Prediction of Personal Glycemic Responses to Food for Individuals With Type 1 Diabetes Through Integration of Clinical and Microbial Data

OBJECTIVE Despite technological advances, results from various clinical trials repeatedly showed that many individuals with type 1 diabetes (T1D) do not achieve their glycemic goals. One of the major challenges in disease management is the administration of an accurate amount of insulin for each meal which will match the expected postprandial glycemic response (PPGR). RESEARCH DESIGN AND METHODS We recruited individuals with T1D using continuous glucose monitoring (CGM) and continuous subcutaneous insulin infusion (CSII) devices simultaneously to a prospective cohort and profiled them for 2 weeks. Participants were asked to report real-time dietary intake using a designated mobile app. We measured their PPGRs and devised machine-learning algorithms for PPGR prediction, which integrate glucose measurements, insulin dosages, dietary habits, blood parameters, anthropometrics, exercise, and gut microbiota. Data of the PPGR of 1,057 healthy individuals to 47,863 meals were also integrated into the model. The performance of the models was evaluated using 10-fold cross validation. RESULTS 121 individuals with T1D, 75 adults and 46 children, were included in the study. PPGR to 6,377 meals was measured. Our PPGR prediction model substantially outperforms a baseline model emulating standard of care (correlation of R=0.59 compared to R=0.40 for predicted and observed PPGR respectively, p <10−10). The model was robust across different subpopulations. Feature attribution analysis revealed that glucose levels at meal initiation, glucose trend 30 minutes prior to meal, meal carbohydrate content and meal’s carbohydrate/fat ratio were the most influential features to the model. CONCLUSIONS Our model enables a more accurate prediction of PPGR and therefore may allow a better adjustment of the required insulin dosage for meals. It can be further implemented in closed-loop systems and may lead to rationally designed nutritional interventions personally tailored for individuals with T1D based on meals with expected low glycemic response.

Download Full-text

Blood Glucose Level Forecasting on Type-1-Diabetes Subjects during Physical Activity: A Comparative Analysis of Different Learning Techniques

Bioengineering ◽

10.3390/bioengineering8060072 ◽

2021 ◽

Vol 8 (6) ◽

pp. 72

Author(s):

Benedetta De Paoli ◽

Federico D’Antoni ◽

Mario Merone ◽

Silvia Pieralice ◽

Vincenzo Piemonte ◽

...

Keyword(s):

Physical Activity ◽

Type 1 Diabetes ◽

Blood Glucose ◽

Mean Squared Error ◽

Glucose Monitoring ◽

Online Training ◽

Machine Learning Algorithms ◽

Glucose Levels ◽

Blood Glucose Levels

Background: Type 1 Diabetes Mellitus (T1DM) is a widespread chronic disease in industrialized countries. Preventing blood glucose levels from exceeding the euglycaemic range would reduce the incidence of diabetes-related complications and improve the quality of life of subjects with T1DM. As a consequence, in the last decade, many Machine Learning algorithms aiming to forecast future blood glucose levels have been proposed. Despite the excellent performance they obtained, the prediction of abrupt changes in blood glucose values produced during physical activity (PA) is still one of the main challenges. Methods: A Jump Neural Network was developed in order to overcome the issue of predicting blood glucose values during PA. Three learning configurations were developed and tested: offline training, online training, and online training with reinforcement. All configurations were tested on six subjects suffering from T1DM that held regular PA (three aerobic and three anaerobic) and exploited Continuous Glucose Monitoring (CGM). Results: The forecasting performance was evaluated in terms of the Root-Mean-Squared-Error (RMSE), according to a paradigm of Precision Medicine. Conclusions: The online learning configurations performed better than the offline configuration in total days but not on the only CGM associated with the PA; thus, the results do not justify the increased computational burden because the improvement was not significant.

Download Full-text

Metabolic control and hypoglycaemia in people with type 1 diabetes: insulin pump therapy vs. intensified insulin therapy in an unselected cohort in routine care

Diabetology & Metabolic Syndrome ◽

10.1186/s13098-021-00700-0 ◽

2021 ◽

Vol 13 (1) ◽

Author(s):

Guido Kramer ◽

Christof Kloos ◽

Ulrich A. Müller ◽

Gunter Wolf ◽

Nadine Kuniss

Keyword(s):

