The Verge reviewer feeds metabolic data to Twin Health digital twin for 100 days
The Verge senior reviewer Victoria Song says she spent about 100 days feeding metabolic data to a Twin Health digital twin, which used continuous glucose monitor and other data to predict food responses with red, orange and green labels. Twin Health remains experimental and is available through corporate or private health plans.
Song wrote that her mornings began by "feeding" the digital twin. She logged body composition data from a smart scale into the Twin Health app, then logged lemon water taken with medications. On Mondays, she added calming breaths and blood pressure readings. When considering breakfast, she consulted the twin. A bowl of protein Cheerios and milk was orange alone but green if she took her daily walk afterward, she wrote, and she then logged the meal in the app. After a few hours, the app generated an insight from her glucose response and the twin became more accurate.
The twin was built from data she provided and from a continuous glucose monitor. Twin Health is a digital twin program intended to help people "heal" their metabolisms, including goals such as weight loss, diabetes management and better sleep, according to The Verge. It is not available for widespread download. Digital twins remain experimental, and the service is currently accessible only through corporate or private health plans, typically aimed at people with chronic metabolic issues.
Song's metabolic twin was "born" in June when a large box arrived from Twin Health. Building it required 90 days' worth of Dexcom G7 CGMs, a smart scale, a smart blood pressure cuff, three bottles of multivitamins and magnesium glycinate, and her Apple Watch. She then visited Quest Diagnostics, where a phlebotomist drew blood for further analysis.
A few days later, she met on Zoom with her Twin Health coach Amanda, a human, and a doctor to review lab results. The doctor and coach presented numbers and formulas on a shared screen. Song wrote that her liver was inflamed, which she said was understandable given a diagnosis of moderate nonalcoholic fatty liver. She also had insulin resistance despite ideal HbA1c and fasting glucose tests. Her pancreas appeared to be working overtime, but her insulin-producing beta cell count seemed fine "for now." Her triglycerides were high, while the rest of her lipids looked normal.
Song wrote that she knew she had inherited a "bunk metabolism," a conclusion reinforced by a year of testing CGMs as a non-diabetic. After the doctor logged off, Amanda asked her to define her goals. Song described a vague desire to reduce the number of medications she needed to keep her middle-aged metabolism functioning, and asked whether a digital twin could help. Amanda logged the goal, set up a follow-up call in two weeks, and encouraged her to log as many meals as possible during the first 30-day period. At the 90-day mark, another blood draw was scheduled.
Digital twins are a relatively new concept in physical health, both in medicine and sports, The Verge reported. They are virtual models of a person built from data points such as labs, wearable data, images and genetic profiles, and are continuously updated over time. Once built, clinicians can use AI to "test" how a patient might react to specific medicines or treatments on the digital twin before applying them to the actual patient. In theory, this could help healthcare providers assess injury risk, predict disease progression, and implement preventive care while minimizing adverse effects. The article described metabolic coaching as perhaps the lowest-stakes, highest-reward version of the technology that can be tested in the real world now.
Song wrote that she was relieved her digital twin did not have a visible avatar. She preferred the abstract color labels and found novelty in checking whether a cup of pineapple or a turkey hot dog with ketchup and relish would be orange or green. She said the experience at times reminded her of Kazuo Ishiguro's dystopian novel, though the provided article text was truncated.
Editor's Summary Twin Health's digital twin program is an experimental service that uses continuous glucose monitoring and other personal data to predict food reactions. The Verge's 100-day account illustrates how such models may support personalized metabolic coaching, while also showing that access is limited and clinical benefits are still being tested.