Keywords

Diabetes • HbA1c • ROC curve • Mexican population • Depression

Abstract
Background Depression is present in individuals with type 2 diabetes mellitus.

Objective We evaluated the sociodemographic, anthropometric, and clinical features of Mexican individuals with clinical depressive symptoms in order to identify predictors for depression symptoms; finally, we evaluated if HbA1c levels could be used as a probable diagnostic of depressive symptoms in individuals with type 2 diabetes mellitus.

Methods The population studied consisted of 376 Mexicans with diabetes who were interviewed to collect information about comorbidities and habits. The evaluation of depressive symptoms was performed using the Hamilton scale. For a possible clinical association between HbA1c levels and depressive symptoms, we performed chi-square tests (χ2)/ t tests. A multivariate logistic regression model with the backward conditional method was used to identify predictors of depressive symptoms. An ROC curve was plotted to assess the possible role of HbA1c as a diagnostic predictive test of depression symptoms.

Results A total of 42.8% (n = 161) individuals showed clinical depressive symptoms. When comparing individuals with and without depression symptoms, those with depression symptoms showed higher levels of glucose and HbA1c; additionally, gender (p = 0.04), age (p = 0.006), HbA1c (p < 0.01), and complications related to diabetes were predictive factors for clinical depressive symptoms (p < 0.01). However, HbA1c showed a low diagnostic accuracy for depressive symptoms, with an area under the ROC curve of 0.59.

Conclusions Our findings provided evidence of the sex, age, HbA1c levels, and medical complications as predictors of clinical depressive symptoms in individuals with diabetes mellitus. Nevertheless, HbA1c levels are not useful as a diagnostic instrument for depressive symptom severity in these Mexican individuals

Download PDF View PDF