Sympathetic skin response in fibromyalgia patients with small fibre pathology Authors: Elena Ammendola, Giulia Paparella, Emmanuella Ladisa, Giusy Tancredi, Giulia Paparella, Grazia Devigili, Marianna Diletta Delussi, Katia Ricci, Livio Clemente, Marina de Tommaso Published: 2026 ABSTRACT AbstractObjective: Fibromyalgia (FM) is a chronic disease characterised by small fibre pathology and autonomic dysfunction. Sympathetic Skin Response (SSR) can be easily obtained in clinical practice as a marker of autonomic dysfunction. This study aims to assess SSR abnormalities in small fibre pathology associated with FM.Methods: This was a retrospective, observational, single-centre, cross-sectional study including 393 patients with FM, who underwent skin biopsy from the thigh and ankle, SSR obtained with a standard electromyographic active electrode attached to the right palm and sole, and clinical evaluation.Results: SSR at the hand and foot was similarly absent in patients with both normal and abnormal skin biopsy results. Hand amplitude was reduced in patients with denervation (Wilcoxon z = 1.89, p = 0.029), while latencies were similar between groups (z = 1.12, p = 0.13). Foot latency was prolonged in patients with abnormal skin biopsy (z = 2.17, p = 0.014), while amplitudes were similar between groups (chi-square = 2.28, p = 0.21). SSR showed low predictive performance for skin biopsy results (r2 = 0.04). Latency of SSR at the foot and Intraepidermal Nerve Fibre Density (IENFD) correlated with the score for associated symptoms. Amplitude of SSR at the foot also negatively correlated with anxiety, depression, and fatigue scores.Conclusions: SSR abnormalities did not correspond to small fibre impairment detected by skin biopsy in patients with FM.Significance: SSR and skin biopsy may capture different aspects of the complex clinical picture of FM, including psychopathological features and fatigue, and therefore represent complementary yet distinct biomarkers of the disease.Keywords: Fibromyalgia, Sympathetic skin response, Autonomic dysfunction, Small fibre pathology Introduction Fibromyalgia is a chronic disease characterised by widespread pain, fatigue, cognitive impairment, sleep disturbance, and associated symptoms such as headache, bowel dysfunction, and depression (Wolfe et al., 2016). Vegetative impairment is a well-known feature of fibromyalgia, with changes in swallowing, breathing, and cardiac rhythm adaptation, which are frequently not attributed to a primary autonomic system disorder, but rather to stress-related dysfunction (Papadopoulou et al., 2022). In a study conducted by our group involving 50 FM patients, we found that SSR latency was increased in FM compared to controls, and that it correlated with anxiety scores and fibromyalgia-related disability. In a significant proportion of patients with neuropathic pain symptoms, as assessed by the DN4 (Douleur Neuropathique 4), the SSR was absent or poorly recognizable (de Tommaso et al., 2017). However, small fibre pathology is present in about 50% of FM patients and may cause changes in autonomic functions (Marshall et al., 2025), although the clinical phenotype of FM does not resemble that of small fibre neuropathy (SFN) (Leone et al., 2023).There are neurophysiological biomarkers that could be easily used in clinical practice to assess possible autonomic dysregulation in FM, such as heart rate variability (HRV) (Ladisa et al., 2025) and sympathetic skin response (SSR) (Martínez-Martínez et al., 2014). A recent study on neuropathic pain symptoms and signs in FM found that a significant number of patients presented with absent SSR, reduced amplitude, or increased latency (Elsawy et al., 2025). However, the SSR abnormalities are not sensitive in small fibres diseases, in respect to other neurophysiological tests (Lefaucheur et al., 2015), and recent recommendation about neurophysiological and clinical assessment of painful neuropathies, included only skin biopsy among the most reliable procedures (Truini et al., 2023). In fact, SSR reflects a complex reflex that is highly sensitive to non-specific factors such as arousal and emotional state in the context of pain. (Salameh et al., 2022). However, SSR is easily obtained in clinical practice (Devigili et al., 2023), so the present study aims to re-evaluate SSR features in a group of FM patients in light of skin biopsy results to assess its predictive value for small fibre pathology associated with the disease. The primary outcome of the study was the comparison of SSR parameters in patients with and without skin biopsy abnormalities, also considering the group of patients with distal small fibre denervation. The secondary outcome was to correlate the main SSR characteristics with the degree of epidermal nerve fibre density and clinical aspects of the disease, including pain diffusion, disease-related disability, associated symptoms, anxiety, and depression. Methods Participants and clinical assessment This was a