Machine learning research results published in Frontiers in Digital Health

Machine Learning


The findings confirm that engagement data provides clinical signals that directly impact ROI.

Analysis further reveals specific measurement frequencies that drive clinical outcomes and bend the cost curve

new york, March 10, 2026 /PRNewswire/ — Dario Health Co., Ltd. (NASDAQ: DRIO) (the “Company”, “DarioHealth” or “Dario”), a global digital health leader, today announced the publication of new peer-reviewed research results in the United States. Frontiers of digital health Significant and sustained glycemic improvements were demonstrated among users of the Dario platform.

dario logo
dario logo

An observational study titled Machine learning and engagement insights for personalized blood sugar management,” The research team analyzed real-world data from 22,414 adults with type 2 diabetes who had baseline blood sugar levels in the high-risk range. Researchers used advanced machine learning (“ML”) models and longitudinal mixed-effects analysis to identify distinct glycemic trajectories moderated by demographic, clinical, and engagement factors.

By applying a generalized linear mixed-effects tree model, the researchers uncovered key moderating factors that influence glycemic improvement. Importantly, the results suggest broadness across diverse user populations, as demonstrated by the lack of significant body mass index (“BMI”) differences between races. Higher levels of digital engagement, particularly frequent blood glucose monitoring and lifestyle activity tags, were associated with stronger and more sustained blood sugar improvements. Important and actionable insights identify 12 measurements per month as the threshold for blood sugar improvement.

“Our findings confirm that engagement is a clinical signal, not just a usage metric,” said Iffat Hershkovitz, Ph.D., vice president of clinical science at Dalio and senior author of the study. “We observed significant early reductions in blood sugar levels followed by sustained stabilization, especially among users who consistently monitored and utilized lifestyle tools. Machine learning allows us to translate these patterns into adaptive, data-driven strategies that optimize long-term diabetes management.”

“This study shows that digital health platforms can move beyond a one-size-fits-all approach,” said Dr. Omar Manejwala, Dalio’s chief medical officer. “By applying machine learning to real-world data at scale, we can identify which users respond best, when interventions are most effective, and how engagement behaviors impact outcomes. These insights allow us to dynamically personalize support and improve glycemic management in clinically meaningful ways.”



Source link