Diabetes distress mediates the relationship between depressive symptoms and glycaemic control among adults with type 2 diabetes: Findings from a multi-site diabetes peer support intervention

AimsDiabetes distress is positively associated with HbA(1c) and may mediate the relationship between depressive symptoms and HbA(1c). This study examined these relationships in a geographically, socioeconomically, and ethnically diverse sample of adults with type 2 diabetes. MethodsUsing data from five US sites evaluating peer support for diabetes management (n = 917), Structural Equation Modeling (SEM) examined whether diabetes distress (four items from Diabetes Distress Scale) mediated the relationship between depressive symptoms (PHQ-8) and HbA(1c). Sites compared interventions of varying content and duration with control conditions. Time from Baseline Assessment to Final Assessment varied from six to 18 months. Site characteristics were controlled by entering site as a covariate along with age, sex, education, diabetes duration, insulin use, and intervention/control assignment. ResultsDepressive symptoms, diabetes distress, and HbA(1c) were all intercorrelated cross-sectionally and from Baseline to Final Assessment (rs from 0.10 to 0.57; ps <0.05). In SEM analyses, diabetes distress at Final Assessment mediated the relationship between Baseline depressive symptoms and HbA(1c) at Final Assessment (indirect effect: b = 0.031, p < 0.001), controlling for Baseline HbA(1c) and covariates. Parallel analysis of whether depressive symptoms mediated the relationship between Baseline diabetes distress and HbA(1c) at Final Assessment was not significant. ConclusionsIn this diverse sample, diabetes distress mediated the influence of depressive symptoms on HbA(1c) but the reverse, depressive symptoms mediating the effect of distress, was not found. These findings add to the evidence that diabetes distress is a worthy intervention target to improve clinical status and quality of life among individuals with type 2 diabetes.