Clinical Characteristics and Epidemiology of Cardiovascular Autonomic Neuropathy in Peoples With Long-Standing Diabetes

Abstract

Background

Cardiovascular autonomic neuropathy (CAN) is a frequently underdiagnosed complication of diabetes mellitus that contributes to increased mortality and morbidity rates. This multicenter study investigated the epidemiology and clinical characteristics of CAN in patients with long-term diabetes.

Methods

Retrospective data were collected from 884 individuals with diabetes who were assessed for CAN across eight hospitals in Korea. CAN was diagnosed using electrocardiography and cardiovascular autonomic function tests, including the heart rate–corrected QT interval and Ewing’s method. This study evaluated the clinical characteristics, cardiovascular autonomic nerve function test results, and risk factors associated with CAN.

Results

Among the 884 patients, 778 (88%) were diagnosed with CAN (DCAN), while 106 (12%) were not (non-DCAN). Individuals with DCAN were older, had a longer duration of diabetes, and had higher creatinine levels. The DCAN group showed decreased time and frequency parameters of heart rate variability (HRV). A multiple logistic regression analysis revealed that a longer diabetes duration, older age, and higher creatinine level were significant risk factors for CAN.

Conclusion

CAN was associated with advanced age, a longer duration of diabetes, higher creatinine levels, and lower time and frequency HRV parameters.

Graphical Abstract

INTRODUCTION

Diabetes mellitus (DM) is a chronic metabolic disorder that has reached epidemic proportions worldwide whose complications feature significant morbidity and mortality. According to the 2021 World Health Organization fact sheet, the prevalence of type 2 DM (T2DM) is increasing globally, including in Korea.1 In fact, the number of adults with T2DM in Korea increased from 7.5% in 1980 to 16.7% in 2020.1 Cardiovascular autonomic neuropathy (CAN), among the most insidious and least recognized complications of DM,2,3 is characterized by damage to autonomic nerve fibers that innervate the heart and blood vessels that leads to abnormal heart rate control and vascular dynamics. CAN significantly increases the mortality risk among patients with DM owing to its association with silent myocardial infarction, arrhythmia, and sudden cardiac death.2 CAN manifests as a broad subclinical to clinical status.3 While decreased heart rate variability (HRV) is a subclinical sign of CAN, resting tachycardia, exercise intolerance, and orthostatic hypotension are detectable and assessable clinical symptoms.3 CAN causes cardiac functional impairment, which can be fatal.3

The prevalence of CAN varies widely across populations and is influenced by factors such as diabetes duration, glycemic control, other diabetic complications, and non-glycemic factors such as hypertension and hyperlipidemia.4,5 In patients with long-standing T2DM, particularly those with a disease duration > 10 years, the prevalence of CAN is reportedly as high as 50%.6 This finding underscores the importance of early CAN detection and management in the prevention of adverse cardiovascular outcomes.

Despite its clinical significance, CAN is often asymptomatic in the early stages and commonly remains underdiagnosed.7 When symptoms do manifest, they are often non-specific, such as exercise intolerance or lightheadedness, complicating its early diagnosis.7 Traditional diagnostic methods include cardiovascular reflex tests (CARTs), HRV analyses, and measurement of the corrected QT (QTc) interval on electrocardiography (ECG).7,8,9 However, the lack of standardized diagnostic criteria and CAN presentation variability can complicate its diagnosis and epidemiological assessment.

In Korea, data are lacking regarding the epidemiology and clinical characteristics of CAN, particularly among individuals with long-term diabetes. Thus, understanding the prevalence, risk factors, and clinical manifestations of CAN in this population is crucial for improving patient management and outcomes.

Here we hypothesized that this population has a high prevalence of CAN, which features certain clinical factors such as diabetes duration, glycemic control, and nephropathy. By identifying its key risk factors and clinical features, we aimed to investigate the epidemiology and clinical characteristics of CAN among patients with long-term diabetes.

METHODS

Study design, participants, and data collection and measurements

This retrospective observational multicenter study collected the electronic medical records of patients who visited the hospital between January 1, 2015, and June 30, 2018. It included patients with type 1 diabetes mellitus (T1DM) or T2DM for ≥ 10 years. Only individuals who were tested for peripheral and cardiovascular autonomic neuropathies were included. Baseline information questionnaires were administered to all patients. Information on hypertension, smoking status (ex-, current, or never smoker), alcohol consumption, and medical history were collected from the questionnaires. Hypertension was defined as a systolic blood pressure (SBP) ≥ 140 mmHg, a diastolic blood pressure (DBP) ≥ 90 mmHg, or the use of medication to treat high blood pressure.7,8 The patients were asked to identify their smoking status. Alcohol intake was defined as 1–2 drinks per day.

