Abstract
Purpose :
To characterize Bayesian adaptive estimation of the contrast sensitivity function (BCSF) as a novel assessment in diabetic (DM) and nondiabetic (nonDM) individuals across a wide range of diabetic retinopathy (DR) and diabetic macular edema (DME) severity.
Methods :
DM and nonDM subjects underwent undilated binocular BCSF testing by a trained technician using a Sentio system with NEC P463 professional-grade flat panel which displays contrast levels from 0.2%-100% and stimulus sizes from 1-27 cycles/degree (cpd) at 3m. Area under the letter/contrast size curve (AULCSF) and Contrast Sensitivity Function (CSF) Acuity were estimated using 25 trials and a broad spectrum of spatial frequencies (1, 1.5, 3, 6, 12 and 18 cpd). Optos 200° fundus photos and spectral domain OCT were performed at the same visit and graded for DR and DME severity by graders masked to BCSF outcomes.
Results :
Of 111 participants, 61 had type 1 or 2 DM and 50 were age matched without DM. BCSF test-retest repeatability was excellent in both DM and nonDM. AULCSF decreased significantly with age and DM, but was not related to gender, DM type, DM duration or A1c. Compared to DM participants without DME, those with DME were more likely to have reduced AULCSF (1.08 vs 1.46, p <0.0001), CSF Acuity (1.22 vs 1.36, p=0.003), and CS across all spatial frequencies (p<0.0001-0.01). BCSF parameters were incrementally reduced by DME in no, 1 or both eyes: AULCSF (1.46 vs 1.26 vs 0.87, p <0.001), CSF Acuity (1.36 vs 1.30 vs 1.12, p=0.0002) and CS across all spatial frequencies (p=<0.0001-0.006). In multivariable models adjusting for age and visual acuity in the worse or better-seeing eye, all binocular CSF parameters remained significantly associated with DME status.
Conclusions :
Current CS methods such as the Pelli Robson Chart, test limited spatial frequencies and do not correlate well with diabetes or diabetic retinal disease. The BCSF method provides precise CSF estimation over a wide range of spatial frequencies, resulting in a two-dimensional contour defining the lowest contrast distinguished at each spatial frequency. These data suggest that BCSF may be a sensitive method to detect DME-induced visual function changes independent of age and visual acuity. Further study is warranted to determine if BCSF can be used as an early functional marker of diabetic retinopathy or response to therapy.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.