The chronic complications of diabetes are thought to be caused by an interaction between hyperglycemia, or other metabolic consequences of insulin deficiency, and independent genetic or environmental factors that are poorly defined. Several potentially relevant biochemical sequelae to hyperglycemia have been identified in tissue susceptible to diabetic complications. Among these, a rise in tissue sorbitol secondary to concentration-dependent activation of polyol pathway activity by glucose, and an accompanying fall in tissue myo-inositol and Na-K-ATPase activity have recently been linked to a self-reinforcing cyclic metabolic defect that accounts for rapidly reversible slowing of conduction in peripheral nerve in diabetes. Impaired Na-K-ATPase activity also appears to be responsible for intracellular Na+ accumulation and resultant localized axonal paranodal swelling that characterizes diabetic neuropathy in both humans and laboratory animals. These swellings are thought to be responsible for the subsequent disruption of the nodal apparatus (axo-glial disjunction) and some component of the loss of large and small myelinated fibers. Recent studies have suggested that microvascular insufficiency may also contribute to diabetic neuropathy, especially in non-insulin-dependent diabetes. Aldose reductase activity is concentrated in endoneurial vessels, and similar biochemical mechanisms (ie, sorbitol accumulation, myo-inositol deficiency, and impaired Na-K-ATPase activity) are thought to be operative in the endoneurial microvessels in diabetes. Administration of an aldose reductase inhibitor to patients with diabetic neuropathy is associated with repair of damaged nerve fibers and the appearance of newly generated fibers, presumably secondary to metabolic correction within the nerve fibers themselves or their supporting microvasculature.(ABSTRACT TRUNCATED AT 250 WORDS)