AbstractBackground and objectives:
Health-related quality of life (HR-QOL) and quality of life (QOL) are increasingly being examined as outcomes in assessments among patients with type 2 diabetes mellitus. However, there is a lack of standardization in interpreting the two outcomes and insufficient appreciation of the differences between HR-QOL and QOL. This study reports relationships between two instruments of HR-QOL and an instrument of QOL in a cross-sectional study of patients with type 2 diabetes.Methods:
Patients with type 2 diabetes at the outpatient clinics of a university hospital completed measures of generic health status (12-item Short-Form Health Survey [SF-12], version 2 and EQ-5D) and diabetes-specific QOL (Audit of Diabetes Dependent Quality of Life [ADDQoL]). Patient-reported data were merged with retrospective clinical data including glycosylated hemoglobin (HbA1c), co-morbidities, diabetes complications score, body mass index (BMI), and others, obtained from electronic medical records. A path model of our hypothesized relationships between the physical and psychological components of HR-QOL, overall HR-QOL, and QOL was tested in addition to examining bivariate correlations between these constructs. The fit of the path model was assessed using multiple indexes of fit, including an overall chi-squared (χ2) test, the Comparative Fit Index (CFI), the Tucker-Lewis Index (TLI), and the Root Mean Square Error Approximation (RMSEA). The differences in the association between clinical, medical history and sociodemographic variables with HR-QOL and QOL were explored employing univariate t-tests and ANOVAs as well as multiple regression models.Results:
The usable response rate was 44.3% (n = 385). The mean HbA1c of respondents was 7.2% (±1.4), mean duration of diabetes was 10.2 (±9.1) years, and 62.1% were obese (BMI ≥30 kg/m2). About 49% of respondents were taking oral medications only, 31.7% were taking oral medications and insulin, and 9.4% were taking insulin only. Spearman correlations of the EQ-5Dindex were 0.640 with the SF Physical Component Score (PCS)-12, 0.534 with the SF Mental Component Score (MCS)-12, and 0.316 with the ADDQoL (all p < 0.001). A path analytic model relating SF-12 scores with EQ-5Dindex and ADDQoL scores exhibited good fit (χ2 = 1.32; p = 0.250; CFI = 0.99; TLI = 0.99; RMSEA = 0.03). Insulin use and diabetes-related complications were significantly associated with poorer scores on all measures. Only ADDQoL scores were significantly better among those with the American Diabetes Association-recommended HbA1c level of <7.0% (p = 0.002). Obesity was significantly associated with impaired SF-12 and EQ-5Dindex scores but not ADDQoL scores, while depressive symptoms were significantly associated with poorer scores on all these measures. The included explanatory variables explained a greater proportion of the variance in HR-QOL (PCS-12, MCS-12, EQ-5Dindex) than in QOL (ADDQoL) scores.Conclusion:
The study found that HR-QOL measures showed small correlations with the impact of diabetes on QOL. The fit statistics supported the hypothesized relationships in the path model, and provided empirical evidence that HR-QOL is a subset of QOL. In comparison to HR-QOL, QOL was less explained by the included explanatory variables, suggesting a greater influence on QOL by factors not accounted for in the present study.