retrospective, observational, monocentric, cross-sectional study. Written informed consent for the use of anonymised data for scientific purposes was obtained from all participants. The Ethical Committee of Bari Policlinico General Hospital approved the informed consent and the use of retrospective data for the study (Ethical Committee of Bari Policlinico General Hospital, Comitato Etico Indipendente Locale N°1076/CE). The study was conducted in accordance with the Declaration of Helsinki. We included consecutive FM patients who attended their first visit at our Chronic Pain Centre from 2018 to 2024, were diagnosed according to Wolfe et al. (2016) criteria, and underwent skin biopsy. Exclusion criteria for neurophysiological assessment were: age under 18 years, comorbidity with any cause of small fibre pathology other than FM, presence of electroneurography abnormalities suggesting concomitant sensory polyneuropathy, current therapy with CNS-acting drugs, psychiatric diagnosis of psychosis or major depression according to DSM-V (American Psychiatric Association, 2022), or inability to sign the informed consent. Patients who attended our Centre underwent a neurological examination using the ACR 2016 criteria (Wolfe et al., 2016), which include the Widespread Pain Index (WPI) and Symptom Severity (SS) scales. We also used the Fibromyalgia Impact Questionnaire (FIQ) (Bidari et al., 2014), Zung's anxiety and depression scales (Zung, 1971, Zung, 1965), and the Multidimensional Assessment of Fatigue (Belza, 1995). The Numerical Rating Scale from 0 to 10 was used to record the subjective impression of pain during the past week. SSR recording The SSR was studied using the standard method (Shahani et al., 1984). Skin temperature was maintained at 32 °C, with room temperature stabilised at 25–26 °C. A standard electromyographic active electrode was attached to the right palm and sole, and the reference electrode to the dorsum of the hand and foot. The stimuli were single electrical pulses applied to the right wrist at 0.01 A and 0.100 s duration. This stimulation procedure was standardised in previous studies on FM and correlated syndromes (Devigili et al., 2008; Elie and Guiheneuc, 1990). Stimuli were delivered unexpectedly at random intervals between 30 and 60 s. Five consecutive stimuli were administered. Latency was measured from the onset of the stimulus artefact to the onset of the first negative deflection and expressed in seconds. Amplitude was measured from the baseline to the maximal negative peak and expressed in μV. The response was considered absent if no consistent voltage change occurred using a sensitivity of 50 uV per division after three trials at maximum stimulus intensity. Response latencies were considered pathological when exceeding two standard deviations above the mean latency of the control group. Skin biopsy All patients were offered skin biopsy. According to previous studies (Vecchio et al., 2022), the biopsy was performed using a 3-mm disposable punch on the thigh, 20 cm below the anterior superior iliac spine, and on the distal leg (10 cm above the lateral malleolus in the area of the sural nerve), after intradermal injection of 1% xylocaine. Samples were fixed in 2% paraformaldehyde lysine sodium periodate at 4 °C overnight, cryopreserved, serially sectioned with a cryostat into 50-μm-thick sections, and immunostained free-floating with polyclonal anti-protein gene product 9.5 (Ultraclone Ltd) according to a standardised protocol. The methodology was described in detail in Devigili et al. (Devigili et al., 2008) and Lauria et al. (Lauria et al., 2010). IENFD was calculated on three non-consecutive central sections by brightfield microscopy using a stereology workstation (Olympus BX50, PlanApo oil-objective 40×/NA = 1.0) and compared with sex- and age-adjusted normal values. We divided FM patients into three groups according to skin biopsy results: normal skin biopsy, IENFD reduction at the proximal site, and IENFD reduction at both proximal and distal sites. Statistical analysis The Shapiro test showed that SSR data and IENFD density data were not normally distributed. The primary outcome was split into two analyses: the comparison of latency and amplitude of SSR at the foot and hand between the three groups, performed with the Kruskal- Wallis test, and the contingency analysis with the chi-square test, considering the absence or presence of SSR at the hand and foot. An a priori sample size calculation was performed using G*Power (version 3.1). For the omnibus comparison across three independent groups using a one-way fixed-effects analysis, we assumed an effect size of Cohen's f = 0.20, a two-sided significance level of α = 0.05, and a desired power of 0.95. The required total sample size was N = 390. In addition, for the comparison of a binary categorical variable – absence or presence of SSR at the upper and lower limb – across the same three groups (3 × 2 contingency table, df = 2), we assumed Cohen's w = 0.163, α = 0.05, and power = 