Omega‐3 supplementation is neuroprotective to corneal nerves in dry eye disease: a pilot study

Purpose To investigate whether oral, long‐chain omega‐3 (ω‐3) essential fatty acid (EFA) supplementation, for 3 months, induces changes to the central corneal sub‐basal nerve plexus in dry eye disease and whether nerve alterations correlate with clinical findings. Methods This prospective, comparative study involved the final 12 participants enrolled in a randomised, double‐masked, placebo‐controlled clinical trial of 60 participants with moderate dry eye disease. Participants received either placebo (olive oil 1500 mg/day; n = 4) or ω‐3 EFA supplements (~1000 mg/day eicosapentaenoic acid + ~500 mg/day docosahexaenoic acid; n = 8) for 90 days. The main outcome measure was the mean change in central corneal sub‐basal plexus nerve parameters between days one and 90, quantified using in vivo confocal microscopy. Secondary outcomes included mean change in tear osmolarity, corneal dendritic cell density and basal epithelial cell density. Results Compared with baseline, the reduction in OSDI score and tear osmolarity at day 90 were greater in the ω‐3 EFA group than the placebo group (OSDI: ω‐3 EFA, mean ± SEM: −15.6 ± 2.8 vs placebo: −2.8 ± 4.1 units, t5 = 2.6, p = 0.04; tearosmolarity: ω‐3 EFA: −22.63 ± 5.7 vs placebo: −8 ± 2.7 mOsmol/L, t9 = 2.3, p = 0.04). At day 90, corneal total nerve branch density (CTBD: 91.1 ± 8.6 vs 45.1 ± 13.4 branches/mm2, F1,10 = 14, p = 0.004) and corneal nerve branch density on the main fibre (CNBD: 63.4 ± 6.5 vs 27.9 ± 11.5 branches/mm2, F1,10 = 6, p = 0.03) were higher in the ω‐3 EFA group compared with placebo. Relative to day 1, CNBD (branches/mm2) increased at day 90 in the ω‐3 EFA group (+20.0 ± 9.2, t8 = 3.2 p = 0.01) compared with placebo (−10.8 ± 3.2). Similar changes were evident for corneal nerve fibre length (CNFL, mm/mm2), which increased from baseline at day 90 in the omega‐3 EFA group (+2.9 ± 1.6, t8 = 3.4 p = 0.01) compared with placebo (−2.7 ± 0.5). There was a negative correlation between CTBD and tear osmolarity (r10 = −0.70, p = 0.01). No significant changes were observed for basal epithelial cell or corneal dendritic cell density. Conclusion These pilot study findings suggest that ω‐3 EFA supplementation imparts neuroprotective effects in the corneal sub‐basal plexus that correlate with the extent of tear osmolarity normalisation.