Interpreting blood GLUcose data with R package iglu

AI4Food, a feasibility study for the implementation of automated devices in the nutritional advice and follow up within a weight loss intervention

BACKGROUND & AIMS

The widespread prevalence of NCDs calls for an improvement in their prevention and treatment. Wearable technologies can be an important asset in the development of precision nutrition strategies, for both health professionals and patients. However, their clinical use is hindered by a lack of validation against current methodologies or appropriate tools to deliver nutritional strategies based on their data. Our study includes manual and automatic data capture methods within a weight loss intervention with the aim to create an essential asset for the implementation, validation, and benchmarking of AI-based tools in nutritional clinical practice.

METHODS

This is a feasibility prospective and crossover controlled trial for weight loss in overweight and obese participants, randomized into two groups: Group 1 used manual data collection methods based on validated questionnaires for the first two weeks; while Group 2 started with automatic data collection methods consisting of wearable sensors. After two weeks, the two groups switched data collection methods. Lifestyle data, anthropometric measurements and biological samples were collected from all participants.

RESULTS

A total of 93 participants completed the nutritional intervention designed for weight loss, achieving a mean reduction of 2 kg (V1: 84.99 SD ± 13.69, V3: 82.72 SD ± 13.32, p < 0.001). Significant reductions were observed in body mass index, visceral fat, waist circumference, total cholesterol, and HbA1c levels. The use of electronic devices proved satisfactory among the participants (System Usability Scale score 78.27 ± 12.86). We also report the presence of distinct patient groups based on continuous glucose measurements.

CONCLUSION

This study has yielded a large amount of data and has showcased how automatic data collection devices can be employed to gather data in the context of a nutritional intervention. This will enable the implementation of AI-based tools in nutritional clinical practice.

CLINICAL TRIAL REGISTRATION

ClinicalTrials.gov, NCT05807243.

A pretrained transformer model for decoding individual glucose dynamics from continuous glucose monitoring data

Continuous glucose monitoring (CGM) technology has grown rapidly to track real-time blood glucose levels and trends with improved sensor accuracy. The ease of use and wide availability of CGM will facilitate safe and effective decision making for diabetes management. Here, we developed an attention-based deep learning model, CGMformer, pretrained on a well-controlled and diverse corpus of CGM data to represent individual's intrinsic metabolic state and enable clinical applications. During pretraining, CGMformer encodes glucose dynamics including glucose level, fluctuation, hyperglycemia, and hypoglycemia into latent space with self-supervised learning. It shows generalizability in imputing glucose value across five external datasets with different populations and metabolic states (MAE = 3.7 mg/dL). We then fine-tuned CGMformer towards a diverse panel of downstream tasks in the screening of diabetes and its complications using task-specific data, which demonstrated a consistently boosted predictive accuracy over direct fine-tuning on a single task (AUROC = 0.914 for type 2 diabetes (T2D) screening and 0.741 for complication screening). By learning an intrinsic representation of an individual's glucose dynamics, CGMformer classifies non-diabetic individuals into six clusters with elevated T2D risks, and identifies a specific cluster with lean body-shape but high risk of glucose metabolism disorders, which is overlooked by traditional glucose measurements. Furthermore, CGMformer achieves high accuracy in predicting an individual's postprandial glucose response with dietary modelling (Pearson correlation coefficient = 0.763) and helps personalized dietary recommendations. Overall, CGMformer pretrains a transformer neural network architecture to learn an intrinsic representation by borrowing information from a large amount of daily glucose profiles, and demonstrates predictive capabilities fine-tuned towards a broad range of downstream applications, holding promise for the early warning of T2D and recommendations for lifestyle modification in diabetes management.

Continuous Glucose Monitoring Use in Youth with Type 2 Diabetes: A Pilot Randomized Study

Objective: Continuous glucose monitoring (CGM) enhances diabetes self-management in insulin-treated individuals. However, the feasibility, acceptability, and benefits/burdens in youth-onset type 2 diabetes (Y-T2D) who are on infrequent self-monitoring of blood glucose (SMBG) regimens remain unclear. Research Design and Methods: In Y-T2D prescribed SMBG less than or equal to twice daily, we conducted a 12-week randomized 2:1 parallel pilot trial of CGM versus fingerstick monitoring (Control). Control participants had an optional 4-week extension period to use CGM (Control-CGM). Feasibility was defined as recruitment, study participation, and retention >60% of individuals. Acceptability was defined as an individual CGM wear time of ≥60% at the end of the study. Diabetes distress and the benefits/burdens of CGM scores, hemoglobin A1c (HbA1c), and CGM-derived glycemic variables were compared at baseline and at the end of the intervention. Results: The recruitment rate was 54% (52 screened eligible, 18 CGM, 10 Control; 82% female, 68% Black, 14.9 ± 3.8 years, body mass index: 36.2 ± 7.7 kg/m2, HbA1c: 7.4 ± 2.4% (mean ± standard deviation [SD]), and 8 entered the optional Control-CGM group. The most commonly cited reason for declining study participation was reluctance to wear the device (50%). The participation rate was 91% and 75%, and retention was 100% and 75% for CGM and Control-CGM, respectively. A majority of Y-T2D had ≥60% wear time at the end of the study (CGM: 56% and Control-CGM: 83%). Wear time declined during the study (1st month: 71 ± 31% vs. 2nd month: 55 ± 32% vs. 3rd month: 38 ± 34%, P = 0.003). There were no significant changes in glycemia, CGM burden/benefits, or diabetes distress scores (P > 0.05). Minor sensor adhesion adverse events were common (75%) causes of reduced wear time. Conclusion: CGM was a feasible and acceptable adjunct to diabetes self-care among >50% of Y-T2D prescribed infrequent SMBG monitoring. Unwillingness to wear a device and social stigma impeded device use. Additional research is needed to mitigate the high rates of skin adhesion-related adverse events in this population.

Comparison of Computational Statistical Packages for the Analysis of Continuous Glucose Monitoring Data with a Reference Software, "Ambulatory Glucose Profile," in Type 1 Diabetes

Objective: To compare the accuracy of commonly used continuous glucose monitoring (CGM) analysis programs with ambulatory glucose profile (AGP) and Dexcom Clarity (DC) in analyzing CGM metrics in patients with type 1 diabetes (T1D). Research Methods: CGM data up to 90 days from 152 adults using the same CGM and automated insulin delivery system with T1D were collected. Six of the 19 CGM analysis programs (CDGA, cgmanalysis, Glyculator, iglu, EasyGV, and GLU) were selected to compare with AGP and DC. Metrics were compared etween all tools with two one-sided t-tests equivalence testing. For the equivalence test, the acceptable range of deviation was set as ±2 mg/dL for mean glucose, ±2% for time in range (TIR), ±1% for time above range (TAR), time above range level 1 (TAR1), time above range level 2 (TAR2), and coefficient of variation (CV). Results: All packages were compared with each other for all CGM metrics, and most of them had statistically significant differences for at least some metrics. All tools were equivalent to AGP for mean glucose, TIR, TAR, TAR1, and TAR2 within ±2 mg/dL, ±2%, ±1%, ±1% and 1%, respectively. CDGA, Glyculator, cgmanalysis, and iglu were not equivalent to AGP for CV within ±1%. All tools were equivalent to DC for mean glucose, TIR, and TAR2 within ±2 mg/dL, ±2%, and ±1%, respectively. Glyculator was not equivalent for TAR1, TAR, and CV. CGDA, cgmanalysis, and iglu were not equivalent to DC for TAR1 and TAR. EasyGV and GLU were not equivalent for TAR within ±1%. Conclusions: CGM analysis programs reported CGM metrics statistically differently, but these differences may not be applicable in clinical practice. The equivalence test also confirmed that the differences are negligible for TIR and mean glucose, while they can be important for hyperglycemic ranges and CV. A standardization for CGM data handling and analysis is necessary for clinical studies reporting CGM-generated outcomes.

