Biomarkers for Diabetes Mellitus in advanced Peripheral Arterial Disease. Diagnostic performance and outcome prediction of HbA1c, fasting plasma glucose and the oral glucose tolerance test

Abstract

Background: At present, the diagnosis of diabetes mellitus (DM) is based on the result from either of three tests; HbA1c, fasting plasma glucose (FPG) or 2-hour post glucose-load value following an oral glucose tolerance test (OGTT). Although DM is one of the major risk factors for peripheral arterial disease (PAD), undiagnosed DM is common in PAD patients. Also, epidemiological studies have shown that FPG, the OGTT and HbA1c largely classify different patients as having DM and pre-DM. Studies have reported a negative association of abdominal aortic aneurysms (AAA) with DM. However, inconsistent results regarding long-term survival in both AAA patient of normal and abnormal glucose status are reported. Aims and Methods: This was a prospective cohort study of patients with advanced vascular disease. The vascular pathologies were carotid stenosis, iliac occlusive disease (IOD), infrainguinal occlusive disease and AAA. In patients of unknown glycaemic status, the aim was to evaluate the performance of HbA1c as a diagnostic test for DM when using the OGTT as the gold standard (Paper I) and to explore the association of HbA1c levels, FPG values and OGTT results with long-term survival (Paper III). In Paper II, data from 66 AAA patients was used to investigate the prevalence of DM and pre-DM. Further, to evaluate the association of glycaemic status assessed by HbA1c, FPG and OGTT measurements with long-term survival in these patients. Results: Paper I The prevalence of newly diagnosed DM was 12% and that of pre-DM was 33% based on OGTT results. An HbA1c value of ≥ 48mmol/mol (6.5%) detected DM with 45.5% sensitivity and 90% specificity. The total prevalence of DM and pre-DM was higher based on HbA1c values than based on the OGTTs. AUC for the association of HbA1c with the OGTT and FPG as diagnostic parameters for DM was 0.73 (95% CI 0.63-0.84). Paper II The total prevalence of known and newly diagnosed DM in 66 AAA patients were 23% based on OGTT results. The prevalence of newly diagnosed DM in 58 AAA patients of unknown glycaemic status at baseline was 12% by OGTT results. HbA1c values classified DM patients according to OGTT results with 43% sensitivity and 90% specificity. The prevalence of pre-DM was 43% based on OGTT results and 72% based on HbA1c values. In 5.9 years follow-up, all-cause mortality in AAA patients of unknown glycaemic status was 43%. HbA1c was a significant independent predictor for mortality in the DM category, adjusted Hazard Ratio (HR) 6.35, (95% CI 1.49-27.1); p=0.01. Paper III Nine-years all-cause mortality in vascular surgery patients of unknown glycaemic status was 40%. HbA1c was an independent predictor for mortality, adjusted HR 1.54, (95% CI 1.03 – 2.32); p=0.04. The OGTT was not a predictor of long-term mortality. Conclusion: Vascular surgery patients have a high prevalence of undiagnosed DM and pre-DM. The OGTT and HbA1c largely identify different patient groups as having DM and pre-DM. The number of vascular surgery patients having DM and pre-DM is higher based on HbA1c values than based on OGTT measurements. In patients with advanced vascular disease and of unknown glycaemic status, HbA1c may be a useful marker to identify patients at high risk of a reduced survival rate following surgical treatment for PAD.