Combining uncertainty-aware predictive modeling and a bedtime Smart Snack intervention to prevent nocturnal hypoglycemia in people with type 1 diabetes on multiple daily injections

OBJECTIVE

Nocturnal hypoglycemia is a known challenge for people with type 1 diabetes, especially for physically active individuals or those on multiple daily injections. We developed an evidential neural network (ENN) to predict at bedtime the probability and timing of nocturnal hypoglycemia (0-4 vs 4-8 h after bedtime) based on several glucose metrics and physical activity patterns. We utilized these predictions in silico to prescribe bedtime carbohydrates with a Smart Snack intervention specific to the predicted minimum nocturnal glucose and timing of nocturnal hypoglycemia.

MATERIALS AND METHODS

We leveraged free-living datasets collected from 366 individuals from the T1DEXI Study and Glooko. Inputs to the ENN used to model nocturnal hypoglycemia were derived from demographic information, continuous glucose monitoring, and physical activity data. We assessed the accuracy of the ENN using area under the receiver operating curve, and the clinical impact of the Smart Snack intervention through simulations.

RESULTS

The ENN achieved an area under the receiver operating curve of 0.80 and 0.71 to predict nocturnal hypoglycemic events during 0-4 and 4-8 h after bedtime, respectively, outperforming all evaluated baseline methods. Use of the Smart Snack intervention reduced probability of nocturnal hypoglycemia from 23.9 ± 14.1% to 14.0 ± 13.3% and duration from 7.4 ± 7.0% to 2.4 ± 3.3% in silico.

DISCUSSION

Our findings indicate that the ENN-based Smart Snack intervention has the potential to significantly reduce the frequency and duration of nocturnal hypoglycemic events.

CONCLUSION

A decision support system that combines prediction of minimum nocturnal glucose and proactive recommendations for bedtime carbohydrate intake might effectively prevent nocturnal hypoglycemia and reduce the burden of glycemic self-management.

Quantifying insulin-mediated and noninsulin-mediated changes in glucose dynamics during resistance exercise in type 1 diabetes

Exercise can cause dangerous fluctuations in blood glucose in people living with type 1 diabetes (T1D). Aerobic exercise, for example, can cause acute hypoglycemia secondary to increased insulin-mediated and noninsulin-mediated glucose utilization. Less is known about how resistance exercise (RE) impacts glucose dynamics. Twenty-five people with T1D underwent three sessions of either moderate or high-intensity RE at three insulin infusion rates during a glucose tracer clamp. We calculated time-varying rates of endogenous glucose production (EGP) and glucose disposal (Rd) across all sessions and used linear regression and extrapolation to estimate insulin- and noninsulin-mediated components of glucose utilization. Blood glucose did not change on average during exercise. The area under the curve (AUC) for EGP increased by 1.04 mM during RE (95% CI: 0.65-1.43, P < 0.001) and decreased proportionally to insulin infusion rate (0.003 mM per percent above basal rate, 95% CI: 0.001-0.006, P = 0.003). The AUC for Rd rose by 1.26 mM during RE (95% CI: 0.41-2.10, P = 0.004) and increased proportionally with insulin infusion rate (0.04 mM per percent above basal rate, CI: 0.03-0.04, P < 0.001). No differences were observed between the moderate and high resistance groups. Noninsulin-mediated glucose utilization rose significantly during exercise before returning to baseline roughly 30-min postexercise. Insulin-mediated glucose utilization remained unchanged during exercise sessions. Circulating catecholamines and lactate rose during exercise despite relatively small changes observed in Rd. Results provide an explanation of why RE may pose a lower overall risk for hypoglycemia.NEW & NOTEWORTHY Aerobic exercise is known to cause decreases in blood glucose secondary to increased glucose utilization in people living with type 1 diabetes (T1D). However, less is known about how resistance-type exercise impacts glucose dynamics. Twenty-five participants with T1D performed in-clinic weight-bearing exercises under a glucose clamp. Mathematical modeling of infused glucose tracer allowed for quantification of the rate of hepatic glucose production as well as rates of insulin-mediated and noninsulin-mediated glucose uptake experienced during resistance exercise.

Assessment of a Decision Support System for Adults with Type 1 Diabetes on Multiple Daily Insulin Injections

Introduction: DailyDose is a decision support system designed to provide real-time dosing advice and weekly insulin dose adjustments for adults living with type 1 diabetes using multiple daily insulin injections. Materials and Methods: Twenty-five adults were enrolled in this single-arm study. All participants used Dexcom G6 for continuous glucose monitoring, InPen for short-acting insulin doses, and Clipsulin to track long-acting insulin doses. Participants used DailyDose on an iPhone for 8 weeks. The primary endpoint was % time in range (TIR) comparing the 2-week baseline to the final 2-week period of DailyDose use. Results: There were no significant differences between TIR or other glycemic metrics between the baseline period compared to final 2-week period of DailyDose use. TIR significantly improved by 6.3% when more than half of recommendations were accepted and followed compared with 50% or fewer recommendations (95% CI 2.5%-10.1%, P = 0.001). Conclusions: Use of DailyDose did not improve glycemic outcomes compared to the baseline period. In a post hoc analysis, accepting and following recommendations from DailyDose was associated with improved TIR. Clinical Trial Registration Number: NCT04428645.

