Adjusting for age removed evidence of correlations between rim area and visual senstivity in healthy eyes

Abstract

 

Purpose

 

There is a clear correlation between structural and functional measures for groups of patients with glaucoma, but whether this extends to healthy eyes is unclear. In this analysis performed on prospective data collected in an observational study, we tested whether the rim area (RA) and visual sensitivity are correlated in participants with healthy eyes.

 

Methods

 

Pearson linear correlation, and Spearman and Kendall rank correlations were obtained between RA and mean sensitivity (MS) and between RA and mean deviation (MD), which is similar to MS after adjusting for age. A dataset of 1162 healthy eyes from 655 participants were selected from the Diagnostic Innovation and Glaucoma Study (DIGS) or the African Descent and Glaucoma Evaluation Study (ADAGES) studies. As a control calculation, we obtained correlations for a dataset of 1136 eyes with glaucoma from 747 patients enrolled in DIGS or ADAGES. Table 1 shows the demographics for the datasets included in this study.

 

Results

 

Table 2 shows the correlations for both the Healthy and the Glaucoma datasets. In healthy eyes, significant correlations of more than 0.05 were found between (non-age-adjusted) MS and RA. Non-significant correlations of less than 0.05 were found for (age-adjusted) MD and RA. In eyes with glaucoma, all correlations between MS and RA and between MD and RA were significant, ranging between 0.19 and 0.31. The correlations between MS and RA and between MD and RA in glaucoma patients were significantly different than those obtained in healthy eyes.

 

Conclusions

 

Once visual sensitivity is adjusted for age, no evidence was found of correlation between RA and visual sensitivity in healthy eyes. Observed RA did not explain more than about 0.5% of the observed visual sensitivity, and vice versa. Future studies of joint structure and function progression should consider a change in association between eyes with glaucoma and healthy eyes. Correlation analyses on other, large datasets with healthy eyes are necessary to corroborate this finding of absence of association between MD and RA.  

 

Table 1. Distribution of age, mean deviation, and rim area for the selected healthy and glaucoma DIGS-ADAGES datasets. Percentiles for 95% interval and the inter-quartile range are also shown.

 

 

Table 2. Pearson, Spearman, and Kendall correlations for structure and function and corresponding p-values and R².