Neural dysregulation in post-COVID fatigue 2023 Baker et al

[Andy]

Abstract

Following infection from SARS-CoV-2, a substantial minority of people develop lingering after-effects known as ‘long COVID’. Fatigue is a common complaint with substantial impact on daily life, but the neural mechanisms behind post-COVID fatigue remain unclear.

We recruited 37 volunteers with self-reported fatigue after a mild COVID infection and carried out a battery of behavioural and neurophysiological tests assessing the central, peripheral and autonomic nervous systems.

In comparison to age and sex matched volunteers without fatigue (n = 52), we show underactivity in specific cortical circuits, dysregulation of autonomic function, and myopathic change in skeletal muscle. Cluster analysis revealed no sub-groupings, suggesting post-COVID fatigue is a single entity with individual variation, rather than a small number of distinct syndromes. Based on our analysis we were also able to exclude dysregulation in sensory feedback circuits and descending neuromodulatory control.

These abnormalities on objective tests may aid in the development of novel approaches for disease monitoring.

Open access, https://academic.oup.com/braincomms/advance-article/doi/10.1093/braincomms/fcad122/7115845

 

[dreampop]

Interesting paper, I've read about half of it. The authors are from Newcastle University, and aim to test if any of a slew of neurological tests can indicate which of 5 potential hypothesized pathologies of fatigue would apply to Long Covid patients (6-26 weeks after infection, average fatigue impact score of around 80/160).

Hypothesis 1. Motoneurons (and the muscles they innervate) are activated by multiple inputs
from motor cortical areas, the brainstem and spinal cord. If any of these systems have reduced
excitability or increased inhibition, as demonstrated in other chronic conditions associated
with fatigue8, this could contribute to a perception of fatigue.

Hypothesis 2. During normal self-generated movements sensory feedback is attenuated9.
Incomplete sensory attenuation during movement could lead to heightened feedback, and an
increase sense of effort10.

Hypothesis 3. At the level of the periphery, other post-viral syndromes (such as Guillain-
Barré and Miller Fisher syndrome) often lead to ineffective signal transmission at the
neuromuscular junction, whereas myopathic changes within the muscle fibres themselves will
cause weakness11-13 requiring stronger voluntary drive to generate force, which could give
rise to an increased perception of effort14.

Hypothesis 4. Monoaminergic neuromodulators are released in the spinal cord and regulate
the gain of motoneuron responses to inputs through the activation of specific membrane
conductances15. If neuromodulatory inputs to motoneurons are affected in post-COVID
fatigue (pCF)16, a stronger synaptic drive would be required for a given level of force. This
could contribute to movements being perceived as more effortful.

Hypothesis 5. Autonomic dysregulation is often a predictor for fatigue in other chronic
illnesses17, and treating dysautonomia has shown promising results in improving the
symptoms of fatigue18. Autonomic dysregulation could also contribute to pCF.

 

[Dolphin]

Press release:
https://www.ncl.ac.uk/press/articles/latest/2023/05/longcovidfatigue/

 

[Hutan]

Tests were carried out on two groups of volunteers—one who self-reporting as suffering from pCF, and a second cohort of control subjects with no fatigue.

Yes, interesting. Lots of overlap, but still some significant differences. I don't have the background to evaluate this, but I liked the systematic approach to testing, and the evaluation against the hypotheses.


Hypothesis 1 - a problem with brain's control of motor neurons - some evidence to support this
There was one measure supporting this, but a number of other measures not supporting the hypothesis. There was a trend to longer visual reaction times which would support this, but the P-values weren't significant after adjustment for multiple comparison.

Voluntary activation of muscles relies on command signals from motor areas of the cortex; the state of cortical circuits has been linked to perception of effort and force output during fatiguing contractions.22,23 By using TMS to assess the function of primary motor cortex, we found that intracortical facilitation (TMS_ICF24) was significantly lower in pCF than controls (conditioned motor evoked potential relative to unconditioned 171 ± 79% versus 258 ± 140%, P < 0.001), suggesting reduced cortical excitability (Hypothesis 1). Other TMS measures also likely to be related to cortical excitability were no different between controls and pCF (the asymptote of the TMS recruitment curve, TMS_asymptote; the recruitment curve slope, TMS_slope; the intensity yielding 50% of the asymptote response amplitude, TMS_I50; active motor threshold, TMS_AMT; resting motor threshold, TMS_RMT). Multiple measures of cortical inhibition showed no significant differences between pCF and controls (short-interval intracortical inhibition, TMS_SICI; metrics of cortical silent period TMS_CSP_AMT, TMS_CSP_slope, TMS_CSP_I50). Possibly consistent with reduced cortical excitability, we also found a trend towards longer visual reaction
times in pCF (in biceps muscle, VRT_Bic, 232 ± 52 versus 210 ± 41 ms; P = 0.026; in first dorsal interosseous muscle, VRT_1DI, 277 ± 61 versus 251 ± 46 ms, P = 0.024 for pCF versus controls, respectively; neither P-value crossed the significance threshold after adjustment for multiple comparisons).

