Integrating Continuous Glucose Monitoring into Personalised Nutrition: Retrospective Insights from Real-World Vively Use

Integrating Continuous Glucose Monitoring into Personalised Nutrition: Retrospective Insights from Real-World Vively Use | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Integrating Continuous Glucose Monitoring into Personalised Nutrition: Retrospective Insights from Real-World Vively Use Michelle R Jospe, Kelli Richardson, Susan M. Schembre This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6960134/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 28 Apr, 2026 Read the published version in JMIR Human Factors → Version 1 posted You are reading this latest preprint version Abstract The rising popularity of apps that sync with continuous glucose monitors (CGMs) reflects growing interest in on-demand, personalised care. Understanding user characteristics and engagement can inform the design of mobile-enabled tools that move healthcare beyond traditional models. Vively delivers personalised lifestyle guidance based on CGM biofeedback and self-monitored behaviour. The users (N = 7,647) were diverse, with a mean baseline glucose of 5.8 mmol/L (SD 1.1). They wore CGMs for a median of 15 days (IQR 14–30); 91.7% logged food, with daily meal logging dropping from 90.8% during wear to 1.5% after removal. In multivariate models, higher baseline glucose predicted longer CGM wear (IRR 1.15, 95% CI 1.13–1.17) but fewer days of food logging (IRR 0.96, 95% CI 0.94–0.98). Smart device syncing and older age were associated with higher engagement in both behaviours (IRRs 1.45 and 1.32). Health sciences/Health care/Disease prevention/Lifestyle modification Health sciences/Health care/Nutrition Health sciences/Health care/Weight management Health sciences/Medical research/Translational research Health sciences/Endocrinology/Endocrine system and metabolic diseases precision health digital health metabolic health personalised nutrition blood glucose self-monitoring biological feedback Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction In recent years, a growing number of companies have begun offering direct-to-consumer web- or app-based programs that are paired with continuous glucose monitors (CGMs) to optimize wellbeing and improve metabolic health. These programs, which typically involve real-time, CGM-based biofeedback reflecting an individual’s glucose responses to meals and other lifestyle behaviours, are increasingly marketed to individuals without diabetes. Users are encouraged to adjust their diet, meal timing, and physical activity based on their personalized guidance to promote weight loss and glycaemic stability, an indicator of metabolic health. 1 Early clinical trials have begun to demonstrate the effectiveness of combining CGM with personalized nutrition therapy on health outcomes among people without diabetes. 2 – 4 In parallel with growing consumer interest, the commercial landscape has rapidly expanded, with dozens of companies now offering mobile health apps that integrate CGM data with personalized feedback. This expansion has been driven by technological advances in CGM technology and artificial intelligence, growing health awareness and the precision health movement, and increasing accessibility of CGM devices. 5 While CGMs were originally developed for insulin-dependent diabetes management, manufacturers have expanded access by releasing over-the-counter versions in several countries, enabling broader use among health-conscious consumers. Paired with mobile apps that translate sensor data into behavioural insights, CGMs are being positioned not only as medical devices but also as tools for health behaviour change. 4 , 6 – 8 Despite the increasing interest in CGM-based mobile health apps, little is known about who is using these tools, how they are used in everyday settings, and patterns of engagement. Most companies do not publicly report user characteristics or behavioural engagement metrics, limiting our understanding of how CGM-based platforms are used outside of clinical contexts. Leveraging observational data from real-world users may enhance our ability to improve uptake, sustained engagement, and overall effectiveness of these tools, as healthcare continues to evolve from traditional models toward personalized, mobile-enabled care. This study aims to address that gap by analysing patterns of engagement with the Vively app (Vively Health Pty Ltd, Australia), a commercial, CGM-integrated platform focused on personalized nutrition and lifestyle modification. We describe user demographics, CGM usage patterns, food and activity tracking, and identify predictors of sustained engagement. These data offer a rare window into how individuals interact with CGM-based feedback tools in a real-world consumer setting. Methods Study Design This was a retrospective observational study of Vively app users between August 2021 and February 2025. All data were de-identified prior to analysis. The Georgetown University Institutional Review Board reviewed the study and determined it was exempt from human subjects oversight (IRB# STUDY00008558). In accordance with Vively’s Health Privacy Policy, users consented to the use of de-identified data for research by using the app and could opt out at any time. Study Sample To allow all users at least three months of potential engagement, we limited app initiation to November 2024 and included data through February 2025. Users were eligible for inclusion if they had at least one day of CGM sensor wear. The final analytical sample included 7647 users. Vively CGM Program The Vively app is a commercially available digital health application that integrates with the 14-day Abbott FreeStyle Libre 2 and 3 CGM. It delivers real-time glucose feedback, meal scoring, and behavioural guidance to support metabolic health, weight management, and disease prevention. Designed for users with and without diabetes, the app combines CGM data with self-reported food and activity logs, allowing users to observe how lifestyle factors affect glucose levels (Fig. 1). Vively includes meal-level scoring based on both nutrient quality and postprandial glucose responses. Users are encouraged to log meals during and between CGM wear periods to receive personalized feedback and track changes over time. The app can also sync with wearable devices to capture data on physical activity, sleep, and stress. Optional support from a registered dietitian is available as a paid add-on, offering one-on-one feedback for users seeking additional guidance. Users typically begin their Vively experience by wearing a CGM sensor for 14 days to establish a baseline of glucose patterns and responses to food, activity, and lifestyle factors. During this period, users are encouraged to log meals and workouts to help the app generate personalized insights. After this initial wear period, users may continue to engage with the app without wearing a sensor. During these times, users can still log meals and receive guidance based on their previous CGM data and overall dietary patterns. Users are advised to repeat CGM wear approximately every three months as a refresher and to receive updated feedback. This intermittent approach is intended to balance the benefits of biological feedback with long-term usability and cost. Figure 1. Screenshots of the Vively app Study Outcomes User engagement was assessed using two outcomes: daily food logging and CGM wear duration. Daily food logging was treated as a continuous variable (total food logging days) and as a binary outcome indicating whether a user ever logged food. CGM wear duration reflects the total number of days with CGM data. In addition to modelling total CGM wear duration, we categorized wear duration to reflect the intended, intermittent, CGM use pattern (2 weeks, every 3 months): less than one full wear duration (< 13 days), one full wear duration (13–15 days), one to two wear durations (16–27 days), and two or more wear durations (≥ 28 days). A CGM wear duration of 13–15 days will herein be referred to as 1 CGM wear and a CGM wear duration of 15 or more days will be referred to as 2 (or more) CGM wears. Food logging served as our primary measure of app engagement, since users received feedback mainly on this behaviour, while CGM sensor wear was used as a secondary measure of engagement. Diet quality and diet glucose scores, as derived by the Vively app, were extracted as indicators of the average nutritional and glycaemic of meals logged over the cumulative CGM wear and food logging duration. The diet quality score (range: 0–10) reflects the nutritional composition of meals, including macronutrient distribution, degree of food processing, and whether the food contained alcohol. The diet glucose score (range: 0–10) is derived from CGM data and captures postprandial glucose responses over a two-hour window following each reported meal, and includes area under the curve, time-in-target range, and peak glucose levels. For each user, average scores were calculated across all eligible meals to summarize how meals aligned with the app’s metabolic goals. Higher scores for each diet-related outcome are more favourable. Physical activity data were extracted from the Vively app as steps or workouts. Users with physical activity data were those who synced a connected device, such as a smartwatch or smartphone to the Vively app. Step counts were inferred from any non-null step data linked to a third-party source (e.g., Apple Health, Garmin). Workout data was collected from both third-party sources and user-logged events and were analysed descriptively as a proxy for physical activity engagement. Statistical Analysis We used descriptive statistics to summarize user characteristics, CGM wear, and app usage, reporting means and standard deviations or medians and interquartile ranges, as appropriate for the data distribution. Diet scores, step counts, and workout frequency were summarized descriptively and not included in modelling. Step counts were first restricted to a plausible range (500–100,000 steps per day), then participant-specific outliers were removed. Outliers were defined as days with step counts more than three standard deviations below an individual’s mean. CGM wear and food logging were the outcomes of interest, modelled separately as predictors of app engagement. To identify factors associated with food logging, we used a hurdle negative binomial model. The hurdle model accounts for zero inflation and overdispersion in the count of logging days by modelling two components: (1) a logistic regression estimating the odds of ever logging food (≥ 1 day vs. none), and (2) a zero-truncated negative binomial regression estimating the number of logging days among users who logged at least once. To reduce the influence of extreme values, food logging days were winsorised at the 99th percentile prior to modelling. Predictors in both components included age, sex (reference: female), BMI, baseline mean glucose, and connected device use. Total CGM wear days were modelled using a standard negative binomial regression, using the same set of predictors. Food logging days were excluded from this model to avoid adjusting for a potential downstream behaviour. All continuous predictors were modelled linearly and checked for collinearity. Exponentiated coefficients are reported as odds ratios (ORs) or incidence rate ratios (IRRs), with corresponding 95% confidence intervals. User characteristics and engagement outcomes of interest were compared across the pre-defined categories of CGM use using pairwise tests. For continuous variables, pairwise independent t-tests were conducted; for categorical variables, pairwise Fisher’s exact tests were used. All p-values were adjusted for multiple comparisons using the Bonferroni correction. Effect size was calculated using Eta-squared (η 2 ) for continuous variables and Cramér’s V for categorical variables. Users with missing values for any of the predictors were excluded from regression models; no imputation was performed. All analyses were conducted in R (v4.2.2), using the pscl and MASS packages for count models. Results Vively user characteristics The analysis included a total of 7647 individuals who used Vively with at least one day of CGM sensor wear (Table 1 ). The cohort was predominantly female, with roughly equal proportions in the normal (34.7%), overweight (34.2%), and obesity (29.4%) BMI categories. Age ranged from 18 to 88 years. Most users were based in Australia (93.7%), where Vively Health is headquartered. Table 1 User, engagement, and behavioural characteristics across CGM wear categories. Mean (SD) reported, unless otherwise specified. Number of CGM wears All < 1 1 1–2 ≥ 2 p Effect size 1 n (%) 7647 (100) 1298 (17.0) 3263 (42.7) 771 (10.1) 2315 (30.3) CGM wear duration (days) 8.5ᵃ (3.1) 14.6ᵇ (0.7) 22.0ᶜ (3.7) 59.1ᵈ (55.6) < 0.001 η² = 0.32 Sex (%) < 0.001 V = 0.13 Male 2120 (27.7) 379 (29.2) 876 (26.8) 208 (27.0) 657 (28.4) Female 4782 (62.5) 760 (58.6) ab 2126 (65.2) a 459 (59.5) ab 1437 (62.1)ᵇ Age (years) 44.4 (10.9) 41.2ᵃ (10.4) 43.8ᵇ (11.0) 44.2ᵇ (10.7) 47.0ᶜ (10.6) < 0.001 η² = 0.03 BMI (kg/m²) 27.8 (6.1) 28.1ᵃ (5.8) 27.4ᵇ (5.7) 28.1ᵃ (6.1) 28.0ᵃ (6.6) < 0.001 η² = 0.00 Mean baseline CGM glucose (mmol/L) 1 5.8 (1.1) 5.7ᵃ (1.0) 5.7ᵃ (0.8) 5.7ᵃ (1.1) 6.0ᵇ (1.4) < 0.001 η² = 0.01 CGM days with food logging (%) 89.1 (18.5) 90.5ᵃ (18.4) 93.5ᵇ (11.5) 85.2ᶜ (20.7) 83.5ᶜ (23.3) < 0.001 η² = 0.06 Average meals per day 3.5 (1.4) 3.2ᵃ (1.5) 3.7ᵇ (1.5) 3.4ᶜ (1.3) 3.4ᶜ (1.3) < 0.001 η² = 0.02 Vively diet quality score 2 8.1 (0.7) 8.0ᵃ (0.9) 8.1ᵇ (0.7) 8.1ᵇ (0.7) 8.2ᶜ (0.7) < 0.001 η² = 0.01 Vively diet glucose score 3 8.7 (0.8) 8.8ᵃ (0.7) 8.7ᵇ (0.7) 8.8ᵃ (0.7) 8.6ᵇ (0.9) < 0.001 η² = 0.01 Users with connected device (%) 4668 (61.0) 685ᵃ (52.8) 1886ᵇ (57.8) 483 c (62.6) 1614 d (69.7) < 0.001 V = 0.13 Average steps per day 3831 (4024) 3769 (3999) 3803 (4084) 3954 (4216) 3854 (3910) 0.902 η² = 0.00 1: Calculated from hours 24–96 of continuous glucose monitor (CGM) wear; users without at least 96 hours of data were excluded from these summary statistics. 2: A high diet quality score indicates better nutrient composition and less food processing; 3: A high diet glucose score indicates a more stable and controlled glucose response, while a low score suggests potentially more rapid and significant spikes in blood sugar. 4: Effect size calculated using Eta-squared (η2) for continuous variables and Cramér’s V for categorical variables. These reflect different types of association and are not directly comparable. For η²: small ≈ 0.01, medium ≈ 0.06, large ≈ 0.14; For Cramér’s V (df = 3): small ≈ 0.06, medium ≈ 0.17, large ≈ 0.29. Values with different superscript letters indicate statistically significant differences between groups within each outcome variable, based on pairwise independent t-tests or Fisher’s exact tests with Bonferroni correction. CGM wear patterns Users wore a CGM for a median of 15 days (25th percentile = 14 days, 75th percentile = 30 days). There were two peaks in total wear durations: 42.7% (n = 3263/7647) of users had 1 CGM wear (13–15 days) and 30.3% (n = 2315/7647) of users had two or more CGM wears (≥ 28 days), Fig. 2 . For those with 2 CGM wears (either partial or full wears, 16–30 days; 20.7%, n = 1583/7647), the median number of days between CGM wears was 31 days (25th percentile = 6 days, 75th percentile = 85 days) and patterns varied widely (Fig. 3 ). There were 141 users (1.8%) with 10 or more CGM wears (≥ 140 days). Reasons for partial CGM wears (< 14 days; 27.1%, n = 2069/7647) were not recorded. Food logging Most users (n = 7013/7647, 91.7%) logged at least one food item. The median number of entries was 47 items (25th percentile = 18, 75th percentile = 91) over a median of 12 days (25th percentile = 6, 75th percentile = 19). One hundred users logged on at least 100 days, and 8 users logged food on more than 300 days. As with CGM wear, the most common cumulative food logging duration was 13–15 days (n = 960/7647, 12.6%). On days with food entries, users reported an average of 3.5 meals per day (SD 1.4). Food logging remained high during periods of CGM wear but fell sharply once the CGM was removed. During the first 14 days of the initial CGM wear period, 90.8% (SD 8.9) of users logged at least one food item per day. In contrast, during the first 14 days after sensor removal, only 1.5% (SD 2.0) of users logged any food item per day. Diet scores were presented to users for each logged meal and were scored from 0 to 10, with higher scores suggesting better food choices. Diet quality scores, which were based on nutrient composition and degree of food processing, were available for 76.8% of logged meals. To receive a score, a meal needed to contain at least one ingredient with nutrient information. The cumulative, per-user mean diet quality score was 8.1 (SD 0.7). Diet glucose scores, based on postprandial glucose response, were available for 87.3% of logged meals. The cumulative, per-user mean diet glucose score was 8.7 (SD 0.8). Vively user, engagement, and behavioural characteristics across CGM wear categories When comparing Vively user’s, engagement, and behavioural characteristics across CGM wear categories, there were some significant between-group differences with small to medium effect sizes ( Table 2 ). Compared to users with 1 total CGM wear, users with 2 or more CGM wears were older (47.0 (SD 10.6) years vs. 43.8 (SD 11.0) years, p < 0.05), they had the highest mean baseline CGM glucose, they reported fewer meals per day (on average), and had slightly higher Vively diet quality scores. Users with less than 1 CGM wear were the youngest users, they were least likely to have a connected device synced with Vively, they reported the fewest number of meals per day and had the lowest Vively diet quality score. Users with 1–2 CGM wears were not markedly different from users with 1 CGM wear or 2 or more CGM wears. Steps and workouts Most users (n = 4668/7647, 61.0%) connected Vively to a third-party device. The most common devices were the Apple Watch (n = 3026/7647, 39.6%) and Garmin watch (n = 625/7647, 8.2%). Among those with connected devices, step counts were recorded on average for 80.6 days (SD 79.2), with a mean daily step count of 3831 (SD 4024). Exercise logging was high, with n = 6029/7647 (78.8%) users logging at least one workout. The number of workouts per user varied widely (median = 28, 25th percentile = 7, 75th percentile = 93). Predictors of food logging and CGM wears Figure 4 depicts the predictors of app engagement measured as food logging (total days) and CGM wear duration (total days). Baseline mean glucose showed divergent associations across outcomes, predicting greater CGM wear but reduced food logging. Device syncing, older age, and being female predicted higher engagement for both outcomes. Associations with BMI were small and inconsistent across models. Among the 7013 Vively