Abstract
OBJECTIVE: To identify predictors of mental health disorders and self-care worsening in patients with diabetes through 15 months of COVID-19 pandemic. SUBJECTS AND METHODS: Prospective study following patients with type 1 and type 2 diabetes during the COVID-19 pandemic in Southern Brazil. Participants were evaluated through phone calls in two moments: first three months of the outbreak, and 15 months later. The outcomes were the assessment of worsening in mental health disorders (increase ≥ 10% in the total score of the Self-Report Questionnaire), the assessment of emotional distress related to diabetes (increase ≥ 10% in the total score of the Brazilian version of the Problem Areas in Diabetes), and worsening in self-care parameters (reduction ≥ 3 points in the Self-Care Inventory-Revised). Logistic regression models were used to determine the odds ratio (OR) and their respective confidence intervals. Point-biserial correlation coefficients (r(pb)) were used to measure the relationship between the variation in scores and patient characteristics. RESULTS: In total, 150 adults were enrolled (54.6 ± 13.9 years old, 58.7% female, 85.9% white), out of which 118 remained during follow up. After 18 months, 34,7% of them (52.2 ± 14.8 years old, 53.7% female, 87.5% white) worsened mental health scores. An increase in mental health disorders was experienced by patients with lower middle-income [OR 4.2 (1.2-15.0)], and those who reported greater difficulty managing diabetes [OR 3.2 (1.4-7.1); r(pb) 0.32, P < 0.01]. In contrast, those who perceived an improvement in diabetes control showed a reduction in their mental health scores [OR 0.3 (0.1-0.8)]. For self-care, there was a score worsening in patients with longer duration of diabetes [OR 1.1 (1.0-1.1)] and in those using insulin [OR 8.3 (1.7-41.4); r(pb) 0.23, P = 0.01]. Conversely, those who followed the social distancing guidance had an improvement in self-care [OR 0.4 (0.1-0.9); r(pb) 0.18, P = 0.05]. CONCLUSION: Some clinical and socioeconomic characteristics may be suitable for identifying patients at higher risk of mental health and self-care worsening, signaling who needs to be monitored more closely during crisis situations.