Glycemic Risk Assessment Diabetes Equation Research Articles

Development and Validation of a Machine Learning Model to Predict Weekly Risk of Hypoglycemia in Patients with Type 1 Diabetes Based on Continuous Glucose Monitoring.

Aim: The aim of this study was to develop and validate a prediction model based on continuous glucose monitoring (CGM) data to identify a week-to-week risk profile of excessive hypoglycemia. Methods: We analyzed, trained, and internally tested two prediction models using CGM data from 205 type 1 diabetes patients with long-term CGM monitoring. A binary classification approach (XGBoost) combined with feature engineering deployed on the CGM signals was utilized to predict excessive hypoglycemia risk defined by two targets (time below range [TBR] >4% and the upper TBR 90th percentile limit) of TBR the following week. The models were validated in two independent cohorts with a total of 253 additional patients. Results: A total of 61,470 weeks of CGM data were included in the analysis. The XGBoost models had an area under the receiver operating characteristic curve (ROC-AUC) of 0.83-0.87 (95% confidence interval; 0.83-0.88) in the test dataset. The external validation showed ROC-AUCs of 0.81-0.90. The most discriminative features included the low blood glucose index, the glycemic risk assessment diabetes equation (GRADE), hypoglycemia, the TBR, waveform length, the coefficient of variation and mean glucose during the previous week. This highlights that the pattern of hypoglycemia combined with glucose variability during the past week contains information on the risk of future hypoglycemia. Conclusion: Prediction models based on real-world CGM data can be used to predict the risk of hypoglycemia in the forthcoming week. The models showed good performance in both the internal and external validation cohorts.

The Relationship Between Dietary Patterns and Glycemic Variability in People with Impaired Glucose Tolerance

BackgroundDiurnal glucose fluctuations are increased in prediabetes and might be affected by specific dietary patterns. ObjectivesThe present study assessed the relationship between glycemic variability (GV) and dietary regimen in people with normal glucose tolerance (NGT) and impaired glucose tolerance (IGT). MethodsForty-one NGT (mean age: 45.0 ± 9.0 y, mean BMI: 32.0 ± 7.0 kg/m2) and 53 IGT (mean age: 48.4 ± 11.2 y, mean BMI: 31.3 ± 5.9 kg/m2) subjects were enrolled in this cross-sectional study. The FreeStyleLibre Pro sensor was used for 14 d, and several parameters of GV were calculated. The participants were provided with a diet diary to record all meals. ANOVA analysis, Pearson correlation, and stepwise forward regression were performed. ResultsDespite no difference in diet patterns between the 2 groups, GV parameters were higher in IGT than in NGT. GV worsened with an increase in overall daily carbohydrate and refined grain consumption and improved with the increase in whole grain intake in IGT. GV parameters were positively related [r = 0.14–0.53; all P < 0.02 for SD, continuous overall net glycemic action 1 (CONGA1), J-index, lability index (LI), glycemic risk assessment diabetes equation, M-value, and mean absolute glucose (MAG)], and low blood glucose index (LBGI) inversely (r = −0.37, P = 0.006) related to the total percentage of carbohydrate, but not to the distribution of carbohydrate between the main meals in the IGT group. A negative relationship existed between total protein consumption and GV indices (r = −0.27 to −0.52; P < 0.05 for SD, CONGA1, J-index, LI, M-value, and MAG). The total EI was related to GV parameters (r = 0.27–0.32; P < 0.05 for CONGA1, J-index, LI, and M-value; and r = −0.30, P = 0.028 for LBGI). ConclusionsThe primary outcome results showed that insulin sensitivity, calories, and carbohydrate content are predictors of GV in individuals with IGT. Overall, the secondary analyses suggested that carbohydrate and daily consumption of refined grains might be associated with higher GV, whereas whole grains and daily protein intake were related to lower GV in people with IGT.

710-P: In Patients with Type 2 Diabetes Mellitus (T2DM) , Continuous Glucose Monitoring System (CGMS) Usage Does Not Alter Multiple Measures of Glycemic Variability (GV)