users (91.7% of 7647) who logged food at least once, females logged 14% more days than males (IRR 1.14, 95% CI 1.09–1.19). Each additional decade of age was associated with an 10% increase in logging days (IRR 1.10, 95% CI 1.08–1.12), and each 1 mmol/L higher mean glucose was associated with 4% fewer logging days (IRR 0.96, 95% CI 0.94–0.98). Users with a connected device logged 45% more days than those without (IRR 1.45, 95% CI 1.39–1.51). Logging days were marginally lower with higher BMI (IRR 0.995, 95% CI 0.992–0.998). CGM wear duration was positively associated with baseline glucose (IRR 1.15, 95% CI: 1.13–1.17). Similar to food logging, use of a connected device was associated with more CGM days (IRR 1.32, 95% CI: 1.28–1.37). CGM wear days were 8% higher among women than men (IRR 1.08, 95% CI: 1.04–1.12) and were slightly higher among users with higher BMI (IRR 1.01, 95% CI: 1.00–1.01) and older age (IRR 1.18 per decade, 95% CI: 1.16–1.20). Discussion This large-scale observational analysis of Vively users offers a rare window into real-world use of a digital health app that integrates continuous glucose monitoring (CGM) with self-tracked behaviours. The cohort of 7,647 users was predominantly female and middle-aged, with a broad range of BMIs and mostly normoglycemic glucose levels. Most users wore a CGM for 15 days, with distinct peaks in wear duration at 13 to 15 and 28 to 30 days, reflecting one or two sensor periods, respectively. Users with repeat CGM use (two or more wears) tended to be older and had higher baseline glucose levels. Food logging, used as an indicator of app engagement, was high during CGM wear but declined sharply after removal. Several user characteristics predicted engagement, with some variation by outcome. Higher glucose levels were linked to longer CGM wear but fewer food logging days. Consistent predictors of greater engagement across both behaviours included device syncing, older age, and identifying as female. BMI showed small and inconsistent associations with engagement. The level of engagement among this cohort of Vively app users, specifically in food logging, was striking, even when accounting for potential self-selection bias. The observed level of self-directed adherence suggests a strong motivational factor, likely driven by the immediacy and personal relevance of glucose feedback. This aligns with prior research showing that combining food logging with self-monitoring of behavioural outcomes, such as self-weighing during weight loss efforts, is associated with higher engagement and greater goal attainment. 9 Other studies have found that personalized feedback independently improves adherence to food logging. 10 , 11 CGM may offer an even stronger motivator, as it provides immediate, personalised insights into the physiological consequences of food choices. Food logging itself is a key behaviour in achieving health goals and has been identified as one of the strongest predictors of weight loss. 9 , 12 Nevertheless, logging declined sharply within days of CGM removal, with daily logging rates falling from over 90 percent to just 1.5 percent. This pattern suggests that while CGM may sustain engagement during active wear, its motivational influence diminishes once biological feedback is no longer available. Several factors consistently predicted overall engagement with the app, including both food logging and CGM wear. Device syncing emerged as one of the strongest predictors across both outcomes, suggesting that integration of multiple data sources may enhance engagement by providing more complete and rewarding feedback. Previous research has found that integrating third-party devices, such as wearables, into intervention improves engagement. 13 As wearable technology becomes more sophisticated and apps increasingly integrate multiple devices and data streams, the role of device connectivity in sustaining user engagement warrants further investigation. Older age was also associated with higher engagement in our sample, though this relationship is inconsistently observed across studies of nutrition app use. 14 Women were more likely to engage in both food logging and CGM wear. However, as is common in behavioural research, our sample included a greater proportion of women, which may have influenced these findings. We found that baseline glucose showed divergent effects on engagement. While users with higher glucose were more likely to persist with CGM wear, they were paradoxically less likely to log food, potentially reflecting lower motivation for active diet tracking or greater burden of behaviour change. Users with higher glucose levels may perceive greater value in the sensor data itself, focusing on monitoring rather than on modifiable behaviours such as dietary tracking. This finding mirrors those of Kumbara and colleagues, where those with baseline mean glucose > 10 mmol/L (180 mg/dL) were less likely to log food intake, as well as exercise, sleep, and medication use. 15 Böhm et al. similarly observed that users of a diabetes management app showed greater engagement with modules requiring automated data entry, such as CGM, compared to those relying on manual input like food and medication logs. 16 Our findings support this distinction, with CGM wear showing more sustained engagement than food logging. Together, these results suggest that passive monitoring, like CGM, may appeal more to users managing chronic conditions, while active tracking, like food logging, may require stronger behavioural support in those with chronic conditions. This study has several limitations. As an observational dataset of paying users, the sample is self-selected and skewed toward women and Australian residents. Lack of data on comorbidities, medications, or socioeconomic status limits interpretability. Furthermore, step count was based on data from users' smartwatches or phones, without confirmation that devices were worn continuously, leading to potential underestimation when devices were removed or worn inconsistently. However, the study’s strengths include its large sample size and ecological validity, providing rare insight into real-world use of a CGM-integrated digital health app outside of research or clinical settings. The availability of granular engagement data across both passive and active behaviours also enables a nuanced understanding of how users interact with biological feedback tools. The rising popularity of apps that sync with continuous glucose monitors (CGMs) reflects growing interest in on-demand, personalized care. This large-scale, real-world analysis of Vively users underscores how user characteristics and engagement patterns can inform the design of mobile-enabled tools that move healthcare beyond traditional models. As sensors become more affordable and longer-lasting, future research should explore not only how to sustain engagement beyond CGM use but also how long sensor wear continues to influence behaviour. Identifying the behavioural drivers and specific features that support sustained engagement will be essential for developing interventions that are both scalable and effective. This work should also expand to other apps and digital health platforms that use biofeedback to support personalized nutrition, in order to understand which tools and contexts most effectively promote lasting dietary behaviour change. Translating lessons from early adopters to broader, more diverse populations will require attention to access, usability, and behavioural support, especially for those who may benefit most from metabolic feedback but are less intrinsically motivated to actively self-monitor. Declarations Author Contribution MRJ, SMS, and KMR contributed to the concept, design, and execution of the study. MRJ curated the data, conducted all analyses, and prepared the figures. MRJ and SMS co-wrote the manuscript. All authors contributed to manuscript review and editing. Data Availability The data that support the findings of this study are not publicly available due to commercialisation restrictions. References Suh, S. & Kim, J. H. Glycemic Variability: How Do We Measure It and Why Is It Important? Diabetes Metab. J. 39, 273–282 (2015). Basiri, R. & Cheskin, L. J. Personalized Nutrition Therapy without Weight Loss Counseling Produces Weight Loss in Individuals with Prediabetes Who Are Overweight/Obese: A Randomized Controlled Trial. Nutrients 16, 2218 (2024). Dixon, W. et al. Novel Glucose Metric ‘Latest Spike Time’ Correlated with Weight Loss at Six Months in People with Obesity Using the Signos System. Diabetes Technol. Ther. 27, 19–26 (2025). Richardson, K. M. et al. The efficacy of using continuous glucose monitoring as a behaviour change tool in populations with and without diabetes: a systematic review and meta-analysis of randomised controlled trials. Int. J. Behav. Nutr. Phys. Act. 21, 145 (2024). Didyuk, O., Econom, N., Guardia, A., Livingston, K. & Klueh, U. Continuous Glucose Monitoring Devices: Past, Present, and Future Focus on the History and Evolution of Technological Innovation. J. Diabetes Sci. Technol. 