Sunday, January 19, 2014

"When is anything that artificial ever a good idea?"

Dividing everything into "natural" v "artificial" and then arguing that all the natural things are good and all the artificial things are bad fails the sanity check.

For example, home made vegetable soup is artificial. It's made by humans, it doesn't occur in nature, so it's artificial. But most people would feel that it is clearly a natural thing. At least, it's made from all natural things, like vegetables, herbs, water, cream etc. so it must be natural. But everything is made from natural things. You can't create or destroy matter (the most you can do is transform it to energy, and you can't create or destroy that either) so the most "artificial" thing you can think of is, at the end of the day, made from something that occurred naturally.

Perhaps you feel I am just arguing about semantics, and I must understand what people mean when they say something is "natural" or "artificial". So for now I'll accept a (fuzzy and ill-defined) meaning of the word "natural". But I still cannot accept that everything natural is good and everything artificial is bad.

Natural things that are good: strawberries, hugs, a baby's laughter
Natural things that are dangerous: belladonna, tigers, arsenic, drowning
Artificial things that are dangerous: guns, methamphetamine, Zyklon B
Artificial things that are good: vegetable soup, living in a house, pacemakers, looking things up on the internet

So we can't find out if something is good or dangerous simply by deciding whether it is natural or artificial. So how do we find out if something is good or dangerous? Use the fact-checking method I recommended in an earlier blog post

Sanity Checking

A persistent food myth is that "margarine is one molecule away from being plastic". You can use the method I recommended in an earlier blog post to fact check this one, but there's another way which is "sanity checking".

If I told you that putting three small pebbles in a line on the dashboard of your car would make your car go faster, you would think I was insane. You don't need to Google if it's really true, or look it up on Snopes or Google Scholar, because it fails the sanity check. You may not understand in perfect detail how an internal combustion engine works (I certainly don't), but you know enough to know that lining up pebbles on the dashboard is not going to influence the speed.

How does the "plastic margarine" claim do when subjected to the sanity check? It fails on two points.


  1. Margarine is made of a mixture of stuff, all mixed together (source: I looked at the ingredients on a tub of margarine). It isn't one single molecule to start with, so how can it be "one molecule away from being plastic"? That's like saying "chicken casserole is one molecule away from being plastic". It obviously can't be true.
  2. Unlike margarine, plastic isn't a mixture of things. If you have a plastic cup, for example, the plastic in that cup is made up of lots of molecules all alike. But there are lots of different kinds of plastic made of different kinds of molecules (source: the Wikipedia page about plastic). So the plastic in a plastic cup might be different from the plastic in a plastic bag or a plastic food container and so on. So it doesn't make any sense to say margarine is one molecule away from being plastic. Which plastic? Polystyrene? Nylon? PTFE? It's so vague it's nonsense.

You can use Wikipedia and Snopes and Google Scholar and Google to fact check the "plastic margarine" claim if you like. I strongly recommend you do - remember don't take my word for it, never take any single source of facts on faith. But I'll eat my own hat if you find that it's true. It's not a sane claim; it can easily be dismissed using basic sanity checking.

Saturday, January 18, 2014

Diet Fizzy Drinks

As I said in the last blog post. I was recently involved in an online discussion about the advice given by Change 4 Life. Some people expressed disbelief that Change 4 Life (which is an NHS healthy lifestyle campaign) would advise people to swap sugary drinks for diet drinks with artificial sweeteners. They found this incredible because they said artificial sweeteners "are disastrous for your health".

Is that really true? Are artificial sweeteners disastrous for your health? Where is the evidence?

The people I was talking with had probably seen articles and websites such as this one which claims a whole host of terrible effects of aspartame including "Aspartame poisoning is commonly misdiagnosed because aspartame symptoms mock textbook ‘disease’ symptoms, such as Grave’s Disease." The website has lots of links to the author's book and to a "detoxification programme" she hopes to sell you.

These kinds of articles are so widespread that Snopes.com has a whole page about aspartame written by David Hattan  Acting Director of the Division of Health Effects Evaluation in the United States Food & Drug Administration (USFDA) Center for Food Safety and Applied Nutrition. You can (and should) read it yourself. he gives a detailed rebuttal of many of the most common claims of the risks of aspartame, and concludes "The legitimate attempts that have been made to confirm and replicate allegations of adverse reactions from aspartame ingestion have not been successful and the USFDA continues to consider this to a be among the most thoroughly tested of food additives and that this information continues to confirm the safety of aspartame."

But we never trust just one source of information, so let's keep checking. Wikipedia has a page about aspartame and it also has a page just about the controversy over aspartame's safety with 69 separate references, many of them to peer reviewed journals. The articles seem to confirm Hattan's claim that aspartame has been very thoroughly and repeatedly tested, due to the ongoing public concern, but that the evidence seems to be that it is a very safe additive.

