The amplitude of the electroretinogram (ERG) has long been reported to be reduced in highly myopic eyes. The first report of ERG changes in myopia is usually attributed to Karpe,⁸ who reported abnormal ERGs in four eyes with high myopia. Several early reports on the ERG in myopia showed that ERG amplitudes were generally within normal limits up to 8 diopters of correction provided that there were no degenerative alterations in the retina but were abnormal with more severe myopia and associated retinopathy.^1,4,5,7 There was also some evidence in the early literature suggesting that photopic responses were affected earlier, and possibly to a greater degree, than scotopic responses,^1,2 although this observation has not been replicated in subsequent studies. With the development of more sophisticated and sensitive recording techniques, ERG abnormalities have been observed with less severe myopia.

Myopia is generally associated with change in eye shape resulting from elongation of the optic axis. Several studies have shown that ERG amplitudes are inversely proportional to axial length.^15,16,21 Pallin¹⁵ demonstrated a significant negative correlation between ERG b-wave amplitude and the length of the optic axis, which was measured with ultrasonography. This study also demonstrated a high positive correlation between refractive error and the length of the optic axis (figure 50.1), with b-wave amplitudes higher in hyperopic and emmetropic eyes than in myopic eyes. Sex differences in ERG amplitudes have also been attributed to gender differences in average axial lengths.¹⁵ The dependency of ERG amplitudes on axial length was confirmed in subsequent studies,^16,21 although the shape of the posterior segment, rather than the size and axial length alone, appeared to be the critical determinant of the saturated amplitude in a study reported by Chen et al.³

Figure 50.1.  

The relationship between refractive error (in diopters of correction) and the length of the optic axis measured with ultrasonography. (Data from Pallin O.¹⁵)

Some studies have reported selective losses in the b-wave of the electroretinogram in myopia, a finding that would imply differential effects on signal transmission from the photoreceptors to the proximal retina. However, deficits at the level of the b-wave seem to be more common in patients with high myopia who also have chorioretinal degeneration, atrophy, and thinning of the posterior sclera.^2,17 Perlman et al.¹⁶ reported that all myopes demonstrate subnormal amplitudes but a normal waveform morphology, defined by a normal ratio between a- and b-wave amplitudes. However, the characterization of ERG responses in hypermetropic eyes in the Perlman et al.¹⁶ study was more complicated. While all myopic eyes had a- to b-wave amplitude ratios that were within normal limits despite generalized amplitude reductions, hypermetropic eyes had a- to b-wave amplitude ratios that were either subnormal, normal, or hypernormal. For example, the group of patients with subnormal a- to b-wave amplitude ratios was characterized by a relatively large a-wave and a subnormal b-wave, whereas the reverse applied to the group with hypernormal a- to b-wave amplitude ratios. These findings in the hypermetropic eyes were consistent with variable effects on the transmission of signals from the outer retina. However, no definite relationship was found between axial length of the eye and the a- to b-wave amplitude ratios, although there was an inverse relationship between axial length and the amplitude of the dark-adapted ERG evoked by a dim stimulus within each group. Westall et al.²¹ did not show a selective loss of b-wave amplitude across a broad range of refractive error in patients who were carefully screened to exclude pathological defects.

Previous studies have reported a selective loss of short-wavelength cone (S-cone) spectral sensitivity and reduced cortical evoked potentials to short-wavelength light in myopia.^10,11 To investigate the retinal contribution to the S-cone deficits that had previously been documented, Yamamoto et al.²² recorded cone-mediated ERGs to different chromatic stimuli in myopic and normal eyes to elicit short-wavelength-sensitive (S-), and mixed long- (L) and middle- (M) wavelength-sensitive responses. The S-cone and L,M-cone b-wave amplitudes decreased progressively with increasing myopia and were significantly lower in high myopia compared with emmetropic eyes. The S-cone and the L- and M-cone ERGs were almost equally affected in the myopic eye, with S-cone function decreasing at a slightly slower rate compared to L,M-cone ERGs as refractive error increased. These results suggest that the reduction of S-cone sensitivity may originate from inner retinal or higher-order changes that would not be reflected in the conventional ERG.

