Individuals with Diabetes Mellitus Have a Dry Eye Phenotype Driven by Low Symptom Burden and Anatomic Abnormalities

Dry eye disease is an umbrella term that includes a variety of symptoms and signs. A link between diabetes mellitus and dry eye disease exists, but the associated phenotype needs further examination. Thus, our aim was to determine how diabetes mellitus relates to the dry eye disease phenotype. A prospective, cross-sectional study was conducted at the Miami Veteran Affairs Medical Center ophthalmology clinic between October 2013 and September 2019. Participants included a volunteer sample of 366 South Florida veterans with one or more symptoms or signs of dry eye disease [Dry Eye Questionnaire-5 ≥ 6 OR tear break-up time ≤ 5 OR Schirmer's test score ≤ 5 OR corneal fluorescein staining ≥ 2]. Participants were divided into three groups: (1) individuals without diabetes mellitus (controls); (2) individuals with diabetes mellitus but without end-organ complications; and (3) individuals with diabetes mellitus and end-organ complications. Dry eye metrics were compared across groups. The main outcome measures included ocular symptom questionnaires [e.g., 5-item Dry Eye Questionnaire, Ocular Surface Disease Index, and ocular pain assessment] and clinical parameters obtained from an ocular surface evaluation. A total of 366 individuals were included (mean age 59 ± 6 years; 89% males; 39% White; 11% diabetes mellitus and end-organ complications; 15% diabetes mellitus but without end-organ complications). Individuals with diabetes mellitus and end-organ complications had lower symptom scores on the dry eye disease and pain-specific questionnaires compared to individuals with diabetes mellitus but without end-organ complications and controls (Ocular Surface Disease Index: 42.1 ± 24.5 vs. 38.9 ± 25.1 vs. 23.6 ± 16.2; p < 0.001; numerical rating scale of ocular pain intensity: 4.9 ± 3.2 vs. 4.3 ± 2.7 vs. 3.5 ± 2.7; p = 0.02). Eyelid laxity was also more severe in the group with diabetes mellitus and end-organ complications (0.69 ± 0.64 vs. 0.73 ± 0.72 vs. 1.08 ± 0.77; p = 0.004) compared to the two other groups. The diabetic dry eye disease phenotype is driven by signs more so than by symptoms, with anatomic eyelid abnormalities being more frequent in individuals with diabetes mellitus and end-organ complications. Given this, ocular surface abnormalities in individuals with DM may be missed if screened by symptoms alone. As such, individuals with DM should undergo a slit lamp examination for signs of ocular surface disease, including anatomic abnormalities.