Sudomotor Dysfunction as a Measure of Small Fiber Neuropathy in Type 1 Diabetes

Abstract

Background

This study evaluated whether measuring the electrochemical skin conductance (ESC) may be a reliable surrogate for early cardiovascular autonomic neuropathy (CAN).

Methods

Longitudinal study included 37 type 1 diabetes (T1D) subjects (mean age 38±13 years, duration 15±7 years, HbA1c 7.9±1.1%, no known complications at baseline), and 40 age-matched healthy control (HC) subjects. Mean hands ESC (ESChands) and feet (ESCfeet) were measured with the SUDOSCAN (Impeto Medical, France). CAN was assessed with heart rate variability (HRV) studies (ANSAR Inc., PA), cardiovascular autonomic reflex tests (deep-breathing, Valsalva, postural test), and positron emission tomography with sympathetic analogue [11C]meta-hydroxyephedrine. Associations between measures of CAN and ESC were estimated using Spearman correlations. Longitudinal changes were analyzed using paired t-test.

Results

At baseline, there were no differences between T1D and HC in ESChands (69±14 vs. 69±13μS; P=0.84) or ESCfeet (82±8 vs. 78±9μS; P=0.12), while some indices of HRV and Valsalva ratio were significantly lower in T1D vs. HC. T1D subjects experienced a significant decline in both ESChands and ESCfeet at 12 months (mean change −7.2±11.6, P=0.0006; −2.8±7.3, P=0.023 respectively). No significant correlations were consistently found between ESC and measures of CAN at baseline or at 12 months.

Conclusion

Comparing patients with T1D to controls, no differences in ESC were observed at baseline. The associations between ESC and established measures of CAN were inconsistent, which does not support ESC as a reliable surrogate for CAN. While both hands and feet ESC declined over time, the significance of this finding is unclear and warrants further reliability testing.

Introduction

The continuous increase in the prevalence of diabetes mellitus and related complications worldwide is associated with significant burdens on patient’s morbidity and quality of life, and with staggering health care costs (1). Diabetic neuropathy (DN) is one of the most prevalent diabetic complications affecting up to 60% of patients over their lifetime (2; 3). Although DN is rare in early stages of type 1 diabetes (T1D), its prevalence increases with disease duration and poor glycemic control (2; 4–6).

Besides glucose control, pathogenic therapies to treat DN are still missing, possibly because there are limited tools to detect DN in its earlier stages (3; 4; 6) that are most susceptible to therapeutic interventions. Therefore, developing sensitive, specific, non-invasive, and affordable tests to detect the earliest stages of autonomic dysfunction and of DN in general, which are most susceptible to therapeutic interventions, are important in the development of viable therapies to prevent or reverse the progression of this complication.

DN has a broad spectrum of manifestations, but the most common forms are distal symmetrical polyneuropathy and cardiovascular autonomic neuropathy (CAN) (3; 6; 7). One of the earliest signs of CAN is reduced heart rate variability (HRV) (3) and changes in several cardiovascular reflex tests (CARTs) (3; 8–11). Although these measures have been shown to be sensitive, non-invasive and reproducible, they are time-consuming and require trained personnel (3; 8–11).

Sudomotor function, which is largely regulated by the same small, postganglionic, unmyelinated cholinergic sympathetic C-fibers that also regulate cardiovascular autonomic function, may be impaired earlier in patients with DN (8). The current methods currently used to assess sudomotor function are expensive, time consuming, and require specialized trained personnel (12–14). The recently developed SUDOSCAN device (Impeto Medical; Paris, France) provides a simple and non-invasive method that indirectly assesses sudomotor function by measuring the electrochemical skin conductance (ESC) (15–19). This is based on the electrochemical reaction between the chloride ions in sweat and stainless steel plate electrodes on which the subject’s hands and feet are placed by combining direct current stimulation and reverse iontophoresis.

The hypothesis was that the ESC measured with SUDOSCAN, may also detect earliest stages of cardiovascular autonomic dysfunction in T1D. Thus, the main objectives of this study were to evaluate the associations between established measures of CAN and ESC as an indirect measure of sudomotor dysfunction in patients with T1D, and whether this measure could also be used as a reliable tool to evaluate longitudinal changes in the small fiber function in these patients. In addition, we aimed to evaluate the potential role of several risk factors in promoting these deficits.