0.80, yielding a required total sample size of N = 363. To ensure adequate power for both analyses, we planned to recruit the larger sample size (N = 390).A potential risk of bias is the differing number of FM patients in the three skin biopsy groups. We also compared the number of patients with SSR parameters differing from normative values using the chi-square test. Using the same number of cases for comparison between two groups with and without small fibre denervation ensured high statistical power (>95%) to detect small-to-moderate effect sizes (Cohen's d ≥ 0.3) in SSR parameters between groups at α = 0.05. For the secondary outcome, we used non-parametric correlation (Kendall test) to assess the relationship between SSR amplitude and latency with IENFD and main clinical features (FIQ, MAF, anxiety and depression scores, WPI, SS, NRS). As an additional analysis, we employed a neural network model to explore the predictive performance of the main SSR features in relation to skin biopsy abnormality. JMP software, 2025 Student version, was used. Results Participants and skin biopsy results Among 765 patients who attended our Pain Centre during the study period, 681 were diagnosed with FM. Of these, 140 were excluded from the present evaluation due to concurrent CNS-acting drug therapy at their first visit (including 25 with major depression or psychosis), 120 had comorbidities with possible causes of small fibre pathology, such as diabetes, renal insufficiency, or paraneoplastic diseases, and 28 did not give consent for skin biopsy and/or neurophysiological examination. We therefore analysed data from 393 patients (322 females, mean age 49.18 ± 12.32 years). Skin biopsy showed reduced IENFD at the proximal site in 221 patients; 32 patients showed denervation at both proximal and distal sites; 140 patients had normal skin biopsy results. The IENFD was 11.26 ± 3.63 at the proximal site (15.3 ± 2.11 in the normal skin biopsy group, 9.32 ± 2.40 in the proximal denervation group, 8.26 ± 2.12 in the distal denervation group; ANOVA p < 0.001) and 8.40 ± 3.18 at the distal site (10 ± 3 in the normal skin biopsy group, 7.98 ± 2.72 in the proximal denervation group, 4.35 ± 1.9 in patients with distal denervation; ANOVA p < 0.001). Sympathetic skin response Among control subjects selected for SSR laboratory normative data, we included 30 healthy individuals matched for sex and age with the FM patients (26 females, mean age 47.43 ± 14.6 years). Normal values were 1294.64 ms ± 333.65 at the hand and 1989.05 ± 373.69 ms at the foot for latency, and 263.84 μV ± 214 at the hand and 98.45 ± 70.8 μV at the foot for amplitude [6]. The SSR was absent in both the hand and foot in one control subject, in 54 FM patients at the hand (chi-square 4.37, p = 0.036), and in 159 patients at the foot (chi-square 25.74, p < 0.0001). Primary outcome: Comparison of SSR across the 3 FM groups Taking into account the macro group with small fibre denervation, absence of SSR at the hand and foot was equally represented in both groups (SSR hand: chi-square 1.87, p = 0.17; SSR foot: chi-square 1.03, p = 0.3).Considering the presence of proximal and/or distal small fibre denervation, the absence of SSR at the hand was similar across the three FM groups: those with normal skin biopsy, proximal denervation, and both proximal and distal denervation (chi-square 1.9, p = 0.38). The absence of SSR at the foot also did not differ among the three groups (chi-square 1.60, p = 0.58) (Fig. 1). Distribution of fibromyalgia (FM) patients with absent sympathetic skin response (SSR) at the hand A) and at the foot B) among the 3 subgroups (normal skin biopsy, proximal denervation and proximal and distal denervation). Mean latency values were 1383.35 ± 211 ms at the hand and 2020 ± 388 ms at the foot. Assigning a value of 0 to absent SSR, amplitudes were 0.23 ± 0.21 μV at the hand and 0.074 ± 0.09 μV at the foot. Considering the two groups, with and without skin biopsy anomalies, hand amplitude was reduced in patients with denervation (Wilcoxon z = 1.89, p = 0.029), while latencies were similar between the groups (z = 1.12, p = 0.13). Foot latency was prolonged in patients with abnormal skin biopsy (z = 2.17, p = 0.014), while amplitudes were similar between the groups (z = 2.28, p = 0.21). Dividing the analysis into three groups, foot latency was significantly prolonged in patients with proximal denervation compared to those with normal skin biopsy (Kruskal-Wallis test: chi-square 5.58, p = 0.06; Wilcoxon test between normal skin biopsy and proximal denervation groups: z = 2.33, p = 0.019). In the hand, patients with denervation at both proximal and distal sites showed reduced amplitude compared to those with normal skin biopsy (Wilcoxon 4.31, p = 0.11; normal skin biopsy vs proximal plus distal denervation z = 2.30, p = 0.020). (Fig. 2, A,B; Fig. 3, A,B). Amplitudes of sympathetic skin response (SSR) in the three fibromyalgia (FM) groups, divided by normal skin biopsy, proximal denervation, and