Visual-Inertial Odometry (VIO) is commonly used for unmanned aerial vehicle (UAV) navigation in GNSS-denied areas. However, drift due to sensor noise and cumulative errors presents a significant challenge, especially in environments with sparse features or long trajectories. In this work, we introduce a tightly-coupled fusion approach that fuses visual, inertial and Ultra-Wideband (UWB) distance measurements for UAV navigation. We specifically utilize UWB with a single anchor as the global sensor, effectively reducing localization drift and enhancing accuracy. Furthermore, we innovate the use of distance data with a “derivative-enhanced” approach, coupling distance derivatives extracted from distance data with positions and velocities. Simulation experiments were conducted using the public EuRoC dataset with simulated UWB measurements. Our method, when compared with the latest single anchor visual-inertial-UWB odometry, demonstrates a notable improvement in reducing odometry system drift and enhances localization accuracy by 14.7%.