Investigating predictors of well-being in type 2 diabetes mellitus patients: the role of undiagnosed depression.

Abstract

Type 2 diabetes mellitus (T2DM) is a common metabolic disorder with various medical and psychological adverse effects. Well-being in patients with T2DM is often compromised. The aim of the present study was to investigate clinicodemographic predictors of well-being in patients with T2DM with no known psychiatric history and explore the mediatory role of undiagnosed anxiety and depression. We recruited 175 outpatients with T2DM (54.3% males, aged 34-79 (mean 59.9) years) followed-up at the Diabetes Center of the General Hospital of Nikaia-Peiraeus in Athens. Patients included had no severe diabetes-related complications or known psychiatric history. Well-being was measured with the Mental Health Continuum Short-Form (MHC-SF), a novel 14-item tool measuring the emotional (EWB), social (SWB) and psychological (PWB) dimensions of well-being, as well as a total score of well-being (WBT). Hospital Anxiety and Depression Scale (HADS) was used for screening for undiagnosed anxiety (HADS-A) and depression (HADS-D). Patients' demographics, Body Mass Index (BMI), glycemic control (HbA1c), T2DM duration, comorbid hypertension or dyslipidemia and type of antidiabetic medication were investigated as predictors of well-being or its dimensions in stepwise linear regression models, also including or excluding HADS-A and HADS-D. Mediational effects of HADS-A and HADS-D were explored in structural equation models through path analyses. Results showed that 21.1% of participants had comorbid depression (HADS-D≥11) and 5.1% comorbid anxiety disorder (HADS-A≥11). In the models without HADS, higher WBT as well as EWB and PWB were significantly predicted by lower HbA1c (all p=0.001) and lower BMI (p=0.015, 0.019 and 0.030, respectively). After being included in the model, HADS-A and HADS-D significantly predicted WBT and every dimension of well-being, but the effects of HbA1c and BMI were no longer statistically significant. In path analyses, the indirect effects of HbA1c and BMI on well-being via HADS-D were statistically significant, while the direct and indirect effects via HADS-A were not. Therefore, the effects of HbA1c and BMI on EWB, PWB and WBT were completely mediated by HADS-D. Concludingly, this is the first study using MHC-SF to measure well-being in patients with T2DM. High levels of undiagnosed depression were recorded, in agreement with other studies. Depression was predicted by HbA1c and BMI and finally predicted well-being. Undiagnosed depression fully explained the effects of HbA1c and BMI on well-being. The interplay of glycemic control and positive mental health should be further investigated.