Cardiac autonomic changes in middle-aged women: identification based on principal component analysis

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Heart rate variability (HRV) analysis is a non-invasive method to assess autonomic nervous system function.
• Age-related changes in cardiovascular regulation can be reflected in HRV parameters.
• Distinguishing age groups using HRV requires robust analytical techniques.

Purpose of the Study:

• To evaluate the effectiveness of Principal Component Analysis (PCA) on Power Spectral Density (PSD) of RR intervals (iRR) for discriminating between young and middle-aged healthy women.
• To assess the potential of PCA-derived features from iRR PSD for age classification.

Main Methods:

• Collected 10-minute iRR sequences from 30 healthy, non-smoking women (13 young, 17 middle-aged) under controlled conditions.
• Estimated PSD using an autoregressive model on a 5-minute iRR segment with minimal variance.
• Extracted five principal component coefficients from PSD signals and used a Mahalanobis distance classifier.

Main Results:

• A logistic regression model achieved 100% specificity, 83.2% sensitivity, and 93.3% overall accuracy in separating age groups.
• PCA successfully classified women into young (20-25 years) and middle-aged (40-60 years) groups based on iRR PSD features.

Conclusions:

• PCA applied to iRR PSD is a promising technique for non-invasively assessing age-related physiological differences in women.
• This approach offers a high-accuracy method for discriminating between age groups using simple heart rate data.