Quantifying the impact of physical activity on future glucose trends using machine learning

Prevention of hypoglycemia (glucose <70 mg/dL) during aerobic exercise is a major challenge in type 1 diabetes. Providing predictions of glycemic changes during and following exercise can help people with type 1 diabetes avoid hypoglycemia. A unique dataset representing 320 days and 50,000 + time points of glycemic measurements was collected in adults with type 1 diabetes who participated in a 4-arm crossover study evaluating insulin-pump therapies, whereby each participant performed eight identically designed in-clinic exercise studies. We demonstrate that even under highly controlled conditions, there is considerable intra-participant and inter-participant variability in glucose outcomes during and following exercise. Participants with higher aerobic fitness exhibited significantly lower minimum glucose and steeper glucose declines during exercise. Adaptive, personalized machine learning (ML) algorithms were designed to predict exercise-related glucose changes. These algorithms achieved high accuracy in predicting the minimum glucose and hypoglycemia during and following exercise sessions, for all fitness levels.

Predicting and Preventing Nocturnal Hypoglycemia in Type 1 Diabetes Using Big Data Analytics and Decision Theoretic Analysis

Background: Despite new glucose sensing technologies, nocturnal hypoglycemia is still a problem for people with type 1 diabetes (T1D) as symptoms and sensor alarms may not be detected while sleeping. Accurately predicting nocturnal hypoglycemia before sleep may help minimize nighttime hypoglycemia. Methods: A support vector regression (SVR) model was trained to predict, before bedtime, the overnight minimum glucose and overnight nocturnal hypoglycemia for people with T1D. The algorithm was trained on continuous glucose measurements and insulin data collected from 124 people (22,804 valid nights of data) with T1D. The minimum glucose threshold for announcing nocturnal hypoglycemia risk was derived by applying a decision theoretic criterion to maximize expected net benefit. Accuracy was evaluated on a validation set from 10 people with T1D during a 4-week trial under free-living sensor-augmented insulin-pump therapy. The primary outcome measures were sensitivity and specificity of prediction, the correlation between predicted and actual minimum nocturnal glucose, and root-mean-square error. The impact of using the algorithm to prevent nocturnal hypoglycemia is shown in-silico. Results: The algorithm predicted 94.1% of nocturnal hypoglycemia events (<3.9 mmol/L, 95% confidence interval [CI], 71.3-99.9) with an area under the receiver operating characteristic curve of 0.86 (95% CI, 0.75-0.98). Correlation between actual and predicted minimum glucose was high (R = 0.71, P < 0.001). In-silico simulations showed that the algorithm could reduce nocturnal hypoglycemia by 77.0% (P = 0.006) without impacting time in target range (3.9-10 mmol/L). Conclusion: An SVR model trained on a big data set and optimized using decision theoretic criterion can accurately predict at bedtime if overnight nocturnal hypoglycemia will occur and may help reduce nocturnal hypoglycemia.

Dual-Hormone Closed-Loop System Using a Liquid Stable Glucagon Formulation Versus Insulin-Only Closed-Loop System Compared With a Predictive Low Glucose Suspend System: An Open-Label, Outpatient, Single-Center, Crossover, Randomized Controlled Trial

OBJECTIVE

To assess the efficacy and feasibility of a dual-hormone (DH) closed-loop system with insulin and a novel liquid stable glucagon formulation compared with an insulin-only closed-loop system and a predictive low glucose suspend (PLGS) system.

RESEARCH DESIGN AND METHODS

In a 76-h, randomized, crossover, outpatient study, 23 participants with type 1 diabetes used three modes of the Oregon Artificial Pancreas system: 1) dual-hormone (DH) closed-loop control, 2) insulin-only single-hormone (SH) closed-loop control, and 3) PLGS system. The primary end point was percentage time in hypoglycemia (<70 mg/dL) from the start of in-clinic aerobic exercise (45 min at 60% VO_2max) to 4 h after.

RESULTS

DH reduced hypoglycemia compared with SH during and after exercise (DH 0.0% [interquartile range 0.0-4.2], SH 8.3% [0.0-12.5], P = 0.025). There was an increased time in hyperglycemia (>180 mg/dL) during and after exercise for DH versus SH (20.8% DH vs. 6.3% SH, P = 0.038). Mean glucose during the entire study duration was DH, 159.2; SH, 151.6; and PLGS, 163.6 mg/dL. Across the entire study duration, DH resulted in 7.5% more time in target range (70-180 mg/dL) compared with the PLGS system (71.0% vs. 63.4%, P = 0.044). For the entire study duration, DH had 28.2% time in hyperglycemia vs. 25.1% for SH (P = 0.044) and 34.7% for PLGS (P = 0.140). Four participants experienced nausea related to glucagon, leading three to withdraw from the study.

