AI Digital Twins Help Manage Diabetes, Obesity Without Expensive Drugs

Digital twins replace expensive weight loss drugs

Startup Twin Health is helping employees lose weight and reverse chronic conditions using AI-powered metabolic replicas to replace expensive GLP-1 medications. The Mountain View-based company uses a combination of wearable sensors and predictive algorithms to manage diabetes, prediabetes, and obesity without the need for weekly injections.

The program offers a direct alternative to drugs like Ozempic and Wegovy, which currently cost employers between $1,000 and $1,500 per person every month. Companies are struggling to maintain these high healthcare expenses as demand for weight loss medication surges across the United States. Twin Health provides a non-medication pathway for users to reduce their reliance on these drugs or avoid them entirely.

Rodney Buckley, the 55-year-old mayor of Third Lake, Illinois, lost 100 pounds in less than a year using the platform. Buckley weighed 376 pounds in March 2023 and had struggled with traditional dieting for decades. He joined the program through his wife’s employer and transitioned from being unable to walk a mile to completing 6.5 miles every morning.

Sensors track metabolic health in real time

Twin Health ships a comprehensive hardware kit to every user to gather baseline biological data. This kit includes several connected devices that feed information into a central mobile application. The company uses these data points to build a "digital twin" of the user's specific metabolism.

The hardware kit includes the following components:

• Continuous glucose monitors (CGM) to track blood sugar spikes
• Blood pressure cuffs for cardiovascular monitoring
• Smart scales to measure body fat percentage and weight
• Fitness trackers to monitor sleep, stress, and physical activity

The app analyzes these inputs to generate personalized recommendations for nutrition and movement. It predicts how a specific user’s blood sugar will react to certain foods before they eat them. This allows the system to suggest immediate behavioral changes, such as reducing a portion size or taking a walk after a meal to blunt a glucose spike.

Users log their meals by scanning barcodes, taking photos of their plates, or using voice commands. The AI categorizes these items into a color-coded system of green, yellow, and red foods. As a user’s metabolic health improves, the system updates these categories, occasionally moving a "red" restricted food into the "yellow" or "green" categories.

Clinical trials prove the technology works

The Cleveland Clinic recently conducted a randomized controlled trial to test the effectiveness of the Twin Health platform. Staff endocrinologist Kevin Pantalone recruited 150 participants with type 2 diabetes and an average A1C level of 7.2 percent. The study aimed to see if digital interventions could achieve clinical targets with fewer medications than standard care.

The 12-month trial produced the following results:

• 71 percent of Twin Health users achieved an A1C below 6.5 percent while reducing their medication.
• Only 2 percent of the control group achieved the same result.
• Twin Health users lost an average of 8.6 percent of their body weight.
• The control group lost 4.6 percent of their body weight.

The study also showed a significant reduction in GLP-1 drug usage among the participants. At the start, 41 percent of the Twin Health group used GLP-1 medications, but that number dropped to 6 percent by the end of the year. In contrast, GLP-1 usage in the control group increased from 52 percent to 63 percent during the same period.

The results, published in the New England Journal of Medicine Catalyst, suggest that real-time feedback is more effective than traditional nutrition counseling. Pantalone noted that patients often feel overwhelmed by static advice, whereas the app provides individualized reinforcement for behavior changes. One participant in the study lost 25 pounds and stopped all diabetes medications entirely.

Employers save money on healthcare costs

Twin Health currently works with nearly 200 employers, including the asset management firm Blackstone. These organizations use the service to curb the rising costs of metabolic health treatments. The company operates on a performance-based payment model, meaning it only collects fees when users hit specific health milestones.

Twin Health receives payment when users achieve the following outcomes:

• Lowered A1C levels or blood sugar stabilization
• Significant weight loss sustained over time
• Reduction or elimination of expensive metabolic medications

Jahangir Mohammed founded the company in 2018 after seeing the impact of type 2 diabetes on his own family. He designed the business model to align the company’s revenue with the actual health improvements of the participants. This approach appeals to large self-insured employers who are looking for measurable ROI on their health benefits packages.

For users like Buckley, the financial savings for the employer mirror the personal health benefits. Buckley had taken blood pressure medication for decades before his doctor recently reduced his dosage. His body fat percentage and blood pressure have continued a downward trend, providing the data-driven motivation he needs to maintain his new lifestyle.

Privacy and the future of metabolic health

The platform requires the collection of highly personal biometric data, which can raise privacy concerns for some users. Twin Health handles this data as a healthcare provider and must comply with HIPAA regulations and state-level privacy laws. The company undergoes third-party security assessments to ensure the integrity of the information it stores.

Employers do not see the granular, identifiable health data of individual employees. Instead, they receive aggregated and anonymized reports that show the overall enrollment numbers and the collective health outcomes of their workforce. This allows companies to track the financial impact of the program without infringing on the medical privacy of their staff.

Diabetes experts like Bernard Zinman of the University of Toronto see digital twins as a way to scale early intervention. Zinman believes that providing this technology to people in the early stages of diabetes could potentially reverse the condition or prevent it from developing. He suggests that the real-time nature of digital health tools makes them more effective than traditional periodic doctor visits.

The technology is expected to expand beyond diabetes management into broader categories of overweight and obesity treatment. As access to these interventions improves, more patients may be able to achieve significant weight loss without the high costs or side effects of pharmaceutical interventions. The shift toward digital twins represents a move toward continuous, data-driven care that adapts to the unique biology of every individual.