This paper presents analytical results on the BER degradation of QPSK signals for BAN antennas for the case where the combined outcome of shadowing and multipath fading emerge simultaneously and different types of human walking models are employed in the analysis. The results show that a probability of a low signal level is higher than that for the Rayleigh distribution owing to shadowing caused by the movement of the arms, resulting in the significant impact of arm-swinging on the BER performance of body-attached BAN devices. Furthermore, statistical measurements of human walking motion using a number of human subjects are conducted. Using the measured data, dynamic analytical models are created to show that different types of human walking models have appreciable effects on the BER performance of BAN antenna systems.