Cardiac autonomic neuropathy (CAN) is an often underdiagnosed and serious microvascular complication of type 2 diabetes mellitus (T2DM). Heart rate variability (HRV) provides a non-invasive assessment of autonomic nervous system (ANS) function. However, conventional linear indices often identify abnormalities only at later stages. Nonlinear HRV metrics have emerged as sensitive markers of early ANS impairment. This study aimed to compare time-, frequency-, and nonlinear-domain HRV indices to determine whether nonlinear metrics enhance early detection of autonomic dysfunction in T2DM. This is one of the first studies in Vietnam to comprehensively evaluate multidomain HRV analysis in T2DM patients.

A cross-sectional study was conducted in patients with T2DM. Five-minute resting HRV recordings were obtained using the Polar H10 sensor and analyzed with Kubios Premium. HRV indices included SDNN and RMSSD (time-domain); LF, HF (frequency-domain); and SD1, SD2, and SD2/SD1 ratio (nonlinear). Clinical parameters included BMI, HbA1c, lipid profile, and eGFR. Descriptive statistics and Pearson correlations were applied, with significance set at p < 0.05.

In total, 75 patients participated in the study. Overall, they showed globally reduced HRV, with marked parasympathetic reduction (RMSSD 28.4 ± 12.6 ms; SD1 24.6 ± 10.5 ms). Patients with eGFR <60 mL/min/1.73m² were significantly older (p <0.01) and exhibited a trend toward lower HRV, although differences were not statistically significant. Nonlinear indices demonstrated strong correlation with linear measures, including RMSSD–SD1 (r = 0.95, p <0.01) and SD2–LF (r = 0.92, p <0.01). The SD2/SD1 ratio was positively correlated with BMI (r = 0.24, p <0.05) and the triglyceride–HDL-C ratio (r = 0.24, p <0.05), suggesting a link between ANS imbalance and metabolic burden. No significant associations were found between HbA1c and primary HRV indices.

HRV indices across all domains were reduced in T2DM, with the greatest declines in parasympathetic-related markers (RMSSD, HF, SD1). Nonlinear metrics—particularly SD1 and SD2/SD1—showed strong correlation with linear measures and greater sensitivity to metabolic parameters, supporting their utility in detecting early ANS dysfunction. Lower eGFR was associated with further HRV reduction, suggesting renal impairment may aggravate autonomic decline. Incorporating nonlinear HRV analysis may enhance early identification of subclinical CAN and improve cardiometabolic risk stratification in diabetes.