Effects of four-weeks porcine-collagen hydrolysate consumption on glucose concentrations, glycemic variability, and fasting/postprandial cardiometabolic risk markers in men and women with overweight or obesity: A randomized, controlled trial

BACKGROUND

Different collagen hydrolysate sources have reduced fasting glucose concentrations. Although porcine-derived collagen hydrolysate predicts in vitro the highest potency for improving glucose metabolism, these effects have not been studied in humans.

AIM

To evaluate the effects of porcine-derived collagen hydrolysate on continuously monitored glucose concentrations in real-life conditions in individuals with overweight/obesity. Additionally, postprandial responses following a mixed meal test were examined.

METHODS

Fifty-six men and women participated in this randomized placebo-controlled parallel trial. After a two-week run-in period, participants consumed daily for four weeks 10 g porcine-derived collagen hydrolysate or placebo (erythritol). The primary outcome parameter was the interstitial glucose area under the curve (AUC) during daytime (07:00 to 22:00) measured during three consecutive days. In addition, glycemic variability (GV) was quantified. For this, a continuous glucose monitor (Freestyle Libre ProiQ, Wiesbaden, Germany) was used at the end of the run-in and intervention periods. Postprandial glucose, insulin, and triacylglycerol concentrations were also evaluated after a mixed meal tolerance test. Furthermore, fasting glucose, insulin, hemoglobin A1c (HbA1c), homeostatic model assessment for insulin resistance (HOMA-IR), HOMA of β-cell function (HOMA-β), and triacylglycerol changes were analyzed. Physical activity profiles and dietary intakes were monitored to exclude confounding by these lifestyle factors.

RESULTS

Collagen hydrolysate consumption did not significantly affect daytime interstitial glucose AUC concentrations (95%CI for the effect size: -5.1, 30.0 mmol/(L∗h); p-value = 0.159), but increased several GV metrics: standard deviation (95%CI: 0.0, 0.2 mmol/L; p-value = 0.011), continuous overall net glycemic action (CONGA-4) (95%CI: 0.1, 0.4 mmol/L; p-value = 0.015), coefficient of variation (95%CI: 0.1, 3.0 %; p-value = 0.036), M-value (95%CI: 0.2, 1.8; p-value = 0.036), and mean amplitude of glycemic excursions (MAGE) (95%CI: 0.2, 1.8 mmol/L; p-value = 0.036). Furthermore, the postprandial glucose AUC after the mixed meal test significantly increased (95%CI: 0, 103 mmol/L∗4-h; p-value = 0.049), as well as fasting insulin concentrations (p-value = 0.005), HOMA-IR (p-value = 0.008), and HOMA-β (p-value = 0.009). Other parameters, anthropometrics, physical activity, and energy/nutrient intakes were not significantly changed.

CONCLUSION

Four-week collagen hydrolysate intake did not change free-living glucose concentrations, but increased GV, postprandial glucose AUC, fasting insulin, HOMA-IR, and HOMA-β. However, these changes were small with limited clinical relevance. Therefore, it can be concluded that this porcine-derived collagen hydrolysate does not improve glucose metabolism or other cardiometabolic risk markers.

CLINICAL TRIAL REGISTRATION

This clinical trial was registered in November 2021 as NCT05282641.

A Processing Algorithm to Address Real-World Data Quality Issues With Continuous Glucose Monitoring Data

Continuous glucose monitoring (CGM) data stored in data warehouses often include duplicated or time-shifted uploads from the same patient, compromising data quality and accuracy of resulting CGM metrics. We developed a processing algorithm to detect and resolve these errors. We validated the algorithm using two weeks of CGM data from 2038 patients with diabetes. Duplication errors were identified in 528 patients, with 25.7% showing significant differences in at least one metric (Time in Range, Coefficient of Variation, Glycemic Management Indicator, or Glycemic Episode counts) between raw and processed data. Eleven patients crossed clinically meaningful thresholds in one or more metrics after processing. Our results underscore the importance of real-world CGM data processing to maintain accurate and reliable CGM metrics for research and clinical care.

Impact of Dietary Intake and Cardiorespiratory Fitness on Glycemic Variability in Adolescents: An Observational Study

Background

Cardiorespiratory fitness (CRF), estimated by maximum oxygen consumption (VO2 max) during exercise, is worsening among adolescents and associated with a decline in metabolic health into adulthood. Glycemic patterns may provide a mechanism between CRF and health.

Objectives

This study assessed the feasibility of measuring glycemic patterns using continuous glucose monitoring (CGM) in adolescents, aged 14-22 y, to estimate the relationship between VO2 max and glucose patterns.

Methods

Healthy adolescents (n = 30) were recruited for a treadmill VO2 max test and to complete the following activities for 7-10 d: 1) wear a Dexcom G6 CGM, 2) complete ≥3 24-h dietary recalls, and 3) complete 1 at-home oral glucose tolerance test (OGTT, 75 g glucose). Glycemic patterns were extracted as mean glucose, the coefficient of variance, the mean amplitude of glycemic excursions, and the mean of daily differences. The 2-h glucose responses to the OGTT and individual meals were extracted. Statistical analyses evaluated the relationship between VO2 max and 1) overall glycemic patterns and 2) the maximum glucose level and AUC response to OGTT and meals, stratified by sex.

Results

Participant feasibility demonstrated that 90% completed CGM data (n = 27), 87% ≥7 d of CGM data (n = 26), 97% attempted OGTT (n = 29), and 93% completed ≥3 dietary recalls (n = 28). Most participants had normal BMI (70%) with an even distribution of sex (44% male). Males exhibited an inverse relationship between VO2 max and overall mean glucose (ß= -7.7, P = 0.04). Males demonstrated an inverse relationship between VO2 max and 1) maximum glucose (ß = -29, P = 0.006) and AUC (ß = -2702, P = 0.001) in response to the OGTT and 2) AUC (ß = -1293, P = 0.03) in response to meals. No association was observed between VO2 max and glucose patterns in females.

Conclusions

A sex-specific relationship between VO2 max and glycemic patterns was observed, suggesting a unique metabolic capacity during late adolescence by sex.This trial was registered at clinicaltrials.gov as NCT05845827.

A case study on using a large language model to analyze continuous glucose monitoring data

Continuous glucose monitors (CGM) provide valuable insights about glycemic control that aid in diabetes management. However, interpreting metrics and charts and synthesizing them into linguistic summaries is often non-trivial for patients and providers. The advent of large language models (LLMs) has enabled real-time text generation and summarization of medical data. The objective of this study was to assess the strengths and limitations of using an LLM to analyze raw CGM data and produce summaries of 14 days of data for patients with type 1 diabetes. We first evaluated the ability of GPT-4 to compute quantitative metrics specific to diabetes found in an Ambulatory Glucose Profile (AGP). Then, using two independent clinician graders, we evaluated the accuracy, completeness, safety, and suitability of qualitative descriptions produced by GPT-4 across five different CGM analysis tasks. GPT-4 performed 9 out of the 10 quantitative metrics tasks with perfect accuracy across all 10 cases. The clinician-evaluated CGM analysis tasks had good performance across measures of accuracy [lowest task mean score 8/10, highest task mean score 10/10], completeness [lowest task mean score 7.5/10, highest task mean score 10/10], and safety [lowest task mean score 9.5/10, highest task mean score 10/10]. Our work serves as a preliminary study on how generative language models can be integrated into diabetes care through data summarization and, more broadly, the potential to leverage LLMs for streamlined medical time series analysis.

