Research Papers

A Data-Driven Personalized Model of Glucose Dynamics Taking Account of the Effects of Physical Activity for Type 1 Diabetes: An In Silico Study

[+] Author and Article Information
Jinyu Xie

Mechanical and Nuclear Engineering,
315 Leonhard Building,
Penn State University,
University Park, PA 16802
e-mail: jxx123@psu.edu

Qian Wang

Mechanical Engineering,
325 Leonhard Building,
Penn State University,
University Park, PA 16802
e-mail: quw6@psu.edu

1Corresponding author.

Manuscript received January 30, 2018; final manuscript received August 26, 2018; published online October 17, 2018. Assoc. Editor: Guy M. Genin.

J Biomech Eng 141(1), 011006 (Oct 17, 2018) (12 pages) Paper No: BIO-18-1057; doi: 10.1115/1.4041522 History: Received January 30, 2018; Revised August 26, 2018

This paper aims to develop a data-driven model for glucose dynamics taking into account the effects of physical activity (PA) through a numerical study. It intends to investigate PA's immediate effect on insulin-independent glucose variation and PA's prolonged effect on insulin sensitivity. We proposed a nonlinear model with PA (NLPA), consisting of a linear regression of PA and a bilinear regression of insulin and PA. The model was identified and evaluated using data generated from a physiological PA-glucose model by Dalla Man et al. integrated with the uva/padova Simulator. Three metrics were computed to compare blood glucose (BG) predictions by NLPA, a linear model with PA (LPA), and a linear model with no PA (LOPA). For PA's immediate effect on glucose, NLPA and LPA showed 45–160% higher mean goodness of fit (FIT) than LOPA under 30 min-ahead glucose prediction (P < 0.05). For the prolonged PA effect on glucose, NLPA showed 87% higher FIT than LPA (P < 0.05) for simulations using no previous measurements. NLPA had 25–37% and 31–54% higher sensitivity in predicting postexercise hypoglycemia than LPA and LOPA, respectively. This study demonstrated the following qualitative trends: (1) for moderate-intensity exercise, accuracy of BG prediction was improved by explicitly accounting for PA's effect; and (2) accounting for PA's prolonged effect on insulin sensitivity can increase the chance of early prediction of postexercise hypoglycemia. Such observations will need to be further evaluated through human subjects in the future.

Copyright © 2019 by ASME
Your Session has timed out. Please sign back in to continue.


