At this point in the book, I move off the edge of the charts and into unexplored lands maked ‘Here be dragons.’ For while it’s true that the bulk of the mainstream still supports the diet part of the diet-heart hypothesis, there are many researchers who have long since given up on the idea – including such notables as Professor Michael Oliver. So I am far from being a lonely traveller in that area.
But when it comes to attacking the second part of the hypothesis, namely that raised cholesterol levels cause heart disease, I find myself in the wilderness. There are a few others moving around in this landscape, true, but not many. To be frank, a number of them have a tendency to write in green ink and regularly howl at the moon, whereas I only howl at a full moon.
I am also fully aware that when I talk about raised cholesterol levels causing heart disease I am in the realm of the ‘known fact’ – a fact that has apparently been proven beyond the slightest doubt, time and time again.
However, despite that fact that hardly anyone else agrees with me, I believe firmly that the cholesterol hypothesis is wrong. By the time you have finished this chapter, I hope to have convinced you of this fact too.
Before starting on the demolition job I must admit that, for many years, I too believed that a raised cholesterol level caused heart disease. On the face of it the evidence seemed overwhelming, and it also seemed to make sense. The most powerful facts, at least to me, were the following:
| Fact One: | Atherosclerotic plaques contain a lot of cholesterol, which must have come from the blood. So heart disease had to have something to do with cholesterol-containing lipoproteins. |
| Fact Two: | People with familial hypercholesterolaemia (FH) die very young from heart disease, sometimes as young as five. |
| Fact Three: | Statins lower cholesterol levels and protect against heart disease. |
| Fact Four: | ‘Normal’ people (without FH) with higher cholesterol levels are more likely to die from heart disease. |
These facts seemed concrete and inarguable. Every time I opened a journal, or read a paper, they were confirmed. Again and again.
But these facts are really only partially true. They are rather like the false-fronted buildings used in Westerns. If you look at them from dead ahead, you see what looks like an entire town laid out in front of you. But if you move sideways, just a little bit, you can see that the supposedly solid buildings are just four-inch-thick plywood with nothing behind them at all.
And so it is with the second part of the cholesterol hypothesis – ‘raised cholesterol/LDL causes CHD’. Seen from one angle, the facts look solid. But once you decide to quit the ‘opinion leader guided tour’, you get a completely different view. And so, ladies and gentlemen, it is time for a backstage trip around the cholesterol hypothesis.
Gasp as you see the real facts exposed for the first time!
Scream as the fearsome Austrian study bares it claws! Prepare to be amazed as the awesome three-headed Honolulu trial eats LDL in front of your very eyes!
Roll up, roll up. Only two and thruppence for adults, and children under six are free!
CHOLESTEROL LEVELS AND STROKES
I shall start the discussion by moving sideways for a moment to talk about a slightly different manifestation of cardiovascular disease – strokes. Why am I talking about strokes instead of heart disease? Well, strokes and heart disease are part of the family known as cardiovascular disease (CVD). People with heart disease are far more likely to get strokes, and vice versa. Strokes also kill very nearly as many people as heart attacks, so this is not some minor problem.
A stroke happens when blood supply to a part of the brain is cut off. The brain tissue downstream dies, and the victim will lose some brain function. A small stroke is sometimes known as a transient ischaemic attack (TIA); a big stroke can be fatal, or leave the victim with severe disability.
The most common cause of a stroke is the development of a big, nasty atherosclerotic plaque at the base of neck, in the carotid arteries. Clots form over these plaques. The clots can then break off and travel into the brain, where they get jammed into an artery and block the blood flow.
Given the fact that both strokes and heart disease are caused by the development of atherosclerotic plaques, you would think, would you not, that if a raised cholesterol level is a risk factor for heart disease, it would also be a risk factor for stroke? But it is not.
In 1995, The Lancet published a massive study that looked at 450,000 people over a period of 16 years who suffered, between them, 13,000 strokes. This represented 7.3 million person-years of observation. Frankly, that’s quite long enough for anybody. And the conclusions thereof: ‘There was no association between blood cholesterol and stroke.’
More recently, a pan-European study known as EUROSTROKE, published in 2002, asked the same question. The result: ‘This analysis of the EUROSTROKE project could not disclose an association of total cholesterol with fatal, non-fatal, haemorrhagic or ischaemic stroke.’
There are many other studies showing exactly the same thing. So, you have two conditions – stroke and heart disease – that are both fundamentally a form of arterial disease. Yet, raised cholesterol is a risk factor for one, but not the other. Listed below is a slightly shortened list of risk factors for stroke from the American Stroke Association:
• High blood pressure
• Tobacco use
• Diabetes
• Carotid or other artery disease
• Other heart disease – people with coronary heart disease or heart failure have a higher risk of stroke
• Physical inactivity and obesity
• Excessive alcohol intake
• Some illegal drugs – intravenous drug abuse carries a high risk of stroke
• Cocaine use has been linked to strokes and heart attacks
• Increasing age
• Sex (gender) – stroke is more common in men than in women
• Prior stroke or heart attack – someone who has had a stroke is at much higher risk of having another one. If you’ve had a heart attack, you’re at higher risk of having a stroke too
Given that these are precisely the same risk factors as for heart disease (in fact, some of them are heart disease), where is cholesterol in this list? Even more critical for this discussion, how can lowering cholesterol with statins reduce the risk of stroke (which they do), if a raised cholesterol level isn’t a risk factor for stroke? This most certainly does not make sense.
Actually, it would make perfect sense if you believe that any benefit gained from taking a statin has nothing to do with lowering cholesterol levels. But this explanation cannot be allowed by the medicine world at large, or else the entire cholesterol hypothesis crumbles to the ground.
In fact, in its quiet, academic sort of way, this was a crisis! The inner keep of the castle was under fire, the enemy had got in unnoticed through an underground tunnel. No one had expected an attack from this direction. The cholesterol hypothesis had to be protected. But how? How indeed. Tricky, this one. Very tricky.
But you know, the cholesterol brotherhood has some very clever boffins on its side. Although, in this area I think we should trust the instincts of Montaigne:
I prefer the company of peasants because they have not been educated sufficiently to reason incorrectly.
Michel de Montaigne
By the way,
I quote others only in order the better to express myself.
Michel de Montaigne
Step one to protect the cholesterol hypothesis was to split strokes into two basic types.
• Ischaemic
• Haemorrhagic
An ischaemic stroke is caused when a small blood clot travels into the brain, then gets jammed as the arteries narrow. The bigger the clot, the bigger the artery that gets blocked and the bigger the stroke. Around 75 per cent of strokes are ischaemic.
A haemorrhagic stroke happens when an artery in the brain bursts. This causes blood to escape into the brain tissue and cause damage. Haemorrhagic strokes are, generally, more deadly than ischaemic strokes.
Once you have split strokes into two types, you then state that an increased cholesterol level causes ischaemic strokes. On the other hand, a low cholesterol level does not cause haemorrhagic strokes, it is merely associated with haemorrhagic strokes. (Yet another ad-hoc hypothesis plucked from thin air.)
Then – goodness me, this is getting complicated – if you lower cholesterol levels, you will prevent ischaemic strokes, but you will not cause an increase in haemorrhagic strokes. Which is why lowering cholesterol levels with statins can reduce the overall rate of stroke – even if a raised cholesterol level is not a risk factor for stroke. Phew! I’m glad we sorted that one out.
At which point you can add in the one missing risk factor from the American Stroke Association list. Yes, it’s cholesterol. I cut it out of the list – what a naughty boy. But I did it for a reason. Now you can read it with open eyes:
A high level of total cholesterol in the blood (240 mg/dL or higher [about 6 mmol/l]) is a major risk factor for heart disease, which raises your risk of stroke. Recent studies show that high levels of LDL [‘bad’] cholesterol (greater than 100 mg/dL [about 3 mmol/l]) and triglycerides (blood fats, 150 mg/dL or higher [about 4.5 mmol/l]) increase the risk of stroke in people with previous coronary heart disease, ischemic stroke or transient ischemic attack (TIA).
http://www.americanheart.org/presenter.jhtml?identifier=4716
This is so carefully crafted that if you read it without prior knowledge you would think it said that increased cholesterol levels increase the risk of stroke. But it doesn’t. What it says – if you read it very carefully – is that a high cholesterol level is a major risk for heart disease – and heart disease, in turn, increases your risk of stroke. Hmmmm! This is teleoanalysis again:
• A (a raised cholesterol level) leads to B (heart disease).
• B (heart disease) causes C (stroke).
• A isn’t a risk factor for C in any study.
• But A acting through B causes C.
• Thus, A does cause C – huzzah! Faultless logic.
The American Stroke Association then goes on to state that raised LDL levels increase the risk of stroke – in people who already have heart disease, or who have already suffered a stroke. Now, I could chase myself round in circles trying to dissect the logic in that passage. But I will just make one point. What this passage does not say is that raised cholesterol is a risk factor for stroke. Why not? Because that would be a lie. And the powers that be do not lie – they just ensure that the truth lies sleeping atop a very high tower, guarded by fierce beasties with awfully sharp teeth.
As a general point, I will just say that it is a damn sight easier to create ad-hoc hypotheses, and pluck conjectures from the sky, than it is to disprove them. Usually, by the time you have managed to do so, everyone’s eyes have glazed over. Or the whole argument has become so complex that you forget where you started.
But I am going to hunt this one down, because I believe it is kind of critical. What I am going to show you is that a low cholesterol level is actually associated with a massive increase in death from stroke, and may even be a cause. Let’s start with a helpful little passage:
Epidemiological data are generally consistent with the animal experiments, they indicate that diets which are very low in fat increase the occurrence of some forms of stroke. Societies with a low intake of fat and animal protein, such as traditional Japan, tend to have high rates of haemorrhagic stroke. An elevated risk of stroke is found among segments of the Japanese population with low levels of serum cholesterol, particularly among those with high blood pressure.
In a large, screened population of men in the USA, those with the lowest serum cholesterol levels had an elevated risk of haemorrhagic stroke.
This suggests that a low cholesterol level may actually cause haemorrhagic stroke. Is this effect powerful enough overcome the theoretical benefits of low cholesterol in preventing ischaemic stroke? To answer this question, we need to move to Japan, land of the rising sun and the falling stroke. While the Japanese have always had a low rate of heart disease, they used to have the highest rate of strokes in the world. At one time their rate of stroke was 30 times their rate of heart attacks.
In fact, death from stroke represented such a huge health problem that, in the not-too-distant past, the Japanese were being actively encouraged to raise their fat intake to prevent so many of them dying of strokes. This was probably good advice, as confirmed by a 15-year Japanese study published in Stroke in 2004:
The risk of death from [cerebral] infarction [AKA stroke] was reduced by 64% in the high cholesterol consumption group, compared with the low cholesterol consumption group… Animal protein was not significantly associated with [cerebral] infarction after adjustment for animal fat and cholesterol…
This study suggests that in Japan, where animal product intake is lower than in Western countries, a high consumption of animal fat and cholesterol was associated with a reduced risk of cerebral infarction death.
Compare and contrast this hugely positive result with the miserable failure in any trial to show that reducing saturated fat in the diet prevents heart disease. I know that I am supposed to have moved on from discussing the diet-heart hypothesis, but hey! This is just too good to resist! A high consumption of saturated fat reduces the stroke rate by 64 per cent. Reducing saturated fat in the diet reduces the risk of heart disease by 0 per cent.
Perhaps as a result of the advice to increase fat consumption, or perhaps as a result of enemy infiltration by fast food restaurants, in the last 50 years fat, and saturated fat, consumption has gone up in Japan, as have cholesterol levels. (See table below.)
| 1958 | 1999 | |
| Total Calories | 2,837 | 2,202 |
| Carbohydrate intake % calories | 84 | 62 |
| Protein intake % calories | 11 | 18 |
| Fat intake % calories | 5 | 20 |
Virtually a doubling of protein intake, and a quadrupling of fat intake. Oh my God, what happened to the cholesterol levels? They went up by 20 per cent:
Cholesterol levels 1958 = 3.9mmol/l
Cholesterol levels 1999 = 4.9mmol/l
The poor devils, they must have started to drop like flies from heart disease. Ah, no.
Fig. 20 CHD mortality in Japanese men, 1965–95
And just look what has happened to the rate of stroke:
Fig. 21 Death rates from stroke in Japanese men (aged 60–69), 1950–95
Good golly, Miss Molly! I think we have another paradox on our hands. A double paradox, no less. If a raised cholesterol level does actually cause ischaemic stroke, and 75 per cent of strokes are ischaemic, then a 20 per cent rise in cholesterol levels across the board ought to – really ought to – increase the rate of stroke. (Note: this graph does not distinguish between the two types of stroke.) Instead, between 1965 and 1995 the rate of stroke fell from 1,334 to 226 (per 100,000/year). This is a 5.9-fold reduction. Five point nine. You know, that’s very nearly six.
And if you are wondering why I chose the age group 60–69, it wasn’t because this particular age make my case stronger. It just seemed a reasonable age to look at. Had I chosen men aged 55–59, the rate of stroke fell from 463 to 81 (per 100,000/year). If I had chosen 75–79, the rate of stroke fell from 3,470 to 851 (per 100,000/year). These figures represent pretty much the same proportional drop. And it is gigantic. In fact, it is the greatest fall in death rates I have ever seen for any disease in any population – ever.
So what does this prove? Well, it doesn’t prove anything, because epidemiological data can only suggest a connection, or a lack of a connection. However, in Japan, as cholesterol levels went up, death rates from two of the main cardiovascular diseases fell dramatically. Ergo, these data very strongly suggest a causal connection between raised cholesterol levels and cardiovascular disease is [ ] unlikely. (Insert adverb of your choice into the brackets above.)
You can splutter all you like about paradoxes, and that one country ‘does not prove anything’. Japan may be just one country, but it is a country of 115 million people. So, in reality, it is 115 million inscrutable paradoxes. I think if I did a clinical trial on 115 million people, most scientists would consider that to be adequately ‘powered’ to disprove the null hypothesis (don’t worry, just a bit of statistical jargon).
My takeaway point in this section on strokes is as follows. According to mainstream thinking, ischaemic strokes are caused by raised cholesterol levels, and ischaemic strokes represent 75 per cent of all strokes. However, over the last 50 years, cholesterol levels have risen by 20 per cent in Japan, and the rate of stroke has fallen off the edge of a cliff – dropping 600 per cent. And the rate of heart disease has also fallen dramatically. Gentlemen, try to fit those pieces of a jigsaw puzzle together. (Here’s a hint. Some of the pieces may, currently, be upside down.)
