Prognostic accuracy of continuous glucose monitoring in the prediction of diabetes mellitus in children with incidental hyperglycemia: receiver operating characteristic analysis

Postprandial Plasma Glucose between 4 and 7.9 h May Be a Potential Diagnostic Marker for Diabetes

Postprandial glucose levels between 4 and 7.9 h (PPG4-7.9h) correlate with mortality from various diseases, including hypertension, diabetes, cardiovascular disease, and cancer. This study aimed to assess if predicted PPG4-7.9h could diagnose diabetes. Two groups of participants were involved: Group 1 (4420 participants) had actual PPG4-7.9h, while Group 2 (8422 participants) lacked this measure but had all the diabetes diagnostic measures. Group 1 underwent multiple linear regression to predict PPG4-7.9h using 30 predictors, achieving accuracy within 11.1 mg/dL in 80% of the participants. Group 2 had PPG4-7.9h predicted using this model. A receiver operating characteristic curve analysis showed that predicted PPG4-7.9h could diagnose diabetes with an accuracy of 87.3% in Group 2, with a sensitivity of 75.1% and specificity of 84.1% at the optimal cutoff of 102.5 mg/dL. A simulation on 10,000 random samples from Group 2 revealed that 175 participants may be needed to investigate PPG4-7.9h as a diabetes diagnostic marker with a power of at least 80%. In conclusion, predicted PPG4-7.9h appears to be a promising diagnostic indicator for diabetes. Future studies seeking to ascertain its definitive diagnostic value might require a minimum sample size of 175 participants.

Continuous glucose monitoring and HbA1c in the evaluation of glucose metabolism in children at high risk for type 1 diabetes mellitus

AIMS

Continuous glucose monitoring (CGM) parameters, self-monitored blood glucose (SMBG), HbA1c and oral glucose tolerance test (OGTT) were studied during preclinical type 1 diabetes mellitus.

METHODS

Ten asymptomatic children with multiple (⩾2) islet autoantibodies (cases) and 10 age and sex-matched autoantibody-negative controls from the Type 1 Diabetes Prediction and Prevention (DIPP) Study were invited to 7-day CGM with Dexcom G4 Platinum Sensor. HbA1c and two daily SMBG values (morning and evening) were analyzed. Five-point OGTTs were performed and carbohydrate intake was assessed by food records. The matched pairs were compared with the paired sample t-test.

RESULTS

The cases showed higher mean values and higher variation in glucose levels during CGM compared to the controls. The time spent ⩾7.8mmol/l was 5.8% in the cases compared to 0.4% in the controls (p=0.040). Postprandial CGM values were similar except after the dinner (6.6mmol/l in cases vs. 6.1mmol/l in controls; p=0.023). When analyzing the SMBG values higher mean level, higher evening levels, as well as higher variation were observed in the cases when compared to the controls. HbA1c was significantly higher in the cases [5.7% (39mmol/mol) vs. 5.3% (34mmol/mol); p=0.045]. No differences were observed in glucose or C-peptide levels during OGTT. Daily carbohydrate intake was slightly higher in the cases (254.2g vs. 217.7g; p=0.034).

CONCLUSIONS

Glucose levels measured by CGM and SMBG are useful indicators of dysglycemia during preclinical type 1 diabetes mellitus. Increased evening glucose values seem to be common in children with preclinical type 1 diabetes mellitus.

Glycemic variability in patients with Wolfram syndrome is lower than in type 1 diabetes

AIMS

Wolfram syndrome (WFS) is diagnosed as coexistence of diabetes mellitus and optic atrophy, where pancreatic beta cell destruction is associated with neurodegeneration. Typically, WFS necessitates insulin treatment similar to type 1 diabetes (T1D), but the mechanism of beta cell mass reduction leading to hyperglycemia is different.

METHODS

The aim of the study was to assess glycemic variability using the continuous glucose monitoring (CGM) system in seven pediatric patients with genetically confirmed WFS and compare the results with data obtained from 21 propensity score-matched patients with T1D. The "GlyCulator" application was used for the calculation of glycemic variability indices.

RESULTS

CGM recordings showed similarities in glycemic variability among WFS patients, but differing from those of the T1D group. Coefficient of variation (%CV), CONGA4h, and GONGA6h were significantly (p < 0.05) lower in WFS patients (28.08 ± 7.37, 54.96 ± 11.92, and 55.99 ± 10.58) than in T1D patients (37.87 ± 14.24, 74.12 ± 28.74, p = 0.02, and 80.26 ± 35.05, respectively). In WFS patients, the percentage of values above 126 mg/dL was 69.79 (52.08-77.43), whereas in patients with T1D, the percentage was significantly lower-47.22 (35.07-62.85, p = 0.018). Curiously, a tendency toward a lower percentage of measurements below 70 mg/dL was noted in the WFS group [0 (0-7.29)] in comparison with the T1D group [6.25 (0-18.06), p = 0.122]. WFS patients had a significantly higher C-peptide level (0.31 ± 0.2 ng/mL) than T1D patients (0.04 ± 0.04 ng/mL; p = 0.006).

CONCLUSIONS

Patients with WFS show smaller glycemic variability than individuals with T1D, and this may be associated with persistent residual insulin secretion.