Weight-neutral early time-restricted eating improves glycemic variation and time in range without changes in inflammatory markers

Early time-restricted eating (eTRE) is a dietary strategy that restricts caloric intake to the first 6-8 h of the day and can effect metabolic benefits independent of weight loss. However, the extent of these benefits is unknown. We conducted a randomized crossover feeding study to investigate the weight-independent effects of eTRE on glycemic variation, multiple time-in-range metrics, and levels of inflammatory markers. Ten adults with prediabetes were randomized to eTRE (8-h feeding window, 80% of calories consumed before 14:00 h) or usual feeding (50% of calories consumed after 16:00 h) for 1 week followed by crossover to the other schedule. Using continuous glucose monitoring, we showed that eTRE decreased glycemic variation (mean amplitude of glycemic excursion) and time in hyperglycemia greater than 140 mg/dL without affecting inflammatory markers (erythrocyte sedimentation rate and C-reactive protein). These data implicate eTRE as a candidate dietary intervention for the weight-independent management of dysglycemia in high-risk individuals.

Examining Share plus-A Continuous Glucose Monitoring Plus Data-Sharing Intervention in Older Adults and Their Care Partners: Protocol for a Randomized Control Study

BACKGROUND

Older adults with type 1 diabetes (T1D) are increasingly turning to care partners (CPs) as resources to support their diabetes management. With the rise in diabetes technologies, such as continuous glucose monitoring (CGM), there is great potential for CGM data sharing to increase CP involvement in a way that improves persons with diabetes' glucose management and reduces distress.

OBJECTIVE

The specific aims of this paper are to (1) evaluate the feasibility, usability, and acceptability of the Share plus intervention compared to the CGM Follow app plus diabetes self-management education and support; (2) evaluate the effect of the Share plus intervention on time-in-range (TIR; primary outcome) and diabetes distress (secondary outcome); and (3) explore differences between groups in person with diabetes and CP dyadic appraisal and coping, quality of life, diabetes self-care, and CP burden at 12 and 24 weeks and associations of dyadic variables on outcomes.

METHODS

This is a protocol for a feasibility, pilot randomized controlled trial. Older adults with T1D and their CP (N=80 dyads) will be randomized 1:1 to the Share plus intervention or Follow app plus diabetes self-management education. The trial will include a 12-week active intervention to determine the change in primary (TIR) and secondary (diabetes distress) outcomes, followed by a 12-week, observation-only phase to examine maintenance effects. The evaluation is guided by the Dyadic Coping Model. Patient-level effectiveness outcomes (TIR, hemoglobin A1c [HbA1c], diabetes distress, diabetes appraisal, coping, quality of life, diabetes self-care behaviors, and CP burden) will be assessed, using patient-reported outcomes measures and a home HbA1c test kit. Patient- and CP-level acceptability and feasibility will be assessed using surveys and interviews. Quantitative feasibility, acceptability, and usability data will be described using frequencies and percentages. Acceptability will be summarized based on Likert questions and open-ended questions. Usability will be examined separately for the intervention and control groups based on the System Usability Scale, with a study benchmark of ≥68 indicating good usability. TIR will be computed based on 2 weeks' worth of data at baseline (prior to intervention) and 2 weeks each after the intervention (week 12) and at follow-up (week 24).

RESULTS

Recruitment started in August 2023 and enrollment began in November 2023. To date, 24 participants have been enrolled in this study. We expect to conclude this study in March 2026 and expect to disseminate results in March 2026.

CONCLUSIONS

To our knowledge, this will be the first pilot randomized controlled trial to evaluate both feasibility and effectiveness outcomes for the web-based, platform-delivered Share plus intervention for older adults with T1D and their CP. This research has implications for CGM data sharing in other age groups with T1D and type 2 diabetes.

TRIAL REGISTRATION

ClinicalTrials.gov NCT05937321; https://clinicaltrials.gov/study/NCT05937321.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

DERR1-10.2196/60004.

Phenome-wide associations of human aging uncover sex-specific dynamics

Aging varies significantly among individuals of the same chronological age, indicating that biological age (BA), estimated from molecular and physiological biomarkers, may better reflect aging. Prior research has often ignored sex-specific differences in aging patterns and mainly focused on aging biomarkers from a single data modality. Here we analyze a deeply phenotyped longitudinal cohort (10K project, Israel) of 10,000 healthy individuals aged 40-70 years that includes clinical, physiological, behavioral, environmental and multiomic parameters. Follow-up visits are scheduled every 2 years for a total of 25 years. We devised machine learning models of chronological age and computed biological aging scores that represented diverse physiological systems, revealing different aging patterns among sexes. Higher BA scores were associated with a higher prevalence of age-related medical conditions, highlighting the clinical relevance of these scores. Our analysis revealed system-specific aging dynamics and the potential of deeply phenotyped cohorts to accelerate improvements in our understanding of chronic diseases. Our findings present a more holistic view of the aging process, and lay the foundation for personalized medical prevention strategies.

Identification of gut microbiome features associated with host metabolic health in a large population-based cohort

The complex relationship between the gut microbiome and host metabolic health has been an emerging research area. Several recent studies have highlighted the potential effects of the microbiome's diversity, composition and metabolic production capabilities on Body Mass Index (BMI), liver health, glucose homeostasis and Type-2 Diabetes (T2D). The majority of these studies were constrained by relatively small cohorts, mostly focusing on individuals with metabolic disorders, limiting a comprehensive understanding of the microbiome's role in metabolic health. Leveraging a large-scale, comprehensive cohort of nearly 9000 individuals, measured using Continuous Glucose Monitoring (CGM), Dual-energy X-ray absorptiometry (DXA) scan and liver Ultrasound (US) we examined the functional profile of the gut microbiome, and its relation to 38 metabolic health measures. We identified 145 unique bacterial pathways significantly correlated with metabolic health measures, with 86.9% of these showing significant associations with more than one metabolic health measure. Furthermore, 87,678 unique bacterial gene families were found to be significantly associated with at least one metabolic health measure. Notably, "key" bacterial pathways such as purine ribonucleosides degradation and anaerobic energy metabolism demonstrated multiple robust associations across various metabolic health measures, highlighting their potential roles in regulating metabolic processes. Our results remained largely unchanged after adjustments for nutritional habits and for BMI they were replicated in a geographically independent cohort. These insights pave the way for future research and potentially the development of microbiome-targeted interventions to enhance metabolic health.

Relationship Between Sensor-Detected Hypoglycemia and Patient-Reported Hypoglycemia in People With Type 1 and Insulin-Treated Type 2 Diabetes: The Hypo-METRICS Study

OBJECTIVE

Use of continuous glucose monitoring (CGM) has led to greater detection of hypoglycemia; the clinical significance of this is not fully understood. The Hypoglycaemia-Measurement, Thresholds and Impacts (Hypo-METRICS) study was designed to investigate the rates and duration of sensor-detected hypoglycemia (SDH) and their relationship with person-reported hypoglycemia (PRH) in people living with type 1 diabetes (T1D) and insulin-treated type 2 diabetes (T2D) with prior experience of hypoglycemia.

RESEARCH DESIGN AND METHODS

We recruited 276 participants with T1D and 321 with T2D who wore a blinded CGM and recorded PRH in the Hypo-METRICS app over 10 weeks. Rates of SDH <70 mg/dL, SDH <54 mg/dL, and PRH were expressed as median episodes per week. Episodes of SDH were matched to episodes of PRH that occurred within 1 h.

