Comparative Utility of Tear Film and Ocular Surface Parameters in Differentiating Sjögren's Disease Dry Eye From Non-Sjögren's Dry Eye.

PURPOSE: Tear production and ocular surface damage are equally weighted in the 2016 classification criteria for Sjögren's disease (SjD). We compared Schirmer testing versus ocular surface staining to distinguish SjD from non-SjD dry eye.

METHODS

Patients with SjD dry eye (n = 20), non-SjD dry eye (n = 47), and controls (n = 29) were evaluated in this cross-sectional study. Schirmer testing assessed tear production; ocular surface staining was graded using the Sjögren's International Collaborative Clinical Alliance (SICCA) and National Eye Institute (NEI) systems. Analyses included age-stratified comparisons, receiver operating characteristic (ROC) analyses, and multivariable regression.

RESULTS

Overall, mean Schirmer values did not differ between SjD and non-SjD dry eye but were lower in SjD among participants <60 years (8.78 ± 6.24 vs. 18.00 ± 10.04 mm, P = 0.018), not ≥60 years (15.27 ± 11.01 vs. 14.00 ± 8.28 mm, P = 0.699). Schirmer ≤5 mm threshold differentiated SjD only in the younger group (<60 years: 55.6% vs. 5.0%, P = 0.014; ≥60 years: 18.2% vs. 25.9%, P = 0.145), and Schirmer testing showed poor overall discrimination, particularly among older participants (area under the curve [AUC] = 0.602; <60 years 0.772; ≥60 years 0.453). By contrast, SICCA corneal staining demonstrated consistently stronger discrimination (overall AUC 0.803; <60 years 0.872; ≥60 years 0.754). In multivariable models, NEI conjunctival staining (odds ratio [OR] = 1.23, P = 0.008) and SICCA corneal staining (OR = 1.83, P = 0.046), but not Schirmer testing (OR = 1.01, P = 0.888), were associated with SjD.

CONCLUSIONS

Ocular surface staining was more strongly associated with classified SjD status in this cohort, whereas Schirmer testing showed age-dependent performance.

TRANSLATIONAL RELEVANCE: Ocular surface staining may be a more reliable marker of Sjögren's dry eye than Schirmer testing, with implications for clinical evaluation.