Type 1 Diabetes ◽

Insulin Therapy ◽

Metabolic Control ◽

Continuous Glucose Monitoring ◽

Treatment Satisfaction ◽

Severe Hypoglycaemia ◽

Glucose Monitoring ◽

Routine Care ◽

Insulin Dosage

Abstract Aims The aim of this study was to compare individuals with type 1 diabetes with continuous subcutaneous insulin infusion (CSII) and intensified insulin therapy (ICT) in routine care regarding metabolic control and treatment satisfaction. Methods Individuals with type 1 diabetes (CSII n = 74; ICT n = 163) were analysed regarding metabolic control, frequency of hypoglycaemia and treatment satisfaction (DTSQs range 0–36). Results Individuals with CSII (duration of CSII: 14.1 ± 7.2 years) were younger (51.1 ± 15.8 vs. 56.2 ± 16.2 years, p = 0.023), had longer diabetes duration (28.7 ± 12.4 vs. 24.6 ± 14.3 years, p = 0.033), lower insulin dosage (0.6 ± 0.2 vs. 0.7 ± 0.4 IU/kg, p = 0.004), used more frequently short-acting analogue insulin (90.5% vs. 48.5%, p < 0.001) and flash/continuous glucose monitoring (50.0% vs. 31.9%, p = 0.009) than people with ICT. HbA1c was similar between CSII and ICT (7.1 ± 0.8%/54.4 ± 9.1 mmol/mol vs. 7.2 ± 1.0%/55.7 ± 10.9 mmol/mol, p = 0.353). Individuals with CSII had higher frequency of non-severe hypoglycaemia per week (in people with blood glucose monitoring: 1.9 ± 1.7 vs. 1.2 ± 1.6, p = 0.014; in people with flash/continuous glucose monitoring: 3.3 ± 2.2 vs. 2.1 ± 2.0, p = 0.006). Prevalence of polyneuropathy (18.9% vs. 38.0%, p = 0.004) and systolic blood pressure (138.0 ± 16.4 vs. 143.9 ± 17.1 mmHg, p = 0.014) was lower in CSII. Satisfaction with diabetes treatment (26.7 ± 7.3 vs. 26.0 ± 6.8, p = 0.600) did not differ between CSII and ICT. Conclusions CSII and ICT yielded comparable metabolic control and treatment satisfaction but CSII was associated with higher incidence of non-severe hypoglycaemia and lower insulin dosage.

Download Full-text

Use of a Mobile App to Facilitate Blood Glucose Monitoring in Adolescents With Type 1 Diabetes: Single-Subject Nonrandomized Clinical Trial

JMIR Diabetes ◽

10.2196/diabetes.8357 ◽

2018 ◽

Vol 3 (1) ◽

pp. e3

Author(s):

Edward J Bellfield ◽

Lisa K Sharp ◽

Yinglin Xia ◽

Ben S Gerber

Keyword(s):

Clinical Trial ◽

Type 1 Diabetes ◽

Blood Glucose ◽

Glucose Monitoring ◽

Mobile App ◽

Blood Glucose Monitoring ◽

Single Subject

Download Full-text

Type 1 Diabetes and Addison’s Disease: When the Diagnosis Is Suggested by the Continuous Glucose Monitoring System

Children ◽

10.3390/children8080702 ◽

2021 ◽

Vol 8 (8) ◽

pp. 702

Author(s):

Francesco Vinci ◽

Giuseppe d’Annunzio ◽

Flavia Napoli ◽

Marta Bassi ◽

Carolina Montobbio ◽

...

Keyword(s):

Type 1 Diabetes ◽

Adrenal Insufficiency ◽

Continuous Glucose Monitoring ◽

Young Patients ◽

Glucose Monitoring ◽

Addison’S Disease ◽

Insulin Dosage ◽

Adrenal Crisis ◽

Addison's Disease

Our objective is to emphasize the important role of continuous glucose monitoring (CGM) in suggesting adrenal insufficiency in patients affected by type 1 diabetes. We describe an adolescent girl with type 1 diabetes and subsequent latent Addison’s disease diagnosed based on a recurrent hypoglycemic trend detected by CGM. In patients with type 1 diabetes, persistent unexplained hypoglycemic episodes at dawn together with reduced insulin requirement arouse souspicionof adrenal insufficiency. Adrenal insufficiency secondary to autoimmune Addison’s disease, even if rarely encountered among young patients, may be initially symptomless and characterized by slow progression up to acute adrenal crisis, which represents a potentially life-threatening condition. Besides glycometabolic assessment and adequate insulin dosage adjustment, type 1 diabetes needs prompt recognition of potentially associated autoimmune conditions. Among these, Addison’s disease can be suspected, although latent or paucisymptomatic, through periodic and careful evaluation of CGM data.

Download Full-text