Effect of exercise on plasma insulin levels in individuals with type 1 diabetes: A systematic review and meta-analysis.

Abstract

Current evidence of the impact of acute exercise on insulin levels in individuals with type 1 diabetes remains controversial. Therefore, we conducted a systematic review and meta-analysis to explore exercise-induced changes in insulin levels. We conducted a systematic review (until 05 November 2023) and meta-analysis exploring the effect of exercise on insulin concentration in individuals with type 1 diabetes. We included randomised cross-over studies for rapid-acting insulin and pre- and post-studies for long-acting insulin in individuals with type 1 diabetes performing any type of acute exercise and had a control condition. The exercise-induced change in insulin levels was the outcome of interest. When possible, the mean differences (MDs) in insulin levels were pooled using the DerSimonian and Laird random effect method. Risk of bias was assessed for each included study. Seventeen trials, encompassing 186 participants with type 1 diabetes, were included in the systematic review. Twelve out of 17 studies included participants on rapid-acting insulin regimens and used a cross-over design, whereas five out of 17 single-arm studies included participants on (ultra)long-acting insulin. Seven out of 12 studies on rapid-acting insulins and all the single-arm studies were at high risk of bias. Results suggest a statistically significant, small-to-moderate increase of rapid-acting insulin after 30 min of exercise (MD of 18.44 [95% CI 0.02; 36.86; I2 0%] pmol/L); meanwhile, findings on (ultra)long-acting insulin were inconclusive. A small-to-moderate increase of insulin levels in studies including rapid-acting insulin was found after a bout of physical exercise in individuals with type 1 diabetes. However, current gaps in high-quality evidence challenge our understanding of insulin kinetics around exercise.