CHOLESTEROL/LDL AND TOTAL MORTALITY
Having looked at stroke, and the evidence that the greatest risk factor for stroke is a low cholesterol level, not a high cholesterol level, I think that it is time to introduce the concept of ‘total mortality’.
You see, it is actually possible to die of things other than heart disease, although to hear a cardiologist speak you would sometimes think not. They are utterly obsessed with cardiovascular deaths. Benefits in this area are trumpeted to the very skies. Yet overall mortality is often overlooked; in some trials these data isn’t even published at all.
Speaking personally, I think that total mortality data are by far the most important thing. I’m not that bothered about exactly how people die. Nor, I suspect, are most people. It’s the dying bit we are all trying to prevent, or avoid. Indeed, to be perfectly honest, a massive heart attack seems preferable to dying slowly of cancer. Maybe you think not. It’s probably a matter of personal taste. So I think it is interesting to look rather more closely at the association between cholesterol levels and total mortality. That is, mortality from everything. Heart disease, cancer, respiratory diseases, digestive diseases – the works.
Time, then, to look at the Conference on Low Blood Cholesterol and Mortality, which gathered together the data from 523,737 men and 124,814 women, and reported back in 1992. I think you should probably go and make yourself a cup of coffee at this point, because I do not think you are going to believe the data that I am about to present. So steady yourself.
Firstly, the overall mortality data from women:
Fig. 22 Risk of death at various cholesterol levels in the next five years – women
I think that’s pretty clear, it is not? The healthiest cholesterol level is somewhere around about 5.5mmol/l. I know that this data is on total cholesterol, not LDL. But I can assure you that the two things are tightly bound. In study after study, total cholesterol was as good a predictor of death as LDL alone – if not better. And a higher total cholesterol level means, 99 per cent of the time, a higher LDL level.
Now, someone like me might look at that data and wonder why the current recommendations are that we should all strive to get the cholesterol level below 5.0mmol/l. Indeed, the most recent guidelines recommend that we should be aiming to get below 4.0mmol/l. To be frank, one look at the diagram on female mortality should tell you everything you need to know about that idea. (Top tip: look to the left and upwards on the graph.)
Anyway, on to men and total mortality:
Fig. 23 Risk of death at various cholesterol levels in the next five years – men
Not quite the same pattern as with women, more of a U-shaped curve. But it’s still hardly a graph that suggests cholesterol is a deadly killer, as the highest mortality rate is to be found at the lowest cholesterol level.
Just for the heck of it, below is a graph showing the rate of non-cancer, non-cardiovascular mortality in women. This one just keeps going down as cholesterol levels go up:
Fig. 24 Risk of non-cancer non-cardiovascular death at various cholesterol levels in the next five years – women
But you shouldn’t worry about low cholesterol levels. Why not? Because no one at the conference did:
Most participants considered it to be likely that many of the statistical associations of low or lowered TC (total cholesterol) level are explainable by confounding in one form or another. The conference focused on the apparent existence and nature of these associations and on the need to understand their source, rather than on any pertinence of the finding for public health policy.
Dr Yusuf, who was a major player at the conference, noted that the excess of non-cardiac deaths was: ‘… of borderline statistical significance, was spread over a number of causes, and was not related to the strength of the intervention’. Yusuf interpreted these findings as ‘biologically implausible and probably due to chance’. And thus was any association between low cholesterol levels and increased rate of death airily waved away. No need for the public to worry their pretty little heads about such matters.
But you know, perhaps the public should worry their pretty little heads. Because a key finding from the Framingham Study was the following.
There is a direct association between falling cholesterol levels over the first 14 years [of the study] and mortality over the following 18 years (11% overall and 14% CVD death rate increase per 1mg/dl per year drop in cholesterol levels).
Yes, you did just read that. Those people whose cholesterol levels fell, were at a greatly increased risk of dying – and at an even greater risk of dying of cardiovascular disease. I shall expand on these figures a bit.
The figures on total mortality show an 11 per cent overall increase of death for each 1mg/dl drop in cholesterol levels, which doesn’t sound that bad. But remember that mg/dl are titchy little US units. To convert into the magnificent jumbo-sized units used in the UK – mmol/l – you need to multiply by 39. So, a quick translation of the Framingham results gives the following: a 1mmol/l fall in cholesterol levels is equal to a (39 x 11 per cent) increase in the risk of total mortality. Which is 429 per cent.
To put this into a real-life context, if your total cholesterol were to fall from 5 to 4 mmol/l, your risk of dying would increase by more than 400 per cent. Not only that, but your risk of dying of a cardiovascular disease would increase by 39 x 14 per cent = 546 per cent.
This might seem so incredible that you may not believe that you read it. But I can assure you that it is there, in black and white, in the Journal of the American Medical Association, 24 April 1987, pages 2176 to 2180: ‘Cholesterol and mortality. 30 years of follow-up from the Framingham Study’.
I hope you recognise by now that I make up nothing. All facts and data that I use come from peer-reviewed, high-impact journals that can be found by looking in the database, www.pubmed.org – a fantastic resource that is absolutely free.
The interpretation of those facts, however – that’s a completely different matter. A few statistical models here, a bit of meta-analysis there, just a sprinkling of confounding variables, a few ‘probably due to chances’ thrown into the mix and, bibbity, bobbity, boo! A circle turns into a square, and cholesterol turns into a deadly killer.
But it is time to return the main point of this particular story, which is that a low cholesterol level, especially after the age of 50, significantly increases your risk of dying. One massive long-lasting study that looked specifically at cholesterol levels and mortality in older people, was carried out in Honolulu and published in August 2001 in The Lancet. And the findings thereof:
Our data accord with previous findings of increased mortality in elderly people with low serum cholesterol, and show that long term persistence of low cholesterol concentration actually increases the risk of death. Thus, the earlier that patients start to have lower cholesterol concentrations, the greater the risk of death.
Their interpretation:
We have been unable to explain our results. These data cast doubt on the scientific justification for lowering cholesterol to very low concentrations.
This study, by the way, was immediately attacked from all sides. I think my favourite attack included the word ‘irresponsible’. Things have come to a pretty pass when publishing a well-designed medical study in The Lancet is considered irresponsible. I mean, people might learn the truth and then there is no way of knowing what will happen. Panicking in the streets, law and order breaking down, the playing of loud and licentious music, egg yolks and meat pies consumed in public places…
But you know, it is not only in the elderly that a low-cholesterol diet is associated with a higher mortality rate. The Austrians carried out a study of 149,650 men and women, looking at cholesterol levels and cardiovascular and all-cause mortality. It was entitled: ‘Why Eve is not Adam: prospective follow-up in 149,650 women and men of cholesterol and other risk factors related to cardiovascular and all-cause mortality’. This study lasted 15 years and looked at nearly 70,000 men, and more than 80,000 women ranging from 20 to 95 years of age who underwent, between them, more than 450,000 examinations. This was a huge study. One of the biggest and longest ever. And I am willing to bet a large sum of money that you have never heard of it.
One of the reasons for this is that it ended up being published in the Journal of Women’s Health. Not that I have anything against this journal – how could I? I had never heard of it before tracking down this study. But why was this not published in the BMJ, or The Lancet or the New England Journal of Medicine? It was of huge public-health significance, yet it ended up in a journal with a relatively low ‘impact’ factor and was thus, effectively, buried.
However, what this study confirmed is that a low cholesterol level after the age of 50 (and under 50, if you are a man) is significantly associated with all-cause mortality:
In men, across the entire age range… and in women from the age of 50 onward only, low cholesterol was significantly associated with all-cause mortality, showing significant associations with death through cancer, liver diseases, and mental diseases.
You can’t get clearer than that. If you have a low cholesterol level, you are at a much greater risk of death.
Perhaps you would prefer a British study? This from the BMJ in 1995:
Low serum cholesterol concentrations (<4.8mmol/l), present in 5% of the men, were associated with the highest mortality from all causes, largely due to a significant increased in cancer deaths.
Or perhaps you would like a Finnish study? This was a report produced 25 years into the Seven Countries Study, published in the American Journal of Epidemiology in 1992:
During the first ten years of follow-up… men with high cholesterol levels had lower all-cause mortality… because of their low cancer mortality and residual mortality.
What about a study in the very old? The oldest old – those over 85. The following was published in The Lancet in 1998:
Each 1mmol/l increase in total cholesterol corresponded to a 15% decrease in mortality.
Or how about this one from France, published in The Lancet in 1989? A small study, admittedly, but quite amazing nonetheless. Ninety-two women living in a nursing home, most of whom died over the next five years. The lowest mortality rate was at an average cholesterol level of 7.0mmol/l, and the highest mortality rate was at an average cholesterol level of 4.0mmol/l. At this level, the mortality rate was 5.2 times higher than at 7.0mmol/l. You probably thought that anyone with a cholesterol level of seven had died fifty years earlier of a heart attack. Not so.
Enough already, I hear you cry. OK, enough already. I shall merely summarise the data on overall mortality:
• Under the age of 50, your cholesterol level doesn’t really make much difference to your risk of dying. However, if your cholesterol level starts falling, watch out. You are at a terrible risk – a 429 per cent increased risk of death per 1mmol/l cholesterol drop, according to the Framingham Study.
• After the age of 50, a low cholesterol level is associated with a significantly greater overall mortality. The older you get, the more dangerous it is to have a low cholesterol level.
Does this mean that a low cholesterol level is, itself, deadly? No, I don’t think so. I do not believe that a low or a high cholesterol/LDL level actually causes anything except, perhaps, haemorrhagic stroke – if the level is very low. I think it is mainly a disease ‘marker’ of a kind. Although, in general, it seems much more dangerous to have a low level than a high level.
Of course, I am not the only person in the world to have noticed that low cholesterol levels are associated with increased mortality. The mainstream research community also picked up on this one. Perhaps, to be more accurate, I should say the mainstream research community has failed to sweep this fact under the carpet. (Or maybe they have, since no one I speak to is ever aware of this fact.) What is their explanation? It is as follows. A falling, or low cholesterol level, is a sign of an underlying disease. Thus it is not the low cholesterol level that kills you, it is the underlying disease.
It is true that certain diseases – e.g. advanced cancer – can create a low cholesterol level, as can liver diseases such as chronic Hepatitis B. This makes it likely that some people with low cholesterol levels are suffering from a serious underlying disease. Therefore, this is one ad-hoc hypothesis with which I am in a certain amount of agreement.
The leading proponent of this hypothesis is a researcher called Carlos Iribarren. I think he was the first to propose the idea that a low cholesterol level indicates underlying disease, and he bangs on about it regularly. Whether it was his original idea or not, the rest of the scientific community fell upon this concept gratefully, and now repeat it as their new mantra. Thus, everyone can reassure themselves with the knowledge that a raised cholesterol really, truly, is deadly. Even when it’s low – perhaps especially when it’s low.
Time to quote from one of Iribarren’s studies, published in the Journal of the American Medical Association (JAMA). This study was designed to prove that a low cholesterol level was not an independent risk factor for death. The conclusion of the study:
We conclude that the excess mortality at low TC [total cholesterol] levels can be partially explained by confounding with other determinants of death and by pre-existing disease at baseline… In our study TC level was not associated with increased cancer or all-cause mortality in the absence of smoking, high alcohol consumption, and hypertension.
So there you go. Once you add in smoking, high alcohol consumption and high blood pressure, you find that low cholesterol levels disappear as a risk factor.
Now, I was explaining that, according to mainstream researchers, a low cholesterol level is not a risk factor for dying, because it is, in turn, caused by an underlying disease, and it’s the underlying disease that kills you – not the low cholesterol level. Maybe in some cases this is true. However, I find the idea that cancer can cause a low cholesterol level – before the cancer can even be detected – somewhat bizarre.
An early stage cancer is smaller than a grain of rice – far smaller. The possibility that 0.1g, or thereabouts, of tumour mass can have a discernible effect on cholesterol levels seems utterly bizarre. How could it? Of course, when you have advanced cancer, this knackers the entire metabolic system. But can cancer do this five or ten years before diagnosis? I think I will go as far as to say that this is impossible.
In fact, I don’t need to rely on such theoretical arguments, because this ad-hoc hypothesis has actually been disproved. The Framingham research team had also noted a high mortality rate in over-50s who had low cholesterol levels. They too wondered if the low cholesterol levels were caused by an underlying illness:
Similar results from several modified analyses make low cholesterol levels due to a severe illness an unlikely explanation for our results.
Sorry about that tortured passage, but I do try to use the exact words of the researchers, rather than put my words in their mouths. (For some reason, people seem to find this more believable.) However, I shall translate. Those with low cholesterol levels did not have a severe underlying illness. They just had long-term low cholesterol levels followed by a much higher mortality rate. And however many ‘modified analyses’ were used, they just couldn’t sweep this association under the carpet without leaving a big bump sticking up in the middle.
The Honolulu researchers also looked carefully at their findings in the light of the Iribarren ad-hoc hypothesis:
Iribarren and colleagues suggested that a decline in serum cholesterol might occur over a decade before diagnosis of a disease [yeah, right – my words], and such long-term morbidity could be attributable to chronic subclinical infections with Hepatitis B, or to chronic respiratory diseases.
… our data suggest that those individuals with a low serum cholesterol maintained over a twenty-year period will have the worst outlook for all cause mortality.
Our present analysis suggest that this [Iribarren’s] hypothesis is implausible and is unlikely to account for the adverse effects of low cholesterol levels over twenty years.
This is as close as one set of researchers will ever come to telling another set of researchers that they are talking complete bollocks. At least in public, anyway.
Just to ram home the point, the Austrian researchers also analysed their data to see if underlying diseases caused the low cholesterol levels:
For the first time, we demonstrate that the low cholesterol effect occurs even among younger respondents, contradicting the previous assessments among cohorts of older people that this is a proxy or marker for frailty occurring with age.
In a way, it’s a shame. I rather like Iribarren’s hypothesis in a kind of last-desperate-throw-of-the-dice kind of a way. Low cholesterol levels are caused by early stage diseases, so early that you can’t actually detect them. So how do we know they are there? Well we can’t, obviously… duh! They’re undetectable, stupid. But we know they must be there, otherwise these people wouldn’t have low cholesterol levels. Yes, it’s the good old circular argument again.
Q: ‘Why have these people got low cholesterol levels?’
A: ‘Because of an underlying disease.’
Q: ‘How do you know they have an underlying disease?’
