Between-eye correlation of ocular parameters

Abstract

Objective: It is accepted that ocular parameters are more correlated with the fellow eye. This poses a challenge in ophthalmic research. There is a relative scarcity of data concerning the extent of correlation. The aim of this study was to analyze and quantify the correlation of different ocular parameters.

Design: Historical registry analysis.

Participants: All patients examined in a 10-year period (2011-2021) in one academic tertiary medical centre in central Israel.

Methods: Data from optical coherence tomography examinations and biometry of a single examination of both eyes taken at the same time from each patient was included. Pearson's r values were calculated to estimate the extent of correlation.

Results: A total of 17,212 patients were included. Mean age was 73 ± 12.7 years, and 54.5% were female. All examined parameters were highly statistically significantly correlated between eyes (all with p < 0.1 × 10^-36), whereas the strength of correlation differed. Very weak correlation was seen in central macular thickness (r = 0.189), and a weak correlation was seen in anterior-chamber depth (r = 0.379) and retinal nerve fibre layer thickness (r = 0.479). A strong correlation was seen in central corneal thickness (r = 0.754), and a very strong correlation was seen in axial length (r = 0.900).

Conclusions: In a retrospective analysis of optical coherence tomography and biometry measurements of >10,000 patients, all examined parameters were highly statistically significantly correlated. Correlation magnitude varied, with structural characteristics more correlated than functional ones. Including both eyes in an outcome analysis likely will introduce bias. We recommend adjusting for inter-eye correlation in all studies assessing ocular outcome measures.