Many people have good results from acupuncture treatment, but critics say that acupuncture is no more than a placebo. I want to explore some of the evidence and theories about placebos and see how it relates to acupuncture.
What is a placebo?
Placebo literally means “I will please”. It is where an inactive treatment (sugar pill, saline injection etc) is given to someone who believes they are being given an active treatment. If the person’s symptoms or condition improves, the improvement is considered to be a result of ‘the placebo effect’ (or they would have spontaneously improved anyway).
What can a placebo do?
In the later years of the Second World War, American anaesthetist Henry Beecher ran out of morphine at a field hospital as he was preparing to treat a wounded soldier. Out of desperation, one of the nursing staff used a saline injection instead and the patient was able to undergo surgery as if he had received the painkiller. Beecher went on to be one of the first to conduct research into the placebo effect.
In the 1950s an operation known as internal mammary ligation was used to treat angina patients. It involved blocking some of the arteries supplying the heart in the hope that new vessels would sprout to supply the heart muscle. When no evidence of new vessels was found in those who had received the operation, two teams of researchers (including Cobb et al, 1959) investigated. They found that patients who received a sham (fake) operation (they had their chests cut into and the arteries exposed but not ligated) responded at least as well as those who received the real surgical procedure. Other research has shown similar effects with some kinds of knee surgery (Moseley et al, 2002).
Over time, the placebo effect proved so ubiquitous that medical research required a placebo control (i.e. one group receives the real treatment while another group receives a placebo). More recently the placebo group has often been replaced by a group receiving the current best treatment due to ethical concerns.
How does a placebo work?
This is not entirely clear but there are a few theories. Some symptoms (such as pain) are subjective, so your expectations and feelings about the sensations you experience will determine to a certain extent the amount of pain you feel.
Nicholas Humphrey, the British Psychologist suggests an evolutionary theory of the placebo. He suggests that our bodies have a much greater capacity for self-repair than we generally use, however these repair mechanisms require a lot of resources (vitamins and minerals, white blood cells etc) to function. Our bodies conserve many of these resources in case of a protracted fight against illness and also create symptoms (such as pain and swelling) to prevent us doing further damage. When we receive medical help, there is a suggestion that we are on the road to recovery and can release these resources for healing. We might expect that when we are stressed we conserve our resources for a similar reason, then as we return to a safer space and can relax our bodies will begin to repair. Essentially, receiving treatment is like getting permission from an authority figure to spend the resources on healing ourselves, which we would not otherwise be able to do.
Dylan Evans in his book, Placebo: Mind over matter in modern medicine, identifies certain symptoms and conditions that tend to respond to placebo (while others don’t appear to). He theorises that the placebo response is linked to the ‘acute phase response‘, and is in fact predominantly about suppressing the immune response (which is slightly different from Nicholas Humphries’ idea that the body releases more resources) in order to relieve symptoms such as pain, swelling and nausea.
There are also interesting links between placebo and hypnosis. Under hypnosis, responsive subjects have been able to withstand sitting in an ice bath without feeling any discomfort (although they do respond physiologically to the cold), and there are people who have used hypnosis for pain relief during, for example, dental surgery. Furthermore, hypnosis requires that a subject accept certain suggestions in order to be effective. In essence, the placebo response is produced when someone accepts the suggestion that they will feel better.
What is the evidence for acupuncture being a placebo?
When acupuncture is subjected to placebo-controlled trials the results are often quite mixed. It is quite common for both real and sham acupuncture to produce a significant improvement. Real acupuncture is usually better than sham acupuncture but the difference between the two is sometimes quite small and not always statistically significant. This leads many to the conclusion that if there’s not much difference between real and sham acupuncture, it must be a placebo. One good example is the response to the research by Vickers et al, 2012 which I commented on in this post on Bias in Reporting Acupuncture Research.
Furthermore, many of the conditions for which there is good evidence for acupuncture being effective are the kinds of conditions that tend to respond well to placebos.
Finally, research by Bausell et al, 2005 found that patients who believed they were in the real acupuncture treatment group experienced greater pain reduction following dental surgery than those who believed they were in the sham treatment group, regardless of which group they were actually in. This suggests that expectation makes a big difference to the result of acupuncture treatment.
What is the evidence for acupuncture being more than a placebo?
Firstly, brain imaging studies reveal there are neurobiological differences between real acupuncture and sham acupuncture (for example, Pariente et al, 2005). In this study there were three groups: real acupuncture, sham acupuncture and overt placebo – the first two groups expected some treatment effect, while the third group did not. The PET scan revealed some areas of the brain were activated in both real and sham conditions that were not activated in the overt placebo group, but interestingly there were other areas activated only in the real acupuncture condition.
Not only this, but there are also point-specific effects (Minagawa, Munenori et al, 2013, Choi, Jiang and Longhurst, 2012) where needling different acupuncture points has different effects on areas of brain activation as well as haemodynamic responses such as heart rate variability and blood pressure, and gastric electrical activity.
