Microneurography Reveals Unmyelinated Small Nerve Fiber Dysfunction in Long COVID, 2025, Ribeiro et al.

[SNT Gatchaman]

Microneurography Reveals Unmyelinated Small Nerve Fiber Dysfunction in Long COVID

Spoiler: Authors

Ana Ribeiro; Shahrzad Hadavi; Nick Gall; Robert Dm Hadden; Jordi Serra

OBJECTIVE
To review the microneurography findings of long coronavirus disease 2019 (COVID) patients who presented to the clinic with multisystem involvement affecting neurological, cardiovascular, gastrointestinal, genitourinary, pulmonary, and immunological domains.

METHODS
We analyzed 36 consecutive long COVID patients using microneurography. We evaluated abnormalities in C nociceptors, including spontaneous activity, peripheral sensitization, multiple spikes, conduction failure, and alterations in activity-dependent slowing of conduction velocity. Sympathetic nerve fiber function was assessed using the recovery cycle of excitability. Results were compared with a large normative database.

RESULTS
The mean age was 40.9 ± 9.2 years (range 17–60 years), with a female predominance (30/36, 83.3%). Patients were seen from 15 to 61 months after onset of symptoms (35.7 ± 11.3 months). All patients presented with neuropathic symptoms, mainly pain and orthostatic intolerance. A total of 32 patients (88.9%) had objective electrophysiological abnormalities in peripheral C fibers, including spontaneous nociceptor activity (61.1%), peripheral sensitization (27.8%), and multiple spikes (11.1%). Long COVID patients also showed a significant shift in C nociceptor populations, with a higher prevalence of type 1B mechano-insensitive C nociceptors compared with healthy controls. Changes in activity-dependent slowing of conduction velocity differed in opposite directions between mechano-sensitive and mechano-insensitive C nociceptors. Postganglionic sympathetic fibers also showed abnormal recovery cycles with a lack of supernormality, suggesting impaired neuronal homeostasis.

INTERPRETATION
This study provides novel electrophysiological evidence linking small nerve fiber dysfunction to long COVID. These findings align with previous histological evidence of small nerve fiber loss, reinforcing the hypothesis that peripheral nerve dysfunction contributes to the multisystem symptoms of long COVID.

Web | PDF | Annals of Neurology | Paywall

 

[SNT Gatchaman]

In summary, we have described for the first time widespread abnormalities in the function of unmyelinated C fibers, both somatosensory and sympathetic, that could represent a common pathophysiological mechanism explaining the multiplicity of symptoms in long COVID.

From the Department of Clinical Neurophysiology, King’s College Hospital, London, UK; Department of Cardiology, King’s College Hospital, London, UK; Department of Neurology, King’s College Hospital, London, UK; and Department of Basic and Clinical Neuroscience, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, UK

 

[SNT Gatchaman]

Summary quotes from introduction —

SFN involves damage or dysfunction of the small unmyelinated (C fibers) and/or thinly myelinated (Aδ fibers) peripheral nerve fibers, which play a crucial role in pain sensation, autonomic regulation, and immune interactions.

There are several lines of evidence of SFN in long COVID. For example, psychophysical quantitative sensory testing has shown variable degrees of sensory dysfunction, and skin biopsies have demonstrated reduced intraepidermal nerve fiber density ranging from 46.2% to 78%.

The diagnosis of SFN is particularly difficult, because the function of unmyelinated axons cannot be assessed using routine nerve conduction studies. Also, skin biopsies suffer from rather low sensitivity, with many false negatives. Other specialized tests with higher sensitivity, such as microneurography, are not widely available.

 

[SNT Gatchaman]

Summary quotes from methods (1) —

retrospective analysis of a cohort of consecutive patients referred for microneurography to the Department of Clinical Neurophysiology at King’s College Hospital, London, UK, between June 2022 and February 2025. The patients were seen as part of their clinical study to rule out small nerve fiber involvement.

All patients were diagnosed with long COVID after clinically proven SARS-CoV-2 infection […] following the definition of the World Health Organization of onset of symptoms within 3 months of SARS-CoV-2 infection, and duration of symptoms that last for at least 2 months that cannot be explained by an alternative diagnosis.

Symptoms were grouped by affected organ system into 6 main categories: cardiovascular/dysautonomia (tachycardia, dizziness, syncope or presyncope, autonomic dysfunction), musculoskeletal (fatigue, fibromyalgia syndrome), gastrointestinal (nausea, abdominal pain, diarrhea and/or constipation, gastroesophageal reflux), genitourinary (bladder control issues, irregular menstrual cycles), neurological—headaches and cognitive function (headaches/migraines, brain fog) and neurological—somatosensory symptoms (burning pain, paresthesia, and numbness), and their location.