But let's still not rest there. We can use Google scholar to check the primary evidence ourselves. I used the search terms "aspartame safety" which returned 13,100 results. The first was called
Effect of aspartame loading upon plasma and erythrocyte amino acid levels in phenylketonuric heterozygotes and normal adult subjects You need to know that phenylketonuria (PKU) is a genetic disorder which makes the body unable to process the amino acid phenylalanine, present in aspartame but also found found naturally in other foods such as breastmilk. If you have two genes for PKU you have the condition and you would definitely know about it - untreated it causes learning disability. But what if you only have one PKU gene? Then you wouldn't have the condition, but maybe it would make you more susceptible to ill-effects from aspartame, especially if you were consuming a lot of it. It's a reasonable hypothesis, they tested it, and found that if you have one gene for PKU you are slightly less able to process shedloads of aspartame than someone with no PKU genes, but you can still process it adequately, and so aspartame is as safe for you as for anyone else. Without paying moeny you can only read the abstract for this article, which is all I have done. The article was published in 1979 (that's pretty old) in The Journal of Nutrition which Wikipedia tells me is a peer reviewed journal (that's good - you can only publish an article there if other nutrition experts have read it and think it's good science) with an impact factor of 3.916 which isn't bad for a niche journal.

How safe is it for anyone else? The second hit on Google scholar addresses that. It's a review paper which is good because systematic reviews are at the top of the hierarchy of evidence. I keep telling you that you shouldn't trust just one source of evidence, even a well-conducted scientific study. A systematic review gathers together all the scientific studies on a particular topic, weights them according to how well-conducted and therefore reliable they are, and weighs up the balance of evidence from all the available research. So a good, recent systematic review is about as dependable evidence as you can ask for. This review is called Aspartame: review of safety and again all you can see is the abstract unless you pay for the full paper. It was published in 2002 which isn't bang up to date but it's not really old either. It was published in Regulatory Toxicology and Pharmacology which Wikipedia tells me is a peer reviewed journal with an impact factor of 2.132 which isn't terriible for a niche journal. The review concludes "The safety testing of aspartame has gone well beyond that required to evaluate the safety of a food additive. ... it is clear that aspartame is safe, and there are no unresolved questions regarding its safety under conditions of intended use." Seems conclusive ... until I spotted that the email address for the lead author is harriett.h.butchko@nutrasweet.com - she works for the NutraSweet company, who make artificial sweeteners. Obviously, that seriously undermines the trust I want to place in this review. That doesn't mean I automatically believe the opposite of what this review says, just that it carries very little weight in making up my mind whether I believe aspartame is safe or not.

I could keep on looking, but even discounting that last research paper all the dependable evidence seems to be pointing in the same direction - aspartame is a very safe food additive. All the evidence in the opposite direction seems much less reliable - unscientific and trying to sell you detox programs and books.  Change 4 Life are probably right to advise people to stop drinking sugary drinks (which are empty calories) and switch to diet drinks instead (which are safe and won't make you fat). If you don't drink sugary drinks but instead prefer to drink water, then just keep doing that.

Fact Finding

I recently engaged in a Facebook debate about healthy diets and the safety of various foods including artificial sweeteners and margarine. I was asked to dig up some facts because I am an academic. But I have decided against using the university library to find papers on the topic, for two reasons:

  1. Most of the papers I would find that way would be behind paywalls, and inaccessible to the people who asked me to find them
  2. I'd prefer to help those people know how they can check facts for themselves

If you read an article which makes claims, for example an article which claims that artificial sweeteners are terrible for your health, or an article which claims that all homeopathy is bunkum, the very first thing you should do is ask yourself is that really true? Always ask where is the evidence? If you just do that, instead of swallowing everything you read, you're already way ahead of the game.

How can you find out if something is really true? There are a variety of online resources you can use and ideally you should use a combination of them if you want to be really sure of something.

Wikipedia: You should never trust any single source of information and that definitely goes for Wikipedia. Wikipedia is a terrible place to stop looking for information but it can be a pretty good place to start. Wikipedia pages should always give references for any claims of fact, so check that there is a reference, then check out the reference to see if it really does say what the Wiki article claims it says, and whether you think it is a reliable source of information or not. Also check the edit history of the page to see if they thing you want to know about is in the middle of a furious revert war or something like that. If the last edit to this information was made a while ago and the sources seem to be sound and it's an actively edited page (where vandalism would be caught and reverted early) then you can have a fair bit of faith in the content.

Snopes: Snopes checks the veracity of urban myths like The Hook or The Microwaved Pet but it also debunks loads of food myths such as margarine is one molecule away from plastic or the dangers of aspartame. It's carefully researched and it gives sources. But remember, no one site is the last word in trustworthiness.

Google Scholar: This is a terrific and little-known resource where you can search scholarly literature from peer reviewed journals even if you are not a student or an academic. You can get many full text papers, and many more abstracts (a short summary of the full paper written by the author of the paper, not summarised by someone else). It works just like the google search engine we're all familiar with. Type in some keywords and get a bunch of results, then start reading.

Just plain old Google: But use with caution. Don't just see if you can find a website that contains the information you hoped to find. Think about what kind of a site it is on. If it is just some random blog (like this one) you can't necessarily trust the information on there. It might be reliable (like this one!!) or it might be a load of hooey. Keep looking until you find a credible source of information.