Fibromyalgia patients and HCs were recruited from a single university-based hospital, Bucheon St. Mary’s Hospital. The inclusion criteria for fibromyalgia patients were as follows: (1) fulfilled all 1990, 2010, and 2016 American College of Rheumatology classification criteria for fibromyalgia [1,2,13], (2) over 18 years of age, and (3) disease duration of less than 5 years. Fibromyalgia patients who had concomitant inflammatory arthritis (e.g., rheumatoid arthritis or spondyloarthritis), other systemic autoimmune diseases (e.g., systemic lupus erythematosus, primary Sjogren’s syndrome, systemic sclerosis, inflammatory myositis, or mixed connective tissue disease), or cancer were excluded. As a control group, age- and sex-matched HCs were included. HCs who had potentially pain-inducing disorders such as osteoarthritis, inflammatory arthritis, systemic autoimmune diseases, cancer, and psychological disorders (e.g., major depression disorder or somatoform disorder) were excluded. Both fibromyalgia patients and HCs underwent SUDOSCAN, EMG, and NCS on the day of enrolment. All enrolled patients with fibromyalgia were naïve to fibromyalgia medication, which was started after the aforementioned examinations were performed. A good response to tricyclic antidepressants (TCAs) was defined as a decrease in the fibromyalgia impact questionnaire (FIQ) by more than 30% or a score below 40 after 1 month of TCA administration. Other fibromyalgia medications were allowed in cases of non-response to TCA therapy. The study was conducted in accordance with the Declaration of Helsinki (1964, and its later amendments). Written informed consent was obtained from each participant before enrolment. The study was approved by the Institutional Review Board of the Catholic University of Korea, Bucheon St. Mary’s hospital (approval number: HC15DISI0040).

The clinical, demographic, and laboratory data of fibromyalgia patients and HCs were collected at the time of enrolment. The age, sex, body mass index (BMI), serum creatinine level, and estimated glomerular filtration rate of fibromyalgia patients and HCs were collected. To assess quality of life, EQ5D and EQ5D visual analogue scale were used [14]. Several conventional tools for evaluating neuropathy other than NCS, EMG, and SUDOSCAN were also used. The Toronto Clinical Neuropathy Score was used to measure the symptom degree of sensorimotor polyneuropathy [15]. Vibration perception threshold by biothesiometer and deep pressure pain perception threshold by algometer were measured in the distal plantar surface of great toe/abductor pollicis brevis or thenar eminence/abductor halluces/nail bed of first toe. Serum levels of neurotransmitters (serotonin, substance P, and tryptophan) were measured with enzyme-linked immunosorbent assay for all participants on the day of enrollment. The FIQ [16] and fibromyalgia disease duration were only collected in fibromyalgia patients.

Viking Electromyography System (Nicolet-Viasys Biomedical Inc., Madison, WI, USA) was used for NCS with the following settings: 20 Hz for the lower filter, 3 kHz for the upper filter for sensory studies, 5–10 kHz for motor studies, and a pulse duration of 0.1 ms [9]. NCS tests included the sural sensory, superficial peroneal sensory, median sensory, ulnar sensory, radial sensory, tibial motor, and median motor nerves of one side. The minimum F-response latencies from both tibial nerves and H-reflex latencies from tibial nerve stimulation were collected. EMG needling was done by inserting the monopolar needle into selected lower extremity muscles (tibial and median motor nerves). NCS of bilateral medial plantar (MP) sensory, dorsal sural (DS) sensory, and medial dorsal cutaneous (MDC) sensory nerves were also conducted to evaluate distal sensory neuropathy, which is known to show abnormal findings in early diabetic neuropathy [17]. All NCS and EMG tests were performed by experienced physicians.

SUDOSCAN was performed in fibromyalgia and HC groups to measure ESC of hands/feet and CAN-RS on the same day that NCS and EMG were performed. A separate examiner who was blind to the patient’s information performed SUDOSCAN before NCS and EMG. SUDOSCAN is a simple and non-invasive procedure that takes only 3 minutes. The detailed procedure of SUDOSCAN was described in a previous study [9]. ESC, expressed in microsiemens (μS), was calculated by analyzing electrochemical currents generated in response to low-voltage stimulation. Lower ESC indicated sweat gland dysfunction (sudomotor dysfunction), which could be interpreted as positive for SFN. Patients were divided into three groups according to the following cut-off values of the mean ESC (mESC) for hands and feet: (1) no SFN (hands mESC ≥ 60 μS and feet mESC ≥ 70 μS), (2) moderate SFN (hands mESC 40–60 μS and feet mESC 50–70 μS), and (3) severe SFN (hands mESC < 40 μS and feet mESC < 50 μS) [18]. The CAN-RS is presented as percentage and automatically calculated by the algorithm included in the SUDOSCAN device software. ESC, BMI, and age are used for calculating the CAN-RS, and a higher CAN-RS indicates a higher risk of cardiac autonomic abnormalities.

deviation or median and interquartile range after a normality test of each continuous variable was performed by the Shapiro–Wilk test. Categorical variables were presented as number and percentage, and significant differences between the fibromyalgia and HC groups were determined by the chi-square or Fisher’s exact test. Correlations between ESC versus fibromyalgia disease duration and serum neurotransmitter levels were measured by the Spearman correlation coefficient. p values < 0.05 were considered statistically significant. All analyses were performed using SPSS statistics software (version 23; IBM Corp., Armonk, NY, USA).

