The Placebo That Knew It Was a Placebo

In 2010 a team led by Ted Kaptchuk of Harvard Medical School gave eighty patients with irritable bowel syndrome either a bottle of pills or nothing but the same attention from the staff. The bottle was labelled placebo. The patients were told plainly that the pills contained no medicine, that they were like sugar pills, and that such pills had been reported to help through the link between mind and body. After three weeks, 59 percent of those taking the pills said that their symptoms had been adequately relieved, against 35 percent of those who received no pills.

The result should not have happened. A placebo has always been understood as a kind of deception, a pill that helps only because the patient takes it for real medicine. The word is Latin for will please, and it entered the medical dictionaries in the eighteenth century for pills that a doctor gave to patients he could not otherwise help.

Kaptchuk's trial cracked that assumption. If the sugar worked without deceit, then belief in the pill's chemistry was not what produced the benefit. The candidates for what did are few. One is conditioning. A patient who has swallowed pills for years and felt better afterwards may respond to the act of swallowing as Pavlov's dogs responded to a bell. Another is ritual and care, since a pill arrives with a clinic, a bottle, a schedule and a person who takes the complaint seriously.

If the sugar worked without deceit, then belief in the pill's chemistry was not what produced the benefit.

Sceptics note that the trial was small, that every patient knew what was being given, and that the outcome was reported by the patients themselves. A patient who has been told that such pills help may say that she feels better in order to please the clinician, which is the old placebo again in a new costume. The objection is fair and only partly answered. Later trials have reported similar benefits in chronic back pain and in the fatigue that cancer survivors report, though most of these are small too.

The wider puzzle is what it means to be treated. Medicine has long assumed that the molecule heals and the setting merely surrounds it. Yet the setting seems to carry a force of its own, and that force survives even when the patient understands the trick. If a sugar pill that announces itself can still quiet a gut, then the question of what healed the patient cannot be answered by listing the ingredients, and nobody has yet said what the patient was persuaded by.