Diabetic peripheral neuropathy is a length-dependent, "dying-back" axonopathy: the longest fibres (here, a ~1.2 m tibial-nerve-scale axon running from the spinal cord to the toe) degenerate first and worst, producing the classic distal "stocking-glove" sensory loss. Two chronic-hyperglycemia mechanisms drive it, both real and well-established:
- Polyol-pathway / AGE injury to Schwann cells — excess glucose is shunted through aldose reductase to sorbitol, depleting myelin-maintaining metabolites and thinning the myelin sheath at nodes of Ranvier.
- Vasa nervorum microangiopathy — the small vessels feeding the nerve itself narrow, and the most distal, poorly-perfused segments become ischemic first, compounding demyelination with outright axonal loss.
Segment health is modelled as a length-dependent probability of injury:
burden = clamp((HbA1c − 5.5) × years × k, 0, 1)
risk(x) = burden × (x / L)^2 [x = distance from spinal cord, dying-back]
health(x) = 1 − risk(x) (ischemia toggle steepens the distal falloff)
For myelinated fibres, conduction is saltatory — the action potential jumps node-to-node, with velocity approximated by Rushton's rule v ≈ 6·d (d = axon diameter in μm, v in m/s). Demyelination lengthens the effective delay at each node (slowing v); once local health falls below a threshold the internode can no longer regenerate the impulse and conduction blocks — the fibre goes numb distal to that point, which is exactly why diabetic neuropathy first presents as loss of sensation in the toes, not the thigh. This 2D view plots conduction velocity directly against distance along the fibre, alongside the animated impulse.
Controls: the two sliders set cumulative metabolic injury; the ischemia toggle adds distal microvascular damage; "Fire action potential" sends a real impulse down the fibre so you can watch it slow through demyelinated internodes and stop (or succeed) at the point degeneration has reached.