Abstract
Purpose :
Previous studies have shown that corneal sensitivity and tear production is reduced in diabetes, and that meibomian gland dysfunction is associated with diabetes. This observational clinical study aimed to determine if there was a reduction in symptoms which led to an underdiagnosis of diabetic dry eye disease, as well as determine similarities and differences between the diabetic and the non-diabetic dry eye disease.
Methods :
This was a multicenter study with data collection from 7 different sites. All subjects were asked to complete the Ocular Surface Disease Index (OSDI), and Pain Sensitivity Questionnaire (PSQ), lissamine green staining graded on a 0-4 scale across 6 areas of the conjunctiva and 5 corneal zones, meibography, meibomian gland expression, Non-Invasive Breakup Time (NIBUT) (Oculus Keratograph 5M) and Hemoglobin A1c (HbA1c)(A1cNow PTS Diagnostics).
Results :
81 patients were enrolled if they were either previously identified as diabetic or pre-diabetic (DM), had dry eye disease (DED), or both (DMDED). In the DM cohort (n=37), 19 were found to have dry eye based on DEWS II criteria. (OSDI ≥13 and at least 9 spots of conjunctival staining) an undiagnosed rate of 51.3% of diabetic patients. In the DMDED cohort, there was significantly less self reported redness compared to DED cohort (DMDED mean (µ)= 0.82 Standard Deviation(σ)=0.90, DED µ=1.33 σ=0.91; p=0.05). DMDED cohort also had significantly worse conjunctival staining (DMDED µ=13.42 σ=10.13, DED µ=8.33 =6.04; p=0.03). There was an increased directional trend with self-reported itching in compared to the DED cohort (DMDED µ=1.65 =1.29, DED µ=1.05 σ=1.07; p=.07). In the DMDED cohort, higher HbA1c was moderately associated with OSDI (r=0.52, p=0.0077). There was no statistical difference between DM, DED, and DMDED patients with respect to pain sensitivity, meibomian gland structure, or NIBUT.
Conclusions :
Although there is a strong overlap in signs and symptoms of diabetes associated with DED and DMDED patients, there are differences in staining and self-reported redness. Decreased redness may be related to vascular changes seen in diabetes and may merit further study. Our study also showed a large percentage of dry eye in patients with diabetes goes undiagnosed based on DEWS II criteria. The addition of a questionnaire for dry eye diagnosis may be useful in regular clinical practice for patients with diabetes.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.