Determinants Of Retinal Oxygen Saturation In Healthy Subjects

Abstract

Purpose: : Retinal oximetry is a novel imaging modality that non-invasively measures retinal oxygen saturation. As an emerging technology, little is known about normal oxygen measurements in healthy adults. In this study, we have evaluated a cohort of healthy individuals to obtain normative values and determine which factors influence retinal oxygen saturation (SO₂).

Methods: : Both eyes of 60 healthy subjects from European, African, Hispanic and Asian descent underwent retinal oximetry (Oxymap, Reykjavik, Iceland) and color fundus photography (Carl Zeiss, Dublin, CA). The SO₂ values in retinal arterioles (SaO₂) and venules (SvO₂) were measured in four retinal quadrants: superotemporal, inferotemporal, superonasal, and inferonasal (the optic nerve was used as the center of the retina). The global SO₂ was defined as the average of the four quadrant measurements in either arterioles or venules. The correlation between SO₂ and age was assessed using simple linear regression. Multivariate regression analysis was performed to determine the association between global SO₂ and ethnicity, iris color, age, gender, eye laterality, refraction, history of smoking and arterial hypertension. Only one eye randomly selected per patient was included in the analysis.

Results: : There were 15 subjects from each ethnic group. The average age was 41.8±14.1 years. The average global SaO₂ was 93.4±3.0% and global SvO₂ was 64.3±4.6%. Global SO₂ did not differ between ethnic groups in either arterioles (p = 0.72) or venules (p = 0.38). Statistically significant differences in SaO₂ and SvO₂ were observed between quadrants (both p < 0.001). Linear regression analysis indicated that SO₂ decreases with increasing age, but this effect was significant only in subjects older than 50 years for SaO₂ (R² = 0.37, p = 0.004) and SvO₂ (R² = 0.58 , p < 0.001). In multivariate regression analysis, history of systemic hypertension was the only significant determinant of both global SaO₂ (β = -1.19, p = 0.027) and global SvO₂ (β = -2.37, p = 0.004) and quadrant SaO₂ (β = -1.30 to -1.97, all p <0.01) and quadrant SvO₂ (β = -2.27 to -3.34, all p < 0.01).

Conclusions: : As a first step in evaluating this emerging technology, we have established a preliminary, multi-ethnic normative database in retinal oximetry. We observed significant variation in SO₂ among retinal quadrants. Although older people had lower SO₂, history of systemic hypertension was the only significant independent predictor of SO₂. These factors should be taken into account when translating retinal oximetry measurements into a clinically meaningful interpretation. Future studies will build on these normal values from which to compare retinal oximetry measurements in disease states.

Keywords: retina • oxygen • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)