Author: Abraham Olvera-Barrios (United Kingdom)
Co-authors: Yuka Kihara, Philipp Mueller, Katie Williams, Aaron Lee, Catherine Egan, Adnan Tufail
Purpose
To examine sociodemographic associations, and ocular and early life determinants of optical coherence tomography (OCT) derived foveal curvature (FC) in a large cohort.
Setting/Venue
The UK Biobank is a national research resource with the aim of improving the prevention, diagnosis and treatment of a wide range of diseases. The study recruited more than 500,000 people aged 40 to 69 between 2006 to 2010 from across the United Kingdom. A subset of this population received enhanced ophthalmic assessment with macular OCT.
Methods
Cross-sectional study. We included UK Biobank participants, who as part of the enhanced ophthalmic examination, had macular spectral domain OCT. We developed a deep leaning model to extract the internal limiting membrane (ILM) boundaries, detect the foveal center, and fitted a 2-D polynomial curve to the central foveal slice OCT B-Scan. OCT-derived FC perfectly agreed with manual ordering of FC tertiles by a retina specialist. Multilevel linear regression models examined FC associations with sociodemographic, ocular and early life factors. The model reported adjusts for age, gender, ethnicity, height, visual acuity, refraction (spherical equivalent [SE]), corneal astigmatism, intraocular pressure, macular curvature, center point foveal thickness (CFT), fluid intelligence, higher education, deprivation (Townsend index), annual income, maternal age, birth weight, birth order, and breastfed as a baby category to examine associations with FC. A random effect for person was included in the models to allow for the right- and left-eye data from the same person to contribute to the analysis.
Results
A total of 122,384 eyes of 72,229 participants (46% male) were included for the analysis. Mean age was 56.5 ± standard deviation (SD) 8 years. Mean fovea curvature was 0.073 ± 0.020 SD. Males had on average steeper foveal curvature (0.078; 95% CI 0.077 - 0.078) than females (mean 0.068; 95% CI 0.068 – 0.069). Whites showed the steepest foveal curvature (mean 0.073; 95% CI 0.0731 – 0.0734), followed by Mixed (mean 0.068; 95% CI 0.067 – 0.069) and Chinese (mean 0.067; 95% CI 0.065 – 0.068) ethnic groups. Individuals with the lowest annual average income (<18K Great British Pounds [GBP]) had flatter fovea curvatures (mean 0.071, 95% CI 0.0719-0.0714) than individuals with the average annual income of 31-59.9K GBP (mean 0.073, 95% CI 0.0726-0.0731). Multilevel linear regression showed that older age (β= 0.04 per decade), male gender (β= 0.51), increasing height (β= 0.02 per 5cm), and steeper macular curvature (β= 0.26 per 0.01) were associated with steeper foveal curvature. Worse visual acuity (β= -0.01 per 0.1 logMAR), higher spherical equivalent (β= -0.04 per diopter), increasing corneal astigmatism (β= -0.06 per diopter), and thicker CFT (β= -0.01 per μm) showed an inverse association with foveal curvature.
Conlusions
There are ethnic differences in FC with whites having steeper foveas. Ocular factors and low-income show associations with FC. Further work is needed to incorporate our findings to foveal development and genetics in order to achieve a better understanding of this highly important visual region.
Financial Disclosure
None
Comments
-