A: ‘Well, just look at the low cholesterol levels. They couldn’t have such a low level if they did not have an underlying disease.’
What I find perhaps most amusing about this area is the ‘clash of the mighty ad-hoc hypotheses’. On one hand we have Iribarren explaining that a low cholesterol level is caused by underlying diseases. So a low cholesterol level is a sign of being completely knackered. On the other hand, we have another group of researchers – led by the mighty Law and Wald, quelle surprise – explaining that ‘primitive’ peoples have very, very low levels of blood cholesterol, and this is exceedingly healthy. ‘You cannot have a cholesterol level that is too low. Statinate, statinate!’
So a low cholesterol level in the West is a sign of desperate illness, but a low cholesterol level among primitive peoples is a sign of glowing health – something we should all aspire to achieve. Go figure, as they say. I say, go look at the life expectancy of primitive peoples and then tell me how healthy a low cholesterol level might be. ‘Oh, but they have such a high mortality rate because they die of things other than heart disease.’ (Well they would, wouldn’t they – see everything written above.)
Maybe all groups of researchers should get together and try to put together a story that doesn’t keep contradicting itself all the time. Fat chance. To be honest, I don’t think that they are even aware that they are arguing directly against each other in their attempts to defend the cholesterol hypothesis. One lot are digging a hole, and the others are frantically filling it in again. Still, it’s probably good for the GDP.
Moving on from Iribarren, and all the other desperate ad-hoc hypotheses that I really don’t have time to mention, the simple fact is this: a low cholesterol level increases the risk of death in men and women. This is one fact that has never been contradicted by any study. It is also a fact that is so well hidden that no one I have ever spoken to is aware of it. Indeed, when I mention it, no one actually believes me.
It is also rather important. The fact that a low cholesterol level is unhealthy may even make you think about your cholesterol level in a whole new way. Is it around 5.5mmol/l and above? Good. Below 4.0mmol/l? Watch out.
WOMEN AND HEART DISEASE
Now it is time to look directly at cholesterol levels and heart disease. I will start by looking at women and heart disease, an area of research that may otherwise be referred to as ‘The case of the mysterious disappearing fact’ starring female sex hormones, the ever-popular menopause, evil LDL and our plucky hero HDL. Along with a full supporting cast of ad-hoc hypotheses – as always.
It has been recognised for many years that women, generally, suffer much less heart disease than men – especially younger women. The difference is normally about 300 per cent. This is despite the fact that women have higher average cholesterol levels. The widest gap I found was in New Zealand in the 1970s. Here, women aged 45–55 had one-tenth the mortality rate of men. Now that’s what I call a gap.
Women, therefore, present a problem for the cholesterol hypothesis. Higher cholesterol levels than men, but much lower rates of heart disease. This must mean that…
Eager schoolboy: ‘Sir, sir… it must mean that raised cholesterol levels don’t cause heart disease.’
Teacher: ‘You stupid boy. We know that raised cholesterol levels cause heart disease. Anybody else?’
Teacher’s pet: ‘It means that women must be protected against a high cholesterol level, sir.’ (Smug grin.)
Teacher: ‘Well done, Snodgrass, that is the correct answer.’
At this point it is worth presenting some data. These data come, once again, from 1992 and the ‘Report of the Conference on Low Blood Cholesterol: Mortality Associations’, published in Circulation. The researchers looked at all available data on women from 11 major studies or trials, representing 124,818 women. Their conclusions:
Many findings for women were discrepant from those for men. Of particular importance in women was considered to be the essentially flat relation of total cholesterol to total mortality, total cardiovascular, and total cancer mortality.
See graph overleaf (Fig. 25).
Yes, it’s the female paradox.
Don’t worry, you see this is not actually a paradox. (‘Phew, for a minute there you had me worried.’) Women, you see, are protected against a raised cholesterol level by their sex hormones. Ad-hoc hypothesis no. 8,396,249.
For many years, I too, believed that women were protected by their sex hormones. Everyone said it, everyone believed it. After all, protection disappeared after the menopause, doesn’t it? I’m sure I’ve read that many times.
Fig. 25 Risk of cardiovascular death at different cholesterol levels – women
Well, in 1963, a study was carried out on women who had had hysterectomies. Half of the women had their ovaries removed at the same time – thus they had no sex hormones – and half retained their ovaries. (I think I should make it clear that the removal, or retention, of the ovaries was done purely on medical need; if not, this would have been one of the world’s least ethical studies…)
The results:
We found no difference in the prevalence of coronary heart disease in the oopherectomised [both ovaries removed] and hysterectomised [no ovaries removed] women.
The finding of no difference in the arteriosclerotic heart disease rates in the two groups suggests that some factor, or factors, apart from ovarian function are responsible for the relative freedom from coronary heart disease in women as compared to men.
I think that this was the first time anyone actually put the sex hormone ad-hoc hypothesis to the test, and it failed utterly and completely. Of course, the results had no discernible effect on anyone, or anything. Eyes tightly closed to the available evidence, researchers continued to study female sex hormones and the menopause – in relation to heart disease – and they continued to find the same thing:
The normal menopause, which causes a gradual decrease in oestrogen production, was not associated with any increase in the risk of coronary heart disease.
That quote from the New England Journal of Medicine in 1987. Ten years before, a study had appeared in the BMJ that looked at heart disease in relation to age, sex and the menopause:
… nevertheless, the idea of male sex hormones putting men at extra risk is more plausible than that of female sex hormones being protective, since large doses of oestrogen given to men for prostatic cancer, and the use of oral contraceptives containing oestrogen and progesterone have been shown to increase the risk of dying from coronary heart disease… Furthermore, the idea that female sex hormones protect against coronary heart disease should probably be abandoned.
In fact, there has never been a study – ever – showing that female sex hormones protect against heart disease in humans. Depressingly, however, without the slightest scrap of data to feed on, this hypothesis managed to gain power and credence. It’s the ‘Blob’ analogy again: ‘Mere facts have no effect on the sex-hormone hypothesis! Run for the hills before we are all engulfed!’ – or should that be ‘enblobbed’?
Indeed, so powerful did the sex-hormone ad-hoc hypothesis become that, by the 1990s, millions of women were actively being prescribed hormone replacement therapy (HRT) to reduce the risk of heart disease. Several GP colleagues mutter and go red when I mention this to them. Others have wiped their memory banks of ever having done such a thing. ‘Does not compute, does not compute. I was merely trying to prevent osteoporosis.’
So what changed? What changed was that someone finally decided to put the sex-hormone hypothesis to the test in a proper, grown-up, clinical trial: the heart and estrogen/progestin study – or HERS. This was randomised, placebo-controlled, and all those other things that can actually prove, or disprove, a causal relationship rather than relying solely on teleoanalysis.
I think you can guess the results by now – if you didn’t know them already. Basically, HRT increases the rate of heart disease. As of today, the American Heart Association, a bastion of conventional thinking, recommends strongly against using HRT to protect against heart disease.
How many women died from heart disease having been prescribed HRT? I only ask in the spirit of disinterested scientific discovery. I would never dream of suggesting that any advice given by the ‘establishment’ could possibly ever have been harmful. Opinion leaders are infallible, don’tcha know.
In fact, I think the idea that female sex hormones protect against heart disease represents, possibly, the most perfect example of the pure ad-hoc hypothesis in the history of medicine. It came into existence for one reason, and one reason only. To provide an explanation for the alleged female ‘protection’ against raised cholesterol levels. It was based on no evidence whatsoever. In fact, every time it was studied it was disproved, yet it still failed to die. It was only a very large, controlled clinical study that finally killed it.
Now that it is dead, what is left to account for female protection? My explanation is as follows:
A high cholesterol level does not cause heart disease. It is caused by other things. Which means that there is no need to ask why women are protected against high cholesterol levels, because, you see, there is nothing from which to be protected.
And wouldn’t accepting this possibility make life easier?
If there is no connection between cholesterol levels and heart disease, then there is no need to try and explain why, as cholesterol levels rose in Japan, the rate of heart disease fell. There is no need to explain why women have a lower rate of heart disease than men, despite having higher cholesterol levels. Because there is nothing to explain.
But if you accept this interpretation of the facts, you have just destroyed the cholesterol hypothesis. And that would never do. ‘Off with his head,’ bellowed the Queen of Hearts. And so another ad-hoc hypothesis was rapidly wheeled into place – one that had been prepared earlier. ‘You see, my dear boy, it never was female sex hormones that protected women as I, ahem, said all along. Instead, it is the fact that women, generally, have higher HDL levels, and this protects against heart disease.’ (Ad-hoc hypothesis no. 3 billion, and counting.) To quote a study in Geriatrics:
High-density lipoproteins and triglyceride levels are independent predictors of CVD in women. Cholesterol screening guidelines should be re-evaluated to reflect the importance of HDL and triglycerides in determining CVD risk in women.
To quote Dr Malcolm Kendrick, your author:
Oh for God’s sake, can you not just give up and admit that you are wrong?
Actually, the above quote from Geriatrics came before the HERS finally impaled the sex-hormone conjecture. Indeed, I believe the only reason why the HERS was widely accepted, rather than subjected to the usual rubbishing, is that mainstream researchers in heart disease were already jumping ship from sex hormones to HDL. Spurred on, no doubt, by the imminent arrival of HDL-raising agents – with all the lucrative ‘swimming-pool-enhancing’ clinical trials that would ensue.
A large part of me would rather not face the effort of chasing down this hypothesis and stomping it to death. But another part of me knows that the entire HDL ‘good’ cholesterol idea is in need of a serious kicking – as we used to say in Bonnie Scotland. After all, it has started to become the dominant hypothesis. ‘Look deeply into my eyes… forget total cholesterol, forget LDL… think only of HDL.’ Four legs good, two legs better.
What is a high density lipoprotein? It is a lipoprotein that is smaller, and denser, and contains more cholesterol, pound for pound, than any other type of lipoprotein. No one is certain where they come from, but the liver seems the most likely place.
What do they do? What are they for? They seem to be cholesterol ‘scavengers’. When a cell dies, cholesterol is released and floats about in the spaces between the cells. A passing HDL molecule can ‘incorporate’ this floating cholesterol, then transfer it to a VLDL or LDL, from whence it can be reabsorbed back into the liver and reprocessed. This is the so-called ‘reverse cholesterol transport system’.
It also seems that HDL can, through some system or other, remove excess cholesterol from within cells themselves. I’m not sure if I quite believe it, but incomprehensible tomes have been written on the subject. However, any argument in this area would require discussion of things like microtubular transcytosis – so I am not going there, however fascinating microtubular transcytosis may be.
Moving on – once HDL has hoovered up excess cholesterol, it is stimulated to transfer this cholesterol to an LDL or a VLDL by an enzyme known as lecithin cholesterol acyltransferase (LCAT). I only mention this rather arcane fact because several trials are underway on drugs designed to block LCAT, so that the HDL level will be increased. Although if you think this one through it makes absolutely no sense whatsoever. ‘Let’s block one of the central actions of the “reverse cholesterol transport” system, thought to protect against heart disease. That way we will have more HDL molecules in the bloodstream, but they won’t be able to do anything, such as transferring cholesterol back to the liver.’
Still, with the lucrative 22-year patents starting to run out on all statins, the pharmaceutical companies desperately need another angle beyond mere LDL lowering, and raising HDL levels looks like it might hit the jackpot.
To my mind, this is all a case of history repeating itself, only this time as farce. Today, everyone firmly believes HDL to be protective, just like they believed in sex hormones yesterday. And yes, there were complex theories in place to explain all the mechanisms by which sex hormones achieved their protective effect, just as there are with HDL. Yet, as with sex hormones, how many randomised, controlled, clinical trials on raising or lowering HDL have there been? Have a wild guess. You’re right. The answer is none. To put it another way – NONE! Or, to quote the Journal of the American College of Cardiology, January 2005:
Epidemiologic evidence has shown that HDL-C is inversely related to coronary heart disease (CHD) risk. However, the evidence for reducing CHD risk by raising HDL-C is thin, predominantly due to the paucity of effective and safe HDL-increasing drugs.
‘Thin’… that’s a good scientific word. How about ‘nonexistent’? That’s a better one.
In this particular area, though, it is interesting to look again at the HERS. Why? Because one of the central reasons why sex hormones were thought to be protective is because they raised the HDL level. And, as expected, the HDL levels did rise in the HERS trial. One slight problem, though: as HDL levels rose, so did the risk of heart disease.
Or, to put this another way, when ‘good’ cholesterol levels went up, so did the risk of heart disease. Perhaps we need to redefine ‘good’ as ‘bad’. George Orwell to the rescue again, I think. ‘Freedom is slavery, war is peace, I’m a little teapot…’
What the HERS researchers found was a relatively small 3 per cent increase in heart disease risk for every 5.4mg/dl rise in HDL. Converting this to UK units, a 0.14mmol/l rise in HDL increased CHD risk by 3 per cent. On that basis, a 1mmol/l rise in HDL would increase the risk of heart disease by 21 per cent.
This, I admit, is not a very scientific calculation, and you’re never really going to see a 1mmol/l rise in HDL anyway. But what the heck, mainstream researchers are allowed to use teleoanalysis, so I think I can get to use a little multiplication.
Despite these results, and however tempting it may be, I am not going to claim that HDL is damaging. Again, I think HDL is a marker of some kind. Almost certainly a marker of deranged carbohydrate metabolism, diabetes, insulin resistance and suchlike. Which means that a low HDL is, potentially, a worrying sign that something is going wrong with your metabolism. But it is only a sign, nothing else.
Of course, it is true that one key function of HDL is to transport cholesterol out of tissues and back to the liver via VLDL and LDL. But there is a huge difference between absorbing cholesterol that is floating about inside cells, or in the spaces between cells, and sucking cholesterol out of an atherosclerotic plaque.
Firstly, atherosclerotic plaques are almost universally covered over by a lining, or cap separating the plaque from the bloodstream, and this cap is impermeable to HDL. Secondly, a great deal of the cholesterol in plaque is in clefts, even crystals (how do you think Virchow recognised it 150 years ago?). It is not free and floating about inside a plaque, you would need a pneumatic drill to extract it, and I can’t see HDL wielding a pickaxe to a cholesterol cleft.
Thirdly, no one has explained, or identified, any sort of mechanism by which HDL gets cholesterol out of a plaque. It just sort of… does it. Speaking personally, I always like to see some sort of plausible biological mechanism to explain why something works. But on this one we have an almost total silence. Actually, the silence is not almost total, it is total.
