Investigation of the relationship between HbA1c and tumor markers in type 2 diabetes mellitus

Abstract

Aim: Tumor markers are frequently measured in patients with type 2 diabetes mellitus (T2DM), yet their interpretation is challenging because hyperglycemia may influence marker levels. This study aimed to examine the relationship between hemoglobin A1c (HbA1c) and commonly used tumor markers in T2DM.

Methods: In this retrospective cross-sectional study, 557 T2DM patients were included. HbA1c and tumor markers—carcinoembryonic antigen (CEA), alpha-fetoprotein (AFP), prostate-specific antigen (PSA), carbohydrate antigen 19-9 (CA19-9), cancer antigen 15-3 (CA15-3), cancer antigen 72-4 (CA72-4), and cancer antigen 125 (CA125)—were measured concurrently. Patients were compared by HbA1c <7% vs ≥7%. Spearman correlations with 95% confidence intervals and Holm–Bonferroni-adjusted p values were calculated. Multivariable linear regression models for log-transformed markers were adjusted for age, sex, smoking, and estimated glomerular filtration rate.

Results: The cohort (33.4% female; mean age 68.3±12.7 years) had a median HbA1c of 6.96%. Median CEA and CA72-4 levels were slightly higher in the HbA1c ≥7% group. HbA1c showed a weak positive correlation with CEA, while correlations with CA19-9 and CA72-4 were very weak and lost significance after correction. In adjusted regression analyses, HbA1c was not an independent predictor of AFP, PSA, CA19-9, CA15-3, or CA125. Its association with CEA was borderline and explained little variance; CA72-4 results were unstable due to limited data.

Conclusions: In T2DM, HbA1c demonstrates only a weak association with CEA and no independent relationship with other tumor markers. Therefore, elevated tumor markers in T2DM should not be attributed solely to poor glycemic control, and appropriate diagnostic evaluation should be maintained.