Abstract
Purpose :
Retinal nerve fibre Layer (RNFL) loss occurs in glaucoma and neurodegenerative diseases, but almost all studies are clinic-based and cross-sectional. The aim of this study was to identify determinants of RNFL loss in a population-based longitudinal twin cohort.
Methods :
Peripapillary global average RNFL thickness was measured in 4076 volunteers from the TwinsUK cohort using non-mydriatic SD-OCT scans (Optovue iVue 100, Optovue, Freemont, CA) between 2014 and 2018. 697 participants underwent repeat measurements in 2020-22. Autorefraction and intraocular pressure (IOP) were measured concurrently (Visionix). The right eye was used to calculate the rate of RNFL thinning, as an average change per year (CPY), with left eye data used when images were missing or poor quality (63 participants). Intraclass correlation coefficients were performed for monozygotic (MZ) and dizygotic (DZ) twin pairs for comparison. Multivariable regression analyses examined associations with age, IOP and spherical equivalent.
Results :
Data from 584 participants (mean age 63.1 SD 13.27, 90% female) were included, with an average time of 4.7 (±0.92) years between measurements. The average RNFL change per year (CPY) was -0.39 microns (95%CI -0.45 to -0.32). 201 twin pairs had complete data; 119 pairs were monozygotic (MZ) and 82 dizygotic (DZ). Twin-pair correlations were weak, and the MZ pairs were not more correlated than DZ pairs (r=0.14 c.f. r=0.19), in contrast to cross-sectional RNFL thickness where the correlations were stronger (MZ: r=0.74 DZ: r=0.35). While lower SphE was associated with thinner RNFL as expected, CPY was not (β=0.006, p=0.46). IOP was weakly associated (β=-0.03, p=0.06). After adjusting for sex, IOP and SphE, age was associated with RNFL loss (β=-0.007 p=0.006), with rate of change increasing per quartile of age (β=-2.5 p=0.03 for upper quartile).
Conclusions :
In an unselected non-glaucomatous population-based cohort, mean age 63, RNFL appears to thin at around 0.4 microns per year. In contrast to cross-sectional RNFL thickness, RNFL loss does not appear heritable, meaning that environmental and stochastic events are more important (as in many ageing traits). There was a significant association in RNFL thinning with age, a weak association with IOP (higher IOP results in greater RNFL loss) and none with myopic refractive error. Larger, long-term studies are needed to identify risk factors for RNFL thinning in pre-glaucomatous populations.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.