Methods

Study Design

Longitudinal prospective study in 37 T1D subjects followed for 12 months and 40 age-matched healthy control (HC) subjects.

Study population

T1D subjects were recruited from an ongoing clinical trial (NCT01170832) designed to evaluate the natural history of CAN and myocardial dysfunction in T1D at the University of Michigan Health System. Recruitment for this study began in August 2010 and ended in May 2013. Major inclusion criteria for the T1D subjects were: T1D as defined by the American Diabetes Association diagnostic criteria (20), duration >5 years, age 18–65 years, and no history of any complications at baseline. HC were age-and-gender matched with the T1D subjects at the start of the SUDOSCAN study, but the controls had normal glucose tolerance, normal blood pressure, and normal lipid profile.

Major exclusions were pregnancy or nursing, history of any cardiovascular disease (confirmed by stress test and cardiology evaluations), history of previous kidney, pancreas or cardiac transplantation, history of drug/alcohol abuse, and current use of any agents or drugs that interfere with the uptake or metabolism of catecholamines.

The study was approved by the Institutional Review Board of the University of Michigan and all subjects signed a written informed consent.

Evaluations

All evaluations were performed in the morning after an overnight fast. Height was measured in centimeters (cm) using a stadiometer, and weight in kilograms (kg) using a standardized calibrated scale. Sitting systolic and diastolic blood pressure (BP) was obtained after at least 10 minutes of rest. Fasting blood samples were obtained for glucose, metabolic and lipid panel [total cholesterol, calculated high-density lipoprotein cholesterol (HDLc), triglycerides, calculated low-density lipoprotein cholesterol (LDLc)] and processed by the University of Michigan Health System laboratory via Advia 1800 (Siemens, Erlangen, Germany), and hemoglobin A1c (HbA1c) was measured by high performance liquid chromatography (TOSOH Bioscience, San Francisco, CA).

Assessment of Sudomotor Function

Electrochemical skin conductance was evaluated in the T1D subjects at baseline and 12 months, and in healthy control subjects once with the SUDOSCAN device (Impeto Medical: Paris, France) as previously described (15–19; 21–24). Briefly, ESC is measured via reverse iontophoresis, an electrochemical reaction between the chlorides in sweat and stainless steel electrodes at low direct current (DC) voltage. The SUDOSCAN is composed of 2 sets of electrodes for hands and feet connected to a computer for recording and data management. Measurements are performed while the participants stand for 3 minutes placing their hands and feet on the plate electrodes. Four combinations of 15 different DC incremental voltages (<4V) are applied. The device measures an ESC score for each hand and foot and provides a percent asymmetry between left and right limbs. No special preparations of subjects are required. All data were collected with two SUDOSCAN devices operated by the same trained technicians in the same fashion (37 T1D and 9HC with one device and 31 HC with the second device) During the study Impeto Medical remotely monitored these devices and updated their software.

Assessment of CAN

CAN was assessed at baseline in both groups, and at 12 months in the T1D patients. Subjects were asked to avoid caffeine and tobacco products for at least 8 hours prior to testing, and hold any medication (except for basal insulin) until testing was completed. Subjects who experienced a hypoglycemic episode after midnight [blood glucose ≤ 50mg/dl (2.77mmol/l)] prior to the testing were rescheduled.

Left Ventricle (LV) sympathetic innervation with positron emission tomography (PET) with sympathetic analogue [11C] meta-hydroxyephedrine (HED)

The PET scans were done at baseline in T1D as previously described (25; 26). All PET studies were performed on a Siemens/ECAT Exact HR+ PET scanner. After positioning the subject in the PET scanner (Siemens, Erlanger, Germany) gantry, 2.0 mCi of [13N] ammonia was injected intravenously (IV) and a brief PET scan acquired to visualize the heart. Thirty minutes later, 20 mCi of [11C] HED was injected IV as a 60 min dynamic PET data acquisition sequence was started as previously described (25; 26).

