Endocrine Abstracts

Introduction: The impact of ethnicity and socioeconomic status (SES) on glycaemic control in children with type 1 diabetes (T1D) is poorly understood in England and Wales.

Methods: We studied 18 478 children and young people with T1D aged <19 years attending diabetes clinics in England and Wales and included in the 2012–2013 National Paediatric Diabetes Audit (NPDA). Self-identified ethnicity was categorized as white, Asian, black, mixed, other and ‘not stated’ (those that chose not to divulge ethnicity). A small area measure of SES was estimated using postcode and the Index of Multiple Deprivation (grouped into quintiles). Multivariable linear regression was used to assess impact of ethnicity and SES on glycaemic control (mean HbA_1c levels) accounting for age, gender and diabetes duration. Associations between SES and HbA1c were also tested in models stratified by ethnicity. The impact of insulin pump use on the ethnicity/SES-HbA1c associations was tested in 13 962 children.

Results: All ethnic minorities had higher mean HbA_1c compared to white children, with the largest differences observed in black and mixed ethnicity children (7.84 mmol/mol, 95% CI 5.07–10.6 and 6.81 mmol/mol, 4.55–9.08 respectively). Lower SES was associated with higher mean HbA_1c with a dose effect. The lowest SES group (quintile 5) had on average 6.78 mmol/mol (5.63–7.95) higher mean HbA1_c compared to the highest SES group, adjusted for ethnicity. Estimates for ethnicity were attenuated but remained significant on adjustment for SES. Having a lower SES was associated with higher mean HbA_1c irrespective of ethnicity in stratified analyses. However, being in the lowest SES group and of Asian (6.90 mmol/mol, 2.52–11.28), mixed (11.26 mmol/mol, 6.31–16.21) and other (8.85 mmol/mol, −0.05–17.5) ethnicity was associated with higher mean HbA_1c compared to being in the corresponding lowest SES group and white ethnicity (6.03 mmol/mol, 4.72–7.34). An interaction test between ethnicity and SES was statistically significant (P=0.005). Less non-white (white 20.3 vs Asian 12.1 vs black 5.5%) and deprived (least deprived 21.1 vs most deprived 13.2%) children were on insulin pump therapy. Both ethnicity and SES remained significant predictors of HbA_1c after accounting for insulin pump use.

Conclusion: The effect of ethnicity independent to deprivation on glycaemic control persists after adjustment for pump use, indicating that an alternative approach to intensive insulin therapy for the treatment of glycaemic control is required in these vulnerable children.