proximal and distal denervation at the hand (A) and at the foot (B). In (A), values for patients with distal denervation were significantly lower than those with normal skin biopsy (p = 0.020). Latencies of sympathetic skin response (SSR) in the 3 fibromyalgia (FM) groups, divided for normal skin biopsy and proximal and proximal and distal denervation at the hand A) and at the foot B). In B) values of patients with proximal denervation were significantly longer than those with normal skin biopsy (p = 0.019). Other analyses: among patients with valid SSR, 25 had latency values exceeding normal limits at the hand, and 30 at the foot. The number of patients with abnormal amplitude and latency was not distributed differently among the three skin biopsy groups. Training a neural network to classify patients with and without abnormal skin biopsy, considering absence, latency, and amplitude of SSR at the foot, resulted in low predictive performance in the validation session (r2 = 0.04) (Fig. 4). Receiver operating characteristic (ROC) curve on validation data for predicting abnormal skin biopsy from latency and amplitude of sympathetic skin response (SSR) at the foot (r2 0.04). Secondary outcome: Correlation between SSR characteristics with skin biopsy and clinical aspects of the disease The Kendall test showed significant correlations between SSR latency at the foot and the score for associated symptoms, which were also correlated with denervation at both proximal and distal sites. SSR amplitude at the foot was negatively correlated with anxiety, depression, and fatigue scores. SSR latency and amplitude and IENFD were not correlated (Table 1). Table 1Non-parametric correlation Kendall test between results of intraepidermal nerve fibre density (IENFD) at proximal and distal sites, latency and amplitude of sympathetic skin response (SSR), and clinical variables (SS associated symptoms, SAS Zung anxiety score, SDS Zung depression score, MAF: Multidimensional Assessment of Fatigue). Only significant results are reported.IENFD proximalSS−0.13270.0010IENFD distalSS−0.09000.0250SRR Latency FootSS0.10940.0471SSR Amplitude - FootSAS−0.10050.0278SSR Amplitude - FootSDS−0.12330.0071SSR Amplitude - FootMAF−0.09700.0416 Non-parametric correlation Kendall test between results of intraepidermal nerve fibre density (IENFD) at proximal and distal sites, latency and amplitude of sympathetic skin response (SSR), and clinical variables (SS associated symptoms, SAS Zung anxiety score, SDS Zung depression score, MAF: Multidimensional Assessment of Fatigue). Only significant results are reported. The main clinical, skin biopsy, and neurophysiological data are summarised in the Supplementary Table (Supplementary Data). Discussion The present analysis in a large group of FM patients confirmed that, in a significant number of cases, SSR was absent in both the hand and the foot, with particularly notable results for the foot compared with normal subjects. Previous studies were conducted in small case series, with Papadopoulou et al. (2022) reporting SSR latency and amplitudes similar between patients and controls, and our previous study (de Tommaso et al., 2017) finding increased SSR latency. The present results, obtained in a large number of patients, indicated that while absent SSR is expected, especially in the foot, amplitude and latency values fall outside normal limits in only a minority of cases, due to the variability of these features in normal controls (Elie and Guiheneuc, 1990). However, our normal ranges were similar to those described in the literature, and differences in normal amplitude values were due to our measurement from the baseline to the first peak, rather than using peak-to-peak values. Primary outcome: Comparison of SSR across the 3 FM groups Considering the criterion of absence or presence of SSR, we did not observe significant differences between patients with normal or abnormal skin biopsy results, including those with distal or both distal and proximal denervation. To date, the origin of small fibre impairment in FM and the usefulness of performing skin biopsy remain subjects of debate (Sommer and Üçeyler, 2025). The general consensus is that FM has a more severe outcome in patients with small fibre pathology (Quitadamo et al., 2023), which should, in any case, be distinguished from small fibre neuropathy, characterised by a typical distal distribution of pain and other sensory symptoms (Leone et al., 2023; Sommer and Üçeyler, 2025). Skin biopsy has been recommended for patients with FM, along with other procedures to investigate autonomic impairment and conduction along thermo-nociceptive pathways, such as laser-evoked potentials, heart rate variability (HRV) (Ladisa et al., 2025) , and SSR (Devigili et al., 2023).The present results confirm that SSR and skin biopsy findings do not correspond, but likely reflect different aspects of the complex clinical picture of FM. While sudomotor dysfunction is common in patients with small fibre neuropathy (Kool et al., 2024), SSR is not recommended as a criterion for diagnosing small fibre pathology (Cheshire, 2020) due to its variability and dependence on both central and peripheral factors.