Yes, I know that HDL is part of the reverse cholesterol transport system. Big deal. You can babble about this process all you like, and it does not explain how an inanimate molecule penetrates a fibrous cap then sucks cholesterol from a plaque before returning, back through the fibrous cap, ‘unharmed’ to the bloodstream. ‘The name is HDL… James HDL. Licensed to cure.’
To my mind, a good hypothesis should start with a theory as to how a thing may happen, based on a sound knowledge of physiology and biochemistry and the underlying science. But the HDL hypothesis only exists to plaster over yet another contradiction to the central cholesterol/LDL hypothesis. It’s not really a hypothesis, it’s really just an excuse, disguised as science.
However, my objections to the HDL hypothesis are not just theoretical. Now I shall introduce you to another black swan. ‘A black swan,’ did I hear you say? Oh yes indeedy. You see, there are two basic schools of thought in scientific research. There are the ‘weight of evidence’ scientists who seem to believe that if, for example, you find a high HDL level and a low rate of heart disease in ten studies, and a high HDL level yet a high rate of heart disease in two studies, you should place your faith in the ten studies. Such people are what I would call scientific ‘democrats’: whichever finding is supported by the greatest number of studies is the winner.
My view, and in this I am a follower of Karl Popper, is that such people are not truly scientists. The true scientific method is to propose a hypothesis in such a way that it can be refuted. You then set up experiments designed to refute the hypothesis. If you can’t, the hypothesis is likely to be correct. But if you can find a refutation, the hypothesis is wrong. And it doesn’t matter how many positive studies you have, they are all trumped by one contradictory study.
To use an example from Popper. A biologist offers the conjecture that all swans are white. If a black swan is discovered, his conjecture is wrong, and it doesn’t matter how many white swans there are relative to black swans. Ten to one, fifty to one, a million to one. Find one black swan and you have to accept that swans come in colours other than white. There are, of course, ways round this. To quote Popper again:
… when black swans are discovered in Australia he [the biologist] says that his conjecture is not refuted. He insists that black swans are a new kind of bird since it is part of the defining property of a swan that it is white.
Popper, K. Popper Selections
Anyone recognise this technique?
Anyway, black swan number one was the HERS. This showed that as HDL went up, so did the risk of heart disease. Now it is time to introduce a second and third black swan.
The second black swan was a study done in Poland and the USA ten years ago. At that time, the rate of heart disease was going up very rapidly in Poland and down in the USA. Researchers wanted to know if HDL levels might be the cause of this difference. The assumption behind the study was that the HDL levels would be high in the USA and low in Poland. Just for the record, they measured three different subtypes of HDL. (Yes, even HDL fragments into smaller and smaller fractions.) Not that it actually made any difference to the study, it just makes it considerably more difficult to understand the results.
The results:
In Polish subjects levels of HDL-C, HDL2, and HDL3, both unadjusted and adjusted for age and lifestyle factors, were higher than in US subjects. These differences contrast sharply with rising CHD rates in Poland and suggest either that other risk factors account for this trend or that the relationship between HDL-C and CHD risk may differ between the two countries.
In short, Polish men have high HDL levels and a high rate of heart disease.
In Russia, we find our third black swan:
High density lipoprotein (HDL) cholesterol was inversely related to mortality in US women, but there was no association of HDL cholesterol with mortality in Russian women. The absence of an association between HDL cholesterol and mortality in the Russian sample should be investigated further.
American Journal of Epidemiology, 15 February 1994; 139(4): 369–79
Are these studies, plus the HERS, enough to demolish the hypothesis that HDL protects against CHD? In my opinion the answer is yes. If your hypothesis is that HDL protects against CHD and you can find, without trying too hard, three pieces of directly contradictory data, then your hypothesis has just been shot dead.
Or has it? For if I have discovered one thing about this area, it is that HDL protection hypothesis truly cannot be killed. For example, a community in Italy was discovered with very low HDL levels and yet a very low of heart disease:
Thirty years ago, researchers showed that a family living in a northern Italian town, Limone sul Garda, lived to be very old and were extraordinarily resistant to heart attacks. Lots of people live to be one hundred years old and do not suffer heart attacks, but these people were extremely unusual because they had extremely low blood levels of the good HDL cholesterol that prevents heart attacks…
So, here we have a group with low HDL levels and low rates of heart disease, yet this fact had no impact on the ‘protective HDL’ hypothesis. In fact, it has actually managed to strengthen it. That noise you can hear is me beating my head against a wall.
HDL protects against heart disease – check. We find a population with a very low HDL level – check. They don’t die of heart disease – check. This is the strongest proof ever that HDL protects against heart disease – check.
I’m a little teapot – check…
Short and stout – check…
But hey! How silly of me to question this. You see, it has now been established that these people have a super-special form of HDL known as ApoA-1 Milano, no less. A stylish, two-door coupé form of HDL with that indefinable Italian flair and high performance. And it’s protective, even at low levels – in fact, especially at low levels. Once again, low has become today’s new high.
Researchers have now taken ApoA-1 Milano, cloned it in a laboratory, and started infusing it into people with heart disease, claiming results of such magnificent wonderfulness that the Emperor himself is going to clothe himself in them.
You know what? I really wouldn’t hold my breath waiting for this wonder cure. Remember that ApoA-1 Milano represents an ad-hoc hypothesis that only exists because of a previous ad-hoc hypothesis, which only exists because of an ad-hoc hypothesis prior to that. I am referring to the following:
• Raised LDL is supposed to cause CHD – but not in women.
• Women have higher HDL levels – so HDL is hypothesised to be protective against raised LDL.
• A population was found with low HDL levels and low rates of heart disease – so their HDL is hypothesised to be super-protective – even at low concentrations
Thus, a whole new branch of medicine opens up. And do you know how many people from the small Italian village of Limone sul Garda this protective HDL research is based on? Thirty-eight!
ApoA-1 Milano was patented by Esperion therapeutics, and the study on this form of HDL was done by Steven Nissen who is a research collaborator with Dr Eric Topol. Topol, in turn, runs a major cardiovascular website, called www.theheart.org entirely pharma-ceutical company sponsored.
And what did the www.theheart.org have to say about the ApoA-1 Milano study?
Who would believe that with five weeks of therapy we could actually remove significant quantities of plaque from the coronaries?
Please remember that this research was non-randmised, and conducted by professionals working closely with pharmaceutical companies.
* * * * *
Anyhow. At this point I shall attempt to summarise the evidence on HDL. What is the evidence to support the fact that HDL is protective?
• People with high HDL levels tend to have a lower rate of heart disease. And that’s it.
What is the evidence against?
• In the HERS – the only study done in which proper outcomes were measured, e.g. death from heart disease – when HDL levels went up so did the rate of heart disease.
• You can find populations with a high HDL level and a high rate of heart disease, and vice versa.
• It is clear that HDL levels reflect other things, e.g. alcohol consumption, which do have a direct effect on heart disease.
What’s the most important fact?
• No controlled, randomised study has even been done in which raising HDL levels has reduced the rate of heart disease.
In short, HDL should be relegated to the same status as female sex hormones. It’s an ad-hoc hypothesis, pure and simple. Which is kind of critical to the female heart-disease discussion. Because once you remove HDL from the equation, there is nothing left to explain the alleged protection that women have against heart disease.
This is just as well, really. For I am now going to present evidence that women are not actually protected against a high cholesterol level at all. Women, even young women, can suffer the same rate of heart disease as men, if not higher. Yet another little-known fact.
At this point I think it would be interesting to compare British women with French men. I know they live in different countries, but so what? Why should that matter? Unless, of course, you think that the French are ‘genetically protected against heart disease’. In which case you should beat yourself with a large club and dismiss yourself from the discussion.
Or perhaps you think that risk factors cannot be compared between different countries? If not, why not? If, as we are endlessly informed, cholesterol levels, blood pressure, smoking, saturated-fat consumption, age and sex are the most important risk factors for heart disease in the UK and the US, then why not in France? He who lives by the risk factor should also be prepared to die by the risk factor.
Now, if we do compare the countries we can see that firstly, French men eat more saturated fat. They also have marginally higher cholesterol levels, a greater percentage have hypertension, and French men also smoke considerably more than British women. (See table below, featuring the most recent figures from MONItor Trends in CArdiovascular Disease – or MONICA.) These are considered the critical risk factors by the major medical organisations, and they are used to calculate your risk of heart disease (along with age and sex, which are cancelled out in this comparison).
| RISK FACTOR | BRITISH WOMEN | FRENCH MEN |
| Saturated fat | 13.6% | 15.5% |
| % total calories | ||
| % with systolic BP | 7% | 11% |
| >160mmHg | ||
| Total cholesterol level | 5.6mmol/l | 5.7mmol/l |
| Percentage who smoke | 25 | 31 |
Yet, if we look at heart-disease rates (ages 35–74 per 100,000/year), every ten years since 1968, when statistics first started, they are as follows:
| Year | CHD rate British Women | CHD rate French Men |
| 1968 | 175 | 152 |
| 1978 | 180 | 154 |
| 1988 | 156 | 118 |
| 1998 | 97 | 85 |
Quelle horreur! French men have a lower rate of death from heart disease than British women, and always have done, despite having higher cholesterol levels and a greater burden of other ‘risk factors’. What does this mean? What indeed.
If you want a more spectacular example of the lack of female protection, we can look at Russian women and British men:
| Russian women | British men | |
| Rate of smoking | 0% | 27% |
| Average cholesterol levels | 5.4mmol/l | 6.0mmol/l |
| Average systolic BP | 132 | 134 |
| Saturated-fat consumption | 8.2% of calories | 13.6% of calories |
| Death rate from heart disease (200) | 267/100,0100/year | 229/100,00/year |
As you can see, British men smoke almost three times as much as Russian women, they have 10 per cent higher cholesterol levels, slightly higher blood pressure and eat 40 per cent more saturated fat. And yet their heart disease rate is 14 per cent lower.
At this point I think I should highlight the fact that French men have far more risk factors than Russian women, and 300 per cent less heart disease. What the HDL is going on?
You may feel that it is unscientific for me to make comparisons between different countries (although I would be interested to hear your reasoning). If so, I shall look within the same countries.
In Brazil, in 1989, women suffered a higher rate of heart disease than men. Admittedly, this was the only year when women had a higher rate. However, in general, the difference between men and women in Brazil is, and remains, tiny. Below are their respective risk factors:
| Risk Factor | Brazilian Men | Brazilian women |
| % with hypertension | 19 | 27 |
| Average cholesterol level | 4.92 mmol/l | 5.10 mmol/l |
| % who smoke | 24 | 18 |
| % who are obese | 48 | 39 |
OK, so their risk factors are pretty similar. But then again, risk factors for men and women are pretty similar in most countries. Yet, on average, women have one-third the rate of heart disease. In fact in the UK it is one-third, in France it is one-quarter. In New Zealand, at one point, it was one-tenth. In Brazil there is no difference.
But I am not going to stop here, because there are countries where women suffer more heart disease than men. For example, an Indian study in Delhi in 1993 showed that:
• The overall incidence of CHD was 19.7 per 1,000
• Men: 17.3 per 1,000
• Women: 21.0 per 1,000
In New Zealand in the 1970s, it was found that Maori women had more than twice the rate of heart disease of men, as revealed in a study in the New Zealand Medical Journal:
This paper reports the prevalence of coronary heart disease (CHD) and its relationship with several standard risk factors in samples of New Zealand Maoris… The prevalence rates of CHD are: 16.1 percent, and 7.3 percent in Maori females and males respectively.
Confused yet? If so, I would like to state that this is really not my fault. I think everything is quite simple, it’s the endless ad-hoc hypotheses developed to protect the cholesterol hypothesis that have created this current unholy mess of sex hormones, HDL, ApoA-1 Milano and their like.
In the end, to cut through the confusion, you have to take this argument down to basics. Women are either protected against a high cholesterol level or they are not. If women are protected against a high cholesterol level, how can you explain the fact that there are populations where women have lower cholesterol levels than men, yet suffer more heart disease? Where is the protection here? Where’s it gone?
On the other hand, if women are not protected against a high cholesterol level, why do they have much less heart disease than men – in most countries – when they have higher cholesterol levels? Run these arguments any way you like, and they keep breaking down as logic snaps under the strain.
In fact, there is only one conclusion that can be drawn from this unholy mess: cholesterol levels have no effect on heart disease rates in women. No other explanation fits the facts, but this explanation fits perfectly without the need for any ad-hoc hypothesis. Or, indeed, any other explanation at all. Remove cholesterol from the equation and all confusion disappears. Simple, isn’t it?
You might then ask, well, why do women generally get much less heart disease than men? That, of course, is the $64,000 question, and one that I shall answer in due course.
RAISED CHOLESTEROL LEVELS AND HEART DISEASE IN MEN
At this point, things are beginning to thin out somewhat. A raised cholesterol level doesn’t cause strokes, but a low cholesterol level may well do. A raised cholesterol level doesn’t increase overall mortality, but a low cholesterol level does. A raised cholesterol level does not cause heart disease in women.
What’s left? Does a raised cholesterol level cause heart disease in men?
Here are two facts with which I fully agree.
1: In men under the age of 50, a raised cholesterol level is associated with an increased risk of heart disease. (Note that I didn’t say ‘caused’, I said ‘associated’.)
2: Within countries/populations, a higher cholesterol level in men is associated with a higher rate of heart disease.
Does this mean that a high blood-cholesterol level causes heart disease in men? I dinnae think so, laddie. You see, there is far, far too much directly contradictory evidence out there.
Time, I think, to introduce you to Australian Aboriginal men for the first time. This group has one of the highest – possibly the highest – rates of heart disease in the world. They have a rate that currently stands at 1,100 per 100,000/year. This is about four times the rate in the UK, and more than ten times the rate in France. (It is a stunning 50 times the rate in French women.)
The average blood-cholesterol level in Aboriginal men is 4.9mmol/l, contrasting with 6.1mmol/l in the UK. Their average blood pressure is 125/77 – considerably lower than men in the UK. Their average HDL level is 1.1mmol/l, which is 0.2mmol/l lower than the UK. Their average body mass index (BMI) is 23.2, which makes them considerably less obese than British men.
The only conventional risk factor where they truly lead the way is smoking, which stands at just over 80 per cent. (Slightly higher than the rate in Japan where, incidentally, the rate of heart disease is 20 times lower. That’s right, 20 times.)
The main reason for bringing up the Australian Aboriginals is to compare and contrast their rate of heart disease, and average cholesterol levels, with countries from the MONICA study discussed previously. This study has been going on for ages now. It was set up by the WHO to look at heart disease rates and risk factors around the world.