Following image reconstruction, software was used to reorient and reslice the raw transaxial PET data into cardiac short-axis view data sets. The LV wall is divided into 60 regions to generate 480 independent LV regions. The measured [11C]HED radioactivity concentration in each sector in the final image frame (40–60 min) was normalized to the calculated integral of the total radioactivity in the blood pool throughout the PET study to obtain a [11C]HED “retention index” (RI, units: mL blood/min/mL tissue) for each LV sector. Polar maps of regional [11C]HED retention was generated and saved for visual inspection of [11C]HED retention deficits. A quantitative measure of the degree of cardiac denervation in each subject was generated by statistically comparing the measured [11C]HED RI value of each sector in the subject’s [11C]HED polar map to the mean and standard deviation of the RI data for that sector in our database of healthy non-diabetic subjects (age 20 to 70 years, 18 male, 18 female). Using this standard ‘z-score analysis’ (25), sectors in the subject’s [11C]HED polar map with RI values more than 2.5 standard deviations (SD) below the healthy control mean value was considered to be regions with ‘abnormal’ retention.

CARTs and HRV Studies

All T1D study participants underwent the standardized CARTs, considered the gold standard tests for CAN, (8; 9; 27) and HRV studies as described (28) at baseline and 12 months.

The ECG recordings were obtained in the supine position using a physiologic monitor (Nightingale PPM2, Zoe Medical Inc.) and data were collected during a resting state (5 minutes) and during the standardized CARTs obtained under paced breathing (R-R response to deep breathing, Valsalva maneuver and postural changes) as previously described (28; 29).

HRV data were analyzed according to current guidelines (9–11) using the continuous wavelet transform methods with the ANX 3.1 (ANSAR Inc., PA). This method incorporates respiratory activity in the formula and is reported to be superior for the analysis of nonstationary signals compared to Fourier transform.

Measures of CAN

The following measures of CAN were predefined as outcomes of interests and analyzed: the [11C]HED RI by PET, expiration/inspiration (E/I) ratio, Valsalva ratio and 30:15 ratio, and time and frequency indices of HRV analyzed at rest and during CARTs that included: the standard deviation of the normal RR interval (SDNN) (msec), the root mean square of the differences of successive RR intervals (rmsSD), low frequency power (Lfa) (0.04 to 0.10 Hz), high frequency power (Rfa) (0.15 to 0.4 Hz), Lfa/Rfa ratio.

Statistical Analyses

Statistical analyses were performed using SAS software (SAS Institute Inc., Cary, NC). Differences among the T1D and healthy controls were evaluated using Mann-Whitney-Wilcoxon Rank Sum test due to the presence of non-normally distributed variables. Descriptive statistics are reported as mean and standard deviations for the continuous variables and n (%) for the categorical variables. Association between indices of HRV, CARTs ratios, global [11C] HED RI and hand and feet ESC were estimated using Spearman correlation in T1D patients. Changes in ESC in μS, HbA1c and measures of HRV in T1D over time were analyzed using paired t-tests. P-values are reported with no correction for multiple testing. The subjects with T1D were further categorized into those who had a decline in their hands and feet ESC (progressors, n=14) and those who had no decline or improved in both ESC (non-progressors, n=14) after 12 months. The differences in clinical and CAN parameters between progressors and non-progressors were assessed by two-sample t-tests.

Results

Characteristics of the T1D (62% females, mean age 38±13 years, mean diabetes duration 15±7 years, mean HbA1c 7.9 ±1.1%) at baseline and HC are presented in Table 1. Fasting blood glucose, HbA1c, BMI, and SBP were significantly higher in T1D vs HC. In addition, T1D participants had significantly higher heart rate, and significantly lower deep breathing Lfa/Rfa ratio and Valsalva ratio at baseline compared with HC (Table 1). However, there were no differences between T1D and HC in ESChands (69±14 vs. 69±13μS; P=0.84), or ESCfeet (82±8 vs. 78±9μS; P =0.12), in spite of a mean duration of T1D of 15 years.

Table 1.

Selected general characteristics and measures of HRV, CARTs and ESC in HC and T1D at first evaluation