Patients with the following were excluded: 1) gestational or secondary DM; 2) infection, liver cirrhosis, thyroid disease, vitamin B₁₂ deficiency, and malignancies; 3) use of anti-arrhythmic agents; 4) pacemaker implantation; 5) hypoglycemia (glucose < 70 mg/dL) or fasting hyperglycemia (glucose ≥ 200 mg/dL); and 6) older age (> 80 years).

Cardiovascular autonomic function tests

The diagnosis of CAN was based on a battery of CARTs according to established guidelines as follows:

1. HRV – HRV was assessed using 5-minute ECG. Time- and frequency-domain measures of HRV were calculated, including the standard deviation of normal-to-normal intervals (SDNN) and the low- to high-frequency ratio (LF/HF ratio). Reduced HRV is indicative of autonomic dysfunction;

2. Blood Pressure Response to Postural Change (Orthostatic Hypotension) – Blood pressure was measured with each patient in the supine position after a 5-minute rest and then in the standing position at 1 and 3 minutes after standing. Orthostatic hypotension was defined as a drop in SBP ≥ 20 mmHg or drop in DBP ≥ 10 mmHg upon standing;

3. Valsalva Ratio – The Valsalva maneuver was performed and the Valsalva ratio (ratio of longest RR interval after to shortest RR interval during the maneuver) was calculated. A Valsalva ratio < 1.20 was considered abnormal;

4. Heart Rate Response to Deep Breathing – The difference between the maximum and minimum heart rates during deep breathing (6 breaths/min) was calculated. A difference of ≤ 10 beats/min was considered abnormal; and

5. QTc Interval on ECG – QTc interval was measured using a standard 12-lead ECG and corrected for heart rate using Bazett’s formula (QTc = QT/√RR). A prolonged QTc interval (> 440 ms in men, > 460 ms in women) is considered a marker of autonomic dysfunction.

DCAN was assessed using five standard CARTs according to Ewing’s protocol.9,10 Three of these measurements assessed parasympathetic function using the heart rate response to deep breathing (beat-to-beat variation), standing (30:15 ratio), and Valsalva maneuver. The other two tests assessed sympathetic function, changes in blood pressure in response to standing, and sustained handgrip strength. The heart rate response to deep breathing, upright standing, and the Valsalva maneuver was automatically assessed from ECG recordings using the DCAN evaluation system (Medicore Co., Ltd., Seoul, Korea). DCAN severity was quantified by summing the points obtained from each of the five tests (0, 0.5, or 1 point each depending on whether it was a normal, borderline, or abnormal value, respectively).9,10 DCAN was defined as the presence of at least two abnormal results or autonomic neuropathy points ≥ 1.11 For the HRV analysis, the resting beat-to-beat heart rate was measured for 5 minutes. A time-domain analysis was performed based on the SDNN of normal RR intervals and the root mean square of successive differences (RMSSD). A frequency-domain analysis was performed using low-frequency (LF; 0.04–0.15 Hz) and high-frequency (HF; 0.15–0.40 Hz) HRV as well as the LF/HF ratio.

The final diagnosis of DCAN was based on the ECG QTc interval or Ewing’s test in which five CARTs were evaluated through deep breathing, lying-to-standing, sustained handgrip strength, and Valsalva tests. The patients were divided into two groups (non-DCAN vs. DCAN) for the investigation of the clinical characteristics of CAN.

Clinical and laboratory data

Initially, all patients were evaluated based on detailed demographic data, clinical laboratory parameters, and medical histories obtained from electronic medical records. Blood samples were collected after an 8-hour overnight fast. Post-prandial blood samples were collected 2 hours after a normal meal. Fasting glucose, glycated hemoglobin (HbA1c), fasting C-peptide, 2-hour post-prandial C-peptide, serum creatinine, urine albumin, total cholesterol, triglycerides, high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol levels as well as urine albumin-to-creatinine ratio (UACR) were measured.

Assessment of diabetic complications

The presence of other complications such as diabetic retinopathy, nephropathy, and peripheral neuropathy was assessed through comprehensive eye examinations, UACR, and clinical neurological assessments.

