It sounds more like a question for philosophy than for medicine: if you know a placebo is a placebo, is it still a placebo?

Definitions of placebo often involve it being a ‘sham’ or a ‘dummy’ treatment. But a Boston-based team has been trying to see if they can elicit a placebo effect when their patients are told quite openly they are getting nothing more than an inert pill.

Ted Kaptchuk, of Harvard Medical School, Massachusetts, and his colleagues gave 80 people with irritable bowel syndrome (IBS) either placebo pills or no treatment at all. All patients had the same “warm supportive patient-practitioner relationship”, they report in PLoS ONE.

The results are pretty ambiguous. There seemed to be a difference in reported ‘global improvement scores’. (This asks: “Compared to the way you felt before you entered the study, have your IBS symptoms over the past 7 days been: 1) Substantially Worse, 2) Moderately Worse, 3) Slightly Worse, 4) No Change, 5) Slightly Improved, 6) Moderately Improved or 7) Substantially Improved.”)

Those given placebos reported a 21-day improvement score of 5.0 (plus or minus 1.5). Those denied the chance to swallow nothing of medical significance reported improvement scores of 3.9 (plus or minus 1.3). Obviously there’s some potential for overlap in those figures.

“Nevertheless,” says Kaptchuk, “these findings suggest that rather than mere positive thinking, there may be significant benefit to the very performance of medical ritual.”

He cites the fact that 22 of the 37 people in the placebo arm reported adequate symptom relief at the end of the trial vs 15 in the 43-person control group.

An informed placebo effect could be useful in practice. Lying to your patients is an ethical no-no for most doctors and this has limited the use of true placebos. (Well, that’s the theory – there is evidence that some doctors are quite happy to prescribe placebos; not to mention the fact that handing out antibiotics like sweets for pestering patients with viral infections is not unknown.)

If you could give a patients a placebo, tell them it was a placebo, and still show that was a useful intervention it could be a godsend for doctors.

The use of placebos is an area of keen interest, and one that is still deeply mysterious. Consider, for instance, that as this recent Nature Medicine piece points out, “Red placebo pills are more likely to act as stimulants compared with blue placebo pills, because red is interpreted as ‘hot’ and ‘danger’. More expensive placebo treatments produce significantly more placebo analgesia than less expensive ones.”

There is, however, an even deeper question related to this study: were the placebos that people were told were placebos really placebos?

Those given pills got gelatine capsules filled with inert ingredient Avicel, a form of cellulose. They were told it was a placebo. But they were also told that “placebo pills have been shown in clinical studies to produce significant improvement in IBS symptoms through mind-body self-healing processes”.

So they were given placebos, that they were told were placeboes, but which they were also told might help their condition. Which brings us back to where we began: if you know a placebo is a placebo, is it still a placebo if you think it might work?

If all this is making your head hurt, try taking a sugar pill. It’s been shown to help.