RESULTS

Median [interquartile range] rates of hypoglycemia were significantly higher in T1D versus T2D; for SDH <70 mg/dL (6.5 [3.8-10.4] vs. 2.1 [0.8-4.0]), SDH <54 mg/dL (1.2 [0.4-2.5] vs. 0.2 [0.0-0.5]), and PRH (3.9 [2.4-5.9] vs. 1.1 [0.5-2.0]). Overall, 65% of SDH <70 mg/dL was not associated with PRH, and 43% of PRH had no associated SDH. The median proportion of SDH associated with PRH in T1D was higher for SDH <70 mg/dL (40% vs. 22%) and SDH <54 mg/dL (47% vs. 25%) than in T2D.

CONCLUSIONS

The novel findings are that at least half of CGM hypoglycemia is asymptomatic, even below 54 mg/dL, and many reported symptomatic hypoglycemia episodes happen above 70 mg/dL. In the clinical and research setting, these episodes cannot be used interchangeably, and both need to be recorded and addressed.

Early Dysglycemia Is Detectable Using Continuous Glucose Monitoring in Very Young Children at Risk of Type 1 Diabetes

OBJECTIVE

Continuous glucose monitoring (CGM) can detect early dysglycemia in older children and adults with presymptomatic type 1 diabetes (T1D) and predict risk of progression to clinical onset. However, CGM data for very young children at greatest risk of disease progression are lacking. This study aimed to investigate the use of CGM data measured in children being longitudinally observed in the Australian Environmental Determinants of Islet Autoimmunity (ENDIA) study from birth to age 10 years.

RESEARCH DESIGN AND METHODS

Between January 2021 and June 2023, 31 ENDIA children with persistent multiple islet autoimmunity (PM Ab+) and 24 age-matched control children underwent CGM assessment alongside standard clinical monitoring. The CGM metrics of glucose SD (SDSGL), coefficient of variation (CEV), mean sensor glucose (SGL), and percentage of time >7.8 mmol/L (>140 mg/dL) were determined and examined for between-group differences.

RESULTS

The mean (SD) ages of PM Ab+ and Ab- children were 4.4 (1.8) and 4.7 (1.9) years, respectively. Eighty-six percent of eligible PM Ab+ children consented to CGM wear, achieving a median (quartile 1 [Q1], Q3) sensor wear period of 12.5 (9.0, 15.0) days. PM Ab+ children had higher median (Q1, Q3) SDSGL (1.1 [0.9, 1.3] vs. 0.9 [0.8, 1.0] mmol/L; P < 0.001) and CEV (17.3% [16.0, 20.9] vs. 14.7% [12.9, 16.6]; P < 0.001). Percentage of time >7.8 mmol/L was greater in PM Ab+ children (median [Q1, Q3] 8.0% [4.4, 13.0] compared with 3.3% [1.4, 5.3] in Ab- children; P = 0.005). Mean SGL did not differ significantly between groups (P = 0.10).

CONCLUSIONS

CGM is feasible and well tolerated in very young children at risk of T1D. Very young PM Ab+ children have increased SDSGL, CEV, and percentage of time >7.8 mmol/L, consistent with prior studies involving older participants.

Impact of blood glucose on cognitive function in insulin resistance: novel insights from ambulatory assessment

BACKGROUND/OBJECTIVES

Insulin resistance (IR)-related disorders and cognitive impairment lead to reduced quality of life and cause a significant strain on individuals and the public health system. Thus, we investigated the effects of insulin resistance (IR), and blood glucose fluctuations on cognitive function under laboratory and free-living conditions, using ecological momentary assessment (EMA).

SUBJECTS/METHODS

Baseline assessments included neuropsychological tests and blood analysis. Individuals were classified as either insulin-sensitive (<2) or insulin-resistant (≥2), based on their Homeostatic Model Assessment (HOMA-IR) values. Continuous glucose monitoring (CGM) using a percutaneous sensor was performed for 1 week. Using multiple linear regression, we examined the effects of HOMA-IR and CGM metrics on cognitive domains. Working memory (WM) performance, which was assessed using EMA, 4 times a day for 3 consecutive days, was matched to short-term pre-task CGM metrics. Multilevel analysis was used to map the within-day associations of HOMA-IR, short-term CGM metrics, and WM.

RESULTS

Analyses included 110 individuals (mean age 48.7 ± 14.3 years, 59% female, n = 53 insulin-resistant). IR was associated with lower global cognitive function (b = -0.267, P = 0.027), and WM (b = -0.316; P = 0.029), but not with executive function (b = -0.216; P = 0.154) during baseline. EMA showed that higher HOMA-IR was associated with lower within-day WM performance (β = -0.20, 95% CI -0.40 to -0.00). CGM metrics were not associated with cognitive performance.

CONCLUSIONS

The results confirm the association between IR and decrements in global cognitive functioning and WM, while no effects of CGM metrics were observed, making IR a crucial time point for intervention. Targeting underlying mechanisms (e.g., inflammation) in addition to glycemia could be promising to minimize adverse cognitive effects. Registered under https://drks.de/register/de identifier no. DRKS00022774.

Real-life hypoglycaemia partially blunts the inflammatory response to experimental hypoglycaemia in people with type 1 diabetes

AIM

To determine whether recent repeated exposure to real-life hypoglycaemia affects the pro-inflammatory response during a hypoglycemia episode.

MATERIALS AND METHODS

This was a post hoc analysis of a hyperinsulinaemic normoglycaemic-hypoglycaemic clamp study, involving 40 participants with type 1 diabetes. Glucose levels 1 week before the clamp were monitored using a Freestyle Libre 1. Blood was drawn during normoglycaemia and hypoglycaemia, and 24 hours after resolution of hypoglycaemia for measurements of inflammatory responses and counterregulatory hormone levels. We determined the relationship between the frequency and duration of spontaneous hypoglycaemia, and time below range (TBR) and the inflammatory response to experimental hypoglycaemia.

RESULTS

On average, participants experienced 0.79 (0.43, 1.14) hypoglycaemia episodes per day, with a duration of 78 (47, 110) minutes and TBR of 5.5% (2.8%, 8.5%). TBR and hypoglycaemia frequency were inversely associated with the increase in circulating granulocyte and lymphocyte counts during experimental hypoglycaemia (P < .05 for all). A protein network consisting of DNER, IF-R, uPA, Flt3L, FGF-5 and TWEAK was negatively associated with hypoglycaemia frequency (P < .05), but not with the adrenaline response. Neither other counterregulatory hormones, nor hypoglycaemia awareness status, was associated with any of the inflammatory parameters markers.

CONCLUSIONS

Repeated exposure to spontaneous hypoglycaemia is associated with blunted effects of subsequent experimental hypoglycaemia on circulating immune cells and the number of inflammatory proteins.

AGATA: A Toolbox for Automated Glucose Data Analysis

BACKGROUND

Analyzing continuous glucose monitoring (CGM) data is a mandatory step for multiple purposes spanning from reporting clinical trial outcomes to developing new algorithms for diabetes management. This task is repetitive, and scientists struggle in computing literature glucose control metrics and waste time in reproducing possibly complex plots and reports. For this reason, to provide the diabetes technology community a unified tool, here we present Automated Glucose dATa Analysis (AGATA), an automated glucose data analysis toolbox developed in MATLAB/Octave.

METHODS

Automated Glucose dATa Analysis is an open-source software program to visualize and preprocess CGM data, compute glucose control metrics, detect adverse events, evaluate the effectiveness of users' prediction algorithms, and compare study arms. Automated Glucose dATa Analysis can be used as a standalone computer application accessible through a dedicated graphical user interface, particularly suitable for clinicians, or by integrating its functionalities in user-defined MATLAB/Octave scripts, which fits the need of researchers and developers. To demonstrate its features, we used AGATA to analyze CGM data of two subjects extracted from a publicly available data set of individuals with type one diabetes. Finally, AGATA's features are compared against those of 12 noncommercial software programs for CGM data analysis.

