Assessing Mealtime Macronutrient Content: Patient Perceptions Versus Expert Analyses via a Novel Phone App

The Association Between Diet Quality and Glycemic Outcomes Among People with Type 1 Diabetes

Background

The amount and type of food consumed impacts the glycemic response and insulin needs of people with type 1 diabetes mellitus (T1DM). Daily variability in consumption, reflected in diet quality, may acutely impact glycemic levels and insulin needs.

Objective

Type 1 Diabetes Exercise Initiative (T1DEXI) data were examined to evaluate the impact of daily diet quality on near-term glycemic control and interaction with exercise.

Methods

Using the Remote Food Photography Method, ≤8 d of dietary intake data were analyzed per participant. Diet quality was quantified with the Healthy Eating Index-2015 (HEI), where a score of 100 indicates the highest-quality diet. Each participant day was classified as low HEI (≤57) or high HEI (>57) based on the mean of nationally reported HEI data. Within participants, the relationship between diet quality and subsequent glycemia measured by continuous glucose monitoring (CGM) and total insulin dose usage was evaluated using a paired t-test and robust regression models.

Results

Two hundred twenty-three adults (76% female) with mean ± SD age, HbA1c, and body mass index (BMI) of 37 ± 14 y, 6.6% ± 0.7%, and 25.1 ± 3.6 kg/m2, respectively, were included in these analyses. The mean HEI score was 56 across all participant days. On high HEI days (mean, 66 ± 4) compared with low HEI days (mean, 47 ± 5), total time in range (70-180 mg/dL) was greater (77.2% ± 14% compared with 75.7% ± 14%, respectively, P = 0.01), whereas time above 180 mg/dL (19% ± 14% compared with 21% ± 15%, respectively, P = 0.004), mean glucose (143 ± 22 compared with 145 ± 22 mg/dL, respectively, P = 0.02), and total daily insulin dose (0.52 ± 0.18 compared with 0.54 ± 0.18 U/kg/d, respectively, P = 0.009) were lower. The interaction between diet quality and exercise on glycemia was not significant.

Conclusions

Higher HEI scores correlated with improved glycemia and lower insulin needs, although the impact of diet quality was modest and smaller than the previously reported impact of exercise.

Factors Affecting Reproducibility of Change in Glucose During Exercise: Results From the Type 1 Diabetes and EXercise Initiative

AIMS

To evaluate factors affecting within-participant reproducibility in glycemic response to different forms of exercise.

METHODS

Structured exercise sessions ~30 minutes in length from the Type 1 Diabetes Exercise Initiative (T1DEXI) study were used to assess within-participant glycemic variability during and after exercise. The effect of several pre-exercise factors on the within-participant glycemic variability was evaluated.

RESULTS

Data from 476 adults with type 1 diabetes were analyzed. A participant's change in glucose during exercise was reproducible within 15 mg/dL of the participant's other exercise sessions only 32% of the time. Participants who exercised with lower and more consistent glucose level, insulin on board (IOB), and carbohydrate intake at exercise start had less variability in glycemic change during exercise. Participants with lower mean glucose (P < .001), lower glucose coefficient of variation (CV) (P < .001), and lower % time <70 mg/dL (P = .005) on sedentary days had less variable 24-hour post-exercise mean glucose.

CONCLUSIONS

Reproducibility of change in glucose during exercise was low in this cohort of adults with T1D, but more consistency in pre-exercise glucose levels, IOB, and carbohydrates may increase this reproducibility. Mean glucose variability in the 24 hours after exercise is influenced more by the participant's overall glycemic control than other modifiable factors.

Artificial Intelligence and Machine Learning for Improving Glycemic Control in Diabetes: Best Practices, Pitfalls, and Opportunities

OBJECTIVE

Artificial intelligence and machine learning are transforming many fields including medicine. In diabetes, robust biosensing technologies and automated insulin delivery therapies have created a substantial opportunity to improve health. While the number of manuscripts addressing the topic of applying machine learning to diabetes has grown in recent years, there has been a lack of consistency in the methods, metrics, and data used to train and evaluate these algorithms. This manuscript provides consensus guidelines for machine learning practitioners in the field of diabetes, including best practice recommended approaches and warnings about pitfalls to avoid.

