Group I projections from intrinsic foot muscles to motoneurones of leg and thigh muscles in humans.

1. Group I projections from intrinsic plantar muscles to motoneurones (MNs) of human leg and thigh muscles were investigated. Changes in firing probability of single motor units (MUs) in the tibialis anterior (TA), peroneus brevis (Per brev), soleus (Sol), gastrocnemius medialis (GM), vastus lateralis (VL), semitendinosus (ST) and biceps (Bi) were studied after electrical stimuli applied to: (i) the tibial nerve (TN) at ankle level, (ii) the corresponding homonymous nerve, and (iii) the skin of the heel, to mimic the TN-induced cutaneous sensation. 2. Homonymous facilitation, attributable to monosynaptic Ia excitation, was found in all the sampled units. Early heteronymous excitation elicited by TN stimulation was found in many MUs. Later effects (3-5 ms central delay) were bigger and more frequently observed: excitation in most TA and Per brev MUs, and inhibition in most Sol, GM and Bi MUs and in many ST and VL MUs. The low threshold (approximately 0.5-0.6 x motor threshold) and the inability of a pure cutaneous stimulation to reproduce these effects (except the late excitation in TA MUs) indicate that they were due to stimulation of group I muscle afferents. 3. The early excitation was accepted to be monosynaptic when its central delay differed from that of the homonymous Ia excitation by less than 0.5 ms. Such a significant TN-induced monosynaptic Ia excitation was found in MUs belonging to all leg and thigh motor nuclei tested. Although its mean strength was relatively weak, it is argued that these monosynaptic connections might affect already depolarized MNs. 4. The late excitation found in TA and Per brev MUs is argued to be mediated through interneurones located rostral to MNs. 5. The late suppression, found in most Sol, GM and Bi MUs, and in many ST and VL MUs, was the dominant effect. It was accompanied by an inhibition of the Sol and quadriceps H reflexes at rest, and therefore reflects an inhibition directed to MNs. Its long latency is argued to reflect transmission by interneurones located rostral to MNs (the inhibitory counterpart of non-monosynaptic excitation). 6. The functional implications of these connections are discussed with respect to the requirements of the stance phase of human walking and running.