I am a big fan of the MONICA study, by the way. It generates huge volumes of data that can be relied upon to be accurate and objective. So three cheers to the WHO. The interpretation of their data may often be exceedingly dodgy, but the data themselves are trustworthy. MONICA is where I found the data on saturated fat consumption across Europe.
For years, MONICA can remain silent then, every so often, it bestirs itself and out plops a golden egg. One of the latest golden eggs came from its review of cholesterol levels across Europe. I related these data to its published death rates from heart disease. And to this list I have added in the Australian Aboriginals and drawn a graph (Fig.26):
Fig. 26 Comparison between heart-disease rates in men aged 35–74 and average cholesterol levels in 15 populations
As you can see, ahem, a very clear pattern emerges. As average cholesterol levels rise heart disease rates fall, then go up, then fall, then go up, then fall, then fall a bit more, then go up, then fall. I suppose the general trend is that as cholesterol levels rise, heart-disease rates fall. But I would not even attempt to make such a claim from that graph.
The point I want to make is that there is a complete and utter dissociation between cholesterol levels and heart disease. And no, I didn’t choose these countries to make my point. I just took all the countries that appeared on page 77 of the European cardiovascular disease statistics from the WHO MONICA Project:
Mean total blood cholesterol and percentage with levels of 6.5mmol/l and above, adults aged 35–64, by sex, latest available data, MONICA Project populations. [NB: I only used data on men in this graph.]
This was published in the International Journal of Epidemiology. Go look it up if you don’t believe me. The only countries I left out are Yugoslavia and East Germany, which were in the most recent MONICA statistics available on cholesterol levels – somewhat surprising, as these countries didn’t exist at the time, and neither did any statistics on deaths from heart disease.
I suppose that I didn’t really need to add in the figures from Australian Aboriginals to make my point. But they represent, as far as I am aware, the most outrageous cholesterol ‘paradox’. Lowest average cholesterol levels, highest rate of heart disease. Compare this to the Swiss – highest average cholesterol levels, second lowest rate of heart disease. Or the Russians, second lowest cholesterol level, highest rate of heart disease in Europe. Take your pick. Every single country is a ‘paradox’.
And so, an entire flock of black swans wheels overhead, nearly blocking out the sun. Below us, the cardiovascular opinion leaders eagerly tend to their one white swan, subjecting it to intense scientific scrutiny. ‘I think we can confirm that all swans are white.’ They intone. ‘Could someone please shoot those strange black birds above us; they are distracting us from our work.’
Somewhat closer to home are ‘Emigrant Asian Indians’. The term is not mine, and it is rather confusing. Emigrant Asian Indians, in the context of heart-disease research, means anyone who emigrates from Pakistan, India, Sri Lanka and/or Bangladesh. It has been known for years that Emigrant Asians, as I shall call them, suffer catastrophically high rates of heart disease. This is true wherever they emigrate to.
And what of their risk factors for heart disease? A study in Bradford, looking at ‘Ethnic differences in risk markers for heart disease in Asian immigrants’, found that, in comparison to the surrounding non-Asian population:
• They had lower total cholesterol levels
• They had lower LDL levels
• They had lower blood pressure
• They smoked less
• They were slightly less obese
So there you have it. The reason why Asian Indians in the UK have higher rates of heart disease is because they have lower LDL and total cholesterol levels, smoke less and have lower blood pressure. A study of Emigrant Asian Indians in the USA came to the following conclusions:
Asian Indians have the highest rates of coronary artery disease (CAD) of any ethnic group studied, despite the fact that nearly half of this group are life-long vegetarians [my emphasis]. CAD occurs early in age and generally follows a malignant course, although the incidence of classic risk factors is low.
Enas A Clin Cardiol, March 1995
In more detail, the figures from Asian Indians in the USA were as follows:
| Risk Factor | Asian Indians | Caucasians |
| Rate of obesity | 4.2% | 22.6% |
| Rate of Hypertension | 14.2% | 19.1% |
| Percentage with high LDL | 13.7% | 22.3% |
| Smoking rate | 1.3% | 27.1% |
Different country, same findings. But of course (sound of me slapping my forehead in the background) Asian Indians are genetically susceptible to developing diabetes, and diabetes causes a huge increase in the rate of heart disease. How silly of me to forget. So what you’re saying is that Australian Aboriginals and Asian Indians must have descended from the same gene pool. That famous prehistoric group of ‘diabetics who drop dead of heart disease while having low cholesterol levels’. Caused by a gene found on chromosome twelve.
The descendants of this genetic line, then, must also have fought their way across the Pacific to North America, because another major population that has low cholesterol levels and a catastrophically high rate of heart disease are Native Americans. Look up the ‘Strong Heart Study’ on the internet. They sure get about, these ‘low-cholesterol high-heart-disease offspring’, considering how young they die.
If we are tracing genetic trees, I suppose we need to include Russia, because a major study was done in Russia in response to the rapid rise of heart disease-related deaths during the latter half of the 20th century. It was called ‘Increased risk of coronary heart disease death in men with low total and low density lipoprotein cholesterol in the Russian Lipid Research Clinics Prevalence Follow-up Study’. If you want to look it up, the main author of the report was Shestov, of the Institute of Experimental Medicine, Russian Academy of Medical Sciences, St Petersburg. As you can see from the title, Shestov discovered that in Russia, in a significant number of men, a low LDL level was the most important risk factor for dying of heart disease:
The results disclose a sizeable subset of hypo cholesterolaemics [my emphasis] in the population at increased risk of cardiac death.
So, Australian Aboriginals, Native Americans, Emigrant Asian Indians and a large percentage of Russian men have low cholesterol levels and high rates of heart disease. Hmmm. And the Swiss and the French have very high cholesterol levels and very low rates of heart disease. Who’s next? What’s next?
What’s next, I think, is to make the following point. Heart disease is primarily a disease of older people. The rate of heart disease in 65-year-old men is approximately 10 times that of 45-year-old men. Yet, while a raised cholesterol level is associated with heart disease in younger men, the association disappears as men get older. Here is a summary of the findings of a study published in the Journal of the American Medical Association in 1995:
Our findings do not support the hypothesis that hyper-cholesterolemia or low HDL-C are important risk factors for all-cause mortality, coronary heart disease mortality, or hospitalization for myocardial infarction or unstable angina in this cohort of persons older than 70 years.
This is supported by another study published in the Journal of the American Geriatric Society in 1991:
Elevated total cholesterol was not found to be associated with CHD mortality in older men.
And just for luck, here’s another one from the National Centre for Health Statistics:
Although coronary heart disease remains a leading cause of death and disability in old age, the relationship of serum cholesterol level to risk of coronary heart disease in old age is controversial. Data for 2,388 white persons aged 65–74… were examined to determine the relationship of serum cholesterol level to coronary heart disease incidence… there was no overall relationship between serum cholesterol level and coronary heart disease risk in either men or women…
How can a risk factor for a disease stop being a risk factor at the age when the disease kills the greatest number of people? What on earth is the reasoning here? That raised cholesterol levels only do damage when people are younger? Actually, I have yet to hear an explanation for this. It’s another fact that is just quietly swept under the carpet, as if it were of minimal importance.
But it is critical. This would be like finding that smokers suffer an increased risk of lung cancer at 45, but by the time they are 65, smoking ceases to be a risk factor – so you might as well keep on puffing away. How much sense would that make? That’s right, it would make no sense at all…
Hold on. Unless… unless a high cholesterol level didn’t actually cause heart disease, but just acted as a heart-disease ‘marker’ in younger men. Yes, that would make sense. Hmmm.
* * * * *
At this point I should remind you of the data from the Framingham Study demonstrating that if your cholesterol level falls, your rate of cardiovascular disease increases:
… there is a direct association between falling cholesterol levels over the first 14 years [of the study] and mortality over the following 18 years (11% overall and 14% CVD death rate increase per 1mg/dl per year drop in cholesterol levels).
So, a falling cholesterol level causes strokes and heart disease? How does that work, then? Maybe a falling cholesterol level drags the arteries down with it? Perhaps the discussion should just stop here as I draw together the salient facts:
• A raised cholesterol level is associated with heart disease in younger men – within a country.
• There is no association at all between average cholesterol levels and the heart disease rate between countries.
• Over the age of about 50, the association between cholesterol levels and heart disease disappears.
• A falling cholesterol level is associated with a greater risk of heart disease.
If you are not convinced by now that raised cholesterol levels do not cause heart disease, nothing could possibly convince you. But before moving on to look at familial hypercholesterolaemia and such matters, I would like to introduce you to a major study that I find unintentionally hilarious.
The European Society of Cardiology runs a very major pan-European study called EUROASPIRE, which looks at risk factors for heart disease across Europe, and also analyses the management of those risk factors. A couple of years ago they found that:
… smoking, previous coronary heart disease and diabetes proved significant predictors of total, cardiovascular (CVD) and coronary heart disease (CHD) mortality. Obesity, low education, raised blood pressure, elevated total cholesterol and low HDL cholesterol [my emphasis], however, were not significantly associated with higher mortality rates.
But they had an explanation for the fact that raised blood pressure, raised cholesterol and low HDL cholesterol were not associated with heart-disease mortality:
Failure to find statistically significant associations between other classical risk factors, such as blood pressure and plasma lipid levels, and mortality may be related to the extensive use of antihypertensive and lipid-lowering drugs in this cohort.
Perhaps they should have read the conclusion of a sister EUROASPIRE paper on the management of said risk factors:
This European survey of coronary patients shows a high prevalence of unhealthy lifestyles, modifiable risk factors and inadequate use of drug therapies to achieve blood pressure and lipid goals [my emphasis].
The actual figures were that 58 per cent had total cholesterol concentrations over 5.5 mmol/l and more than 50 per cent had blood pressure above 150/90. In short, there was not ‘extensive use of antihypertensive and lipid-lowering drugs in this cohort’. Or if there was, they weren’t doing much to lower blood pressure or cholesterol levels. Another instant ad-hoc hypothesis bites the dust.
What EUROASPIRE actually proved, rather nicely, is that smoking, already having heart disease and having diabetes are true risk factors for heart disease. Whereas a high cholesterol level, and a high blood pressure, are not. Did the authors of this study consider this possibility? They did not. In this area, the ability of researchers to ignore the results of their own research is quite mind-bending.
Men occasionally stumble over the truth, but most of them pick themselves up and hurry off as if nothing had happened.
Winston Churchill
As a quick aside, discovering that ‘already having heart disease’ is a risk factor for dying of heart disease hardly ranks alongside the discovery of penicillin.
FAMILIAL HYPERCHOLESTEROLAEMIA
Enough of men and heart disease, and endless contradictions to the cholesterol hypothesis. Time now to look at a group of people who have extremely high levels of LDL in their bloodstream and who can die, aged five, of heart disease. Yes, I am talking about people with familial hypercholesterolaemia (FH).
Most doctors that I know find the evidence on FH particularly powerful and utterly convincing. It’s a kind of ‘A-ha!’ type of fact: ‘A raised cholesterol level must cause heart disease, because people with FH die very young of heart disease, and the only problem they have is a high LDL/cholesterol level. A-ha!’
In fact, I sometimes think of FH as the ‘Lourdes’ of the cholesterol believer. ‘Are you feeling dispirited? Have you too begun to feel that the cholesterol hypothesis may be falling apart? Visit Familial Hypercholesterolaemialand.
‘Here you can see children with severe FH dying as young as five from heart disease. Here men and women with FH can have a 9,686 per cent increase in the risk of dying of heart disease. In Hypercholesterolaemialand, men barely make it past 40 before keeling over and dying.’ Of course! How could I ever have doubted the cholesterol hypothesis? It’s true, it’s true!
Here, for instance, are some figures taken from the Simon Broome Register Group, set up in the UK to study and help manage people with FH:
The cohort was followed up for 2,234 person years during 1980–9… The excess mortality from this cause (CHD) was highest at age 20–39 (standardised mortality ratio 9,686).
A 9,686 per cent increased risk of dying of heart disease! Take that! Now Dr Kendrick you must give in, abandon, buckle under, capitulate, cave in, cede, commit, concede, consign, cry uncle, deliver up, eat crow, eat dirt, entrust, fall, fold, forego, give in, go down, go under, hand over, knuckle, knuckle under, leave, let go, lump it, pack in, part with, play dead, quit, relinquish, renounce, resign, roll over, say uncle, submit, succumb, waive and yield.
I never had any intention of doing so. Because, as with everything in the strange, distorted world of the cholesterol hypothesis, facts are not exactly what they seem – even when they are true.
For instance, here is a later paper from the same Simon Broome study:
A study showing that although mortality from heart disease is increased in FH, cancer mortality is reduced by nearly half, so that the overall mortality is no higher than in the general population of England and Wales.
So FH is a deadly killer disease, but it doesn’t actually kill you. Perhaps a nigh-on 10,000 per cent increase in the risk of dying of heart disease is not exactly what you thought it was. Don’t worry, it never is.
Time, I think, to reveal a little game that researchers play with statistics in the world of heart disease. It’s called ‘Data Inflation – the Revenge’. By the way, as you will see later, the statin researchers have taken this game to infinity – and beyond.
To give you an example of how to warp statistics into weapons of mass disinformation, I shall pose a simple question. What are the chances of winning the jackpot in the lottery? About 1 in 15 million per week. If I were able to improve your odds from 1 in 15 million to a massive 1.5 in 15 million, I could claim to have increased your chances of winning by 50 per cent. Fifty per cent, however, represents the relative increase in your chances. The absolute increase is 0.5 in 15,000,000. Or 1 in 30 million. Or, in percentage terms, 0.000003 per cent.
If I were trying to sell my lottery-enhancing service to you, which figure do you think I would use? I see an advert coming on: ‘Do you want to be a MILLIONAIRE? If so, the world-renowned lottery expert and Nobel prize-winning mathematician Dr Kendrick can increase your chance of winning the lottery by 50 per cent – yes, FIFTY PER CENT! You CAN be a winner! Just send £10 in a stamped addressed envelope to the following address to receive your advice! This is NO CON TRICK, this advice is guaranteed to work.’
