HbA1c tracking and bio-psychosocial determinants of glycaemic control in children and adolescents with type 1 diabetes: retrospective cohort study and multilevel analysis

Abstract

To explore the association between HbA1c 6 months after diagnosis (6 m-HbA1c) and long-term glycaemic control in children with type 1 diabetes, accounting for other bio-psychosocial determinants. This retrospective cohort study, included 155 children (≤16 yr) from the North of Scotland, diagnosed between January 1993 and August 2011, and receiving care between November 2008 and August 2012. Multilevel analysis explored the relationships between 6 m-HbA1c, other persistent or dynamic variables, and HbA1c. Patterns of glycaemic control were identified by cluster-analysis. 6 m-HbA1c was positively associated with diabetic ketoacidosis at diagnosis, shorter duration of partial-remission, female gender, and psychosocial adversity. In multilevel analysis the effects of 6 m-HbA1c on subsequent HbA1c trajectories remained significant after adjusting for patient- and observation-level predictors. An increase in 6 m-HbA1c of 10 mmol/mol (0.9%) was associated with an average increase in HbA1c levels of 5.3 (95% CI: 4.5-6.2) mmol/mol, or 0.48% (0.41 to 0.57%; p < 0.001) over the follow-up period. Coefficients for linear and quadratic growth identified sustained effects of 6 m-HbA1c on glycaemic control (p < 0.001). Higher average levels or accelerated increases in HbA1c were associated with age at diagnosis, falling BMI (in girls > boys), mental health diagnosis, major adverse life-events, single-parenting, child welfare concerns, neighbourhood deprivation, and clinic non-attendance. Cluster-analysis identified groups with poor or deteriorating control, characterized by older age at diagnosis, multiple psychosocial adversities, and maladaptive healthcare use. Early HbA1c predicted future glycaemic control across childhood. Trajectories were further modified by biological factors, exposures to psychosocial adversity, and healthcare use.