Science and FODMAP

One of the lessons of economics is that markets only work well if consumers are able to make informed choices. For a variety of reasons, including fundamental flaws in the human thought process, it is hard to make informed choices about issues of diet and health. A natural result of this is that the market for food and diet advice is filled with worthless or dangerous nonsense. Very often the people selling or spreading this junk honestly believe what they are saying or selling, which makes it even more difficult to find that it is nonsense.

In this post, I will demonstrate the basics of how to research food and diet advice with Google Scholar, using the low FODMAP diet as an example of the search process. When researching anything that might be a fad diet, never use a standard web search. You will be buried under an avalanche of junk. There are thousands of nonsense health claims that have thousands of cultlike web-savvy followers, and these people are often very convincing.

People have been believing dumb things about food for all of human history. It is only in the past few centuries that we have developed a tool to find the truth: the scientific method. Many people do not understand what science is, which means that the trappings of science, like labs and machines and big Latin words, are often used to sell junk. Real science is the independent experimental testing of ideas.

Your friend trying the diet and then feeling better does not count as an independent experiment. Health can change at random, the placebo effect makes people feel better when they think they are being cured, and confirmation bias makes people remember only what confirms their existing beliefs. Good scientific experiments correct for these problems with a variety of techniques. Whenever someone tells you some 'fact' or theory about food, you need to check, using http://scholar.google.com/, that multiple independent experiments have tested this claim before you make any major changes to your diet based on it. Often you will not be able to read the full text of the articles unless you are at a university or library, but that is ok. Reading the title and abstract and looking at the authors will often tell you what you need to know.

The first result that comes up when I type 'fodmap' into Google Scholar is a review article, summarizing the evidence in favor of using the low FODMAP diet as a treatment for irritable bowel syndrome. This looks like a proper scientific article, written in the right style and using the right words in the right way. I will not go over these, because quacks have learned the 'magic words' and use them as often as they can. I will instead point out a very strong signal that you are dealing with real science: the 'Limitations and potential concerns' section:

The diet is not a panacea for patients with FGID. It provides good relief of symptoms in about 75% of patients, but has little benefit in some. Studies have yet to identify predictive factors of benefit apart from dietary adherence. Intermittent symptoms remain, albeit at a now tolerable level, in many patients since the underlying FGID is not directly addressed by the diet. Patients should not be given expectations of a 'cure'.

...

While these suggestions are all unsubstantiated, they do provide a reminder that this dietary intervention is established for those with functional gut symptoms and is not a diet for otherwise healthy people.

People who are trying to sell fad diets will claim that their diet cures everything in everybody, and this claim will be spread in cult-like fashion by the adherents of the diet.

However, I noticed while reading the review article that the evidence base for the diet is thin. They only cite is only one double-blind placebo-controlled trial of the diet, it only involved a small number of people, and it was run by the same person who wrote the review article. The existence of this one trial is more evidence than most fad diets ever muster, but the fact that there are not more tests is a warning sign.

After reading a little more, I found that there is a small team at a university in Australia that is researching and promoting the diet. They are probably good researchers trying to investigate a promising treatment, but in a situation like that it is very easy for groupthink to emerge and start biasing results. Independent replication is the key to science, and nothing should ever be called scientific until many independent teams have tested it.

So I looked for studies of the diet by people who were not associated with the Australia team. I ignored every article written by any of the authors of the review article. The review article was one strong piece of evidence that FODMAP can help people with irritable bowel syndrome, but nothing else written by the same people should be counted as additional evidence in favor of the diet. It would not be an independent confirmation, just the same people saying the same thing in different places.

Most of the other articles were other nutritionists discussing the diet, and most of them were positive. This is weak evidence in favor of the diet. It is good that other health professionals have evaluated the Australia team's evidence and agree with their conclusions, but nutritionists often have a herd mentality. The history of medicine is littered with worthless things that were believed by thousands of health professionals. The key is independent experimental evidence.

In the end, out of about a hundred articles, I found about half a dozen independent experimental studies of the diet. One of them was in feeding tube patients, and found improvements due to the diet. Another involved testing the gut response to FODMAPs in healthy volunteers, and verified the theory behind the diet. There were tests of the diet in Britain and Norway, although they were not placebo-controlled, that showed positive results.

None of these articles would be considered good evidence in isolation, but they all served as independent confirmation of the science and results presented by the Australia team, and I could not find any article pointing to flaws in their research. I also saw nothing to indicate that there is a better way of treating irritable bowel syndrome. Therefore, I feel comfortable recommending that people with irritable bowel syndrome try out the FODMAP diet.

There is a caveat, however. The diet suggests restricting five types of carbs: fructose, lactose, fructans, galactans, and polyols. The placebo trials only tested the response to fructose and fructans, and the other trials followed the advice of the Australia team to restrict all five. The restriction of the other three is based on theory, not evidence. It may be that their restriction is unnecessary, and it may be that something else also needs to be restricted. The diet in its current form should be considered a first draft, and we need to be open to the possibility that further research will change things. With only a few real studies telling us what we now know, there is a lot of opportunity for things to change.

This is how science works. We collect data in an attempt to understand a complicated world, and we rely on other people to check our work for mistakes. Knowledge advances one experiment at a time. Although we would like to have more data, the data we do have suggests that the FODMAP diet has a good chance of helping people with irritable bowel syndrome. It has passed far more tests than most of the other diets out there.