Background: The RCT data on CGMS and GV in patients with T2DM have used limited measure of GV metrics. In patients with T2DM on CGMS, we examined whether the use of blinded or unblinded CGMS will affect their GV across a broad swatch of metrics. Methods: A randomized control study involving 40 subjects with T2DM was done [50% females, mean (SD) age 59 years (12.5) , BMI 34.kg/m2 (8.1) , mean (SD) screening HbA1c 7.59% (1.2) ]. Patients were randomized (1:1) to either 1 unblinded (Freestyle Libre) +1 blinded (Freestyle Libre Pro) CGMs or 2 blinded CGMS (Freestyle Libre Pro) for 16 weeks. Sensors were replaced every 2 weeks with an in-person visit. Multiple GV metrics were calculated from the blinded CGM at baseline (Week 0-Week2) and end-intervention (Week 14-16) . Participants randomized to the Unblinded/Blinded CGM group were instructed to use the CGM as they wished. Participants randomized to the Blinded CGM group were instructed to continue checking their fingerstick glucose as usual. Measured GV metrics included the following: mean glucose, standard deviation (SD) , and coefficient of variation (CV) , continuous overlapping net glycemic action (CONGA) , liability Index, J-index, low and high blood glucose index (LBGI/HBGI) , glycemic risk assessment diabetes equation (GRADE) , mean of daily differences (MODD) , average daily risk range (ADRR) , mean amplitude of glycemic excursions (MAGE) , mean of daily differences (MODD) , Area Under Curve (AUC) , time spent in range (TIR) , Glucose Management Indicator (GMI) , number of episodes below 55 and glycemic variability percentage (GVP) . Results: In this study, we found that was no significant difference in all GV metrics between the blinded/ blinded CGMS group and unblinded/blinded CGMS group at baseline and end intervention. Conclusion: The novelty of this study is the extensive analysis of GV across multiple metrics in patients with T2DM on CGMS and CGMS usage did not significantly alter GV. Disclosure Y.Lee: None. A.Mehfooz: None. Q.Wang: None. L.S.Chow: Other Relationship; Dexcom, Inc.

699-P: Time-Restricted Eating Did Not Alter Glycemic Variability in Humans Who Are Overweight and without Diabetes

Background: Time-restricted eating (TRE) is increasingly popular for weight loss, yet its effects on glycemic measures remain unclear. Using continuous glucose monitor (CGM) worn for 2 weeks at baseline and end-intervention, we hypothesized that TRE may reduce glycemic variability compared to unrestricted eating (non-TRE) in patients who are overweight and without diabetes. Method: Participants (17 women and 3 men; mean [SD] age: 45.5 [12.1] years; BMI 34.1 [7.5] kg/m2) without self-reported diabetes (HbA1c 5.5 [0.4]) who had prolonged eating window (15.4 [0.9] hours) were randomized to a 12 week intervention of either TRE (n=11: 8-hour window, unrestricted eating within window) versus non-TRE (n=9: unrestricted eating) . Each participant wore a blinded CGM (Freestyle Libre Pro) for 2 weeks prior to randomization and at end-intervention (week 11-12) . At each time point, multiple measures of glycemic variability were calculated from the raw CGM data. This is a secondary analysis of a previously published study where 12 weeks of TRE did not alter HbA1c. Result: Between the TRE and non-TRE groups in the pre and end-intervention periods, there was no statistical difference in any of the calculated glycemic variability measures: including, time in range (TIR) , standard deviation (SD) , coefficient of variation (CV) , continuous overlapping net glycemic action (CONGA) , liability, Index, J-index, low and high blood glucose index (LBGI/HBGI) , glycemic risk assessment diabetes equation (GRADE) , mean of daily differences (MODD) , average daily risk range (ADRR) , mean amplitude of glycemic excursions (MAGE) , mean of daily differences (MODD) , Area Under Curve (AUC) , and Glucose Management Indicator (GMI) . Conclusion: In patients who are overweight without diabetes, TRE did not alter CGM-measured glycemic variability. Further studies regarding TRE effects on glycemic variability should focus on participants with diabetes. Disclosure A.Mehfooz: None. Y.Lee: None. Q.Wang: None. L.S.Chow: Other Relationship; Dexcom, Inc.

Continuous Glucose Monitoring for the Detection of Hypoglycemia in Patients With Diabetes of the Exocrine Pancreas.

Detailed evaluations of hypoglycemia and associated indices based on continuous glucose monitoring (CGM) are limited in patients with diabetes of the exocrine pancreas. Our study sought to evaluate the frequency and pattern of hypoglycemic events and to investigate hypoglycemia-specific indices in this population. This was a cross-sectional study comprising 83 participants with diabetes of the exocrine pancreas. CGM and self-monitoring of blood glucose (SMBG) were performed on all participants for a minimum period of 72 hours. The frequency and pattern of hypoglycemic events, as well as hypoglycemia-related indices, were evaluated. Hypoglycemia was detected in 90.4% of patients using CGM and 38.5% of patients using SMBG. Nocturnal hypoglycemic events were more frequent (1.9 episodes/patient) and prolonged (142 minutes) compared with day-time events (1.1 episodes/patient; 82.8 minutes, P < 0.05). The mean low blood glucose index was 2.1, and glycemic risk assessment diabetes equation hypoglycemia was 9.1%. The mean time spent below (TSB) <70 mg/dL was 9.2%, and TSB <54 mg/dL was 3.7%. The mean area under curve (AUC) <70 mg/dL was 1.7 ± 2.5 mg/dL/hour and AUC <54 mg/dL was 0.6 ± 1.3 mg/dL/hour. All of the CGM-derived hypoglycemic indices were significantly more deranged at night compared with during the day (P < 0.05). Patients with diabetes of the exocrine pancreas have a high frequency of hypoglycemic episodes that are predominantly nocturnal. CGM is superior to SMBG in the detection of nocturnal and asymptomatic hypoglycemic episodes. CGM-derived hypoglycemic indices are beneficial in estimating the risk of hypoglycemia.