On the other hand, I should most likely not advertise my services thus: ‘The desperate, lonely, and rather hard-up Dr Kendrick can very slightly increase your miserably small chance of winning the lottery by a meagre, and almost unnoticeable 0.000003 per cent. Yes, pitiful isn’t it, you pathetic, snivelling toe-rag. But if you feel like wasting £10, send a stamped addressed envelope…’
You have to admit that 50 per cent sounds rather more impressive than 0.000003 per cent. However, both figures are true. Ah yes, the truth. A slippery little worm is it not. Oh, by the way, the best way to increase your chances of winning the lottery by 50 per cent is to buy three tickets between two people. Worth £10 of anyone’s money, that advice.
To return to the Simon Broome study. In the entire population there were actually only 24 deaths, 15 of which were from heart disease. So, how many people died of heart disease in the 20–39 age group? If memory serves, it was six. From this they calculated a 9,686 per cent increase in risk. Hot dang! Aren’t statistics fun? And no, I am not going to run the calculation for you. I am talking broad brush strokes here.
So while people claim massive increases in heart-disease risk with FH, the figures may not be quite what they seem – to say the least. Another reason for the overestimation of heart-disease rates in FH is that the people with FH who have been studied are the ones that already have heart disease.
The risk of early death and premature cardiovascular disease in familial hypercholesterolaemia may have been overestimated because previous studies have looked only at patients and families who sought medical attention. Sijbrands et al traced all members of a Dutch pedigree dating from 1800 to 1989 and calculated their risk of death. They found that many untreated patients had normal life spans.
What happened to the people with FH who did not die early of heart disease? Well, because they didn’t have heart disease, no one actually knew they had FH. So the vast majority just carried on with their lives – undetected. How many such people are there? Well, difficult to tell, as they are undetected… Doh!
Luckily – for me, anyway – they are not actually always undetected. A study was done in South Africa, looking at genetic inheritance of FH. Researchers found that different members of the same family could have the gene, yet experience a very different heart-disease outcome. For example:
One individual did not develop coronary heart disease (CHD) by age 84, despite having the FH Afrikaner-1 mutation, while his son who inherited the same gene, developed CHD before age 50 and had to undergo bypass surgery.
So, did the FH cause premature CHD in the son? If so, why did it not cause CHD in the 84-year-old father? Same gene, same cholesterol level, completely different outcome.
At the risk of flogging a dead horse here, I would like to re-emphasise one point. If the only member of the family who had come across the medical profession had been the son, his condition would have been put down as ‘yet more evidence’ that FH kills people early from heart disease. The father would have slipped past unnoticed. With FH, it’s a one-way evidence valve, allowing in only evidence to support your hypothesis.
Given this, do I think that FH can cause heart disease? In fact, I do. Although I believe that it is not nearly as deadly as most people believe. After all, people can live into their eighties, even nineties, and even up to 103 – in the Simon Broome study. Maybe they would have made it to 104 if they hadn’t been suffering from that deadly killer FH.
However, before moving on to talk about how FH may cause heart disease, I would like to knock one thing on the head. Namely, the evidence of very early death from heart disease in homozygous FH. That is, in people who get the gene from both parents. This fact is true, but completely irrelevant to any sensible discussion.
Some years ago I was working on a medical ward when a long-term psychiatric patient was admitted. He was unconscious and very nearly died. His sodium level was 100 – which will mean nothing much to you, but we thought it was a world-record low level. Guinness World Records time indeed. Disappointingly, I then discovered lower levels. So my 15 minutes of fame have gone. What had happened to this patient? He had become very upset at the death of a friend and had started drinking water. Gallons of water. This diluted his blood to the point at which his brain nearly packed in.
I also noticed recently that drinking too much water, and then dying, is emerging as a risk factor in marathon runners who ‘overhydrate’, i.e. drink too much water. This from a paper in the New England Journal of Medicine:
Hyponatremia has emerged as an important cause of race-related death and life-threatening illness among marathon runners.
Well, there you go. If you look hard enough, you can find that water is a deadly dangerous substance. And so what? Can I then argue that, because a vast excess of water is deadly, that water is deadly at any level of intake? I could try, but I don’t think I would get very far.
You may remember that I mentioned Smith-Lemli-Opitz Syndrome (SLOS) earlier. This is a condition characterised by very, very low cholesterol levels, and catastrophic effects on health – stillbirth, death from multi-organ failure, visual loss, congenital heart disease and the like. This syndrome is much more deadly than FH.
Can I use the extreme example of SLOS to argue that moderately low cholesterol levels are also deadly? No, I cannot, and will not. If I did, this would be exactly the same as people claiming that premature heart disease – brought about by LDL levels in excess of 15mmol/l – somehow also provides proof that moderately raised cholesterol levels also causes heart disease.
Once you reach such extremes you are talking about a completely different pathological state. There is no substance I can think of that would not kill you if it reached five times the ‘normal’ level in the blood. Most things, such as sodium or potassium, will wipe you out if they vary by more than 20 per cent.
In fact, on this basis we would have to view LDL as a very benign substance indeed. You can have five times as much as normal in the bloodstream, and you can still live into your 20s, or 30s – even your 50s.
But what of more moderate FH – heterozygous FH, where LDL levels are about double normal? This still increases the rate of heart disease, although by how much I am not certain, and I don’t think anyone else is either. However, even if it does, I believe that LDL may have no direct role to play.
Time now to draw the curtain aside and reveal a whole different world of research into heart disease. A world where it can be explained how FH raises the risk of dying of heart disease – but not through a raised LDL level. In another life I may have avoided talking about this world until rather later. However, I can’t really say that FH raises the risk of heart disease through another mechanism, without some explanation. Otherwise, I suspect you might not believe me. Of course, you might not believe me anyway. But I rather hope that you may.
Now, let us venture into a world where researchers think that heart disease is, basically, a response to injury. The ‘response to injury hypothesis’ goes something like this:
• A factor (or more likely, several factors acting in unison), damages the endothelium (single-celled lining of the artery wall).
• This stimulates a small blood clot to form over the area of damage.
• Endothelium re-grows over the blood clot, ‘repairing’ the area of damage.
• With endothelium on top of it, the blood clot is effectively drawn into the artery wall, and is then broken down by white blood cells – thus disappearing.
That, anyway, is the healthy response to endothelial injury – if any degree of endothelial injury can actually be seen as healthy, that is. However, if the endothelium keeps getting damaged (for whatever reason), clots keep forming, and more and more blood clots keep getting drawn into the artery wall. At which point the healing responses cannot cope. The area of damage becomes permanent. A plaque forms that can grow and grow. In many cases a fibrous ‘cap’ forms over the plaque. This can rupture, triggering such a big blood clot that it fully blocks the artery. Heart-attack time.
That, anyway, is roughly what Austrian pathologist Carl Freiherr von Rokitansky said. And he said it more than 150 years ago. Yes, you may be surprised to learn that this is in no way a new theory. It has been around for at least as long as the cholesterol hypothesis. Perhaps even a few years longer. In fact, Rokitansky was a compatriot of Rudolf Virchow – also a man who, in my opinion, got it right. Indeed, I believe that the evidence for the ‘Rokitansky’ model of heart disease is overwhelming. Heart disease – or to be more accurate, ‘atherosclerotic plaque formation’ – is, basically, a response to injury. And it consists of two basic processes. Firstly, endothelial damage, then blood-clot formation.
As you may recall from earlier on in this book, this is exactly the research direction that Pfizer was heading off in nearly 15 years ago, before multi-billion-dollar statin profits rather distracted their attention. It is also the direction that Kilmer McCully was heading in – when he left his appointment at Harvard Medical School and Massachusetts General Hospital.
It is also the direction that Linus Pauling took for the last 20 years of his life or so. Linus Pauling is the only person, ever, to win two individual Nobel prizes. In my book, this makes him quite clever. In his later years, though, his standing within the scientific community waned as he focused increasingly on vitamin C. He claimed that it could cure just about anything, and everything. I don’t believe this, but when it comes to vitamin C and heart disease he was on a very interesting track indeed. Pauling had teamed up with a certain Dr Matthias Rath on this subject. I believe Dr Rath’s ideas on heart disease are brilliant. Despite a reputation for outspoken and unpopular views on anti-retroviral drugs, (which have been met with opposition from the AIDS lobby, the South African government and parts of the pharmaceutical industry), I would like to cite Dr Rath and his studies. One of Dr Rath’s theories that so enamoured Linus Pauling. It was as follows. Humans are one of very few animal species that cannot make their own vitamin C. Included in this exclusive group are guinea pigs, fruit bats and most other large apes. Which means that the only way we humans can get sufficient vitamin C is to eat it, and we need to eat quite a lot – although perhaps not nearly as much as the two grams a day recommended by Pauling – although he did live to very nearly 100. (Maybe he had the ApoA-1 Paulingo mutation.)
If we don’t eat enough vitamin C, one of the first things to happen is that our blood vessels spring leaks as the supporting collagen structure breaks down around them. Then we start to bleed, with bleeding gums being one of the early signs of scurvy. Left untreated, scurvy leads to death from internal bleeding.
It is speculated that during the ice age, our ancestors found it rather tricky to get hold of enough vitamin C, so a lot of them suffered from scurvy, and many died. However, some of them had a trick up their sleeves. This was to manufacture a lipoprotein called lipoprotein (a) – or Lp(a) – which you may remember me mentioning earlier.
Lp(a) is a form of LDL with an extra protein stuck on the side called apolipoprotein(a). This extra protein has several interesting properties. Firstly, it is attracted to areas of endothelial damage. Secondly, it sticks firmly to those areas, and helps to form a very strong and difficult-to-remove blood clot.
Thirdly, and perhaps most important of all, apolipoprotein(a) is almost identical in structure to an enzyme known as plasminogen. This enzyme is well known to the medical profession. It is incorporated into blood clots as they form. When it is activated, plasminogen turns into plasmin, and then acts like a little tiny stick of dynamite within the clot, cleaving it apart. Without plasminogen all blood clots would most likely remain in place for ever – which is not a great idea if the clot is completely blocking an artery.
Plasminogen, in turn, is activated by an enzyme known as plasminogen activator, or tissue plasminogen activator (tPA). If you are having a heart attack, you could well be given an injection of tPA to stimulate the plasminogen to blow apart the blood clot blocking your coronary arteries; tPA is thus also know as a ‘clot-buster’.
Still with me? Good. Because you see, the important point here is that when Lp(a) becomes incorporated in a blood clot, it blocks the action of tPA, and makes the clot much more difficult to shift. This, of course, was probably a good thing if you were an ice-age cavemen with arteries leaking due to vitamin C deficiency. You probably needed strong ‘plugs’ to prevent excess blood loss, and you didn’t want plasminogen blowing clots apart. But today, when no one really has vitamin C deficiency, a high level of Lp(a) may be an important factor in creating big, difficult-to-shift blood clots in your arteries. So, a high level of Lp(a) could well be a risk factor for causing the development of atherosclerotic plaques – could it not?
Of course it could. Just to give one example, a study in Spain demonstrated that people with two genes for Lp(a) expression, and high Lp(a) levels, had a 650 per cent greater risk of developing heart disease. And, increasingly Lp(a) is being isolated as an important risk factor in Emigrant Asian Indians – the ones with low LDL levels and high rates of heart disease.
I’ll add one example from the Journal of Lipid Research. Rather a lengthy quote, I’m afraid, but I think it is fascinating nonetheless.
A high incidence of coronary heart disease (CHD) has been observed among Asian Indians immigrating to the USA and among native people remaining within the Indian subcontinent. The mortality rate for CHD in Asian Indians from Singapore is 4 times higher than in Chinese residents from Singapore, and 20 times higher than in blacks from South Africa Moreover, CHD in Asian Indians occurs prematurely and is often more severe than in Europeans…
Traditional risk factors for CHD such as obesity, insulin-dependent diabetes mellitus, smoking, hypertension, and elevated plasma Chol or low density lipoprotein cholesterol (LDL-Chol) levels do not [my emphasis] explain the observed increase in CHD incidence among Asian Indians… [however, as the article goes on to say]… This study suggests that elevated plasma Lp(a) confers genetic predisposition to CHD in Asian Indians.
By the way, forget the genetic bit, and concentrate only on the Lp(a) bit.
Finally getting back eventually to FH, is it possible that people with FH not only have high LDL levels, but that they also have high Lp(a) levels, given that they are basically the same molecule? Why, funny that you should ask. For a study was done in Austria and South Africa and the researchers found that…
This leaves little doubt that LDL-R mutations that result in FH with elevated LDL also result in hyperlipoprotein (a).
This may all seem, I suppose, a bit theoretical. Surely people have looked at atherosclerotic plaques and found LDL within them, not Lp(a) – end of counter-hypothesis. Well, superficially at least, it seems to be true that LDL has been found in plaques, along with their key identification protein apolipoprotein B-100. But, of course you must remember that Lp(a) is just a form of LDL, and it too has apolipoprotein B-100 attached to it. So how would you know if you were really finding LDL within a plaque, and not Lp(a)? There is only one way to tell, really. See if you can find apolipoprotein(a) within plaques. If you can, it can only have come from Lp(a), not LDL.
Dr Rath decided to look for apolipoprotein(a) in areas of plaque formation, along with a team of other researchers, and they published their findings in the European Heart Journal. And guess what they found? That’s right, a high concentration of apolipoprotein(a) within atherosclerotic plaques.
And Dr Rath is not the only one to find apolipoprotein(a) in plaques. A paper published in Nature in 1989 found exactly the same thing. In their words, which I will simplify afterwards:
We report here that apolipoprotein(a) interferes with endothelial cell fibrinolysis by inhibiting plasminogen binding and hence plasmin generation. In addition, we demonstrate lipoprotein(a) accumulation in atherosclerotic lesions. These findings may provide a link between impaired cell surface fibrinolysis and progressive atherosclerosis.
In short, this team also found apolipoprotein(a) in atherosclerotic plaques. They also confirmed that apolipoprotein(a) inhibits both the binding, and activity, of plasminogen within a blood clot.
This is fascinating research, but it is also research that is frowned upon by the powers that be. ‘As we already know what causes heart disease, what exactly – young man – is the point of the research you are proposing? Grant application denied. And by the way, as a matter of interest, do you still enjoy working here?’
I hope it has become clear by now that Dr Matthias Rath is, in fact, a very clever chap. He worked out an entirely new hypothesis about heart disease. One that fits many of the known facts, and makes a lot more sense, frankly, than the LDL hypothesis. It’s not difficult to see why Linus Pauling thought that Rath was on the right track. Personally, I think Rath has only a piece of the jigsaw puzzle in his hands. But it is a critically important piece.
My aim at this point, however, is not to delve too deeply into the Matthias Rath theory of heart disease. What I wanted to do was to give you enough information on an alternative view to accept that FH could cause heart disease, but not by raising LDL levels. And this is not just wild theorising – there is a lot of solid research behind it.
