A novel randomized trial taking on a vexing issue around one of the world’s most commonly prescribed medications has concluded that frequently intolerable statin side effects, such as muscle weakness or pain, are almost entirely a nocebo effect, the placebo effect’s darker cousin.
The many patients who report such symptoms while taking statins are indeed probably feeling them, but they are a result of taking the pills rather than any pharmacologic effects, conclude researchers based on their 60-patient study, Self-Assessment Method for Statin Side-effects or Nocebo (SAMSON).
“SAMSON leaves no doubt that patients really do get side effects from statin tablets, but what it shows us is that 90% of this symptomatic burden is elicited by placebo tablets too,” said James P. Howard, MB, PhD, Imperial College London, United Kingdom, when presenting the results November 15 during the American Heart Association (AHA) Scientific Sessions 2020 virtual meeting. They were published simultaneously in the New England Journal of Medicine.
Studies have shown that in practice, “more than half of patients abandon statins completely within 2 years. And yet, in placebo-controlled trials, no more people stop statins than placebo,” Howard said. “The most important message from SAMSON is that side effects from statin tablets are very real, but they are mainly caused by the act of taking the tablets, not by the statin that is contained within them.”
Patients in the trial, all of whom had a history of dropping statins because of side effects, each took atorvastatin 20 mg/day, a placebo, or neither pill for 1 month, alternating the regimens in randomized order over 1 year so that each was followed a total of 4 months. They used a smartphone app to record the severity of any side effects, not necessarily just pain, on a scale of 0 to 100.
Symptom intensity scores averaged 16.3 for atorvastatin and 15.4 for placebo, for a nonsignificant difference, but only 8.0 for no-pill months (P < .001 compared to the statin or placebo).
Because such symptoms seem to be based on patient expectations from statin therapy, positive communication about what the drugs can achieve and how the next treatment steps are described can play a big role in their continued use.
For example, “Changing them to another statin is a very reasonable thing to do, but as soon as you start trying people on lower doses and working up, you’re sort of telling them that you’re expecting at some dose that they are going to get side effects,” cautioned Howard at a media briefing on SAMSON.
“The most important thing is to explain the evidence, and what our expectations are, maybe be a bit more optimistic about statins, and tell them they’re very unlikely to suffer from side effects,” he explained, “because the nocebo effect can only really rear its head if the patients are expecting to feel worse — just like the placebo effect will only work if people are expecting to feel better.”
Amit Khera, MD, who moderated the media briefing, said he always tells such patients, “Yes, 1 in 10 patients report having muscle ache. But first and foremost, 9 in 10 don’t. The vast majority of patients don’t get muscle aches. I think that’s really an important part of the communication.”
Now, after SAMSON, “I have an additional point that I’m going to tell them: out of the patients that get muscle aches, probably 90% of that is the anticipation of getting the statin, the nocebo effect,” said Khera, who directs the preventive cardiology program at UT Southwestern Medical Center, Dallas, Texas.
In practice, however, many patients who report adverse statin effects do so later than 2 weeks after starting therapy, “so these findings cannot be generalized to them,” proposed Francine K. Welty, MD, PhD, Beth Israel Deaconess Medical Center, Boston, Massachusetts, as the invited discussant after Howard’s presentation.
All 60 patients recruited for SAMSON had previously stopped taking a statin because of side effects that arose within 2 weeks of their first dose. That requirement was intended to boost chances that any further symptoms during the trial would arise within a month of starting each new round of pills, Howard said.
So the trial’s results, Welty said, “are limited to those subjects who develop symptoms within 2 weeks of starting a statin.”
Including only such patients may have created bias toward a nocebo effect, she said, because “non–drug-related side effects of medications are often greatest during the initial weeks of treatment and tend to abate over time.” For example, “Metformin causes diarrhea and β-blockers cause fatigue, but subjects do adapt and generally tolerate them very well.”
The patients, 25 women and 35 men, 90% of whom were white, received four pill bottles, each with a month’s supply of atorvastatin, four bottles each with 1 month of placebo, and four empty bottles each, to be used double-blind for a month in randomized order.
Patients used the smartphone app to document their symptom scores, which ranged from 0 for no symptoms to 100 for symptoms that were the “worst imaginable,” the published report notes. Patients who experienced symptoms so severe as to be intolerable could stop the 1-month regimen they were then following, with instructions to resume the regimens in order starting the next month.
Eleven patients were unable to complete all 12 one-month segments of the trial.
The study’s overall “nocebo ratio” of 0.90 was calculated as the difference between symptom intensity scores on placebo and on no treatment divided by the difference between symptom intensity on the statin and on no treatment. The interpretation: 90% of the symptom burden felt by patients receiving atorvastatin was also felt during placebo use.
Thirty of the patients, contacted 6 months after the trial concluded, had resumed taking a statin, while “4 planned to do so and one could not be contacted,” the report notes. The 25 other patients weren’t receiving a statin and had no plans to take one.
In an important part of the trial, Howard said, at its conclusion the patients were shown their pattern of symptoms in relation to whether they were taking the statin, placebo, or neither. “Participants could see as clearly as we could the surprisingly powerful magnitude of the nocebo effect. And this led to half of our patients happily restarting statins.”
The implications of SAMSON, Welty said, “are very important, in that those developing symptoms within 2 weeks of starting a statin should be reassured that approximately half will be able to successful restart the statin.”
SAMSON was funded by the British Heart Foundation. Howard had no disclosures. Welty discloses chairing the data safety monitoring committee for Empagliflozin International Clinical Trials, supported by Boehringer Ingelheim.
American Heart Association Scientific Sessions 2020. LBS.04. Presented November 15, 2020.
N Engl J Med. November 15, 2020. Abstract