(Nice to see fatigue listed as musculoskeletal for a change, rather than neuropsychological.)

Microneurography recordings from the superficial peroneal nerve were obtained following pre-established methods. All parameters extracted from the long COVID cohort were compared with an internal normative database comprised of 256 healthy volunteers with 1262 C fibers recorded using identical methods.

 

[SNT Gatchaman]

Summary quotes from methods (2) —

Subpopulations of peripheral unmyelinated C fibers were identified based on their profile of activity-dependent slowing of conduction velocity (ADS), as detailed previously. C-fiber type nomenclature follows that used in previous studies.

(Reformatting)

Type 1 nerve fibers correspond to C nociceptors, further subdivided between
type 1A (mechano-sensitive C nociceptors) and
type 1B (mechanoinsensitive C nociceptors);

type 2 corresponds to C-thermoreceptors;
type 3 represents the class of (C-low threshold mechanoreceptors; and
type 4 refers to efferent sympathetic nerve fibers

We analyzed the microneurography recordings, looking for the 4 main types of abnormalities in C nociceptors previously described in peripheral neuropathic pain patients:

1. Spontaneous activity: […] In summary, when stimulating at the baseline rate of 0.25 Hz, C nociceptors have a stable latency, producing almost flat lines on the raster plots. To be confident that an abnormally jittery latency profile in a raster plot is caused by spontaneous activity, the histogram of latency changes between successive sweeps should be positively skewed and should extend to a latency increase expected for 2 extra spikes. These fluctuations produce a characteristic ‘saw-tooth’ when stimulating at 0.25 Hz.

2. Abnormalities in transductor properties to natural stimuli: […] monofilaments exerting 3 different forces of 128, 256, and 512 mN […] were applied […]. Mechanically responsive units produce a transductor-mediated response that adds ADS on the latency plot lines […] Peripheral sensitization was defined as the appearance of a mechanical response in a type 1B mechano-insensitive C nociceptor, or by long-lasting responses exceeding the duration of the actual mechanical stimulus, typically during several minutes. Contrarily, lack of responsiveness of type 1A mechano-sensitive C nociceptors was defined as the lack of any response to the stronger 512-mN force stimulus.

3. Multiple spikes: […] whereby unidirectional conduction failure at a branch point can lead to a paradoxical increase in the number of impulses reaching the central nervous system. […] Multiple spikes can be facilitated by sprouting, which is seen in degeneration/regeneration processes.

4. Activity-dependent slowing: The measures of C nociceptor ADS obtained from the patient cohort were compared with an internal large normative database previously obtained from 256 healthy volunteers.

 

[SNT Gatchaman]

Summary quotes from results (1) —

Ages ranged from 17 to 60 years (median 42; 40.9 ± 9.2 years). A female predominance was seen in this group (30/36, 83.3%).

All except 1 patient had suffered from mild COVID-19 infection that did not require hospitalization […] The 1 patient who required hospital admission had signs of COVID-19 pneumonitis, but did not require intensive care

Patients presented with a heterogeneous clinical picture, with multiple body systems involved. Cardiovascular symptoms […] (91.7%) […] A total of 15 out of 36 (41.7%) of these patients had received a formal diagnosis of postural tachycardia syndrome. Fatigue was the second most common symptom […] (86.1%). Headaches and migraines […] (61.1%), and brain fog […] (44.4%)

Gastrointestinal symptoms (23/36, 63.9%) and genitourinary symptoms (19/36, 52.8%) […] Somatosensory symptoms, such as paresthesia (21/36, 58.6%) and numbness (7/36, 19.4%) […] Finally, patients with pain were categorized into 3 main groups: (1) burning pain or painful dysesthesia, 11 out of 36 (30.6%) patients; (2) widespread deep aching pain consistent with fibromyalgia syndrome, 10 out of 36 patients (27.8%); and (3) combination of (1) and (2), 8 out of 36 patients (22.2%).

Large fiber nerve conduction studies were performed in 7 patients and were all reported as normal. Only 1 patient had skin biopsy, with normal IENFD. Two patients had been diagnosed with Lyme disease in the past, but they were asymptomatic for years before

Type 1 C nociceptors were the most frequent class, representing 209 of the recorded C fibers. Of these, 109 were type 1A mechano-sensitive units, and 95 were type 1B mechano-insensitive units. Interestingly, 5 type 1 units could not be classified as either 1A or 1B, and are therefore referred to as type 1?

