Lid-Parallel Conjunctival Folds and Their Ability to Predict Dry Eye

Heiko Pult, Stefan Bandlitz

Research output: Contribution to journalArticlepeer-review


PURPOSE: The observation of lid-parallel conjunctival folds (LIPCOF) is reported to be useful in dry eye diagnoses. Around 70% of the European clinicians use this test in a dry eye clinic. This study investigated the sensitivity and specificity of LIPCOF to predict dry eye.

METHODS: Ocular Surface Disease Index (OSDI) outcome, noninvasive break-up time (NIKBUT) using the Keratograph (Oculus, Wetzlar, Germany), and LIPCOF of 148 randomized selected patient reports (88 females; mean age=37.0±12.9 years) from 3 different sites were analyzed. Subjects were divided into dry eye and non-dry eye groups by OSDI only, named symptomatic dry eye (OSDI scores ≥15) and by a composite score (OSDI scores ≥15 and NIKBUTaverage of ≤9 sec).

RESULTS: Lid-parallel conjunctival folds (temporal, nasal, Sum [=temporal+nasal]) were significantly correlated to OSDI and to NIKBUTaverage (Spearman; r>-0.185, P<0.013). NIKBUTaverage (r=-0.322; P<0.001) was significantly correlated to OSDI scores. Based on the OSDI questionnaire, 37.2% of the subjects were symptomatic and 14.7% were positive for dry eye based on the composite score. Lid-parallel conjunctival folds were a significant discriminator between healthy and dry eye subjects (P≤0.001). The areas under the receiver operating characteristic curve for temporal LIPCOF, nasal LIPCOF, and LIPCOF Sum were 0.716, 0.737, and 0.771, respectively, for the symptomatic dry eye group and 0.771, 0.719, and 0.798, respectively, for the composite dry eye group.

CONCLUSIONS: As LIPCOF demonstrated high predictive ability of dry eye; it appears to be a promising test in the diagnoses of dry eye.

Original languageEnglish
Pages (from-to)S113-S119
JournalEye & contact lens
Publication statusPublished - 1 Nov 2018


Dive into the research topics of 'Lid-Parallel Conjunctival Folds and Their Ability to Predict Dry Eye'. Together they form a unique fingerprint.

Cite this