Comparison of CGM-Derived Measures of Glycemic Variability Between Pancreatogenic Diabetes and Type 2 Diabetes Mellitus.

To compare glycemic variability (GV) indices between patients with fibrocalculous pancreatic diabetes (FCPD) and type 2 diabetes mellitus (T2D) using continuous glucose monitoring (CGM). We measured GV indices using CGM (iPro™2 Professional CGM, Medtronic, USA) data in 61 patients each with FCPD and T2D who were matched for glycated hemoglobin A1c (HbA1c) and duration of diabetes. GlyCulator2 software was used to estimate the CGM-derived measures of GV (SD, mean amplitude of glycemic excursion [MAGE], continuous overall net glycemic action [CONGA], absolute means of daily differences [MODD], M value, and coefficient of variance [%CV]), hypoglycemia (time spent below 70 mg/dL, AUC below 70 mg/dL, glycemic risk assessment diabetes equation hypoglycemia, Low Blood Glucose Index), and hyperglycemia (time spent above 180 mg/dL at night [TSA > 180], AUC above 180 mg/dL [AUC > 180], glycemic risk assessment diabetes equation hyperglycemia, High Blood Glucose Index [HBGI], and J index). The correlation of GV indices with HbA1c, duration of diabetes, and demographic and biochemical parameters were also assessed. All the CGM-derived measures of GV (SD, MAGE, CONGA, MODD, and %CV), except M value, were significantly higher in the FCPD group than in the T2D group (P < 0.05). Measures of hyperglycemia (TSA >180, AUC >180, HBGI, and J index) were significantly higher in the FCPD group than in the T2D group (P < 0.05). The measures of hypoglycemia were not significantly different between the two groups. All the hyperglycemia indices showed a positive correlation with HbA1c in both groups. FCPD is associated with higher GV than is T2D. The findings of higher postprandial glycemic excursions in patients with FCPD could have potential therapeutic implications.

Evolution of Glycemic Control and Variability After Kidney Transplant.

The evolution of glycemic changes after kidney transplantation has not been described. We prospectively examined glycemic control and variability over time from transplantation using continuous glucose monitoring (CGM). Continuous glucose monitoring devices were fitted for 3 to 5 days at time of transplant, month 3, and month 6 posttransplant. Indices of glucose control (mean glucose, percent time in hyperglycemic range, and Glycemic Risk Assessment Diabetes Equation score) and variability were calculated. An oral glucose tolerance test was performed at month 3. Twenty-eight patients (mean age, 45 ± 15 years) were enrolled, 64% male, 75% white, receiving tacrolimus, mycophenolate, and prednisolone (93%). Of 24 patients with complete CGM data at month 0, 3 had prior diabetes and 6 (25%) developed new-onset diabetes after transplant (NODAT). Hyperglycemia (>11.1 mM) was evident in 79% during days 0 to 3 posttransplant, particularly between 1 and 9 PM. Compared with recipients without diabetes, recipients with prior diabetes had higher mean glucose (7.8 mM; 95% confidence interval [CI], 7.4-8.2 vs 9.9 mM; 95% CI, 8.9-10.8; P < 0.001), Glycemic Risk Assessment Diabetes Equation (GRADE) score (4.5; 95% CI, 3.7-5.4 vs 7.8; 95% CI, 5.6-10.4; P = 0.003) and percent time with hyperglycemia. Glycemic control was also inferior in those that subsequently developed NODAT (mean glucose, 8.8 mM; 95% CI, 8.2-9.4; P = 0.004, GRADE: 6.2, 95% CI, 5.2-7.7; P = 0.04 vs no diabetes). Glucose variability was increased in patients with prior diabetes (glucose standard deviation, 1.99; 95% CI, 1.72-2.27 vs 2.97; 95% CI, 2.27-3.67; P = 0.006) but not in NODAT. All measures of glucose control and variability significantly improved over time after transplantation (P < 0.001). Dysglycemia is very common after renal transplantation, exhibiting a distinct diurnal pattern of afternoon and evening hyperglycemia. The magnitude of hyperglycemia and variability are maximal in recipients with preexisting diabetes and significant in those who go on to develop NODAT. Dysglycemia improves with time.

