Variability of Retinal Oxygen Metrics in Healthy and Diabetic Subjects

Abstract

Purpose: Previous studies have reported alterations in total retinal blood flow (TRBF), oxygen delivery (DO2), oxygen metabolism (MO2), and oxygen extraction fraction (OEF) due to retinal diseases. The purposes of the current study were to determine variabilities and establish normal confidence intervals (CIs) for these metrics.

Methods: A total of 22 healthy and 14 diabetic subjects participated in the study. Retinal vascular oxygen saturation (SO2) and TRBF were measured by oximetry and Doppler optical coherence tomography, respectively. DO2, MO2, and OEF were calculated from SO2 and TRBF measurements. Means, standard deviations (SDs), and CIs of metrics were determined in healthy subjects. Intra-visit variability was determined by the mean SDs of repeated measurements. Inter-visit variability was determined by the difference of measurements between two visits.

Results: TRBF was 44 ± 15 µL/min (95% CI, 37-51) in healthy subjects. Intra-visit variabilities of TRBF were 5 µL/min and 6 µL/min in healthy and diabetic subjects, respectively. Inter-visit variability of TRBF was 3 µL/min in diabetic subjects. DO2, MO2, and OEF were 8.3 ± 2.9 µLO2/min (95% CI, 7.0-9.6), 3.2 ± 0.9 µLO2/min (95% CI, 2.8-3.6), and 0.40 ± 0.08 (95% CI, 0.36-0.43), respectively, in healthy subjects. Inter-visit variabilities of DO2, MO2, and OEF were 0.6 µLO2/min, 0.1 µLO2/min, and 0.03, respectively, in diabetic subjects.

Conclusions: The findings established variabilities and normal baselines for TRBF, DO2, MO2, and OEF measurements in a small cohort of subjects.

Translational relevance: The variability and normal baselines of retinal oxygen metrics may be useful for diagnosing and monitoring patients with retinal diseases.

Figure 1.

(A) Example of an en face OCT image of the optic nerve head in a diabetic subject. Blood flow measurements were obtained in veins detected in multiple en face planes. (B) A retinal image acquired in the same subject. Retinal vascular oxygen saturation was measured in blood vessel segments within the concentric rings.

Figure 2.

Inter-visit variability of TRBF displayed using the Bland–Altman plot. The difference in measurements between visits is plotted as a function of the average of measurements at two visits. The solid line shows the mean difference (bias) and the dashed lines denote upper and lower 95% limits of agreement.

Figure 3.

Inter-visit variability of DO₂ and MO₂ (A) and OEF (B) displayed using Bland–Altman plots. The differences in measurements between visits are plotted as a function of the average of measurements at two visits. The solid line indicates mean difference (bias), and the dashed lines denote upper and lower 95% limits of agreement.