Can we do without mydriasis in multifocal ERG recordings?

The purpose of this study is to determine whether mfERG recordings with natural pupils approximate recordings obtained in mydriasis if the stimulus luminance is adjusted. If so, a bright monitor could obviate pupil dilation, rendering clinical management considerably easier. In 11 eyes of six subjects, we took four mfERG recordings with a resolution of 61 scaled hexagons per eye, two with natural pupils and two in mydriasis with a luminance of 150 cd/m(2) and 500 cd/m(2), respectively. Conditions will be abbreviated as "LoNat, HiNat, LoDil, and HiDil," where Lo/Hi stands for low vs. high luminance and Nat/Dil stands for natural vs. dilated pupils. Response densities and peak times were averaged for the central seven and the remaining 54 hexagons. An ANOVA found highly significant effects for both amplitude and peak time between the four recording conditions (P always <0.001). However, for the two "middle" conditions, HiNat and LoDil, amplitudes and peak times did not differ markedly. Retinal illuminances of the conditions LoNat and HiNat each differed from the conditions LoDil and HiDil, respectively, by a factor of 5.4. Amplitudes increased and peak times decreased in the order of conditions LoNat, HiNat, LoDil, HiDil. When data were corrected for the Stiles-Crawford effect, there was an inversion of effective retinal illuminances of the recordings HiNat and LoDil. The mfERG guidelines correspond to the condition LoDil; for all practical purposes, the amplitudes and peak times are the same when the HiNat condition is used. Thus, when clinical constraints render pupil dilation undesirable, the mfERG can be recorded with natural pupils when monitor luminance is increased by five times (if available).