what’s the science behind the placebo effect?
In: 9
Oh hey, I did my thesis on this topic!
The ELI5 is that placebos don’t work for everything but can be quite powerful for particular conditions or people. They usually work because of expectations (I believe I will feel better) or learned association (I take a placebo with an active medication repeatedly, and then later the placebo alone can make me feel better). This is also similar for the ‘nocebo’ effect where I might feel worse or get bad side effects to taking something for the same reasons.
A bit more beyond the ELI5:
Placebos can have different effects: subjective changes in perception of symptoms (e.g. nothing measurably changes but I feel better anyway), or objectively measurable physical effects.
As mentioned above there are two main hypothesised mechanisms of action: expectation and conditioning.
Expectation is what it sounds like: when we believe that we have been given an effective treatment, we may feel better or even have physiological changes. The evidence is good for subjective effects from verbal suggestion alone, but a bit spotty and inconsistent for objectively measurable effects.
Conditioning (classical conditioning in this case) is learning by association: e.g. the taste of coffee becomes associated with the caffeine hit through repeated exposure, so even decaf coffee or the first smell/taste before the caffeine is actually active can still sometimes make us feel more alert.
A striking example of a conditioned placebo response is from a few studies looking at conditioned immunosuppression. Basically people were given a weird drink and some active immunosuppressant drugs together for a while, then waited a few days so the active drug was out of the system. Then when given the weird drink without the drug, their blood samples showed immunosuppression as if they had taken the drug.*
*Goebel et al. (2002). Behavioral conditioning of immunosuppression is possible in humans.
Wirth et al. (2011). Repeated recall of learned immunosuppression: Evidence from rats and men.
Oh hey, I did my thesis on this topic!
The ELI5 is that placebos don’t work for everything but can be quite powerful for particular conditions or people. They usually work because of expectations (I believe I will feel better) or learned association (I take a placebo with an active medication repeatedly, and then later the placebo alone can make me feel better). This is also similar for the ‘nocebo’ effect where I might feel worse or get bad side effects to taking something for the same reasons.
A bit more beyond the ELI5:
Placebos can have different effects: subjective changes in perception of symptoms (e.g. nothing measurably changes but I feel better anyway), or objectively measurable physical effects.
As mentioned above there are two main hypothesised mechanisms of action: expectation and conditioning.
Expectation is what it sounds like: when we believe that we have been given an effective treatment, we may feel better or even have physiological changes. The evidence is good for subjective effects from verbal suggestion alone, but a bit spotty and inconsistent for objectively measurable effects.
Conditioning (classical conditioning in this case) is learning by association: e.g. the taste of coffee becomes associated with the caffeine hit through repeated exposure, so even decaf coffee or the first smell/taste before the caffeine is actually active can still sometimes make us feel more alert.
A striking example of a conditioned placebo response is from a few studies looking at conditioned immunosuppression. Basically people were given a weird drink and some active immunosuppressant drugs together for a while, then waited a few days so the active drug was out of the system. Then when given the weird drink without the drug, their blood samples showed immunosuppression as if they had taken the drug.*
*Goebel et al. (2002). Behavioral conditioning of immunosuppression is possible in humans.
Wirth et al. (2011). Repeated recall of learned immunosuppression: Evidence from rats and men.
[disclaimer: Not a neuroscientist, just an interested layman] One of the major theories to explain the placebo effect and many other observations is something called “predictive coding” or “predictive processing” [0]. The basic idea is something like this: Your brain receives perceptions from your senses, combined with a confidence level, and it then compares them with its predictions for what it expects to perceive (also with confidence levels). Now, if predictions and perceptions disagree, there are several options:
If the prediction has a low confidence and the perception a high one, the prediction is simply discarded and the brain just considers the the perception [1]. (Imagine you’re sitting in a cafe, your brain predicts (with low confidence) that there will be no movement visible in the corner of your eye, but if someone walks by, you take notice of it)
If the prediction has a high confidence and the perception a low one, then your brain goes with the prediction and discards the perception (you probably didn’t notice me using “the the” earlier, did you?) And lastly, if both conflict and have a high confidence, this triggers “surprisal”
This applies very well to the placebo effect: You have some noisy perceptions of pain, but your brain predicts that the pain should get better, because it has taken “medicine” and so the perceptions get overruled.
[0]: [https://www.cell.com/neuron/fulltext/S0896-6273(14)00192-5](https://www.cell.com/neuron/fulltext/S0896-6273(14)00192-5)
[1]: Technically, the part with lower confidence is not discarded, but both are combined in a way where the part with the higher confidence dominates the end result.
[disclaimer: Not a neuroscientist, just an interested layman] One of the major theories to explain the placebo effect and many other observations is something called “predictive coding” or “predictive processing” [0]. The basic idea is something like this: Your brain receives perceptions from your senses, combined with a confidence level, and it then compares them with its predictions for what it expects to perceive (also with confidence levels). Now, if predictions and perceptions disagree, there are several options:
If the prediction has a low confidence and the perception a high one, the prediction is simply discarded and the brain just considers the the perception [1]. (Imagine you’re sitting in a cafe, your brain predicts (with low confidence) that there will be no movement visible in the corner of your eye, but if someone walks by, you take notice of it)
If the prediction has a high confidence and the perception a low one, then your brain goes with the prediction and discards the perception (you probably didn’t notice me using “the the” earlier, did you?) And lastly, if both conflict and have a high confidence, this triggers “surprisal”
This applies very well to the placebo effect: You have some noisy perceptions of pain, but your brain predicts that the pain should get better, because it has taken “medicine” and so the perceptions get overruled.
[0]: [https://www.cell.com/neuron/fulltext/S0896-6273(14)00192-5](https://www.cell.com/neuron/fulltext/S0896-6273(14)00192-5)
[1]: Technically, the part with lower confidence is not discarded, but both are combined in a way where the part with the higher confidence dominates the end result.
[disclaimer: Not a neuroscientist, just an interested layman] One of the major theories to explain the placebo effect and many other observations is something called “predictive coding” or “predictive processing” [0]. The basic idea is something like this: Your brain receives perceptions from your senses, combined with a confidence level, and it then compares them with its predictions for what it expects to perceive (also with confidence levels). Now, if predictions and perceptions disagree, there are several options:
If the prediction has a low confidence and the perception a high one, the prediction is simply discarded and the brain just considers the the perception [1]. (Imagine you’re sitting in a cafe, your brain predicts (with low confidence) that there will be no movement visible in the corner of your eye, but if someone walks by, you take notice of it)
If the prediction has a high confidence and the perception a low one, then your brain goes with the prediction and discards the perception (you probably didn’t notice me using “the the” earlier, did you?) And lastly, if both conflict and have a high confidence, this triggers “surprisal”
This applies very well to the placebo effect: You have some noisy perceptions of pain, but your brain predicts that the pain should get better, because it has taken “medicine” and so the perceptions get overruled.
[0]: [https://www.cell.com/neuron/fulltext/S0896-6273(14)00192-5](https://www.cell.com/neuron/fulltext/S0896-6273(14)00192-5)
[1]: Technically, the part with lower confidence is not discarded, but both are combined in a way where the part with the higher confidence dominates the end result.
Latest Answers