Continuous Glucose Monitoring Glycemic Variability Profiles Are More Favorable for Dapagliflozin plus Saxagliptin Compared with Insulin Glargine in Type 2 Diabetes

Glycemic excursions, variability, and hypoglycemia may impact therapeutic effectiveness even when A1C reduction is comparable between therapies. We analyzed masked continuous glucose monitoring (CGM) data from patients with type 2 diabetes (T2D) inadequately controlled on metformin ± sulfonylurea during an international randomized, 24-week, open-label, non-inferiority trial of dapagliflozin plus saxagliptin add-on (DAPA + SAXA, n = 141) compared to titrated insulin glargine add-on (INS, n = 142). Six-day CGM (288 readings/day), A1C and patient-reported outcomes (PRO) were measured at baseline and Weeks 2, 12 and 24. At baseline, patients had mean ± SD age = 55.2 ± 9.4 years; A1C = 9.1 ± 1.1%; diabetes duration = 9.8 ± 6.1 years; 54.1% male. Using linear mixed models, mean ± SE changes in A1C for DAPA + SAXA vs. INS at Week 24 were comparable (−1.75 ± 0.09% vs. −1.63 ± 0.09%, P = 0.33). However, CGM summary analytics were more favorable for DAPA + SAXA: more time in glycemic range, lower glucose variability and less time with nocturnal hypoglycemia (Table 1a). Improvements in CGM at Week 24 correlated with better patient satisfaction and quality of life (n = 225, Table 1b). This analysis of CGM data showed superior diurnal glycemic profiles associated with better PRO for add-on DAPA + SAXA vs. INS in patients with T2D, even though A1C changes were comparable.Table 1a: Treatment Differences in CGM Summary Analytics --- Table 1b: Correlations of CGM with PROTable 1a. CGM Change from Baseline [Mean (SE)]DAPA + SAXA + MetforminINS + MetforminDifferenceP-valueTable 1b. Correlation of Change in CGM with Change in PRO at Week 24 P-valueMean 24-hr glucose mg/dL - Week 2−48.53 (2.51)−28.54 (2.51)−19.99 (3.55)&amp;lt; 0.0001Mean 24-hr glucose mg/dLSatisfaction&amp;lt;0.0001% time glucose ≤ 70 mg/dL - Week 120.52 (0.38)2.29 (0.39)−1.77 (0.54)0.001% time &amp;gt; 70 &amp; ≤ 180 mg/dLSatisfaction&amp;lt;0.0001% time glucose ≤ 70 mg/dL 12 to 6 am - Week 240.60 (0.53)2.66 (0.53)−2.(0.75)0.007Glycemic Risk Assessment Diabetes Equation (GRADE)Satisfaction&amp;lt;0.0001% time glucose &amp;gt; 70 and ≤ 180 mg/dL - Week 2434.28 (1.89)28.50 (1.91)5.78 (2.69)0.033AUC &amp;gt; 180 mg/dLSatisfaction0.001Mean Amplitude Glycemic Excursion (MAGE) mg/dL - Week 24−12.73 (2.26)−5.01 (2.28)−7.72 (3.21)0.017High Blood Glucose Index (HBGI)Satisfaction0.001Within-Day SD (mg/dL) - Week 24−5.93 (0.85)−0.50 (0.86)−5.43 (1.21)&amp;lt; 0.0001Within-Day SD (mg/dL)QOL Weight Interference0.001Low Blood Glucose Index (LBGI) - Week 240.20 (0.05)0.41 (0.05)−0.20 (0.08)0.008Within-Day SD (mg/dL)QOL Composite0.020 Disclosure D.C. Simonson: Advisory Panel; Self; GI Windows, Inc.. Stock/Shareholder; Self; GI Windows, Inc.. Stock/Shareholder; Spouse/Partner; Phase V Technologies, Inc. T. Vilsbøll: None. E. Ekholm: None. E.K. Johnsson: Employee; Self; AstraZeneca. M.A. Testa: Stock/Shareholder; Spouse/Partner; GI Windows, Inc.. Stock/Shareholder; Self; Phase V Technologies, Inc.. S. Jabbour: None. M. Lind: Consultant; Self; AstraZeneca. Research Support; Self; AstraZeneca. Consultant; Self; Novo Nordisk Inc.. Advisory Panel; Self; Novo Nordisk Inc.. Research Support; Self; Novo Nordisk Inc.. Consultant; Self; Eli Lilly and Company. Research Support; Self; Dexcom, Inc.. Consultant; Self; Eli Lilly and Company. Advisory Panel; Self; MSD K.K.. Research Support; Self; Pfizer Inc.. Consultant; Self; Medtronic.