(Devigili et al., 2019). In fact, patients in the groups with abnormal skin biopsy showed longer latency and lower amplitude compared to those with normal IENFD, although the comparison among the three groups with normal skin biopsy, proximal denervation, and both proximal and distal denervation indicated significant results only in the post-hoc comparison and not in the overall test, which reduced the statistical relevance. As an adjunctive analysis, we trained a neural network that considered skin biopsy results and main SSR features. The predictive value of SSR for the skin biopsy result was very low, confirming that the two tests may be complementary but not substitutive, at least in the evaluation of small fibres and autonomic impairment in FM.. The lack of correspondence between the SSR pattern and skin biopsy confirms that SSR is a complex response linked to vegetative arousal in conditions of pain, so it is not directly connected to peripheral small fibre involvement but is highly influenced by central autonomic circuits (Salameh et al., 2022). Secondary outcome: Correlation between SSR characteristics with skin biopsy and clinical aspects of the disease We found that some SSR features could reflect complex aspects of FM, such as anxiety and depression, with scores correlating with reduced amplitude of reflex responses. In general, anxious patients present with higher arousal and increased SSR, while in depression the reflex responses appear markedly inhibited or differently modulated compared to normal controls (Markiewicz et al., 2022; Moscovitch et al., 2010; Romano et al., 2017). In our previous study on SSR in a small group of FM patients, we found that anxiety was correlated with increased SSR latency (de Tommaso et al., 2017), a finding not confirmed in the present, larger FM group. We can now confirm that psychopathological aspects of FM may interact in a complex manner with small fibre impairment, resulting in abnormal autonomic reactions and arousal responses to external stimuli. Impairment in vegetative behaviour was also correlated with the fatigue score, which is reasonable, as fatigue is a complex symptom caused by pain but also facilitated by the inability to adapt to motor and cognitive demands (Tanaka et al., 2015). Both increased SSR latency and reduced IENFD correlated with the scores of associated symptoms of FM, calculated based on sleep disorders, fatigue, brain fog, headache, depression, and bowel discomfort, confirming that they may be considered distinct biomarkers of the complex clinical presentation of FM. Interestingly, neither IENFD nor SSR showed any correlation with pain intensity or its distribution in the body, suggesting that bothare mainly related to the associated symptoms that comprise FM syndrome. Limitations Although the number of patients was notable, the study was retrospective and conducted during routine clinical practice. Therefore, we did not have a comparable number of control subjects and instead used the normative values from our laboratory, which, for a variable test such as the SSR, could underestimate SSR latency and amplitude abnormalities in the patient group.Another cause of bias could be treating absent responses as zero values in amplitude analyses, as this calculation may not be accurate.The three groups of patients identified by skin biopsy have different numbers of cases, with few cases in the group with distal IENFD reduction.For technical issues, present data focused on a single neurophysiological procedure for autonomic functions, such as the SSR. Other tests for small fibres, such as laser-evoked potentials and heart rate variability, could also be analysed in the future to assess possible correspondence with skin biopsy results. Conclusions The SSR abnormalities did not correspond with small fibre impairment as shown by skin biopsy in FM patients, but instead appear to reflect different aspects of the complex clinical picture, including psychopathological features and fatigue. With caution due to the retrospective nature of the present data, SSR and skin biopsy could be considered distinct biomarkers of the symptom constellation, which could potentially be used to monitor disease progression and treatment effects. Declaration of generative AI and AI-assisted technologies in the manuscript preparation process Artificial intelligence provided by Google was used to search for references, and all selected studies were verified by the authors. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements The authors would like to express their appreciation to all the participants who took part in this study. Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.cnp.2026.06.001. Supplementary data Supplementary material References Associated Data Supplementary Materials Supplementary material