Statistical analysis

Descriptive data are shown as mean ± standard deviation for continuous variables, number and percentage for categorical variables, and median for non-normally distributed variables. Two different groups (non-DCAN vs. DCAN) were analyzed using an independent samples t-test for continuous variables and the Chi-squared or Fisher’s exact test for categorical variables. A multiple logistic regression analysis was performed to identify the associations between each factor and CAN using adjusted P values among the variables. All statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA).

Ethics statement

This study was approved by the Institutional Review Board (IRB) of Jeonbuk National University Hospital (IRB no. 2020-01-016), which waived the requirement for informed consent.

RESULTS

Clinical characteristics of patients with CAN

This study included 884 patients, a small proportion of whom had T1DM (17). The non-DCAN group included 106 (12%) patients, while the DCAN group included 778 (88%) patients. The mean patient age was 56.77 ± 10.78 years in the non-DCAN group versus 62.38 ± 9.90 years in the DCAN group (P < 0.0001). The non-DCAN and DCAN groups included 64 (60%) and 446 (57%) men, respectively (P = 0.580), and 42 (40%) and 329 (43%) women, respectively. The CAN group had a significantly longer mean DM duration (12.65 ± 3.97 years vs. 13.69 ± 4.21 years, P = 0.016). Mean body mass index did not differ between the two groups (non-DCAN, 25.31 ± 4.47 kg/m²; DCAN, 25.19 ± 4.15 kg/m²; P = 0.791). Although the prevalence of a history of smoking was not significantly different (40.43% vs. 37.64%, P = 0.600), the alcohol consumption rate was significantly higher in the non-DCAN group (51.58% vs. 38.51%, P = 0.014).

DM nephropathy was significantly more common in the DCAN versus non-DCAN group (15.68% vs. 7.55%, P = 0.027). Other comorbidities and diabetic complications did not differ significantly between groups. The prevalence of hospitalization was significantly higher in the DCAN versus non-DCAN group (1.21 ± 1.33 vs. 0.78 ± 0.75, P = 0.012).

Fasting C-peptide (2.39 ± 2.13 ng/mL vs. 1.81 ± 1.44 ng/mL, P = 0.001), creatinine (1.05 ± 3.21 mg/dL vs. 0.81 ± 0.20 mg/dL, P = 0.047), urine albumin (117.70 ± 506.05 mg/dL vs. 45.99 ± 221.77 mg/dL, P = 0.022), and UACR (197.95 ± 804.69 mg/g creatinine vs. 86.83 ± 155.16 mg/g creatinine, P = 0.002) were significantly higher in the DCAN versus non-DCAN group (Table 1).

Table 1. Baseline characteristics of peoples with diabetes by CAN status (non-DCAN vs. DCAN)13 .