The second most common class were type 4 sympathetic nerve fibers (48/266). Type 2 thermoreceptors (8/266) and type 3 C-low threshold mechanoreceptors (1/266) were less common, and their properties were not analyzed. […] there was a clear shift in the relative proportion of C nociceptors, with an increase in the frequency of type 1B mechano-insensitive units in long COVID patients

patients had a shift to the left in the histogram of fiber density recorded from a single intraneural site, with a high proportion of patients having only 1 or 2 fibers per recording site. […] the relative frequency of finding 1 or a maximum of 2 nerve fibers per recording site was 19.4% in healthy volunteers, as opposed to 41.1% in long COVID patients […] (Mann–Whitney, p = 0.0024). Interestingly, when we plotted the density of nerve fibers against long COVID duration, we did not observe any correlation

Microneurography recordings revealed abnormalities in most patients. Overall, there was evidence of small nerve fiber dysfunction in 32 out of 36 patients (88.9%). […] seen exclusively in somatosensory axons (including nociceptors) in 13 out of 36 patients (36.1%), exclusively in sympathetic fibers in 8 out of 36 patients (22.2%), and on both somatosensory and sympathetic nerve fibers in 11 out of 36 patients (30.6%).

 

[SNT Gatchaman]

Summary quotes from results (2) —

Characteristic ‘saw-tooth’ raster plots fulfilling the definition of spontaneous activity were present in 22 out of 36 patients (61.1%; Fig 3). Overall, 55 out of 209 (26.3%) C nociceptors showed signs of spontaneous activity in 22 out of 36 patients (61.1%). […] Spontaneous activity was recorded in 18 out of 109 (16.5%) of type 1A, 33 out of 95 (34.7%) of type 1B, and 4 out of 5 (80%) of type 1? C nociceptors. In comparison, only 1.5% of C nociceptors were engaged in spontaneous activity in 3.1% of the healthy volunteers (1.1% of type 1A, 2.5% of type 1B; Fig 4).

Overall, long COVID patients had a statistically significantly greater proportion of spontaneous units (Fisher’s exact test, p < 0.001) […] the combined presence of spontaneous activity, peripheral sensitization, and multispikes (24/36, 66.7%) was significantly higher in long COVID patients reporting any type of pain than in those without pain (22/29 [75.9%] vs 2/7 [28.6%]; p = 0.029

We observed signs of peripheral sensitization of C nociceptors in 10 patients (27.8%), all in type 1B fibers. Of all the type 1B C nociceptors, 34.0% responded to mechanical stimuli as low as 128 mN, often with long-lasting responses, which on many occasions persisted for several minutes, clearly outlasting the brief application of the mechanical stimulus (Fig 3).

Interestingly, none of the type 1A mechano-sensitive C nociceptors showed signs of mechanical sensitization, but 34.1% of them were in fact ‘mechano-insensitive’; that is, they did not respond even to the stronger 512-mN force. This was suggestive of either a loss of transductor properties or a dying-back of nerve terminals with ‘disconnection’ from the skin.

We detected multiple spikes in 4 patients (11.1%) in 4 C nociceptors; 1 was a type 1A mechano-sensitive unit, and 3 were Type 1B mechano-insensitive units (Fig 5). All the multispikes were double spikes. As a comparison, multispikes occurred in only 1.56% of the healthy volunteers.

Resting conduction velocity; that is, the conduction velocity of the first spike evoked by the electrical stimulus after a period of rest, was significantly faster in type 1A units in long COVID […] no statistically significant differences in resting conduction velocity in type 1B units

There were widespread changes in all the measures of ADS affecting both classes of C nociceptors. Surprisingly, the observed changes occurred in opposite directions in type 1A versus type 1B units (Fig 6). Type 1A mechano-sensitive C nociceptors slowed more during repetitive stimulation at 2 Hz, and recovered more slowly after it than type 1A units from healthy volunteers.

The changes in type 1B mechano-insensitive C nociceptors appeared like a mirror image […] they slowed less during the 2 Hz stimulation, and recovered faster after it. These opposite-direction changes were striking […] Finally, there was a change in the amount of slowing at very low frequencies, also in opposite directions between type 1A and 1B units. A significant proportion of type 1A units presented a paradoxical acceleration at 0.25 Hz, whereas type 1B units showed an abnormally pronounced slowing from rest to 0.25 Hz.

