Baseline β-Cell Secretory Reserve and Its Association with Glycaemic Control and Long-Term Outcomes Across Diabetes Phenotypes

Residual β-cell secretory function plays a central role in diabetes pathophysiology; however, long-term comparative data describing β-cell trajectories from diagnosis across diabetes phenotypes remain limited. In this prospective observational study, 393 adults with newly diagnosed diabetes underwent assessment of fasting and glucagon-stimulated C-peptide at diagnosis. The cohort included individuals with type 1 diabetes mellitus (T1DM), encompassing both classical adult-onset autoimmune diabetes and latent autoimmune diabetes in adults (LADA), as well as individuals with type 2 diabetes mellitus (T2DM). A subgroup of 89 participants underwent follow-up visit after a mean of seven years. Glucagon stimulation testing was not repeated at follow-up in patients with T1DM and LADA for clinical and safety reasons; therefore, longitudinal analyses in these groups are based on fasting C-peptide measurements. At diagnosis, fasting and glucagon-stimulated C-peptide concentrations differed markedly between phenotypes (median fasting C-peptide: 0.87 ng/mL in T1DM, 1.53 ng/mL in LADA, and 2.64 ng/mL in T2DM; stimulated C-peptide: 1.35, 1.86, and 4.60 ng/mL, respectively; all <i>p</i> < 0.001). During follow-up, patients with T1DM exhibited a pronounced decline in fasting C-peptide (from 0.95 to 0.10 ng/mL), whereas individuals with T2DM showed preserved or increased stimulated responses (from 4.37 to 5.46 ng/mL). Participants with LADA displayed intermediate baseline values and a gradual decline in fasting C-peptide (from 1.41 to 0.31 ng/mL). Higher baseline C-peptide concentrations were associated with more favourable long-term metabolic profiles, including lower insulin resistance and better glycaemic control. These findings demonstrate that the early dynamic assessment of β-cell reserve using the glucagon stimulation test complements fasting C-peptide by revealing biologically meaningful heterogeneity in disease trajectories, thereby refining phenotypic classification and prognostic stratification at the time of diabetes diagnosis.