Friday, May 29, 2015

Chocolate helps weight loss, a hoax

Health and dietary science is clearly one of the softest disciplines out there but sometimes, one may still be surprised that it's softer than he has thought.

The story is the following: Exactly two months ago, media in 20 countries brought their fat readers good news. Scientists have found out that you may accelerate your weight loss if you eat chocolate with lots of cocoa on a daily basis.

The most important article for masses appeared in Der Bild, Europe's highest-circulation journal: Wer Shokolade isst, bleibt schlank! The supertitle said "Diese Studie schmeckt uns" i.e. "This study tastes well to us". It surely does, as I will discuss below.

The English-language media were somewhat underrepresented but the reports appeared in The Express, The Daily Star, The Irish Examiner, The Huffington Post, Times of India, some TV news in Texas and Australia, and others.

Because the statement sounds surprising, a sane person should ask what is the basis for the claims.

Well, all of it boiled to an article in The Institute of Diet and Health, a website where you can publish your articles for a fee. Dr Johannes Bohannon, the lead author, apparently published an article and a press release over there. The "technical" article has been removed by now.

Their method is simple: they divided 15 fat humans on a diet to 3 groups. One didn't change its diet, one lowered carbohydrates, and one lowered carbohydrates but added a chocolate bar with a high cocoa content every day. The last two groups were losing weight but the third group was losing it faster, by 10%.

It sounded great to tons of journalists.

Well, the lead author Johannes Bohannon PhD – whose real name is John Bohannon PhD but his PhD is in in the science of bacteria – is actually a journalist of a sort these days. The hoax study was masterminded by two other journalists/filmmakers and they picked Bohannon because he has previously succeeded in similar hoaxes. He has published nonsensical diet studies in technical journals years ago – but they were not widely reported.

This time, he needed to pay just EUR 600 to the fee-based journal – which claims to be peer-reviewed but if the grammar is OK and if you pay, they clearly publish everything without the change of a single word. And that was enough for the results of this "science" to be uncritically trumpeted to millions of people in the world.

Everyone may read Bohannon's I fooled millions... report.

You should appreciate the huge gap between how little is needed to produce this bogus science; and how huge its impact usually is. It took $650 or so to make this nonsense published. But millions of people could have bought dozens of extra chocolates for that. If the hoax were engineered by Nestle rather than Bohannon, that company would have earned tens of millions of dollars in profits for the price of $650. An incredible investment, indeed! Everything that is needed for it to work is a bunch of imbeciles distributed across the world, the so-called health journalists.

And you need so little to see that the paper is almost certainly nonsense. First of all, if someone has followed Science or made a search for the words "Bohannon, fake research", he would have easily found a 2013 article in Science that tells us that earlier in 2013, Bohannon had actually submitted a whopping 304 fake research papers to open access journals. About 1/2 of them were accepted.

You know all the people in many corners who worship the concept of "open access journals". In most cases, it just means "open access for junk journals". It's that simple.

The content of the paper is untrustworthy, too. The sample was so small that false positives are virtually guaranteed, so even if they observed what was reported, it could have easily been a coincidence, not solid evidence supporting the surprising effect of chocolate.

Your weight loss is 500% accelerated if you eat a pound of jam every day.

The main trick is that if you eat "something special", like chocolate a day, and especially because of the flavonoids in cocoa, you may get surprising results in your diet. This is exactly the kind of "magic" that journalists and their readers want to believe. If you serve it to them, they will eat it.

But the whole "type of faith" is pretty much irrational. It doesn't matter much in which form you swallow the carbohydrates. What matters is primarily the quantity! In fact, I would argue that even the separation of food to carbohydrates, fats, and perhaps even proteins is heavily overrated. What primarily matters for one's weight is how much food – or how many calories – he eats in total. To a large extent, your body converts one form to another. Not all fat people eat too much but most of those who do and who are unhappy about it sometimes eat "lots of calories" that are unaccounted for, that they overlook, and that's the actual main reason of their weight gain.

And flavonoids are "secondary" (not quite essential) metabolites for plants but animal bodies like ours largely ignore them. They only add "flavor" to food and perhaps help to kill some germs.

Bohannon's hoax is another example of the incredible laziness and the irrationally uncritical approach of the health and science journalists. If you formulate your text in a certain way so that "certain people want to hear such things" and many people are ready to say that "it sounds convincing", you may make it publish almost anywhere even though it's complete bunk.

Now, Bohannon's primary goal was a hoax. There are lots of other authors whose primary goal is to become famous, rich, or who genuinely want to contribute to science but they're too stupid to do it well. But if they fix the grammar and increase the proportion of "what the readers want to see", they may always succeed with pretty much arbitrarily nonsensical junk. In fact, under certain circumstances, the more nonsensical and absurd certain reports are, the more enthusiastically certain people will spread it!

The recent wave of coverage of the warp drive and EM drive crackpottery is a truly childish example of that correlation. But even when we discuss the daily doses of junk science, like the constant counterrevolutions undoing quantum mechanics, the mechanisms making it possible for this junk to spread are similar. It has OK enough grammar, the worthless enough website where it initially appears is a good enough symbol of infallibility for too many people, and most importantly, the conclusions match the template of what the millions of readers want to see.

Because hundreds of millions of people not only believe similar nonsense but they are literally thirsty and willing to read something like that, it is clear that the bulk of the mankind doesn't live in a scientific epoch.

One very close relative of mine (...) is obsessed with similar health reports "this is good/bad for you" but you may see that he or she is changing the opinion by pretty much 180 degrees on almost every question every year or two. "Have you read that "this" (...) is exactly the opposite than previously thought?" we hear regularly. Yes, no, cute, I don't really care, it's junk. Why can't such people see that the changes show that the previous reports were untrustworthy and the new ones don't seem qualitatively different so they're probably also untrustworthy – and one should just conclude that we don't know what the answer is? Or we should rely on our experience and common sense instead of sensationalist news? Or maybe the people don't want to make such a conclusion? Is it better to believe in things that are exciting but constantly changing, hoaxes, or noise than to admit ignorance?

No comments:

Post a Comment