The authors surveyed 112 U.S. adults living with type 1 or type 2 diabetes who had used a Dexcom or Abbott Libre continuous glucose monitor within the preceding 12 months. The cohort had a mean age of 54.5 years and was 60.7% female. The study was approved by the Georgetown University Institutional Review Board (approval number STUDY000010019), and participants provided electronic informed consent before completing the survey. Data and codebooks for the analysis were made available in a Zenodo repository, as stated by the study team.
Respondents rated 26 mechanism-of-action (MoA) statements that the investigators mapped onto the Capability, Opportunity, Motivation-Behaviour (COM-B) framework. Participants endorsed a median of 18 MoAs. Endorsements clustered most often around items related to knowledge (awareness of glucose patterns), skills (interpreting data), and perceived consequences of behaviours on glucose—domains the authors report as frequently invoked by CGM users. The survey captured the breadth of perceived mechanisms but did not report physiological or objective behavioural measures; the source did not provide additional qualitative excerpts or examples from open-text responses.
Survey respondents identified specific app features they believed contributed to the MoAs they endorsed. Real-time glucose readings and graphical displays were most commonly credited with supporting multiple mechanisms. The report specifies that these features were frequently attributed to changes in knowledge and skill-based MoAs, and to users’ recognition of short-term behavioural consequences. The preprint did not report detailed ranking of every application feature beyond noting that graphs and real-time values were prominent contributors, nor did it provide a full feature-by-MoA matrix in the abstract.
Participants reported CGM-attributed dietary change at high frequency (82.1%) and physical activity change less frequently (46.4%). Logistic regression models tested associations between the number of MoAs endorsed and self-reported behaviour change. Each additional MoA endorsed was associated with higher odds of reporting dietary change (odds ratio 1.27, 95% CI 1.15–1.44) and physical activity change (odds ratio 1.27, 95% CI 1.16–1.43). When the authors examined the COM-B domains as predictors, Motivation emerged as making the strongest distinct contribution to reported behaviour change. The preprint presents these associations as perceived links between CGM use, the array of psychological and informational mechanisms users name, and self-reported modification of diet and activity.
The investigators declared competing interests for several authors, including prior consultations and institutional support involving industry partners; the preprint lists these disclosures explicitly. The study team reported NIH P30CA051008 among funders. The authors deposited the de-identified dataset and codebook in a Zenodo repository (DOI listed in the source), allowing independent access to the materials used for analysis.
This report is a cross-sectional, self-report survey and the preprint cautions that the findings have not been peer reviewed. The data reflect participants’ perceptions and attributions; the source did not report objective behavioural or glycaemic endpoints linked to MoA endorsement in the abstract. The study sample size and demographic summary are reported, but the preprint abstract does not provide additional subgroup analyses or longitudinal follow-up data. The authors do not claim causality; instead, they report associations between the number of perceived mechanisms and odds of self-reported behaviour change. Specific app feature efficacy, the durability of reported changes, and how these perceived mechanisms translate into measurable clinical outcomes remain unanswered by the material presented in the preprint. Future work, including peer review and studies with objective behavioural or glycaemic endpoints, would be necessary to determine whether the user-reported mechanisms correspond with sustained clinical benefit.