Association of the glucose metabolism continuum (fasting plasma glucose/HbA1c) with tear-film stability and secretion: ocular surface evidence across non-diabetes, prediabetes, and diabetes

Background: Whether gradations in glycemia across non-diabetes, prediabetes, and diabetes relate to tear-film homeostasis and dry eye disease (DED) remains uncertain, and explicitgraded association evidence is limited. Methods: = 100 per stratum). Same-day hemoglobin A1c (HbA1c) and fasting plasma glucose (FPG) classified strata and were also modeled continuously. Co-primary outcomes were non-invasive tear break-up time (NIBUT) and Schirmer I (no anesthesia). Secondary outcomes included fluorescein tear break-up time (TBUT), ocular surface disease index (OSDI), National Eye Institute (NEI) staining, meiboscore, tear meniscus height, bulbar redness, tear osmolarity, and matrix metalloproteinase-9 (MMP-9). DED was defined as OSDI ≥13 plus ≥1 sign [NIBUT <10 s, osmolarity ≥308 mOsm/L or intereye difference ≥8 mOsm/L, or NEI staining score ≥2 (beyond trace)]. Multivariable analysis of covariance (ANCOVA) and logistic regression adjusted for prespecified demographic, clinical, and environmental covariates. Restricted cubic splines tested non-linearity, and Benjamini-Hochberg false discovery rate (FDR) controlled multiplicity for secondary endpoints. Results: = 0.001). Osmolarity abnormality and MMP-9 positivity rose across strata and with HbA1c, remaining significant after FDR control. Using FPG instead of HbA1c yielded concordant effects. Conclusions: Higher glycemic status, including prediabetes, was associated with shorter NIBUT, lower Schirmer, and higher odds of DED and inflammatory signs, although the absolute between-stratum differences were modest in magnitude. Longitudinal studies are needed to assess the temporality and the practical relevance of these cross-sectional associations.