Main Findings of Our Study
The American Academy of Neurology summary of evidence-based guidelines for clinicians recommends that the combination of autonomic screening tests in the CASS should be considered to achieve the highest diagnostic accuracy of CAN [
26]. To the best of our knowledge, our work is the first study to adopt cardiovagal and adrenergic subscores of the CASS as a measure of CAN, and examine the feasibility of ESC as a screening service of CAN in patients with PD.
Our data showed significantly different clinical scales among the three subgroups of patients with PD. The PIGD group had the worst function, the TD group the best, while the AR group was placed in between. The finding was consistent with the original report suggesting such phenotype classification [
3]. Our data did not show a similar difference of autonomic function among these subgroups. Although the TD subgroup had higher values for all autonomic parameters and lower CASS, suggesting better autonomic function than the other two subgroups, the difference was not statistically significant. Further study is warranted to answer the question of whether autonomic function differs among different phenotypes of PD. This study also showed the severity of CAN, demonstrated by CASS, neither significantly correlated with clinical scores nor the ESC values.
There are only a few clinical studies using Sudoscan to evaluate CAN in patients with PD. The study by Xu et al. enrolled 43 patients having later stage PD (mean UPDRS score = 38.5 ± 9.0) with mean age 66.0 ± 8.9 years, and 42 healthy controls [
18]. This study showed that ESC of the limbs, especially the hands, was significantly lower in patients with PD than in controls, and was significantly correlated with the results of the scale of autonomic symptoms (SCOPA-AUT) in PD. The report by Popescu et al. enrolled 67 patients with PD (mean UPDRS score = 21.2 ± 11.9) whose mean age was 74.3 years, and 66 aged-match controls [
17]. This study showed no significant reduction in ESC in patients with PD compared with controls, and further found weak correlation between foot ESC and modified Hoehn and Yahr Scale. Our study enrolled 63 patients with PD (the median UPDRS score was 18.0) whose mean age was 65.4 ± 9.8 years. According to our results, there was no significant difference of either hand or foot ESC between patients with CAN and without CAN. Our patients were in a relatively early stage, and their cardiovascular impairment was mild in severity. The discrepancy among these three studies may be attributed to the different methodologies (e.g., comparisons between patents with PD and controls, and between PD patients with and without CAN), assessments of CAN (SCOPA-AUT vs. CASS), PD stages of patients (later or earlier stage), and statistical methods.
CAN risk score integrates ESC with age and BMI to estimate the individual’s current CAN risk [
16]. Although the algorithm of the risk score was developed in patients with diabetes, it may also be applied to patients with PD according to our data. The CAN risk score had higher correlation coefficients with cardiovascular autonomic function (HR_DB, VR, and BRS_seq) than the original ESC value. On contrary to ESC value, there was significant difference between patients with and without CAN (
p = 0.029). Our study also showed that CAN risk score could provide good diagnostic accuracy for the presence of CAN in patients with PD [
27]. The area under the ROC curve for CAN risk score in predicting the presence of CAN was 0.704.
The results of our study showed no significant correlation between ESC and cardiovascular subscores of CASS. This finding implies that the deterioration of autonomic function in patients with PD may be unparalleled in different domains (sudomotor and cardiovascular). The notion is supported by a previous report by Kim et al. who studied autonomic function, including cardiovagal, adrenergic, and sympathetic sudomotor domains, in patients with PD [
9]. They showed that the proportion of abnormal patients with cardiovagal function (HR_DB and VR) differed among the groups divided by each HY stage, whereas adrenergic (HUT) and sympathetic sudomotor function (measured by quantitative sudomotor axon reflex test (QSART)) did not differ significantly among groups. As for the comparison between Sudoscan and other established methods for sudomotor function measurement, Novak has shown that ESC correlated with skin nerve fiber density and sweat gland nerve fiber density [
28], but studies thus far have shown little or no correlation between ESC and QSART [
29]. Although these tests all evaluate sympathetic sudomotor or small-fiber function, the applied mechanisms and physiologies are totally different for each test.
Previous studies have shown that autonomic pathology in PD is primarily postganglionic [
11]. The reduction of ESC was expected to follow a length-dependent pattern. Foot ESC may be more sensitive in showing abnormalities than hand ESC. Nevertheless, it is surprising that our data revealed UPDRS and parameters of cardiovagal function correlated with hand ESC instead of foot ESC although the correlation coefficient indicates a weak positive linear relationship (r < 0.4). The results are consistent with a report by Xu et al. which showed that a decrease in ESC was greater in the hands than in feet for patients with PD [
18]. We suggest some possible explanations. First, measurement of ESC may be confounded by local skin factors such as skin temperature, increased thickening of stratum corneum, and the dirt on the skin. These factors are more likely to occur on the feet than on hands thus preventing distinct measurement of ESC. Furthermore, ESC, strictly speaking, is a measurement of sweat gland function, which is assumed to reflect autonomic nerves innervating the gland. However, a study by Duchesne et al. showed weak correlation of ESC with skin biopsy. Therefore, they suggested that mechanisms other than the loss of innervating fibers may be responsible for sweat gland dysfunction in polyneuropathies [
30]. Finally, the pathophysiology underlying autonomic neuropathy in PD is not totally understood. A severe non-length-dependent reduction of substance
p-immunoreactive intraepidermal nerve fibers in patients with PD has been reported by Doppler et al. [
31].
There are some limitations in this study. First, although both the clinical assessment and autonomic testing were done in the “off” state, 8–12 h after the latest dose of anti-Parkinsonian agents, lingering effects from the medication on autonomic function, especially sudomotor function cannot be excluded. Second, this was a cross-sectional study, and thus it did not answer the question of whether ESC measured by a Sudoscan is effective as a tool to follow-up sudomotor function for patients with PD. In addition, different ethnic groups may affect the normal values of ESC, which seem to be lower in the Chinese population [
32]. A more comprehensive study including normal healthy controls is mandatory. Finally, the patient number in the TD subgroup was limited. Although our results showed that the TD subgroup had better autonomic function and lower CASS, it remained inconclusive whether autonomic function differs among different phenotypes of PD.