Variables	Non-DCAN (n = 106)	DCAN (n = 778)	P value
Age, yr	56.77 ± 10.78	62.38 ± 9.90	< 0.001
Sex, male, %	64 (60.38)	446 (57.55)	0.580
Waist circumference, cm	84.44 ± 6.00	87.99 ± 8.28	0.242
BMI, kg/m2	25.31 ± 4.47	25.19 ± 4.15	0.791
Smoking, %	38 (40.43)	277 (37.64)	0.600
Alcohol drinking, %	49 (51.58)	285 (38.51)	0.014
DM duration, yr	12.65 ± 3.97	13.69 ± 4.21	0.016
Hypertension, %	71 (66.98)	563 (72.37)	0.248
Hyperlipidemia, %	85 (80.19)	587 (75.45)	0.284
Stroke, %	2 (1.89)	49 (6.30)	0.068
CAD, %	27 (25.47)	242 (31.11)	0.237
T2DM, %	101 (96.19)	766 (98.84)	0.054
DM retinopathy, %	18 (16.98)	131 (16.84)	0.971
DM nephropathy, %	8 (7.55)	122 (15.68)	0.027
DM neuropathy, %	51 (48.11)	414 (53.21)	0.324
Diabetic foot, %	1 (0.94)	14 (1.80)	> 0.999
PAOD, %	5 (4.72)	42 (5.40)	0.770
Hospitalization	0.78 ± 0.75	1.21 ± 1.33	0.012
FPG, mg/dL	155.41 ± 116.60	144.91 ± 56.87	0.370
Total cholesterol, mg/dL	166.11 ± 134.61	149.25 ± 40.20	0.211
Triglyceride, mg/dL	142.05 ± 104.69	146.33 ± 107.43	0.706
HDL-C, mg/dL	49.84 ± 12.56	47.37 ± 13.07	0.075
LDL-C, mg/dL	83.13 ± 32.99	79.25 ± 29.36	0.252
HbA1c, %	7.98 ± 1.85	7.91 ± 1.73	0.707
Fasting C-peptide, ng/mL	1.81 ± 1.44	2.39 ± 2.13	0.001
PP2 C-peptide, ng/mL	2.26 ± 1.35	14.75 ± 169.10	0.301
Cr, mg/dL	0.81 ± 0.20	1.05 ± 3.21	0.047
Urine albumin, mg/dL	45.99 ± 221.77	117.70 ± 506.05	0.022
UACR, mg/gCr	86.83 ± 155.16	197.95 ± 804.69	0.002
Life style modification	6 (5.66)	11 (1.41)	0.011
Insulin	31 (29.25)	216 (27.76)	0.750
Metformin	76 (71.70)	574 (73.78)	0.649
SU	37 (34.91)	337 (43.32)	0.100
DPP4 inhibitors	52 (49.06)	461 (59.25)	0.046
GLP1R agonist	1 (0.94)	3 (0.39)	0.401
SGLT2 inhibitors	26 (24.53)	117 (15.04)	0.013
TZD	13 (12.26)	83 (10.67)	0.620
α-glucosidase inhibitors	1 (0.94)	19 (2.44)	0.497
Meglitinide	0 (0.00)	9 (1.16)	0.610
ACEi	5 (4.72)	57 (7.33)	0.324
ARB	55 (51.89)	398 (51.16)	0.888
CCB	16 (15.09)	219 (28.15)	0.004
Diuretics	9 (8.49)	121 (15.55)	0.054
α-blocker	2 (1.89)	14 (1.80)	> 0.999
ASA	22 (20.75)	275 (35.35)	0.003
Clopidogrel	10 (9.43)	94 (12.08)	0.427
Statin	75 (70.75)	534 (68.64)	0.659
Anti-depressant	3 (2.83)	11 (1.41)	0.230

Continuous variables of independent t-test and categorical variables of Chi (κ) square or Fisher’s exact test for two different groups (non-CAN vs. CAN) were analyzed. Statistically significant values are indicated in Bold (P < 0.05). Data are expressed as number (%) or mean ± standard deviation.

BMI = body mass index, DM duration = diabetes mellitus duration, CAD = coronary artery disease, T2DM = type 2 diabetes mellitus, DM retinopathy = diabetes mellitus retinopathy, DM nephropathy = diabetes mellitus nephropathy, DM neuropathy = diabetes mellitus neuropathy, PAOD = peripheral artery occlusive disease, FPG = fasting plasma glucose, HDL-C = high-density lipoprotein cholesterol, LDL-C = low-density lipoprotein cholesterol, HbA1c = glycated hemoglobin A1c, PP2 C-peptide = 2 hour post-prandial C-peptide, Cr = creatinine, UACR = urine albumin-to-creatinine ratio, SU = sulfonylurea, DPP4 inhibitor = dipeptidyl 4 inhibitor, GLP1R agonist = glucagon-like peptide-1 receptor, SGLT2 inhibitor = sodium glucose co-transport 2 inhibitor, TZD = thiazolidinedione, ACEi = angiotensin-converting enzyme inhibitor, ARB = angiotensin II receptor blocker, CCB = calcium channel blocker, ASA = acetylsalicylic acid.

Cardiovascular autonomic nerve function test characteristics in patients with CAN

The mean SBP values in the supine and standing positions were significantly higher in the DCAN versus non-DCAN group (124.63 ± 18.42 mmHg vs. 115.54 ± 14.43 mmHg, P < 0.0001; 122.71 ± 18.45 mmHg vs. 117.30 ± 14.74 mmHg, P = 0.001, respectively). The mean DBP with a position change did not differ significantly between groups (supine: DCAN, 77.85 ± 10.44 mmHg vs. non-DCAN, 76.15 ± 9.36 mmHg, P = 0.112; upright: DCAN, 78.22 ± 10.82 mmHg vs. non-DCAN, 79.20 ± 9.84 mmHg, P = 0.380) (Table 2). Other autonomic parameters such as the SDNN, RMSSD of the RR interval, total power, LF HRV, and HF HRV were significantly lower in the DCAN versus non-DCAN group (Table 2). The DCAN group had a decreased HRV (4.51 ± 12.88 vs. 4.95 ± 14.25, P = 0.744), whereas the intergroup LF/HF ratio did not differ significantly (5.09 ± 13.85 vs. 6.32 ± 14.75, P = 0.394). The DCAN versus non-DCAN group had a prolonged QTc interval (431.16 ± 30.04 ms vs. 420.71 ± 18.95 ms, P < 0.0001) (Table 2).

