User engagement behavior and health check purchasing decisions are crucial for e-health service platforms (e-HSPs) to enhance service design and improve profitability. However, the relationship between user engagement behaviors and the factors motivating users to make health check purchasing decisions remains unclear. In this study, we employ a Markov chain modeling approach to understand the underlying mechanisms of user behaviors and their transition states on an e-health service platform. We tracked 9,903 users with 38,603 behavioral activities who had experience purchasing health-check products on an e-health service platform offering various health services. We analyzed the dynamics of user behaviors by modeling the transition probabilities across different participation states. Our analysis identifies three distinct behavioral transition patterns: one-way transitions, cyclical transitions, and self-cyclical transitions. Additionally, we identified three e-health literacy development paths: health knowledge learning, self-health management, and support exchange. Among these paths, the support exchange path has the highest transition probability leading to health check purchasing decisions. Users who actively search for health information are more likely to make health check purchases than those who passively receive health information. Our study offers valuable practical implications for the design and management of online health service platforms.