Medication adherence trajectories and association with risk factors and clinical outcomes in type 2 diabetes treatment

Abstract

Background Monitoring medication adherence and main clinical outcomes changes in type 2 diabetes (T2D) is essential for optimal patient management. This study aimed to compute and evaluate medication adherence in its three process phases among T2D patients treated with oral antidiabetic (OAD) medications assessing association with main clinical outcomes. Methods This retrospective cohort study included newly diagnosed T2D patients initiating OAD therapy within 2015–2019 identified from the Nivel Primary Care Database (Nivel-PCD) in the Netherlands. Initiation was operationalised as ≥2 OAD prescriptions within follow-up. Implementation was quantified monthly using the Continuous Multiple-interval Measure of Medication Availability (CMA9). Group-based trajectory modelling (k-means via AdhereR) identified adherence patterns over 12 months. Persistence was assessed with a permissible-gap rule on the same 12-month series. Clinical measures (HbA1c, LDL cholesterol, blood pressure, BMI) were evaluated at baseline and at 12 months, and HbA1c trends were analysed over time. Results Among 3,404 T2D patients started an OAD treatment, four distinct adherence clusters were identified: perfect adherence (70.1%), slow decline (13.3%), low adherence (10.6%), and slow increase (6.0%). Overall initiation was 99%. One-year persistence ranged from 26% (low adherence group) to 98.8% (perfect adherence group). Patients in the perfect-adherence group had higher baseline HbA1c but showed lower HbA1c and LDL levels over time compared with other groups. Blood pressure improved slightly in all groups. Changes in BMI were minimal. Conclusions Adherence trajectories provide a dynamic view of patients’ behaviour and are associated with better glycaemic and lipid control. High-risk patients were often in the perfect-adherence group and appeared to gain the highest clinical benefit. These results support routine adherence monitoring and targeted, trajectory-informed interventions in T2D care.