Beyond the “Gold Standard”: Diet, Cancer Prevention, and the Randomized Clinical Trial

gold-barsLast week’s soy-breast cancer study out of China received a fair amount of press. But some of that coverage contained basic assumptions about the nature of cancer research that aren’t accurate.

Take this passage from Time.com:

…. the study was not a randomized clinical trial of soy consumption. That is, rather than randomly assigning breast-cancer survivors to consume or not consume various amounts of soy, then following those participants to see whether they developed recurrent tumors, the study looked retrospectively at women’s self-determined soy-eating habits.

So far so good.

But then came this next bit:

The randomized clinical trial is the gold standard upon which medical practice is determined, and the only kind of trial that gives scientists confidence that other variables are not confounding their results.

Yeah, that’s … not always true. Not when you’re studying something as complex as the human diet, and a disease that can take many years to develop, like cancer.

When it comes to studying diet, lifestyle and cancer prevention, the randomized clinical trial (RCT) is one tool investigators use, but it can’t – and shouldn’t – be considered the be-all and end-all, the “gold standard” in all situations.

After the jump:  The difference between studying cancer treatment and studying cancer prevention.

Consider the case of tobacco and lung cancer. No responsible doctor or researcher today would deny that smoking raises risk for lung cancer.  Yet there never has been, and never will be, an RCT that demonstrates that tobacco causes lung cancer.

There are many reasons for that:

Ethical: Scientists can’t randomly assign substances they suspect might by harmful to healthy subjects.

Logistical: In any RCT, researchers have to hope that the dose that they’ve chosen to assign to subjects is large enough to have an influence, and that they’ve given it enough time to have an effect. That’s difficult, because ….

Biological: Cancer can take years, even decades, to develop.

Despite the fact that the smoking-cancer link has never been demonstrated by an RCT, we know that link exists.

We know it exists because it emerges from the huge amount of evidence from many different kinds of study involving tobacco and lung cancer – investigations that have taken place in clinics, in animal models and in cell cultures.  When all of this evidence is examined collectively and systematically, the findings point to the same conclusion.

What an RCT is For … And What it Isn’t

The RCT is a powerful research design, but it was created to study very specific substances (like cancer drugs) for their very specific effects (like halting tumor growth) in very specific populations (like cancer patients) for a very specific period of time.

That makes it perfect to study the role of an isolated nutrient, or the effects of a new cancer treatment pill.

But the human diet doesn’t consist of isolated nutrients and pills. Think of those women in that Chinese study, who ate whole soy foods as part of their overall diet, throughout their lifetimes. The RCT can’t address itself to that situation, but it can get close.

If someone were now to conduct an RCT that, for example, examined the effect of a specific soy component among a select group of breast cancer survivors, it would be useful.  Those findings would add to the scientific discussion.  But they wouldn’t bring that discussion to a close – no single study, no matter what its design, could do that.

The thing to keep in mind whenever reading about research on diet, lifestyle and cancer prevention is this:

There is no “gold standard.”

It’s only by looking systematically at the accumulated evidence from many different kinds of study, and noting the agreement among them, that experts can determine which factors raise cancer risk, and which factors protect against it.

That’s what our expert report did. It’s why we know what we know about the steps people can take to lower their risk.

Read More:


Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>