Evaluating the agreement of continuous glucose monitoring system with venous methods for glycemic index determination

Continuous glucose monitoring systems (CGMS) are widely used in diabetes care, but their reliability for standardized food glycemic index (GI) measurement remains unclear. Most previous studies used capillary blood rather than venous glucose as the reference.

This study aimed to verify the consistency between the CGMS and standardized venous sampling methods for measuring postprandial incremental area under the curve (IAUC) and GI values.

This study was conducted in accordance with the Chinese GI testing standard WS/T 652-2019. Three measurement methods were compared: venous sampling (Method A), CGMS with all data points adopted (Method B), and CGMS data aligned with the standard venous time points (Method C). One-way analysis of variance (ANOVA) and Lin's concordance correlation coefficient (CCC) were used for statistical analysis. The Parkes Error Grid analysis was employed to ascertain the clinical acceptability of glucose discrepancies within low, medium, and high concentration strata.

The 12 enrolled participants (six males and six females) had a mean age of 29.50 ± 10.12 years, mean BMI of 21.5 ± 1.7 kg/m², mean HbA1c of 5.07 ± 0.21%, and all lipid, glycemic, hepatic and renal function indicators were within normal reference ranges. For 120-min IAUC, the mean values were 186.93 ± 87.29 (Method A), 166.00 ± 76.46 (Method B) and 164.24 ± 77.48 (Method C). ANOVA showed no significant inter-method difference (F = 1.178, p = 0.311). CCC analysis with Method A as the reference revealed good agreement: CCC = 0.8818 (95% CI: 0.8068-0.9289) for A vs. B and CCC = 0.8727 (95% CI: 0.7921-0.9234) for A vs. C, with a high Pearson's ρ (>0.91) and bias correction factors (Cb, >0.95 fot both). For GI values (n = 12 for each method), the mean values were 55.07 ± 15.57 (Method A), 54.28 ± 12.24 (Method B) and 53.68 ± 15.04 (Method C). ANOVA showed no significant inter-method difference (F = 0.028, p = 0.972), and CCC analysis indicated moderate agreement (CCC ≈ 0.79) with minimal systematic bias (Cb > 0.97).

CGMS methods demonstrate good-to-moderate concordance with standardized venous testing for IAUC and GI determination. CGMS may represent a feasible, non-invasive alternative for food GI evaluation in nutritional and diabetes-related research among healthy individuals.