Indeed, once you start looking into this area, you find that FH is associated with a whole raft of other abnormalities, mostly to do with abnormal blood clotting. Just a couple of quick quotes on the matter from Atherosclerosis and the British Haematology Journal respectively:
The results suggest the hypercoagulability may play a role in the pathogenesis of coronary heart disease in patients with familial hypercholesterolaemia.
Plasma fibrinogen [a clotting factor] was elevated in FH…
In short, it is fully possible that FH causes an increase in the risk of heart disease, not by raising LDL levels, but through its impact on blood-clot formation.
However, I still think that the strongest argument against FH causing heart disease is that most people who die of heart disease do not have raised LDL levels. And most people with raised LDL levels do not die of heart disease, even people with FH. I think it is reasonable to ask, how can a raised LDL level be the cause of heart disease… when it is not present? And how can it be present and not cause heart disease? You’re right, it can’t.
HOW, EXACTLY, IS LDL SUPPOSED TO CAUSE HEART DISEASE?
Finally, in this chapter, I think it is reasonable to ask the question, ‘If the establishment is so sure that LDL causes heart disease, how does it do it? What is the mechanism – where’s the biological plausibility?’ It’s no good saying a thing causes heart disease then failing to provide a half-decent mechanism of action.
I think you should always bear in mind that the cholesterol hypothesis started life as the diet-heart hypothesis, with cholesterol in the diet as the major culprit substance. However, even Ancel Keys gave up on cholesterol in the diet pretty quickly. And, although saturated fat clings on in most people’s minds, I hope you are convinced by now that it has no role. Which means that the first half of the hypothesis is dead.
But never mind, even with its legs cut off, researchers were still left with the ‘fact’ that raised cholesterol levels caused heart disease. Although what, exactly, is supposed to raise the cholesterol levels now is not clear. Even if were clear, it still can’t be cholesterol in the blood that causes heart disease, because you don’t have any cholesterol floating free in the blood.
Indeed, it was only quite late on that cholesterol was gently swept under the carpet, and LDL was introduced as the killer lipoprotein. At which point, presumably, people must have asked themselves something along the lines of: ‘Finally we know what causes heart disease – it’s LDL. But how does it do it?’
No one seems to have questioned how it can be that a hypothesis can go through several major changes, yet still somehow manage to remain the correct answer. All everyone wanted to do, it seems, was to hammer cholesterol into the jigsaw puzzle in some way – even if it had transformed itself into LDL along the way.
Personally, I would have more faith if the hypothesis had started life the right way round. The present thinking reminds me of the story of Keppler, who was determined to understand the orbits of planets around the sun. In his time it was decreed by the great thinkers that the planetary orbits had to fit within the ‘perfect shapes’ of the Greeks. For example, a perfect circle. Why did everyone think this? Because it had been decreed so by the Greeks, and they were all-knowing, and right about everything. You could not even dream of questioning their ancient Mumu wisdom.
Somewhat hampered by the requirement to get the facts to agree with the already-known answer, Keppler battled for 20 years to fit the observations of the 16th-century astronomer Tyco Brahe within the Greek ideal of perfect shapes. Unsurprisingly, he failed. Finally, Keppler gave up on the Greek idea of perfection, and thus were born Keppler’s laws of planetary motion. Simultaneously, at least one section of the ancient wisdom of the Greeks was exposed for what it was – ridiculous dogmatic twaddle. You mean wise men in flowing gowns with beards aren’t always right? Surely not.
Despite my philosophical objections, you may still think that everything had been sorted out by now, and that scientists with electron microscopes had watched LDL sticking to artery walls, or battling bravely through artery walls and then building up into a big plaque. Or something of the sort. The reality, however, is that researchers are still trying to work out how LDL creates atherosclerosis. It is true that huge tomes have been written on this subject outlining countless enzymes, and co-factors, and Lox-1 receptors and intracellular transportation systems. There is, literally, no end to it all. However, if it really is LDL that causes heart disease, I would like you to consider the following question. Why do we now hear so much about antioxidants? What’s all that about? Why are they supposed to be so good for you? What have they got to do with LDL?
Well, you see, things have moved on a little. Now it is not actually LDL that’s damaging, it is ‘oxidised’ LDL. Oxidised LDL is LDL that has reacted with oxygen, and can be thought of as slightly ‘damaged’. A bit like leaving out meat uncovered overnight. Oxygen gets at it, reacts with it, and turns it into nasty wrinkly stuff. Of course, raised LDL is still important, but you must focus on oxidised LDL at the same time. Yet when you do you will find that it has miraculously changed shape back to LDL again. This area is a bit like quantum physics. The moment you focus on LDL it flips into oxidised LDL, and then when your attention slips, it flips back again. It is a veritable will o’ the wisp that will never fully take form in front of you.
Indeed, if you ever start reading about this area you may find yourself filled with a desire to scream. ‘Well, is it LDL, or oxidised LDL, will you make up your damned mind?’ No chance: in general, those who write papers in this area give the impression that it is – sort of – both. By taking two positions simultaneously, this negates the tiresome requirement of formulating a hypothesis that can actually be disproved. You can keep on leaping backwards and forwards from one hypothesis to the other. I will simply make the general point that the ‘oxidised’ LDL hypothesis was primarily developed to explain how people with low LDL levels can get heart disease – while still claiming that LDL has a key role in the process. It’s the old ‘low is high’ concept again.
Despite the fact that you cannot really get a handle on the ‘oxidised LDL hypothesis’, I think it does need some further explanation. However, I am not going to delve into this area in any great depth as it rapidly becomes incomprehensible.
To keep things as simple as possible then, the oxidised LDL hypothesis goes something like this (depending on which version you read). Endothelial cells posses a greater number of receptors for ‘oxidised’ LDL than receptors for normal LDL. This harmful, damaged, oxidised LDL is removed from the circulation by locking on to Lox-1 receptors on the surface endothelial cells. The oxidised LDL is then drawn into endothelial cell, transported through it, then ejected into the artery wall behind – although why an endothelial cell would want to do this is not clear. (And how an endothelial cell might do this is even more opaque.) After all, the liver has millions of receptors for oxidised LDL (Lox-1 receptors), and it sucks oxidised LDL from the circulation almost instantly – which is why there is very little oxidised LDL in the bloodstream: the liver doesn’t like damaged goods in the blood. But if there is very little oxidised LDL in the bloodstream, then how… I know, I’m not going there.
Ignoring the fact that no one knows why endothelial cells would choose to eject oxidised LDL into the arterial wall behind it – or even how they do it – the theory states that once oxidised LDL starts to build up in the artery wall, white blood cells, called monocyctes, are attracted to the area to clear up the ‘damaged’ LDL. These monocytes then travel between the endothelial cells (converting themselves into macrophages) and set about absorbing the oxidised LDL and then… Well, they absorb so much oxidised LDL that they explode, as they have no off-switch to tell them that they have absorbed too much. No, I am not making this up. (OK, the technical term is ‘rupture’, but ‘explode’ sounds much more fun.)
Soon, more monocytes are attracted to the area. They too convert into macrophages, then over-fill to the point where they explode, and you are left with a big mass of dead macrophages, bits of oxidised LDL and lots of cholesterol, all floating about in a blob of goo. And that, ladies and gentlemen, is how an atherosclerotic plaque forms.
One question that no one seems to bother answering is this: how could such a process ever actually stop? If macrophages keep filling up and then exploding, we would seem to have a positive feedback loop on our hands. The more oxidised LDL there is, the more monocytes are attracted to the area, they convert into macrophages, gorge themselves to the point where they explode, releasing more oxidised goo. More monocytes are attracted, more explosions, more goo. This process does not appear to have any off switch. Or if there is, researchers are keeping remarkably quiet about it. It’s just another one of the many fault lines in heart-disease research where people – rather than answering the question – start using high-speed jargon, change the subject, or say things like, ‘Moving on, it is now clear that…’
Anyway, according to the oxidised LDL theory it is not the level of LDL that matters, it is the level of oxidised LDL in the bloodstream that is important. Thus, if you can find substances that act as antioxidants, e.g. vitamin E, beta-carotene and vitamin C – the sort of things found in green-leaf tea – you will stop LDL from getting oxidised and prevent heart disease. Frankly, if you believe this, you will believe anything.
I don’t even need to believe it. Because no study on antioxidants has managed to unearth any difference, whatsoever, in the rate of heart disease between those taking the antioxidants and those taking the placebo. Of course, this has made no difference to anything at all. ‘It was the wrong sort of antioxidants, you see.’ Don’t worry, until a positive study appears (which is going to happen by chance at some point), it always will be the wrong sort of antioxidants. Then, suddenly, it will be just the right sort.
The oxidised LDL hypothesis is not the only new hypothesis to spring up in an attempt to shoehorn in LDL as the primary cause of heart disease. We now have not one but three mainstream competing hypotheses. This hardly suggests that things are heading towards a speedy resolution, with only a few loose ends to tidy up.
The simple fact of the matter is that after many decades, and hundreds of millions of pounds spent, no one truly has the faintest idea exactly how LDL causes heart disease. But you would never believe this from listening to the experts talk, and reading what they write. LDL still rules supreme.
Personally, I find it rather weird that as I listen to yet another opinion leader outlining complex discussions on LDL receptor down-regulation of this, and microtubular transcytosis system that, I too come to think that this has all been proven. Bullshit truly does baffle brains. Yes, now you put it like that… Yes… I can see the Emperor’s clothes. They truly are magnificent. Thank God I am finally able to see what everyone else can see.
At times like this, I always try to pull myself back and think what question an intelligent child might ask about all of this. What are the ‘stand-out’ problems with the LDL hypothesis that just cannot be explained, no matter how hard you try? I think that there are three killer questions. And they are these:
• Why don’t veins develop atherosclerosis?
• Why does atherosclerosis develop in discrete (separate) plaques?
• If a high LDL level causes atherosclerosis, how can people with low LDL levels get the same disease?
Why don’t veins develop atherosclerosis?
You may think that veins and arteries are very different. However, in general structure, arteries and veins both have a thin inner lining, one cell thick, known as the endothelium. Behind this lies a thicker layer made up of muscle and connective tissue – known as the media. Wrapped around this, and holding the blood vessel together, is the externa.
In basic structure, therefore, arteries and veins are identical. Indeed, they start life exactly the same way in the embryo. The only difference is that arteries are thicker than veins because they have to deal with a higher blood pressure. You might think of a vein as a puny artery that has not done enough exercise. (Proof of this later.)
Given that fact that veins and arteries have exactly the same structure, and are exposed to exactly the same level of LDL, oxidised or otherwise, you would think, would you not, that if LDL causes plaques to develop in an artery it would also cause atherosclerotic plaques to develop in a vein? If it is a case of LDL somehow passing though the endothelium into the artery wall, and this is a function of the level of LDL in the blood. Yet atherosclerosis never develops in veins. Ponder that thought for a moment, and see if you can come up with an answer that involves LDL as a cause. Because I can’t.
Actually, I have to admit that I haven’t been entirely truthful here. If you take a vein from the leg and transplant it to the heart – as is done in a coronary artery bypass graft – the ‘vein’ rapidly develops severe atherosclerosis. Which means that veins can develop atherosclerosis, if you make them do the job of an artery. Perhaps of even greater interest is that one group of researchers did the reverse to a bypass graft: they took a bit of an artery and grafted it into a vein. OK, they did in rabbits, not humans, but I still think it is enlightening to see what happened:
Three months after surgery, grafted arteries possess similar structures as that of veins. The artery interposed to vein did not develop atherosclerosis and underwent atrophic remodeling.
Effectively – in rabbits, at least – if you insert a bit of artery into a vein, the artery wall narrows and thins until you are left with something that looks identical to a vein – in fact, it is a vein. Not only that, the converted artery/vein is immune to developing atherosclerosis.
What does this prove? Well, think about it logically:
• Veins and arteries are exposed to identical levels of LDL.
• Arteries develop atherosclerosis.
• Veins don’t.
• Replace an artery with a vein, and the vein develops atherosclerosis.
• Replace a vein with an artery, and the artery is protected against atherosclerosis.
Conclusion: something about the position of arteries within the body or the job they do causes atherosclerosis to develop. That something cannot be LDL, or oxidised LDL, because this factor remains constant throughout the circulatory system.
Why does atherosclerosis develop in discrete plaques?
Another major problem with the LDL hypothesis is the fact that, even in severely diseased individuals, most arterial walls are completely unaffected. If the level of LDL, oxidised or otherwise, is the main cause, how come some bits of artery don’t get touched?
To my mind, if raised LDL causes atherosclerosis through some form of excess exposure, then finding discrete patches of atherosclerosis is akin to lying in the sun all day, yet only getting sunburned in a few small patches, the rest of the skin remaining unaffected. How likely does this seem? Not very, I would suggest.
This, I suppose, is actually a similar problem to the vein/artery conundrum. Why don’t veins get atherosclerosis, and why are some areas of arteries vulnerable, while others are not? If LDL is leaking through the endothelium, or being transported through endothelium (or whatever is supposed to be happening), and then entering the arterial wall behind, this should happen everywhere, in all arteries. Unless… some bits of the arterial tree – e.g. coronary arteries, carotid arteries, or bits where blood flow changes direction quickly – are exposed to greater flow turbulence, and this causes damage to these areas, so plaques form at these places. If you are thinking this, then I would agree with you.
But what you would probably then go on to ponder is that plaques start at areas of endothelial damage – fine, no problem with that – but LDL does not damage the endothelium, no matter what the level. What damages the endothelium therefore, has to be something else. Ergo, something else causes heart disease.
If a high LDL level causes atherosclerosis, how can people with low LDL levels get the same disease?
Now to look at my final problem. How can high levels of cholesterol cause heart disease if people with low levels get exactly the same disease? This would be the only example in recorded history of a factor causing a disease when it isn’t even there.
There are two mainstream ad-hoc hypotheses designed to deal with this problem. The first is to claim that it is oxidised LDL that is the problem – see above for my response to this incoherent ad-hoc hypothesis thingy. The second is to claim that no one in the West has a low LDL level – that we all have high levels. Yes, every single one of us. Ho-hum, here we go again.