METHODS

Algorithmic approaches are reviewed and benefits of different algorithms are discussed including importance of clinical accuracy, explainability, interpretability, and personalization. We review the most common features used in machine learning applications in diabetes glucose control and provide an open-source library of functions for calculating features, as well as a framework for specifying data sets using data sheets. A review of current data sets available for training algorithms is provided as well as an online repository of data sources.

SIGNIFICANCE

These consensus guidelines are designed to improve performance and translatability of new machine learning algorithms developed in the field of diabetes for engineers and data scientists.

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.

The Type 1 Diabetes and EXercise Initiative: Predicting Hypoglycemia Risk During Exercise for Participants with Type 1 Diabetes Using Repeated Measures Random Forest

Objective: Exercise is known to increase the risk for hypoglycemia in type 1 diabetes (T1D) but predicting when it may occur remains a major challenge. The objective of this study was to develop a hypoglycemia prediction model based on a large real-world study of exercise in T1D. Research Design and Methods: Structured study-specified exercise (aerobic, interval, and resistance training videos) and free-living exercise sessions from the T1D Exercise Initiative study were used to build a model for predicting hypoglycemia, a continuous glucose monitoring value <70 mg/dL, during exercise. Repeated measures random forest (RMRF) and repeated measures logistic regression (RMLR) models were constructed to predict hypoglycemia using predictors at the start of exercise and baseline characteristics. Models were evaluated with area under the receiver operating characteristic curve (AUC) and balanced accuracy. Results: RMRF and RMLR had similar AUC (0.833 vs. 0.825, respectively) and both models had a balanced accuracy of 77%. The probability of hypoglycemia was higher for exercise sessions with lower pre-exercise glucose levels, negative pre-exercise glucose rates of change, greater percent time <70 mg/dL in the 24 h before exercise, and greater pre-exercise bolus insulin-on-board (IOB). Free-living aerobic exercises, walking/hiking, and physical labor had the highest probability of hypoglycemia, while structured exercises had the lowest probability of hypoglycemia. Conclusions: RMRF and RMLR accurately predict hypoglycemia during exercise and identify factors that increase the risk of hypoglycemia. Lower glucose, decreasing levels of glucose before exercise, and greater pre-exercise IOB largely predict hypoglycemia risk in adults with T1D.

Enabling fully automated insulin delivery through meal detection and size estimation using Artificial Intelligence

We present a robust insulin delivery system that includes automated meal detection and carbohydrate content estimation using machine learning for meal insulin dosing called robust artificial pancreas (RAP). We conducted a randomized, single-center crossover trial to compare postprandial glucose control in the four hours following unannounced meals using a hybrid model predictive control (MPC) algorithm and the RAP system. The RAP system includes a neural network model to automatically detect meals and deliver a recommended meal insulin dose. The meal detection algorithm has a sensitivity of 83.3%, false discovery rate of 16.6%, and mean detection time of 25.9 minutes. While there is no significant difference in incremental area under the curve of glucose, RAP significantly reduces time above range (glucose >180 mg/dL) by 10.8% (P = 0.04) and trends toward increasing time in range (70-180 mg/dL) by 9.1% compared with MPC. Time below range (glucose <70 mg/dL) is not significantly different between RAP and MPC.

Modeling risk of hypoglycemia during and following physical activity in people with type 1 diabetes using explainable mixed-effects machine learning

BACKGROUND

Physical activity (PA) can cause increased hypoglycemia (glucose <70 mg/dL) risk in people with type 1 diabetes (T1D). We modeled the probability of hypoglycemia during and up to 24 h following PA and identified key factors associated with hypoglycemia risk.

METHODS

We leveraged a free-living dataset from Tidepool comprised of glucose measurements, insulin doses, and PA data from 50 individuals with T1D (6448 sessions) for training and validating machine learning models. We also used data from the T1Dexi pilot study that contains glucose management and PA data from 20 individuals with T1D (139 session) for assessing the accuracy of the best performing model on an independent test dataset. We used mixed-effects logistic regression (MELR) and mixed-effects random forest (MERF) to model hypoglycemia risk around PA. We identified risk factors associated with hypoglycemia using odds ratio and partial dependence analysis for the MELR and MERF models, respectively. Prediction accuracy was measured using the area under the receiver operating characteristic curve (AUROC).

