To test the hypothesis that latencies of evoked potentials in IDDM patients are delayed compared with healthy control subjects during euglycemia, and that insulin-induced hypoglycemia causes further latency delays of evoked potentials to occur.
We recruited 23 IDDM patients (27.9 ± 1.6 yr of age, HbA1c 6.7 ± 0.3%, without sensory or autonomic neuropathy) and 26 unequivocally healthy subjects who were carefully matched for sex, age, and body mass index to serve as the control group (18 men and 8 women, 28.4 ± 1.6 yr of age, 22.6 ± 0.7 kg/m2), for a controlled, prospective study. Sequential euglycemichypoglycemic clamps were performed with stable glycemic plateaus of 5.6, 3.3, 2.2, and 1.7 mM, at which the patients' and healthy control subjects' neurophysiological functions were evaluated. The methodological armamentarium included the measurement of brainstem auditory, middle-latency auditory, and somatosensory evoked potentials that assessed conduction velocity in corresponding neural structures and information processing in the midbrain and auditory cortex.
Multiple analysis of variance revealed a significant overall difference of brainstem auditory evoked potential latencies during euglycemia between the study group and healthy control group (F = 3.41, P < 0.03), which was mainly attributable to latency delays of wave III (F = 6.60, P < 0.02), V (F = 9.19, P < 0.01), and interpeak latency 1-V (F = 2.82, P < 0.07). Repeated analysis of variance measures detected a significant latency delay of the major wave Pa of the middle-latency auditory evoked potentials during hypoglycemia (F = 4.4, P < 0.02), which rapidly returned to normal after reinstitution of euglycemia.
In IDDM patients, chronic, structural CNS changes already appear at the brainstem level during euglycemia. Functional, reversible CNS changes, however, seem to emerge during acute deviation from glucose homeostasis in more rostral brain regions.