AI could help people keep from developing diabetes

Publicly released:
International
CC:0
CC:0

Researchers from the US say that an AI-based diabetes intervention program was as good as a human coach-led program. The researchers randomly assigned 368 people with overweight or obesity and prediabetes to either an AI or a human-led diabetes prevention program. They found 31.7% of participants in the AI group and 31.9% of participants in the human group were able to achieve the 'primary composite outcome' - a 5% weight loss, 4% weight loss plus 150 minutes of physical activity per week, or a blood sugar reduction relative to holding off diabetes - after 12 months.

Attachments

Note: Not all attachments are visible to the general public. Research URLs will go live after the embargo ends.

Research JAMA, Web page The URL will go live after the embargo ends
Journal/
conference:
JAMA
Research:Paper
Organisation/s: Johns Hopkins University School of Medicine, USA
Funder: This study was fully funded by the NIDDK, which covered participant costs for both the human and AI-based DPPs (R01DK125780). AWwas supported by a grant from NIH/NIA K01AG076967. This study was made possible by the Johns Hopkins Institute for Clinical and Translational Research (ICTR), which is funded in part by Grant Number UL1TR001079 from the National Center for Advancing Translational Sciences (NCATS), a component of the NIH, and the NIH Roadmap for Medical Research.
Media Contact/s
Contact details are only visible to registered journalists.