Hypothesis 1 proposes that circuits providing inputs to motoneurons are less active in pCF; this could lead to weaker contractions, and an increased sense of effort. In support of this proposed mechanism, intra-cortical facilitation, a measure of intracortical glutamatergic function,45 was reduced in pCF.45 Other metrics of cortical state, which included measures of intra-cortical inhibition were not different—for example reduced facilitation was not countered by a concomitant reduction in intra-cortical GABAergic or cholinergic inhibition, suggesting a rebalancing of cortical activity and excitability to a lower level. As a result, corticospinal neurons could fire less vigorously for the same input from other upstream cortical areas, and hence plausibly lead to an increased sense of effort and fatigue. In agreement with these results, visual reaction times tended to be slower in pCF. This result also suggests that fatigue can affect cortical circuits differently in different cohorts as in a previous study,8 we instead found evidence for increased intra-cortical inhibition and normal intra-cortical facilitation.

Hypothesis 2 - a problem with sensory feedback processing - no evidence to support this

Disturbances in sensory feedback processing have been previously hypothesized to contribute to an increased perception of effort25 (Hypothesis 2). However, the attenuation of sensory input during movement (SAT), short-latency afferent inhibition (TMS_SAI) and the different components of the cutaneomuscular reflex (CMR_E1, CMR_I1, CMR_E2) all showed no significant differences. This suggests that sensory abnormalities are unlikely to be a contributing factor to pCF in our cohort.

Hypothesis 2 suggests that fatigue results from an impairment of sensory attenuation during movement. If sense of effort is judged from the level of feedback, this could make a movement feel more effortful than it actually was, and hence lead to fatigue.7 Importantly, a direct measure of sensory attenuation was unaffected in pCF; indeed, all measures related to sensory processing appeared normal. While this mechanism may contribute to fatigue in other pathologies (e.g. after stroke, see Kuppuswamy10), it does not appear important in pCF.

From the reference Kuppuswamy:

Therefore, in chronic pathological fatigue, simple activities feel effortful due to the brains inability to ignore the afferent somatosensory consequences of movement.

Hypothesis 3 - a problem with the function of muscles - evidence to support this
On the basis of increased peripheral fatigue during a prolonged maximal contraction, the authors concluded that this could indicate metabolic changes in muscle fibres after prolonged activity, leading to reduced force output. That leads to a stronger voluntary drive and a higher perceived effort. The authors didn't think that the connections between the motoneurons to the muscle fibres was a problem.

Fatigue could arise from a reduced ability of the neuromuscular apparatus to generate force; a given movement would then require stronger voluntary drive and perceived effort would rise. Changes could arise in the muscles themselves,26 due to a weakened connection from motoneurons to muscle fibres,27 or because motoneurons are less excitable. We found that maximal grip strength (Grip) was not significantly reduced in pCF, suggesting no deficit in force production for brief contractions. The efficacy of transmission at the neuromuscular junction (assessed using repetitive nerve stimulation, RNS), and intrinsic motoneuron excitability (assessed by estimating the peak firing rate of single motor units, SMU_peakF and the after-hyperpolarization of motoneurons, SMU_AHP) were also not significantly different between our two cohorts. However, when we tested changes during a prolonged maximal contraction, we found pCF subjects had an increased level of peripheral fatigue (size of maximum twitch evoked by direct electrical stimulation of the muscle after a sustained contraction compared with baseline, TI_PeriphFatigue, 48.5 ± 30.8% in pCF versus 67.1 ± 25.2% in controls, P = 0.003). This suggests that people with pCF develop metabolic changes in muscle fibres after prolonged activity, leading to reduced force output (Hypothesis 3).

