Trick or Treatment (9 page)

Haygarth was suggesting that patients be treated with tractors made from Perkins’ special alloy and with fake tractors made of ordinary materials to see if there was any difference in outcome. The results of the trial, which was conducted in 1799 at Bath’s Mineral Water Hospital and Bristol Infirmary, were exactly as Haygarth had suspected – patients reported precisely the same benefits whether they were being treated with real or fake tractors. Some of the fake, yet effective, tractors were made of bone, slate and even painted tobacco pipes. None of these materials could conduct electricity, so the entire basis of Perkins’ tractors was undermined. Instead Haygarth proposed a new explanation for their apparent effectiveness, namely that ‘powerful influences upon diseases is produced by mere imagination’.

Haygarth argued that if a doctor could persuade a patient that a treatment would work, then this persuasion alone could cause an improvement in the patient’s condition – or it could at least convince the patient that there had been such an improvement. In one particular case, Haygarth used tractors to treat a woman with a locked elbow joint. Afterwards she claimed that her mobility had increased. In fact, close observation showed that the elbow was still locked and that the lady was compensating by increasing the twisting of her shoulder and wrist. In 1800 Haygarth published
Of the Imagination as a Cause and as a Cure of Disorders of the Body
, in which he argued that Perkins’ tractors were no more than quackery and that any benefit to the patient was psychological – medicine had started its investigation into what we today would call the
placebo effect
.

The word
placebo
is Latin for ‘I will please’, and it was used by writers such as Chaucer to describe insincere expressions that nevertheless can be consoling: ‘Flatterers are the devil’s chaplains that continually sing placebo.’ It was not until 1832 that
placebo
took on its specific medical meaning, namely an insincere or ineffective treatment that can nevertheless be consoling.

Importantly, Haygarth realized that the placebo effect is not restricted to entirely fake treatments, and he argued that it also has a role to play in the impact of genuine medicines. For example, although a patient will derive benefit from taking aspirin largely due to the pill’s biochemical effects, there can also be an added bonus benefit due to the placebo effect, which is a result of the patient’s confidence in the aspirin itself or confidence in the physician who prescribes it. In other words, a genuine medicine offers a benefit that is largely due to the medicine itself and partly due to the placebo effect, whereas a fake medicine offers a benefit that is entirely due to the placebo effect.

As the placebo effect arises out of the patient’s confidence in the treatment, Haygarth wondered about the factors that would increase that confidence and thereby maximize the power of the placebo. He concluded that, among other things, the doctor’s reputation, the cost of the treatment and its novelty could all boost the placebo effect. Many physicians throughout history have been quick to hype their reputations, link high cost with medical potency and emphasize the novelty of their cures, so perhaps they were already aware of the placebo effect. In fact, prior to Haygarth’s experiments, it seems certain that doctors had been secretly exploiting it for centuries. Nevertheless, Haygarth deserves credit for being the first to write about the placebo effect and bringing it out into the open.

Interest in the placebo effect grew over the course of the nineteenth century, but it was only in the 1940s that an American anaesthetist named Henry Beecher established a rigorous programme of research into its potential. Beecher’s own interest in the placebo effect was aroused towards the end of the Second World War, when a lack of morphine at a military field hospital forced him to try an extraordinary experiment. Rather than treating a wounded soldier without morphine, he injected saline into the patient and suggested to the soldier that he was receiving a powerful painkiller. To Beecher’s surprise, the patient relaxed immediately and showed no signs of pain, distress or shock. Moreover, when morphine supplies ran low again, the sly doctor discovered that he could repeatedly play this trick on patients. Extraordinarily, it seemed that the placebo effect could subdue even the most severe pains. After the war, Beecher established a major programme of research at Harvard Medical School, which subsequently inspired hundreds of other scientists around the world to explore the miraculous power of placebos.

As the twentieth century progressed, research into placebo responses threw up some rather shocking results. In particular, it soon became clear that some well-established treatments benefited patients largely because of the placebo effect. For example, in 1986 a study was conducted with patients who had undergone tooth extraction, and who then had their jaw massaged by an applicator generating ultrasound. These sound waves, whose frequency is too high to be heard, could apparently reduce post-operative swelling and pain. Unknown to the patients or the therapists, the researchers tampered with the apparatus so that there was no ultrasound during half of the sessions. Because nobody can hear ultrasound, those patients not receiving ultrasound did not suspect that anything was wrong. Astonishingly, patients described similar amounts of pain relief regardless of whether the ultrasound was on or off. It seemed that the effect of the ultrasound treatment was wholly or largely due to the placebo effect and had little to do with whether the equipment was working. Thinking back to Haygarth’s criteria for a good placebo, we can see that the ultrasound equipment fits the bill – dentists had promoted it as effective, it looked expensive and it was novel.

An even more startling example relates to an operation known as internal mammary ligation, which was used to relieve the pain of angina. The pain is caused by a lack of oxygen, which itself is caused by insufficient blood running through the narrowed coronary arteries. The surgery in question was supposed to tackle the problem by blocking the internal mammary artery in order to force more blood into the coronary arteries. Thousands of patients underwent the operation and afterwards stated that they suffered less pain and could endure higher levels of exercise. However, some cardiologists became sceptical, because autopsies on patients who eventually died revealed no signs of any extra blood flow through the remaining coronary arteries. If there was no significant improvement in blood flow, then what was causing the patients to improve? Could the relief of symptoms be due simply to the placebo effect? To find out, a cardiologist named Leonard Cobb conducted a trial in the late 1950s that today seems shocking.

