Study design
A randomized crossover study in adolescents with type 1 diabetes, directly comparing CarboZen with traditional manual carbohydrate counting.
ATTD 2026
CarboZen presented at ATTD
CarboZen was presented at ATTD through an e-poster by Dr. Andrea Scaramuzza, with a controlled study comparing the app with traditional carbohydrate counting in adolescents with type 1 diabetes.
Scientific presentation
A randomized study in a controlled setting
The abstract presented at ATTD describes a randomized crossover study conducted on 15 adolescents during a supervised diabetes camp. Participants used both CarboZen and traditional carbohydrate counting on consecutive days, with dietitian calculation based on weighed portions used as the reference.
The results showed that CarboZen significantly reduced the absolute carbohydrate counting error compared with the traditional method, while also reducing the potential insulin dosing error and achieving high perceived usability.
Why it matters
Participation at ATTD marks an important milestone for CarboZen: not only as a digital product, but as a solution observed in an international clinical and scientific setting.
Main results
CarboZen showed a significantly lower average carbohydrate counting error than the traditional method, with a clinically meaningful reduction in the risk of insulin bolus dosing errors.
Clinical relevance
This type of evidence strengthens the credibility of the project and shows how AI can provide practical, usable and meaningful support in daily meal management.