Retinal Thickness Is Not Associated with Renal Function in Diabetes

Introduction: Spectral domain optical coherence tomography (SD-OCT) has made possible in vivo investigation of the segmented neural retinal layers of the eye. Following previous reported associations between retinal segmentation and choroidal thickness with renal function in a healthy population, we assessed associations in a cohort with multiple co-morbidities related to diabetes, kidney and cardiovascular disease.

Methods: Cross-sectional data was collected from adults attending nuclear cardiology and renal medicine clinics. All participants were imaged with SD-OCT (Heidelberg Spectralis) using Enhanced Depth Imaging. Measurements included macular, retinal nerve fibre layer and choroidal thickness, and macular volume. Blood and urinary measurements were taken, including estimated glomerular filtration rate (eGFR), urinary protein and plasma C-reactive protein (CRP). Linear regression was used to assess associations with adjustment for age, mean arterial blood pressure (MABP), gender, CRP, BMI, LDL, and retinopathy status.

Results: Of 241 participants, 93 (39%) had diabetes mellitus, a mean age of 65 years (standard deviation [SD] 9), MABP of 99 mmHg (SD: 13) and mean eGFR of 67 ml/min/1.73m² (SD: 25). Diabetes was associated with lower retinal thickness across the macula and reduced choroidal volume (B = -0.07, 95% confidence interval (CI) -0.14, -0.01, p = 0.04) although failed to remain significant in the fully adjusted model. In individuals with diabetes, a thinner retina was associated with lower eGFR although this also did not remain significant following adjustment for confounders.

Conclusion: In this cohort of high disease burden, diabetes was associated with lower retinal thickness and choroidal volume but did not remain significant following adjustment for confounders. Similarly, lower retinal thickness was associated with reduced eGFR in individuals with diabetes, but these associations also did not remain significant following adjustment for associated confounders.