The effects of gender and age on the range of the normal human electro-oculogram

Abstract

Purpose

To define the normal ranges for the slow oscillations (SO) and fast oscillations (FO) of the electro-oculogram (EOG) recorded to International Society for Clinical Electrophysiology of Vision (ISCEV) standards. The effects of age and gender on the EOG ranges were examined.

Methods

ISCEV standard SOs and FOs were recorded from 121 subjects (51 % male) aged from 7 to 72 years. Study variables for the SO were dark trough (DT) and light peak (LP) amplitudes (µV), times to DT and LP (min), and the Arden ratio (LP/DT amplitude). The FO was fit by a sine wave and peak-to-peak amplitude (µV), phase (°), and peak-to-trough (PT) ratios derived. The effects of age, gender and pupil size on EOG parameters were examined by multiple regression analysis.

Results

The average Arden ratio was 2.5. Arden ratio decreased with age at a rate of 0.13 per decade of age (R ² = 0.14, P < 0.0001). The 5th percentile of the Arden ratio decreased from 2.0 to 1.7 between 10 and 60 years of age. Median time to LP was 9 min (interquartile range 8–9 min). Time to LP was age-dependent and increased by 2 min for subjects over 55 years of age compared with those less than 25 years. EOG amplitudes were greater in women than in men (P < 0.005). The average PT ratio was 1.18, which was not affected by age or gender. Time to reach the light trough of the FO was 40 s, which increased with age (1.1 s/decade). No correlation was observed between Arden ratio and PT ratio.

Conclusions

The major strength of this study is the definition of the normal range and associated lower limits of ISCEV standard EOGs based on recordings from 121 subjects balanced by gender and spanning the 1st through 8th decades of life. Decreased Arden ratio and increased time to LP are associated with aging, which is likely due to the intricate mechanisms involved in generation of the light rise. Differences between the FO and SO with respect to the effects of aging are consistent with separate generation of these two EOG signals.

Appendices

Appendix 1

This appendix outlines the algorithm used by the UTAS E-2000 software (as detailed in the Manufacturer’s Operation Manual) for calculating average saccade amplitude during recording of the SO. For each 15-s interval: (1) A 5-point median filter was applied to the waveform to eliminate spike artefacts. With the median filter, a group of 5 points centered on the point in question were sorted and the value of the point in question replaced by the middle value in the group; (2) the peak to trough amplitude was found for each saccade using the following algorithm. A local maximum (minimum) was defined as a point that was higher (lower) than its 24 nearest neighbors (12 on either side) and that exceeds a predefined threshold. Saccade amplitude was determined from the difference between the maximum and minimum amplitudes; (3) Saccade amplitudes for a 15 s sweep were sorted and the largest and smallest amplitudes discarded; (4) the average saccade for each 15 s sweep was calculated from the remaining values.

Appendix 2

This appendix details the analysis performed by the UTAS E-2000 software (as detailed in the Manufacturer’s Operation Manual) for deriving FO parameters from the fit of a sine wave to the absolute amplitude data recorded during the FO protocol.

The amplitude of the sine and cosine components of the FO was determined using a discrete Fourier transform:

$$A_{sin} = \frac{2}{N}\mathop \sum \limits_{n = 0}^{N - 1} X_{n} \sin (2\pi fT)\quad A_{cos} = \frac{2}{N}\mathop \sum \limits_{n = 0}^{N - 1} X_{n} \cos (2\pi fT)$$

(3)

where N = total number of samples, T is the time for a full saccade, $X_{N}$ is the amplitude of the FO at time nT and f is the frequency of the FO. For our FO recording conditions, N = 450 points, T = 2 s for a full saccade and f = 0.00666 Hz

Peak to trough amplitude (µV), of the sine wave was estimated from:

$$A_{p - t} = 2\sqrt ( A_{cos}^{2} + A_{sin}^{2} )$$

(4)

The mean amplitude of the standing FO was estimated as

$$\bar{X} = \frac{1}{N}\mathop \sum \limits_{n = 0}^{N - 1} X_{n}$$

(5)

The results from Eqs. 2 and 3 are used to calculate the peak-to-trough ratio:

$$R_{p - t} = \frac{{\bar{X} + \frac{{A_{p - t} }}{2}}}{{\bar{X} - \frac{{A_{p - t} }}{2}}}$$

(6)

The phase of the sine wave ($\theta$, deg) measured relative to the time the light turns on in the light phase is estimated from:

$$\theta = \frac{180}{\pi } \tan^{ - 1} \left( {\frac{{A_{sin} }}{{A_{cos} }}} \right)$$

(7)