Factors explaining variation in self-esteem among persons with type 1 diabetes and elevated HbA1c

Objectives To investigate associations between perceived autonomy support from health-care professionals, autonomy-driven motivation, diabetes self-perceived competence and self-esteem in adults (age 18–55 yrs) with suboptimally regulated type 1 diabetes mellitus (T1DM) with at least one HbA1c≥8.0% (≥64 mmol/mol) during the past year, and whether these factors could predict decrease in self-esteem over time. Methods A cross-sectional population-based survey was performed, and 9 months follow-up data were collected. Data collection comprised clinical and socio-demographic variables, blood sampling (HbA1c) and self-report questionnaires; the Health Care Climate Questionnaire (HCCQ), Treatment Self-Regulation Questionnaire (TSRQ), the Perceived Competence in Diabetes Scale (PCDS), and the Rosenberg Self-esteem Scale (RSES). We fitted block-wise linear regression models to assess associations between RSES and variables of interest (HCCQ, TSRQ, PCDS, HbA1c, clinical and socio-demographic variables) and linear regression models to assess predictors of change over time. Findings In this study sample, aged 36.7 (±10.7) mean HbA1c 9.3% (±1.1), 31.5% had long-term complications and 42.7% had experienced severe hypoglycemia within the previous 12 months. In the final regression model the association between PCDS and RSES was strongly significant (B = 1.99, p<0.001) and the associations between HCCQ, TSRQ and RSES were reduced to non-significance. All predictor variables combined explained 42% of the variability of RSES (adjusted R2 = 0.423) with PCDS contributing 18% to explained variance (R-square change = 0.184, p<0.001). The strongest predictors of change in RSES over time were long-term complications (B = 2.76, p