Twenty-three subjects with early perimetric defects (MD ≤ 6 dB) in the worse eye were enrolled. The average age of the subjects was 66.8 ± 7.06 years. There was no significant difference between the ages of normal and glaucomatous subjects (P = 0.5; t-test). Six participants were men, and the rest were women. HFA defects were bilateral in 6 subjects and unilateral in 18; 29 eyes of 23 patients were analyzed. The average MD of eyes with HFA defects was −1.75 ± 1.8 dB, with average PSD of 3.9 ± 2.12.
The mean amplitude (±SD) of eyes with HFA defects detected by BonY, HLA, and LLA stimulation were 655 ± 142, 713 ± 120, and 569 ± 127 nV, respectively (
Fig. 3). These represented reductions of 14.6% for LLA, 15.5% for BonY, and 8.6% for HLA compared to average amplitudes of the normative database. With all three stimuli, amplitudes of eyes with glaucoma were significantly lower than those of the normative database (
P < 0.0001,
P = 0.0007, and
P = 0.04 for BonY, LLA, and HLA, respectively).
Multivariate analysis using multiple linear regression was performed on the combined (normal and glaucomatous subjects) dataset, to analyze the effects of glaucoma, age, and sex on the amplitudes of each stimulus condition. For all three stimuli, amplitudes were significantly affected by glaucoma (P < 0.0001, P = 0.004, P = 0.036 for BonY, LLA, and HLA, respectively), but not by age (P = 0.8, 0.4, and 0.5 for BonY, LLA, and HLA, respectively) or sex (P = 0.3, 0.5, and 0.5 for BonY, LLA, and HLAZ, respectively).
Table 1 shows the performance of each of the tested mfVEP stimuli, in comparison to defects identified by HFA. Repeated-measures ANOVA revealed significant differences in ASI scores between the three groups (
P = 0.006). A Tukey post hoc multiple-comparison test revealed significant differences between HLA and LLA (
P < 0.05) and between HLA and BonY (
P < 0.05), but no significant differences between LLA and BonY (
P > 0.05).
Table 2 compares results of BonY versus LLA stimulation for hemifields that were abnormal on subjective visual field testing.
Figures 4 and
5 are case examples.
Figure 4 shows consistent defects identified on all methods of stimulation, while in
Figure 5, the defect was not detected by HLA stimulation.