Variables	HC N = 40	T1D N = 37	P values (Mann-Whitney test)
Age, years	39 ± 8	38 ± 13	0.47
Gender Females n, (%)	24 (60%)	23 (62%)	0.85
Diabetes Duration, years	NA	15 ± 7	NA
HbA1c, %	5.4 ± 0.3	7.9 ± 1.1	<0.0001
Glucose, mg/dl	85 ± 7	168 ± 56	<0.0001
BMI, kg/m2	22.8 ± 2.6	26.5 ± 3.9	<0.0001
SBP, mmHg	110 ± 11	118 ± 11	0.006
DBP, mmHg	66 ± 10	69 ± 11	0.26
Resting HR, bpm	59 ± 10	66 ± 12	0.006
Triglyceride, mg/dl	73 ± 29	68 ± 36	0.41
HDLc, mg/dl	65 ± 16	63 ± 22	0.73
LDLc, mg/dl	90 ± 25	88 ± 26	0.81
Hands ESC, μS	69 ± 13	69 ± 14	0.84
Feet ESC, μS	78 ± 9	82 ± 8	0.12
Resting Lfa/Rfa ratio	3.3 ± 6.5	2.6 ± 2.4	0.23
Deep Breathing Lfa/Rfa ratio	2.04 ± 4.22	0.37 ± 0.96	0.001
Standing Lfa/Rfa ratio	9.4 ± 8.6	8.9 ± 8.3	0.70
sdNN, ms	65.2 ± 36.7	51.4 ± 24.4	0.12
rmsSD, ms	45.5 ± 28.3	36.1 ± 26.1	0.13
30:15 Ratio	1.3 ± 0.2	1.2 ± 0.21	0.88
E/I ratio	1.2 ± 0.1	1.2 ± 0.1	0.60
Valsalva ratio	1.5 ± 0.3	1.3 ± 0.2	0.0001

HR: heart rate; BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure; HDLc: calculated high density lipoprotein; LDLc: calculated low density lipoprotein; sdNN: standard deviation of normal R–R intervals; rmsSD: root-mean square of the difference of successive R–R intervals; Lfa: low frequency power; Rfa: high frequency power; E/I ratio: expiration/inspiration ratio; CARTs: Cardiovascular autonomic reflex tests; HRV: heart rate variability. All data are reported as either mean ± SD, or n (%).

In patients with T1D, both hands and feet mean ESC declined significantly from the first evaluation to 12 months (mean change −7.2 ± 11.6, P=0.0006; −2.8 ± 7.3, P=0.023 respectively), although no consistent significant changes were observed in measures of CAN (HRV and CARTs) in this interval (Table 2).

Table 2.

Mean change of characteristic variables in T1D (12-month-baseline)

Variables	Change in T1D (12 month-baseline)	P value (paired t-test)
Age, years	1 ± 0.4	NA
Diabetes Duration, years	1.2± 0.6	NA
HbA1c, %	−0.1 ± 0.8	0.63
Glucose, mg/dl	2.5 ± 81	0.85
BMI, kg/m2	−0.3 ± 1.7	0.24
SBP, mmHg	−2.8 ± 10.6	0.12
DBP, mmHg	−1.6 ± 8.8	0.28
Resting HR, bpm	−1.1 ± 8.3	0.42
Mean Hands ESC, μS	−7.2 ±11.6	0.0006
Mean Feet ESC, μS	−2.8 ±7.3	0.023
Resting Lfa/Rfa ratio	0.4±2.9	0.42
Deep Breathing Lfa/Rfa ratio	0.01±1.21	0.98
Standing Lfa/Rfa ratio	1.3±9.4	0.40
Valsalva Lfa/Rfa ratio	1.3 ± 6.8	0.07
30:15 Ratio	−0.01±0.17	0.85
E/I ratio	−0.01±0.11	0.75
Valsalva ratio	0.1±0.3	0.13
sdNN, ms	−4.6± 21.7	0.21
rmsSD, ms	−2.2±19.2	0.49

We then examined the potential driving factors for these changes in the T1D cohort as a group. There were no significant associations between any of the established risk factors including HbA1c, fasting glucose, BMI, lipids, BP, HR, and CAN measures, and the changes in hands and feet ESC.

We next examined the associations between indices of HRV, CARTs and ESC using Spearman correlation. Although feet ESC and hand ESC were significantly correlated with E/I ratio (r=0.37, P= 0.02) and Valsalva ratio (r=0.34, P=0.04) respectively at first assessment, these associations did not persist at 12 months. No other consistent associations between feet or hands ESC and the rest of the measures of CAN were found at both evaluations (data not shown). Similarly, at baseline, there were no significant correlations between the global left ventricle [11C] HED RI and the hands or feet ESC (data not shown).

In an attempt to understand the phenotype associated with progressive ESC decline, we also performed subgroup analysis stratifying by T1D subjects who had a decline in their hands and feet ESC (progressors, n=14) and T1D subjects who had either no decline or improvement in both ESC (non-progressors, n=14) at follow-up. There were no significant differences in any of the established risk factors including the HbA1c, blood pressure, lipids, BMI, ESC or CAN measures between these groups (data not shown, p>0.5 for all). In addition, there was no change in any of the measures of CAN from baseline to 12-month in the ESC progressors compared with non-progressors (data not shown).

