Abstract
This study utilizes a custom, ECG-validated IoT PPG device to investigate the correlation between urban pollutants (PM2.5, PM10, CO₂) and heart rate variability (HRV) among young adults in Almaty. Data collection across three microclimates revealed a "green zone paradox," where high humidity retained particulate matter and triggered significant sympathetic nervous system activation despite the natural setting. The analysis identified PM2.5 as the primary driver of physiological stress, while an evaluation of multiple machine learning models showed that XGBoost achieved a peak classification accuracy of over 91%. These findings demonstrate the power of integrating wearable sensing with advanced analytics to provide real-time cardiovascular risk assessments in polluted urban environments.