Today Sceptic Soc, our humble society, joined forces with MedSoc to welcome Petro Dobromylskyj to our hallowed [concrete] halls to talk about his scepticism about cholesterol (In fact the reason for MedSoc’s involvement was entirely to attract numbers, and the entire business of inviting the speaker and organising lunch was skilfully organised by Theo, two posters were engineered by James and I set about the fun task of propaganda and publicity). I actually wasn’t expecting much – basing my predictions on the homeopathy speaker we had last time I was fully expecting the proceedings to consist predominantly of Biology teachers hurling intellectual abuse at an unfortunate speaker. In fact, it all went pretty well.
For starters, Theo got the guy’s name right when introducing him: generally considered a good start. Dobromylskyj rapidly launched into a concise list of reasons (consisting of a single element) why he would be unable to expand on the theory part of his talk: time. With the committee feeling slightly guilty, he pressed on, and what he had to say was not only impressive, but actually quite convincing.
Firstly, about his background, it’s clear from his blog that he isn’t just some nutter who simply believes what he wants to believe, hoping that backing from a non-insignificant scientific community justifies his beliefs. All of the posts I skimmed over seemed scientific in nature, although my background in Biology consists mostly of unhappy memories of unwillingly making posters, edible models and various other ‘fun’ activities in the wee hours of the morning for Dr Burnett’s homeworks, so anything more than an iGCSE answer and/or colourful A3 poster is beyond me. He’s also highly qualified: he has a BSc in physiology and a veterinary degree. Alone this doesn’t count for much – a highly qualified professor can still be wrong, taking Steven Hawking’s professor as an example (a documentary-drama had him make a mistake so it must be true…) – however it was clear from our conversation with him over lunch that he was extremely learned and well-read in this area, and I think he really takes this belief seriously and scientifically, thus with good ground. Put a different way: I hadn’t a clue what he was on about so I assume it’s correct. Theo seemed to understand and agree at any rate.
His talk was mostly statistical and he attempted to explain to us the mechanisms of some of the flawed experiments which support the lipid hypothesis. Most of his arguments were statistical – much like the entire global warming debate, the data is actually incredibly dubious: fat intake can have both a positive and negative correlation with cardiovascular events depending on the set of countries considered. He pointed out some graphing mistakes – particularly the case in which the graph was an ‘arbitrary unit’ against frequency graph. On replotting it as a correct cholesterol v frequency density graph the correlation appeared … somewhat less convincing.
Apart from criticising bad statistics he also attempted to explain why the experiments were flawed. I say ‘attempted’ because time was short and proper explanations like the ones he gave involving biological molecules tend to get lost on me. Admittedly therefore I can’t really say very much intelligently about this apart from the fact that most of the studies utilised methods of artificially changing cholesterol levels which themselves affected so many other factors that it is impossible to isolate the effects of the change in cholesterol levels. He then finished up with proposing some alternative correlations – links with stress levels, sugar levels, age, lifestyle, etc.
Although my unbiological ramblings don’t do Mr Dobromylskyj any justice, his talk was actually one of the most mind-changing talks I have ever attended, and it really changed my views on the subject dramatically. When I was putting up posters in the library, upon interrogation I suggested to the librarian that the man was a nutter and he wouldn’t have any reasoning behind these ideas. In fact, I now have a completely different view of cholesterol and the lipid hypothesis. The statistics just don’t seem to stand up to close scrutiny and the obvious and therefore often over-simplistic and often quoted graph of cholesterol levels against heart attack rate just doesn’t quite seem so obviously true any more. Dobromylskyj has successfully convinced me that a large part of the lipid hypothesis is built on poorly designed experiments whose results are further twisted by equally poorly conceived statistics. With the media consistently blowing small experimental fluctuations out of proportion and into ‘fact’, the theory ends up looking like a poorly spun web of lies. I still think there must be some truth to it – he even indirectly acknowledged some sort of correlation since a fatty diet is correlated with a sugary one (think along the lines of weak willpower) – so I shan’t be pigging out on lard anytime soon; but this episode really did get me thinking along new and different lines. Thanks, Theo, for organising this.
Next up on Sceptic Soc: a religious person will come and try to justify God … if and only if we can sort out a date (I have a horrible feeling I told James to tell the speaker it’ll be next week) and be organised enough to make it happen in 7 days – a situation which is highly unlikely and which would falsify my so-far-accurate null hypothesis that we are incompetent at organising.