Exercise training with dietary restriction enhances circulating irisin level associated with increasing endothelial progenitor cell number in obese adults: an intervention study

Participants

Participants aged between 18 and 40 years were recruited from the Shenzhen Sunstarasia Weight Loss Camp. Participants were included that met the following initial eligibility requirements: (a) obesity status, as assessed by body mass index (BMI) ≥30 kg/m²; (b) absence of unstable angina pectoris, cardiomyopathy, severe lung diseases, or renal failure. Informed consent was obtained from each individual prior to measurements. This study was conducted according to the Declaration of Helsinki and was approved by the Ethics Committee of Guangzhou Sport University (approval No. GSU20160012). Participant recruitment and follow-up were conducted between March 2016 and September 2016. The trial was registered in ISRCTN registry (ISRCTN83594346). The protocol for this trial and supporting TREND checklist are available as supporting information (see S1 Protocol and S1 Checklist).

Eligible participants from a traditional weight loss camp (located at a remote district of Huizhou city, Guangdong, China) self-selected to join this program. The closed camp provided uniformly controlled accommodation, diet, and physical training during an 8-week intervention period. The camp participants were housed in the same building and could not come and go freely during their stay. The location, residential setting, and level of obesity (BMI = 37.8 ± 5.0 kg/m²) prevented identification of an appropriate control group.

Diet restriction

Participants were provided with calorie-restricted diets that contained 1,300–2,200 kcal/day based on their weight. During the study, the menu was changed weekly, and the diet was adjusted each week according to an individual’s updated weight. The energy percentages provided by protein, fat, and carbohydrate were 20%, 20%, and 60%, respectively. Energy distributions at breakfast, lunch, and dinner were 30%, 40%, and 30%, respectively. All meals were prepared and supervised by registered professional dietitians during the diet intervention.

Exercise training

Subjects performed a training program six days/week for eight weeks. The program was primarily comprised of endurance exercise such as bicycling, walking, running, dancing, and ball games for five hours/day. It was also supplemented by strength training. The endurance exercises involved an equivalent combination of moderate (i.e., four metabolic equivalents of tasks (METs)) and high intensity (eight METs) physical training. We used the following equation to calculate the exergy expenditure during activity (Pinheiro Volp et al., 2011): energy expenditure (kcal/min) = 0.0175 × METs × weight (kg). The exercise program was specifically designed to induce an energy expenditure of 1,500–2,500 kcal/day. The intensity of moderate-intensity exercise was set at 70–85% of the subject’s maximum heart rate (HR_max), which was calculated via the formula of 208 − (0.7× age). Heart rate was continuously monitored by Polar heart rate monitors and recorded by researchers. The high-intensity training (∼90% of HR_max) was alternated with low-intensity exercise (∼60% of HR_max) during training. Strength training was implemented at 40–50% maximal strength for 2–3 sets of 12–15 RM with 2–3 min of rest between sets. Qualified trainers supervised the subjects during the program. All measurements were obtained before and after the 8-week training program.

The primary outcome measure for this study was endothelial function, which was assessed via flow-mediated dilation (FMD). Secondary outcomes were anthropometry, resting heart rate, blood pressure, body composition, aerobic fitness, maximal strength, EPC number and function, as well as multiple biochemical markers (lipid profiles, insulin, tumor necrosis factor-alpha, high-sensitivity C-reactive protein, superoxide dismutase, vascular endothelial growth factor, endothelial nitric oxide synthase, adiponectin, and irisin).

Anthropometry and body composition

Height and weight were measured to calculate the BMI (kg/m²). Body composition was determined using a body composition analyzer (Inbody 370; Biospace, Seoul, Korea).

Resting heart rate and blood pressure

After subjects rested for at least 10 min, resting heart rate and peripheral brachial systolic/diastolic blood pressure were measured in triplicate. The average of three readings was recorded.

Aerobic fitness

Aerobic fitness was assessed using the Physical Working Capacity test on a cycle ergometer (Ergoselect 100; Ergoline, Bitz, Germany) at a heart rate of 150 or 170 beats/min (PWC150 or PWC170). Seat height was adjusted to individual satisfaction of each subject, and toe clips with straps were used to prevent feet from slipping off the pedals. Prior to the tests, subjects were instructed that they should pedal with a steady cadence of 60 revolutions/min. The start power was 50 watts (W) and was followed by a gradual increase of 25 W for PWC150 or 50 W for PWC170 every 2 min until the targeted heart rate (150 or 170 beats/min, respectively) was achieved and maintained at a steady state. PWC150 or PWC170 were calculated as the power corresponding to heart rate of 150 or 170 beats/min and expressed as W per kg of body mass (W/kg).

Maximum strength test

The maximum strength was determined by a one-repetition maximum (1RM) bench press test, which was administered after each subject performed two warm-up sets. After all warm-up sets were completed, the subject then attempted the 1RM.