Metrics to Evaluate Quality of Glycemic Control: Comparison of Time in Target, Hypoglycemic, and Hyperglycemic Ranges with "Risk Indices".

We sought to cross validate several metrics for quality of glycemic control, hypoglycemia, and hyperglycemia. We analyzed the mathematical properties of several metrics for overall glycemic control, and for hypo- and hyperglycemia, to evaluate their similarities, differences, and interrelationships. We used linear regression to describe interrelationships and examined correlations between metrics within three conceptual groups. There were consistently high correlations between %Time in range (%TIR) and previously described risk indices (M100, Blood Glucose Risk Index [BGRI], Glycemic Risk Assessment Diabetes Equation [GRADE], Index of Glycemic Control [IGC]), and with J-Index (J). There were also high correlations among %Hypoglycemia, Low Blood Glucose Index (LBGI), percentage of GRADE attributable to hypoglycemia (GRADE%Hypoglycemia), and Hypoglycemia Index, but negligible correlation with J. There were high correlations of percentage of time in hyperglycemic range (%Hyperglycemia) with High Blood Glucose Index (HBGI), percentage of GRADE attributable to hyperglycemia (GRADE%Hyperglycemia), Hyperglycemia Index, and J. %TIR is highly negatively correlated with %Hyperglycemia but very weakly correlated with %Hypoglycemia. By adjusting the parameters used in IGC, Hypoglycemia Index, Hyperglycemia Index, or in MR, one can more closely approximate the properties of BGRI, LBGI, or HBGI, and of GRADE, GRADE%Hypoglycemia, or GRADE%Hyperglycemia. Simple readily understandable criteria such as %TIR, %Hypoglycemia, and %Hyperglycemia are highly correlated with and appear to be as informative as "risk indices." The J-Index is sensitive to hyperglycemia but insensitive to hypoglycemia.

Measures of Glycemic Variability in Type 1 Diabetes and the Effect of Real-Time Continuous Glucose Monitoring.

To report the impact of continuous glucose monitoring (CGM) on glycemic variability (GV) indices, factors predictive of change, and to correlate variability with conventional markers of glycemia. Data from the JDRF study of CGM in participants with type 1 diabetes were used. Participants were randomized to CGM or self-monitored blood glucose (SMBG). GV indices at baseline, at 26 weeks in both groups, and at 52 weeks in the control group were analyzed. The associations of demographic and clinical factors with change in GV indices from baseline to 26 weeks were evaluated. Baseline data were available for 448 subjects. GV indices were all outside normative ranges (P < 0.001). Intercorrelation between GV indices was common and, apart from coefficient of variation (CV), low blood glucose index (LBGI), and percentage of glycemic risk assessment diabetes equation score attributable to hypoglycemia (%GRADEhypoglycemia), all indices correlate positively with HbA1c. There was strong correlation between time spent in hypoglycemia, and CV, LBGI, and %GRADEhypoglycemia, but not with HbA1c. A significant reduction in all GV indices, except lability index and mean absolute glucose change per unit time (MAG), was demonstrated in the intervention group at 26 weeks compared with the control group. Baseline factors predicting a change in GV with CGM include baseline HbA1c, baseline GV, frequency of daily SMBG, and insulin pump use. CGM reduces most GV indices compared with SMBG in people with type 1 diabetes. The strong correlation between time spent in hypoglycemia and CV, LBGI, and %GRADEhypoglycemia highlights the value of these metrics in assessing hypoglycemia as an adjunct to HbA1c in the overall assessment of glycemia.

Glucose variability indices in type 1 diabetes: parsimonious set of indices revealed by sparse principal component analysis.

Continuous glucose monitoring (CGM) time-series are often analyzed, retrospectively, to investigate glucose variability (GV), a risk factor for the development of complications in type 1 diabetes (T1D). In the literature, several tens of different indices for GV quantification have been proposed, but many of them carry very similar information. The aim of this article is to select a relatively small subset of GV indices from a wider pool of metrics, to obtain a parsimonious but still comprehensive description of GV in T1D datasets. A pool of 25 GV indices was evaluated on two CGM time-series datasets of 17 and 16 T1D subjects, respectively, collected during the European Union Seventh Framework Programme project "Diadvisor" (2008-2012) in two different clinical research centers using the Dexcom(®) (San Diego, CA) SEVEN(®) Plus. After the indices were centered and scaled, the Sparse Principal Component Analysis (SPCA) technique was used to determine a reduced set of metrics that allows preserving a high percentage of the variance of the whole original set. In order to assess whether or not the selected subset of GV indices is dataset-dependent, the analysis was applied to both datasets, as well as to the one obtained by merging them. SPCA revealed that a subset of up to 10 different GV indices can be sufficient to preserve more than the 60% of the variance originally explained by all the 25 variables. It is remarkable that four of these GV indices (i.e., Index of Glycemic Control, percentage of Glycemic Risk Assessment Diabetes Equation score due to euglycemia, percentage Coefficient of Variation, and Low Blood Glucose Index) were selected for all the considered T1D datasets. The SPCA methodology appears a suitable candidate to identify, among the large number of literature GV indices, subsets that allow obtaining a parsimonious, but still comprehensive, description of GV.

