Studies on the stability of the clinical electro-oculogram.

The electro-oculogram is variable as a clinical test, but the recording technique has not been standardized, and differences in the protocols for adaptation and stimulation among laboratories may contribute to the variability. To analyze some of these factors, we performed more than 100 electro-oculograms on a single subject under different conditions. Both the Arden ratio and the ratio of light peak amplitude to a stable dark-adapted baseline were independent of pupillary dilation but linearly related to retinal illumination measured in trolands. Between 3.0 and 4.0 log trolands, the values began to level off, but they were difficult to interpret because of subject discomfort above 3.5 log trolands. The Arden ratios were influenced by preadaptation light levels and were roughly 20% higher above 90 cd/m2 than below 45 cd/m2. There was a circadian rhythm in dark trough, baseline and light peak values, but the composite Arden ratio and light peak/baseline ratio showed little circadian effect. The light peak/baseline ratios were slightly more stable than the Arden ratios; the variability (defined as [standard deviation x 100]/mean) was 6.6% and 12.5%, respectively. Our results suggest that stimulus intensities for clinical electro-oculographic testing should be between 3.0 and 3.5 log trolands; pupil dilation is optional. This translates into light levels of 141-447 cd/m2 for a 3-mm pupil and 20-63 cd/m2 for an 8-mm pupil. The reduction in variability and independence from preadaptation achieved by using the light peak/baseline ratio instead of the Arden ratio must be weighed against the time required to achieve a stable dark baseline.