RESULTS

Using AGATA, we easily preprocessed, analyzed, and visualized CGM data in a handy way, in compliance with the requirements and the standards defined in the literature. Compared to the other considered software programs, AGATA offers more functionalities and capabilities.

CONCLUSION

Automated Glucose dATa Analysis is easy to use and reduces the burden of CGM data analysis. It is freely available in GitHub at https://github.com/gcappon/agata.

Continuous Glucose Monitor Metrics Are Associated with Emergency Department Visits and Hospitalizations for Hypoglycemia and Hyperglycemia, but Have Low Predictive Value

Objective: Determine whether continuous glucose monitor (CGM) metrics can provide actionable advance warning of an emergency department (ED) visit or hospitalization for hypoglycemic or hyperglycemic (dysglycemic) events. Research Design and Methods: Two nested case-control studies were conducted among insulin-treated diabetes patients at Kaiser Permanente, who shared their CGM data with their providers. Cases included dysglycemic events identified from ED and hospital records (2016-2021). Controls were selected using incidence density sampling. Multiple CGM metrics were calculated among patients using CGM >70% of the time, using CGM data from two lookback periods (0-7 and 8-14 days) before each event. Generalized estimating equations were specified to estimate odds ratios and C-statistics. Results: Among 3626 CGM users, 108 patients had 154 hypoglycemic events and 165 patients had 335 hyperglycemic events. Approximately 25% of patients had no CGM data during either lookback; these patients had >2 × the odds of a hypoglycemic event and 3-4 × the odds of a hyperglycemic event. While several metrics were strongly associated with a dysglycemic event, none had good discrimination. Conclusion: Several CGM metrics were strongly associated with risk of dysglycemic events, and these can be used to identify higher risk patients. Also, patients who are not using their CGM device may be at elevated risk of adverse outcomes. However, no CGM metric or absence of CGM data had adequate discrimination to reliably provide actionable advance warning of an event and thus justify a rapid intervention.

Glucose Color Index: Development and Validation of a Novel Measure of the Shape of Glycemic Variability

BACKGROUND

Standard continuous glucose monitoring (CGM) metrics: mean glucose, standard deviation, coefficient of variation, and time in range, fail to capture the shape of variability in the CGM time series. This information could facilitate improved diabetes management.

METHODS

We analyzed CGM data from 141 adults with type 2 diabetes in the Hyperglycemic Profiles in Obstructive Sleep Apnea (HYPNOS) trial. Participants in HYPNOS wore CGM sensors for up to two weeks at two time points, three months apart. We calculated the log-periodogram for each time period, summarizing using disjoint linear models. These summaries were combined into a single value, termed the Glucose Color Index (GCI), using canonical correlation analysis. We compared the between-wear correlation of GCI with those of standard CGM metrics and assessed associations between GCI and diabetes comorbidities in 398 older adults with type 2 diabetes from the Atherosclerosis Risk in Communities (ARIC) study.

RESULTS

The GCI achieved a test-retest correlation of R = .75. Adjusting for standard CGM metrics, the GCI test-retest correlation was R = .55. Glucose Color Index was significantly associated (p < .05) with impaired physical functioning, frailty/pre-frailty, cardiovascular disease, chronic kidney disease, and dementia/mild cognitive impairment after adjustment for confounders.

CONCLUSION

We developed and validated the GCI, a novel CGM metric that captures the shape of glucose variability using the periodogram signal decomposition. Glucose Color Index was reliable within participants over a three-month period and associated with diabetes comorbidities. The GCI suggests a promising avenue toward the development of CGM metrics which more fully incorporate time series information.

Leveraging Large Language Models to Analyze Continuous Glucose Monitoring Data: A Case Study

Continuous glucose monitors (CGM) provide patients and clinicians with valuable insights about glycemic control that aid in diabetes management. The advent of large language models (LLMs), such as GPT-4, has enabled real-time text generation and summarization of medical data. Further, recent advancements have enabled the integration of data analysis features in chatbots, such that raw data can be uploaded and analyzed when prompted. Studying both the accuracy and suitability of LLM-derived data analysis performed on medical time series data, such as CGM data, is an important area of research. The objective of this study was to assess the strengths and limitations of using an LLM to analyze raw CGM data and produce summaries of 14 days of data for patients with type 1 diabetes. This study used simulated CGM data from 10 different cases. We first evaluated the ability of GPT-4 to compute quantitative metrics specific to diabetes found in an Ambulatory Glucose Profile (AGP). Then, using two independent clinician graders, we evaluated the accuracy, completeness, safety, and suitability of qualitative descriptions produced by GPT-4 across five different CGM analysis tasks. We demonstrated that GPT-4 performs well across measures of accuracy, completeness, and safety when producing summaries of CGM data across all tasks. These results highlight the capabilities of using an LLM to produce accurate and safe narrative summaries of medical time series data. We highlight several limitations of the work, including concerns related to how GPT-4 may misprioritize highlighting instances of hypoglycemia and hyperglycemia. Our work serves as a preliminary study on how generative language models can be integrated into diabetes care through CGM analysis, and more broadly, the potential to leverage LLMs for streamlined medical time series analysis.

Statistical Packages and Algorithms for the Analysis of Continuous Glucose Monitoring Data: A Systematic Review

BACKGROUND

Continuous glucose monitoring (CGM) measures glucose levels every 1 to 15 minutes and is widely used in clinical and research contexts. Statistical packages and algorithms reduce the time-consuming and error-prone process of manually calculating CGM metrics and contribute to standardizing CGM metrics defined by international consensus. The aim of this systematic review is to summarize existing data on (1) statistical packages for retrospective CGM data analysis and (2) statistical algorithms for retrospective CGM analysis not available in these statistical packages.

METHODS

A systematic literature search in PubMed and EMBASE was conducted on September 19, 2023. We also searched Google Scholar and Google Search until October 12, 2023 as sources of gray literature and performed reference checks of the included literature. Articles in English and Danish were included. This systematic review is registered with PROSPERO (CRD42022378163).

RESULTS

A total of 8731 references were screened and 46 references were included. We identified 23 statistical packages for the analysis of CGM data. The statistical packages could calculate many metrics of the 2022 CGM consensus and non-consensus CGM metrics, and 22/23 (96%) statistical packages were freely available. Also, 23 statistical algorithms were identified. The statistical algorithms could be divided into three groups based on content: (1) CGM data reduction (eg, clustering of CGM data), (2) composite CGM outcomes, and (3) other CGM metrics.

CONCLUSION

This systematic review provides detailed tabular and textual up-to-date descriptions of the contents of statistical packages and statistical algorithms for retrospective analysis of CGM data.

OGTT Metrics Surpass Continuous Glucose Monitoring Data for T1D Prediction in Multiple-Autoantibody-Positive Individuals

CONTEXT

The value of continuous glucose monitoring (CGM) for monitoring autoantibody (AAB)-positive individuals in clinical trials for progression of type 1 diabetes (T1D) is unknown.

OBJECTIVE

Compare CGM with oral glucose tolerance test (OGTT)-based metrics in prediction of T1D.

METHODS

At academic centers, OGTT and CGM data from multiple-AAB relatives were evaluated for associations with T1D diagnosis. Participants were multiple-AAB-positive individuals in a TrialNet Pathway to Prevention (TN01) CGM ancillary study (n = 93). The intervention was CGM for 1 week at baseline, 6 months, and 12 months. Receiver operating characteristic (ROC) curves of CGM and OGTT metrics for prediction of T1D were analyzed.

