OBJECTIVE: We assessed the function of rod/cones and melanopsin in type 1 (T1DM) and type 2 diabetes mellitus (T2DM) with and without non-proliferative diabetic retinopathy (NPDR).

METHODS: We performed pupillometry on 22 healthy controls and four diabetic groups: 12 T1DM patients without NPDR and 12 with moderate NPDR, and 16 T2DM patients without NPDR and 12 with moderate NPDR. Monocular stimulations of 20 seconds with red (λ = 633 nm) and blue light (λ = 463 nm) at ~15 log quanta/cm2 /second were performed. The primary outcome was the melanopsin-mediated late redilation phase of postillumination pupillary light response (PIPRLate ) to blue light. The secondary outcomes were the mixed rod/cone and melanopsin responses, that is maximal pupil constriction and the early redilation phase of PIPR (PIPREarly ).

RESULTS: Late redilation phase of PIPR (PIPRLate ) to blue and red light stimuli was not significantly different between healthy control and the four diabetic groups (n.s.). The maximal pupil contractions to blue light stimulus were significantly reduced in T1DM patients as well as in T2DM patients with NPDR (p ≤ 0.02), whereas for red light stimuli, the maximal pupil constriction was only reduced in T2DM with NPDR (p < 0.01). Early redilation phase of PIPR (PIPREarly ) to blue and red light stimuli was not significantly different between healthy controls and diabetic patients (n.s.).

CONCLUSION: Neither the PIPREarly nor the PIPRLate was significantly reduced in diabetics with or without NPDR compared to healthy controls. The reduced maximal pupil constrictions in diabetics with NPDR indicate decreased mixed rod/cone and melanopsin responses.