2273-PUB: Translational Analytics Clustering Type 2 DM Patients Treated on Wholesome Millet in Reducing HbA1c

Aim: To study the effect of whole grain millet based food in type 2 DM patients on HbA1c reduction.

Methodology: The contexts of the datasets of patients were studied as they arrive at Coimbatore Diabetes Foundation for over a period of 3 years, were classified based on initial visit according to the HbA1c results. High risk > 8%, moderate risk 7% ≤ 8% and low risk < 7%, among 100 patients, were further clustered (as 1,2 and 3 containing 41, 49 and 10 patients respectively) based on age, BMI, FBS, PPBS, and HbA1c, using K-Mean cluster. The translational analytics focus on the conclusions of the datasets from the passive deduction scope to active prescriptive scalability for real time practice on the effect of wholesome millet diet (WMD), millets are small seeded grasses grown as a cereal crops. This approach of real time application of clinical practice in fundamental research that can be easily practiced for treatment of DM, despite the natural limitation of hereditary and aging process. The patients who were following WMD were tested for HbA1c periodically. The Translational analytics: The Interval Plot and individual control chart with the analysis of variance examines the significance of the reduction in the average level of HbA1c% for each cluster.

Findings: 1. Significant reduction of HbA1c between first and second visit from 11.7 to 7.4% (at P value = 0) for cluster group 3 after a lapse of 106 days. This was sustained in the 4th visit up to 480 days 2. In cluster 2 the average reduction from 8.9 to 7.1% in the interval of 79 days and further significant reduction was visible in 4 th visit at 6.6% (at P value 0) up to 335 days. 3. In cluster 1 the significant reduction of HbA1c was noticed at the fourth visit at 6.8% from 7.4% after a lapse of 361 days (P value at 0.007).