Abstract
Statement of Problem Type 2 diabetes mellitus (T2DM), a public health crisis fueled by obesity, poor diet, and sedentary lifestyle, affects nearly 38 million adults in the United States, with lack of time cited as a primary reason for physical inactivity (Gildea et al., 2021; Riebe et al., 2018). Given that aerobic, resistance, and interval exercise have all been shown to significantly improve markers of blood glucose regulation, high-intensity interval training (HIIT) offers an efficient way of potentially inducing significant and clinically meaningful blood glucose improvements in patients with T2DM. While existing meta-analyses have assessed various exercise interventions in T2DM patients, none have separated submaximal from supramaximal HIIT, so it is unclear whether certain HIIT interval intensities may be more effective than others in combating hyperglycemia in T2DM. The purpose of this study was to determine if submaximal or supramaximal HIIT reduces blood glucose or improves markers of insulin resistance more than the other in T2DM patients using a systematic review and meta-analytical approach.
Methods Embase/MEDLINE, Cochrane Central Register of Controlled Trials (CENTRAL), and PubMed databases were systematically searched in July 2021, with a repeat search conducted in April 2022, using specific combinations of keywords and search filters to find published studies using submaximal or supramaximal HIIT interventions on individuals with T2DM. Seventy-eight articles were compared to the defined inclusion and exclusion criteria, and a total of twenty-one articles (eighteen from the July 2021 search and three from the April 2022 search) were selected for this meta-analysis. Nineteen used a submaximal HIIT exercise protocol, while two used a supramaximal HIIT protocol. Comprehensive Meta-Analysis (CMA) was used for data management and calculations. Six glycemic control outcome variables were identified, including HbA1c, fasting blood glucose (FBG), fasting insulin, Homeostasis Model of Insulin Resistance (HOMA-IR), and two-hour glucose, and for each study, pre- and post-HIIT means and standard deviation for each reported outcome variable of interest were entered into CMA. Effect sizes were corrected and weighted, and Hedges’ g values were reported along with 95% confidence intervals and p values (Lipsey & Wilson, 2001; Thomas & French, 1986).
Results Using a multivariate meta-analytical approach, submaximal HIIT was found to produce a decrease of -0.399 (95% CI [ 0.572, 0.226], p = 0.000) in blood glucose metrics, while supramaximal HIIT produced a similar, although non-significant, effect size of -0.366 (95% CI [-0.916, 0.184], p = 0.192). Across five of six outcome variables, submaximal HIIT had a significant small to medium effect, while supramaximal HIIT had a negligible to medium effect. For both HbA1c (%) and fasting blood glucose (FBG), supramaximal HIIT produced effect size values 48% and 67.4% larger, respectively, when compared to submaximal HIIT. Although none of the supramaximal effect sizes were statistically significant, those for HbA1c (%) and FBG trended towards significance (p = 0.059 and 0.082, respectively). The lack of supramaximal HIIT effect size statistical significance must be interpreted within the context of the small number (n = 2) of supramaximal HIIT studies, which limited the statistical power and decreased the likelihood of finding significant results. Graphing HIIT intensity and training load against HbA1c effect size revealed R2 values of 0.08 and 0.2, respectively, indicating that neither variable fully explains the observed variations in glycated hemoglobin.
ConclusionSubmaximal HIIT is an effective way of reducing markers of high blood glucose among individuals with T2DM. Supramaximal HIIT may offer greater blood glucose improvements but requires more research. Finally, neither exercise intensity nor training load alone can explain improvements to blood glucose homeostasis as measured by HbA1c.