New insights from continuous glucose monitoring into the route to diabetes

AimType 2 diabetes mellitus (T2DM) is preceded by a period of impaired glucoregulation. We investigated if continuous glucose monitoring system (CGMS) (1) could improve our capacity to predict the development of T2DM in subjects at risk. (2) Find out if impaired fasting glucose/impaired glucose tolerance differentiation through CGMS would also elucidate differences in clinical phenotypes. Material and methodsObservational study of 209 hypertensive patients, aged 18 to 85years who wore at entry a CGMS. Two CGMS metrics, percent of time under the 100mg/dL glycaemic threshold (TU100) (impaired fasting glucose surrogate phenotype) and area above the 140 mg/dL glycemic threshold (AO140) (impaired glucose tolerance surrogate phenotype) were measured. The median follow-up was 32months (6-72mo), and there were 17 new cases of T2DM. ResultsIn a multivariate Cox proportional hazard survival analysis including the conventional prediabetes-defining criteria and the 2 CGMS-derived variables, only TU100 and HbA(1c) were significant and independent variables in predicting T2DM development. An increase in 0.1 in TU100 resulted in a 0.69 (95% CI, 0.54-0.88; P<.01) odds ratio of developing T2DM. With cut-off points of 0.5 for TU100 and 5.7% for HbA(1c), the test TU<0.5 and HbA(1c)>5.7% had a sensitivity of 0.81 (SD, 0.10), a specificity of 0.83 (SD, 0.03), and a likelihood ratio of 4.82 (SD, 1.03) for T2DM development. ConclusionsContinuous glucose monitoring system allows for a better T2DM risk-development categorization than fasting glucose and HbA(1c) in a high-risk population. Continuous glucose monitoring system-derived phenotyping reveals clinical differences, not disclosed by conventional fasting plasma glucose/HbA(1c) categorization. These differences may correlate with distinct pathophysiological mechanisms. In this observational study of 209 hypertensive patients aged 18-85 years, we investigated if continuous glucose monitoring system (CGMS) could improve our capacity to predict the development of T2DM in patients at risk. Two CGMS metrics, percent of time under the 100-mg/dL glycaemic threshold (TU100) (impaired fasting glucose surrogate phenotype) and area above the 140-mg/dL glycaemic threshold (AO140) (impaired glucose tolerance surrogate phenotype), were measured. glycaemic threshold (AO140) (IGT surrogate-phenotype) were measured. The median follow-up was 32 months (6-72 months) and there were 17 new cases of T2DM. In a multivariate Cox proportional hazard survival analysis only TU100 and HbA1c were significant and independent variables in predicting T2DM development.