Adaptive optics imaging of parafoveal cones in type 1 diabetes.

PURPOSE

To evaluate the parafoveal cone density in patients with Type 1 diabetes mellitus (DM1).

METHODS

Adaptive optics retinal images of the photoreceptor mosaic were acquired from 11 DM1 patients (study group) and 11 age-matched healthy subjects (control group). Cone density was analyzed, along the horizontal and vertical meridian, at 230-µm, 350-µm, and 460-µm eccentricity from the fovea. Central retinal thickness was measured using a Spectralis spectral-domain optical coherence tomography. A multiple regression model was performed to determine the relationships between the explanatory variables (age, glycohemoglobin level, presence of diabetic retinopathy, duration of diabetes, and central retinal thickness) and cone density.

RESULTS

Patients had a diagnosis of DM1 in the past 9 years to 21 years. Of these, five patients had a diagnosis of no diabetic retinopathy and six had mild nonproliferative diabetic retinopathy. On average, cone density was 10% lower in the study than in the control group at each retinal eccentricity along the horizontal and vertical meridians (analysis of variance, P < 0.001). The central retinal thickness was thicker in DM1 eyes than in the control eyes (278 ± 20 µm and 260 ± 13 µm; P < 0.05). The model explained 61% (P < 0.01) of the variance of cone density in the population, with the variables representing an abnormal glucose metabolism, that is, a higher glycohemoglobin level, the presence of diabetic retinopathy, and a chronic diabetes, having the highest influence on cone density decline.

CONCLUSION

A subtle decrease of parafoveal cone density was found in DM1 patients in comparison with age-matched control subjects via high-resolution adaptive optics retinal imaging. The cone density decline was moderately associated with a disturbance in the glucose metabolism.