A Review of Indoor Thermal Comfort Studies on Older Adults in China

This review systematically examines research on indoor thermal comfort for older adults conducted in China since 2010. It highlights several existing research gaps, including the lack of a systematic understanding of environmental and individual influences, limitations of thermal comfort models, challenges in their optimization, and inadequate integration of intelligent technologies. Results indicate that environmental factors usually exert a greater impact on the elderly’s neutral temperature than individual factors. Thermal comfort models differ in predictive accuracy, data requirements, and applicability. The adaptive predicted mean vote (aPMV) model is better suited for group-level assessments. Machine learning (ML) models, featuring high flexibility and accuracy, are more appropriate for personalized predictions. In addition, physiological parameters could play a critical role in thermal assessments. When integrated with ML models, physiological parameters could further improve predictive accuracy. When integrated into artificial intelligence (AI) and Internet of Things (IoT) systems, forehead and back skin temperatures could act as early-warning indicators during heat exposure, while lower-limb temperatures are more indicative of thermal discomfort during cold exposure. Overall, this review summarizes current progress and limitations, offering a reference for the development of user-friendly modeling and intelligent temperature-control systems.