Thabit, H. , Tauschmann, M. , Allen, J. M. , Leelarathna, L. , Hartnell, S. , Wilinska, M. E. , Acerini, C. L. , Dellweg, S. , Benesch, C. , Heinemann, L. , Mader, J. K. , Holzer, M. , Kojzar, H. , Exall, J. , Yong, J. , Pichierri, J. , Barnard, K. D. , Kollman, C. , Cheng, P. , Hindmarsh, P. C. , Campbell, F. M. , Arnolds, S. , Peiber, T. R. , Evans, M. L. , Dunger, D. B. , and Hovorka, R. , 2015, “ Home Use of an Artificial Beta Cell in Type 1 Diabetes,” N. Engl. J. Med., 373(22), pp. 2129–2140. [CrossRef] [PubMed]
Doyle, F. J. , Huyett, L. M. , Lee, J. B. , Zisser, H. C. , and Dassau, E. , 2014, “ Closed-Loop Artificial Pancreas Systems: Engineering the Algorithms,” Diabetes Care, 37(5), pp. 1191–1197. [CrossRef] [PubMed]
Kudva, Y. C. , Carter, R. E. , Cobelli, C. , Basu, R. , and Basu, A. , 2014, “ Closed-Loop Artificial Pancreas Systems: Physiological Input to Enhance Next-Generation Devices,” Diabetes Care, 37(5), pp. 1184–1190. [CrossRef] [PubMed]
Wang, Q. , Xie, J. , Molenaar, P. , and Ulbrecht, J. S. , 2015, “ Model Predictive Control for Type 1 Diabetes Based on Personalized Linear Time-Varying Subject Model Consisting of Both Insulin and Meal Inputs: An In Silico Evaluation,” J. Diabetes Sci. Technol. 9(4), pp. 941–942. [CrossRef] [PubMed]
Mallad, A. , Hinshaw, L. , Schiavon, M. , Dalla Man, C. , Dadlani, V. , Basu, R. , Lingineni, R. , Cobelli, C. , Johnson, M. L. , Carter, R. , Kudva, Y. C. , and Basu, A. , 2015, “ Exercise Effects on Postprandial Glucose Metabolism in Type 1 Diabetes: A Triple-Tracer Approach,” Am. J. Physiol. Endocrinol. Metab., 308(12), pp. E1106–E1115. [CrossRef] [PubMed]
Breton, M. D. , 2008, “ Physical Activity the Major Unaccounted Impediment to Closed Loop Control,” J. Diabetes Sci. Technol., 2(1), pp. 169–174. [CrossRef] [PubMed]
Goodyear, L. J. , and Kahn, B. B. , 1998, “ Exercise, Glucose Transport, and Insulin Sensitivity,” Annu. Rev. Med., 49(1), pp. 235–261. [CrossRef] [PubMed]
Metcalf, K. M. , Singhvi, A. , Tsalikian, E. , Tansey, M. J. , Zimmerman, M. B. , Esliger, D. W. , and Janz, K. F. , 2014, “ Effects of Moderate-to-Vigorous Intensity Physical Activity on Overnight and Next-Day Hypoglycemia in Active Adolescents With Type 1 Diabetes,” Diabetes Care, 37(5), pp. 1272–1278. [CrossRef] [PubMed]
Roberts, A. J. , and Taplin, C. E. , 2015, “ Exercise in Youth With Type 1 Diabetes,” Curr. Pediatr. Rev., 11(2), pp. 120–125. [CrossRef] [PubMed]
Garcia-Garcia, F. , Kumareswaran, K. , Hovorka, R. , and Hernando, M. , 2015, “ Quantifying the Acute Changes in Glucose With Exercise in Type 1 Diabetes: A Systematic Review and Meta-Analysis,” Sports Med., 45(4), pp. 587–599. [CrossRef] [PubMed]
Bergman, R. N. , Ider, Y. Z. , Bowden, C. R. , and Cobelli, C. , 1979, “ Quantitative Estimation of Insulin Sensitivity,” Am. J. Physiol., 236(6), pp. E667–E677. https://www.physiology.org/doi/abs/10.1152/ajpendo.1979.236.6.E667 [PubMed]
Dalla Man, C. , Breton, M. D. , and Cobelli, C. , 2009, “ Physical Activity Into the Meal Glucose Insulin Model of Type 1 Diabetes: In Silico Studies,” J. Diabetes Sci. Technol., 3(1), pp. 56–67. [CrossRef] [PubMed]
Dalla Man, C. , Rizza, R. A. , and Cobelli, C. , 2007, “ Meal Simulation Model of the Glucose-Insulin System,” IEEE Trans. Biomed. Eng., 54(10), pp. 1740–1749. [CrossRef] [PubMed]
Roy, A. , and Parker, R. S. , 2007, “ Dynamic Modeling of Exercise Effects on Plasma Glucose and Insulin Levels,” J. Diabetes Sci. Technol., 1(3), pp. 338–347. [CrossRef] [PubMed]
Ewings, S. M. , Sahu, S. K. , Valletta, J. J. , Byrne, C. D. , and Chipperfield, A. J. , 2015, “ A Bayesian Network for Modelling Blood Glucose Concentration and Exercise in Type 1 Diabetes,” Stat. Methods Med. Res., 24(3), pp. 342–72. [CrossRef] [PubMed]
Eren-Oruklu, M. , Cinar, A. , Rollins, D. K. , and Quinn, L. , 2012, “ Adaptive System Identification for Estimating Future Glucose Concentrations and Hypoglycemia Alarms,” Automatica, 48(8), pp. 1892–1897. [CrossRef] [PubMed]
Turksoy, K. , Bayrak, E. S. , Quinn, L. , Littlejohn, E. , Rollins, D. , and Cinar, A. , 2013, “ Hypoglycemia Early Alarm Systems Based on Multivariable Models,” Ind. Eng. Chem. Res., 52(35), pp. 12329–12336. [CrossRef]
Turksoy, K. , Quinn, L. T. , Littlejohn, E. , and Cinar, A. , 2014, “ An Integrated Multivariable Artificial Pancreas Control System,” J. Diabetes Sci. Technol., 8(3), pp. 498–507. [CrossRef] [PubMed]
Turksoy, K. , Quinn, L. T. , Littlejohn, E. , and Cinar, A. , 2014, “ Multivariable Adaptive Identification and Control for Arti- Ficial Pancreas Systems,” IEEE Trans. Biomed. Eng., 61(3), pp. 883–891. [CrossRef] [PubMed]
