Diabetes intervention study designs require knowledge of within-subject variability of outcome measures. We evaluated the variability of measures of β-cell response assessed across 9-12 weeks in 3 nondiabetic adult populations relevant to type 1 diabetes (T1D) research: relatives of individuals with T1D who are single islet autoantibody (Ab) +, first degree relatives (FDR) with TID who are Ab -, and Ab- adults with no family history of T1D.
C-peptide and glucose levels were measured in repeated mixed meal tolerance tests (MMTT) and hyperglycemic clamps separated by means of 2.8 months and 3.0 months. The 12 participants (22-49 y; 50% male) included 3 single Ab + FDRs, 5 Ab-FDRs, and 4 with no family history of T1D. Fasting proinsulin:C-peptide (PI:C) ratios were calculated as molar ratios. For MMTTs and clamps, stimulated C-peptide measures were calculated. Between-visit coefficients of variation and intraclass correlations were calculated.
Test-retest variability differed between measures (Table). Test-retest correlations were highest for fasting PI:C and first-phase C-peptide response measures.
In nondiabetic adults with or without known T1D risk factors, test-retest variability ranged from 10-30% for measures of β-cell response. These observations will inform study design including choice of methodology and sample sizes needed to detect treatment effects using these methods.
A. Neyman: None. C. Evans-Molina: None. L. DiMeglio: Research Support; Self; Amgen Inc., Caladrius Biosciences, Inc., Janssen Research & Development, Medtronic, Sanofi. Other Relationship; Self; Dexcom, Inc. R. Mirmira: None. K.J. Mather: Advisory Panel; Self; Roche Diabetes Care. Research Support; Self; Abbott, Merck & Co., Inc., Novo Nordisk A/S, Sanofi. T.S. Hannon: Advisory Panel; Self; Eli Lilly and Company. E.K. Sims: None.