Maternal serum clusterin in established gestational diabetes mellitus: association with the need for insulin therapy...

Clusterin is a multifunctional glycoprotein involved in lipid transport, complement regulation, and oxidative/inflammatory stress responses, pathways relevant to insulin resistance and metabolic stress in gestational diabetes mellitus (GDM). However, its role in established GDM remains insufficiently defined. This study aimed to compare maternal serum clusterin in established GDM and normoglycemic controls and explore associations with the need for insulin therapy and perinatal outcomes.

This prospective single-center case-control study included 88 singleton pregnancies at 24-28 weeks' gestation (44 GDM, 44 controls). Fasting serum clusterin was measured by enzyme-linked immunosorbent assay after completion of GDM screening and, in women who subsequently required insulin, before insulin initiation.

Clusterin was lower in GDM than in controls (1.46 [0.76-2.81] vs 2.56 [1.31-3.98] ng/mL; p = 0.018). Within GDM, women who subsequently required insulin had lower clusterin than those managed with diet alone (1.14 [0.54-1.63] vs 2.26 [1.32-3.98] ng/mL; p = 0.005). In exploratory Firth penalized regression, higher clusterin remained inversely associated with the need for insulin therapy after adjustment for selected clinical and glycemic covariates (adjusted odds ratio per 1 ng/mL increase, 0.50; 95% CI 0.27-0.91; p = 0.024). Clusterin alone showed modest discrimination for GDM (AUC 0.646, 95% CI 0.537-0.745), whereas its discrimination for insulin therapy was higher but imprecise because of the small subgroup and wide confidence interval (AUC 0.748, 95% CI 0.595-0.867). Clusterin did not significantly discriminate the composite adverse perinatal outcome in either the overall cohort or the GDM subgroup.

Maternal serum clusterin was reduced in established GDM and was associated with the subsequent need for insulin therapy, suggesting a possible link with metabolic severity. These exploratory findings are hypothesis-generating and require validation before use in treatment stratification.