Background Children and young people with type-1 diabetes (T1D) have excess all-cause hospital admissions, particularly younger children with lower socioeconomic status. Education on managing alcohol consumption is given to teenagers with T1D in paediatric diabetes services, but little is known about alcohol-related harm. We compare the risk of alcohol-related hospital admission (ARHA) in children with T1D over 18.5 years with that of the general population for the same birth years.

Methods We extracted data for 1,794,559 individuals born between 1979 and 2014 with a GP registration in Wales and record-linked these to wholly attributable ARHA between 1998 and June 2016 within the Secure Anonymised Information Linkage Databank (SAIL). Diabetes status was assessed by record-linking to a national register (Brecon Cohort), containing 3,577 children diagnosed since 1995 with T1D before the age of 15 years. Linking to the Welsh Demographic Service dataset provided information on age, sex and the lower super output areas (LSOAs) of residence, including moves. To each LSOA we linked the Welsh Index of Multiple Deprivation 2008 quintiles. We censored for death or leaving Wales. We estimated hazard ratios (HRs) with 95% confidence intervals (95% CIs) for the risk of (multiple) ARHA for sex, age and deprivation quintiles (both time-varying) using recurrent-event models. We also included interaction terms between age group, and separately deprivation fifth, and diabetes status.

Results There were 37,930 (multiple) admissions and 19.1 million person-years of follow up. Individuals with T1D had 252 admissions (up to 4 admissions each), and overall had an 80% higher risk of ARHA (HR 1.8; 95% CI 1.60 to 1.99) compared to those without, having adjusted for age group, sex and deprivation fifth. In diabetic individuals the risk of ARHA was highest aged 14–17 years (HR 6.03; 95% CI 4.70 to 7.75), six times higher than the reference group of those without T1D aged 11–13. In the general population the highest risk was in those aged 18–22 (HR 2.23, 95% CI 2.14 to 2.32) compared to the same reference group. The deprivation gradient in those with T1D was less pronounced than in the comparison population.

Conclusion Young people with T1D have increased risks of ARHA, highest at school age (14–17 years) and earlier than the peak at student age (18–22 years) in the comparison population. Interventions aiming to reduce alcohol-related harm in T1D are needed, particularly in teenagers, and structured education on managing alcohol consumption may need revision.