A novel refined classification system for type 2 diabetes in adults: A Chinese retrospective cohort study.

AIMS: We propose a simple type 2 diabetes mellitus (T2DM) classification method based on fasting C-peptide (FCP) levels and examined its feasibility and validity. METHODS: Adult T2DM patients first diagnosed in our tertiary care centre from January 2009 to January 2020 were included. Patients were followed until January 2021; their clinical characteristics, chronic complications, treatment regimen, and glycaemic control were compared. RESULTS: In total, 5644 T2DM patients were included. Three subgroups were established based on FCP levels: subtype T1 (FCP ≤ 1.0 µg/L), 1423 patients (25.21%); subtype T2 (FCP 1.0-2.5 µg/L), 2914 patients (51.63%); and subtype T3 (FCP ≥ 2.5 µg/L), 1307 patients (23.16%). T1 was characterised by older age, lower body mass indices, higher initial glycosylated haemoglobin (HbA1c) levels, and the lowest homoeostatic model assessment 2 estimates of ß-cell function (HOMA2-ß) and HOMA2-insulin resistance at baseline. The T3 group's clinical characteristics were opposite to those of T1. T3 patients showed higher incidence rates and risks of diabetic kidney disease, diabetic peripheral vascular disease, and non-alcoholic fatty liver, while the risks of diabetic retinopathy and diabetic peripheral neuropathy were highest in T1. Insulin, glycosidase inhibitors, and thiazolidinedione were the most frequently used drugs, but the use of metformin, dipeptidyl peptidase-4 inhibitor, and insulin secretagogue drugs was slightly lower in T1. T1 maintained higher HbA1c levels throughout follow-up. Overall HbA1c fluctuations were more significant in T3 than in T1 and T2. CONCLUSIONS: The new adult T2DM classification is simple and clear and will help classify different T2DM clinical characteristics and guide treatment plans.