Clinical trials short changing us?

By Stephen Daniells

Yet again major randomised clinical trials report that vitamin and mineral supplements don’t work to prevent disease, so should we give up and just eat chips and chocolate?

No! The studies are important, no one is questioning that, but their true value is as a way of enforcing that perhaps randomised clinical trials (RCTs) are not suitable as a stand-alone way of determining if nutrients are beneficial or not.

We are, after all, talking about trials that, in essence, pull a nutrient out of context and follow the same methodology as used for the testing of drugs.

But let's not forget that by following the drug model we are supplementing the diet with one or two nutrients, each at a single dose, for a set period of time. RCTs work by randomly assigning a group of volunteers to receive an active compound, be it a drug or nutrient(s), or a non-active comparison, be it an inactive form of the active compound or a placebo.

For food items that do not normally form part of the usual food chain, RCTs are the best of the best because such compounds can be tested and retested successfully. Probiotics, for example, could fit into this category. But micronutrients that are found in a wide array of foods, in different food matrices, and in (synergistic?) combinations with other nutrients, do not ally themselves to a form of study that isolates one at a fixed dose for a set period of time.

This week’s publication of the results of the Selenium and Vitamin E Cancer Prevention Trial (SELECT) and the Physicians' Health Study II (PHSII) in the Journal of the American Medical Association, are prime examples of the dubious application of the drug model to nutrients.

Guess what the results showed: No effect.

Cue the media headlines, such as the BBC’s “Vitamins 'do not cut cancer risk'”. This is an over-simplification of the situation, and a damaging one. Will consumers read this and can their vitamins as ineffective? (Questions over the quality of the journalism should be put on hold for another day.)

I am no expert, and I don’t pretend to be, but it amazes me time and again when people pump millions of dollars, pounds, or euros into studies that pull nutrients out of context.

The same questions jump to my mind all the time: Where’s the control group? Are the people in the placebo group actually taking supplements on the side? How long is the latency period for the disease in question?

Let’s address these one by one: For many nutritional studies a true control group doesn’t really exist. You cannot remove vitamin C, for example, totally from someone’s diet. It’s both unethical and practically impossible.

The PHSII study authors admit that the population was well-nourished, so could we ever have expected to see an effect? They also admit that this only applies to one dose. So is a study with one dose of one nutrient in a well-nourished population enough to conclude that selenium, vitamin E, and/or vitamin C are ineffective against prostate cancer?

Putting it a different way, ultimately we’re comparing someone with a good diet, with someone with a good diet plus a little extra vitamin C. What if people in the control group had a real love of kiwi fruit?

Next up is the question of whether the placebo group participants are also taking supplements on the side. This may sound silly, but there are plenty of examples in the literature where we see this. After all, if people believe calcium supplements are good for them, then you can’t stop someone from taking them. This is exactly what we saw with the WHI study a few years back that showed no differences between people in the calcium plus vitamin D group, and the placebo group. Turns out many of the participants in the placebo group were also taking calcium pills. Oops!

And then the question of latency is critical. If it can take upwards of 10 to 15 years for a disease to develop, as is the case with colorectal cancer, for example, and the study participants are free of the disease at baseline, then can we be hugely surprised or disappointed if we see no benefits after six years of supplementation? I think not.

But don’t think that just because we are aware of these issues and the limitations that this will have any impact on the view of RCTs.

Let’s look at Europe: the European Food Safety Authority (EFSA) is pinning all the health claims regulations on data from the ‘gold standard’ randomised clinical trial. The message is clear: If you have no RCT data, don’t even think about applying for a health claim.

So where does that leave us?

I would say we are at an important moment in time, and things may be moving in the right direction. At the recent Supply Side West, Professor Jeffrey Blumberg gave an excellent presentation about the over-reliance on evidence based nutrition, based on evidence based medicine, and called for science to be taken in its entirety, including in vitro, animal, and epidemiological studies.

It will be a long way to achieve the paradigm shift but there is hope. Until then, recommendations remain to limit the chips and chocolate.

Stephen Daniells is the science editor for NutraIngredients.com and FoodNavigator.com. He has a PhD in chemistry from Queen's University Belfast and has worked in research in the Netherlands and France.