RESULTS

Five of 7 OGTT metrics and 29/48 CGM metrics but not HbA1c differed between those who subsequently did or did not develop T1D. ROC area under the curve (AUC) of individual CGM values ranged from 50% to 69% and increased when adjusted for age and AABs. However, the highest-ranking metrics were derived from OGTT: 4/7 with AUC ∼80%. Compared with adjusted multivariable models using CGM data, OGTT-derived variables, Index60 and DPTRS (Diabetes Prevention Trial-Type 1 Risk Score), had higher discriminative ability (higher ROC AUC and positive predictive value with similar negative predictive value).

CONCLUSION

Every 6-month CGM measures in multiple-AAB-positive individuals are predictive of subsequent T1D, but less so than OGTT-derived variables. CGM may have feasibility advantages and be useful in some settings. However, our data suggest there is insufficient evidence to replace OGTT measures with CGM in the context of clinical trials.

A case study of glucose levels during sleep using multilevel fast function on scalar regression inference

Continuous glucose monitors (CGMs) are increasingly used to measure blood glucose levels and provide information about the treatment and management of diabetes. Our motivating study contains CGM data during sleep for 174 study participants with type II diabetes mellitus measured at a 5-min frequency for an average of 10 nights. We aim to quantify the effects of diabetes medications and sleep apnea severity on glucose levels. Statistically, this is an inference question about the association between scalar covariates and functional responses observed at multiple visits (sleep periods). However, many characteristics of the data make analyses difficult, including (1) nonstationary within-period patterns; (2) substantial between-period heterogeneity, non-Gaussianity, and outliers; and (3) large dimensionality due to the number of study participants, sleep periods, and time points. For our analyses, we evaluate and compare two methods: fast univariate inference (FUI) and functional additive mixed models (FAMMs). We extend FUI and introduce a new approach for testing the hypotheses of no effect and time invariance of the covariates. We also highlight areas for further methodological development for FAMM. Our study reveals that (1) biguanide medication and sleep apnea severity significantly affect glucose trajectories during sleep and (2) the estimated effects are time invariant.

Hypoglycaemia induces a sustained pro-inflammatory response in people with type 1 diabetes and healthy controls

AIM

To determine the duration and the extension of the pro-inflammatory response to hypoglycaemia both in people with type 1 diabetes and healthy controls.

MATERIALS AND METHODS

Adults with type 1 diabetes (n = 47) and matched controls (n = 16) underwent a hyperinsulinaemic-euglycaemic hypoglycaemic (2.8 ± 0.1 mmoL/L [49.9 ± 2.3 mg/dL]) glucose clamp. During euglycaemia, hypoglycaemia, and 1, 3 and 7 days later, blood was drawn to determine immune cell phenotype, monocyte function and circulating inflammatory markers.

RESULTS

Hypoglycaemia increased lymphocyte and monocyte counts, which remained elevated for 1 week. The proportion of CD16+ monocytes increased and the proportion of CD14+ monocytes decreased. During hypoglycaemia, monocytes released more tumour necrosis factor-α and interleukin-1β, and less interleukin-10, after ex vivo stimulation. Hypoglycaemia increased the levels of 19 circulating inflammatory proteins, including high sensitive C-reactive protein, most of which remained elevated for 1 week. The epinephrine peak in response to hypoglycaemia was positively correlated with immune cell number and phenotype, but not with the proteomic response.

CONCLUSIONS

Overall, despite differences in prior exposure to hypoglycaemia, the pattern of the inflammatory responses to hypoglycaemia did not differ between people with type 1 diabetes and healthy controls. In conclusion, hypoglycaemia induces a range of pro-inflammatory responses that are sustained for at least 1 week in people with type 1 diabetes and healthy controls.

Investigating the value of glucodensity analysis of continuous glucose monitoring data in type 1 diabetes: an exploratory analysis

Introduction

Continuous glucose monitoring (CGM) devices capture longitudinal data on interstitial glucose levels and are increasingly used to show the dynamics of diabetes metabolism. Given the complexity of CGM data, it is crucial to extract important patterns hidden in these data through efficient visualization and statistical analysis techniques.

Methods

In this paper, we adopted the concept of glucodensity, and using a subset of data from an ongoing clinical trial in pediatric individuals and young adults with new-onset type 1 diabetes, we performed a cluster analysis of glucodensities. We assessed the differences among the identified clusters using analysis of variance (ANOVA) with respect to residual pancreatic beta-cell function and some standard CGM-derived parameters such as time in range, time above range, and time below range.

Results

Distinct CGM data patterns were identified using cluster analysis based on glucodensities. Statistically significant differences were shown among the clusters with respect to baseline levels of pancreatic beta-cell function surrogate (C-peptide) and with respect to time in range and time above range.

Discussion

Our findings provide supportive evidence for the value of glucodensity in the analysis of CGM data. Some challenges in the modeling of CGM data include unbalanced data structure, missing observations, and many known and unknown confounders, which speaks to the importance of--and provides opportunities for--taking an approach integrating clinical, statistical, and data science expertise in the analysis of these data.

Continuous glucose monitoring in patients with post-bariatric hypoglycaemia reduces hypoglycaemia and glycaemic variability

AIM

To determine whether continuous glucose monitoring (CGM) can reduce hypoglycaemia in patients with post-bariatric hypoglycaemia (PBH).

MATERIALS AND METHODS

In an open-label, nonrandomized, pre-post design with sequential assignment, CGM data were collected in 22 individuals with PBH in two sequential phases: (i) masked (no access to sensor glucose or alarms); and (ii) unmasked (access to sensor glucose and alarms for low or rapidly declining sensor glucose). Twelve participants wore the Dexcom G4 device for a total of 28 days, while 10 wore the Dexcom G6 device for a total of 20 days.

RESULTS

Participants with PBH spent a lower percentage of time in hypoglycaemia over 24 hours with unmasked versus masked CGM (<3.3 mM/L, or <60 mg/dL: median [median absolute deviation {MAD}] 0.7 [0.8]% vs. 1.4 [1.7]%, P = 0.03; <3.9 mM/L, or <70 mg/dL: median [MAD] 2.9 [2.5]% vs. 4.7 [4.8]%; P = 0.04), with similar trends overnight. Sensor glucose data from the unmasked phase showed a greater percentage of time spent between 3.9 and 10 mM/L (70-180 mg/dL) (median [MAD] 94.8 [3.9]% vs. 90.8 [5.2]%; P = 0.004) and lower glycaemic variability over 24 hours (median [MAD] mean amplitude of glycaemic excursion 4.1 [0.98] vs. 4.4 [0.99] mM/L; P = 0.04). During the day, participants also spent a greater percentage of time in normoglycaemia with unmasked CGM (median [MAD] 94.2 [4.8]% vs. 90.9 [6.2]%; P = 0.005), largely due to a reduction in hyperglycaemia (>10 mM/L, or 180 mg/dL: median [MAD] 1.9 [2.2]% vs. 3.9 [3.6]%; P = 0.02).

CONCLUSIONS

Real-time CGM data and alarms are associated with reductions in low sensor glucose, elevated sensor glucose, and glycaemic variability. This suggests CGM allows patients to detect hyperglycaemic peaks and imminent hypoglycaemia, allowing dietary modification and self-treatment to reduce hypoglycaemia. The use of CGM devices may improve safety in PBH, particularly for patients with hypoglycaemia unawareness.

