Towards the unity of pathological and exertional fatigue: A predictive coding model, 2020, Greenhouse-Tucknott et al.

Our nerves do not function quickly enough for motor control feedback to occur in real-time, hence we have a predictive model of motor control (partially located to the supplementary motor area of the brain). Without this model, we would not be able to precisely control the force and location of our limbs. Obviously this is not a conscious process.

Feedback from afferent proprioception modulates this model pretty much all the time, however the impact of feedback is limited over time (which is why it takes time to learn skilled motor tasks) when actual motor function continues to undershoot over time, this signals to the brain that the model as failed and this is perceived as a sense of weakness/fatigue. The purpose of this is to say, hey, maybe we can't continue doing this task forever!

There is a separate mechanism which tries to front-run this failure, the type 3 and 4 afferents which measure local conditions - mechanical and metabolic. These feedback directly through spinal feedback (causing "central fatigue), and can also signal sensations associated with fatigue (muscle soreness for example, which we may consciously interpret as predicting fatigability), but is not fatigue itself.

The brain does not have a predictive model of these afferents and hence this is not a part of the models described in the above papers. Nonetheless, "active inference" still occurs to minimise these sensations, namely behaviour to slow down/stop.
Notably, there is no body-wide sense of allostasis, and the local measure of metabolic homeostasis has no direct input into the proprioceptive model in the brain, instead perceived indirectly as a loss of expected force (fatigue) due to the direct spinal feedback.

 

Thanks, @Snow Leopard, I'm not clear on some of the technical stuff you mention, but from a cognitive psychological point of view. it seems to me that the predictive coding model is not that applicable to explaining how we modify bodily systems in response to activity demands, and how we experience feedback that physical demands are are starting to outstrip our capacity to meet them. For one, the original idea of predictive coding was supposed to explain learning that generalises to new situations, but this is more of a dynamic, event-specific scenario. Also, it seems that the predictive bit is not doing any new explanatory work that isn't already being done by the concept of feedback loops.

The other problem I have with it is that it seems to have taken the bit about prediction error being aversive rather too literally. Its obviously not the case that all experiences of prediction error are unpleasant. Some surprises are lovely, as are some unexpected events, and sometimes having our predictions disconfirmed is a source of delight in its own right - like in the unexpected punchline of a good joke.

I read Edwards et al's stuff on FND a few years ago, and I think you give it way too much credit. Basically, the claim is that people with FND predict that their motor behaviour will be aberrant (that is, they believe at some level that they have a motor dysfunction), when in fact they have no dysfunction at all. This will obviously result in prediction error, because the person is likely to find that they they are able to move in ways that do not align with their predictions, and experiencing this prediction error is supposed to be aversive. The idea is that these people are so wedded to their prediction (their belief in having a motor disorder) that they will resist revising it, but instead will deal with the prediction error by modifying their behavior to fit the prediction. That is, they "create" the motor problem to fulfil their prediction. Not necessarily all at a conscious level.

Its is not made clear why these individuals hold so strongly to their incorrect predictions (aka false beliefs) that they will resist revising them at great cost to themselves. That seems to go against the core principles of predictive coding, which is designed to explain how we revise our understanding in the face of surprising information. The idea seems extremely insulting to people with FND.

 

Be afraid; be very afraid. He is co-opting ME for his own ends, and sees it only as "chronic fatigue" because that's what suits his purposes.

 

I'm definitely wary about anything he's involved with. We have a couple of threads about studies that he is involved with, if anybody want's to read up on his work.

Persistent fatigue induced by interferon-alpha: A novel, inflammation-based, proxy model of Chronic Fatigue Syndrome, 2018, Pariante et al

UK: Brighton and Sussex University Hospital and Medical School - Chronic Pain and Fatigue

 

The purpose of this sort of work seems to be to subsume ME/CFS within their model of FND.

 

I have a growing impression that most FND patients are pwME, or similar, patients who were simply defined by an arbitrary "main symptom". So I'd say it's already the case, basically all the neurological symptoms are folded into the FND concept, probably applied to those who have less or no fatigue as main symptoms.

It's the "primary complaint of X" that is the universal feature, despite the fact that there is rarely such a thing but they just pick and choose. Probably anchored from whatever the patient first talked about, all other symptoms are basically assumed to be, somehow, sub-symptoms, or whatever they tell themselves.

They just lump us up based on superficial features and try to explain things post hoc from whatever they threw in the drawer, the models are simply invented afterward.

 

They have reduced ME to chronic fatigue and left out all the neurological problems that meant it was classified as a neurological disease. So now the people who have classic ME with fatiguability, variability and a long list of other symptoms are being rediagnosed as FND.

I think it was PW who said there were no neurological problems in ME when they were formulating the NICE guidelines.

 

Published open access here, https://link.springer.com/article/10.3758/s13415-021-00958-x