Spontaneous activity, peripheral sensitization, and multispikes frequently occurred in the same patients in different nerve fibers, or in the same patient in the same fiber.

Of the 36 patients with long COVID, 33 (91.7%) presented with dysautonomia, especially orthostatic intolerance. Of these, the majority (29/33; 87.9%) had abnormal electrophysiological findings in both somatosensory and sympathetic C fibers. This included 19 patients showing changes in the recovery cycle of excitability in sympathetic nerve fibers, 3 patients showing normal recovery cycles of excitability in sympathetic fibers, and 7 patients where no sympathetic fibers could be recorded.

The recovery cycle curves of long COVID patients characteristically had ‘absent’ or ‘flat’ supernormal periods, in sharp contrast to the recovery cycles from healthy volunteers

 

[SNT Gatchaman]

Summary quotes from discussion —

The exact mechanism underlying the generation of spontaneous activity remains unknown, but it is possible that slight increases in background depolarization or channel expression may render neurons more prone to spontaneous firing.

Another abnormality found in our long COVID patients was peripheral sensitization, with long-lasting bouts of ongoing activity triggered by a single mechanical stimulus […] can be induced by posttranslational modifications in transduction properties, membrane trafficking of ion channels, and neuroimmune modulation. There has been extensive discussion about the role of persistent inflammation in long COVID, and it is reasonable to think that some of these mechanisms could affect unmyelinated C fibers

some patients showed multiple spikes, which have been described as a possible mechanism for peripheral amplification of sensory inputs in peripheral neuropathy.

The multiple abnormalities in C nociceptors of long COVID patients, mirroring those in SFN and fibromyalgia, may suggest small nerve fiber dysfunction as a shared pathophysiological mechanism.

An intriguing finding also observed previously in fibromyalgia patients is the paradoxical acceleration of conduction velocity in type 1A fibers at very low rates of stimulation. The exact mechanism of ADS is debated and probably multifactorial. We know ADS is almost entirely a property of the final terminal arborization of the axons due to a cumulative effect of sodium channel slow inactivation, accumulation of intra-axonal sodium, and probably hyperpolarization by the activation of the Na+/K + pump.

There exists a fine interplay between Nav1.7 and Nav1.8 channels in C nociceptors. Nav1.8 is the main sodium channel responsible for the upstroke of the action potential and has more depolarized activation thresholds than Nav1.7, which probably exerts a major role during depolarization closer to the resting membrane potential. Nav1.9 channels have also been implicated in the excitability of C nociceptors, but because of their activating/inactivating kinetics, their role is probably to set the membrane potential toward more depolarized levels.

Another possible explanation for this paradoxical acceleration is that in the resting state, these fibers are less excitable due to inactivation of Nav1.7 caused by resting depolarization. However, upon stimulation, activity-dependent hyperpolarization may allow Nav1.7 to recover from inactivation. Because the channels have a hyperpolarized activation threshold, they become reactivated more easily, leading to a positive feedback loop that increases excitability.

It is also highly intriguing that the observed changes in ADS differed between type 1A mechano-sensitive and type 1B mechano-insensitive C nociceptors. […] Some human nociceptors that express the MRGPRD and CALCA genes can also express angiotensin-converting enzyme 2 mRNA.

We also detected abnormalities in the RCE of postganglionic sympathetic axons. The RCE reflects the interplay of multiple ionic mechanisms, including sodium channel availability, potassium conductance, and electrogenic pumping activity, which together shape how the fiber responds to prior activity, and it is highly sensitive to membrane potential.

The sympathetic fibers recorded in long COVID patients lacked a supernormal period, which can be explained by a relative membrane depolarization. Interestingly, there is also evidence of a reduction of supernormality in C nociceptors in a patient with erythromelalgia with a gain-of-function mutation of Nav 1.7, which might be due to a similar mechanism.

some patients experiencing spontaneous pain lacked spontaneous C nociceptor activity, whereas some pain-free patients showed such activity. Nevertheless, there was a strong correlation between the presence of microneurographic abnormalities that could plausibly underlie pain

the spatial and temporal characteristics of the peripheral afferent activity required to evoke pain perception may not be met in every patient at every time point. Central sensitization may also develop even in the presence of minimal peripheral activity, and low-level nociceptor firing in painfree patients may remain below the threshold for pain perception. Interestingly, such subthreshold nociceptor activity could contribute to nonspecific symptoms, such as brain fog or anxiety-related hypervigilance, as suggested by recent animal studies.