CGMap: Characterizing continuous glucose monitor data in thousands of non-diabetic individuals

Despite its rising prevalence, diabetes diagnosis still relies on measures from blood tests. Technological advances in continuous glucose monitoring (CGM) devices introduce a potential tool to expand our understanding of glucose control and variability in people with and without diabetes. Yet CGM data have not been characterized in large-scale healthy cohorts, creating a lack of reference for CGM data research. Here we present CGMap, a characterization of CGM data collected from over 7,000 non-diabetic individuals, aged 40-70 years, between 2019 and 2022. We provide reference values of key CGM-derived clinical measures that can serve as a tool for future CGM research. We further explored the relationship between CGM-derived measures and diabetes-related clinical parameters, uncovering several significant relationships, including associations of mean blood glucose with measures from fundus imaging and sleep monitoring. These findings offer novel research directions for understanding the influence of glucose levels on various aspects of human health.

Factors associated with hemoglobin glycation index in adults with type 1 diabetes mellitus: The FGM-Japan study

AIMS/INTRODUCTION

The discrepancy between HbA1c and glucose exposure may have significant clinical implications; however, the association between the hemoglobin glycation index (HGI) and clinical parameters in type 1 diabetes remains controversial. This study aimed to find the factors associated with HGI (laboratory HbA1c - predicted HbA1c derived from the continuous glucose monitoring [CGM]).

MATERIALS AND METHODS

We conducted a cross-sectional study of adults with type 1 diabetes (n = 211, age 50.9 ± 15.2 years old, female sex = 59.2%, duration of CGM use = 2.1 ± 1.0 years). All subjects wore the CGM for 90 days before HbA1c measurement. Data derived from the FreeStyle Libre sensor were used to calculate the glucose management indicator (GMI) and glycemic variability (GV) parameters. HGI was defined as the difference between the GMI and the laboratory HbA1c levels. The participants were divided into three groups according to the HGI tertile (low, moderate, and high). Multivariate regression analyses were performed.

RESULTS

The female sex ratio, HbA1c, and % coefficient of variation (%CV) significantly increased over the HGI tertile, while eGFR and Hb decreased over the HGI tertile. In multivariate analysis, the factors associated with HGI were %CV and eGFR, after adjusting for HbA1c level and sex (R²  = 0.44).

CONCLUSIONS

This study demonstrated that HGI is associated with female sex, eGFR, and some glycemic variability indices, independently of HbA1c. Minimizing glycemic fluctuations might reduce HGI. This information provides diabetic health professionals and patients with personalized diabetes management for adults with type 1 diabetes.

Validation of a Hypoglycemia Risk Stratification Tool Using Data From Continuous Glucose Monitors

This cohort study uses data from continuous glucose monitoring to validate a hypoglycemia risk stratification tool.

Cardiometabolic health improvements upon dietary intervention are driven by tissue-specific insulin resistance phenotype: A precision nutrition trial

Precision nutrition based on metabolic phenotype may increase the effectiveness of interventions. In this proof-of-concept study, we investigated the effect of modulating dietary macronutrient composition according to muscle insulin-resistant (MIR) or liver insulin-resistant (LIR) phenotypes on cardiometabolic health. Women and men with MIR or LIR (n = 242, body mass index [BMI] 25-40 kg/m², 40-75 years) were randomized to phenotype diet (PhenoDiet) group A or B and followed a 12-week high-monounsaturated fatty acid (HMUFA) diet or low-fat, high-protein, and high-fiber diet (LFHP) (PhenoDiet group A, MIR/HMUFA and LIR/LFHP; PhenoDiet group B, MIR/LFHP and LIR/HMUFA). PhenoDiet group B showed no significant improvements in the primary outcome disposition index, but greater improvements in insulin sensitivity, glucose homeostasis, serum triacylglycerol, and C-reactive protein compared with PhenoDiet group A were observed. We demonstrate that modulating macronutrient composition within the dietary guidelines based on tissue-specific insulin resistance (IR) phenotype enhances cardiometabolic health improvements. Clinicaltrials.gov registration: NCT03708419, CCMO registration NL63768.068.17.

Continuous Glucose Monitoring Time Series Data Analysis: A Time Series Analysis Package for Continuous Glucose Monitoring Data

The R package Continuous Glucose Monitoring Time Series Data Analysis (CGMTSA) was developed to facilitate investigations that examine the continuous glucose monitoring (CGM) data as a time series. Accordingly, novel time series functions were introduced to (1) enable more accurate missing data imputation and outlier identification; (2) calculate recommended CGM metrics as well as key time series parameters; (3) plot interactive and three-dimensional graphs that allow direct visualizations of temporal CGM data and time series model optimization. The software was designed to accommodate all popular CGM devices and support all common data processing steps. The program is available for Linux, Windows, and Mac at GitHub.

Software Packages and Tools for the Analysis of Continuous Glucose Monitoring Data

The advancement of technology in the field of glycemic control has led to the widespread use of continuous glucose monitoring (CGM), which can be nowadays obtained from wearable devices equipped with a minimally invasive sensor, that is, transcutaneous needle type or implantable, and a transmitter that sends information to a receiver or smart device for data storage and display. This work aims to review the currently available software packages and tools for the analysis of CGM data. Based on the purposes of this work, 12 software packages have been identified from the literature, published until December 2021, namely: GlyCulator, EasyGV (Easy Glycemic Variability), CGM-GUIDE^© (Continuous Glucose Monitoring Graphical User Interface for Diabetes Evaluation), GVAP (Glycemic Variability Analyzer Program), Tidepool, CGManalyzer, cgmanalysis, GLU, CGMStatsAnalyser, iglu, rGV, and cgmquantify. Comparison of available software packages and tools has been done in terms of main characteristics (i.e., publication year, presence of a graphical user interface, availability, open-source code, number of citations, programming language, supported devices, supported data format and organization of the data structure, documentation, presence of a toy example, video tutorial, data upload and download, measurement-units conversion), preprocessing procedures, data display options, and computed metrics; also, each of the computed metrics has been analyzed in terms of its adherence to the American Diabetes Association (ADA) 2017 international consensus on CGM data analysis and the ADA 2019 international consensus on time in range. Eventually, the agreement between metrics computed by different software and tools has been investigated. Based on such comparison, usability and complexity of data management, as well as the possibility to perform customized or patients-group analyses, have been discussed by highlighting limitations and strengths, also in relation to possible different user categories (i.e., patients, clinicians, researchers). The information provided could be useful to researchers interested in working in the diabetic research field as to clinicians and endocrinologists who need tools capable of handling CGM data effectively.

Comprehensive self-tracking of blood glucose and lifestyle with a mobile application in the management of gestational diabetes: a study protocol for a randomised controlled trial (eMOM GDM study)

INTRODUCTION

Gestational diabetes (GDM) causes various adverse short-term and long-term consequences for the mother and child, and its incidence is increasing globally. So far, the most promising digital health interventions for GDM management have involved healthcare professionals to provide guidance and feedback. The principal aim of this study is to evaluate the effects of comprehensive and real-time self-tracking with eMOM GDM mobile application (app) on glucose levels in women with GDM, and more broadly, on different other maternal and neonatal outcomes.

METHODS AND ANALYSIS

This randomised controlled trial is carried out in Helsinki metropolitan area. We randomise 200 pregnant women with GDM into the intervention and the control group at gestational week (GW) 24-28 (baseline, BL). The intervention group receives standard antenatal care and the eMOM GDM app, while the control group will receive only standard care. Participants in the intervention group use the eMOM GDM app with continuous glucose metre (CGM) and activity bracelet for 1 week every month until delivery and an electronic 3-day food record every month until delivery. The follow-up visit after intervention takes place 3 months post partum for both groups. Data are collected by laboratory blood tests, clinical measurements, capillary glucose measures, wearable sensors, air displacement plethysmography and digital questionnaires. The primary outcome is fasting plasma glucose change from BL to GW 35-37. Secondary outcomes include, for example, self-tracked capillary fasting and postprandial glucose measures, change in gestational weight gain, change in nutrition quality, change in physical activity, medication use due to GDM, birth weight and fat percentage of the child.