I have to admit, though, that the ‘everyone in the West has a high cholesterol level’ argument is a cracker. It’s complete rubbish, of course, but it carries a kind of eerie power. Primarily because ‘everyone in the West’, it seems to me, is truly convinced that we have all fallen from the higher state of grace granted to primitive people who are ‘at one’ with nature. We listen not to the great goddess Gaia as we plunder the world of its riches, isolated in our metal carapaces (cars), disconnected from the ebb and flow of the seasons and nature itself. Numb to the wondrous circle of life. Yes, I know, I feel terribly guilty too. But I can still comfort myself, just a little, with my vastly improved life expectancy, freedom from nasty infections and infestations, Bose hi-fi system, air-con in the car and suchlike. The occasional malt whisky also helps.
The high priest of the ‘primitive is best’ philosophy is Dr JH O’Keefe Jr. This, from a recent paper produced by Dr O’Keefe along with four of his colleagues:
Optimal low-density lipoprotein is 50 to 70 mg/dl: lower is better and physiologically normal
The normal low-density lipoprotein (LDL) cholesterol range is 50 to 70 mg/dl for native hunter-gatherers, healthy human neonates, free-living primates, and other wild mammals (all of whom do not develop atherosclerosis). Randomized trial data suggest atherosclerosis progression and coronary heart disease events are minimized when LDL is lowered to <70 mg/dl. No major safety concerns have surfaced in studies that lowered LDL to this range of 50 to 70 mg/dl. The current guidelines setting the target LDL at 100 to 115 mg/dl may lead to substantial undertreatment in high-risk individuals.
By the way, 50–70mg/dl in US units converts to 1.3–1.8mmol/l in UK units. This is considerably less than half the current average LDL level in the UK. For the vast majority of us, the only way you could get your LDL level this low would be to take a high-dose statin for the rest of your life. Based on current figures, this would mean 99 per cent of the population of the western world taking statins – forever. On that basis, I make Pfizer a buy.
Perhaps, at this point, it would be tempting to name other names and outline how much money is spent on first-class flights, slap-up dinners etc by the statin manufacturers. Indeed, I would like to, but my publisher say’s it is too hazardous. Of course disclosure of doctors’ links to pharmaceutical companies is freely available if you care to look.
Indeed you would surely find it easier to pass through the eye of a needle than to find an eminent cardiologist who has not been paid… (Sorry, they are not paid, they are given honoraria. Start again.) You can hardly find an eminent cardiologist who has not been given honoraria by at least one pharmaceutical company that manufactures statins. And the sums involved are far from small. We are talking, in many cases, hundreds of thousands of dollars per year. I know, for I – ahem – have signed some of the cheques.
By way of illustrating the connections between industry and opinion leaders, it might be interesting to look at the American National Cholesterol Education Program (NCEP). This is a hugely influential body in the USA that has developed three sets of guidelines on LDL lowering, each time revising their treatment level for LDL further downwards. Where they lead, all others follow.
The last set of NCEP guidelines actually caused a bit of an outcry. Here is a section of an article that appeared in the Washington Post at the time:
On July 13, the National Cholesterol Education Program (NCEP), part of the National Institutes of Health, unveiled tougher guidelines for cholesterol levels – guidelines so stringent that millions of Americans at risk of heart disease would have to take costly statin drugs to meet the new lower limits. What the NCEP didn’t unveil was that most panel members who helped write the recommendations had financial ties to the pharmaceutical companies that stood to gain enormously from increased use of statins.
Critics immediately complained about the hidden financial ties, and demanded disclosure. Within days, the highly respected sponsors of the cholesterol guidelines – the NIH, the American Heart Association (AHA) and the American College of Cardiology (ACC) – posted the disclosures on the NCEP’s web site. The extent of the connections was stunning: of the nine members of the panel that wrote the guidelines, six had each received research grants, speaking honoraria or consulting fees from at least three and in some cases all five of the manufacturers of statins; only one had no financial links at all.
If all the members with conflicts had recused themselves, in fact, only two would have been left.
That didn’t look too good, so a day or so later, another note appeared on the site, attempting to make the guidelines seem more credible. It explained that the panel’s draft proposals had been ‘subjected to multiple layers of scientific review,’ first by the NCEP’s coordinating committee, ‘consisting of 35 representatives of leading medical, public health, voluntary, community, and citizen organizations and Federal agencies,’ and then by the scientific and steering committees of the heart association and the college of cardiology. ‘Altogether approximately 90 reviewers scrutinized the draft,’ the note said. So the message to the public: No need to worry about pro-industry bias.’
Jerome Kassirer [Editor in Chief emeritus of the New England Journal of Medicine and a professor at the Tufts University School of Medicine.]
Actually, there was another bit of this article that I really enjoyed. It was part of a discussion about how bias can creep into collective decisions: ‘When companies with identical interests are underwriting virtually all the researchers, decision makers can become susceptible to “group think.” The military has a name for this sort of trap – “incestuous amplification.”’
I love the concept of incestuous amplification. Never was a truer phrase coined than this. I knew that there had to be some psychological explanation for the collective madness surrounding LDL levels and treatment with statins. My phrase for it was the ‘tyranny of conformity’. But I prefer ‘incestuous amplification’. It sounds much more pathological and in need of treatment.
This all concerns me greatly. These guidlines carry real weight with the medical proffession and the public and it is therefore essential that the public have confidence in the process that produces them. Where else in an area of critical endeavour would such a potential conflict of interest be allowed?
Anyway, to return to the main point, which is the ad-hoc hypothesis that ‘everyone in the West has a high cholesterol level’. I shall just run through the argument again. In comparison to ‘primitive man’, everyone in the West has a cholesterol level that is far too high. Thus, everyone in the West who dies of heart disease automatically must have have a high cholesterol level.’ Huzzah! You’ve got to admit, this knocks teleoanalysis into a cocked hat. Everyone is abnormal, and all shall be treated.
But how can this argument actually be supported? We have no idea what the ‘healthy’ cholesterol levels of our free-range ancestors might have been – surely we must be guessing? Not so. You see, according to JH O’Keefe Jr, we should look at the surviving communities of hunter-gatherers left in the world, animals in the wild, and infants as yet ‘uncontaminated’ by a western lifestyle. By analysing these groups, we should be able to see what a natural, healthy, ‘primitive’ LDL level should be – the level that we should strive to attain. In his words:
The normal low-density lipoprotein (LDL) cholesterol range is 50 to 70 mg/dl [1.3–1.8mmol/l] for native hunter-gatherers, healthy human neonates, free-living primates, and other wild mammals (all of whom do not develop atherosclerosis).
While on the surface this sounds sort of reasonable, in reality it is complete balls. Let’s just open up this statement a little bit more. Firstly, I would like to point out that a healthy human neonate is a child under the age of four weeks. What can their LDL level tell us about healthy adult levels? Precisely nothing.
Would O’Keefe like to argue that neonatal blood pressure also represents the ‘ideal’ for adult blood pressure? The average blood pressure of a neonate is about 80/40. If you found that level in an adult it would indicate massive blood loss, shock and imminent death. So maybe it’s not that healthy after all. In fact, maybe healthy neonates are just a smidge different to healthy adults, and the two should not really be compared.
Moving on to the free-living primates, and other wild animals. The animals he chose were:
• Baboon
• Howler monkey
• Horse
• Boar
• Peccary
• Black rhinoceros
• African elephant
I know that at this point you must think I am starting to make this up. Oh gracious me, no. I have spent many an evening amusing myself by reading his paper. It ranks right up there with teleoanalysis in my pantheon of great medical papers with which you can use to light your barbecue. If you want to read the full version merely type in the following web address: http://www.thepaleodiet.com/articles/JACC%20LDL%20Final.pdf
I thought it might be interesting to match the cholesterol levels, and life expectancy, of the animals O’Keefe chose against those of humans:
| Boar: | Cholesterol level 1.5mmol/l – life expectancy 4–5 years |
| Baboon: | Cholesterol level 2.1mmol/l – life expectancy 30 years |
| Adult American: | Cholesterol level 5.8mmol/l – life expectancy 72 years |
Frankly, I gave up at that point. Mainly because I thought that using the cholesterol levels of other species to make a point about human health was utter and complete… words fail me… the world is going dark. Rosebud… rosebud.…
Where was I? Oh yes, the hummingbird, for example, has a ‘healthy’ blood-sugar level that is five times that of humans. So what? So absolutely nothing at all. A full-grown bull African elephant weighs about six tons. Is that a healthy weight for a human being?
Perhaps you feel that it is still useful to look at the hunter-gatherer communities that O’Keefe chose. They were: the Hazda, the Inuit, the IKung, the Pygmy and the San. Average cholesterol levels were about 2.7mmol/l among these five groups. And yes, it’s true, they also had a very low rate of heart disease.
But I thought it might also be educational to look at the reported life expectancy of these five groups:
| San | = 36 years |
| Hazda | = 32.5 years |
| IKung | = 30 years |
| Pygmy | = 17 years |
| Inuit | = ‘Inuit have the lowest life expectancy of all Aboriginal peoples [in Canada], followed by those living on-reserve.’[I couldn’t get an accurate figure, I think it’s about 55 years.] |
Should we really aspire to reach their super-low cholesterol levels? If we did, could we also look forward to achieving their super-low life expectancy? You think these things are unconnected? Iribarren would argue strongly that very low cholesterol levels are a sign of serious underlying illness, which is why people in the West with low cholesterol levels have such a terrible life expectancy. (By the way, of course hunter-gatherers have very low rates of heart disease – before they reach the age at which heart disease is likely to strike, they are already dead.)
I believe we should ignore extreme examples of horribly unhealthy populations with terrible life expectancies, and look instead at healthy populations with a long life expectancy, e.g. those in North America, Japan and all countries in western Europe. Here, it is blindingly clear that a cholesterol level of 2.7mmol/l is neither normal, nor optimal. It is a sign of ill health and imminent death.
When cholesterol levels in Japan rose from an unhealthy 3.9mmol/l to a much more ‘normal’ and healthy 4.9 mmol/l, life expectancy increased, and death from all forms of cardiovascular disease fell dramatically. But hey, if you prefer to support your argument by analysing the cholesterol levels of two-week-old children, baboons, black rhinoceroses, elephants, the IKung and the Hazda, that is entirely up to you.
In reality, using the ‘totality of the evidence’, it seems clear that a cholesterol level between something like 5.0mmol/l and 6.5mmol/l is normal for healthy adult humans, as it is associated with the greatest life expectancy. On this basis, something around 5.5–6.0mmol/l is average.
And if we use the figure of about 5.7mmol/l as bang-on average, the facts would say that approximately 50 per cent of people who die of heart disease have high cholesterol levels and 50 per cent are below average. Which removes cholesterol as a risk factor entirely.
The Framingham risk score
Just before signing off on this section, I thought I should mention something else: the ‘Framingham risk score’. This has been developed over the years using the major risk factors discovered in the population of Framingham, near Boston, to calculate your risk of dying of heart disease.
To look at this scoring system, you would get the impression that absolutely everything about risk of heart disease had been worked out to the nearest fraction of a percentage point. There is nothing left to discover. Plug yourself into the Framingham risk calculator and you can work out your risk to within a 1 per cent tolerance. To do this, you use the following criteria:
• Sex (gender)
• Age
• Total cholesterol levels
• Smoking status
• HDL level
• Blood pressure (systolic)
(Some tables also add in whether or not you are diabetic.)
Each of the above criteria has points attached, e.g. if you are a man aged between 20–34, this age gives you a score of −9. If you are a man aged between 50–54, this age gives you a score of +6. If you are a man aged between 50–54 and you have a cholesterol level between 5.0mmol/l and 5.5mmol/l, your cholesterol level at this age scores an additional 4 points – and so on.
Once you have added up all your points, you then reference the handy percentage risk ready-reckoner. Then you find, just to give a couple of examples, that a fifteen-point total equates to a three per cent risk of developing heart disease over the next ten years. Nineteen points equates to a ten-year risk of eight per cent.
This is all incredibly precise – is it not? Well, if you are a white, Protestant, American male living near Boston, it is. But what about other populations? A study in Italy used the Framingham risk score and found that:
The estimated number of coronary events by the Framingham function was 2,425 in women while that observed was only 1,181 (ratio 2.1). In men 9,919 events were expected and only 3,706 were observed (ratio 2.7).
In short, if you want to make the male Framingham risk score work in Italy, you have to divide by 2.7. If you want to make the Framingham risk score work in France, you actually need to divide by 4. So we have this fantastically accurate calculator of cardiovascular risk, which overestimates risk in Italy by 2.7 and in France by 4.
Even in the UK, which is much more like the US in overall heart-disease risk, the Framingham risk data was hardly accurate.
When the Framingham Risk Equation using cholesterol levels was applied to British men for ten years, it was found that 84% of the heart disease occurred in the men classified as low risk!
Furthermore, 75% of the men classified as high risk using the Framingham Risk data were still free of heart disease ten years later. It seems the equation is still missing a few important variables.
BMJ, 29 November 2003; 327(7426): 1267.
Looking in the other direction, if you apply the Framingham risk calculator to Australian Aboriginals it is capable of underestimating the risk by up to thirtyfold:
The observed incidence was about four and three times the predicted incidence for age groups <35 and 35–44 years, respectively, and about twice the predicted incidence for those over 45 years of age. The Framingham function was a particularly unreliable predictor for women, especially younger women, in whom the observed CHD rate was 30 times the predicted rate [my emphasis].
http://www.mja.com.au/public/issues/182_02_170105/wan10439_fm.html
In summary: if you use the Framingham risk score for French men, it overestimates risk by 400 per cent. If you use it for younger Australian Aboriginal women, it can underestimate the risk by 3,000 per cent. Which gives a tolerance of 12,000 per cent. Let’s put it this way: I hope the people who developed the Framingham risk score aren’t designing aeroplanes. At least, not any aeroplane that I am going to fly on.
However, my main takeaway point from this is the following. The Framingham risk score, with its exact measurements, and exact calculations, gives the very strong impression that the mainstream research community now knows all the risk factors for heart disease in exact detail, fully understands them, and can control them. Everything important has been discovered. And if you were to follow the advice of the experts you could probably avoid heart disease altogether. The reality, however, is that almost nothing is explained by the conventional risk factors. If an Australian Aboriginal man can have exactly the same risk factors as a French man, and yet have 16 times the rate of heart disease, this means that the conventional risk factors can only explain a maximum of 6.25 per cent of the total risk of heart disease, which leaves another 93.75 per cent lurking out there, yet to be discovered.
The high priests of heart disease may wish to give the impression that they have the required wisdom to keep death at bay. The reality is that they haven’t got a clue. You might as well resort to blood sacrifices and drilling holes in the side of your head. Did someone call for leeches?
