Single human fingertip mechanoreceptive afferents simultaneously encode multidimensional aspects of touch

Touch using the hands is essential for recognising surface features and manipulating objects, where different aspects are encoded by four main types of low-threshold mechanoreceptors (LTMs) in glabrous skin. Although factors such as movement, vibration and pressure are often studied individually, touch requires their integration to produce complex sensations. We investigated these processes jointly by recording single-unit activity via microneurography from human LTM afferents in the median nerve as periodic gratings slid across their receptive fields, varying the normal force and sliding velocity per trial. Mixed-effects models revealed that fast-adapting type 1 (FA-1) afferent firing was influenced by all three stimulus dimensions - force, velocity and spatial period. Slowly adapting type 1 (SA-1) afferent firing was driven primarily by force and, to a lesser extent, velocity. These findings support the view that FA-1 afferents encode stimulus intensity in an approximately linear manner, while SA-1 afferents function mainly as force detectors, demonstrating that mechanoreceptive afferents provide a complementary and multidimensional representation of the stimulus features during sliding contact. This distributed encoding expands the notion that LTMs are dedicated to single stimulus features, suggesting that tactile information is represented across LTM populations where each class contributes differently weighted inputs to capture skin-object interactions. KEY POINTS: Mechanoreceptors in the skin are divided into different classes that respond well to various aspects of touch, such as pressure and vibration. We recorded directly from low-threshold mechanoreceptors in human peripheral nerves while different dynamic stimuli were moved across their receptive fields, changing the velocity, force and spatial period. Computational modelling of fingertip mechanoreceptor responses showed multiplexity in their encoding strategy, yielding partial convergence in feature representation across afferent classes. The study expands our knowledge on how touch is encoded, where information is represented across mechanoreceptor populations and each class contributes differently weighted inputs to capture skin-object interactions.