A new approach for investigating the relative contribution of basal glucose and postprandial glucose to HbA1c

The study addresses how basal glucose (BG) and postprandial glucose (PPG) contribute to HbA1c and overall hyperglycaemia, a key issue for optimizing glycaemic control and reducing diabetes-related complications. Prior work links both fasting and postprandial glycaemia to cardiovascular risk, but the relative contributions to HbA1c vary across studies due to methodological differences. Building on Monnier’s concept of partitioning fasting and postprandial contributions to HbA1c, the authors hypothesize that using a physiologically derived baseline from a normal glucose tolerance (NGT) 24-hour glucose curve will more accurately quantify the relative contributions of BG and PPG than using fixed fasting thresholds (5.6 or 6.1 mmol/L). The purpose is to develop and validate a more accurate evaluation method using continuous glucose monitoring (CGM) across NGT, IGT, newly diagnosed T2DM (NDDM), and treated T2DM populations.

The paper reviews 17 studies (up to Jan 1, 2021) that estimated relative contributions of BG and PPG to HbA1c using diverse populations, treatments, and measurement methods (fingerstick profiles vs CGM), with differing baseline definitions: 11 used 5.5/5.6 mmol/L (ADA upper normal fasting glucose), 4 used 6.1 mmol/L (WHO upper normal), and some only reported correlations without partitioning. One prior study used the NGT 24-h glucose profile as baseline and suggested 6.1 mmol/L can overestimate PPG contribution by ~10–20% relative to the NGT curve. Findings across studies generally indicate higher PPG contribution at lower HbA1c and increasing BG contribution as HbA1c rises, but the magnitude varies widely, likely due to methodological differences (measurement modality, baseline choice, populations, treatments).

Design: Prospective, real-world study (ClinicalTrials.gov NCT02648685) using CGM to quantify relative contributions of BG and PPG to HbA1c across different glucose tolerance categories and with three baseline criteria. Population: Adults aged 18–75 years recruited from communities in Chengdu (Chenghua, Xindu, Longquanyi districts) and West China Hospital, Sichuan University. Groups: NGT (n=92), IGT (n=36), newly diagnosed T2DM, drug-naïve (NDDM; n=131), and drug-treated T2DM (n=231); total n=490 (age range 21–77 years). Inclusion criteria for NGT included absence of obesity, hyperlipidaemia, diabetes, or hypertension; BMI 18.5–24.9 kg/m2; normal BP (90–139/60–89 mmHg) and lipid profiles (TG <2.22 mmol/L; TC <6.22 mmol/L). Drug-treated T2DM met WHO criteria and had ≥6 months diagnosis with stable oral therapy ≥3 months. Exclusions: Significant hepatic/renal dysfunction, conditions causing severe glucose fluctuations (e.g., malignancy, HIV, acute infection, trauma, surgery), and treatments specifically affecting PPG (insulin, DPP-4 inhibitors, GLP-1 receptor agonists, acarbose) to avoid bias. Assessments: Medical history, anthropometrics, OGTT. Laboratory measurements included plasma glucose (glucose oxidase), HbA1c (HPLC; Bio-Rad Variant II), lipids, liver and renal function (Roche analyzer). CGM captured 24-hour glucose profiles over multiple days. Diet: Individualized mixed-meal plans per China Medical Nutritional Therapy Guideline for Diabetes (2013), with macronutrient ratios: carbohydrates 45–60%, protein 15–20%, fat 25–35% (≤30% in overweight/obese). Standardized nutritional counseling was provided. Computation of relative contributions: Using CGM-derived 24-h glucose curves, the relative contribution of PPG to overall hyperglycaemia was computed under three baseline definitions: Method A (6.1 mmol/L), Method B (5.6 mmol/L), and Method C (the 24-h NGT glucose profile). Area-under-the-curve (AUC)-based partitioning was used to separate PPG increments from basal hyperglycaemia, with statistical comparisons across methods and groups. Statistical analyses included Wilcoxon Signed Ranks tests; significance set at p<0.05. Baseline characteristics: Mean HbA1c across all subjects 6.96 ± 1.26% (range 4.5–12.7%). From IGT to NDDM, age, BMI, TG, TC, and LDL-C generally increased (p<0.05).

- Relative PPG contributions differed significantly across the three baseline criteria in all participants (p<0.01). - Using Methods A (6.1 mmol/L), B (5.6 mmol/L), and C (NGT 24-h curve), mean relative PPG contributions to HbA1c were: • IGT: 80.85%, 72.36%, 67.68% • NDDM: 48.76%, 39.63%, 37.85% • T2DM: 39.46%, 33.22%, 31.47% • NDDM + T2DM: 42.82%, 35.54%, 33.78% All pairwise method comparisons were significant (mostly p<0.001; see Table 3). - Trend: PPG contribution decreased progressively from IGT to NDDM to T2DM, while BG contribution increased correspondingly, regardless of baseline criterion. - Overestimation relative to Method C (NGT baseline): • In IGT, using 6.1 mmol/L overestimated PPG contribution by ~13.16%. • In T2DM, overestimation was 9.04% with 6.1 mmol/L and 1.76% with 5.6 mmol/L (p<0.01). - Clinical characteristics: Across groups, HbA1c and cardiometabolic risk factors differed as expected; DBP showed no significant differences; several lipid parameters varied across groups with statistical significance. - Overall, using a fixed fasting threshold (especially 6.1 mmol/L) tends to overestimate PPG’s contribution compared with the physiologic NGT 24-h curve baseline.

Findings confirm a progressive shift in the relative contributions to overall hyperglycaemia: PPG predominates at lower HbA1c (e.g., IGT), whereas BG contributes more as HbA1c worsens (NDDM and T2DM). Methodological differences in past literature (CGM vs fingerstick profiles, baseline definitions, populations, treatments) likely account for conflicting estimates of BG vs PPG contributions. This study demonstrates that a physiologically derived baseline—the 24-h NGT glucose curve—more accurately accounts for normal postprandial fluctuations, avoiding bias introduced by fixed thresholds. Using 6.1 mmol/L notably overestimates PPG’s contribution compared with the NGT curve, with smaller bias using 5.6 mmol/L. The CGM-based approach captures full diurnal variability, improving partitioning accuracy versus limited-point SMBG. Clinically, recognizing how contributions vary with HbA1c supports tailored therapy: targeting PPG in prediabetes/low HbA1c states and prioritizing BG control as HbA1c rises, while considering patient factors (e.g., triglycerides, waist circumference) that may influence fasting hyperglycaemia via insulin resistance.

A progressive shift occurs in the relative contributions of BG and PPG to overall hyperglycaemia with increasing HbA1c, with BG becoming predominant at higher HbA1c levels. The 24-h NGT glucose curve is the most accurate baseline for assessing BG and PPG contributions because it incorporates physiological postprandial fluctuations. When the NGT curve is unavailable, 5.6 mmol/L is an acceptable alternative, whereas 6.1 mmol/L tends to overestimate PPG’s contribution. These insights inform individualized treatment strategies: emphasize PPG control in prediabetes/lower HbA1c, and prioritize basal hyperglycaemia control as HbA1c increases.

- The analysis primarily compared different baseline criteria across glucose tolerance groups and did not stratify results by detailed HbA1c categories. - Single-center design may limit generalizability to broader Chinese or international populations. - Patients on insulin, GLP-1 RAs, DPP-4 inhibitors, or more complex regimens and those with diabetes-related complications were excluded, limiting applicability to such populations. - Further prospective studies are needed to evaluate impacts of diverse oral and injectable regimens on BG and PPG contributions across HbA1c ranges.