Multiscale computational models of the neuromuscular system provide a better understanding of the phenomena involved in motor control, as they are based on well-known biophysical, neurophysiological, and biomechanical mechanisms, whose quantifiers are available for analysis. The use of these models for the study of peripheral neuropathies can be useful to obtain a better understanding of the impacts of these neuropathies, as well as serving as a basis for suggesting new proposals for early diagnosis and follow-up. Despite that, multiscale models of the neuromuscular system have not been widely used to study these types of diseases. In the present work, a phenomenological model of motor and sensory axons is proposed, capable of representing axons of subjects with peripheral neuropathies. These axonal representations were inserted into an existing multiscale model of the neuromuscular system, which was elaborated and validated for healthy subjects by previous researchers. As the starting model represents leg muscles, the initial focus was on neuropathies that mainly affect the lower limbs, especially acute inflammatory demyelinating polyneuropathy, a variant of Guillain-Barré syndrome. The existing axon delay model was modified so that it would be possible to represent the axonal conduction velocity of motor and sensory axons, as well as the number of functional motor units. To obtain a representative model of subjects affected by the neuropathy, the model parameters were strongly based on electrophysiological findings of patients available in the literature. Different neuropathy scenarios were simulated in two motor control tasks (force task and position task). The results suggested that it is possible to detect changes in quantitative features of both torque and surface electromyography signals, in force and position tasks, in patients with the studied polyneuropathy. These findings can complement qualitative and quantitative assessments classically used in clinical neurology, aiding early diagnosis and follow-up of the disease. In addition, the developed model allows straightforward adaptation for other conditions not covered in the present research, being only necessary to make the appropriate changes in the model parameters.