Hate of Irongarm

I am not in the same state as Rant when he said that he hates Irongarm, but I am getting... cautious. I noticed a while ago that people are overly sensitive when their beliefs are challenged. This must be an evolutionary survival mechanism, maintaining the status quo. As the result there is plenty of dogma around.

Issues regarding health and nutrition is the perfect example of such dogma. What's good for you? Everyone knows that saturated fat in food is bad, as well as cholesterol. On the other hand, eating grains, complex carbd, lean meats and plenty of fruits and vegetables is good: it will prevent disease and make your life longer and happier. The only problem is that the evidence for thuis is very, very thin. Often non-existent. For starters, For those really interested I recommend Anthony Colpo's The Cholesterol Con and Gary Taubes Good Calories Bad Calories.

The latest exchange of pleasantries in IGx revolves around longevity issues, namely the Okinawan Study. Methodologically, every type of scientific research has problems. While randomised controlled trials are considered the gold standard, they have serious drawbacks. Observational studies are useful in their own right, but have even more problems. Yet, it is exactly these type of trials that attract attention of the media. They are easier to understand and usually produce more understandable results. For example, the statement such as "most of drivers involved in serious accidents are under the age of 25 years old" is very illustrative and simple and allows to make - practical recommendations. The only problem is that these recommendations are wrong more often than not.

In the Okinawan study they examined and questioned 900 centenarians and summarised the findings. Old okinawans eat this and that and they do this and that. Nothing wrong so far. Where they do go wrong is making far reaching conclusions, particularly about the impact of their lifestyle. The study like this is subject to so called survival bias which in this case can be taken in literal sense: you see only those who survived to the age of 100 and have no idea how many of those following the same lifestyle died at earlier age. These kind of mistakes are made over and over again. One paediatrician once mentioned to me that non-steroidal anti-inflammatory drugs are very bad: all kids that had serious complications were on them. When I suggested that maybe because sicker kids are put on these drugs, maybe it is more severe disease that is responsible, not NSAIDs, he said that, no, NSAIDs are given virtually to every sick kid. So why blame the drugs, obviously vast majority of those who receive it don't develop problems!

Descriptive study on Okinawans suffers from the same fallacy. You only see those who have what we want, in this case the age of 100 and higher. But you have no idea how many people followed the same lifestyle but died earlier, maybe much earlier. Even though the study is ongoing, the survival bias still applies, as there is no comparison.

If you analyse car accidents and find that most car in crashes are red, should we go on and recommend re-painting all red cars in order to reduce the number of road deaths? Another example: most people admitted to hospitals with heart attacks are over the age of 60 and have grey hair. Should you dye your hair in order to reduce the chances of an infarct? Ridiculous, isn't it? So why then should you take the lifestyle of very old people and conclude that this is the reason for their longevity? Australia is the second longest living nation on Earth. Australians diet consists of pies, potato chips, lots of fast food and Tum Tams, sort of concentrated sugar bisquits. The obesity rates in Australia are among the highest in the world. Would you recommend, say, Kenyans, whose longevity is far down the list to adopt Australian diet in order to prolong life? No? Why then it seems logical to recommend Okinawan diet to everyone else?

No doubt, lifestyle factors may be importans for longer survival. Stress is probably number one killer. Moderate physical activity seems to be good as well, though the evidence is less convincing. Diet - here the facts mess the reality completely. A lot of research reveals facts that run opposite to accepted views. Dietary fiber does not prevent colon cancer. Old people with high cholesterol live longer than those with normal one. Saturated fat may have protective effect against cancer while not increasing the risk of cardiac disease. Adding fruits and vegetables to your diet in several studies failed to make a difference to the recurrence of cancer or survival. An interesting post on this here:
http://junkfoodscience.blogspot.com/2009/02/australian-doctors-are-introduced-to.html

If you analyse enough data long enough you will find a lot of interesting associations. However few of them will be true ones. The perfect example is the book The Bible Code. Two respectable researches loaded the Torah into the computer and analysed what is going to happen if you read every second letter, every third, fourth and so on. What will come out if you read the text diagonally, in squares, triangles and what not. Not surprisingly, they found a lot of interesting stuff, such as “Izhak Rabbin will be assassinated”, “the towers will fall” and another similar garbage. Holy Predictions? No, just alphabetical massage. As Nassim taleb mentioned in his book, if you put infinite number of monkey in front of typewriters, it is a certainty that one of them will produce Homer’s Iliad.

The lesson for me has several aspects. Most people are not interested in the discussion of facts, but rather in the selection of facts that confirm their beliefs. If you challenge these beliefs the people in question may become surprisingly uncivilised. Well, I am not there to destroy a church. Blessed are the ignorant. Amen.