Glycemic Variability Measures in a Group of Subjects with Type 1 Diabetes and Repeated Severe and Non-Severe Hypoglycemia

There is some evidence that not only hemoglobin A1c (HbA1c) but also short-term variations in blood glucose could represent an independent risk factor for hypoglycemia.1 Normative ranges for measures of glycemic variability (GV) in normal subjects and in continuous glucose monitoring (CGM) data sets obtained from subjects with type 1 diabetes mellitus (T1DM) have been published.2 We sought to describe GV in T1DM patients with repeated episodes of severe hypoglycemia and non-severe hypoglycemia (SH/NSH). Thirty-six T1DM subjects underwent CGM for up to 72 h. Twenty-five of 36 were considered the hypoglycemic group [(H-group), presenting repeated SH/NSH], while the remaining 11 were evaluated as the control group (C-group). General characteristics, biochemical measurements, the number of hypoglycemic episodes, and the awareness of hypoglycemia (Clarke Sign Test) were evaluated. Glycemic variability was calculated using EasyGV© (by N. R. Hill, University of Oxford, Oxford, UK; available at www.easygv.co.uk) for the following: standard deviation (SD), M-value (mg glucose per kg body weight per min), mean amplitude of glucose excursions (MAGE), average daily risk ratio, lability index, J-Index, low blood glucose index (LBGI), high blood glucose index, continuous overlapping net glycemic action, mean of daily differences, glycemic risk assessment diabetes equation (GRADE), and mean absolute glucose. No differences were found in the entire set of clinical data or in HbA1c (H-group: age 34.5 ± 7.8 years, T1DM duration 16.0 ± 6.3 years, HbA1c 6.6 ± .0 %). Scores for the Clarke Sign Test differed for both groups, with higher values in the H-group. The H-group showed higher percentages of values and areas under the curve (AUC) <70 mg/dl with respect to the C-group (11.3 ± 8.4 vs 5.3 ± 5.3% <70 mg/dl; p < .05 and 2.5 ± 1.8 vs 0.8 ± 1.0 mg/dl AUC <70mg/dl; p < .005 for the H-group and the C-group, respectively). The H-group presented significantly higher scores than the T1DM C-group in LBGI (9.3 ± 5.0 vs 4.3 ± 2.4), GRADE (8.0 ± 3.4 vs 4.8 ± 3.5), GRADE-hypoglycemia (GRADE-hypo) (22.8 ± 22.0 vs 9.2 ± 8.9), and M-value (17.7 ± 6.5 vs 10.3 ± 6.0). Using CGM glucose profiles, we have shown that only the variability measures representing increased glycemic risk related to hypoglycemia and the quality of glycemic control differ significantly in T1DM subjects who are prone to repeated episodes of hypoglycemia. Almost all measurements (EasyGV software) from our entire group of subjects with T1DM were above the mean + 2 SD normative values that are described in a nondiabetic population. We could not find differences in the majority of measures, except for four (LBGI, GRADE, GRADE-hypo, and M-value), in our group of patients with repeated hypoglycemic episodes in comparison with the control group of T1DM patients without significant hypoglycemia and with good metabolic control. We found differences only in those glucose measures that mainly focused on increased glycemic risk coming from hypoglycemic episodes (LBGI and GRADE-hypo) and those mostly representing quality of glycemic control (GRADE and M-value). The only exception for a glucose measure specifically related to variability was MAGE, which was close to significantly higher in the group of patients with repeated hypoglycemia. As it has been claimed using self-monitoring blood glucose information,3 our data suggest that measures of glucose variability and quality of glucose control derived from CGM data could be used as instruments to estimate the risk of hypoglycemia and could help to prevent it.

Hypoglycemia, but not glucose variability, relates to vascular function in children with type 1 diabetes.

