Predictive models of post-prandial glucose response in persons with prediabetes and early onset type 2 diabetes: A pilot study.

Abstract

Objective: Post-prandial glucose response (PPGR) is a risk factor for cardiovascular disease. Meal carbohydrate content is an important predictor of PPGR, but dietary interventions to mitigate PPGR are not always successful. A personalized approach, considering behaviour and habitual pattern of glucose excursions assessed by continuous glucose monitor (CGM), may be more effective.

Research design and methods: Data were collected under free-living conditions, over 2 weeks, in older adults (age 60 ± 7, BMI 33.0 ± 6.6 kg/m²), with prediabetes (n = 35) or early onset type 2 diabetes (n = 3), together with sleep and physical activity by actigraphy. We assessed the predictive value of habitual CGM glucose excursions and fasting glucose on PPGR after a research meal (hereafter MEAL-PPGR) and during an oral glucose tolerance test (hereafter OGTT-PPGR).

Results: Mean amplitude of glucose excursions (MAGE) and fasting glucose were highly predictive of all measures of OGTT-PPGR (AUC, peak, delta, mean glucose and glucose at 120 min; R² between 0.616 and 0.786). Measures of insulin sensitivity and β-cell function (Matsuda index, HOMA-B and HOMA-IR) strengthened the prediction of fasting glucose and MAGE (R² range 0.651 to 0.832). Similarly, MAGE and premeal glucose were also strong predictors of MEAL-PPGR (R² range 0.546 to 0.722). Meal carbohydrates strengthened the prediction of 3 h AUC (R² increase from 0.723 to 0.761). Neither anthropometrics, age nor habitual sleep and physical activity added to the prediction models significantly.

Conclusion: These data support a CGM-guided personalized nutrition and medicine approach to control PPGR in older individuals with prediabetes and diet and/or metformin-treated type 2 diabetes.