Changes in circulating adiponectin, leptin, glucose and C-peptide in patients with ketosis-prone diabetes.

AIMS: To evaluate circulating adipokines in people with ketosis-prone diabetes, a heterogeneous disorder characterized by unprovoked ketoacidosis in people with previously unrecognized diabetes. METHODS: Patients presenting with ketoacidosis with no previous diabetes diagnosis were compared with patients with previously established Type 1 diabetes. Baseline assessments of autoimmune status (A+/A-), and ß-cell function (B+/B-), as well as leptin and adiponectin levels during a standardized mixed-meal tolerance test of 120 min, were performed. In all, 20 patients with heterogeneous ketosis-prone diabetes and 12 patients with Type 1 diabetes were evaluated at baseline, 12 and 24 months. RESULTS: At baseline, during a mixed-meal tolerance test, glucose and adiponectin concentrations were lower in patients with ketosis-prone diabetes than in those with Type 1 diabetes (P = 0.0023 and P < 0.0001, respectively), whereas C-peptide concentrations were higher, with no significant difference in leptin concentrations. Within 12 months, 11 patients with ketosis-prone diabetes (all A-/B+) were discontinued from insulin treatment (ketosis-prone diabetes - insulin group), while nine patients (four A-B-, four A+B- and one A-B+) were maintained on insulin (ketosis-prone diabetes + insulin group). Fasting C-peptide levels increased significantly over 24 months in the ketosis-prone diabetes - insulin group (P = 0.01), while HbA1c levels decreased (P < 0.0001). Overall, the ketosis-prone diabetes - insulin group had a higher BMI (P = 0.018), yet a lower fasting glucose concentration (P = 0.003) compared with the ketosis-prone diabetes + insulin group. Over 24 months, the mixed-meal tolerance test area-under-the-curve of C-peptide increased in the ketosis-prone diabetes - insulin group, with no change in ketosis-prone diabetes + insulin (P < 0.0001). At 24 months, in spite of the higher BMI in the ketosis-prone diabetes - insulin group, mixed-meal tolerance test glucose and leptin concentrations were significantly lower (P < 0.0001 and P = 0.017, respectively), while adiponectin levels were higher (P = 0.023) compared with the ketosis-prone diabetes + insulin group. CONCLUSIONS: In spite of the higher BMI in the ketosis-prone diabetes - insulin group, lower leptin and higher adiponectin levels may contribute to improved ß-cell function and insulin sensitivity, as evidenced by lower glucose and higher C-peptide levels. This allows insulin therapy to be withdrawn.

Clinical characteristics and long-term follow-up of ketosis-prone diabetes in Thai patients.

INTRODUCTION: Diabetes presenting with ketoacidosis is a heterogeneous disorder. The purpose of this study was to determine whether ketosis-prone diabetes (KPDM) in Thai patients were different from type1 diabetes by assessment of the beta-cell response to a standardized mixed meal and pancreatic autoantibodies. MATERIAL AND METHODS: 20 patients who were categorized as ketosis-prone diabetes based on the occurrence of unprovoked DKA after the age of 30 years were compared with 12 type1 diabetic patients. The beta-cell function and pancreatic autoantibodies were followed after resolution of DKA every 6 months for 2 years. RESULTS: Mean (±SD) age at presentation was 38.8±11.5 and 26.7+10.3 years in KPDM and type1 DM, respectively (p<0.05). Median (IQR) fasting plasma C-peptide obtained after resolution of DKA within 2 weeks was 0.90 ng/dl -(0.20-1.30) in KPDM compared with 0.10 ng/dl (0.10-0.45) in type1 diabetes and median peak stimulated plasma C-peptide was 6.80 ng/dl (0.90-9.80) compared with 0.10 ng/dl (0.10-0.75). Based on Aß classification, 4 patients were classified as A+ß-, 12 patients were classified as A-ß+, and 4 patients were classified as A-ß-. No patient was classified as A+ß+ in this study. At the median time of 31 months follow-up (range from 8-44 months), 11 patients from 12 A-ß+ KPDM (92%) could be withdrawn from insulin treatment successfully at median time of 5 months after admission. CONCLUSIONS: Thai KPDM patients had variable clinical course which were different from typical type1 DM. The Aß classification was proven to be useful predictors for consideration of insulin withdrawal after resolution of DKA.