Electroconductivity of Saliva as a Diagnostic Tool for Dry Eye Disease: A Case-Control Study

Abstract

Purpose: To evaluate the diagnostic potential of body fluid electroconductivity for dry eye disease (DED) and compare its accuracy with commonly used DED tests.

Methods: Individuals with dry eye (n = 36) and controls (n = 26) were enrolled in this case-control prospective study. Electroconductivity measurements were performed on blood, serum, tears, urine, and saliva. Dry eye assessments included tear film breakup time (TBUT) and Schirmer test (Schirmer), with symptoms evaluated using the Ocular Surface Disease Index (OSDI). Blood and urine analyses were performed to assess the baseline systemic profiles of both groups.

Results: Among all body fluids, saliva (saliva electroconductivity [ESaliva]) showed the most significant differences in electroconductivity between controls and dry eye individuals (2514.02 ± 329.18 vs. 3262.00 ± 992.47 µS/cm, P < 0.001). ESaliva showed robust diagnostic performance (area under the curve [AUC] = 0.800), comparable to TBUT (AUC = 0.693, P = 0.103) and superior to Schirmer (AUC = 0.536, P < 0.001). OSDI showed a moderate correlation with ESaliva (r = 0.43, P < 0.001), representing the strongest association, followed by TBUT (r = -0.26, P = 0.004) and Schirmer (r = -0.09, P = 0.313). Cross-validation procedure identified ESaliva cutoffs of 2373 µS/cm (95% confidence interval [CI], 2340-2456) for low-to-moderate and 2880 µS/cm (95% CI, 2845-2931) for moderate-to-high DED risk. Net reclassification improvement and integrated discrimination improvement analyses confirmed ESaliva's superior predictive ability. A single cutoff of 2880 µS/cm yielded 64% sensitivity and 89% specificity for DED prediction.

Conclusions: ESaliva effectively distinguishes patients with DED and exhibits superior diagnostic performance.

Translational relevance: ESaliva: offers a noninvasive and self-assessable tool for DED diagnosis.

Figure 1.

ROC curves and AUC for dry eye measures, shown for (A) continuous variables and (B) binary outcomes based on cutoff points. The binary outcomes include TBUT <5 and <10 seconds, Schirmer test <5 mm, OSDI ≥13, and ESaliva >2880 µS/cm.

Figure 2.

Density plot showing the bimodal distribution of ESaliva in control and dry eye groups. Cross-validated cutoff points of 2373 µS/cm (dotted red line) and 2880 µS/cm (dashed red line) stratify individuals into low-, moderate-, and high-risk categories. Controls are predominantly in the low-risk category, while those with dry eye are mainly in the high-risk category. The moderate-risk category shows overlap between groups.

Figure 3.

Bar graphs showing the clinical performance of dry eye measures using single cutoff points: TBUT <5 and <10 seconds, Schirmer test <5 mm, OSDI ≥13, and ESaliva >2880 µS/cm. Bars represent the proportion of true-negative, false-negative, false-positive, and true-positive results (scaled to 100 participants). Rightward shifts indicate stricter thresholds for dry eye diagnosis, and leftward shifts indicate thresholds favoring control classification.