Heart rate variability helps classify phenotype in systemic sclerosis.

Area of Science:

• Cardiology
• Autonomic Nervous System Research
• Medical Diagnostics

Background:

• Systemic sclerosis (SSc) is a complex autoimmune disease with distinct subtypes.
• Accurate classification of SSc subtypes, such as diffuse cutaneous SSc (dcSSc) and limited cutaneous SSc (lcSSc), is crucial for patient management.
• Current diagnostic methods may lack bedside applicability for rapid subtype differentiation.

Purpose of the Study:

• To develop and validate a classifier tool for distinguishing between dcSSc and lcSSc subtypes at the patient's bedside.
• To investigate the utility of heart rate variability (HRV) parameters in differentiating SSc subtypes.
• To identify the most effective HRV-based models for SSc subtype classification.

Main Methods:

• Compared 5-minute resting and orthostatic heart rate variability (HRV) in 58 SSc patients (16 dcSSc, 38 lcSSc).
• Analyzed HRV using time, frequency, and nonlinear domains from beat-to-beat RR intervals.
• Evaluated 341 classification models using various HRV variable combinations and algorithms, assessing performance metrics like F1-score and accuracy.

Main Results:

• The dcSSc group exhibited higher heart rate and lower HRV compared to the lcSSc group under both resting and orthostatic conditions.
• A stand-up maneuver significantly altered specific HRV indices (sd_HR, SD2, CorDim) differently between dcSSc and lcSSc groups.
• The best classification models achieved a maximum F1-score of 0.947 using 4 HRV variables, with sd_HR, SD2, and CorDim being highly discriminative.

Conclusions:

• Developed high-performance, bedside-applicable classification models to differentiate dcSSc from lcSSc.
• The models leverage 1-5 non-linear HRV indices as markers of autonomic influence on cardiac activity.
• This HRV-based approach offers a promising, non-invasive method for SSc subtype classification.