Persistent C-peptide secretion is associated with favourable CGM metrics in adults with type 1 diabetes

Abstract Aims/hypothesis Residual endogenous insulin secretion, reflected by measurable C-peptide, has been linked to improved glycaemic management in type 1 diabetes. We aimed to assess the relationship between random plasma C-peptide levels and continuous glucose monitoring (CGM) metrics in a large, real-world cohort of adults with type 1 diabetes. Methods We conducted a cross-sectional analysis of adults with type 1 diabetes attending a single UK centre. Inclusion criteria were diabetes duration >1 year, random plasma glucose >4 mmol/l at C-peptide sampling and ≥70% data completeness on the Freestyle Libre 2 CGM device within the corresponding month. Associations between C-peptide categories (<50 pmol/l to >400 pmol/l) and CGM/HbA 1c outcomes were assessed using non-parametric tests and multivariable logistic regression. Results In total, 945 adults with type 1 diabetes were included with a median age of 45 years (33–57) and median diabetes duration of 18 years (7–29). Of these, 54% were male and median HbA 1c was 63 mmol/mol (54–73) (7.9% [7.1–8.8]). Higher C-peptide levels were associated with shorter diabetes duration (OR 0.87 per year, p <0.001), older age at diagnosis (OR 1.04 per year, p <0.001) and male sex (OR 1.44, p =0.042). Higher C-peptide was significantly associated with favourable CGM metrics, including lower time below range (2% [1–5] in those with C-peptide <50 pmol/l vs 1% [0–3] in those with C-peptide 101–200 pmol/l), lower glucose variability (glucose CV 37.5% [34.2–42.0] in those with C-peptide <50 pmol/l vs 32.5% [29.0–36.6] in those with C-peptide 101–200 pmol/l), higher time in range (45.0% [32.0–61.0] in those with C-peptide <50 pmol/l vs 55.0% [37.0–66.5] in those with C-peptide 50–100 pmol/l) and favourable hyperglycaemia measures (time above 13.9 mmol/l 20.0% [9.0–36.0] in those with C-peptide <50 pmol/l vs 10.0% [5.5–31.5] in those with C-peptide 50–100 pmol/l) ( p <0.05 for all pairwise comparisons). C-peptide ≥100 pmol/l was independently associated with meeting time below range <4% (OR 5.4, p <0.001), and C-peptide ≥100 pmol/l was also associated with achieving HbA 1c <53 mmol/mol (7%) (OR 1.8, p =0.043. No significant glycaemic differences were seen between individuals with C-peptide <10 pmol/l and 10–49 pmol/l. Conclusions/interpretation Random C-peptide measurement in routine care identifies adults with type 1 diabetes who are more likely to achieve CGM and HbA 1c targets. Differences in glycaemic metrics are clinically meaningful at thresholds ≥100 pmol/l. These findings support efforts to preserve residual beta cell function and highlight the potential value of C-peptide in individualising therapy. Graphical Abstract