Dasanayake, I. S. , Seborg, D. E. , Pinsker, J. E. , Doyle, F. J. , and Dassau, E. , 2015, “ Empirical Dynamic Model Identification for Blood-Glucose Dynamics in Response to Physical Activity,” Conference on Decision and Control (CDC), Osaka, Japan, Dec. 15–18, pp. 3834–3839.
Lee, H. , and Bequette, B. W. , 2009, “ A Closed-Loop Arti- Ficial Pancreas Based on Model Predictive Control: Human friendly Identification and Automatic Meal Disturbance Rejection,” Biomed. Signal Process. Control., 4(4), pp. 347–354. [CrossRef]
Percival, M. W. , Bevier, W. C. , Wang, Y. , Dassau, E. , Zisser, H. C. , Jovanovic, L. , and Doyle, F. J., III, 2010, “ Modeling the Effects of Subcutaneous Insulin Administration and Carbohydrate Consumption on Blood Glucose,” J. Diabetes Sci. Technol., 4(5), pp. 1214–1228. [CrossRef] [PubMed]
Finan, D. A. , Zisser, H. , Jovanovic, L. , Bevier, W. C. , and Seborg, D. E. , 2007, “ Practical Issues in the Identification of Empirical Models From Simulated Type 1 Diabetes Data,” Diabetes Technol. Ther., 9(5), pp. 438–450. [CrossRef] [PubMed]
Xie, J. , and Wang, Q. , 2016, “ A Nonlinear Data-Driven Model of Glucose Dynamics Accounting for Physical Activity for Type 1 Diabetes: An in Silico Study,” ASME Paper No. DSCC2016-9742.
Breton, M. , and Kovatchev, B. , 2008, “ Analysis, Modeling, and Simulation of the Accuracy of Continuous Glucose Sensors,” J. Diabetes Sci. Technol., 2(5), pp. 853–862. [CrossRef] [PubMed]
Wang, Q. , Molenaar, P. , Harsh, S. , Freeman, K. , Xie, J. , Gold, C. , Rovine, M. , and Ulbrecht, J. , 2014, “ Personalized State-Space Modeling of Glucose Dynamics for Type 1 Diabetes Using Continuously Monitored Glucose, Insulin Dose, and Meal Intake an Extended Kalman Filter Approach,” J. Diabetes Sci. Technol., 8(2), pp. 331–345. [CrossRef] [PubMed]
Schiavon, M. , Hinshaw, L. , Mallad, A. , Dalla Man, C. , Sparacino, G. , Johnson, M. , Carter, R. , Basu, R. , Kudva, Y. , Cobelli, C. , and Basu, A. , 2013, “ Postprandial Glucose Fluxes and Insulin Sensitivity During Exercise: A Study in Healthy Individuals,” Am. J. Physiol. Endocrinol. Metab., 305(4), pp. E557–E566. [CrossRef] [PubMed]
Frohnauer, M. K. , Woodworth, J. R. , and Anderson, J. H., Jr ., 2001, “ Graphical Human Insulin Time-Activity Profiles Using Standardized Definitions,” Diabetes Technol. Ther., 3(3), pp. 419–429. [CrossRef] [PubMed]
Kovatchev, B. P. , Breton, M. , Dalla Man, C. , and Cobelli, C. , 2009, “ In Silico Preclinical Trials: A Proof of Concept in Closed-Loop Control of Type 1 Diabetes,” J. Diabetes Sci. Technol., 3(1), pp. 44–55. [CrossRef] [PubMed]
Riddell, M. C. , Zaharieva, D. P. , Yavelberg, L. , Cinar, A. , and Jamnik, V. K. , 2015, “ Exercise and the Development of the Artificial Pancreas: One of the More Difficult Series of Hurdles,” J. Diabetes Sci. Technol., 9(6), pp. 1217–26. [CrossRef] [PubMed]
Tsalikian, E. , Kollman, C. , Tamborlane, W. B. , Beck, R. W. , Fiallo-Scharer, R. , and Diabetes Research in Children Network (DirecNet) Study Group, 2006, “ Prevention of Hypoglycemia During Exercise in Children With Type 1 Diabetes by Suspending Basal Insulin,” Diabetes Care, 29(10), pp. 2200–2204. [CrossRef] [PubMed]
Zakynthinaki, M. S. , 2015, “ Modelling Heart Rate Kinetics,” PLoS One, 10(4), p. e0118263. [CrossRef] [PubMed]
Cameron, F. , Bequette, B. W. , Wilson, D. M. , Buckingham, B. A. , Lee, H. , and Niemeyer, G. , 2011, “ A Closed-Loop Artificial Pancreas Based on Risk Management,” J. Diabetes Sci. Technol., 5(2), pp. 368–379. [CrossRef] [PubMed]
Magni, L. , Raimondo, D. M. , Bossi, L. , Dalla Man, C. , De Nicolao, G. , Kovatchev, B. , and Cobelli, C. , 2007, “ Model Predictive Control of Type 1 Diabetes: An in Silico Trial,” J. Diabetes Sci. Technol., 1(6), pp. 804–812. [CrossRef] [PubMed]
El-Khatib, F. H. , Russell, S. J. , Nathan, D. M. , Sutherlin, R. G. , and Damiano, E. R. , 2010, “ A Bihormonal Closed-loop Artificial Pancreas for Type 1 Diabetes,” Sci. Transl. Med., 2(27), p. 27ra27. [CrossRef] [PubMed]
Finan, D. A. , Palerm, C. C. , Doyle, F. J. , Seborg, D. E. , Zisser, H. , Bevier, W. C. , and Jovanovic, L. , 2009, “ Effect of Input Excitation on the Quality of Empirical Dynamic Models for Type 1 Diabetes,” AIChE J., 55(5), pp. 1135–1146. [CrossRef]
Clarke, W. , and Kovatchev, B. , 2009, “ Statistical Tools to Analyze Continuous Glucose Monitor Data,” Diabetes Technol. Ther., 11(Suppl. 1), pp. S-45–S-54. [CrossRef]
Gore, C. J. , and Withers, R. T. , 1990, “ The Effect of Exercise Intensity and Duration on the Oxygen Deficit and Excess Post-Exercise Oxygen Consumption,” Eur. J. Appl. Physiol., 60(3), pp. 169–174. [CrossRef]
Saalasti, S. , 2003, “ Neural Networks for Heart Rate Time Series Analysis,” Ph.D. dissertation, Department of Mathematical Information Technology, University of Jyväskylä, Jyväskylä, Finland.


