Effect of blood glucose on chromatic and achromatic visually evoked potentials in type I diabetics and normals

Purpose. Selective effects of short-term variations in blood glucose on the S-cone pathway have been demonstrated psychophysically using blue test spots on bright yellow backgrounds (Volbrecht, Schneck and Adams, ARVO 91). Here we extend those findings using an objective measure - the visually evoked potential (VEP) - and stimulus conditions that isolate individual color mechanisms without using chromatic adapting backgrounds. Methods. Stimuli were sinusoidal gratings modulated along 3 axes in MacLeod, Boynton, Krauskopf, Lennie and Derrington color space. Two isoluminant chromatic stimuli (one modulating only S-cones, the other modulating L/M- but not S-cones) and a luminance-varying achromatic stimulus were used. Observers were 8 young (mean age: 29.9 years) Type I diabetics without retinopathy and age-similar normals. Monocular VEP's and blood glucose (BG) levels were recorded repeatedly over several hours, and the relationship between BG and VEP latency for each stimulation axis was determined. Results. As a group, the diabetics showed significantly longer S-axis VEP latencies when BG was high than when BG was low. Individually, 5 of the 8 diabetics showed significant associations between BG level and S-axis latency. There was no group effect of blood glucose on LM or achromatic VEP latency. However, 2 of the 3 diabetics who did not show longer S-axis latency at high BG levels showed shorter achromatic VEP latencies when BG was high. The magnitude of the BG effect was much smaller for achromatic than S stimulation (mean differences, 2.2 vs. 20.4 msec). Non-diabetics' achromatic VEP's are like diabetics' S-axis VEP's in that latency is longer at higher BG. Conclusions. VEP's under neutral adaptation show that elevated BG appears to selectively impair S-pathway function in diabetics without retinopathy, but not in non-diabetics.