Tracking the hidden dynamics of proprioception

Abstract Every movement unfolds with a simple question: Where is my body? The nervous system answers through proprioception—the sense of limb position (static position sense) and movement (dynamic proprioception). Although position sense has been well characterized, dynamic proprioception has remained difficult to isolate and measure. Here we introduce a continuous proprioceptive tracking paradigm, coupled with computational modelling, that captures dynamic proprioception in real time. We first establish that this approach is sensitive, reliable and efficient. Leveraging this method, we then show that dynamic proprioception provides faster and more faithful estimates of limb state than vision, dominates multisensory state estimation when vision is also available, and is not correlated with conventional measures of position sense. Together, these findings provide a new quantitative framework for characterizing dynamic proprioception in health and disease.