They had just become accustomed to being the new stars of the western food industry (and hair and skin care industries) when along comes turmeric. While last week expansion and productivity increases seemed the obvious next step, complimented by artisan coconut plantations pursuing a sideline in eco-tourism, this week those same farmers are frantically consulting Agribusiness for Dummies to see if they can grow turmeric, as middle class Westerners rush out to buy it in products that range from pills to porridge and shampoo to enemas, all while looking to book holidays in eco-lodges on turmeric plantations.
Surely, it can only be a matter of time before Mark Lynas pops up on the Today programme promoting Rothamsted scientists, who will claim to have managed to splice turmeric genes into the humble British sheep, thereby curing cancer for people who eat lamb chops and adding a much needed splash of colour to the Yorkshire Dales and Lakeland fells.
Is turmeric just another in the long-line of food industry-created superfood fads, or is there something to it? Well, a medically trained journalist at the Beeb, Michael Mosley, wanted to answer this question, so, naturally, he organised a trial and brought in some experts.
Could Turmeric Really Boost Your Health?
Bold health claims have been made for the power of turmeric. Is there anything in them, asks Michael Mosley.
So, what of the article and trial?
The trial was, even for a small trial undertaken at the request of a journalist, badly constructed. The 'placebo' control was only a control for those taking the supplement, when there really needed to be a 'placebo' control for those taking the turmeric as part of their diet. A decent trial would have had a randomised group given inert yellow powder and told to use it in cooking and food prep, ideally this should have a similar bitter taste to that associated with turmeric. What we usually refer to as the placebo response is, as Daniel Moerman pointed out in his now classic text on the topic: a meaning response. To control for a meaning response and isolate it you need to replicate the meaning of the practice of which the administration of the apparently effective treatment is part, and that means that the way the control group are talked to needs to be the same. This means understanding what that practice means for the participants and replicating that meaning with the inert turmeric substitute. Having failed to do this in the trial they fell short of gaining warrant for the claims made about turmeric. As things stand, we cannot, with this trial, say whether it was the turmeric, a meaning response ('placebo'), or a healthy diet that is responsible for the results (or, as often transpires to be the case, some combination of all three). Consider one example, perhaps pertinent here: there seems to be reason to believe that diary-keeping in relation to diet brings health benefits: people often eat healthier food, portion better and eat at more regular intervals and optimal times, and, of course, doing so has meaning in our culture.
What is also of interest here is that once again supplements seem to fail. For whatever the flaws in this small trial, the supplements did not produce the outcome that they were looking for. It does seem to be the case in many examples of foodstuffs showing some sort of medicinal or significant health benefit that this is rarely replicated when consumed as supplements.
There are obviously lots of possible explanations for this, some of which might cast doubt on the claims made for the particular foodstuff, others which might point to factors such as that mentioned here in this article: that other things help activate or aid absorption of the active element and those things aren't present when it is consumed in supplement form. Sometimes, however, it seems to point to what we believed were the effects of a wonder-food or wonder-ingredient actually being the effects of good diet and/or a meaning response.
Take two examples: the claims made by the food and supplement industry for antioxidants and Omega 3 fish oil. When the claims made for antioxidants and Omega 3 fish oil are examined alongside the RCT data, any of the claimed are notoriously difficult to isolate from the benefits of a good, balanced diet, for example. Just take a look at the trial data on PubMed. Indeed, this has been the case for over ten years now, yet one still finds anti-oxidants talked about in a way that implies that it is uncontroversial and established that eating food which is high in antioxidants is not only beneficial to health, but staves off cancer and Alzheimers, and so on. Well, while the food industry might like you to think that is the case, the evidence remains inconclusive at best regarding any benefits of nutritional antioxidants.
When we design trials we need to be more inventive with the controls, having understood the nature of meaning responses (placebo), and only then will we have a better understanding of what is really going on. It is, as I've already noted, notoriously difficult.
So far things might look hopeful with turmeric, but bad science is driven by hope (and, in this context, by food marketing) which makes us jump the gun and declare something a superfood. While fads might seem harmless, they can kill, either because people stop taking proven treatment in favour of the latest fad product promoted on their Facebook news feed or because, for example, people damage their liver by ingesting large quantities of something because it is 'good' and 'natural', but which has never been trialled in that form and quantity.
I'll end with an anecdote. I once discovered I had a sensitivity to tea tree oil. I had a mouth wash which contained it and I developed blisters in my mouth when I used this mouthwash. Over the space of a couple of years, and by process of elimination (using different mouth washes, cutting out mouthwash altogether and using tea tree essential oil in water as a mouth rinse) everything pointed to the tea tree oil as the cause of the oral blisters. I then did some investigating and found that there were studies that had indeed demonstrated that for a minority this was a known reaction (at the time, 10 years ago, it was stated as 14% in studies that I consulted, if I recall correctly). I reported my findings to a friend, whose response summed up for me one of the problems: "No", she said, "tea tree oil is good. It can't be that!"
The thing is, what's good in certain doses might well be an irritant or even poisonous in higher doses; what is good for some people, might not be for others; what makes you feel good in the short term might be slowly leading to liver failure in the long term. Whether what you are taking was recommended by the nice guy at the health food co-op, and comes from an organically grown tree, or was prescribed by a medically trained doctor with a poor bedside manner and comes from the labs of Big Pharma, what you need to form a judgement is data from well-designed and well-conducted trials.
Whether or not a trial is well designed is not merely about randomisation, blinding, and sample size, it also depends on the designers understanding the nature of the meaning response and factoring this into the design of the trials. Simply administering a dummy pill and ticking-off the placebo control box on the checklist is a fail, and understanding the meaning response is to understand why it is so.