Patients with angina were divided into two groups, one of which underwent the usual internal mammary ligation, while the other group received sham surgery; this means that an incision was made in the skin and the arteries were exposed, but no further surgery was conducted. It is important to point out that patients had no idea whether they had undergone the real or sham surgery, as the superficial scar was the same for both. Afterwards, roughly three-quarters of the patients in
both
groups reported significantly lower levels of pain, accompanied by higher exercise tolerance. Incredibly, because both real and sham operations were equally successful, then the surgery itself must have been ineffective and any benefit to the patient must have been induced by a powerful placebo effect. Indeed, the placebo effect was so great that it allowed patients in both groups to reduce their intake of medication.

Although this suggests that the placebo effect is a force for good, it is important to remember that it can have negative consequences. For example, imagine a patient who feels better because of a placebo response to an otherwise ineffective treatment – the underlying problem would still persist, and further treatment might still be necessary, but the temporarily improved patient is less likely to seek that treatment. In the case of mammary ligation, the underlying problem of narrowed arteries and lack of oxygen supply still existed in patients, so they were probably lulled into a false sense of security.

So far, it would be easy to think that the placebo effect is restricted to reducing the experience of pain, perhaps by increasing the patient’s pain threshold through placebo-induced will power. Such a view would underestimate the power and scope of the placebo effect, which works for a wide range of conditions, including insomnia, nausea and depression. In fact, scientists have observed real physiological changes in the body, suggesting that the placebo effect goes far beyond the patient’s mind by also impacting directly on physiology.

Because the placebo effect can be so dramatic, scientists have been keen to understand exactly how it influences a patient’s health. One theory is that it might be related to unconscious
conditioning
, otherwise known as the Pavlovian response, named after Ivan Pavlov. In the 1890s Pavlov noticed that dogs not only salivated at the sight of food, but also at the sight of the person who usually fed them. He considered that salivating at the sight of food was a natural or unconditioned response, but that salivating at the sight of the feeder was an unnatural or conditioned response, which existed only because the dog had come to associate the sight of the person who fed it with the provision of food. Pavlov wondered if he could create other conditioned responses, such as ringing a bell prior to the provision of food. Sure enough, after a while the conditioned dogs would salivate at the sound of the bell alone. The importance of this work is best reflected by the fact that Pavlov went on to win the Nobel Prize for Medicine in 1904.

Whilst such conditioned salivation might seem very different from the placebo effect on health, work by other Russian scientists then went on to show that even an animal’s immune response could be conditioned. Researchers worked with guinea pigs, which were known to develop a rash when injected with a certain mildly toxic substance. To see if the rash could be initiated through conditioning, they began lightly scratching the guinea pigs prior to giving an injection. Sure enough, they later discovered that merely scratching the skin and
not
giving the injection could stimulate the same redness and swelling. This was extraordinary – the guinea pig responded to scratching as if it were being injected with the toxin, simply because it had been conditioned to associate strongly the scratching with the consequences of the injection.

So, if the placebo effect in humans is also a conditioned response, then the explanation for its effectiveness would be that a patient simply associates getting better with, for example, seeing a doctor or taking a pill. After all, ever since childhood a patient will have visited a doctor, received a pill and then felt better. Hence, if a doctor prescribes a pill containing no active ingredient, a so-called sugar pill, then the patient might still experience a benefit due to conditioning.

Another explanation for the placebo effect is called the
expectation theory
. This theory holds that if we expect to benefit from a treatment, then we are more likely to do so. Whereas conditioning would exploit our unconscious minds to provoke a placebo response, the expectation theory suggests that our conscious mind might also be playing a role. The expectation theory is supported by a host of data from many lines of research, but it is still poorly understood. One possibility is that our expectations are somehow interacting with our body’s so-called
acute phase response
.

The acute phase response covers a range of bodily reactions, such as pain, swelling, fever, lethargy and loss of appetite. In short, the acute phase response is the umbrella term used to describe the body’s emergency defensive response to being injured. For instance, the reason that we experience pain is that our body is telling us that we have suffered an injury, and that we need to protect and nurture that part of the body. The experience of swelling is also for our own good, because it indicates an increased blood flow to the injured region, which will accelerate healing. The increased body temperature associated with fever will help kill invading bacteria and provide ideal conditions for our own immune cells. Similarly, lethargy aids recovery by encouraging us to get much-needed rest, and a loss of appetite encourages even more rest because we have suppressed the need to hunt for food. It is interesting to note that the placebo effect is particularly good at addressing issues such as pain, swelling, fever, lethargy and loss of appetite, so perhaps the placebo effect is partly the consequence of an innate ability to block the acute phase response at a fundamental level, possibly by the power of expectation.

The placebo effect may be linked to either conditioning or expectation or both, and there may be other even more important mechanisms that have yet to be identified or fully appreciated. While scientists strive to establish the scientific basis of the placebo effect, they have already been able to ascertain, by building on Haygarth’s early work, how to maximize it. It is known, for instance, that a drug administered by injection has a bigger placebo effect than the same drug taken in pill form, and that taking two pills provokes a greater placebo response than taking just one. More surprisingly, green pills have the strongest placebo effect on relieving anxiety, whereas yellow pills work best for depression. Moreover, a pill’s placebo effect is increased if it is given by a doctor wearing a white coat, but it is reduced if it is administered by a doctor wearing a T-shirt, and it is even less effective if given by a nurse. Large tablets offer a stronger placebo effect than small tablets…unless the tablets are very, very small. Not surprisingly, tablets in fancy branded packaging give a bigger placebo effect than those in plain packets.