A total of 44 participants were enrolled, of which 22 were fibromyalgia patients, and 22 were HCs. All enrolled participants were female. The mean ages of the HC and fibromyalgia groups were 49.5 ± 12.2 years and 51.1 ± 8.7 years, respectively. BMI, serum creatinine, and estimated glomerular filtration rate were comparable between the HC and fibromyalgia groups. Regarding serum neurotransmitter levels, the fibromyalgia group exhibited non-significantly lower serotonin levels (139.1 ± 75.1 ng/mL vs. 157.3 ± 69.4 ng/mL, p = 0.430), significantly higher tryptophan levels (16.4 ± 3.5 μg/mL vs. 13.9 ± 3.1 μg/mL, p = 0.023), and non-significantly higher substance P levels (7.0 ± 8.7 ng/mL vs. 5.4 ± 7.6 ng/mL, p = 0.550) than the HC group. EQ5D and EQ5D visual analogue scale scores were significantly lower in the fibromyalgia group than in the HC group, which implies lower health-related quality of life in fibromyalgia patients. Both the total score and each component of the Toronto Clinical Neuropathy Score (symptom, reflex, and sensory test scores) were significantly higher in the fibromyalgia group. The vibration perception threshold and deep pressure pain perception threshold values did not significantly differ between HC and fibromyalgia groups. All demographic, laboratory, and clinical data are summarized in Table 1.

The NCS of proximal sensory nerves (including the right side sural sensory nerve and right superficial peroneal sensory nerve) demonstrated lower sensory nerve action potential (SNAP) amplitude in the fibromyalgia group than in the HC group (11.9 ± 7.1 μV vs. 16.8 ± 6.1 μV, p = 0.021, for the right sural sensory nerve and 9.3 ± 4.4 μV vs. 13.9 ± 3.7 μV, p < 0.001, for the right superficial peroneal sensory nerve). The SNAP amplitudes of the right median sensory wrist and right median sensory palm in the fibromyalgia group tended to be lower than those in the HC group, but without statistical significance. The EMG findings regarding onset latency (ms), distal compound muscle action potential amplitude (mV), conduction velocity (m/s), and F-waves on the right tibial motor and right median motor nerves were comparable between the HC and fibromyalgia groups (Table 2). These NCS and EMG findings demonstrate that neuropathies on proximal sensory and motor nerves are not observed more definitely in fibromyalgia than in HC patients.

Next, we performed NCS on distal sensory nerves because most asymptomatic and early neuropathies first occur on distal sensory nerves prior to proximal sensory or motor neuropathies [17,19]. The SNAP amplitudes of bilateral distal sensory nerves, including MP, DS, and MDC nerves, were significantly lower in the fibromyalgia group than in the HC group. Additionally, onset latencies of right MP, left MDC, and left DS nerves were significantly higher in the fibromyalgia group (Table 3). This implies that distal sensory neuropathies are more frequently found in fibromyalgia patients than HCs, even in fibromyalgia patients without definite neuropathies on proximal sensory or motor nerves.

The mESC values were calculated in bilateral hands and feet. In the HC and fibromyalgia groups, the hands mESC values were 68.8 ± 10.3 μS and 57.6 ± 16.2 μS (p = 0.010), and the feet mESC values were 72.0 ± 8.2 μS and 64.9 ± 11.5 μS (p = 0.025), respectively. In the fibromyalgia group, ten patients (45.5%) exhibited moderate SFN, and five patients (22.7%) exhibited severe SFN. In the HC group, 11 individuals (50.0%) exhibited moderate SFN and none exhibited severe SFN. The CAN-RS values exhibited a tendency to be higher in the fibromyalgia group than in the HC group; however, this was not statistically significant (Table 4).

Next, we evaluated correlations between hands/feet mESC and fibromyalgia disease duration and serum neurotransmitter levels in fibromyalgia patients. Spearman correlation coefficients revealed that hands/feet mESC was negatively correlated with fibromyalgia disease duration (Rho = −0.636, p = 0.006, and Rho = −0.726, p < 0.001, respectively, Fig. 1). Additionally, hands/feet mESC was negatively correlated with the tryptophan-to-serotonin ratio (Rho = −0.464, p = 0.039, and Rho = −0.564, p = 0.010, respectively), and serum serotonin levels were positively correlated with hands/feet mESC (Rho = 0.484, p = 0.031, and Rho = 0.512, p = 0.021, respectively). The total FIQ score, widespread pain index, and pain VAS score did not show a significant correlation with the mESC score of the hands or feet. Only the waking component of the FIQ and symptom severity score showed a significant correlation with hand mESC values (Supplementary Table 1).

We could only evaluate the TCA treatment responses of 20 out of 22 fibromyalgia patients, because follow up data was lacking for two fibromyalgia patients. Among the 20 patients, eight patients were TCA responders and 12 patients were non-responders. The mESC values of hands and feet were significantly higher in TCA responders, and the non-responder group showed a higher prevalence of moderate to severe SFN (11/12, 91.7%) than the TCA responder group (3/8, 37.5%, Table 5).