Discussion

In this pilot study we evaluated the associations between several established and sensitive measures of CAN, and ESC assessed by SUDOSCAN, as an indirect measure of sudomotor function in patients with T1D with an early course of disease. We have also evaluated longitudinal changes in these measures after 12 months. To the best of our knowledge, this is the first study that compared the performance of ESC to very comprehensive evaluations of CAN that included an entire battery of established CAN measures such as CARTs, HRV and PET studies in patients with T1D, and assessed decline over time.

At baseline (first assessment), we found that subjects with T1D presented with a significantly lower deep breathing Lfa/Rfa ratio, a measure of overall sympathetic/parasympathetic balance, and Valsalva ratio, and a significantly higher heart rate compared with HC. These differences may be possibly due to the presence of early subclinical signs of cardiovagal imbalance in this T1D cohort with approximately 15 years of diabetes duration and sub-optimal glucose control, since both factors are major determinants for CAN in T1D as documented by findings from the DCCT/EDIC and other cohorts (4; 5; 29). However, in spite of the above diabetes duration and changes in established CAN measures, the hand and feet ESC in T1D participants were within normal limits based on recently reported data (21; 30), and there were no differences in the hand and feet ESC between T1D and HC. In addition, no significant consistent associations were found between hands and feet ESC and established CAN measures at baseline, or between changes in these measures over 12 months, other than mean feet ESC and hands ESC with E/I ratio and Valsalva ratio respectively at first assessment, which could have been due to chance due to multiple testing. These data thus suggest that ESC may not be a good substitute to detect early autonomic neuropathy or CAN.

There was also a significant decline in both hands and feet ESC in this cohort from baseline to 12-month. A similar decline was not observed in the CAN measures, which implies that the changes in ESC do not necessarily reflect changes in established measures of CAN. Based on published data, a physiological ESC decline with age is expected, which would be in line with the effect of aging on many other measures of neuropathy including sudomotor function (9; 30–32). It is unclear however, from the available data how much decline would be expected over 12 months in a relatively young cohort such as the one evaluated here.

In an attempt to understand the potential drivers of this decline, we further separated the subjects who experienced an ESC decline from the ones who stayed the same or improved, and classified them into progressors vs non-progressors. Although we did not evaluate measures of peripheral neuropathy in this study, none of these progressors had a decline in the CAN measures, and there were no differences in any of the documented risk factors for neuropathy progressions including glucose control, T1D duration, lipids, BMI, BP, or exploratory inflammatory markers. In theory, it could be possible that ESC provides information on other aspects related to autonomic function, with differentiated response to various factors and thus may change quicker than the established CAN measures. However, the observed ESC decline was not explained by any of the traditional risk factors confirmed in many other cohorts of patients with diabetes to be instrumental for nerve function decline.

The finding that there was no difference between the baseline evaluations of ESC in the diabetic subjects after 15 years of diabetes compared to the controls, but a significant decline in the diabetics over one year is inconsistent with expectations. Although the data from the controls are cross-sectional, revisited the controls’ data and looked to see if there was a trend over time in the controls. The controls collected on the instrument used to measure diabetics were concurrent with the first data collection. The controls on the second instrument could be divided into a year prior to, concurrent with the first year of data collection, or a year after (concurrent with the second data collection). The following is a summary of the mean±SD (N) for the concurrent period in the first instrument, the before, concurrent and after periods in the second instrument: Mean hands ESC: 70±11 (9), 66±11 (11), 82±6 (8), 62±14 (12) respectively; Mean feet ESC: 79±11 (9), 73±9 (11), 84±5 (8), 78±9 (12) respectively. This high degree of variability between measurements in the HC over time suggests a need for more reliability testing with this device.

Thus, all these data raise the question whether the ESC decline is due to device measurement error over time, a chance finding, or a real difference in sudomotor function.

These findings are now further discussed in the context of recent reports from other cohorts and groups of investigators. For instance, in contrast with our data, Casellini et al recently reported that hands and feet ESC significantly correlated with several measures of CAN that included E/I ratio, deep breathing Lfa, deep breathing Rfa, sdNN, Valsalva Lfa, and Valsalva Rfa (21). However, that study included predominantly patients with T2D and with more advanced disease, as 72% of the cohort had already documented DN confirmed by the Neuropathy Impairment Score-Lower Legs (NIS-LL). No data were reported on the ESC/CAN correlations in patients without DN.

