Polygenic background contributes to GCK-MODY clinical presentation and glycaemic variability

GCK-MODY (glucokinase MODY) causes lifelong, mild hyperglycaemia with high penetrance. Variation in glycaemic phenotype among carriers remains unexplained. We hypothesised that polygenic background contributes to this variability and that this influence differs from that of HNF1A-MODY.

To test whether polygenic background contributes to the GCK-MODY clinical phenotype, we analysed polygenic risk scores (PGS) for nine diabetes-related traits in 897 clinically referred individuals with GCK-MODY. We compared these to 7645 non-diabetic control participants, 4773 participants with type 2 diabetes and 601 participants with HNF1A-MODY and assessed associations between PGS and glycaemic measures. Additionally, we evaluated 158 clinically unselected GCK variant carriers from the UK Biobank to examine polygenic effects independent of clinical referral.

We observed independent polygenic enrichment for HbA_1c (including both glycaemic and non-glycaemic components), fasting glucose and type 2 diabetes in clinically referred GCK-MODY individuals compared with non-diabetes controls (0.16-0.33 SD higher, all p<0.003), but not for type 1 diabetes. Importantly, HbA_1c and fasting glucose PGSs were higher than in both type 2 diabetes and HNF1A-MODY groups, whereas type 2 diabetes PGS was lower. GCK-MODY and HNF1A-MODY showed distinct patterns of polygenic enrichment, with only the type 2 diabetes PGS contributing independently in HNF1A-MODY (0.33 SD, p<1 × 10⁻²⁷). By contrast, no polygenic enrichment was seen in GCK pathogenic variant carriers from a clinically unselected population-based cohort. In both settings, HbA_1c PGS were associated with measured HbA_1c levels in GCK carriers (clinically referred: β=0.97, clinically unselected: β=0.91, both p<0.009), with effect sizes similar to those in non-carriers. GCK-MODY cases in the top HbA_1c quintile had a 3-to-6-fold risk of exceeding the diabetes diagnostic HbA_1c threshold (≥48 mmol/mol) in clinically selected and clinically unselected cohort respectively.

Our findings suggest that polygenic background and GCK variants interact to modify the glycaemic expression of GCK-MODY, influencing clinical diagnosis despite high penetrance. The pattern of polygenic contribution differs from that of HNF1A-MODY, highlighting the aetiology specific interaction. Our study highlights the importance of integrating both monogenic and polygenic factors to better understand phenotypic variability in monogenic diseases.