Exploring the effect of urban environment on cardiovascular health: a feasibility study using machine learning to predict heart rate variability responses to environmental attributes in CHF patients

The global shift towards urbanization has been associated with a rise in non-communicable diseases, including congestive heart failure (CHF). Understanding the influence of diverse living environments on physiological variables such as heart rate variability (HRV) in patients with chronic cardiac disease may contribute to more effective lifestyle advice and telerehabilitation strategies. This study aims to investigate the feasibility of machine learning (ML) models to predict autonomic nervous system (ANS) responses, quantified through HRV metrics, to environmental attributes in uncontrolled real-world settings. The goal is to validate whether this approach can help ascertain and quantify the presence of a connection between separate environmental attributes and cardiovascular health in patients with CHF.

20 participants, comprising 10 healthy subjects and 10 patients with CHF, irrespective of type or etiology, were enrolled in the study. Over a 3-week period, activities at specific GPS locations, along with heart rate (HR) data, were collected using smartwatches. Environmental attribute quantification was accomplished through semantic sementation of Google Street view (GSV) images. The collected data were utilized to train and test machine learning models for predicting ANS dynamics, as demonstrated by HRV metrics, associated with environmental attributes. The predictive performance of ML models was assessed using Spearman’s correlation coefficient (R), root mean square error (RMSE), prediction intervals, and Bland-Altman analysis.

The findings demonstrate the satisfactory efficacy of ML models in predicting HRV metrics associated with vagal nerve activity (R > 0.8, p < 0.05 for HR; 0.8 > R > 0.5, p < 0.05 for RMSSD and SD1; 0.5 > R > 0.4, p < 0.05 for HF, and LF/HF). However, predictive accuracy for metrics reflecting overall autonomic activity exhibited limitations due to the intricate interplay between sympathetic and parasympathetic modulation.

This study highlights the potential of ML-based models to discern vagal dynamics influenced by living environments in healthy individuals and patients diagnosed with CHF. Ultimately, this strategy could offer rehabilitation and tailored lifestyle advice, leading to improved prognosis and enhanced overall patient well-being in CHF.