Background and aims: Exercise training is considered a cornerstone in the management of type 2 diabetes, which is associated with impaired endothelial function. However, the association of exercise training with endothelial function in type 2 diabetes patients has not been fully understood. This meta-analysis aimed to investigate their associations with focus on exercise types. Methods: Databases were searched up to January 2018 for studies evaluating the influences of exercise training with durations ≥ 8 weeks on endothelial function assessed by flow-mediated dilation (FMD) among type 2 diabetes patients or between type 2 diabetics and non-diabetics. Data were pooled using random-effects models to obtain the weighted mean differences (WMDs) and 95% confidence intervals (CIs). Results: Sixteen databases were included. Exercise training resulted in an overall improvement in FMD by 1.77% (95% CI 0.94–2.59%) in type 2 diabetes patients. Specifically, both aerobic and combined aerobic and resistance exercise increased FMD by 1.21% (95% CI 0.23–2.19%) and 2.49% (95% CI 1.17–3.81%), respectively; but resistance exercise only showed a trend. High-intensity interval aerobic exercise did not significantly improve FMD over moderate-intensity continuous exercise. Notably, the improvement in FMD among type 2 diabetes patients was smaller compared with non-diabetics in response to exercise training (WMD − 0.72%, 95% CI − 1.36 to − 0.08%) or specifically to aerobic exercise (WMD − 0.65%, 95% CI − 1.31 to 0.01%). Conclusions: Exercise training, in particular aerobic and combined exercise, improves endothelial function in type 2 diabetes patients, but such an improvement appears to be weakened compared with non-diabetics. Trial registration PROSPERO CRD42018087376

Background: Prediabetes progression is associated with increased mortality while its regression decreases it. It is unclear whether muscle strength is related to prediabetes progression or regression. This study investigated the associations of muscle strength, assessed by grip strength and chair‐rising time, with prediabetes progression and regression based on the China Health and Retirement Longitudinal Study (CHARLS) enrolling middle‐aged and older adults.Methods: We included 2623 participants with prediabetes from CHARLS, who were followed up 4 years later with blood samples collected for measuring fasting plasma glucose and haemoglobin A1c. Grip strength (normalized by body weight) and chair‐rising time were assessed at baseline and categorized into tertiles (low, middle, and high groups). Prediabetes at baseline and follow‐up was defined primarily using the American Diabetes Association (ADA) criteria and secondarily using the World Health Organization (WHO) and International Expert Committee (IEC) criteria. Multinomial logistic regression analysis was applied to obtain the odds ratios (ORs) and 95% confidence intervals (CIs).Results: The mean age of included participants was 59.0 ± 8.6 years, and 46.6% of them were males. During follow‐up, 1646 participants remained as prediabetes, 379 progressed to diabetes, and 598 regressed to normoglycaemia based on ADA criteria. Participants who progressed to diabetes had lower normalized grip strength than those who remained as prediabetes (0.49 ± 0.15 vs. 0.53 ± 0.15, P < 0.001), but participants who regressed to normoglycaemia showed the opposite (0.55 ± 0.16 vs. 0.53 ± 0.15, P = 0.003). However, chair‐rising time was comparable across different groups (P_overall = 0.17). Compared with participants in low normalized grip strength or high chair‐rising time group, those in high normalized grip strength or low chair‐rising time group had decreased odds of progression to diabetes (OR 0.62, 95% CI 0.44 to 0.87; and OR 0.69, 95% CI 0.51 to 0.93, respectively) after multivariable adjustment. However, both were unrelated to the odds of regression to normoglycaemia (OR 0.94, 95% CI 0.71 to 1.25; and OR 0.84, 95% CI 0.65 to 1.07, respectively). These outcomes remained generally comparable when prediabetes was defined by WHO or IEC criteria. Higher normalized grip strength but not lower chair‐rising time was prospectively associated with lower blood pressure, better glycaemic condition, and lower inflammation (all P ≤ 0.04).Conclusions: High muscle strength is associated with reduced odds of progression to diabetes but does not predict regression to normoglycaemia in prediabetes. Future studies are warranted to assess whether increases in muscle strength promote prediabetes regression.

While step counter use has become popular among type 2 diabetes (T2D) patients, its effectiveness in increasing physical activity (PA) and improving glycemic control has been poorly defined. The aim of this meta-analysis of randomized controlled trials (RCTs) was to evaluate the association of step counter use with PA and glycemic control in T2D patients. Articles were identified by searches of PubMed, Web of Science and Cochrane Library from January 1994 to June 2013. RCTs in the English language were included, if they had assessed the effectiveness of step counters as motivating and monitoring tools in T2D patients, with reported changes in steps per day (steps/d) or glycosylated hemoglobin A1c (HbA1c), or both. Data were independently collected by 2 authors and overall estimates were made by a random-effects model. Of the 551 articles retrieved, 11 RCTs were included. Step counter use significantly increased PA by 1,822 steps/d (7 studies, 861 participants; 95% confidence interval (CI): 751 to 2,894 steps/d) in patients with T2D. Step counter use with a PA goal showed a bigger increase in PA (weighted mean difference (WMD) 3,200 steps/d, 95% CI: 2,053 to 4,347 steps/d) than without (WMD 598 steps/d, 95% CI: −65 to 1,260 steps/d). Further subgroup analysis suggested step counter use with a self-set PA goal (WMD 2,816 steps/d, 95% CI: 1,288 to 4,344 steps/d) made no difference in increasing PA from a 10,000 steps/d goal (WMD 3,820 steps/d, 95% CI: 2,702 to 4,938 steps/d). However, no significant HbA1c change was observed by step counter use (10 studies, 1,423 participants; WMD 0.02%, 95% CI: −0.08% to 0.13%), either with (WMD 0.04%, 95% CI: −0.21% to 0.30%) or without a PA goal (WMD 0.01%, 95% CI: −0.10% to 0.13%). Step counter use is associated with a significant increase in PA in patients with T2D. However, evidence regarding its effect in improving glycemic control remains insufficient. PROSPERO CRD42013005236

Although step counters are popularly employed for physical rehabilitation in chronic obstructive pulmonary disease (COPD) patients, their effectiveness is inconsistent and even questioned. This meta-analysis aimed to investigate whether step counter use increases physical activity or improves exercise capacity in COPD patients. Electronic databases were searched for randomized controlled trials that assessed the efficacy of step counter use in increasing physical activity or in improving exercise capacity. Data were aggregated using a random-effects model to get the overall effect sizes [standard mean difference (SMD) with 95% confidence interval (CI)], and subgroup analyses were performed. A total of 15 trials enrolling 1316 patients with moderate to severe COPD were included. Step counter use increased physical activity compared with controls (SMD = 0.57, 95% CI 0.31–0.84), which is equal to a magnitude of 1026 steps/day in daily steps. It also enhanced exercise capacity with an effect size of 0.30 (95% CI 0.16–0.45), approximating to a magnitude of 11.6 m in the 6-min walking distance. Step counter use could augment physical activity (SMD = 0.64, 95% CI 0.19–1.08) and exercise capacity (SMD = 0.32, 95% CI 0.01–0.62) for patients receiving pulmonary rehabilitation. Yet it cannot enhance physical activity or exercise capacity in patients with severe COPD or among studies with intervention durations ⩾6 months (both p > 0.50). Step counter use increases physical activity and improves exercise capacity in COPD patients, at least in the short term, which supports the notion of recommending step counter use in COPD management.