Walking and running activities (W&RAs), encompassing strolling, slow walking, brisk walking, jogging, and running, hold significant importance as popular forms of exercise within urban parks. Recognized for their efficacy in promoting public health and preventing chronic diseases, understanding the nuanced impact of pathway features on W&RAs is crucial for advancing health-centric urban park planning. Based on extensive, high-frequency field observation data, we utilize multiple OLS regression models and univariate OLS regression models to investigate the relationship between urban park pathway features and variations in W&RAs, specifically examining activity density and intensity. Subsequently, we propose corresponding pathway optimization strategies. Our findings highlight the primary determinants, with vegetation coverage ratio, path type, and security facility density influencing activity density; and control value, time required to reach the nearest entrance, and pavement type influencing activity intensity. Significantly, increased vegetation coverage enhances density, while interconnected spaces and improved accessibility elevate intensity. In conclusion, our study delineates key features that merit prioritization, specifies their optimal ranges and proposes optimization design strategies for urban park pathways. By shedding light on these considerations, our research contributes valuable insights to the realm of health-oriented urban park planning and design.