Subscribe to the OSS Weekly Newsletter!

Register for the OSS 25th Anniversary Event

Placebos are harmless, right? Not always

The 'nocebo' effect: If placebos can reduce pain, they could just as easily increase it under certain circumstances.

This article was originally posted in the Montreal Gazette.

My last column on the placebo effect generated the same comment from two people. They suggested the next article should be on nocebos — that is, placebos that cause harm.

While most people associate the placebo effect with an improvement in symptoms, there is no reason why it couldn’t or shouldn’t possibly make things worse. If the placebo response can reduce pain symptoms, it could just as easily increase pain under certain circumstances.

One common and particularly challenging example is the nocebo effect of statins, the commonly prescribed cholesterol medication. 

Statins are lightning rods for controversy, with muscle pain being the often cited reason for why they are stopped or never started. Statins have been shown to cause muscle damage (rhabdomyolysis) in about one in 10,000 people. And while rare, it can be life threatening.

Where things get interesting is when people have pain but no evidence of muscle damage on blood tests. These pains, termed myalgias, are hard to study because the symptoms are by necessity subjective. They are fairly rare in study settings but in community and “real world” situations they are much more common.

The GAUSS-3 trial, which tested two cholesterol medications — ezetimibe and evolocumb — in statin intolerant patients, provided an interesting insight into why this paradox happens and how common true statin intolerance is.

Before testing these statin alternatives, researchers first had to establish that the participants were truly statin intolerant. They designed an initial placebo crossover trial where people got placebo followed by atorvastatin, or vice versa, to see if they got symptoms with only the drug but not the placebo.

Despite the fact that people had to report an intolerance to two or three different statins before enrolling in the trial, when tested in this double-blind crossover fashion over 60 per cent of these patients did not have true statin intolerance — that is, symptoms with the drug but not the placebo. More than half the people who thought they were getting symptoms from statins actually weren’t and couldn’t differentiate between the medication and the placebo.

The point was reinforced by the open label extension of the ASCOT-LAA trial. This 2003 trial was one of the first to establish that atorvastatin reduced cardiovascular events compared to placebo. Patients were blinded as to their medication assignment so they did not know if they were getting atorvastatin or placebo. During the trial, rates of muscle-related side-effects did not differ between the statin and placebo groups and occurred in about two per cent of patients.

After the trial ended and patients knew if they were getting the medication or not, an interesting thing happened. The rate of muscle-related side-effects diverged and dropped considerably in those not taking atorvastatin, down to one per cent, but remained significantly higher in those taking the medication. Just the possibility that patients might be on a statin seemed to induce more muscle pain, but telling them they weren’t made the pain go away.

The nocebo effect, like the placebo effect, is complex and powerful. Statins are just one example where the nocebo effect could lead someone to falsely attribute their symptoms to a medication.

This matters because when patients stop their medications mortality goes up. Contrary to what some people like to claim, cholesterol does matter for cardiovascular disease. But negative press about statins is linked to people stopping their medication, which then leads to more heart attacks.

Fear of side-effects is the most common reason people refuse or stop statins and that fear is used by people who want to advocate against their use. But the science is not on their side. As with all medications, small numbers of people get side-effects with statins. But these side-effects are rare and may not be due to the medication. They might be due to the nocebo effect.


Back to top