Hypothesis 3 is that pCF leads to myopathy, producing muscle weakness that requires an increased neural drive to generate a given contraction strength. Our results provide partial support for this idea. Individuals with pCF had normal grip strength, and there was no evidence of fatiguing transmission at the neuromuscular junction. As far as we could assess, the intrinsic excitability of motoneurons was also normal (measurements of persistent inward currents, SMU_deltaF and after-hyperpolarisation, SMU_AHP). However, myopathic changes became apparent after a sustained contraction, when the ability of muscle to produce force in response to electrical stimulation was significantly reduced in pCF subjects. This may reflect abnormalities in energy metabolism, leading to a more rapid depletion of muscle energy stores46 but this would need verification with further studies that directly measure muscle metabolic function. Clearly, such deficits could lead to a feeling of fatigue,47,48 although whether muscles are regularly pushed to the regime where such effects become noticeable in everyday life is perhaps debatable.

Hypothesis 4 - a problem with the motorneuron responses - no evidence to support this

We assessed the state of descending neuromodulatory pathways by looking at differences in the recruitment and de-recruitment of motoneurons (SMU_deltaF); the persistent inward currents that mediate this phenomenon are highly sensitive to serotonergic and noradrenergic inputs. We did not find any difference between our cohorts, suggesting that pCF is not associated with significant changes in descending neuromodulatory drive (Hypothesis 4).

Hypothesis 4 relates to the extensive role played by neuromodulators in motoneuron function.15 Recent work has emphasized how active channels in the motoneuron dendrites amplify synaptic currents, and even generate sustained firing and thereby contractions in the absence of synaptic drive. The magnitude of these persistent inward currents is regulated by neuromodulators.49 There is evidence for changes in neuromodulatory centres following other inflammatory50 or autoimmune disorders16; thus even a small reduction in tonic levels of neuromodulators could leave motoneurons relatively unresponsive to descending drive,51 and hence generate feelings of weakness and fatigue. However, assessment of persistent inward currents showed no evidence for a difference in pCF, suggesting that this mechanism does not contribute to fatigue after SARS-CoV-2 infection.

Hypothesis 5 - autonomic dysregulation - some evidence to support this
The increased heart rate finding was consistent with autonomic dysregulation. There were also some trends to a number of other measures supporting autonomic dysregulation.

Autonomic dysregulation is often associated with fatigue in other conditions17,28 and recent studies reported autonomic dysregulation after SARS-CoV-2 infection29-32(although not universally33). We found a significantly increased resting heart rate in pCF (Mean_HR, 74.8 ± 11.1 versus 67.7 ± 8.8 beats/min, P = 0.0016). Other measures of autonomic function (tympanic temperature, Temp, 36.9 ± 0.4 versus 36.7 ± 0.3°C, P = 0.018; heart rate variability, pNN50, 8.8 ± 15.7 versus 20.2 ± 21.1%, P = 0.011; galvanic skin response habituation, GSR_Hab, 25.2 ± 24.5 versus 14.3 ± 12.2%, P = 0.026) also differed between cohorts, but did not pass correction for multiple comparisons. In our cohorts, only a small number of subjects had any medication that could potentially affect heart rate measurements [propranolol (n = 1, Control), atenolol (n = 1, Control) and amlodipine (n = 1, pCF)], and therefore, medications are unlikely to have had a significant impact on our results. These results all point towards a reduced vagal (relative to sympathetic) tone, suggesting at least some of our pCF subjects suffer from a degree of dysautonomia (Hypothesis 5).

They note that these findings, including temperature, and also the reduced oxygen saturations, could be related to long-lasting impacts of the acute infection on lung function and immune activation. I had the impression that any temperature difference in ME/CFS tended to lower temperatures, not higher.

On the oxygen saturations:

However, it should be noted that although they were significant, the differences in SaO2 were small. Many clinical conditions lead to reductions in SaO2 larger than the 2% change we saw here, without producing symptoms of fatigue. It is thus unlikely that SaO2 is the sole driver for the differences in neural measures that we observed.

Interestingly, there was no evidence for more than one cluster within the pCF cohort, as we might expect if pCF originated from multiple causes (which could include a psychogenic origin). This finding should be treated as preliminary, given the relatively small size of our cohort, but it does suggest that treatment of pCF may not require extensive stratification to be successful.

 

[SNT Gatchaman]

Followed up in Recovery of neurophysiological measures in post-COVID fatigue: a 12-month longitudinal follow-up study (2024, Nature Scientific Reports)