Introduction: Type 2 diabetes (T2D) is a heterogeneous disease for which causal mechanisms are incompletely understood and subclassification may improve patient management. In the latest assessment of patients with adult-onset diabetes, Ahlqvist and colleagues (2018) defined five new subgroups: an autoimmune form, two severe forms (insulin-deficient and insulin-resistant diabetes) and two mild forms (obesity and age-related diabetes).
Objectives: To stratify patients with T2D into subgroups and assess the impact of the clusters on outcomes and therapeutics.
Material and Methods: We conducted a cluster analysis of patients with T2D (n=1280) in an Endocrinology department. Clusters were based on three variables (presence of antibodies, age at diagnosis and BMI) and data from patient records was collected on development of complications and therapeutics.
Results: We identified 4 clusters of T2D, with significantly different patient characteristics. Cluster 1 (autoimmune) consisted of 2% of patients, Cluster 3 (obesity-related diabetes) of 63%, Cluster 4 (age-related diabetes) of 13% and Cluster 2 the remaining 22%. Clusters 1 to 4 presented a mean age at diagnosis of 46, 52, 54 and 72, respectively. The remaining clusters presented a mean BMI of 24.4 kg/m2. Cluster 3 presented the highest mean BMI value (31.7 kg/m2). Cluster 1 had the highest mean HbA1c (7.3%), while Cluster 4 the lowest (6.6%) (P-value 0.033). Nephropathy was the most common complication in this population. Retinopathy was most frequent in Clusters 1 (18%) and 2 (16%) (P-value 0,001). Insulin was prescribed to most patients in Cluster 1. Monotherapy was the trend in Cluster 4. Combination therapy was required more often in clusters 2 and 3 (P-value <0.001). The majority of patients in Clusters 2 and 3 (60%) presented family history of diabetes.
Discussion and Conclusions: Clusters 1 and 2 were characterized by early-onset disease, higher HbA1C and low BMI. Furthermore, they presented the highest prevalence of retinopathy. Cluster 3 presented the highest BMI. Combination therapies were more common in clusters 2 and 3. Cluster 4 was characterized by late-onset disease, low HbA1c and low BMI. Also, monotherapy was the treatment of choice. Nephropathy was the most common complication. This new classification is easily replicable in a real word setting, especially amid general practioners. It will be exciting to explore whether individuals respond differently to medications based on the pathway predominantly disrupted or whether they have a variable rate of progression and diabetic complications.
18 - 21 May 2019
European Society of Endocrinology