Relationship between glycaemic variability and hyperglycaemic clamp-derived functional variables in (impending) type 1 diabetes

AIMS/HYPOTHESIS

We examined whether measures of glycaemic variability (GV), assessed by continuous glucose monitoring (CGM) and self-monitoring of blood glucose (SMBG), can complement or replace measures of beta cell function and insulin action in detecting the progression of preclinical disease to type 1 diabetes.

METHODS

Twenty-two autoantibody-positive (autoAb(+)) first-degree relatives (FDRs) of patients with type 1 diabetes who were themselves at high 5-year risk (50%) for type 1 diabetes underwent CGM, a hyperglycaemic clamp test and OGTT, and were followed for up to 31 months. Clamp variables were used to estimate beta cell function (first-phase [AUC5-10 min] and second-phase [AUC120-150 min] C-peptide release) combined with insulin resistance (glucose disposal rate; M 120-150 min). Age-matched healthy volunteers (n = 20) and individuals with recent-onset type 1 diabetes (n = 9) served as control groups.

RESULTS

In autoAb(+) FDRs, M 120-150 min below the 10th percentile (P10) of controls achieved 86% diagnostic efficiency in discriminating between normoglycaemic FDRs and individuals with (impending) dysglycaemia. M 120-150 min outperformed AUC5-10 min and AUC120-150 min C-peptide below P10 of controls, which were only 59-68% effective. Among GV variables, CGM above the reference range was better at detecting (impending) dysglycaemia than elevated SMBG (77-82% vs 73% efficiency). Combined CGM measures were equally efficient as M 120-150 min (86%). Daytime GV variables were inversely correlated with clamp variables, and more strongly with M 120-150 min than with AUC5-10 min or AUC120-150 min C-peptide.

CONCLUSIONS/INTERPRETATION

CGM-derived GV and the glucose disposal rate, reflecting both insulin secretion and action, outperformed SMBG and first- or second-phase AUC C-peptide in identifying FDRs with (impending) dysglycaemia or diabetes. Our results indicate the feasibility of developing minimally invasive CGM-based criteria for close metabolic monitoring and as outcome measures in trials.

A Computational Method to Determine Glucose Infusion Rates for Isoglycemic Intravenous Glucose Infusion Study

The results of the isoglycemic intravenous glucose infusion (IIGI) study need to mimic the dynamic glucose profiles during the oral glucose tolerance test (OGTT) to accurately calculate the incretin effect. The glucose infusion rates during IIGI studies have historically been determined by experienced research personnel using the manual ad-hoc method. In this study, a computational method was developed to automatically determine the infusion rates for IIGI study based on a glucose-dynamics model. To evaluate the computational method, 18 subjects with normal glucose tolerance underwent a 75 g OGTT. One-week later, Group 1 (n = 9) and Group 2 (n = 9) underwent IIGI studies using the ad-hoc method and the computational method, respectively. Both methods were evaluated using correlation coefficient, mean absolute relative difference (MARD), and root mean square error (RMSE) between the glucose profiles from the OGTT and the IIGI study. The computational method exhibited significantly higher correlation (0.95 ± 0.03 versus 0.86 ± 0.10, P = 0.019), lower MARD (8.72 ± 1.83% versus 13.11 ± 3.66%, P = 0.002), and lower RMSE (10.33 ± 1.99 mg/dL versus 16.84 ± 4.43 mg/dL, P = 0.002) than the ad-hoc method. The computational method can facilitate IIGI study, and enhance its accuracy and stability. Using this computational method, a high-quality IIGI study can be accomplished without the need for experienced personnel.

The effectiveness and durability of an early insulin pump therapy in children and adolescents with type 1 diabetes mellitus

BACKGROUND

This study evaluated the predictors of effectiveness and durability of insulin pump therapy in children and adolescents who have initiated continuous subcutaneous insulin infusion (CSII) within 2 years after the diagnosis of type 1 diabetes mellitus (T1DM).

SUBJECTS AND METHODS

The charts of individuals with T1DM using insulin pumps who were treated at our center were reviewed, including subjects with age at onset of <22 years, interval between onset and insulin pump commencement (interval onset-commencement) of <2 years, use of pumps of >1 year, and use of glucose sensors for <4 weeks/year. The primary end point was the mean glycosylated hemoglobin (HbA1c) value (MHbA1c) throughout the follow-up.

RESULTS

From 684 patients treated with insulin pumps, 119 met the inclusion criteria, and 113 were selected for statistical analysis (60 females; age at diabetes onset, 8.9±5.6 years [mean±SD]; follow-up, 4.0±1.8 years; range, 1-8 years; baseline HbA1c, 9.3±1.8%). Only the interval onset-commencement was a linear predictor of the MHbA1c (P=0.01; R(2)=0.089). A significant reduction of the mean yearly HbA1c from baseline throughout all the follow-up was observed (P<0.001). Categorizing the sample into four quartiles on the basis of an increasing interval onset-commencement resulted in levels of MHbA1c significantly lower in the first and second quartiles in comparison with the fourth quartile (7.6±0.8% and 7.8±1.0%, respectively, versus 8.5±0.8%; P<0.001 and P=0.004, respectively).

CONCLUSIONS

The present study suggests that early pump commencement in children and adolescents with T1DM provides lower and more durable HbA1c values than a late commencement. It is possible that an early pump commencement could prolong the honeymoon phase, but we cannot confirm or exclude this hypothesis because the lack of data about C-peptide levels during the follow-up.