Brain-controlled wearable prosthesis system which uses wireless human communication technology provides new ways for the disabled to interact with their surroundings. To realize efficient data transmission between the robot arm and whole body-scattered or implanted sensors, it is necessary to establish reliable human body communication links with the help of human-centered wireless technology. In this paper, the path losses from different locations of the arm surface to the whole body area are calculated and statistically studied based on an anatomical adult male model below 100 MHz. According to the surface wave propagation characteristics, three human body receiving regions, that is, the head, upper-body, and whole-body areas are specified to construct comprehensive and accurate path loss models. The results show that the linear model is more suitable for the head region, the Friis-based logarithmic distance path loss model is appropriate to describe the upper-body area calculated path losses, and the logarithmic distance path loss model with a surface wave-based linear regulation term is more accurate for long-range whole-body area path loss characteristics. The shadow fading is normally distributed around the derived average path loss. The fitted key parameters of the path loss models are also summarized to provide a reliable basis for further research on the brain-controlled robotic arm communication system.