Aims/hypothesis: The genetic architecture of model-derived parameters of beta cell function has never been assessed. Therefore, we estimated heritability (h(2)) for model-derived phenotypes of insulin secretion in twins.
Methods: Thirty-three monozygotic (MZ) and 23 dizygotic (DZ) twin pairs from the Finnish Twin Cohort Study underwent an OGTT (plasma glucose/C-peptide at 0, 30, 60, 90 and 120 min). A subset of the twin pairs (21 MZ/20 DZ) also underwent an IVGTT (frequent sampling of plasma glucose/insulin from 0 to 60 min) followed by a 160-min euglycaemic-hyperinsulinaemic clamp (45 mU.min(-1).m(-2)). Mathematical modelling was applied to the IVGTT and the OGTT to assess first-phase (readily releasable insulin [RRI]) and second-phase (sigma) secretion (IVGTT), and a global index of beta cell performance (OGTT beta index). Intraclass correlation coefficients and genetic and non-genetic components for trait variances were computed to assess the h(2) of model-derived parameters.
Results: The intraclass correlation coefficients in MZ twins were 0.78 for RRI, 0.67 for sigma and 0.57 for OGTT beta index. In DZ twins the correlation coefficients were 0.23, 0.32 and 0.42, respectively. Using the most parsimonious model for each trait, the h(2)--the proportion of variance accounted for by genetic factors--was 76% (95% CI: 53-88%) for RRI, 28% (34-80%) for sigma and 53% (26-72%) for OGTT beta index.
Conclusions/interpretation: Our findings demonstrate that model-derived parameters of insulin secretion have a substantial genetic component and may be used in the search for genetic determinants of beta cell function in humans.