Patrick Holford regularly accuses scientists, doctors and the pharmaceutical industry of withholding or distorting research findings. In a promotional video for Food is Better Medicine than Drugs Holford casts himself and co-author Jerome Burne as Woodward and Bernstein figures battling against omnipotent forces. Holford espouses a belief in:
full spectrum dominance. [The pharmaceutical industry] literally influence the researchers, the journals, the doctors…So the information we get just simply isn’t the truth.I’ve spent the last year with top, award-winning, medical journalist Jerome Burne, exposing the truth about the pharmaceutical industry…
We’ve…found that about 20% of drug prescriptions are given to patients for whom there is no evidence that it works at all. For example, statin drugs, designed to lower cholesterol. Almost 20,000 people have to take a statin for 5 years for one less heart attack-that is, if you haven’t already had a heart attack.
Scary stuff. Did they obtain this information by interviewing whistleblowers in dark garages? Did operatives sneak past a fully-operational Death Star to smuggle out documents from high-security data archives of suppressed findings? No. In Food is Better Medicine than Drugs, they refer you to a URL for a news story. Not the openly-published-in-a-journal-study on which the news story was based but the news story. And Holford gets his account of that story wrong-badly wrong-but what else is new?
If we were discussing lesser researchers, one might think that that’s what happens if you were to rely upon scanning a news item rather than taking time to read the original paper.
The URL given in Food is Better Medicine than Drugs no longer works. However, you can use the information in the rest of the reference CBC news: Improved cholesterol drug prescribing habits, medical team advises; 11 April, 2005 to track down the news item. You might recall that Holford claims that:
almost 20,000 people have to take a statin for 5 years for one less heart attack-that is, if you haven’t already had a heart attack.
That is not what the authors claim: The 2003 Canadian recommendations for dyslipidemia management: Revisions are needed. It’s not even what the news item reports:
The ICES team looked at Canadians aged 18 to 74 between 1988 and 1992 who were considered at low risk for heart disease but who qualify for statin therapy under current guidelines. Their goal was to run a reality check, estimating how many such people would need to take the drugs to save one life.They concluded 19,600 people would have to take the medication for five years to prevent a single death from heart disease. [Emphasis added.]
The news team didn’t report the nuances of the paper accurately (more below) but they did manage to mention that the analysis involved a “low risk” population. However, Holford uses this study as an example of current ineffective and expensive prescribing practices and generalises it to the wider population.
The ICES authors ran a theoretical calculation to estimate how many more people would qualify for statin therapy etc. if new, revised guidelines were to be implemented in an absolutist fashion. More relevantly, the team estimated the Number Needed to Treat (NNT) for an acknowledged low risk or very low risk population, not the more general statin-taking population to which Holford refers, which includes people who may not have had heart attacks but still have higher risk levels, diabetes, kidney disease, relevant family history etc. These are fantastically different populations and might have NNTs that would differ by one to several orders of magnitude.
The ICES authors acknowledge these differences in the paper. What did the authors say about that remarkable, headline-grabbing (and, not to give away the punchline, misreported) calculation?
The number of low-risk people for whom statins are recommended if lifestyle modification fails to achieve target lipid levels increased 10-fold, from 61 000 to 595 000. Yet, the relative benefit of statin therapy for many people in the low-risk group is incredibly small. By our estimates, among the 149 000 Canadians who fall within thehighest-risklowest-risk quartile in this group, the number needed to treat with 5 years of statin therapy to prevent 1 CAD-related death would be 19 600. The number needed to treat to prevent 1 CAD-related death in the low-risk group recommended for statin therapy…would be 1550. [Strikethrough reflects the ICES authors' correction; however, it was immediately obvious from the context in the original. Emphases added.]
So, if you take a group of people who are already low-risk, and you examine the sub-group of these people who have the very lowest risk of all, they have an NNT of 19,600. The NNT for the low-risk group is 1550. So, that’s an order of magnitude of difference within the low-risk group.
Further, the ICES authors estimated the NNT to prevent death. Holford has somehow transformed this into “one less heart attack”. It may seem like a small point, but it isn’t to the person affected or their family: not all heart attacks are fatal. These are very different calculations and would yield (even for this population) a substantially lower NNT than either 19,600 or 1550.