Chronic sustained hyperglycemia unequivocally predicts vascular disease in diabetes. However, the vascular risk of glucose variability, including hypoglycemia, is uncertain. Vascular dysfunction is present in children with type 1 diabetes and is a critical precursor of atherosclerosis. We aimed to evaluate the relationship between glucose variability and vascular function in children with type 1 diabetes. Fifty-two type 1 diabetes subjects (14 [SD 2.7] years old, 25 males) had continuous glucose monitoring that included 48 h of data used to evaluate glucose variability (mean amplitude of glycemic excursions [MAGE] and other measurements) and hypoglycemia indices (glycemic risk assessment diabetes equation [GRADE] hypoglycemia, Low Blood Glucose Index [LBGI], and observed duration of hypoglycemia). Children with type 1 diabetes and 50 age- and gender-matched controls had assessments of vascular function (flow-mediated dilatation [FMD] and glyceryl trinitrate-mediated dilatation [GTN]). Children with type 1 diabetes had lower FMD and GTN than controls (P=0.02 and P<0.001, respectively). GRADE hypoglycemia and LBGI were inversely related to FMD (r=-0.36, P=0.009 and r=-0.302, P=0.03, respectively) but did not relate to GTN. GRADE hypoglycemia was independently related to FMD (regression coefficient=-0.25±0.09, P=0.006). MAGE and other measurements of glucose variability measurements did not relate to FMD or GTN. Hypoglycemia, but not glucose variability, during continuous glucose monitoring relates to impaired vascular endothelial function in children with type 1 diabetes. Hypoglycemia may be an additional risk factor for early cardiovascular disease, but the effect of glucose variability, independent of glycosylated hemoglobin, on vascular function remains uncertain.

Effect of sitagliptin on glucose control in adult patients with Type 1 diabetes: a pilot, double-blind, randomized, crossover trial

Patients with Type 1 diabetes have significantly elevated postprandial glucagon secretion. Dipeptidyl peptidase IV inhibitors improve HbA(1c) by several mechanisms, including increasing glucagon-like peptide 1 and glucose-dependent insulinotropic peptide concentrations, which decreases postprandial rises in glucagon in both Type 1 and Type 2 diabetes. This study evaluates the clinical implications of sitagliptin in adult patients with Type 1 diabetes. This investigator-initiated, double-blind, randomized, crossover, 8-week, pilot study enrolled 20 adult subjects with Type 1 diabetes. Subjects received sitagliptin 100 mg/day or placebo for 4 weeks and then crossed over. Outcomes included 2-h postprandial blood glucose and 24-h area under the curve changes in glucose measurements from continuous glucose monitoring, HbA(1c) , fructosamine and insulin dose. Sitagliptin significantly reduced blood glucose (2-h postprandial and 24-h area under the curve) despite reduced total and prandial insulin dose. Based on continuous glucose monitor findings, sitagliptin improved measures of glycaemic control, including mean blood glucose (-0.6 mmol/l; P = 0.012) and time in euglycaemic range 4.4-7.8 mmol/l (0.4 ± 0.2 h; P = 0.046). Significant reductions were also observed in M100, Glycemic Risk Assessment Diabetes Equation (GRADE) and J-index. After controlling for period, treatment and insulin dose, the HbA(1c) was also significantly reduced [-0.27 ± 0.11% (-2.91 ± 1.16 mmol/mol); P = 0.025] when patients were taking sitagliptin. Sitagliptin significantly improved overall glucose control, including postprandial and 24-h glucose control, in adult patients with Type 1 diabetes, while significantly reducing prandial insulin requirements. Further investigation is warranted in patients with Type 1 diabetes in a larger cohort designed to assess both clinical outcomes and mechanism of action.

Cluster Analysis of Self-Monitoring Blood Glucose Assessments in Clinical Islet Cell Transplantation for Type 1 Diabetes

OBJECTIVECluster analysis was performed on the results of self-monitoring of blood glucose (SMBG) to discriminate islet graft function after islet cell transplantation (ICT) in patients with type 1 diabetes.RESEARCH DESIGN AND METHODSEleven islet recipients were included in this study. The patients visited our clinic monthly after ICT and provided blood samples for fasting C-peptide (n = 270), which were used to evaluate islet graft function. They also provided their SMBG data through an automatic data collection system. The SMBG data for 3 days immediately before each clinic visit were evaluated using the following assessments: M value, mean amplitude of glycemic excursions, J index, index of glycemic control, average daily risk range, and glycemic risk assessment diabetes equation. The cluster analysis was performed for both SMBG assessments and samples. Multivariate logistic regression analysis was used to evaluate the clusters of SMBG for assessing islet graft function.RESULTSAnalysis for SMBG assessments revealed five types of clusters, which showed similar patterns according to functional or dysfunctional islet graft phase. Two clusters, the euglycemia cluster (P < 0.001) and the hypoglycemia cluster (P = 0.001), were significant factors in the logistic model for islet graft function. The SMBG clusters had significant correlations with clinical graft indexes (P < 0.001).CONCLUSIONSCluster analysis of SMBG data as part of an automated data quality system could allow discrimination of islet graft dysfunction after ICT. This approach should be considered for islet recipients.