15, 676–683 (2021). Jospe, M. R. et al. Leveraging continuous glucose monitoring as a catalyst for behaviour change: a scoping review. Int. J. Behav. Nutr. Phys. Act. 21, 74 (2024). Jospe, M. R., Kendall, M., Schembre, S. M. & Roy, M. Real-World Effectiveness of Glucose-Guided Eating Using the Data-Driven Fasting App Among Adults Interested in Weight and Glucose Management: Observational Study. JMIR Form. Res. 9, e65368 (2025). Veluvali, A. et al. Impact of digital health interventions on glycaemic control and weight management. NPJ Digit. Med. 8, 20 (2025). Patel, M. L., Hopkins, C. M., Brooks, T. L. & Bennett, G. G. Comparing Self-Monitoring Strategies for Weight Loss in a Smartphone App: Randomized Controlled Trial. JMIR MHealth UHealth 7, e12209 (2019). Turk, M. W. et al. Self-Monitoring as a Mediator of Weight Loss in the SMART Randomized Clinical Trial. Int. J. Behav. Med. (2012) doi: 10.1007/s12529-012-9259-9 . Hutchesson, M. J. Enhancement of Self-Monitoring in a Web-Based Weight Loss Program by Extra Individualized Feedback and Reminders: Randomized Trial. J. Med. Internet Res. 18, e82 (2016). Turner-McGrievy, G. M. et al. Defining Adherence to Mobile Dietary Self-Monitoring and Assessing Tracking Over Time: Tracking at Least Two Eating Occasions per Day Is Best Marker of Adherence within Two Different Mobile Health Randomized Weight Loss Interventions. J. Acad. Nutr. Diet. (2019) doi: https://doi.org/10.1016/j.jand.2019.03.012 . Rayward, A. T., Vandelanotte, C., Itallie, A. V. & Duncan, M. J. The Association Between Logging Steps Using a Website, App, or Fitbit and Engaging With the 10,000 Steps Physical Activity Program: Observational Study. J. Med. Internet Res. 23, e22151 (2021). Jakob, R. et al. Factors Influencing Adherence to mHealth Apps for Prevention or Management of Noncommunicable Diseases: Systematic Review. J. Med. Internet Res. 24, e35371 (2022). Kumbara, A. B. et al. Impact of a Combined Continuous Glucose Monitoring–Digital Health Solution on Glucose Metrics and Self-Management Behavior for Adults With Type 2 Diabetes: Real-World, Observational Study. JMIR Diabetes 8, e47638 (2023). Böhm, A.-K., Jensen, M. L., Sørensen, M. R. & Stargardt, T. Real-World Evidence of User Engagement With Mobile Health for Diabetes Management: Longitudinal Observational Study. JMIR MHealth UHealth 8, e22212 (2020). Tables Table 2 is not available with this version. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 28 Apr, 2026 Read the published version in JMIR Human Factors → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6960134","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":479088813,"identity":"9bacf918-b5ba-4837-abd5-4dcee1317978","order_by":0,"name":"Michelle R Jospe","email":"","orcid":"","institution":"Georgetown Lombardi Comprehensive Cancer Center","correspondingAuthor":false,"prefix":"","firstName":"Michelle","middleName":"R","lastName":"Jospe","suffix":""},{"id":479088814,"identity":"caeccd2b-77f6-4443-9699-349f85bcee15","order_by":1,"name":"Kelli Richardson","email":"","orcid":"","institution":"Georgetown Lombardi Comprehensive Cancer Center","correspondingAuthor":false,"prefix":"","firstName":"Kelli","middleName":"","lastName":"Richardson","suffix":""},{"id":479088815,"identity":"1c809ee3-7e09-4242-a8b9-f585b60fddf0","order_by":2,"name":"Susan M. Schembre","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA80lEQVRIiWNgGAWjYHCC9J8fDGxQhSQIaHkgLVGQBmUnEKWF8YEEz4fDJGiR7z+cYCBhcN6un7338YuPP2wStzMwH7zNg0eLwY20hIQCg9vJM3uOm1nOSEhL3NnAlmyNV4sET8IBCaAWoF42Y56Ew7kbDvCYSePTIt9//mMDj8G5ZHuQlj9gLfzf8GphOJCQzMBjcMDOQCKN+TEDxBY2vFoMbiSkMUsYJCdInDnGxtiTlla/4TCbseUcvA47kMb44Y+dPX97G/OHHzY2xgbHmx/eeIPPYVCQ2MDAwAaJDmYilIOAPUjtByIVj4JRMApGwQgDADQLT6JvAG1yAAAAAElFTkSuQmCC","orcid":"","institution":"Georgetown Lombardi Comprehensive Cancer Center","correspondingAuthor":true,"prefix":"","firstName":"Susan","middleName":"M.","lastName":"Schembre","suffix":""}],"badges":[],"createdAt":"2025-06-23 23:38:07","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6960134/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6960134/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.2196/80734","type":"published","date":"2026-04-29T00:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":85985577,"identity":"1b17e7e0-712c-486e-901f-4bb67336c7c7","added_by":"auto","created_at":"2025-07-04 03:06:40","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":83474,"visible":true,"origin":"","legend":"\u003cp\u003eScreenshots of the Vively app\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6960134/v1/1fcf185edc0c94b9ebdd8a0e.jpg"},{"id":85985573,"identity":"06cdc81d-5531-4ddc-b434-78126afe061b","added_by":"auto","created_at":"2025-07-04 03:06:40","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":24167,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of total CGM wear days per user (N=7647)\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6960134/v1/c23c57f32264297e08e9a9a5.jpg"},{"id":85985574,"identity":"9f8caca4-d9ea-4d00-9556-b77a90aeff69","added_by":"auto","created_at":"2025-07-04 03:06:40","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":21831,"visible":true,"origin":"","legend":"\u003cp\u003eContinuous glucose monitor (CGM) wear patterns among users over 300 days (N=7647). Each row represents an individual user, ordered by the longest span of CGM wear. Green tiles indicate days with CGM wear, and grey tiles indicate days without data.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6960134/v1/6433d8ec3d8f8f035c308aaa.jpg"},{"id":85985586,"identity":"cc880830-4056-4b27-ad0b-b212482b01a9","added_by":"auto","created_at":"2025-07-04 03:06:40","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":31741,"visible":true,"origin":"","legend":"\u003cp\u003ePredictors of engagement behaviours. IRRs with 95% CIs from negative binomial models of food logging days and CGM wear days. IRRs are shown on a log scale; values \u0026gt;1 indicate positive associations.\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6960134/v1/b973220b28dcd9006c58f474.jpg"},{"id":108524154,"identity":"f832c1bd-4feb-4e05-a7ba-a3b98c155870","added_by":"auto","created_at":"2026-05-05 14:51:04","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":425766,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6960134/v1/26c09f75-6932-4c66-bc62-f6504629585a.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Integrating Continuous Glucose Monitoring into Personalised Nutrition: Retrospective Insights from Real-World Vively Use","fulltext":[{"header":"Introduction","content":"\u003cp\u003eIn recent years, a growing number of companies have begun offering direct-to-consumer web- or app-based programs that are paired with continuous glucose monitors (CGMs) to optimize wellbeing and improve metabolic health. These programs, which typically involve real-time, CGM-based biofeedback reflecting an individual\u0026rsquo;s glucose responses to meals and other lifestyle behaviours, are increasingly marketed to individuals without diabetes. Users are encouraged to adjust their diet, meal timing, and physical activity based on their personalized guidance to promote weight loss and glycaemic stability, an indicator of metabolic health.\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e Early clinical trials have begun to demonstrate the effectiveness of combining CGM with personalized nutrition therapy on health outcomes among people without diabetes.\u003csup\u003e\u003cspan additionalcitationids=\"CR3\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eIn parallel with growing consumer interest, the commercial landscape has rapidly expanded, with dozens of companies now offering mobile health apps that integrate CGM data with personalized feedback. This expansion has been driven by technological advances in CGM technology and artificial intelligence, growing health awareness and the precision health movement, and increasing accessibility of CGM devices.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e While CGMs were originally developed for insulin-dependent diabetes management, manufacturers have expanded access by releasing over-the-counter versions in several countries, enabling broader use among health-conscious consumers. Paired with mobile apps that translate sensor data into behavioural insights, CGMs are being positioned not only as medical devices but also as tools for health behaviour change.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eDespite the increasing interest in CGM-based mobile health apps, little is known about who is using these tools, how they are used in everyday settings, and patterns of engagement. Most companies do not publicly report user characteristics or behavioural engagement metrics, limiting our understanding of how CGM-based platforms are used outside of clinical contexts. Leveraging observational data from real-world users may enhance our ability to improve uptake, sustained engagement, and overall effectiveness of these tools, as healthcare continues to evolve from traditional models toward personalized, mobile-enabled care.\u003c/p\u003e \u003cp\u003eThis study aims to address that gap by analysing patterns of engagement with the Vively app (Vively Health Pty Ltd, Australia), a commercial, CGM-integrated platform focused on personalized nutrition and lifestyle modification. We describe user demographics, CGM usage patterns, food and activity tracking, and identify predictors of sustained engagement. These data offer a rare window into how individuals interact with CGM-based feedback tools in a real-world consumer setting.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eStudy Design\u003c/p\u003e \u003cp\u003eThis was a retrospective observational study of Vively app users between August 2021 and February 2025. All data were de-identified prior to analysis. The Georgetown University Institutional Review Board reviewed the study and determined it was exempt from human subjects oversight (IRB# STUDY00008558). In accordance with Vively\u0026rsquo;s Health Privacy Policy, users consented to the use of de-identified data for research by using the app and could opt out at any time.