There is considerably more amiss with Holford’s claims but it would be more appropriate to discuss them in a separate post (who knew that there would be so much material arising from a short promotional video?).
Holford regularly accuses scientists and researchers of misrepresenting or distorting research and engaging in cover-ups: this is the more remarkable in the light of his own gaffes concerning premature mortality rates. Holford’s constant refrain is that “the information we get just simply isn’t the truth”. The study that is the source for his remarkable claims about statins is freely available to anyone who cares to read it. Holford is amongst the people who have misreported it and inappropriately generalised the findings.
Holford explicitly chose this statins study to promote his book and to support his argument that Big Pharma is deceiving us into accepting and paying for ineffective drug therapy. We’ve already looked into the substantial costs of Holford’s proposed alternatives which he seems to want people to accept without ever having to proffer an NNT.
If you have any concerns about the appropriateness of statins, then you might be well-advised to discuss them with your GP or similar health professional: you may wish to be guided by Holford. The appropriate use of statins is a hotly-debated topic and recommendations for their role in primary prevention are provisional at best. Consequently, healthcare commentators tend to be cautious both about warnings and optimism, tacking between Scylla and Charybdis so as to avoid creating either undue anxiety or complacency. Sensationalising a message may give attention-grabbing headlines or promotional sound-bites but it may have long term consequences for the public understanding of this issue.
Update Jan 22 2008: Further contributions to the hot-topic of the appropriateness of statins: Business Week Do Cholesterol Drugs Do Any Good/
22 Comments
June 6, 2007 at 5:43 pm
Congratulations, Shinga, for your meticulously researched analysis of Patricks comments.
But how sad that the public have to have such a detailed deconstruction to gain the same ‘credibility’ rating as a Holfords broadcast…. I guess he has pursuaded so many people over time that he “is a leading clinical nutritionist” that they actually, lemming like, believe him.
Well done to you – and Jon – for redressing the balance.
You quote Holford as stating:
“the information we get just simply isn’t the truth”
Probably the most truthful statement he’s ever made…..
June 7, 2007 at 8:44 am
Thanks, Anon. It did look like a statement of astonishing candour but in context it was clear that he was casting aspersions on others…
I hope that some readers find it helpful- I have little optimism that the references etc. will convince those who think that it is wrong to question the misrepresentations or bad science behind some of Patrick Holford’s output.
Regards – Shinga
June 7, 2007 at 9:59 am
I can’t help thinking of the title as “Is Drugs Better Food than Medicine?”
June 7, 2007 at 11:41 am
“Is Drugs Better Food than Medicine?”
Now – why do I see a man in a yellow tracksuit and plenty of bling asking that question?
That sounds like it should be an essay question.
Regards – Shinga
June 7, 2007 at 12:19 pm
John Garrow has written a suitably brutal review of ‘food is better medicine than drugs’ in the Jan 2007 HealthWatch newsletter. The reviews takes on this issue of definitions of food/drugs:
“The authors of this book attach an unusual meaning to the words “food” and “medicine”…“Medicine” is normally something you take to help you to prevent, or recover from, an illness…we need to define “food”. If the term includes alcohol and saturated fats then foods certainly do have unpleasant and sometimes deadly side effects…the authors of this book can hardly claim the moral high ground, since they themselves certainly downplay the side effects of food. Indeed they pervert the meaning of the word “food”, by omitting everything undesirable, and incorporating aspects of non-drug lifestyle (such as not smoking, avoiding stress, taking exercise, not hyperventilating, limiting alcohol intake) that cannot reasonably be included under the heading of “food”.”
Confusion of food, medicine and drugs could have problematic effects. I’d imagine that confusing breathing and eating could cause really nasty problems, too ;)
June 9, 2007 at 3:06 pm
I’ve been citing that statins figure in a number of classes. I fell for the “it was in a book it must be true”.
You do know that he complains about the fact that journalists can miss important findings because the findings are “hidden in the body of the paper, which almost nobody ever reads because they depend on the summary”.
