
AI-CGM Platform Engagement Linked to Clinically Meaningful Weight Loss
Key Takeaways
- A within-individual comparison showed ~3-fold greater weekly weight loss during engaged intervals than nonengaged intervals, consistent across obesity classes, sex, and age strata.
- Exclusion of type 1/2 diabetes, eating disorders, and GLP-1 receptor agonist use aimed to isolate the independent association between platform engagement and weight change.
Real-world data show AI coaching plus CGM helps adults with obesity lose up to 7% in 6 months, with engagement driving faster weekly weight loss.
Adults with obesity who actively engaged with an artificial intelligence (AI)– and continuous glucose monitoring (CGM)–integrated digital health platform lost significantly more weight than they did during periods of nonengagement, according to a real-world study published in the Journal of Diabetes Science and Technology.1
“Adults with obesity lost weight nearly 3 times faster during periods when they actively engaged with the Signos System: 1.17% vs 0.44% per week. At 6 months, participants with higher engagement lost ~6% of their body weight, while highly engaged participants with Class III obesity achieved over 7% weight loss,” Stephanie Kim, MD, MPH, principal investigator at Signos, assistant professor of clinical practice in the Department of Metabolism, Endocrinology & Nutrition, and medical director of the Diabetes Institute Clinical Research Unit, at the University of Washington, explained in an email interview.
“This is significant because it suggests that engagement is not simply an app-usage statistic—it may be an important and modifiable part of effective obesity treatment,” she told The American Journal of Managed Care®.
The retrospective cohort study evaluated 3007 adults 18 years or older with a body mass index (BMI) of 30 or higher who enrolled in the Signos System, a mobile platform that pairs Dexcom CGM data with AI-driven behavioral coaching. Individuals with type 1 or type 2 diabetes, eating disorders, or who were taking glucagon-like peptide-1 (GLP-1) receptor agonists were excluded, allowing the study to isolate the effect of platform engagement independent of pharmacotherapy.
Engagement was measured as total meaningful actions per day (TMAPD), defined as intentional behaviors such as logging meals or exercise, completing curriculum content, or responding to in-app insights. Investigators compared within-individual weight change during engaged (TMAPD > 0) vs nonengaged (TMAPD = 0) intervals.
How Much Weight Did Engaged Users Lose? Key Findings From the Study
Weekly weight loss was nearly 3-fold greater during engaged periods than nonengaged periods (mean [SE], 1.17% [0.04%] vs 0.44% [0.03%]; P < .001). This pattern was consistent across all 3 obesity classes, both sexes, and all age groups studied.
Timing of engagement did not appear to influence outcomes. Among participants stratified into early and late engagers, weekly weight loss during engaged periods did not differ significantly (1.18% [0.05%] vs 1.12% [0.07%]; P = .484). The authors noted that this finding diverges from prior digital health literature, which has often linked early adherence to superior outcomes.
“Prior digital health literature has often emphasized that early engagement is critical to long-term success,” Kim explained. “In our study, however, weight loss during periods of active engagement was remarkably similar among early and late engagers—1.18% vs 1.12% per week—suggesting that the benefits of engagement are not confined to a narrow early window.”
In a cross-sectional analysis of 1147 participants with weight data available at 180 days, mean total body weight loss (TBWL) was 5.14% (0.19%) overall. Participants in the high-engagement group (TMAPD ≥ 1.02, the cohort median) achieved 5.90% (0.25%) TBWL compared with 4.38% (0.23%) among participants with lower engagement (P < .001), surpassing the 5% threshold widely recognized as clinically meaningful for cardiometabolic risk reduction. Participants with the highest baseline BMI (≥ 40) who were highly engaged lost 7.09% (0.60%) of body weight, the largest effect observed across subgroups.
Higher engagement also corresponded with greater exercise duration (164.96 vs 131.03 min/wk; P < .001) and substantially more frequent meal logging (10.48 vs 2.39 logs/wk; P < .001), suggesting that app engagement tracked alongside broader behavioral change.
How Does Signos Compare With GLP-1s, and What Are the Study’s Limitations?
The study authors noted that weight loss observed at 6 months among highly engaged users was comparable to outcomes reported at 12 months in real-world studies of GLP-1 receptor agonists, although they emphasized that GLP-1 therapies still produce substantially greater absolute weight loss (15%-20%) in clinical trials conducted over 68 to 72 weeks.1,2 The authors framed engagement-driven digital interventions such as Signos as a potential complement or alternative for patients who face cost, access, or tolerability barriers to pharmacotherapy.1
The retrospective design limits causal inference, and the authors acknowledged that unmeasured factors—such as fluctuating motivation or life events—could confound the relationship between engagement and weight loss. The 180-day cross-sectional analysis captured only 38.1% of the eligible cohort, raising the possibility of selection bias, and medication use was not systematically tracked. The authors called for prospective studies to establish causality and to evaluate longer-term outcomes.
Signos is the only CGM system cleared by the FDA for weight management that does not require a prescription.3 The company stated that the findings support engagement as a modifiable target in digital obesity care, distinct from passive tracking tools that have historically shown more modest results and higher attrition.
“Signos supports people before medication, alongside pharmacotherapy, or as a nonpharmacologic option for those who do not want, cannot access, or cannot tolerate medication,” Kim said. “Lastly, Signos is a valuable complement to GLP-1 therapy and an effective solution for those looking to transition off a GLP-1 while maintaining their progress.”
References
1. Dixon W, Kim S, Shryack GE, et al. Engagement with an AI- and CGM-integrated digital health platform is associated with clinically significant weight loss. J Diabetes Sci Technol. Published online June 12, 2026. doi:10.1177/15209156261459741
2. Jastreboff AM, Aronne LJ, Ahmad NN, et al. Tirzepatide once weekly for the treatment of obesity. N Engl J Med. 2022;387(3):205-216. doi:10.1056/NEJMoa2206038
3. Latest study shows Signos drives up to 7% weight loss over 6 months, rivaling real-world GLP-1 weight loss. News release. Signos, Inc. June 23, 2026. Accessed June 23, 2026.