\u003c/p\u003e \u003cp\u003eStudy Sample\u003c/p\u003e \u003cp\u003eTo allow all users at least three months of potential engagement, we limited app initiation to November 2024 and included data through February 2025. Users were eligible for inclusion if they had at least one day of CGM sensor wear. The final analytical sample included 7647 users.\u003c/p\u003e \u003cp\u003eVively CGM Program\u003c/p\u003e \u003cp\u003eThe Vively app is a commercially available digital health application that integrates with the 14-day Abbott FreeStyle Libre 2 and 3 CGM. It delivers real-time glucose feedback, meal scoring, and behavioural guidance to support metabolic health, weight management, and disease prevention. Designed for users with and without diabetes, the app combines CGM data with self-reported food and activity logs, allowing users to observe how lifestyle factors affect glucose levels (Fig.\u0026nbsp;1).\u003c/p\u003e \u003cp\u003eVively includes meal-level scoring based on both nutrient quality and postprandial glucose responses. Users are encouraged to log meals during and between CGM wear periods to receive personalized feedback and track changes over time. The app can also sync with wearable devices to capture data on physical activity, sleep, and stress. Optional support from a registered dietitian is available as a paid add-on, offering one-on-one feedback for users seeking additional guidance.\u003c/p\u003e \u003cp\u003eUsers typically begin their Vively experience by wearing a CGM sensor for 14 days to establish a baseline of glucose patterns and responses to food, activity, and lifestyle factors. During this period, users are encouraged to log meals and workouts to help the app generate personalized insights. After this initial wear period, users may continue to engage with the app without wearing a sensor. During these times, users can still log meals and receive guidance based on their previous CGM data and overall dietary patterns. Users are advised to repeat CGM wear approximately every three months as a refresher and to receive updated feedback. This intermittent approach is intended to balance the benefits of biological feedback with long-term usability and cost. \u003cb\u003eFigure 1.\u003c/b\u003e Screenshots of the Vively app\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eStudy Outcomes\u003c/p\u003e \u003cp\u003eUser engagement was assessed using two outcomes: daily food logging and CGM wear duration. Daily food logging was treated as a continuous variable (total food logging days) and as a binary outcome indicating whether a user ever logged food. CGM wear duration reflects the total number of days with CGM data. In addition to modelling total CGM wear duration, we categorized wear duration to reflect the intended, intermittent, CGM use pattern (2 weeks, every 3 months): less than one full wear duration (\u0026lt;\u0026thinsp;13 days), one full wear duration (13\u0026ndash;15 days), one to two wear durations (16\u0026ndash;27 days), and two or more wear durations (\u0026ge;\u0026thinsp;28 days). A CGM wear duration of 13\u0026ndash;15 days will herein be referred to as 1 CGM wear and a CGM wear duration of 15 or more days will be referred to as 2 (or more) CGM wears. Food logging served as our primary measure of app engagement, since users received feedback mainly on this behaviour, while CGM sensor wear was used as a secondary measure of engagement.\u003c/p\u003e \u003cp\u003eDiet quality and diet glucose scores, as derived by the Vively app, were extracted as indicators of the average nutritional and glycaemic of meals logged over the cumulative CGM wear and food logging duration. The diet quality score (range: 0\u0026ndash;10) reflects the nutritional composition of meals, including macronutrient distribution, degree of food processing, and whether the food contained alcohol. The diet glucose score (range: 0\u0026ndash;10) is derived from CGM data and captures postprandial glucose responses over a two-hour window following each reported meal, and includes area under the curve, time-in-target range, and peak glucose levels. For each user, average scores were calculated across all eligible meals to summarize how meals aligned with the app\u0026rsquo;s metabolic goals. Higher scores for each diet-related outcome are more favourable.\u003c/p\u003e \u003cp\u003ePhysical activity data were extracted from the Vively app as steps or workouts. Users with physical activity data were those who synced a connected device, such as a smartwatch or smartphone to the Vively app. Step counts were inferred from any non-null step data linked to a third-party source (e.g., Apple Health, Garmin). Workout data was collected from both third-party sources and user-logged events and were analysed descriptively as a proxy for physical activity engagement.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eWe used descriptive statistics to summarize user characteristics, CGM wear, and app usage, reporting means and standard deviations or medians and interquartile ranges, as appropriate for the data distribution. Diet scores, step counts, and workout frequency were summarized descriptively and not included in modelling. Step counts were first restricted to a plausible range (500\u0026ndash;100,000 steps per day), then participant-specific outliers were removed. Outliers were defined as days with step counts more than three standard deviations below an individual\u0026rsquo;s mean.\u003c/p\u003e \u003cp\u003eCGM wear and food logging were the outcomes of interest, modelled separately as predictors of app engagement. To identify factors associated with food logging, we used a hurdle negative binomial model. The hurdle model accounts for zero inflation and overdispersion in the count of logging days by modelling two components: (1) a logistic regression estimating the odds of ever logging food (\u0026ge;\u0026thinsp;1 day vs. none), and (2) a zero-truncated negative binomial regression estimating the number of logging days among users who logged at least once. To reduce the influence of extreme values, food logging days were winsorised at the 99th percentile prior to modelling. Predictors in both components included age, sex (reference: female), BMI, baseline mean glucose, and connected device use.\u003c/p\u003e \u003cp\u003eTotal CGM wear days were modelled using a standard negative binomial regression, using the same set of predictors. Food logging days were excluded from this model to avoid adjusting for a potential downstream behaviour. All continuous predictors were modelled linearly and checked for collinearity. Exponentiated coefficients are reported as odds ratios (ORs) or incidence rate ratios (IRRs), with corresponding 95% confidence intervals.\u003c/p\u003e \u003cp\u003eUser characteristics and engagement outcomes of interest were compared across the pre-defined categories of CGM use using pairwise tests. For continuous variables, pairwise independent t-tests were conducted; for categorical variables, pairwise Fisher\u0026rsquo;s exact tests were used. All p-values were adjusted for multiple comparisons using the Bonferroni correction. Effect size was calculated using Eta-squared (η\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e) for continuous variables and Cram\u0026eacute;r\u0026rsquo;s V for categorical variables. Users with missing values for any of the predictors were excluded from regression models; no imputation was performed. All analyses were conducted in R (v4.2.2), using the pscl and MASS packages for count models.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eVively user characteristics\u003c/p\u003e \u003cp\u003eThe analysis included a total of 7647 individuals who used Vively with at least one day of CGM sensor wear (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The cohort was predominantly female, with roughly equal proportions in the normal (34.7%), overweight (34.2%), and obesity (29.4%) BMI categories. Age ranged from 18 to 88 years. Most users were based in Australia (93.7%), where Vively Health is headquartered.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eUser, engagement, and behavioural characteristics across CGM wear categories. Mean (SD) reported, unless otherwise specified.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNumber of CGM wears\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAll\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1\u0026ndash;2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003ep\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eEffect size\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003en (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7647 (100)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1298 (17.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3263 (42.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e771 (10.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2315 (30.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCGM wear duration (days)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8.5ᵃ (3.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14.6ᵇ (0.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e22.0ᶜ (3.