June 10, 2007 at 7:19 am
Holford and Burne make that ‘original paper’ comment and several others like it in the book. Amongst my favourite: “The debate over statins can get a bit complex, as it is all too easy to become mired in interpretations of trial results and biomedical statistics”. It’s nice to see that Holford and Burne managed to stay mostly out of the mire with this paper…
Regards – Shinga
July 24, 2007 at 4:31 pm
[...] in some parts. E.g., Why IgG Testing for Food Intolerance Is Not As Simple As ABC or Doh Ray Mi Statins and Why Patrick Holford Is Breaking My Heart: Part 2 Never Mind the Research Quality, Feel the Fear: Justifying Homocysteine [...]
August 19, 2007 at 5:46 pm
[...] than Drugs and New Optimum Nutrition for the Mind – finding everything from basic mathematical and statistical errors to the recommendation of pointless (and potentially harmful) medical [...]
September 25, 2007 at 4:43 pm
[...] with outstanding examples like the QLink and the poor understanding of the general literature that Holford exposed in his breathless reporting on statins, one might have some misgivings about the claims of scientific research and rigour behind some of [...]
December 18, 2007 at 6:39 pm
[...] December 18, 2007 · No Comments Professor Patrick Holford of Teesside University and Head of Science and Education at Biocare frequently upbraids professionals and researchers for what he perceives as their lack of up-to-date research. [...]
January 6, 2008 at 5:40 pm
[...] January 6, 2008 · No Comments Professor Patrick Holford of Teesside University and Head of Science and Education at Biocare frequently upbraids professionals and researchers for what he perceives as their lack of up-to-date research. [...]
January 25, 2008 at 3:26 pm
[...] January 25, 2008 · No Comments Professor Patrick Holford of Teesside University and Head of Science and Education at Biocare frequently upbraids professionals and researchers for what he perceives as their lack of up-to-date research. [...]
February 3, 2008 at 10:37 pm
I attended Dr. Mehmet Oz’s keynote at KCBS Health etc. http://www.kcbs.com/pages/234305.php. He’s the doctor on Oprah Winfrey’s show and author of popular books that prescribe consumers taking more control of their health.
I liked the position he takes if the cholesterol is so high that the chances of heart attack are high then statins, but for borderline cases diet and exercise. Read more here – [url deleted as undeclared commercial site: similarly clickthrough link deleted from name for some reasons of non-declaration and transparency]
seems like a sensible approach.
February 5, 2008 at 12:57 am
It is a sensible approach and really no different from standard medical practice (I believe). Take a look at the clinical guidelines for lipid management: Management issues
Diet, exercise, then statins. The trouble with diet and exercise is that people aren’t very good at sticking to them. Which is sad, but understandable.
February 5, 2008 at 3:03 am
Indeed. Except for those people who are prescribed statins because they’ve already had (say) a cardiac incident. However, I know a number of people who are following diet and exercise guidelines from their GP and have reduced their lipids-of-concern quite nicely. Thanks for the link to the management issues – very useful.
February 28, 2008 at 10:34 pm
[...] to have noticed these nuances. This would be unremarkable except that Holford regularly accuses scientists and researchers of hiding or misinterpreting research. Similarly, he claims that there are times when the MSM fail to read and understand a full paper [...]
March 8, 2008 at 4:11 pm
[...] Info Pack (pdf) [4] Patrick Holford, Professor Andre Tylee and the Endorsements That Cannot Die [5] Statins and Why Patrick Holford is Breaking My Heart: Part 2 [6] Ben Goldacre on Patrick Holford and the QLink [7] Myth: The Scientific Support for Chromium and [...]
March 28, 2008 at 3:29 am
[...] be lucrative book deals for Patrick Holford Mis-Reports Papers, Can’t Do Sums, and 101 other gaffes of varying significance. There wouldn’t be any prime-time slots on GMTV or Tonight with [...]
May 18, 2008 at 2:14 am
[...] of Teesside University and Head of Science and Education at Biocare regularly portrays himself as a fearless advocate for scientific accuracy and rigour. Despite an abundance of evidence to the contrary, he is so convinced of his own competence that [...]