Recovery of neurophysiological measures in post-COVID fatigue: a 12-month longitudinal follow-up study, 2024, Maffitt et al.

[SNT Gatchaman]

Recovery of neurophysiological measures in post-COVID fatigue: a 12-month longitudinal follow-up study
Maffitt, Natalie J.; Germann, Maria; Baker, Anne M. E.; Baker, Mark R.; Baker, Stuart N.; Soteropoulos, Demetris S.

One of the major consequences of the COVID-19 pandemic has been the significant incidence of persistent fatigue following resolution of an acute infection (i.e. post-COVID fatigue). We have shown previously that, in comparison to healthy controls, those suffering from post-COVID fatigue exhibit changes in muscle physiology, cortical circuitry, and autonomic function. Whether these changes preceded infection, potentially predisposing people to developing post-COVID fatigue, or whether the changes were a consequence of infection was unclear.

Here we present results of a 12-month longitudinal study of 18 participants from the same cohort of post-COVID fatigue sufferers to investigate these correlates of fatigue over time. We report improvements in self-perception of the impact of fatigue via questionnaires, as well as significant improvements in objective measures of peripheral muscle fatigue and autonomic function, bringing them closer to healthy controls.

Additionally, we found reductions in muscle twitch tension rise times, becoming faster than controls, suggesting that the improvement in muscle fatigability might be due to a process of adaptation rather than simply a return to baseline function.

Link | PDF (Nature Scientific Reports) [Open Access]

 

[SNT Gatchaman]

Follow-up of cohort in Neural dysregulation in post-COVID fatigue (2023, Brain Communications)

 

[Sean]

Additionally, we found reductions in muscle twitch tension rise times, becoming faster than controls, suggesting that the improvement in muscle fatigability might be due to a process of adaptation rather than simply a return to baseline function.

 

[Hutan]

Some key excerpts on Methodology:

In a previous study12 in long COVID after non-severe SARS-CoV-2 infection, where there had been no requirement for acute hospital in-patient care, we demonstrated changes in muscle physiology. In addition, we found reduced responsiveness in some of the neural circuitry governing cortical excitability, and a rise in the resting heart rate probably indicating an altered balance between sympathetic and parasympathetic systems. These changes might explain some of the symptoms of fatigue described by patients. Here we report a longi- tudinal follow-up study in a subset of the same cohort, in whom neurophysiological measurements and the self-reported perception of the impact of fatigue on daily life were repeated twice more, at 6 month intervals. The reported impact of fatigue improved significantly after a year; for most of the neurophysiological metrics there was likewise a return to levels seen in age and sex-matched controls. We therefore suggest that these changes in muscle physiology and autonomic function occur as a consequence of a SARS-CoV-2 infection, rather than a pre-existing phenotype that increased susceptibility to developing post-COVID fatigue.

Measures collected in our initial study12 were used here as baseline data. This included data from a cohort of 37 participants (27 female) who were suffering from pCF by self-report and a second cohort of 52 volunteer controls (37 female) with no symptoms of fatigue. Inclusion criteria were age 18–65 years, with no history of neurological disease. The first visit to the laboratory was made 6–26 weeks after infection for the pCF cohort. In the control cohort, six subjects had reported having mild COVID-19 but with complete recovery and no symptoms of pCF. Of the 37 people with pCF from our initial study, 18 were later recruited to this longitudinal follow-up study (13 female), completing a further two lab visits at intervals of approximately 6 months to yield a total of three visits.

General electrophysiological methods
Electromyographic activity (EMG) was recorded with adhesive surface electrodes positioned over muscles ...

Paired‐pulse TMS
EMG was recorded from the first-dorsal interosseous (1DI); the TMS coil was moved to locate the hot spot for this muscle. The resting motor threshold (RMT) was determined....We then measured the responses to the test stimulus alone, and when preceded by the conditioning stimulus at intervals of 10 ms, corresponding to intracortical facilitation (ICF) (see Baker et al., 2023 Supple- mentary Fig. 1). ..

Heart rate
A single channel electrocardiogram (ECG) recording was made, using a differential recording from either left shoulder and right leg, or left and right shoulders

Measures of muscle physiology
...Subjects sat with their dominant arm and forearm strapped into a dynamometer to measure torque about the elbow; the shoulder was flexed, and the elbow at a right angle, so that the upper arm was horizontal and the forearm vertical. The forearm was supinated. Thin stainless-steel plate electrodes (size 30 × 15 mm) were wrapped in saline-soaked cotton gauze and taped over the belly of the biceps muscle (cathode) and its distal tendon (anode). Electrical stimuli were delivered through these electrodes while monitoring the evoked twitch response recorded by the dynamometer, and the intensity increased until the response grew no further. This supramaximal stimulus was used for all subsequent measurements.

