Review Within person predictors of physical activity & fatigue in long Covid: Findings from an ecological momentary assessment study, 2025, Burton

https://www.sciencedirect.com/science/article/abs/pii/S0022399925000558

Journal of Psychosomatic Research
Available online 6 March 2025, 112091
In Press, Journal Pre-proof
Within person predictors of physical activity and fatigue in long Covid: Findings from an ecological momentary assessment study
Christopher Burton, Helen Dawes, Caroline Dalton
a
School of Medicine and Population Health, University of Sheffield, Sheffield, UK
b
NIHR Exeter BRC, College of Medicine and Health, University of Exeter, Exeter, UK
c
Advanced Wellbeing Research Centre, Sheffield Hallam University, Sheffield, UK
Received 29 May 2024, Revised 1 March 2025, Accepted 5 March 2025, Available online 6 March 2025.


https://doi.org/10.1016/j.jpsychores.2025.112091Get rights and content

Abstract

Objective
We aimed to examine the extent to which current perceived demand for energy and affect predict subsequent physical activity and fatigue in people with Long Covid using an intensive longitudinal method (ecological momentary assessment).

Methods
Analysis of data from a study of 69 adults with self-reported Long Covid combining 3-hourly self-report data perceived energy, and fatigue, on a smartphone app with continuous physical activity recording. We tested three hypotheses derived from cognitive behavioural and neuroscientific models of fatigue. These related to expectation, current affect and recalled emotional demand. Analysis used linear mixed effects models with fatigue and physical activity as outcomes.

Results
Expectation of energy need for the next 3 h was predictive of physical activity, fatigue and recalled demandingness of the period. (p-values 0.005 to <0.0001). Currently feeling positive was predictive of slightly more subsequent physical activity and less fatigue 3 h later (p = 0.01). Feeling negative was not predictive of physical activity or subsequent fatigue but was predictive of subsequent recall of the period being emotionally demanding. Feeling more anxious was predictive of greater fatigue 3 h later (p = 0.001) but not of reduced physical activity. Absolute effects were small: a one-point increase in anticipated demand (on a scale of 1–7) was associated with an extra 2.2 min of moderate or vigorous physical activity and a one standard deviation increase in anxiety was associated with a one-point increase in fatigue (0–100 scale).

Conclusion
In the day-to-day experience of Long Covid expectation and affect have little detectable effect on subsequent physical activity or fatigue.

 

Wouldn’t this mean that the cognitive behavioural and the neuroscientific models of fatigue have no predictive power? I.e. that the models appear to be wrong? Edit: in the case of LC.

 

I am sure it means the models are actually correct but they just need lots and lots more research funding.

 

It'll be interesting to see a bit more info on this when available - how many people, how were they selected, what kind of LC they were experiencing, how long they were tracked and so on.

Was it Sheffield Hallam who put out that silly fatigue management booklet telling people to record and analyse their symptoms and emotions all the time? (sorry, too foggy to track down the thing I'm half-remembering)

 

There's something funny about presenting those mediocre 'results' as if they suggest that there's something, when it's about on par with a fundraising campaign aiming for a million dollar proudly boasting of having raised $7, $5 of which was from someone who accidentally dropped a bill from a pocket while taking out their phone.

But unfortunately, the superpower of psychobehavioral ideology is that their methodologies are extremely bad on purpose, making it easy to simply not bother with negative results. They only count the apparent positive ones, the very same process used for "one take trick" videos where they record however many takes it takes to get it "in one shot". In accounting this would simply be called fraud, it would be like counting only incomes and ignoring expenses, presenting the appearance of solid profits.

Also another clear sign that the academic process has effectively collapsed with almost no one noticing it.

 

The conclusion is very clear that they did not really detect any effects. So I don’t think it’s fair to say that they present the results as ‘something’ when it’s not.

 

And being able to provide care for as many patients as possible so they can accumulate clinical experience as subject matter experts.

 

Might be this one, download at the bottom of the text: https://www.shu.ac.uk/advanced-well...manage-chronic-fatigue-brought-on-by-covid-19

I briefly skimmed it, it looks like a form of GET where gradually increasing activity is supposed to help after you first find a stable baseline.

 

I wonder if these attempts to find a link between emotions and fatigue are due to a misunderstanding of what patients are actually experiencing.

The researchers are, I suspect, misinterpreting descriptions of fatigue by sick patients as various negative emotions, like feeling disappointed, not seeing a positive future or a chance of positive things, feeling overwhelmed by uncertainty and similar things. These emotions can have a component that is sometimes imprecisely described as fatigue in everyday language, but it's not really the same thing as fatigue due to illness.

From this confusion then come attempts to fix fatigue by working on the psyche. The results seem to consistently show null or small effects, often exaggerated by misleading methods.

 

That’s a generous assumption. My thinking is more that the psychosomatic practitioners believe that they have gained access to some powerful insight about the nature of human beings, and that they can use this to fix things for people.

I’ve heard numerous versions of ‘I decided to learn this [something psychosomatic] because I’m interested in complex cases/modern medicine is failing patients’. They buy into the ancient paradigm shift-narrative because it reinforces their belief that they have understood something useful.

 

The patient cohort:

The retreat from the cognitive behavioural model is interesting, even if the authors do favour a predictive-coding model instead:

 

Thank you, @Nightsong ! Can you post the name of reference 12 as well?

This suggests a need for further research in order to better understand the relationship between symptoms such as fatigue and psychological mechanisms and also to provide explanations which are in keeping with the science and are acceptable to patients
​

Acceptable to the patients should not be a criteria. That’s like not wanting to tell people to stop smoking just because some won’t like it. Figure out what’s best first, then figure out how to sell it.

And they also assume that there is some relationship between fatigue and psychosocial mechanisms. We don’t know that at all, and all evidence points towards an insignificant relationship in the psychosocial to fatigue direction. The other way around might have some relationship.

In this context, neuroscientific explanations for central components of fatigue which include the brain’s allocation of internal resources, competition between emotional, mental and physical demands, and involuntary errors in predictive coding due to altered interoception [12] become both plausible and potentially useful.
​

I don’t see any evidence that the mentioned mechanisms are more plausible (or useful) due to the failure of the behavioural model. Only evidence for the models can make them more plausible, because they are not the only options. It’s worrying that they fail to consider other explanations.

 

A previous article by Burton that seemingly had a lot of premature conclusions regarding causality:
https://www.s4me.info/threads/withi...e-longitudinal-study-2023-burton-et-al.31606/

One part:

These findings are in keeping with an embodied predictive interoceptive coding model of symptoms, suggesting that Long Covid is associated with changes to the way the brain processes signals from the body."
​

It seems like he has a thing for the predictive coding model. So there is a possibility that this study was a ‘hit-piece’ targeted at the behavioural model, and that were observing some internal fighting in the psychosomatic field as a whole?

 

Here’s another piece by Burton:
https://www.s4me.info/threads/stigm...amework-synthesis-2024-treufeldt-et-al.38878/

It’s a patient-blaming article about stigma. This about sums it up:

He has also done work with Rosmalen..

 

"The studies will continue until morale improves."

 

[12] is this predictive-processing paper.

 

Thank you again!

That’s this article, which, according to the thread, is quite bad:
https://www.s4me.info/threads/towar...g-model-2020-greenhouse-tucknott-et-al.16228/