Blood glucose monitoring and glycemic control in adolescents with type 1 diabetes: meter downloads versus self-report

Background: Reported frequencies of blood glucose monitoring (BGM) by both adolescents and their caregivers serve as adherence proxies when meter downloads are not available. Yet, correlates of reported BGM frequencies and their predictive utility are understudied. Objective: To identify sociodemographic, psychological, and disease-specific correlates of reported BGM frequencies in adolescents with type 1 diabetes and to explore the predictive utility of BGM indices on glycemic control. Subjects: Study participants included caregivers and adolescents with type 1 diabetes (N = 143, 13-18 yr) receiving diabetes treatment at a tertiary care setting. Methods: At the initial visit, adolescents and caregivers reported on daily BGM frequencies. A sub-sample provided meter downloads. Adolescents also completed a depression inventory. Three months later, adolescents provided blood sampling for A1c assessment. Results: Multivariate general linear modeling identified that older adolescent age and more depressive symptoms were associated with reports of less frequent BGM. Two stepwise multivariate regression models examined the predictive utility of BGM indices (i.e., adolescent-reported BGM, caregiver-reported BGM, meter download) on glycemic control. Caregiver-reported BGM frequency predicted glycemic control in the absence of meter download data (p < 0.001). However, when clinical and contextual variables were included, meter download data were the most robust predictor of glycemic control (p < 0.0001). Conclusions: Meter downloads have the most robust association with glycemic control when contextual variables are considered. Caregiver-reported BGM frequencies can serve as reliable substitutes in the absence of meter download, but they may not be as reliable in adolescents with depressive symptoms.