Type 1 diabetes

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

Diabetes Technol TherPublished Prediction model development study

Authors' conclusion

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.

Study context

Population
Adults with T1D; structured study-specified (aerobic, interval, resistance training videos) and free-living exercise sessions from the T1D Exercise Initiative study
Comparison
Repeated measures random forest (RMRF) vs repeated measures logistic regression (RMLR) for predicting hypoglycemia during exercise
Outcome
Hypoglycemia (CGM value <70 mg/dL) during exercise; model performance by AUC and balanced accuracy
Follow-up
During exercise
Methodological detailsrepeated measures random forest, repeated measures logistic regression
  • Repeated measures random forest
  • Repeated measures logistic regression

Main finding

As reported in the abstract

Prediction of hypoglycemia during exercise
GroupAUCBalanced accuracy
RMRF0.83377%
RMLR0.82577%

AUC

RMRF0.833
RMLR0.825

Balanced accuracy

RMRF77%
RMLR77%

Further findings

Predictors of hypoglycemia

Higher probability of hypoglycemia with lower pre-exercise glucose, negative pre-exercise glucose rates of change, greater % time <70 mg/dL in the 24 h before exercise, and greater pre-exercise bolus insulin-on-board (IOB)

Hypoglycemia by exercise type

Highest probability with free-living aerobic exercise, walking/hiking and physical labor; lowest with structured exercise

Original publication

The Type 1 Diabetes and EXercise Initiative: Predicting Hypoglycemia Risk During Exercise for Participants with Type 1 Diabetes Using Repeated Measures Random Forest. Diabetes Technol Ther. Published 20 June 2023. doi:10.1089/dia.2023.0140

Composed by AI from the published abstract and checked against it. How Briefs are made →

Get them every week

New studies on this topic in My Week and in your Monday e-mail.

We e-mail you a sign-in link and a code. No password, no spam.