The Role of Digital Twin AI Models in Diabetes Care and Remission | Recording-The Role of Digital Twin AI Models in Diabetes Care and Remission

Video Transcription

Video Summary

The webinar, hosted by the ADA Immunology, Immunogenetics, and Transplantation Interest Group and chaired by Maria Bettini (University of Utah), features Dr. Rohit Warrior, a diabetologist and regenerative medicine specialist, discussing how AI “digital twin” models can enable precision systems biology approaches to improve metabolic health.<br /><br />Warrior argues diabetes care must shift from simply lowering glucose to addressing upstream drivers of metabolic dysfunction, including insulin resistance, inflammation, autonomic imbalance, ectopic fat, loss of metabolic flexibility, and especially mitochondrial dysfunction. Because cellular-level biochemical reactions can’t be directly tracked in real time, digital twins integrate continuous patient data (e.g., CGM, heart rate variability, sleep, body composition, meal patterns) to build individualized computational replicas that recognize patterns and generate personalized, adaptive recommendations across nutrition, activity, sleep, and stress reduction. These models provide near-real-time feedback to reinforce “micro-habits” and reduce relapse.<br /><br />He cites evidence including a Cleveland Clinic randomized controlled trial (reported in NEJM Catalyst) and a real-world retrospective study (~2,000 participants) showing improvements in HbA1c, weight, time-in-range, insulin resistance (HOMA markers), medication deprescribing, and comorbidities such as fatty liver and hypertension. In Q&A, he notes growing preventive interest among high-risk but currently healthy individuals.

Keywords

AI digital twin models

precision systems biology

metabolic health optimization

type 2 diabetes care

mitochondrial dysfunction

continuous glucose monitoring (CGM)

insulin resistance and HOMA-IR

personalized lifestyle interventions

HbA1c and time-in-range improvements