Routine Haematological Parameters Associated with HbA1c and Estimated Whole-Blood Viscosity in Diabetes Management: An Exploratory AIC-Based Regression Analysis.

Background

Routine full blood count (FBC) testing is part of the haematological workup in diabetes management. There is limited information regarding the contributions of individual haematological parameters to regression models for glycated haemoglobin (HbA1c), estimated whole-blood viscosity (eWBV) and the resulting blood viscosity complications. Importantly, because association and prediction represent distinct concepts, this study extends previous work with a focus on comparative and exploratory relationships. The objective was to compare FBC parameters between higher and lower HbA1c and eWBV groups and identify variables contributing to the Akaike Information Criterion (AIC)-based regression model among diabetics.

Methods

This laboratory-based mixed quantitative study involved cross-sectional and regression analyses. Fifteen parameters were evaluated, including the following: red blood cell count (RBC) and indices (MCV, MCH, MCHC); platelet count and derived ratios (PRR, PWR, RPR); and white blood cell count (WBC) with lymphocyte ratios (MLR, NLR, PLR). HbA1c and eWBV data were used to create dichotomous subgroups for univariate comparison, followed by exploratory AIC-based model identification of variables.

Results

HbA1c, RDW, MCV, and RPR, differed significantly between HbA1c groups (p < 0.1). Regression analysis identified RDW, MCV, RPR, MCH and RBC as contributors to the HbA1c model. For eWBV, five out of seven parameters (HCT, HB, RBC, WBC, and MLR) showed a significant association.

Conclusions

These findings highlight haematological parameters with potential values for future predictive model development. Overall, the study supports the usefulness of selected FBC variables as adjuncts in diabetes monitoring with potential utility in understanding glycaemia control and blood viscosity-related complications.