Prior studies examining beta‐cell preservation in type 1 diabetes have predominantly assessed stimulated C‐peptide concentrations approximately 10 wk after diagnosis. We examined whether earlier assessments might aid in prediction of beta cell function over time.

Using data from a multi‐center randomized trial assessing the effect of intensive diabetes management initiated within 1 wk of diagnosis, we assessed which clinical factors predicted 90‐min mixed‐meal tolerance test (MMTT) stimulated C‐peptide values obtained 2 and 6 wk after diagnosis. We also studied associations of these factors with C‐peptide values at 1‐ and 2‐year post‐diagnosis. Data from intervention and control groups were pooled.

Among 67 study participants (mean age 13.3 ± 5.7 yr, range 7.8–45.7 yr) in multivariable analyses, C‐peptide increased from baseline to 2 wks and then 6 wk. C‐peptide levels at these times were significantly correlated with 1‐ and 2‐yr C‐peptide concentrations (all p < 0.001), with the strongest observed associations between 6‐wk C‐peptide and the 1‐ and 2‐yr values (r = 0.66 and r = 0.61, respectively). In multivariable analyses, greater baseline and 6‐wk C‐peptide, and older age independently predicted greater 1‐ and 2‐yr C‐peptide concentrations.

C‐peptide assessments close to diagnosis were predictive of subsequent C‐peptide production. Our data demonstrate a clear increase in C‐peptide over the initial 6 wk after diabetes diagnosis followed by a plateau. Our data do not suggest that MMTT assessments performed closer to diagnosis than 6 wk would improve prediction of subsequent residual beta cell function.