Purpose: To determine whether the retinal nerve fiber layer thickness can be used as an indicator for systemic neurodegeneration in diabetes.
Methods: We used existing data from 38 adults with type 1 diabetes and established polyneuropathy. Retinal nerve fiber layer thickness values of four scanned quadrants (superior, inferior, temporal, and nasal) and the central foveal thickness were extracted directly from optical coherence tomography. Nerve conduction velocities were recorded using standardized neurophysiologic testing of the tibial and peroneal motor nerves and the radial and median sensory nerves, 24-hour electrocardiographic recordings were used to retrieve time- and frequency-derived measures of heart rate variability, and a pain catastrophizing scale was used to assess cognitive distortion.
Results: When adjusted for hemoglobin A1c, the regional thickness of the retinal nerve fiber layers was (1) positively associated with peripheral nerve conduction velocities of the sensory and motor nerves (all P < 0.036), (2) negatively associated with time and frequency domains of heart rate variability (all P < 0.033), and (3) negatively associated to catastrophic thinking (all P < 0.038).
Conclusions: Thickness of the retinal nerve fiber layer was a robust indicator for clinically meaningful measures of peripheral and autonomic neuropathy and even for cognitive comorbidity.
Translational relevance: The findings indicate that the thickness of the retinal nerve fiber layer should be studied in adolescents and people with prediabetes to determine whether it is useful to predict the presence and severity of systemic neurodegeneration.