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e59.1ᵈ (55.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eη\u0026sup2; = 0.32\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eV\u0026thinsp;=\u0026thinsp;0.13\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2120 (27.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e379 (29.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e876 (26.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e208 (27.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e657 (28.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4782 (62.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e760 (58.6)\u003csup\u003eab\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2126 (65.2)\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e459 (59.5)\u003csup\u003eab\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1437 (62.1)ᵇ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge (years)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e44.4 (10.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e41.2ᵃ (10.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e43.8ᵇ (11.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e44.2ᵇ (10.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e47.0ᶜ (10.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eη\u0026sup2; = 0.03\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI (kg/m\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e27.8 (6.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e28.1ᵃ (5.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e27.4ᵇ (5.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e28.1ᵃ (6.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e28.0ᵃ (6.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eη\u0026sup2; = 0.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean baseline CGM glucose (mmol/L)\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.8 (1.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.7ᵃ (1.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.7ᵃ (0.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.7ᵃ (1.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e6.0ᵇ (1.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eη\u0026sup2; = 0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCGM days with food logging (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e89.1 (18.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e90.5ᵃ (18.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e93.5ᵇ (11.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e85.2ᶜ (20.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e83.5ᶜ (23.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eη\u0026sup2; = 0.06\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage meals per day\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.5 (1.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.2ᵃ (1.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.7ᵇ (1.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.4ᶜ (1.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e3.4ᶜ (1.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eη\u0026sup2; = 0.02\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVively diet quality score\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8.1 (0.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8.0ᵃ (0.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8.1ᵇ (0.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e8.1ᵇ (0.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e8.2ᶜ (0.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eη\u0026sup2; = 0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVively diet glucose score\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8.7 (0.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8.8ᵃ (0.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8.7ᵇ (0.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e8.8ᵃ (0.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e8.6ᵇ (0.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eη\u0026sup2; = 0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUsers with connected device (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4668 (61.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e685ᵃ (52.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1886ᵇ (57.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e483\u003csup\u003ec\u003c/sup\u003e (62.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1614\u003csup\u003ed\u003c/sup\u003e (69.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eV\u0026thinsp;=\u0026thinsp;0.13\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage steps per day\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3831 (4024)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3769 (3999)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3803 (4084)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3954 (4216)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e3854 (3910)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.902\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eη\u0026sup2; = 0.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e1: Calculated from hours 24\u0026ndash;96 of continuous glucose monitor (CGM) wear; users without at least 96 hours of data were excluded from these summary statistics. 2: A high diet quality score indicates better nutrient composition and less food processing; 3: A high diet glucose score indicates a more stable and controlled glucose response, while a low score suggests potentially more rapid and significant spikes in blood sugar. 4: Effect size calculated using Eta-squared (η2) for continuous variables and Cram\u0026eacute;r\u0026rsquo;s V for categorical variables. These reflect different types of association and are not directly comparable. For η\u0026sup2;: small\u0026thinsp;\u0026asymp;\u0026thinsp;0.01, medium\u0026thinsp;\u0026asymp;\u0026thinsp;0.06, large\u0026thinsp;\u0026asymp;\u0026thinsp;0.14; For Cram\u0026eacute;r\u0026rsquo;s V (df\u0026thinsp;=\u0026thinsp;3): small\u0026thinsp;\u0026asymp;\u0026thinsp;0.06, medium\u0026thinsp;\u0026asymp;\u0026thinsp;0.17, large\u0026thinsp;\u0026asymp;\u0026thinsp;0.29. Values with different superscript letters indicate statistically significant differences between groups within each outcome variable, based on pairwise independent t-tests or Fisher\u0026rsquo;s exact tests with Bonferroni correction.\u003c/p\u003e \u003cp\u003eCGM wear patterns\u003c/p\u003e \u003cp\u003eUsers wore a CGM for a median of 15 days (25th percentile\u0026thinsp;=\u0026thinsp;14 days, 75th percentile\u0026thinsp;=\u0026thinsp;30 days). There were two peaks in total wear durations: 42.7% (n\u0026thinsp;=\u0026thinsp;3263/7647) of users had 1 CGM wear (13\u0026ndash;15 days) and 30.3% (n\u0026thinsp;=\u0026thinsp;2315/7647) of users had two or more CGM wears (\u0026ge;\u0026thinsp;28 days), Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e2\u003c/span\u003e. For those with 2 CGM wears (either partial or full wears, 16\u0026ndash;30 days; 20.7%, n\u0026thinsp;=\u0026thinsp;1583/7647), the median number of days between CGM wears was 31 days (25th percentile\u0026thinsp;=\u0026thinsp;6 days, 75th percentile\u0026thinsp;=\u0026thinsp;85 days) and patterns varied widely (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e3\u003c/span\u003e). There were 141 users (1.8%) with 10 or more CGM wears (\u0026ge;\u0026thinsp;140 days). Reasons for partial CGM wears (\u0026lt;\u0026thinsp;14 days; 27.1%, n\u0026thinsp;=\u0026thinsp;2069/7647) were not recorded.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFood logging\u003c/p\u003e \u003cp\u003eMost users (n\u0026thinsp;=\u0026thinsp;7013/7647, 91.7%) logged at least one food item. The median number of entries was 47 items (25th percentile\u0026thinsp;=\u0026thinsp;18, 75th percentile\u0026thinsp;=\u0026thinsp;91) over a median of 12 days (25th percentile\u0026thinsp;=\u0026thinsp;6, 75th percentile\u0026thinsp;=\u0026thinsp;19). One hundred users logged on at least 100 days, and 8 users logged food on more than 300 days. As with CGM wear, the most common cumulative food logging duration was 13\u0026ndash;15 days (n\u0026thinsp;=\u0026thinsp;960/7647, 12.6%). On days with food entries, users reported an average of 3.5 meals per day (SD 1.4).\u003c/p\u003e \u003cp\u003eFood logging remained high during periods of CGM wear but fell sharply once the CGM was removed. During the first 14 days of the initial CGM wear period, 90.8% (SD 8.9) of users logged at least one food item per day. In contrast, during the first 14 days after sensor removal, only 1.5% (SD 2.0) of users logged any food item per day.\u003c/p\u003e \u003cp\u003eDiet scores were presented to users for each logged meal and were scored from 0 to 10, with higher scores suggesting better food choices. Diet quality scores, which were based on nutrient composition and degree of food processing, were available for 76.8% of logged meals. To receive a score, a meal needed to contain at least one ingredient with nutrient information. The cumulative, per-user mean diet quality score was 8.1 (SD 0.7). Diet glucose scores, based on postprandial glucose response, were available for 87.3% of logged meals. The cumulative, per-user mean diet glucose score was 8.7 (SD 0.8).