Fig. 4: T2D diagnoses are associated with the accumulation of DCSI secondary complications in those with pathogenic and glucose-elevating GCK variants at a rate similar to that of those with T2D in the population at large.

a Cumulative incidence plot with shaded confidence intervals depicting the proportion of individuals with substantial DCSI outcomes (DCSI > = 3) by age for the UKB cohort, grouped by individuals with pathogenic GCK variants and T2D (green line) (HR = 2.4, p = 0.005, n = 67) and glucose-elevating GCK variants with T2D (pink line) (HR = 2.8, p = 0.002, n = 46). Proportions for those with T2D and without a GCK variant are shown in blue. Hazard ratios are calculated using a Cox Proportional Hazard model, with those without glucose-elevating variants and without T2D (yellow line) as a reference group. b) Cumulative incidence plot with shaded confidence intervals depicting the proportion of individuals with substantial DCSI outcomes (DCSI > = 3) by age for the Helix cohort grouped by individuals with pathogenic GCK variants and T2D (green line) (HR = 7.3, p = 0.00007, n = 11). The glucose-elevating group is not plotted due to the low sample size (HR = 1.5, p = 0.6, n = 5). Proportions for those with T2D and without a GCK variant are shown in blue. Hazard ratios are calculated using a Cox Proportional Hazard model with those without glucose-elevating variants and without T2D (yellow line) as a reference group. c) Hazard Ratio (HR as dots; +/− 95%CI as bars) to develop a DCSI score of at least 3 for each cohort, subsetted by GCK genotype and T2D diagnoses, compared to population controls without T2D and Hazard Ratio (HR as dots; +/− 95%CI as bars) for at least 2 in each DCSI secondary complication area, compared to population controls without T2D in either cohort. Hazard ratios are generated from the Cox Proportional Hazard model controlling for the most recent age, sex, BMI and cohort. For display purposes, 3 categories in which there were no cases are filled in with HR of 0.1. As fewer than 5 cases were expected in these groups based on population frequencies of each phenotype and 0 cases were seen, the hazard ratios are in the range of 0.000006–0.00001 and the CIs are too large to plot well. DCSI outcomes are similar regardless of GCK status for the complication categories of peripheral vascular, cardiovascular, and neuropathy and appear depleted in those with GCK variants for ophthalmic and cerebrovascular, and are variable by GCK variant group for nephropathy (too few in the metabolic category to analyze).