Similar to our findings, Yajnik et al (15) who performed standardized CARTs in 265 patients with T2D and more advanced disease (86% reported at least one symptom of neuropathy, and 54% had Michigan Neuropathy Screening Instrument (MNSI) scores > 2 “Clinical Neuropathy”), reported no associations with ESC between any of the CARTs, although lower foot ESC was associated with higher MNSI scores, a measure of peripheral DN (15).

In a recently published study that included 45 T1D subjects with some evidence of either DN or CAN, and 25 HC, Selvarajah and colleagues (23) reported correlations between feet and hands ESC with some measures of CAN, and that feet and hands ESC were significantly lower in patients with DPN and with CAN compared to HC (23). However, the diagnosis of CAN in that cohort was done using an autonomic function test score, which was not clearly defined and not in line with current recommendations. In addition, only pooled T1D findings were reported, with no analyses done in T1D without DN.

Several other studies have evaluated the efficacy of SUDOSCAN in detecting peripheral diabetic neuropathy. For instance, Smith et al investigating the diagnostic utility of the ESC obtained by SUDOSCAN in subjects with peripheral neuropathy of different etiologies, reported that feet ESC was significantly correlated with the Utah Early Neuropathy Scale, the Michigan Neuropathy Scoring Instrument (MNSI), sural sensory amplitude and intraepidermal nerve fiber density (IENFD) from the proximal thigh skin biopsy (22).

The strengths of our study are the comprehensive characterization of CAN in these subjects with state-of-the-art, well-validated, sensitive and specific measures which also included PET studies, the uniform and standardized fashion of performing these evaluations by the same study staff at each outcome assessment, and the longitudinal assessments.

Study limitations include the relatively small sample size, the lack of other small-fiber neuropathy measures such as quantitative sensory testing or IENFD, and the fact that by including patients with no evidence of neuropathy at baseline we could not evaluate sensitivity and specificity of the ESC in detecting established DN. In addition, given the pre-specified study design and the funding limitations, the healthy controls could only be evaluated at baseline, while only the T1D were followed overtime. There are also limitations of the SUDOSCAN device itself. The device simply stimulates sweat glands and measures the electrochemical response in terms of chloride concentration without an actual measure of nerve function in this process. Also, there may have been a shift in the values reported by the instrument over time, whether due to recalibration or use that could explain a decline, which could be completely independent of sudomotor or cardiovascular autonomic dysfunction. Whether the decline in the ESC is a reliable method to assess changes consistent with autonomic dysfunction, will be better assessed with a planned longer follow-up of this cohort that could evaluate whether the ESC in those who declined at 12-month remained in fact the same, declined further, or increased, and whether with longer follow-up associations with measures of CAN may be more consistent.

The fact that we found very few correlations between ESC and measures of subclinical CAN in this cohort with early T1D, which could have been due to chance, raises however the question whether ESC is a reliable measure of early DN or is a measure of a different physiologic change that occurs in diabetics, such as weight gain or loss. ESC appears to have limited utility in the early assessment of CAN.

Conclusion

In patients with T1D and early disease, both hands and feet ESC declined over time. However, the associations between ESC and established measures of CAN were inconsistent, suggesting that if ESC measures a physiologic change in diabetics, it is not a direct measure of CAN. The relationship of ESC to early changes in T1D (subclinical disease) requires further investigation in prospective studies with longer follow up. In addition, studies of the stability of readings of the instrument over time are necessary.

Abbreviation

DN

diabetic neuropathy

CAN

cardiovascular autonomic neuropathy

HC

healthy control

BMI

body mass index

SBP

systolic blood pressure

DBP

diastolic blood pressure

BUN

blood urea nitrogen

RI

retention index

ESC

electrochemical skin conductance

T1D

type 1 diabetes

sdNN

standard deviation of normal R–R intervals

rmsSD

root-mean square of the difference of successive R–R intervals

Lfa

low frequency power

Rfa

high frequency power

CARTs

Cardiovascular autonomic reflex tests

HRV

heart rate variability

HDLc

calculated high-density lipoprotein cholesterol

LDLc

calculated low-density lipoprotein cholesterol

BP

blood pressure

LV

left Ventricle

PET

positron emission tomography

HED

meta-hydroxyephedrine