The following recordings were then made in sequence; this protocol was followed to maintain consistency with our original study. A brief tone cued the subject to produce and hold a maximal voluntary contraction; 2 s after the tone, a stimulus was given to the biceps, and 1 s later a second tone indicated that the subject should relax. Five seconds later, a biceps stimulus was given, followed by a further 55 s rest period. This sequence was repeated three times. A long tone then cued the subject to make a sustained maximal voluntary contraction. This was continued either for 95 s, or until the force exerted fell below 60% of the initial maximal level. During this sustained contraction, the biceps was stimulated every 10 s. After the contraction ended, a final three biceps stimuli were given at rest (inter-stimulus interval 5 s).

From the three stimuli delivered at rest at the start, we averaged the maximal twitch at rest and measured its amplitude, Fbefore. From the three stimuli delivered at rest after the sustained contraction, we measured Fafter.

rest rest Peripheral fatigue (measure TI_PeriphFatigue) was calculated as:

F after TI_PeriphFatigue = rest 100%

This describes the reduced ability of the muscle to generate force after fatigue, even when activation is per- formed independent of the central nervous system by an electrical stimulus to the muscle. Additionally, we measured the time to maximal force generation following direct electrical stimulation to the muscle (measureRiseTime) using the twitch evoked at rest at the start of the procedure, before the fatiguing contraction.

[Blood oxygen saturation]
Blood oxygen saturation was measured using a pulse oximeter placed onto the index finger (SaO2). This was recorded just after the TMS measurements, when the subject had been sitting at rest for around 20 min, and before the measure of peripheral fatigue.

Also data was collected on the self-reported impact of fatigue.

 

[Hutan]

RESULTS
Baseline (Visit 1) was 6 to 26 weeks after infection.
Visit 2 was 6 months later
Visit 3 was 6 months after that (so roughly, one year to a year and a half after infection).

Excerpts:

Fatigue Impact
There was a significant (p < 0.0001, F = 13.3) decrease in the self-reported perception of the impact of fatigue across visits; mean FIS score declined from 80.3 at V1 to 60.2 at V2 and 46.1 at V3 (Fig. 1B).... Overall, the majority (16/18) of participants had improved FIS scores between V1 and V3.

useful data to support the contention that most people's symptoms improve in the first two years

Biological measures

Our earlier work showed that only a small number of measures (TI_PeriphFatigue, TMS_ICF, Mean_HR, SaO2) out of an extensive initial set were significantly different between controls and participants suffering from pCF. Only these measures were therefore repeated during V2 and V3 (Fig. 2). Over time there was a significant change in peripheral oxygen saturation (SaO2, p = 0.033, χ2(2) = 6.813), heart rate (Mean_HR, p = 0.033, F = 3.88), and peripheral fatigue (TI_PeriphFatigue, p = 0.002, χ2(2) = 12.4). In each case, values from people with pCF became more similar to controls (dotted lines in Fig. 2) with time. Using TMS to investigate the excitability of the primary motor cortex, we found that although intracortical facilitation became more similar to controls with each visit, this trend did not reach significance (TMS_ICF, p = 0.179, χ2(2) = 3.44).

Figure 2

Additional measures of muscle physiology
To investigate the potential mechanisms underlying improvements seen with peripheral fatigue over time, we made one further measure of muscle physiology—the rise time of a maximal twitch (RiseTime). This refers to the time taken from direct muscle stimulation to the peak force generated (measured from the biceps, at rest at the start of the twitch interpolation protocol; Fig. 3A). Although, relative to controls, there was no significant difference at V1 (p = 0.410), RiseTime was significantly reduced compared to controls at both V2 (p = 0.010) and V3 (p < 0.001). Furthermore, RiseTime of the pCF cohort significantly fell over time (p = 0.001, χ2(2) = 13.661). Post hoc comparisons showed that at both V2 and V3, RiseTime significantly decreased relative to V1 (− 9.8 ms, p < 0.001 and -11.8 ms, p < 0.001 respectively). RiseTime was not significantly different between V2 and V3.

