A comparative analysis of HbA1c, glycated albumin, and fasting plasma glucose for glycemic assessment in people living with HIV.

Background: People living with HIV (PLWH) are at increased risk for metabolic disorders, including diabetes and prediabetes. While hemoglobin A1c (HbA1c) is widely used for glycemic assessment, its reliability in PLWH is questioned due to altered red blood cell turnover. Glycated albumin (GA) has been proposed as an alternative, but its diagnostic utility remains unclear in PLWH. This study aims to compare the correlations of HbA1c and GA with fasting plasma glucose (FPG), evaluate their diagnostic performance, and identify factors influencing discrepancies between them in PLWH.

Methods: This retrospective cross-sectional study included 236 PLWH with documented FPG, HbA1c, and GA levels. Correlations between glycemic markers were assessed using Pearson's correlation coefficients. Diagnostic performance for prediabetes and diabetes was evaluated using receiver operating characteristic (ROC) curves, and a GA cut-off was determined using the Youden index. Multivariable logistic regression was performed to identify predictors of HbA1c-GA mismatch.

Results: HbA1c showed a moderate correlation with FPG (r = 0.33, p value < 0.001), while GA had a weaker correlation (r = 0.18, p value = 0.005). The area under the ROC curve (AUC) for detecting glycemic abnormalities was 0.66 for HbA1c and 0.57 for GA. The optimal GA cut-off for prediabetes derived from ROC analysis was 12.42 %, improving sensitivity but reducing specificity. Multivariable analysis identified low mean corpuscular volume (MCV <80 fL) as an independent predictor of HbA1c-GA mismatch (odds ratio = 4.94, 95 % confidence interval: 1.95-12.50, pvalue < 0.001).

Conclusion: HbA1c or GA alone do not reliably capture glycemic abnormalities in PLWH. A lower GA cut-off (12.42 %) for prediabetes improves sensitivity but remains suboptimal. A combined approach incorporating FPG is recommended to enhance prediabetes and diabetes screening accuracy in this population.