ETHICS AND DISSEMINATION

The study has been approved by Ethics Committee of the Helsinki and Uusimaa Hospital District. The results will be presented in peer-reviewed journals and at conferences.

TRIAL REGISTRATION NUMBER

NCT04714762.

A New Analysis Tool for Continuous Glucose Monitor Data

BACKGROUND

With the development of continuous glucose monitoring systems (CGMS), detailed glycemic data are now available for analysis. Yet analysis of this data-rich information can be formidable. The power of CGMS-derived data lies in its characterization of glycemic variability. In contrast, many standard glycemic measures like hemoglobin A1c (HbA1c) and self-monitored blood glucose inadequately describe glycemic variability and run the risk of bias toward overreporting hyperglycemia. Methods that adjust for this bias are often overlooked in clinical research due to difficulty of computation and lack of accessible analysis tools.

METHODS

In response, we have developed a new R package rGV, which calculates a suite of 16 glycemic variability metrics when provided a single individual's CGM data. rGV is versatile and robust; it is capable of handling data of many formats from many sensor types. We also created a companion R Shiny web app that provides these glycemic variability analysis tools without prior knowledge of R coding. We analyzed the statistical reliability of all the glycemic variability metrics included in rGV and illustrate the clinical utility of rGV by analyzing CGM data from three studies.

RESULTS

In subjects without diabetes, greater glycemic variability was associated with higher HbA1c values. In patients with type 2 diabetes mellitus (T2DM), we found that high glucose is the primary driver of glycemic variability. In patients with type 1 diabetes (T1DM), we found that naltrexone use may potentially reduce glycemic variability.

CONCLUSIONS

We present a new R package and accompanying web app to facilitate quick and easy computation of a suite of glycemic variability metrics.

Glycemic Variability Patterns Strongly Correlate With Partial Remission Status in Children With Newly Diagnosed Type 1 Diabetes

OBJECTIVE

To evaluate whether indexes of glycemic variability may overcome residual β-cell secretion estimates in the longitudinal evaluation of partial remission in a cohort of pediatric patients with new-onset type 1 diabetes.

RESEARCH DESIGN AND METHODS

Values of residual β-cell secretion estimates, clinical parameters (e.g., HbA1c or insulin daily dose), and continuous glucose monitoring (CGM) from 78 pediatric patients with new-onset type 1 diabetes were longitudinally collected during 1 year and cross-sectionally compared. Circadian patterns of CGM metrics were characterized and correlated to remission status using an adjusted mixed-effects model. Patients were clustered based on 46 CGM metrics and clinical parameters and compared using nonparametric ANOVA.

RESULTS

Study participants had a mean (± SD) age of 10.4 (± 3.6) years at diabetes onset, and 65% underwent partial remission at 3 months. β-Cell residual secretion estimates demonstrated weak-to-moderate correlations with clinical parameters and CGM metrics (r2 = 0.05-0.25; P < 0.05). However, CGM metrics strongly correlated with clinical parameters (r2 >0.52; P < 0.05) and were sufficient to distinguish remitters from nonremitters. Also, CGM metrics from remitters displayed specific early morning circadian patterns characterized by increased glycemic stability across days (within 63-140 mg/dL range) and decreased rate of grade II hypoglycemia (P < 0.0001) compared with nonremitters. Thorough CGM analysis allowed the identification of four novel glucotypes (P < 0.001) that segregate patients into subgroups and mirror the evolution of remission after diabetes onset.

CONCLUSIONS

In our pediatric cohort, combination of CGM metrics and clinical parameters unraveled key clinical milestones of glucose homeostasis and remission status during the first year of type 1 diabetes.

Exploring Minimally Invasive Approach to Define Stages of Type 1 Diabetes Remotely

Objective: New methods are pivotal in accurately predicting, monitoring, and diagnosing the clinical manifestation of type 1 diabetes (T1D) in high-risk children. Continuous glucose monitoring (CGM) is a valuable tool for patients with T1D, but there is still a knowledge gap regarding its utility in the prediction of diabetes. The current study explored whether 10-day CGM or CGM during an oral glucose tolerance test (OGTT) performed in the laboratory or at home (home-OGTT) could be accurate in detecting stages of T1D. Research Design and Methods: Forty-six subjects 4-25 years of age carrying genetic risk for T1D were recruited and classified into the following groups: islet autoantibody (IAb) negative, one IAb, and stages 1-3 of T1D, based on the laboratory OGTT and IAb results at baseline. A 10-day CGM was initiated before the OGTT. Results: In this study, we showed that CGM was sensitive in detecting asymptomatic individuals at stage 3, and dysglycemic individuals in stage 2 of T1D both during OGTT and the 10-day period. CGM also showed significant differences in several variables during the 10-day sensoring among individuals at different stages of T1D. Furthermore, CGM showed different OGTT profiles and detected significantly more abnormal OGTT results when compared with plasma glucose. Conclusions: CGM together with home-OGTT could detect stages of T1D and offer an alternative method to confirm normoglycemia in high-risk individuals.

Comparison of Cgmanalysis, a Free Open-Source Continuous Glucose Monitoring Data Management and Analysis Software, with Commercially Available CGM Platforms: Data Standardization for Diabetes Technology Research

Background: Cgmanalysis is an open-source software based on the R programming language for data management and descriptive analysis of data from continuous glucose monitoring (CGM). We sought to validate the summary measures calculated by cgmanalysis against the results from proprietary software associated with four CGM commercially available models. Methods: Two weeks of data from 188 patients with type 1 diabetes using commercially available CGMs. Freestyle Libre Gen 1 (n = 53), Medtronic Guardian 3 (n = 52), Dexcom G6 reported by Dexcom Clarity (n = 48), and Dexcom G6 reported by Tandem (n = 35) were analyzed using proprietary software and cgmanalysis. Agreement was assessed using scatterplots, Bland-Altman plots, and equivalence tests. Results: Good agreement was obtained for all glycemic summary measures for all CGMs assessed. None of the differences between the cgmanalysis package and the manufacturers' software were outside the prespecified bounds of equivalence. Conclusions: Cgmanalysis is a validated open-source software to analyze commercially available CGM data and can be used to standardize diabetes technology research.

A High Protein Diet Is More Effective in Improving Insulin Resistance and Glycemic Variability Compared to a Mediterranean Diet-A Cross-Over Controlled Inpatient Dietary Study

The optimal dietary pattern to improve metabolic function remains elusive. In a 21-day randomized controlled inpatient crossover feeding trial of 20 insulin-resistant obese women, we assessed the extent to which two isocaloric dietary interventions—Mediterranean (M) and high protein (HP)—improved metabolic parameters. Obese women were assigned to one of the following dietary sequences: M–HP or HP–M. Cardiometabolic parameters, body weight, glucose monitoring and gut microbiome composition were assessed. Sixteen women completed the study. Compared to the M diet, the HP diet was more effective in (i) reducing insulin resistance (insulin: Beta (95% CI) = −6.98 (−12.30, −1.65) µIU/mL, p = 0.01; HOMA-IR: −1.78 (95% CI: −3.03, −0.52), p = 9 × 10⁻³); and (ii) improving glycemic variability (−3.13 (−4.60, −1.67) mg/dL, p = 4 × 10⁻⁴), a risk factor for T2D development. We then identified a panel of 10 microbial genera predictive of the difference in glycemic variability between the two diets. These include the genera Coprococcus and Lachnoclostridium, previously associated with glucose homeostasis and insulin resistance. Our results suggest that morbidly obese women with insulin resistance can achieve better control of insulin resistance and glycemic variability on a high HP diet compared to an M diet.