Westall et al.²¹ performed an elegant study investigating the relationship between axial length, refractive error, and ERG responses. The extensive ERG protocol that was used included stimuli that conformed to the standards set forth by the International Society for Clinical Electrophysiology of Vision, which was the first study to have done this. ERGs from 60 subjects were recorded with varying degrees of myopia but without evidence of myopic retinopathy. The study revealed a significant difference between subjects with high myopia and subjects with small refractive error for V_max, the maximum saturated amplitude, the b-wave of the cone response, and the summed dark-adapted oscillatory potentials. A linear reduction in the logarithmic transform of the ERG amplitude with increasing axial length was demonstrated (figure 50.2). However, there were no differences in the implicit time of peak components, in the ratio of b- to a-wave amplitude derived from the maximal response, and in the semisaturation intensity estimated from the Naka-Rushton fit to the rod intensity series. In addition, they show relatively larger attenuation of later than earlier photopic oscillatory potentials, which suggests anomalies in pathways beyond the photoreceptors, possibly in the OFF-bipolar cell pathway in eyes with progressively greater myopia.

Figure 50.2.  

ERG amplitude plotted against axial length. The y-axis shows the log ERG amplitude and the x-axis the axial length of the eye. Crosses represent individual data points. Regression lines (thick lines) and inner and outer ranges (thin lines) represent the expected value (from regression) plus and minus two standard deviations. Regression equations and R² values are shown for V_max (A), b-wave amplitude of the cone response (B), and the sum of the dark-adapted oscillatory amplitudes (C). (Data from Westall CA, Dhaliwal HS, Panton CM, et al.²¹)

The finding of reduced ERG amplitudes in myopia using conventional full-field stimulation has also been reported when only the macula or adjacent regions are tested. Ishkawa et al.⁶ examined the relationship between the amplitude of the photopic ERG recorded from the macula (focal ERG) and the degree of myopia. The amplitudes of the a- and b-waves of the focal ERG were significantly smaller than those of normal eyes, and the amplitudes were inversely proportional to the degree of myopia. The abnormal amplitude was interpreted by the authors to suggest a reduction in the number of macular cones that were contributing to the signal. More recently, Kawabata and Adachi-Usami⁹ investigated local retinal function in patients with various degrees of myopia who had undergone multifocal electroretinography (mERG). In the mERG recording technique, small areas of the retina are stimulated simultaneously and local contributions to a massed electrical potential are extracted from a continuously recorded ERG (see chapter 17). The technique permits the mapping of the topography of local retinal function. Again, amplitudes were reduced as refractive error increased, the rate of change being slower for N1, the first negative peak of the mERG waveform, than for P1, the first positive peak. The latencies of N1 and P1 were also delayed, but the rates of change were more similar for each component in comparison to the amplitude parameter. What was particularly interesting about this study is the selective loss of function in more peripheral retinal regions in the myopic groups, as evidenced by the more rapid loss of local amplitudes in the parafoveal areas (figure 50.3). This finding suggests that peripheral cones may be more sensitive to changes in eye shape produced by myopia, perhaps related to a more rapid change in the density of cones in these areas.

Figure 50.3.  

A, Multifocal electroretinographic topographies (three-dimensional view). The response densities (nV/degree²) decreased in all measured retinal fields as the refractive errors increased. B, Summed responses from 103 individual hexagons. N1 and P1 amplitudes of the summed response decreased as the refractive error increased. Both L1 and L2 latencies for high myopia were significantly delayed in comparison to those of emmetropia/low myopia. C, Summed responses from six concentric rings centered on the macula. Response amplitudes were suppressed to a greater degree in the more peripheral rings in high myopia. D, Summed responses over four quadrants (superior-nasal, superior-temporal, inferior-nasal, and inferior-temporal regions). (Data from Kawabata H, Adachi-Usami E.⁹)

Implicit time of ERGs are generally within normal limits with high myopia. Perlman et al.¹⁶ reported normal implicit times for both dark- and light-adapted responses. Lemagne et al.¹³ also reported normal implicit times for rod-mediated responses to a single flash after 20 minutes of dark adaptation. Cone ERG implicit times to white flashes have been reported to be normal in patients with refractive errors ranging from −5.0 to −10 diopters,¹² and cone ERGs to chromatic stimuli have also been reported to be within normal limits.²² In a group of patients with gyrate atrophy and high myopia, implicit times were normal or mildly elongated,²⁰ and in a complete form of congenital stationary night blindness, implicit times have been reported to be within normal limits in patients with a mean of −8.0 diopters of myopia.¹⁴ Ishikawa et al.⁶ also reported normal implicit time of macular focal ERGs in eyes showing only tigroid fundus, but those associated with posterior staphyloma involving the macula were significantly delayed. Kawabata and Adachi-Usami⁹ reported mildly delayed latencies for the mERG recordings in myopia. However, Westall et al.²¹ reported no significant difference in implicit times for a broad range of ERG stimuli and refractive errors.

