Against the backdrop of intensifying global climate change, extreme heat events are occurring with increasing frequency, intensity, and duration, posing severe threats to human health, ecosystem stability, and socio-economic development. As the world’s second most populous developing country and a region highly sensitive to climate change, China has experienced a continued expansion in the frequency and impact range of heatwaves, leading to a growing health risk for populations exposed to extreme heat. In recent years, the Chinese government has attached great importance to the prevention and response to extreme heat events to mitigate their impacts on public health and socio-economic systems. Under this context, conducting in-depth research on heatwaves—particularly their population exposure characteristics and associated health risks—holds both significant theoretical value and practical relevance. This study adopts a dual “natural–social” attribute perspective on heatwaves, guided by disaster system theory and a risk management framework. Utilizing multi-source temperature datasets, population data, and socio-economic statistics, it integrates multi-definition comparisons, multi-scale analysis, multi-group assessments, and multi-scenario simulations to systematically reveal the spatiotemporal evolution patterns of heatwaves and population exposure across China during both historical and future periods. Furthermore, a composite index-based health risk assessment framework was developed, incorporating interpretable machine learning methods to elucidate the spatiotemporal dynamics and mechanisms of heatwave-related health risks. Ultimately, this study integrates spatially based risk-dominant typologies with a temporally oriented whole-process risk management framework to propose comprehensive strategies for managing heatwave risks.