Time in range for fully closed-loop systems versus current care during physical exercise in type 1 diabetes: a systematic review and meta-analysis

Abstract

Background and aims Fully closed-loop systems (CLS) are an attempt to replace the physiological role of pancreatic α-cells in people with type 1 diabetes (T1D). CLS have shown to improve glycemic control under environmental conditions, however, during physical exercise the performance of these systems is questionable since high change rates in glucose and varying glucose responses during exercise might be limiting factors. Therefore, the aim of this work is to perform a systematic review and meta-analysis, evaluating if CLS improve the time spent in an euglycemic range (TIR: 70–180 mg/dL [3.9–10.0 mmol/L]) compared to current care during physical exercise in people with T1D. Materials and methods A systematic literature search was conducted in PubMed, EMBASE, ISI Web of Knowledge Cochrane Central and Register of Controlled Trials from January 1950 until January 2020. Studies were extracted, assessed for eligibility, methodological quality, and level of evidence by two independently working researchers. Fit for inclusion were randomized controlled trials in which CLS were compared against current care during physical exercise in people with T1D. In order to perform the meta-analysis, the random effects model and restricted maximum likelihood estimation method were used. Heterogeneity of included studies was assessed using the Cochran’s Q Test (T2) and I2 statistics. Results Six randomized controlled trials with a total of 153 participants with T1D were included. Studies using different CLS or exercise interventions in crossover test designs, were included repeatedly (a-d), doubling the comparisons to an overall number of 266. TIR was significantly higher for CLS compared to current care, showing an absolute mean difference of 6.17%, 95% CI [1.60, 10.75%] (p

Date of Award	2022
Original language	English
Awarding Institution	Medical University Graz
Supervisor	Othmar Moser (Supervisor) & Max Lennart Eckstein (Co-supervisor)