as Prescott demands 800,000 new homes in 15years. Some non-UK readers have asked whatever happened to Old Two-Jags. Well he is alive and kicking. He is scrapping the planning laws that protect green belts and rural areas. Developers who try to adhere to high standards (like the Duchy of Cornwall) are being forced to pack the houses in at a minimum of 30 to the acre. Your bending author has recently moved from a Hampshire village with beautiful views to and from the downs threatened with a huge housing estate to somewhere quite different, a Wiltshire village with beautiful views to and from the downs threatened with a huge housing estate. Oddly enough, you don’t get many MPs, let alone cabinet ministers, queuing up for these rabbit hutches. For those who don’t remember Old Two Jags here is a picture of him on the campaign trail at the last election.
Those of us who habitually place politicians somewhere below platyhelminthes in the natural order of things will be unsurprised by the announcement in the New Labour Guardian that MPs consider the cost of electricity unsustainably low. Wallowing as they do in their vast new salaries and secured futures with generous pensions, it simply does not occur to them that there are million of people for whom the fuel bill is just another serious financial headache (and there will be many more when those who have been robbed of their pensions come to retire). They are concerned about "sustainability", but like Humpty Dumpty they make the word mean what they choose it to mean. For the "sustainable" energy sources are those that are intermittent and unreliable, such as wind and waves. What they wish to sustain is, of course, the great global warming myth. They complain that
Thanks to the "unsustainably low level" of prices for fossil fuels like oil, coal and gas, the increase in renewables over the past decade rose by only 1% to 2.8% of the total British market - and dropped to 2.6% in 2001, a worse record than any other EU state except Luxembourg and Belgium.
So, in order to re-establish the march of the grotesque, noisy, uneconomic bird-slicers across the country, they want the groaning masses to pay even more for the necessities of life and also to pay for the stand-by systems for when the wind does not blow.
As George Osborne (above) said:
For the first time in recent history, the current generation of Britons will be worse off than the one that went before. But not MPs.
The passing of Lord Weinstock certainly deserves the cliché “end of an era”. He single-handedly changed the face of British engineering. He came to the fore in the early sixties with a reverse takeover of GEC by the company he had married into. There followed a period of expansion by takeover and ruthless asset stripping. Many of the great (and admittedly ramshackle) companies disappeared into the GEC maw, including AEI, English Electric and Elliot Automation. Hundreds of fine industrial research laboratories were closed, as well as many factories. Using methods pioneered by Scrooge and Marley, he built up the notorious GEC cash mountain. He was largely responsible for Britain opting out of the burgeoning new technologies, not only semiconductors but, more crucially, the new high quality electronic production engineering being pioneered by the likes of Hewlett and Packard. He was also largely responsible for the infamous brain-drain, which drove many British scientists and engineers abroad. Many could not tolerate the moral climate in which cash was built up by a deliberate policy of late payment of small suppliers. At one time, in fact, orders went out that no small suppliers were to be paid at all unless they were sole-suppliers. The cash mountain was thus built on the bones of many small companies as well as the better-known large ones. Owing to a bizarre alliance with ardent left wing socialist Anthony Wedgewood (later Tony) Benn, Weinstock had the support of the Labour Government’s Industrial Reorganisation Corporation in building his empire. GEC became so dominant in this “Wedgestock” era that it was believed that it could obtain major Government orders simply on the threat of putting a whole town, such as Rochester, out of work. There were many technological disasters, such as the Torpedo Project, but the most egregious was the Nimrod Early Warning system. To the astonishment of many of us associated with the industry at the time, GEC were awarded the contract even though they did not have a working airborne computer. £800 million later, when the project was cancelled, they still did not have one.
The great irony of his life, however, was to come towards the end, when the couple of clowns who were his successors dissipated the whole cash mountain, being suckered by the Internet bubble (see Number of the month a year ago). Not only his creation, but a large part of his personal fortune disappeared.
Sic transit gloria arnie.
The pay off
There is, of course, one form of life way below platyhelminthes in the natural order, and that is the American compensation lawyer. Reader Steve Dillon sends the first hot-off-the-press advert:
Hormone Replacement Therapy (Estrogen plus Progestin)
Hormone Replacement Therapy (HRT) is a common treatment for
women that helps prevent osteoporosis and certain kinds of cancer.. HRT
typically involves drugs such as Prempro® which combine estrogen and
Progestin, a synthetic form of progesterone. Estrogen, when used alone, has
been linked to an increased risk of uterine cancer. When combined with
Progestin the increased risk of uterine cancer is nearly eliminated.