Table 2. Change of BP, heart rate variability and ECG QTc interval of peoples with DM by CAN status (non-DCAN vs. DCAN).

Variables	Non-CAN (n = 106)	CAN (n = 778)	P value
Supine SBP, mmHg	115.54 ± 14.43	124.63 ± 18.42	< 0.001
Upright SBP, mmHg	117.30 ± 14.74	122.71 ± 18.45	0.001
Supine DBP, mmHg	76.15 ± 9.36	77.85 ± 10.44	0.112
Upright DBP, mmHg	79.20 ± 9.84	78.22 ± 10.82	0.380
Supine HR	74.43 ± 9.98	72.58 ± 11.13	0.105
Upright HR	84.04 ± 11.05	80.85 ± 12.81	0.015
Supine SDNN	26.72 ± 12.15	21.34 ± 11.17	< 0.0001
Upright SDNN	23.54 ± 10.12	19.64 ± 10.85	0.001
Supine RMSSD	17.44 ± 16.86	12.38 ± 11.68	0.003
Upright RMSSD	14.60 ± 14.67	11.09 ± 10.89	0.019
Supine TP	572.55 ± 688.62	379.11 ± 531.40	0.006
Upright TP	439.34 ± 410.98	307.64 ± 511.46	0.003
Supine LF, msec2	167.11 ± 311.62	93.60 ± 167.80	0.019
Upright LF, msec2	112.72 ± 108.56	77.55 ± 188.07	0.005
Supine HF, msec2	123.59 ± 187.99	70.90 ± 122.96	0.006
Upright HF, msec2	98.62 ± 199.64	52.07 ± 121.11	0.021
Supine LF/HF ratio	4.95 ± 14.25	4.51 ± 12.88	0.744
Upright LF/HF ratio	6.32 ± 14.75	5.09 ± 13.85	0.394
ECG QTc prolongation	420.71 ± 18.95	431.16 ± 30.04	< 0.0001

Continuous variables of independent t-test and categorical variables of Chi (κ) square or Fisher’s exact test for two different groups (non-CAN vs. CAN) were analyzed. Statistically significant values are indicated in Bold (P < 0.05). Data are expressed as mean ± standard deviation.

SBP = systolic blood pressure, DBP = diastolic blood pressure, HR = heart rate, SDNN = standard deviation of normal-to-normal interval, RMSSD = root mean square of successive differences, TP = total power, LF = low frequency, HF = high frequency, CAN = cardiovascular autonomic neuropathy, ECG = electrocardiography, QTc = corrected QT interval.

Risk factors of CAN in patients with long-duration DM

Next, we analyzed the risk factors of CAN using multivariate logistic regression after adjusting for age, sex, diabetes duration, HbA1c level, fasting glucose level, high blood pressure, smoking, alcohol consumption, and dyslipidemia. The constituents with higher odds ratios were age, fasting C-peptide level, SBP (supine, upright), QTc prolongation, calcium channel blockers, and aspirin. All of these had significant differences with adjusted P values (< 0.0001, 0.005, < 0.0001, 0.015, 0.001, 0.000, and 0.027, respectively). Alcohol consumption, SDNN, RMSSD, supine LF HRV, HF (supine, upright) HRV, and sodium glucose co-transport 2 inhibitors use had lower odds ratios with significantly adjusted P values (Table 3).

Table 3. Multi-logistic regression analysis of odds ratio of factors in non-DCAN group and DCAN group14 .