The cause of the changes in ERG amplitudes with refractive error is not well understood. Increased ocular resistance due to the elongated eyeball in myopia was thought to be the major cause of decreased ERG amplitudes.¹⁵ Pallin¹⁵ argued that the current passing from the retina to the surface of the eye passes through highly resistant conductors consisting of both intraocular and extraocular tissues and that the density of these conducting tissues increases with axial length. Lemagne et al.¹³ designed a specially built system that allowed simultaneous recording of both the ERG and the resistance between the active and reference electrodes used for recording the ERG. The authors reported that ERG amplitudes were negatively correlated with the resistance measurements. However, there was broad variability in resistance measurements that was not correlated with refractive error, and the correlation between the resistance and amplitude measurements was relatively weak. Furthermore, resistance was measured between the electrode on the eye and a reference electrode on the midline of the forehead, which would not have allowed a specific test of resistance of ocular tissue. Whether the correlation between resistance and ERG amplitude holds when the active and reference electrodes are both positioned on the eye, and the relationship of the resistance measurements to axial length, has yet to be determined.

Chen et al.³ dismiss the resistance argument as an explanation of reduced amplitudes because ocular current would be expected to increase with resistance, and according to these authors, a lower, not higher, resistance would be needed to explain the reduced ERG amplitudes in myopia. In contrast, Chen et al.³ tested between two alternative hypotheses about the mechanism of reduced ERG amplitudes in myopia. One hypothesis, called the stretched retina hypothesis, predicted a reduced retinal sensitivity but a normal saturated amplitude because there would be fewer receptors per unit area of retina and more space between receptors as a result of the elongated eye. Thus, as a result of the reduced ability of photoreceptors to capture photons, higher intensities of light would be required to elicit a threshold response (sensitivity), but the saturated amplitude (responsivity) would be unaltered provided that sufficient light is provided. An alternative hypothesis in which function but not retinal spacing varies, predicted reduced saturated amplitudes and normal sensitivity. They reported significant reductions in the saturated amplitude with progressively higher degrees of myopia but normal retinal sensitivity. Thus, the data were consistent with the decreased cell responsivity hypothesis rather than the enlargement of the eye with wider spacing of retinal elements. In addition, the shape of the posterior segment, which was characterized with magnetic resonance images, rather than the size and axial length alone appeared to be the critical determinant of the saturated amplitude. It is not yet understood why retinal cells in myopia would have a lower responsivity. The finding of normal sensitivity and reduced saturated amplitudes has recently been confirmed.²¹

Optical factors have also been implicated as playing a role in the reduction of ERG amplitudes in myopia. Assuming a similar stimulus intensity and pupil size, retinal illuminance may be lower in the myopic eye compared to the normal eye. Retinal illuminance refers to the density of light falling on a unit area of retina, and with an elongated eye, light is spread over a wider area, thereby decreasing retinal illuminance. If the myopic eye functioned to shift the intensity-response curve so that a given light intensity is less effective, then retinal illuminance could be adjusted to equate amplitudes. The sensitivity but not the saturated amplitude of the myopic retina would be expected to be altered. However, several studies have demonstrated that saturated b-wave amplitudes are lower for the myopic eye,^3,9,21 ruling out the possibility that reduced retinal illuminance is the explanation for the reduced ERG amplitudes.

To summarize, ERG amplitudes are negatively correlated with refractive errors that are caused by elongation of the optic axis, with generally subnormal amplitudes for the highly myopic eye without pathological changes. This finding applies to both rod- and cone-mediated vision. Parafoveal and peripheral cone-mediated responses may be affected to a greater degree than more central regions, possibly related to the differences in the packing density of cones, and the deficits are slightly greater for the L- and M-cones compared to S-cones, possibly related to the greater number of L- and M-cones. Implicit timing of the peak components of the ERG are generally reported to be normal regardless of refractive state and with normal appearing fundi. The cause of the reduced amplitudes in myopia is not well understood. Differences in electrical and optical factors and loss of photoreceptor density have been implicated as the cause of the reduced ERG amplitudes in myopia.