Clinical and neuropathological criteria for the diagnosis and staging of diabetic polyneuropathy.

Scored symptoms, neurological deficits, detection threshold of cutaneous sensation and parameters of nerve conduction were compared with quantitated neuropathological abnormalities in the sural nerve in 47 healthy subjects and 36 diabetic patients, 32 with and 4 without neuropathy. The fifth percentile line of a new Index of Pathology, which combines loss of myelinated fibres and abnormality of the remaining fibres, was found to provide a sensitive and reliable minimum neuropathological criterion for the diagnosis of polyneuropathy. Abnormality, as assessed by two clinical evaluations, similarly separated healthy subjects and diabetic patients into those with and without neuropathy. For the detection of diabetic polyneuropathy, vibration sense was more sensitive than touch-pressure or thermal cooling. Abnormalities of nerve conduction were found to be both sensitive and reliable in the detection of polyneuropathy. Velocity was most frequently abnormal, but only slightly more often than F wave latency and amplitude. We conclude that abnormality, as judged independently from two different types of evaluation, provides a sensitive and reliable minimal criterion for the diagnosis of neuropathy. Although symptoms, neurological deficits and abnormalities of nerve conduction are statistically associated, they should be evaluated separately to provide adequate characterization.