Meibomian Gland Dropout, not Distortion, Can Distinguish Dry Eyes from Normal Eyes in Contact Lens Wearers

Abstract

ABSTRACT Purpose To examine the effectiveness of using meibomian gland (MG) dropout and distortion for dry eye disease (DED) diagnosis in a young population with long-term contact lens (CL) wear. Methods Seventy-three CL-wearers and 68 non-CL-wearers (20–28 years old) participated in the study. The diagnosis of DED was based on an ocular surface disease index ≥ 13, and a noninvasive breakup time < 10 seconds or positive corneal fluorescein staining. The MG dropout and distortion in each subject were quantified with non-invasive meibography. Receiver operating characteristic (ROC) curves were created to evaluate the effectiveness of using MG dropout and distortion to determine the presence of DED in CL-wearers and non-CL-wearers. Results In non-CL-wearers, the MG distortion was significantly higher in DED subjects than in normal subjects (3.83 ± 1.81 versus 1.59 ± 1.04). In CL-wearers, MG distortion was not significantly different (5.12 ± 4.11 versus 5.51 ± 3.15 in normal and DED subjects respectively). The area under the ROC curve (AUC) for MG distortion was 0.783 in the non-CL-wearers, but 0.507 in CL-wearers. In non-CL-wearers, MG dropout was significantly higher in DED subjects than in normal subjects (0.22 ± 0.08 versus 0.15 ± 0.06). In CL-wearers, MG dropout was also significantly higher in DED subjects than in normal subjects (0.29 ± 0.11 versus 0.22 ± 0.08). The AUC for MG dropout used to classify DED was 0.740 in the non-CL-wearers and 0.715 in CL-wearers. Conclusions MG dropout was effective in distinguishing subjects with DED from normal subjects in both CL wearers and non-CL wearers. MG distortion can only distinguish DED subjects from normal ones in non-CL wearers, but not in CL wearers.