A cross-sectional study to characterise young-onset diabetes in UK ethnic groups

Abstract

Diabetes subtypes are challenging to differentiate, particularly in people who present at young age. Assigning the correct subtype of diabetes is critical as it informs treatment and misclassification can lead to erroneous management. The impact of ethnicity on diabetes phenotypes is poorly studied but recognition of diabetes subtypes within non-white ethnic groups, is increasingly problematic. Maturity onset diabetes of the young (MODY) is a monogenic form of diabetes that is frequently misdiagnosed in white populations and not widely studied in other ethnicities.

The work in this thesis aims to investigate the phenotypes of young-onset diabetes across white, south Asian and African-Caribbean ethnic groups, with the aim of improving recognition of these phenotypes.

Results from a retrospective analysis of referrals for MODY testing in the UK, showed a significantly lower detection rate of finding a positive MODY mutation in people referred from south Asian ethnicity compared to white European ethnicity. It was not clear if this significant disparity represented a lower prevalence of MODY in south Asian individuals or if there was a bias in testing referrals based on the traditional clinical criteria for MODY.

To fully investigate this finding, the MY DIABETES study was initiated; a multi-centre cross-sectional study that systematically phenotypes young-onset diabetes across south Asian, African-Caribbean and white European ethnic groups. Fasting C-peptide and pancreatic auto-antibodies are used to stratify individuals for genetic testing of all known MODY genes.

Results of the initial 728 participants recruited, showed an overall detection rate of 9.5% that was not significantly different by ethnicity and confirmed pathogenic MODY mutations were detected across all ethnic groups. These results suggest that a systematic approach to stratification in south Asian and African-Caribbean individuals may be warranted if clinical criteria are less sensitive due to the overlapping phenotypes of other types of diabetes in these groups.