Variables	Non-DCAN vs DCAN
	Odds ratio (95% CI)	Unadjusted P value	Adjusted P value
Age	1.053 (1.032–1.073)	< 0.0001	< 0.001
Alcohol consumption	0.588 (0.383–0.903)	0.015	0.017
DM duration	1.073 (1.013–1.136)	0.016	0.423
T2DM	3.372 (1.020–11.148)	0.046	0.601
Fasting C-peptide	1.259 (1.058–1.497)	0.009	0.005
Cr	2.288 (1.093–4.791)	0.028	0.112
Supine SBP	1.035 (1.020–1.049)	< 0.0001	< 0.0001
Upright SBP	1.018 (1.006–1.031)	0.004	0.015
Upright HR	0.981 (0.966–0.996)	0.015	0.160
Supine SDNN	0.965 (0.950–0.981)	< 0.0001	0.001
Upright SDNN	0.973 (0.958–0.989)	0.001	0.007
Supine RMSSD	0.975 (0.962–0.988)	< 0.001	0.008
Upright RMSSD	0.979 (0.964–0.993)	0.004	0.003
Supine TP	1.000 (0.999–1.000)	0.002	0.020
Upright TP	1.000 (0.999–1.000)	0.025	0.135
Supine LF	0.999 (0.998–0.999)	0.001	0.028
Supine HF	0.998 (0.997–0.999)	0.001	0.004
Upright HF	0.998 (0.997–0.999)	0.003	0.006
ECG QTc prolongation	1.017 (1.008–1.026)	< 0.001	0.001
DPP4 inhibitors	1.51 (1.006–2.268)	0.047	0.058
SGLT2 inhibitors	0.545 (0.336–0.884)	0.014	0.046
Life style modification	0.239 (0.087–0.660)	0.006	0.614
CCB	2.204 (1.266–3.836)	0.005	0.000
ASA	2.087 (1.276–3.414)	0.003	0.027

DM = diabetes mellitus, T2DM = type 2 diabetes mellitus, Cr = creatinine, SBP = systolic blood pressure, HR = heart rate, SDNN = standard deviation of normal to normal interval, RMSSD = root mean square of successive differences, TP = total power, LF = low frequency, HF = high frequency, ECG = electrocardiogram, QTc = corrected QT, DPP4 inhibitor = dipeptidyl 4 inhibitor, SGLT2 inhibitor = sodium glucose co-transport 2 inhibitor, CCB = calcium channel blocker, ASA = acetylsalicylic acid.

DISCUSSION

CAN is a significant and often under-recognized complication of DM, particularly in patients with long-standing disease. The results of this study provide valuable insight into the epidemiology, clinical characteristics, and risk factors associated with CAN in individuals with long-term diabetes.

The most common complication of diabetic autonomic neuropathy is DCAN, which leads to cardiac dysfunction and abnormalities in HR control and cardiac dysfunction.9 DCAN is associated with increased mortality and morbidity from cardiovascular diseases among patients with DM.10

Spallone et al.12 reported a prevalence of CAN of 16.6–20%. Zoppini et al.13 reported a prevalence of CAN (from early-stage to confirmation) of 17.2%, whereas that in another study was 52.2%.14 The current study found a prevalence of CAN among patients with long-duration DM of 88% (non-DCAN, 106; DCAN, 778), which was relatively higher than these reported values.

Several risk factors for diabetic neuropathy such as age, DM duration, macroalbuminuria (300 mg/24 h), and sustained albuminuria (≥ 30 mg/24 h) were reported in the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications study.15 A high SBP and advanced chronic kidney disease (chronic kidney disease stage ≥ 3a) are considered risk factors for CAN,15 as both can damage the blood vessels and nerves, leading to its development.15 Similarly, the risk factors for CAN include older age, long DM duration, and obesity.12,16

A prolonged QTc interval has been observed in individuals with DCAN; here we confirmed that finding. A prolonged QT interval on ECG is a known risk factor for the development of arrhythmia, which can cause sudden cardiac death.17 Some studies demonstrated that individuals with versus without diabetes are at higher risk of QTc prolongation on ECG, possibly due to CAN.17,18 Therefore, a QTc interval prolongation may be a risk factor for CAN (Tables 2 and 3).

We found differences in age, disease duration, SBP, and microvascular complications such as DM-related nephropathy, increased urine albumin and creatinine levels, and UACR between the non-CAN and CAN groups. A multiple logistic regression analysis showed that age, SBP, and QTc prolongation had higher odds ratios of CAN development. This finding shows that CAN progresses as people age, SBP increases, and QTc interval abnormalities occur. The intergroup baseline duration of diabetes differed significantly. A longer duration of diabetes indicates a higher tendency toward CAN progression.