Grahic Jump Location
Fig. 1

Three-day training protocol. For the table on top of the timeline, second row indicates the amount of CHO being taken; first row indicates that the amount of insulin bolus = (number in the first row) ×ICR, e.g., “45 * 1” represents insulin bolus = 45×1×ICR, and “80 * 0.8” denotes insulin bolus = 80×0.8×ICR.

Grahic Jump Location
Fig. 2

Two-day validation protocol. For the table on top of the timeline, second row indicates the amount of CHO being taken; first row indicates that the amount of insulin bolus = (number in the first row) ×ICR, e.g., “90 * 0.8” represents insulin bolus = 90×0.8×ICR.

Grahic Jump Location
Fig. 3

Adult #008: 30-min ahead predictions by the NLPA, LPA, and LOPA from 1 h before exercise to about 2-h after exercise. BG corresponds to the CGM output from the simulator. FIT values are computed based on the predictions during and 2-h after exercise: (a) under the training protocol; (b) under the validation protocol.

Grahic Jump Location
Fig. 4

Adolescent #005: 3-h ahead predictions by the NLPA, LPA, and LOPA under the 3-day training protocol. BG corresponds to the CGM output from the simulator. FIT values are calculated based on the predictions during and 24 h after exercise.

Grahic Jump Location
Fig. 5

Adolescent #005: 3-h ahead predictions by NLPA, LPA, and LOPA under the 2-day validation protocol. BG corresponds to the CGM output from the simulator. FIT values are calculated based on the prediction during and 24 h after exercise.

Grahic Jump Location
Fig. 6

Histogram of the model order k (where nb1=nb2=nb3=k) with the lowest AIC, evaluated with 30 subjects



Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In