The recent decision by NICE to no longer recommend acupuncture for lower back pain got me thinking about the placebo effect. It is a bizarre phenomenon: any treatment (regardless of whether it is a real treatment or not) will improve symptoms in some people simply because the recipient believes that it will work. So if we give someone a placebo (a sugar pill for example) and tell them that it can work for their illness, a proportion of patients will feel better. There are so many interesting things about the placebo effect it’s difficult to know where to start.
The more the placebo, the bigger the effect
Take this observation: if we give someone sugar pills as a placebo, then they might see improvement in their symptoms. However, if we give them four sugar pills instead of two, then we will see a bigger improvement. And if a placebo injection is used, then this is more effective again.
It matters what a placebo looks like
Sugar pills with a brand name stamped on them will have a stronger placebo effect than those that don’t. Researchers have also found that sugar pills with a $2.50 price tag ease pain much more effectively than identical pills with a10 cents tag. We actually see the same effect with wine; people rate a wine highly if they think it is more expensive.
You don’t necessarily have to deceive people
Bizarrely, there is still a placebo effect even when the patients know they are getting a placebo. When researchers told people who were on a trial for an irritable bowel syndrome that they were receiving “placebo pills made of an inert substance”, they found that patients still experienced a reduction in symptom severity, even though they knew they were on a placebo. In another study, researchers gave patients a fake pain-relieving gel for 4 days, and then told them the truth, that it was it was actually just dyed vaseline. However, the pain relieving effect still persisted after this, suggesting that conditioning is part of this effect.
The placebo effect is getting stronger with time
Researchers analysing 20 years of clinical trials for pain have noticed a weird pattern: the placebo effect has been increasing over that time (but only in the US). Several explanations for this have been suggested, including that patients are getting more attention and encouragement now compared to the past, so their overall experience in the trial is better. It has also been pointed out that over this time pharmaceutical companies have turned to private companies to run their trials. As these companies are paid to recruit more people, they may inadvertently include less ill patients on the trial in order to boost numbers. This has been shown to result in an apparent treatment effect in patients, contributing to the placebo effect.
There is a placebo effect on animals
This is a really confusing one: animals experience the placebo effect! If this effect is driven by the expectation that the treatment will work, how do we explain that? More about that below.
Placebos can cause side effects
When patients are aware of the side effects of a treatment, they can also experience these while on placebo. For example, if a patient has previously taken opioid pain relief (one related to opium, such as morphine, codeine or vicodin), which can result in respiratory depression, they are then more likely to experience the same symptoms when on a placebo. Furthermore, if someone has been on hormone replacement therapy (HRT) placebo and this is then discontinued, they can experience the same withdrawal symptoms as someone who has actually been on HRT.
And placebos aren’t just a medical thing
The placebo effect exists in other areas of life too. After drinking placebo vodka, people’s IQ drops and they have impaired judgement. This is demonstrated with predictably funny results in a video of a non-alcoholic keg party that you can find here.
The above examples all point to the fact that while the placebo itself is an inert substance, it can still result in real physiological effects through suggestion, expectation and other mental processes.
We really don’t have a good grasp on how the placebo effect works. We know that it usually affects subjective symptoms like pain, depression, hot flushes, and insomnia, but that it very rarely improves the underlying disease mechanisms. Survival from serious forms of cancer has little observable placebo effect for example. However, we do know that it is made up of a mix of different biological and psychological factors.
First and most obvious of these is that mood and belief can have a significant effect on subjective symptoms, both through our perception of those symptoms, but also through real biological mechanisms. When studying the placebo effect in pain research, scientists found that a drug called naloxone, which inhibits opioids, could also inhibit the placebo effect, suggesting that placebo and opioid pain relief share a common mechanism. Other scientists then used brain imaging to suggest that placebo and opioid pain relief both seem to work via the same pathways in the brain, showing that for pain relief at least, part of the placebo effect is caused by a real biological mechanism.
Such biological mechanisms are also at play in systems where mood really matters. The placebo effect can lower the release of stress hormones (through altering someone’s mood) and as such can have an effect on the heart and the immune system. The influence of a placebo in this context is mild, but almost certainly real.
These biological explanations are in the minority however; it is thought that most of the placebo effect is as a result of cognitive biases that we are all prone to. Before I get into that however, I will point out that, placebo or not, if someone feels better after taking something, then it may have legitimate use in medicine. I won’t address this issue here, but will in my next post on this blog. The use of alternative medicine falls into this category, and is an interesting topic.
Back to the placebo effect; the most common cognitive bias that falls into this category is called regression to the mean. The many so-called “cures” for the common cold (echinacea, hot lemon, etc.) are a great example of this. We can show that these none of these remedies have any effect above the placebo. However, when people get better they credit whatever treatment they have taken for this natural improvement in their symptoms. Often, by the time someone starts their herbal remedy (or whatever they decide to go with) their body has already dealt with the illness and they are on the mend. It is very natural for us to see cause and effect where there is none, but most improvements like this are as a result of the placebo effect, specifically regression to the mean.
The animal placebo effect mentioned above can be explained by another bias we are prone to called the caregivers effect. This occurs because animals cannot directly report their subjective symptoms, so people have to observe the animals and decide whether they are responding to the therapy. Often, the humans will perceive improvement even when objective measures show none.
In human trials, other biases occur. People taking a treatment want to get better and want the time and money they have invested to be worthwhile, so when they are asked to rate their symptoms, they tend to over-estimate their improvement. This is known as reporting bias, and also affects scientists, who want their treatments to work and will tend to over-estimate the positive effect in their trial.
It is also known that simply being in a clinical trial can contribute to the placebo effect. Participants tend to take better care of themselves than normal, and also get more medical attention than people not in trials. This causes improvements in people on the trial, not because of any biologically “real” placebo effect, but simply because they are part of the trial. This is known as the clinical trial effect.
The placebo effect is an incredibly fascinating thing. It shines a light on how much we do not understand about our bodies, but also on how much tricks of the mind can affect our perception of ourselves. I began this article by talking about acupuncture, but the role of the placebo effect in alternative medicine is not one I will address in this post. However, I will tackle this, and the question of whether it is ethical to prescribe a placebo, in my next piece on this topic.