Together, these drugs have been used for years to treat the symptoms of
menopause such as hot flashes and reduce the risk of osteoporosis. Common
HRT medications include: Prempro®, Premphase®, Premelle®, Premique®,
In a July 9th press conference, Claude Lenfant, M.D., Director
National Heart, Lung and Blood Institute announced the cancellation of a
clinical study on Hormone Replacement Therapy due to an increased risk of
invasive breast cancer and evidence that overall health risks exceed any
benefits. There are an estimated 50 million post-menopausal women in the
U.S. It¹s estimated that about 6 million of these women take estrogen plus
Research shows that women who used estrogen plus progestin in Hormone
Replacement Therapy experienced higher rates of serious side effects
compared to women taking a placebo. These serious side effects include:
A 41 percent increase in strokes
A 29 percent increase in heart attacks
A doubling of rates of venous thromboembolism (blood clots)
A 22 percent increase in total cardiovascular disease
A 26 percent increase in breast cancer
If you or a loved one has suffered any of the serious side effects listed
above while taking estrogen plus progestin, call now about your legal
Statutes of Limitation Notice
All states have limits on the amount of time that can go by before you file
a lawsuit (called statutes of limitation). To preserve any potential legal
claims for allegedly dangerous and/or defective medical devices or
medications, you should contact an attorney right away. Otherwise, you may
lose your chance to recover monetary damages forever.
The links for a free case assessment have been removed. Number Watch has no intention of aiding their promotion. This is what the authors of the "study" knowingly launched when they published their dramatic announcement of prematurely ending the statistical assessment. Dangerous and destructive or what?
Number of the month 1.0
This is the level of relative risk (RR) that corresponds to zero effect in a statistical survey to establish a correlation. 95% is the level of confidence commonly espoused by epidemiologists, but elsewhere in science regarded as too undemanding. Both these figures appear in the justification for the dramatic termination of the HRT study above. You have to keep repeating it to believe that it happened – the chance that there is no relationship between HRT therapy and breast cancer lies within the rather lax limits set by the investigators.
It also brings into consideration another factor in trying to assess the claims of such studies. For, if a prior and not very demanding criterion is set for terminating a trial, it acts as an additional guarantee that the criterion will be met. If you think of the estimate of the risk as a random walk that goes up and down as the statistics gather, then at the point at which the criterion is met, it might be about to go up or go down. Terminating the trial eliminates the chance of it going down again. It is all very reminiscent of another precipitately terminated trial, the Tamoxifen fiasco, which your bending author described elsewhere:
Tamoxifen is a drug that was thought to be effective in the prevention of breast cancer. Trials were started in the USA by the National Cancer Institute. A British-led study, involving 4,500 women in Britain, Australia, New Zealand, Spain, Finland, Switzerland and Belgium was hoped to reach a total of 7,000 women, with the results not expected for several years. Suddenly the Americans stopped their trial and announced that tamoxifen reduced breast cancer by 45%. The scientists from the rest of the world were outraged. Not only had the Americans ruined the trial by “unblinding” it, but they demonstrated that they had been cheating all along – the whole point of a blind trial is that you do not know how the results are progressing, so that you cannot exercise and influence over them.
Officials at the institute defended their decision to end the 6-year-old tamoxifen drug trial. According to Associated Press (April 7, 1998) Dr Leslie Ford said that the agency was following the standards set for the tamoxifen trial when it started. She said the trial was designed to ask whether tamoxifen could prevent breast cancer. “We all felt the question had been answered”. She said a statistical evaluation showed that there was a clear difference in the incidence of breast cancer among women taking the drug compared to women who were not. Since half of the 13,388 women in the trial were getting placebo, the study was stopped so that all of them could benefit from the drug.
“The Americans have unblinded the trial, which means it will be unbalanced and they will not be able to answer many questions,” said Dr. Trevor Powles of the Royal Marsden Hospital, London. Powles, who in 1986 headed the first pilot study on the tamoxifen‘s supposed anti-carcinogenic effects, added “It looks as though the benefits of the drug are likely to substantially outweigh the risksbut it was too early to be sure. Dangers include an increased risk of getting endometrial cancer – cancer of the lining of the uterus – and blood clots in the lung.”
“You start to wonder what the hidden agenda is,” said Michael Baum of University College Hospital, London, the other co-chairman of the British-led study. “Is the National Cancer Institute of America trying to defend its budget or something like that? And I don’t think this is just sour grapes or British conservatism.”
The fact is that we may now never know the benefits and dangers of this drug. Indeed, the British and Italian trials failed to confirm that there was any benefit at all. Shares in the British-based Zeneca Group, which produces the drug, rose immediately on the American announcement. It is yet another example of scientific publication by press conference. The people involved ought to be drummed out of whatever learned society they belong to