Reduced Glucose Variability Is Associated With Improved Quality of Glycemic Control in Patients With Type 2 Diabetes: A 12-Month Observational Study

Background: Current diabetes management relies mainly on hemoglobin A1C measurement to assess quality of treatment and to adjust therapy. We assessed the long-term effectivity of therapeutic adjustments in type 2 diabetes with specific focus on indices of glucose variability and multiple criteria for quality of glycemic control. Methods: Continuous glucose monitoring data collected during an observational study involving 405 outpatients with type 2 diabetes were analyzed at baseline and after 12 months. We evaluated the following criteria for glucose variability and quality of glycemic control: mean amplitude of glycemic excursions (MAGE), SD around the mean sensor glucose, mean of daily differences (MODD), fasting and mean sensor glucose, time outside specified glucose ranges, the Glycemic Risk Assessment Diabetes Equation (GRADE) score, High Blood Glucose Index (HBGI), and Low Blood Glucose Index (LBGI). Patients were classified according to quartiles of MAGE. Results: Indices of glucose variability were reduced by 15 to 23% (P 8.9 mmol/l decreased by 57% and 36% (P 8.9 mmol/l (R 2 = 0.374 and 0.877, P < 0.001 for both). Conclusions: Reduction of elevated glucose variability in type 2 diabetic outpatients is associated with lower risk of hypo- and hyperglycemia, and lower A1C values.

Translation of Personalized Decision Support into Routine Diabetes Care

The aim of this study was to evaluate the impact of personalized decision support (PDS) on metabolic control in people with diabetes and cardiovascular disease. The German health insurance fund BKK TAUNUS offers to its insured people with diabetes and cardiovascular disease the possibility to participate in the Diabetiva® program, which includes PDS. Personalized decision support is generated by the expert system KADIS® using self-control data and continuous glucose monitoring (CGM) as its data source. The physician of the participating person receives the PDS once a year, decides about use or nonuse, and reports his/her decision in a questionnaire. Metabolic control of participants treated by use or nonuse of PDS for one year and receiving CGM twice was analyzed in a retrospective observational study. The primary outcome was hemoglobin A1c (HbA1c); secondary outcomes were mean sensor glucose (MSG), glucose variability, and hypoglycemia. A total of 323 subjects received CGM twice, 289 had complete data sets, 97% (280/289) were type 2 diabetes patients, and 74% (214/289) were treated using PDS, resulting in a decrease in HbA1c [7.10±1.06 to 6.73±0.82%; p<.01; change in HbA1ct0-t12 months -0.37 (95% confidence interval -0.46 to -0.28)] and MSG (7.7±1.6 versus 7.4±1.2 mmol/liter; p=.003) within one year. Glucose variability was also reduced, as indicated by lower high blood glucose index (p=.001), Glycemic Risk Assessment Diabetes Equation (p=.009), and time of hyper-glycemia (p=.003). Low blood glucose index and time spent in hypoglycemia were not affected. In contrast, nonuse of PDS (75/289) resulted in increased HbA1c (p<.001). Diabetiva outcome was strongly related to baseline HbA1c (HbA1ct0; p<.01) and use of PDS (p<.01). Acceptance of PDS was dependent on HbA1ct0 (p=.049). Personalized decision support has potential to improve metabolic outcome in routine diabetes care.

Interpretation of Continuous Glucose Monitoring Data: Glycemic Variability and Quality of Glycemic Control

There are a large number of measures of glycemic variability, including standard deviation (SD), percentage coefficient of variation (%CV), interquartile range (IQR), mean amplitude of glucose excursion (MAGE), mean of daily differences (MODD), and continuous overlapping net glycemic action over an n-hour period (CONGA(n)). These are all highly correlated with the overall or "total" SD, SD(T). SD(T) is composed of several components corresponding to within-day variability, between-day variability (between daily means and between days-within specified time points), and the interaction of these sources of variability. We identify several subtypes of SD; each is highly correlated with SD(T). Variability may also depend on time of day. Numerous measures of quality of glycemic control have been proposed, including a weighted average of glucose values (M)(e.g., M(100) is M at 100 mg/dL), a measure of quality of glycemic control based on mean and SD (J), the Glycemic Risk Assessment Diabetes Equation (GRADE), the Index of Glycemic Control (IGC), the High Blood Glucose Index (HBGI), the Low Blood Glucose Index (LBGI), the Average Daily Risk Range (ADRR), and percentage of glucose values within specified ranges. These methods usually but not always give consistent results: they can differ widely in terms of their ability to detect responses to therapeutic interventions. Based on review of the advantages and limitations of these measures and on extensive experience in the application of these methods, we outline a systematic approach to the interpretation of continuous glucose monitoring data for use by clinical researchers and clinicians to evaluate the quality of glycemic control, glucose variability including within- and between-day variability, the day-to-day stability of glycemic patterns, and changes in response to therapy.