\u003c/p\u003e \u003cp\u003eVively user, engagement, and behavioural characteristics across CGM wear categories\u003c/p\u003e \u003cp\u003eWhen comparing Vively user\u0026rsquo;s, engagement, and behavioural characteristics across CGM wear categories, there were some significant between-group differences with small to medium effect sizes (\u003cb\u003eTable\u0026nbsp;2\u003c/b\u003e). Compared to users with 1 total CGM wear, users with 2 or more CGM wears were older (47.0 (SD 10.6) years vs. 43.8 (SD 11.0) years, p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), they had the highest mean baseline CGM glucose, they reported fewer meals per day (on average), and had slightly higher Vively diet quality scores. Users with less than 1 CGM wear were the youngest users, they were least likely to have a connected device synced with Vively, they reported the fewest number of meals per day and had the lowest Vively diet quality score. Users with 1\u0026ndash;2 CGM wears were not markedly different from users with 1 CGM wear or 2 or more CGM wears.\u003c/p\u003e \u003cp\u003eSteps and workouts\u003c/p\u003e \u003cp\u003eMost users (n\u0026thinsp;=\u0026thinsp;4668/7647, 61.0%) connected Vively to a third-party device. The most common devices were the Apple Watch (n\u0026thinsp;=\u0026thinsp;3026/7647, 39.6%) and Garmin watch (n\u0026thinsp;=\u0026thinsp;625/7647, 8.2%). Among those with connected devices, step counts were recorded on average for 80.6 days (SD 79.2), with a mean daily step count of 3831 (SD 4024). Exercise logging was high, with n\u0026thinsp;=\u0026thinsp;6029/7647 (78.8%) users logging at least one workout. The number of workouts per user varied widely (median\u0026thinsp;=\u0026thinsp;28, 25th percentile\u0026thinsp;=\u0026thinsp;7, 75th percentile\u0026thinsp;=\u0026thinsp;93).\u003c/p\u003e \u003cp\u003ePredictors of food logging and CGM wears\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e depicts the predictors of app engagement measured as food logging (total days) and CGM wear duration (total days). Baseline mean glucose showed divergent associations across outcomes, predicting greater CGM wear but reduced food logging. Device syncing, older age, and being female predicted higher engagement for both outcomes. Associations with BMI were small and inconsistent across models.\u003c/p\u003e \u003cp\u003eAmong the 7013 Vively users (91.7% of 7647) who logged food at least once, females logged 14% more days than males (IRR 1.14, 95% CI 1.09\u0026ndash;1.19). Each additional decade of age was associated with an 10% increase in logging days (IRR 1.10, 95% CI 1.08\u0026ndash;1.12), and each 1 mmol/L higher mean glucose was associated with 4% fewer logging days (IRR 0.96, 95% CI 0.94\u0026ndash;0.98). Users with a connected device logged 45% more days than those without (IRR 1.45, 95% CI 1.39\u0026ndash;1.51). Logging days were marginally lower with higher BMI (IRR 0.995, 95% CI 0.992\u0026ndash;0.998).\u003c/p\u003e \u003cp\u003eCGM wear duration was positively associated with baseline glucose (IRR 1.15, 95% CI: 1.13\u0026ndash;1.17). Similar to food logging, use of a connected device was associated with more CGM days (IRR 1.32, 95% CI: 1.28\u0026ndash;1.37). CGM wear days were 8% higher among women than men (IRR 1.08, 95% CI: 1.04\u0026ndash;1.12) and were slightly higher among users with higher BMI (IRR 1.01, 95% CI: 1.00\u0026ndash;1.01) and older age (IRR 1.18 per decade, 95% CI: 1.16\u0026ndash;1.20).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis large-scale observational analysis of Vively users offers a rare window into real-world use of a digital health app that integrates continuous glucose monitoring (CGM) with self-tracked behaviours. The cohort of 7,647 users was predominantly female and middle-aged, with a broad range of BMIs and mostly normoglycemic glucose levels. Most users wore a CGM for 15 days, with distinct peaks in wear duration at 13 to 15 and 28 to 30 days, reflecting one or two sensor periods, respectively. Users with repeat CGM use (two or more wears) tended to be older and had higher baseline glucose levels. Food logging, used as an indicator of app engagement, was high during CGM wear but declined sharply after removal. Several user characteristics predicted engagement, with some variation by outcome. Higher glucose levels were linked to longer CGM wear but fewer food logging days. Consistent predictors of greater engagement across both behaviours included device syncing, older age, and identifying as female. BMI showed small and inconsistent associations with engagement.\u003c/p\u003e \u003cp\u003eThe level of engagement among this cohort of Vively app users, specifically in food logging, was striking, even when accounting for potential self-selection bias. The observed level of self-directed adherence suggests a strong motivational factor, likely driven by the immediacy and personal relevance of glucose feedback. This aligns with prior research showing that combining food logging with self-monitoring of behavioural outcomes, such as self-weighing during weight loss efforts, is associated with higher engagement and greater goal attainment.\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e Other studies have found that personalized feedback independently improves adherence to food logging.\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e CGM may offer an even stronger motivator, as it provides immediate, personalised insights into the physiological consequences of food choices. Food logging itself is a key behaviour in achieving health goals and has been identified as one of the strongest predictors of weight loss.\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e Nevertheless, logging declined sharply within days of CGM removal, with daily logging rates falling from over 90 percent to just 1.5 percent. This pattern suggests that while CGM may sustain engagement during active wear, its motivational influence diminishes once biological feedback is no longer available.\u003c/p\u003e \u003cp\u003eSeveral factors consistently predicted overall engagement with the app, including both food logging and CGM wear. Device syncing emerged as one of the strongest predictors across both outcomes, suggesting that integration of multiple data sources may enhance engagement by providing more complete and rewarding feedback. Previous research has found that integrating third-party devices, such as wearables, into intervention improves engagement.\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e As wearable technology becomes more sophisticated and apps increasingly integrate multiple devices and data streams, the role of device connectivity in sustaining user engagement warrants further investigation. Older age was also associated with higher engagement in our sample, though this relationship is inconsistently observed across studies of nutrition app use.\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e Women were more likely to engage in both food logging and CGM wear. However, as is common in behavioural research, our sample included a greater proportion of women, which may have influenced these findings.\u003c/p\u003e \u003cp\u003eWe found that baseline glucose showed divergent effects on engagement. While users with higher glucose were more likely to persist with CGM wear, they were paradoxically less likely to log food, potentially reflecting lower motivation for active diet tracking or greater burden of behaviour change. Users with higher glucose levels may perceive greater value in the sensor data itself, focusing on monitoring rather than on modifiable behaviours such as dietary tracking. This finding mirrors those of Kumbara and colleagues, where those with baseline mean glucose\u0026thinsp;\u0026gt;\u0026thinsp;10 mmol/L (180 mg/dL) were less likely to log food intake, as well as exercise, sleep, and medication use.\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e B\u0026ouml;hm et al. similarly observed that users of a diabetes management app showed greater engagement with modules requiring automated data entry, such as CGM, compared to those relying on manual input like food and medication logs.\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e Our findings support this distinction, with CGM wear showing more sustained engagement than food logging. Together, these results suggest that passive monitoring, like CGM, may appeal more to users managing chronic conditions, while active tracking, like food logging, may require stronger behavioural support in those with chronic conditions.\u003c/p\u003e \u003cp\u003eThis study has several limitations. As an observational dataset of paying users, the sample is self-selected and skewed toward women and Australian residents. Lack of data on comorbidities, medications, or socioeconomic status limits interpretability. Furthermore, step count was based on data from users' smartwatches or phones, without confirmation that devices were worn continuously, leading to potential underestimation when devices were removed or worn inconsistently. However, the study\u0026rsquo;s strengths include its large sample size and ecological validity, providing rare insight into real-world use of a CGM-integrated digital health app outside of research or clinical settings. The availability of granular engagement data across both passive and active behaviours also enables a nuanced understanding of how users interact with biological feedback tools.