RESULTS

The analysis identified risk factors significantly associated with hypoglycemia during and following PA in both MELR and MERF models including glucose and body exposure to insulin at the start of PA, low blood glucose index 24 h prior to PA, and PA intensity and timing. Both models showed overall hypoglycemia risk peaking 1 h after PA and again 5-10 h after PA, which is consistent with the hypoglycemia risk pattern observed in the training dataset. Time following PA impacted hypoglycemia risk differently across different PA types. Accuracy of hypoglycemia prediction using the fixed effects of the MERF model was highest when predicting hypoglycemia during the first hour following the start of PA (AUROCVALIDATION = 0.83 and AUROCTESTING = 0.86) and decreased when predicting hypoglycemia in the 24 h after PA (AUROCVALIDATION = 0.66 and AUROCTESTING = 0.68).

CONCLUSION

Hypoglycemia risk after the start of PA can be modeled using mixed-effects machine learning to identify key risk factors that may be used within decision support and insulin delivery systems. We published the population-level MERF model online for others to use.

Incorporating Glucose Variability into Glucose Forecasting Accuracy Assessment Using the New Glucose Variability Impact Index and the Prediction Consistency Index: An LSTM Case Example

BACKGROUND

In this work, we developed glucose forecasting algorithms trained and evaluated on a large dataset of free-living people with type 1 diabetes (T1D) using closed-loop (CL) and sensor-augmented pump (SAP) therapies; and we demonstrate how glucose variability impacts accuracy. We introduce the glucose variability impact index (GVII) and the glucose prediction consistency index (GPCI) to assess the accuracy of prediction algorithms.

METHODS

A long-short-term-memory (LSTM) neural network was designed to predict glucose up to 60 minutes in the future using continuous glucose measurements and insulin data collected from 175 people with T1D (41,318 days) and evaluated on 75 people (11,333 days) from the Tidepool Big Data Donation Dataset. LSTM was compared with two naïve forecasting algorithms as well as Ridge linear regression and a random forest using root-mean-square error (RMSE). Parkes error grid quantified clinical accuracy. Regression analysis was used to derive the GVII and GPCI.

RESULTS

The LSTM had highest accuracy and best GVII and GPCI. RMSE for CL was 19.8 ± 3.2 and 33.2 ± 5.4 mg/dL for 30- and 60-minute prediction horizons, respectively. RMSE for SAP was 19.6 ± 3.8 and 33.1 ± 7.3 mg/dL for 30- and 60-minute prediction horizons, respectively; 99.6% and 97.6% of predictions were within zones A+B of the Parkes error grid at 30- and 60-minute prediction horizons, respectively. Glucose variability was strongly correlated with RMSE (R≥0.64, P < 0.001); GVII and GPCI demonstrated a means to compare algorithms across datasets with different glucose variability.

CONCLUSIONS

The LSTM model was accurate on a large real-world free-living dataset. Glucose variability should be considered when assessing prediction accuracy using indices such as GVII and GPCI.

The Ambulatory Glucose Profile: Opportunities for Enhancement

The ambulatory glucose profile (AGP) and the frequency distribution for glucose by ranges are well established as standard methods for display, analysis, and interpretation of glucose data arising from self-monitoring, continuous glucose monitoring, and automated insulin delivery systems. In this review, we consider several refinements that may further improve the utility of the AGP. These include (1) display of the AGP together with information regarding dietary intake, medication administration (e.g., insulin), glucose lowering (pharmacodynamic) activity of medications, and physical activity measured by accelerometers or heart rate; (2) display of average time below range (%TBR), time above range (%TAR), and time in range (%TIR) by time of day to indicate timing of hypoglycemic and hyperglycemic episodes; (3) detailed analysis of postprandial excursions for each of the major meals after synchronizing by onset of meals and adjusting for the premeal glucose levels, enabling comparisons of magnitude, shape, and patterns; (4) methods to characterize distinct patterns on different days of the week; (5) display of glucose on a nonlinear scale to improve the balance between deviations associated with hypoglycemia versus hyperglycemia; (6) use of time scales other than midnight-to-midnight to facilitate analysis of time segments of particular interest (e.g., overnight); (7) options to display individual glucose values to assist in the identification of dates and times of outliers and episodes of hypoglycemia and hyperglycemia; and (8) methods to compare AGPs obtained from different individuals or groups receiving alternative interventions in terms of therapy or technology. These refinements, individually or collectively, can potentially further enhance the effectiveness of the AGP for assessment of glucose levels, patterns, and variability. We discuss several questions regarding implementation and optimization of these methods.