Figure 3

 

[Hutan]

DISCUSSION

As a group the reported impact of fatigue decreased, and our metrics returned to or were returning to normal. This suggests that the
changes were mediated by SARS-CoV-2 infection, rather than being a pre-existing long-term trait amongst the pCF sufferers.

The most significant change over time was observed in muscle fatigue. The metric TI_PeriphFatigue increased (signifying less peripheral fatigue) from 37% at the first lab visit to 55% one year later, indicating that muscles undergo significant physiological changes during recovery.

There is speculation as to what might be causing this muscle fatigue e.g. impairment of mitochondrial function, inflammation in the muscle. There's the suggestion that there is a pre-existing subclinical mitochondrial dysfunction unmasked by the infection.

It is thus possible that those who develop pCF might have subclinical mitochondrial dysfunction, subsequently unmasked by SARS-COV-2 infection, leaving them with systemically impaired mitochondrial function, as is typical of the pattern in mitochondrial disease32, potentially explaining the broad spectrum of symptoms expe- rienced by patients suffering from long COVID. 33,34

There is clear evidence for a specific role of mitochondrial dysfunction both in critical illness myopathy and in long COVID9,10. Metabolic adaptations include a shift away from energy generation in mitochondria through oxidative phosphorylation and towards anaerobic glycolysis, where pyruvate is converted into lactate9,35, and phosphocreatine breakdown; these changes in turn lead to the accumulation of metabolites that promote inflammatory processes36 and the generation of muscle fatigue37.

They suggest that an increase in fast twitch muscle fibres is a response to the disease.

Our cohort with pCF showed significantly less peripheral fatigue 6 months and 12 months after they were first assessed. Interestingly, as the pCF cohort recovered from the symptoms of fatigue, their twitch tension rise times became faster than controls (Fig. 3B). This suggests that the improvement over time in muscle fatigability (corresponding to an improved ability of muscle to generate force with continued contraction) might be due partly to a process of adaptation rather than a complete return to baseline physiology. One recent report showed an increase in the proportion of type II muscle fibres in long COVID9, although another failed to find a significant difference in fibre type proportions10. A similar change in fibre type is recognized in mitochondrial myopathy, where histopathology is reported to show an increased ratio of type II to type I fibres38. More recently, proteins involved in mitochondrial fusion/fission have been implicated in the intracellular signalling processes regulating fibre type switching39, as observed in response to exercise40. Type II fibres have high fatigability, because of their higher reliance on anaerobic rather than aerobic metabolism, and faster single fibre twitch times because of a higher rate of cross-bridge cycling and associated ATP breakdown41. A shift towards type II fibres could therefore lead to the shorter muscle twitch time seen here. However, it must be noted that the twitch time was not different between controls and long COVID in the first visit of our study; the difference only developed with time as pCF recovered. A shift in fibre type cannot therefore be the primary underlying cause of pCF.

They suggest that the changes over time could explain the discrepancies in the literature - some studies are early and some studies are later. They note that they don't know what happens to the ratios of muscle fibres later - do the ratios return to normality over time in people who recover, and in people with ongoing fatigue?

Changes in peripheral fatigue had a functional impact. Whereas almost all (92%) of controls were able to hold a sustained contraction above 60% of original maximal force for 95 s, only 40% of people with pCF could do this at V1, but all managed it at V2 and V3.

 

[Hutan]

They suggest that the fatigue PROM FIS is not very accurate.

Although our biological metrics returned to levels that were compatible with controls and reported impact of fatigue reduced (Fig. 1B), we found no significant correlation between changes in these two. This is perhaps not surprising. How an individual copes with fatigue—and thus assesses its impact on their life—will probably be affected by psychological factors such as level of resilience or social considerations, such as varied access to support systems. The magnitude of change in FIS score, measured by a subjective questionnaire, is likely to be influenced at least as much by these factors as the underlying pathology, which may explain the lack of correla- tion with more objective metrics.

They promote the use of the objective measurement of peripheral fatigue:

In particular, the measurement of peripheral fatigue is simple and can be achieved with low-cost equipment. This could deliver an objective assessment, to be interpreted alongside subjective measures such as the FIS score.

(peripheral fatigue being measured as
"the reduced ability of the muscle to generate force after fatigue, ... activation is performed independent of the central nervous system by an electrical stimulus to the muscle." )

Sounds sensible to me.