\u003c/p\u003e \u003cp\u003eThe rising popularity of apps that sync with continuous glucose monitors (CGMs) reflects growing interest in on-demand, personalized care. This large-scale, real-world analysis of Vively users underscores how user characteristics and engagement patterns can inform the design of mobile-enabled tools that move healthcare beyond traditional models. As sensors become more affordable and longer-lasting, future research should explore not only how to sustain engagement beyond CGM use but also how long sensor wear continues to influence behaviour. Identifying the behavioural drivers and specific features that support sustained engagement will be essential for developing interventions that are both scalable and effective. This work should also expand to other apps and digital health platforms that use biofeedback to support personalized nutrition, in order to understand which tools and contexts most effectively promote lasting dietary behaviour change. Translating lessons from early adopters to broader, more diverse populations will require attention to access, usability, and behavioural support, especially for those who may benefit most from metabolic feedback but are less intrinsically motivated to actively self-monitor.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eMRJ, SMS, and KMR contributed to the concept, design, and execution of the study. MRJ curated the data, conducted all analyses, and prepared the figures. MRJ and SMS co-wrote the manuscript. All authors contributed to manuscript review and editing.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe data that support the findings of this study are not publicly available due to commercialisation restrictions.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSuh, S. \u0026amp; Kim, J. H. Glycemic Variability: How Do We Measure It and Why Is It Important? \u003cem\u003eDiabetes Metab. J.\u003c/em\u003e 39, 273\u0026ndash;282 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBasiri, R. \u0026amp; Cheskin, L. J. Personalized Nutrition Therapy without Weight Loss Counseling Produces Weight Loss in Individuals with Prediabetes Who Are Overweight/Obese: A Randomized Controlled Trial. \u003cem\u003eNutrients\u003c/em\u003e 16, 2218 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDixon, W. \u003cem\u003eet al.\u003c/em\u003e Novel Glucose Metric \u0026lsquo;Latest Spike Time\u0026rsquo; Correlated with Weight Loss at Six Months in People with Obesity Using the Signos System. \u003cem\u003eDiabetes Technol. Ther.\u003c/em\u003e 27, 19\u0026ndash;26 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRichardson, K. M. \u003cem\u003eet al.\u003c/em\u003e The efficacy of using continuous glucose monitoring as a behaviour change tool in populations with and without diabetes: a systematic review and meta-analysis of randomised controlled trials. \u003cem\u003eInt. J. Behav. Nutr. Phys. Act.\u003c/em\u003e 21, 145 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDidyuk, O., Econom, N., Guardia, A., Livingston, K. \u0026amp; Klueh, U. Continuous Glucose Monitoring Devices: Past, Present, and Future Focus on the History and Evolution of Technological Innovation. \u003cem\u003eJ. Diabetes Sci. Technol.\u003c/em\u003e 15, 676\u0026ndash;683 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJospe, M. R. \u003cem\u003eet al.\u003c/em\u003e Leveraging continuous glucose monitoring as a catalyst for behaviour change: a scoping review. \u003cem\u003eInt. J. Behav. Nutr. Phys. Act.\u003c/em\u003e 21, 74 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJospe, M. R., Kendall, M., Schembre, S. M. \u0026amp; Roy, M. Real-World Effectiveness of Glucose-Guided Eating Using the Data-Driven Fasting App Among Adults Interested in Weight and Glucose Management: Observational Study. \u003cem\u003eJMIR Form. Res.\u003c/em\u003e 9, e65368 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVeluvali, A. \u003cem\u003eet al.\u003c/em\u003e Impact of digital health interventions on glycaemic control and weight management. \u003cem\u003eNPJ Digit. Med.\u003c/em\u003e 8, 20 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatel, M. L., Hopkins, C. M., Brooks, T. L. \u0026amp; Bennett, G. G. Comparing Self-Monitoring Strategies for Weight Loss in a Smartphone App: Randomized Controlled Trial. \u003cem\u003eJMIR MHealth UHealth\u003c/em\u003e 7, e12209 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTurk, M. W. \u003cem\u003eet al.\u003c/em\u003e Self-Monitoring as a Mediator of Weight Loss in the SMART Randomized Clinical Trial. \u003cem\u003eInt. J. Behav. Med.\u003c/em\u003e (2012) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s12529-012-9259-9\u003c/span\u003e\u003cspan address=\"10.1007/s12529-012-9259-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHutchesson, M. J. Enhancement of Self-Monitoring in a Web-Based Weight Loss Program by Extra Individualized Feedback and Reminders: Randomized Trial. \u003cem\u003eJ. Med. Internet Res.\u003c/em\u003e 18, e82 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTurner-McGrievy, G. M. \u003cem\u003eet al.\u003c/em\u003e Defining Adherence to Mobile Dietary Self-Monitoring and Assessing Tracking Over Time: Tracking at Least Two Eating Occasions per Day Is Best Marker of Adherence within Two Different Mobile Health Randomized Weight Loss Interventions. \u003cem\u003eJ. Acad. Nutr. Diet.\u003c/em\u003e (2019) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jand.2019.03.012\u003c/span\u003e\u003cspan address=\"10.1016/j.jand.2019.03.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRayward, A. T., Vandelanotte, C., Itallie, A. V. \u0026amp; Duncan, M. J. The Association Between Logging Steps Using a Website, App, or Fitbit and Engaging With the 10,000 Steps Physical Activity Program: Observational Study. \u003cem\u003eJ. Med. Internet Res.\u003c/em\u003e 23, e22151 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJakob, R. \u003cem\u003eet al.\u003c/em\u003e Factors Influencing Adherence to mHealth Apps for Prevention or Management of Noncommunicable Diseases: Systematic Review. \u003cem\u003eJ. Med. Internet Res.\u003c/em\u003e 24, e35371 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKumbara, A. B. \u003cem\u003eet al.\u003c/em\u003e Impact of a Combined Continuous Glucose Monitoring\u0026ndash;Digital Health Solution on Glucose Metrics and Self-Management Behavior for Adults With Type 2 Diabetes: Real-World, Observational Study. \u003cem\u003eJMIR Diabetes\u003c/em\u003e 8, e47638 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eB\u0026ouml;hm, A.-K., Jensen, M. L., S\u0026oslash;rensen, M. R. \u0026amp; Stargardt, T. Real-World Evidence of User Engagement With Mobile Health for Diabetes Management: Longitudinal Observational Study. \u003cem\u003eJMIR MHealth UHealth\u003c/em\u003e 8, e22212 (2020).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTable 2 is not available with this version.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"precision health, digital health, metabolic health, personalised nutrition, blood glucose self-monitoring, biological feedback","lastPublishedDoi":"10.21203/rs.3.rs-6960134/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6960134/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe rising popularity of apps that sync with continuous glucose monitors (CGMs) reflects growing interest in on-demand, personalised care. Understanding user characteristics and engagement can inform the design of mobile-enabled tools that move healthcare beyond traditional models. Vively delivers personalised lifestyle guidance based on CGM biofeedback and self-monitored behaviour. The users (N\u0026thinsp;=\u0026thinsp;7,647) were diverse, with a mean baseline glucose of 5.8 mmol/L (SD 1.1). They wore CGMs for a median of 15 days (IQR 14\u0026ndash;30); 91.7% logged food, with daily meal logging dropping from 90.8% during wear to 1.5% after removal. In multivariate models, higher baseline glucose predicted longer CGM wear (IRR 1.15, 95% CI 1.13\u0026ndash;1.17) but fewer days of food logging (IRR 0.96, 95% CI 0.94\u0026ndash;0.98). Smart device syncing and older age were associated with higher engagement in both behaviours (IRRs 1.45 and 1.32).\u003c/p\u003e","manuscriptTitle":"Integrating Continuous Glucose Monitoring into Personalised Nutrition: Retrospective Insights from Real-World Vively Use","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-04 03:06:35","doi":"10.21203/rs.3.rs-6960134/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d7350265-9dd1-455b-855e-0052a76ad37c","owner":[],"postedDate":"July 4th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":50870697,"name":"Health sciences/Health care/Disease prevention/Lifestyle modification"},{"id":50870698,"name":"Health sciences/Health care/Nutrition"},{"id":50870699,"name":"Health sciences/Health care/Weight management"},{"id":50870700,"name":"Health sciences/Medical research/Translational research"},{"id":50870701,"name":"Health sciences/Endocrinology/Endocrine system and metabolic diseases"}],"tags":[],"updatedAt":"2026-05-05T14:50:55+00:00","versionOfRecord":{"articleIdentity":"rs-6960134","link":"https://doi.org/10.2196/80734","journal":{"identity":"jmir-human-factors","isVorOnly":true,"title":"JMIR Human Factors"},"publishedOn":"2026-04-29 00:00:00","publishedOnDateReadable":"April 29th, 2026"},"versionCreatedAt":"2025-07-04 03:06:35","video":"","vorDoi":"10.2196/80734","vorDoiUrl":"https://doi.org/10.2196/80734","workflowStages":[]},"version":"v1","identity":"rs-6960134","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6960134","identity":"rs-6960134","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}