The 2010 study that changed my mind about placebos

I used to think "placebo effect" meant imaginary effect. Lots of people still do. The phrase has been worn smooth by decades of pop-science use, until "it's just a placebo" became a polite way to say "you were fooling yourself."

Then in 2010, a team at Harvard led by Ted Kaptchuk did something that should have ended that framing and mostly didn't.

They took 80 patients with irritable bowel syndrome and split them in half. One group got nothing — standard care, no intervention. The other group got a bottle of pills. The bottle was clearly labeled "placebo pills". The patients were told, explicitly, that the pills contained no active ingredient, were made of inert substance, and were not expected to work pharmacologically. The researchers also told them — and this is the key part — that placebos have been shown in clinical studies to produce significant improvement through mind-body self-healing. Take them twice a day. We'll see you in three weeks.

Three weeks later, the placebo group reported meaningfully greater symptom relief than the no-treatment group. The numbers were comparable to what you'd see from one of the standard IBS drugs. The patients knew. It worked anyway.

The paper is called Placebos without Deception, it's in PLOS ONE, and you can read the whole thing in about twenty minutes. I'll link it at the bottom. I'd encourage you to actually read it rather than take my word for any of this — it's short, the methods section is readable, and the discussion is more cautious than the headlines suggested at the time.

Here is what the study actually proved:

Telling people a pill is inert does not automatically neutralize its effect.
The framing around the pill — "this has been shown to help via mind-body mechanisms" — matters.
The daily ritual of taking something, in context, with an expectation, produces measurable change on subjective symptom scales.

Here is what the study did not prove:

That placebos cure anything. They don't.
That this generalizes to every condition. The follow-up literature is more mixed — pain, fatigue, and certain functional symptoms respond well; objective biological markers, much less so.
That Mycebo will do anything for you. Mycebo wasn't in the study. Our capsules weren't in the study. The study tested a mechanism — belief, ritual, expectation, context — that we happen to be designing a product around. That's the honest framing. We are not selling outcomes. We are selling the mechanism, in bottle form, with the instructions written down.

Kaptchuk's IBS trial wasn't a one-off, either. Cláudia Carvalho replicated the open-label design in chronic low back pain in 2016 — same direction, significant effect. Slavenka Kam-Hansen's migraine crossover study in 2014 found that the label on the pill accounted for roughly half of the total effect, independent of whether the pill was a real drug. Dan Ariely's 2008 JAMA paper showed price alone changes how well a placebo works. John Kelley's 2012 work suggested the practitioner relationship — warmth, attention, ritual context — amplifies the response further.

Five studies. Different conditions, different teams, same direction. A mechanism that keeps showing up across two decades of careful, unsexy research.

I am not a researcher. I'm a guy who read these papers and decided that if this is real — and it really does seem to be real — then somebody should build a product that takes the mechanism seriously instead of pretending it isn't the whole game. So we did.

Read the studies yourself. Decide what you think. The links live on our research page, and the full Kaptchuk paper is here. Next post: the question I get asked most, which is then why are you charging twenty-five dollars for it?

— Anton