Don't believe the marketing droids

I have long been of the opinion that the majority of claims made about the very expensive nutritional and other products that are pushed at you in a massive variety of ways by the food and drink industry are based on either no evidence or very poor evidence. I'm pleased to say that the British Medical Journal now agrees with me and has published a series of papers where they examine the claims made for a variety of nutritional and other products and find them seriously wanting.

BBC News article here

Blog post from a doctor who works in weight control here

BMJ editorial here (long but absolutely damning)


Remember: these companies are selling products with massive markups, for example maltodextrin from myprotein.com costs £12.99 for 5 kg, PSP22 (which is nothing but maltoextrin and flavour, although SIS claim it's "special" maltodextrin) is £8.50 for half a kilo. They market them in a variety of cunning ways: paying athletes to endorse them (remember that just because someone's a professional athlete it doesn't mean they have any morals or an IQ greater than 25 http://www.guardian.co.uk/sport/2012/jul/12/tour-de-france-remy-digregorio), sponsoring races so that you're given their product during the event,  "social" marketing via facebook and twitter (getting people who aren't employed by their company to endorse them) and targetting top age-groupers with free samples in the hope that they'll endorse their product. These things are usually backed up with impressive sounding "sciencey" claims that are usually based on very weak evidence.

Here's an example of the sort of claims they make. Note that they claim very solid scientific support, with links to technical scientific publications. Must work, right? Someone from my club contacted me and asked my opinion as a professional scientist of their evidence. I had a look and this was my reply.

"Briefly, very little of any consequence there. The "publication" is a 1-page published conference proceeding, so nothing approaching a proper research paper. Not clear if it's been peer-reviewed. The sample size is small. The stats are unsuitable for the design (should have used a mixed-effects model). The analysis is not reported properly (no F-statistic or degrees of freedom for the ANOVA, no indication what the error bars on the graphs are). Assuming the stats *are* appropriate, the p-value for the ANOVA for lactate at pmax is 0.049 so it juuuuust scrapes into statistical significance. More importantly, the specific comparison between pmax for placebo and treatment is non-significant on the Tukey test (p=0.062). This means that there is a suggestion that there is an effect on lactate but it's not really possible to draw much of a conclusion - hardly the "off the scale" response the marketing suggests. Even more important than that, ***they don't even report the data for actual performance*** - if they had data indicating (for example) higher wattage at AT or higher endurance for the treatment group you can bet they'd have included it. My conclusion from reading this is that it indicates a possible but very badly supported effect on lactate at exhaustion only (not at the other times they tested) and nothing else. Bear in mind that it's become clear that lactate is only a small part of the biology of fatigue, and is possibly even unrelated to it, so a product which claims to buffer pH changes because of lactate can only have a small effect at best."

That's quite technical in parts, but I think you get the message. This piece of research tells us nothing about whether it works in the manner claimed or not. People might swear blind to you that it makes a difference to them, but bear in mind that we are terrible at making such assessments about our own health and performance - there are also people who will tell you with complete conviction that the zaniest forms of alternative therapy cured their health problems, and just take a look at all those cyclists in the Tour de France making themselves look silly with breath-rite strips on their noses, a product that was shown years ago to have no effect on performance at all.