As an acupuncturist I commonly see people who have had a particular condition for many years and have tried many other types of treatment before having acupuncture. When acupuncture is successful in providing long term relief for their condition, it seems beyond coincidence that they spontaneously recover at the same time as receiving a treatment after so many years. They have also already had other treatments that provide the same expectation of recovery and the associated placebo effect. This is obviously not particularly robust evidence, but these cases do suggest some kind of mechanism beyond expectation (they’ve already had other treatments), spontaneous remission (they’ve had it for such a long time spontaneous remission is unlikely) and regression to the mean (they don’t just get temporary improvement, they are better for the long term).
Finally, returning to the clinical trials (taking the study by Vickers et al as an example), there is a statistically significant difference between real and sham acupuncture. Sham acupuncture does appear to be more effective than some other kinds of placebos though. Why might this be? Well, sham acupuncture may not actually be completely inactive – it still involves some stimulation of the skin, and in some cases is just acupuncture performed in different locations. Secondly, not all placebos are equally effective, and there could be something about acupuncture that makes it a particularly good placebo.
Not all placebos are equal
Research suggests that placebo injections work better than placebo pills, and that placebo surgery works better than either. Discounted placebos are not as effective as full price ones and sham acupuncture has been found to be better than inert pills. Why is this? The simple answer is expectation. The more powerful you believe a treatment to be, the more effective a placebo it will be. It may not be as simple as this though. There’s a theory in psychology known as cognitive dissonance, where holding two opposing viewpoints creates mental stress, and we try to reduce the inconsistency. So if we are undergoing a more painful therapy, we might tell ourselves that it must be more helpful (“no pain, no gain”) because otherwise we wouldn’t be doing it. Similarly, why would we spend more money unless it was more effective?
There are also other aspects to acupuncture treatment that enhance an expectation that it will be helpful. A first acupuncture session is usually quite long and involves detailed questioning of every aspect of your health, as well as some unusual diagnostic methods such as pulse and tongue diagnosis. This shows that the acupuncturist is very interested in your health and wellbeing and demonstrates their expertise.
Acupuncture is also usually quite relaxing. The treatment room is usually much more comfortable than a doctor’s office or hospital, there may be soothing music, softer lighting. You may get a massage or at the least physical contact in the form of pulse diagnosis and during the needling procedure. You then usually get to lie down and relax for around 20 minutes. This all combines to create a much more relaxed state. Returning to the evolutionary theories of placebo, being in a safe and relaxed state should allow you to better free up your healing resources. Thinking of how placebo may be linked with hypnosis and suggestion, relaxation is an important part of most hypnotic induction techniques, which increases people’s response to suggestions. A 20 minute rest (or snooze!) during the day has all sorts of benefits as well. Short naps have been shown to improve memory, performance and wakefulness, and lying down allows the spine to lengthen as the intervertebral discs are not being compressed.
Acupuncture is also quite an exotic ritual, which can enhance the placebo effect for some people and make it much weaker for others, depending whether they find the treatment credible or not.
I think the effectiveness of a placebo can be defined by the following formula:
Placebo response = nature of suggestion presented x degree of belief in suggestion x ability of the body to respond
Suggestion presented: The effect on the body is determined by the nature of the suggestion – if a positive change in symptoms is suggested, the result will be a placebo, if it is a negative suggestion, the result will be a nocebo (placebo’s evil twin). If you expect mild improvement, that is what you will usually see, whereas if you expect full resolution of a problem, that is much more likely.
Belief in suggestion: This part of the formula accounts for much of the individual variability of the placebo response. The more you believe what is being suggested to you, the stronger the effect will be. If you are generally quite a skeptical person, you will be less prone to benefitting from the placebo effect. However, those interventions that you do find credible (probably most conventional medical treatments) will generate a stronger placebo response.
The nature of the suggestion together with the degree of belief in the suggestion could be referred to as expectation.
Ability of the body to respond: This seems to depend on a few factors. Firstly, the change has to be something that a human body is capable of doing. There are certain things a body isn’t capable of fixing – severed spinal cords, (most) removed organs etc. Then the body must have sufficient resources to make the necessary changes. Being well nourished, relaxed and well rested will help, where dehydration, malnutrition and tiredness will make it more difficult for the body to repair itself.
So, is acupuncture a placebo or isn’t it?
The research suggests that acupuncture has a powerful placebo element (for the reasons listed above), but also has a range of ‘specific effects’, including point-specific effects on the brain and vascular system. While biomedicine has yet to clarify the mechanisms by which acupuncture works, it does work better than pure placebo. Those who are skeptical and don’t expect any improvement won’t necessarily benefit from the ‘placebo’ part of the treatment, but I’ve seen plenty of skeptics get great results from treatment nonetheless.