CONCLUSIONS

The glucagon formulation demonstrated feasibility in a closed-loop system. The DH system reduced hypoglycemia during and after exercise, with some increase in hyperglycemia.

Accuracy of the Dexcom G6 Glucose Sensor during Aerobic, Resistance, and Interval Exercise in Adults with Type 1 Diabetes

The accuracy of continuous glucose monitoring (CGM) sensors may be significantly impacted by exercise. We evaluated the impact of three different types of exercise on the accuracy of the Dexcom G6 sensor. Twenty-four adults with type 1 diabetes on multiple daily injections wore a G6 sensor. Participants were randomized to aerobic, resistance, or high intensity interval training (HIIT) exercise. Each participant completed two in-clinic 30-min exercise sessions. The sensors were applied on average 5.3 days prior to the in-clinic visits (range 0.6–9.9). Capillary blood glucose (CBG) measurements with a Contour Next meter were performed before and after exercise as well as every 10 min during exercise. No CGM calibrations were performed. The median absolute relative difference (MARD) and median relative difference (MRD) of the CGM as compared with the reference CBG did not differ significantly from the start of exercise to the end exercise across all exercise types (ranges for aerobic MARD: 8.9 to 13.9% and MRD: −6.4 to 0.5%, resistance MARD: 7.7 to 14.5% and MRD: −8.3 to −2.9%, HIIT MARD: 12.1 to 16.8% and MRD: −14.3 to −9.1%). The accuracy of the no-calibration Dexcom G6 CGM was not significantly impacted by aerobic, resistance, or HIIT exercise.

An artificial intelligence decision support system for the management of type 1 diabetes

Type 1 diabetes (T1D) is characterized by pancreatic beta cell dysfunction and insulin depletion. Over 40% of people with T1D manage their glucose through multiple injections of long-acting basal and short-acting bolus insulin, so-called multiple daily injections (MDI)^1,2. Errors in dosing can lead to life-threatening hypoglycaemia events (<70 mg dl^-1) and hyperglycaemia (>180 mg dl^-1), increasing the risk of retinopathy, neuropathy, and nephropathy. Machine learning (artificial intelligence) approaches are being harnessed to incorporate decision support into many medical specialties. Here, we report an algorithm that provides weekly insulin dosage recommendations to adults with T1D using MDI therapy. We employ a unique virtual platform³ to generate over 50,000 glucose observations to train a k-nearest neighbours⁴ decision support system (KNN-DSS) to identify causes of hyperglycaemia or hypoglycaemia and determine necessary insulin adjustments from a set of 12 potential recommendations. The KNN-DSS algorithm achieves an overall agreement with board-certified endocrinologists of 67.9% when validated on real-world human data, and delivers safe recommendations, per endocrinologist review. A comparison of inter-physician-recommended adjustments to insulin pump therapy indicates full agreement of 41.2% among endocrinologists, which is consistent with previous measures of inter-physician agreement (41-45%)⁵. In silico^3,6 benchmarking using a platform accepted by the United States Food and Drug Administration for evaluation of artificial pancreas technologies indicates substantial improvement in glycaemic outcomes after 12 weeks of KNN-DSS use. Our data indicate that the KNN-DSS allows for early identification of dangerous insulin regimens and may be used to improve glycaemic outcomes and prevent life-threatening complications in people with T1D.

Artificial Intelligence in Decision Support Systems for Type 1 Diabetes

Type 1 diabetes (T1D) is a chronic health condition resulting from pancreatic beta cell dysfunction and insulin depletion. While automated insulin delivery systems are now available, many people choose to manage insulin delivery manually through insulin pumps or through multiple daily injections. Frequent insulin titrations are needed to adequately manage glucose, however, provider adjustments are typically made every several months. Recent automated decision support systems incorporate artificial intelligence algorithms to deliver personalized recommendations regarding insulin doses and daily behaviors. This paper presents a comprehensive review of computational and artificial intelligence-based decision support systems to manage T1D. Articles were obtained from PubMed, IEEE Xplore, and ScienceDirect databases. No time period restrictions were imposed on the search. After removing off-topic articles and duplicates, 562 articles were left to review. Of those articles, we identified 61 articles for comprehensive review based on algorithm evaluation using real-world human data, in silico trials, or clinical studies. We grouped decision support systems into general categories of (1) those which recommend adjustments to insulin and (2) those which predict and help avoid hypoglycemia. We review the artificial intelligence methods used for each type of decision support system, and discuss the performance and potential applications of these systems.