In the January 2019, Pain Journal published an article comparing pain to the Bayes rule in math.
PAIN Symptom Perception, placebo effects, and the Bayesian brain. It suggests the brain can follow a theory of probability in math known as the Bayes rule. In statistics/math, the rule looks at the likelihood of a given hypothesis. It takes into account prior evidence, current evidence, likelihood probability of other related hypotheses and makes a prediction. The brain generates a top-down, out of our awareness, neurally encoded hypothesis about the state of us and the world. So cool!
This top-down hypothesis is met by its bottom-up sensory input. Any mismatch between the two results in a “coding error” or “prediction error” and the brain revises it’s hypotheses and prediction rules for next time. These hypotheses include those built in from evolution and the person’s personal experiences. Throughout our lifespan, they are constantly changing and updating to help predict the next set up sensory inputs. The implication of this theory is that we perceive not how the world actually is, but the brain’s best guess of it. This definitely makes sense when we think about visual perception.
In the case of chronic pain, the brain puts a high emphasis on the hypotheses generated by the brain and less precision on the sensory experience. Slight and otherwise harmless stimulus (or bottom-up experiences) cause the brain to misinterpret these as contributing to the pain and continue to modify the hypothesis to include these. Individuals without chronic pain would tend to ignore these inputs as “noise.” Conditions such as anxiety, threat, and catastrophizing have a tendency to worsen symptoms by maintaining vigilance to these predictor hypotheses.
So how does the Placebo Effect fit into this theory? The Bayesian perspective states relief of symptoms is not necessarily due to restoring bodily function but more about the lack of interference from the bottom into the hypotheses. As sensory input diminishes the hypothesis again changes. This change, however, seems to be slower if the patient does not have external cues that the change is happening. It’s almost like the brain ignores the improvement as “noise” unless there are other cues to say it’s time to revise the hypothesis. This may be why avoiding nociceptive language AND encouraging our patients that they are RESILIENT and CAN DO their important activities is so important. Your body is getting better so go ahead and revise your hypothesis! This may also be why the placebo effect works. Experiments where the patient was given analgesics in a manner that did not let them know the medications were supposed to help reported significantly less pain relief than those who were told what they were receiving.
I don’t think the article was saying anything we haven’t already approached with pain science. However, I appreciate this statistical analysis way of looking at it. Maybe it’s because I live in Seattle and there are tons of tech-driven people here, but I feel like I have one more analogy to help my patients. AND I don’t actually have to calculate any statistics!
Latest posts by Valerie Bobb (see all)
- How to Screen if the Cervical Spine is Impacting Shoulder Pain. (Yes it’s in Your Scope!) – August 20, 2019
- She Added Massage And That Made Migraines Unbelievably Better – February 4, 2019
- Is Pain Really A Math Equation? – January 29, 2019