The broad range of affected systems—neurologic, cardiovascular, gastrointestinal, genitourinary, and musculoskeletal is striking. SARS-CoV-2 may trigger pathogenic mechanisms across multiple organ systems. A persistent immune or inflammatory response has been proposed […] However, an alternative explanation would be […] A generalized dysfunction of unmyelinated somatosensory and sympathetic nerve fibers would lead to widespread symptoms because of their almost universal body distribution, leading to ‘multiorgan symptoms,’ and representing a common underlying mechanism for the multiplicity of symptoms.

 

[Ryan31337]

I had a microneurography investigation with these guys, part of a POTS workup. It was pre-covid, I have a long history of ME/CFS & a negative IENFD skin biopsy.

Without understanding the specifics, my headline findings appeared the same as here: spontaneous activity & abnormal recovery cycles confirming SFN.

 

[Eleanor]

Trying to understand this bit by bit while brainfogged: Is the 'recovery cycle' what happens after the nerve has been stimulated, as it resets itself so it can respond to the next stimulus? So a prolonged recovery cycle could result in either an exaggerated response to one stimulus or a lack of response to the next one? (Or both?)

 

[Cinders66]

Baffling that small fibre neuropathy, that’s been talked of for a while, wasnt rushed through for ME/CFS investigation anywhere. I look forward to kings college switching their narratives from "biopsychosocial of no organic cause" to this type of research.

 

[Cinders66]

I had a microneurography investigation with these guys, part of a POTS workup. It was pre-covid, I have a long history of ME/CFS & a negative IENFD skin biopsy.

Without understanding the specifics, my headline findings appeared the same as here: spontaneous activity & abornomal recovery cycles confirming SFN.

Wow, I'd love to have investigations but being bedridden ain't going to happen. Did they prescribe any medication & did it help please?

 

[Ryan31337]

Wow, I'd love to have investigations but being bedridden ain't going to happen. Did they prescribe any medication & did it help please?

Yes - I'll drop you a PM so as not to go off-topic

 

[Cinders66]

Yes - I'll drop you a PM so as not to go off-topic

Much appreciated thank you!

 

[Jonathan Edwards]

If there is a new better method to detect small Fibre nerve malfunction that would be very welcome. I have had peripheral neuropathy symptoms since Covid which may be old age and entrapment but something seems to have changed.

I am planning to get a consult from a neuroimmunology friend around Xmas time partly to try and raise interest in MECFS. I may know more then.

 

[V.R.T.]

'A persistent immune or inflammatory response has been proposed […] However, an alternative explanation would be […] A generalized dysfunction of unmyelinated somatosensory and sympathetic nerve fibers would lead to widespread symptoms because of their almost universal body distribution, leading to ‘multiorgan symptoms,’ and representing a common underlying mechanism for the multiplicity of symptoms.'

They don't seem to consider the idea that the immune response is causing the nervous system dysfunction.

 

[Utsikt]

I have had peripheral neuropathy symptoms

What kind of symptoms are that?

 

[jaded]

Like @Ryan31337 I’ve also had microneurography done pre-covid at King’s College Hospital with Dr Jordi Serra referred to by Dr Nick Gall. My results were similar. In particular it was reported that my sympathetic and c-fibre nociceptors were partially depolarised. Which makes total sense as my nerves fire excessively with little input in terms of widespread nerve pain, pots/OI and excessive sweating etc. My sudomotor nerves were also going bonkers apparently. Nice to see it objectively on the screen.

Unfortunately covid made my SFN even worse and now it’s spread to my cranial nerves and given me erythromelalgia/flushing in ears, chest and face and vestibular migraine. I’ve had all of the classic ME symptoms post virally for years but they try and suggest all of my symptoms are due to this one “idiopathic cause”.

 

[jaded]

I am planning to get a consult from a neuroimmunology friend around Xmas time partly to try and raise interest in MECFS. I may know more then.

I’ve tried to get nhs support from neuroimmunology despite all of my objective tests including microneurography and a skin biopsy, CPET at Papworth etc but because the underlying cause isn’t known e.g secondary to lupus or vasculitis they’re not interested.

Even when referred to by a very well informed private neurologist (who was an MS NHS lead for years) who feels confident it’s neuroimmune driven as it’s worsened with several distinct immune challenges over the years.

However I appreciate the importance of spreading awareness of ME among such colleagues/friends.

 

[Jonathan Edwards]

What kind of symptoms are that?

Tingling, numbness, sensitivity, a sense of loss of dexterity using fingers and touch registering as discomfort.