9th August 2017
Looking for the contradictions
Here was my mantra. ‘If I can find an absolute contradiction to any hypothesis, I shall discard it, and start again.’ I have to tell you that this shiny, bright eyed scientific idealism has had to bite the dust. Primarily, because it can be very difficult to know what a contradiction looks like – for sure.
If Newton had found that every so often apples did not fall from a tree, instead they accelerated upwards and into space, the theory of gravity would not have been born – because it would have been wrong. If your hypothesis is that all swans are white, then the finding of a single black swan immediately negates your hypothesis.
However, in science, refutations are rarely so clear cut. In biological science, there are so many things going on, so many variables to consider, that we are more in the world of weather forecasting, rather than Newtonian physics. There are few absolutes, no completely hard and fast rules.
This, of course, has allowed those who believe in the ‘cholesterol hypothesis’ to shape shift, twist and turn, and adapt the hypothesis to fit any facts. Never, ever, can they be pinned down. Never, ever, can the hypothesis be refuted by any single fact, or even a combination of facts. Believe me, I have tried. It is like attempting to nail mercury, firmly, to the table.
Take the hypothesis that a raised cholesterol level causes heart disease. Already, I imagine, you can see this fragmenting before your very eyes. What do you mean by a high cholesterol level. Total cholesterol? Low density lipoprotein (LDL) level? The ratio of LDL to HDL? Are you looking at LDL-C or LDL-P. Are you considering VLDL levels, what about oxidised LDL, or small dense LDL, or light and fluffy LDL.
That, without trying, is nine ‘cholesterol’ variables. And the possible combination of nine variable is nine factorial. This allows 362,880 possible combinations of ‘cholesterol’ that could be tested. In truth, I didn’t really try very hard there with ‘cholesterol’. I could add in at least sixteen variants of HDL (that I am aware of), including apoA-1 Milano (the super-protective form of HDL – allegedly). Which give us another sixteen ‘cholesterol variable).
9 + 16 = 25 variables (assuming they act independently)
The factorial of 25 is 1.55×1025 or: 15,511,210,043,330,985,984,000,000.00
As you can see, there is not the remotest possibility, ever, of trying to work out how all the forms of ‘cholesterol’ may interact. Even if you created theoretical models and fed them into a computer, you would be there for a very, very, long time.
Equally, there is no possibility of refuting the causal impact of any single cholesterol factor. And, if you did manage to pin anything down, the broader issue of ‘definition’ will simply be altered.
Just trying to look at the apparently simple concept of a high total cholesterol level itself. You would think it would be possible to say that there is an average level, a high level and a low level. This would allow you to say that the average total cholesterol level of everyone in the world (who has had their cholesterol level tested) is five point three (5.3mmol/l). [I just made this figure up]
Thus, anyone above this figure could be said to have a cholesterol level above average. Or high. And vice-versa. Just as you could measure the height of everyone in the world, and find an average. However, this cannot be done. Well, it could be done, but it has not been done, and I suspect it never will be done. Because, in the case of cholesterol levels, average is most definitely not considered ‘normal.’
Here, for example, is what is said about cholesterol levels on the Benecol website:
‘The government recommends that healthy adults should have a total cholesterol level below 5 mmol/L. In the UK, three out of five adults have a total cholesterol level of 5 mmol/L or above, and the average cholesterol level is about 5.7 mmol/L, which can be a risk factor in the development of coronary heart disease.’1
Thus, the average cholesterol in the UK is not normal. It is ‘high’ enough that it is a risk factor for heart disease. So, average is not normal. Is 5mmol/l normal? Well, Heart UK (The UK cholesterol charity – funded almost entirely by the pharmaceutical industry), makes this statement:
‘Total Cholesterol (TC) – this is the total amount of cholesterol in your blood. Ideally it should be 5 mmol/L or less.’
Which would suggest that anything below 5mmol/l is fine and normal? But if you have diabetes, you should have a cholesterol below 4.0mmol/. Diabetes UK lists the following blood lipid (cholesterol) targets as a guide for people with diabetes:
- Total cholesterol: under 4.0 mmol/l
- LDL levels: below 2.0 mmol/l2
Which means that four is actually better than five – thus five is high? And if you have had a heart attack it is recommended to get cholesterol levels below 4.0mmol/l. Ergo, a level of 5.0mmol/l must be causing the developing of heart disease. So, five is not actually normal. It is high.
The general consensus, though never very clearly stated, is that, whatever your level of cholesterol, you will gain benefit from lowering it. Which, logically, means that any level of cholesterol increases the risk of heart disease. Thus, there is no optimal level. I have seen it argued that the optimal level for cholesterol is 1.5 mmol/l. 3
Setting the level at this point means that, apart from a vanishingly small number of people, everyone in the western world has a ‘high’ cholesterol. Therefore, you can never argue that a high cholesterol does not cause heart disease, because everyone who suffers from heart disease has a high cholesterol level. In contrast, no-one with a ‘normal’ cholesterol level suffers from heart disease.
With cholesterol levels, we have the following situation:
High = high
Average = high
Low = high
Very low = high
Very, very low = high
So low that you cannot find anyone with this level* = normal
When confronted with logic like this, the cholesterol hypothesis is perfectly protected from attack. It is a non-refutable hypothesis. As Karl Popper said, if you cannot construct your hypothesis in such a way that it can be refuted, it is not science a.k.a. nonsense.
Which is why, in the end, I decided on another approach entirely. Replace the cholesterol hypothesis with something that actually fits the facts without the need for endless distortion of facts, and reality. Also, to try to create a hypothesis whereby data could be found to refute it.
At present, just to repeat myself for the final time, the cholesterol hypothesis is that a high cholesterol level causes CVD. This cannot be refuted, because there is no such thing as a normal cholesterol level. All levels are high. Res Ipsa Loquitir.
*or at least, so few people exist that no study could ever be done