Adherence as a Predictor of Glycemic Control Among Adolescents With Type 1 Diabetes: A Retrospective Study Using Real-world Evidence

Purpose Metabolic control among adolescents with type 1 diabetes mellitus (T1DM) is generally poor. Nonadherence is a contributor to this poor glycemic control, leading to adverse outcomes. The findings of studies reporting the association between adherence and glycemic control are conflicting. This study aimed to assess the level of adherence among adolescents with T1DM and its relationship with glycemic control. Methods This was a retrospective, cross-sectional study that was conducted at Sidra Medicine, a state-of-the-art tertiary health care facility for women and children in Qatar. Mean blood or interstitial glucose monitoring frequency (BGMF) was used to assess adherence level among adolescents with T1DM, whereas glycemic control was assessed via documented glycated hemoglobin A 1c (HbA 1c ). Adolescents who had a mean BGMF of ≥4 checks per day were considered adherent, and those who had an HbA 1c level of <7% were considered as having controlled diabetes. Correlational and logistic regression analyses were performed to assess the relationship between adherence and glycemic control, incorporating other covariates into the model. Findings The rate of adherence among adolescents with T1DM in Qatar was 40.9%. Adherent adolescents had significantly lower median HbA 1c levels compared with nonadherent adolescents (9.0% vs. 9.7%; P = 0.002). A significant negative correlation was found between BGMF and HbA 1c level (correlation coefficient r s = −0.325; P < .001). Approximately 97% of nonadherent adolescents compared with 87% of adherent adolescents had suboptimal diabetes control (HbA 1c ≥7%) ( P = .016). Furthermore, nonadherent adolescents were 78% less likely to have controlled diabetes compared with adherent adolescents (adjusted odds ratio = 0.221; 95% CI, 0.063−0.778; P = 0.019). The combined effect of the determinants of glycemic control among adolescents with T1DM that were included in the multiple regression model was able to explain approximately 9% of the variances in glycemic control (Cox and Snell R 2 = 0.092). Implications The current findings suggest that nonadherence was highly prevalent among adolescents with T1DM and was a significant independent predictor of glycemic control, explaining 9% of the variability. This finding warrants further exploration of other possible predictors of poor glycemic control among the adolescent population. Comprehensive interventions, including educational, technological, and health service delivery aspects, aimed at improving adherence and ultimately optimizing glycemic control are warranted in adolescents with T1DM.