Objectives/Goals: The objective of this study was to identify subgroups of young adults with type 1 diabetes (T1D) defined by HbA1c measured 12 months after participating in a randomized controlled trial for a behavioral intervention. Methods/Study Population: This study is a secondary, exploratory analysis of 76 young adults with T1D with complete data who enrolled in the DiaBetter Together RCT. A regression tree model is used to identify subgroups of participants according to HbA1c measured 12-month post-enrollment. Predictors include baseline participant characteristics, questionnaire scores, and clinical characteristics. Model fit is informed by restricted sum of squares, pruned using cost complexity assessment, and number of observations per node. Results/Anticipated Results: The final regression tree predicting 12m HbA1c included two branches and five leaves defined by baseline HbA1c, Diabetes Distress: Friend/Family (DDFF), and Diabetes Strengths: Getting help with diabetes management (DSGMH). Participants with baseline HbA1c 9.91, two unique subgroups were defined by DDFF. The mean 12m HbA1c was 9.99 (SD=1.5) if DDFF <3.55. The highest 12m HbA1c was observed among participants with baseline HbA1c ≥9.91% and DDFF≥3.55 (M=13.0, SD=0.7). Discussion/Significance of Impact: Baseline HbA1c, DDFF, and DSGMH were important predictors of subgroups of patients defined by 12m HbA1c. Psychosocial variables were important predictors beyond demographic and clinical variables. Clinicians and researchers may wish to target interpersonal aspects of diabetes distress and strengths in work with young adults with T1D.
Minard et al. (Wed,) studied this question.