Intensive glycemic control reduces the CAN progression rate.19 Tang et al.20 reported that intensive glucose and blood pressure control reduced CAN risk by 16% and 25%, respectively. This finding suggests that strict glycemic control retards progression. CAN is also a risk factor for cardiovascular disease. Cha et al.21 showed that well-controlled glucose levels could reduce the risk of cardiovascular disease. The mean HbA1c was reportedly 7.91 ± 1.73% vs. 7.98 ± 1.85% in the non-DCAN and DCAN groups, respectively (P = 0.707) (Table 1). Thus, HbA1c alone is not an indicator of CAN risk in this study, unlike the results of different studies.

Fasting C-peptide levels showed a higher odds ratio in the CAN group, indicating that high fasting C-peptide levels could be another risk factor for CAN progression. Another study reported that C-peptide levels were negatively correlated with CAN.22 A lower mean fasting C-peptide level was observed in the non-DCAN group.

Hypertension is associated with CAN and induced by sympathetic activation.23 In this study, we investigated four different blood pressure variables with respect to position (systole, diastole, supine, and standing). Only SBP while lying down or standing was significantly related to CAN, implying that DBP may not correlate with neuropathy.

Alcohol is toxic to the nervous system. Its consumption over various time periods may damage an individual’s cardiovascular autonomic function.24 It remains controversial whether alcohol abuse harms an individual’s autonomic nervous system function. Consistency was lacking in the effects of excessive chronic alcohol consumption on autonomic function or average alcohol intake.24 Alcoholism duration is a risk factor for sympathetic neuropathy; however, other studies considered it unrelated.25 This study showed that alcohol consumption in patients with diabetes was associated with a lower odds ratio (0.588 [range, 0.383–0.903], P = 0.017). Although the results in the literature are conflicting, alcohol itself seemed unrelated to autonomic neuropathy in this study.

The DCAN group had lower SDNN and RMSSD scores than the non-DCAN group. The DCAN group also showed reduced LF and HF HRV. Shah et al.26 reported that cardiac autonomic dysfunction reduces HRV and parasympathetic tone. Reduced HRV may be a predictor of sudden cardiac death and early indicator of diabetic CAN.11,27 Decreased HRV is an alternative indicator of cardiac death following myocardial infarction.27 Our study also showed that the DCAN group exhibited decreased HRV and impaired cardiovascular autonomic function. A strong correlation was observed between decreased HRV and CAN.

This study had several limitations. First, it was retrospective and enrolled only individuals with long-duration DM who were tested for CAN. However, some confounding factors could not be excluded. Second, we included all individuals with T1DM or T2DM, which have different prevalences of CAN and pathogeneses for its development. Patients with T1DM comprised a very small proportion of the study population. It is also known that the prevalence of CAN is higher among people with T1DM versus T2DM. Third, there were variable data collection defects. For example, we selected only patients who consumed alcohol. We also did not determine the duration of alcohol consumption. Instead, we included only patients who consumed alcohol. Several studies have reported a correlation between the total lifetime dose of ethanol and the development of neuropathy.24 Long-term heavy alcohol consumption can lead to neuropathy, resulting in numbness, tingling, and weakness. Thus, it would be helpful to document both consumption and amount of alcohol to determine the risk factors of CAN. The enrolled patients had poor blood sugar control. Further studies are needed to investigate the prevalence of CAN in intensive glucose control groups.

CAN is a significant complication of DM, particularly among individuals with long-duration disease. This study highlighted the high prevalence of CAN in individuals with long-term DM and identified the key risk factors associated with its development, including prolonged diabetes duration, poor glycemic control, hypertension, dyslipidemia, and the presence of other microvascular complications.

The clinical characteristics of patients with CAN, including reduced HRV, orthostatic hypotension, and a prolonged QTc interval, underscore the severity of the autonomic dysfunction. These findings have important implications for the management of patients with DM by emphasizing the need for regular screening, early detection, and a comprehensive approach to DM care that simultaneously addresses multiple risk factors.

Given the significant morbidity and mortality associated with CAN, particularly cardiovascular events, prioritizing it in the clinical management of patients with long-term DM is crucial. Future research should focus on longitudinal studies to further elucidate the temporal relationship between risk factors and CAN development, and on the development of novel therapeutic strategies aimed at preventing or delaying the onset of this debilitating complication.

In conclusion, the findings of this study underscore the need for heightened awareness and proactive management of CAN in patients with long-term DM to improve their outcomes and reduce the complication-related burden.