Endothelial function

Subjects were requested to fast, abstain from exercise and the consumption of alcohol and caffeine, and withhold all medications and supplements known to affect vascular function for at least 12 h prior to testing. They were asked to rest in a quiet and air-conditioned room (22–25 °C) in a supine position for 30 min right before examination. Ultrasound equipment and a high-resolution linear array transducer coupled to computer-assisted analysis software provided one longitudinal and two short-axis images using a 10 MHz H-type probe (UNEXEF38G; UNEX, Nagoya, Japan). This was used to scan the brachial artery in B-mode 5 to 10 cm above the right elbow. This location was marked on the skin of each participant and all subsequent measurements were performed at the same location. When the clearest B-mode image of the intima-media complex had been obtained, a stereotactic probe holder held the transducer at the same point throughout the scan. FMD was measured via A-mode waves as a signal of the intima-media complex that was synchronized with the electrocardiographic R-waves and automatically tracked. After measuring baseline brachial artery diameter, we compressed the brachial artery (at least 50 mm Hg above systolic blood pressure) for 5 min, using a blood pressure cuff placed around the forearm. After compression, the maximum brachial artery diameter was measured after cuff release for 2 min (Miyagi et al., 2014; Stoner & Sabatier, 2012; Tomiyama et al., 2008). The brachial artery peak hyperemic shear rate was calculated as eight times the peak velocity divided by the diameter at the time of peak velocity.

FMD was calculated as the maximum percentage increase in arterial diameter during continuous measurement of arterial diameter following cuff deflation. Subsequently, endothelium-independent dilation was measured by sublingually administering 500 µg of nitroglycerine. The menstrual cycle of female subjects was recorded and they were asked to test on day 1–7 of the menstrual cycle. The same operator collected all measurements.

In our laboratory, we performed repeated FMD measurements on seven healthy adults (20–35 years old, 57% male). The intraclass correlation coefficient for repeated readings was r = 0.88 with a coefficient of variation of 11.4%.

Blood markers

A fasting blood sample was collected into evacuated plastic tubes containing ethylenediaminetetraacetic acid (EDTA). Whole venous blood was also collected in tubes without anticoagulant for serum preparation. Total cholesterol, triglycerides, high-density lipoprotein cholesterol (HDL-c), low-density lipoprotein cholesterol (LDL-c), fasting glucose, and fasting insulin were measured. Insulin resistance was evaluated using the Homeostasis Model of Assessment of Insulin Resistance (HOMA-IR) and was calculated as (fasting insulin (µU/ml) × fasting glucose (mmol/L)) / 22.5. Serum concentrations of vascular endothelial growth factor (VEGF), endothelial nitric oxide synthase (eNOS), adiponectin, tumor necrosis factor-alpha (TNF-α), high-sensitivity C-reactive protein (hsCRP), and superoxide dismutase (SOD) were analyzed using ELISA Kits (Cusabio, Biotech. Co., Ltd., Wuhan, China), following the manufacturer’s instructions for each kit. Serum irisin concentrations were measured using an irisin ELISA kit (Cat. EK-067-29; Phoenix Pharmaceuticals, Burlingame, CA, USA), which had previously been validated via MS spectrometry analysis (Huh et al., 2015; Polyzos & Mantzoros, 2015; Zhang et al., 2014). Cross reactivity was validated via western blot. The minimum detectable concentration was 1.43 ng/ml. Intra- and interassay variations were below 10% and 15%, respectively (Benedini et al., 2017).

Flow cytometric quantification of EPCs

EPCs are defined as CD34⁺/KDR⁺ cells. EDTA-anticoagulated whole blood samples (100 µl) were stained for 10 min at room temperature with APC-labeled anti-human CD34 (eBioscience, San Diego, CA, USA) and AlexaFlour488-labeled anti-human KDR (BioLegend, San Diego, CA, USA) monoclonal antibodies. Fluorescent isotype-matched antibodies were used as controls. EPCs were measured using a Cytomics FC500 flow cytometer (Beckman Coulter, Brea, CA, USA), and acquisition was stopped after 40,000 events. Data were analyzed using CXP software 2.0 (Fernandes et al., 2012).

EPC culture

EPCs were isolated and cultured, following previously described protocols (Fernandes et al., 2012; Junhui et al., 2008; Tobler et al., 2010). Briefly, peripheral blood (10 ml) was obtained from obese subjects, and total mononuclear cells were isolated via density gradient centrifugation (400 g for 30 min) with Histopaque-1077 (Sigma, St. Louis, MO, USA). Then, cells were cultured on fibronectin-coated 6-well plates in M199 medium supplemented with 20% FBS, 100 U/ml penicillin, 100 µg/ml streptomycin, and 10 ng/ml VEGF. After four days in culture, nonadherent cells were removed. Adherent cells were maintained until day seven and then used for EPC functional assays.

EPC migration assay

EPC migration was determined with a modified Boyden chamber (Costar Transwell assay, 8 µm pore size; Corning, NY, USA) (Junhui et al., 2008; Xia et al., 2012). Briefly, a total of 2 × 10⁴ isolated EPCs were resuspended in 250 µl serum-free M199 medium and placed in the upper chamber. The chamber was placed in a 24-well culture dish containing 500 µl M199 medium supplemented with 10 ng/ml VEGF. After a 24-hour incubation at 37 °C, the membrane was briefly washed with phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde. The membrane was then stained using 0.1% crystal violet solution and carefully removed. The transmigrated cells were manually counted in three random microscopic fields (×200) by independent, blinded investigators.

EPC adhesion assay

The EPC adhesion assay was conducted according to previously described techniques (Junhui et al., 2008; Xia et al., 2012). Briefly, EPCs were isolated and resuspended in M199 medium with 5% FBS. An equivalent amount of cells were placed on a fibronectin-coated 96-well plate and incubated at 37 °C in 5% CO₂ for 6 h. After two gentle washes with PBS, adherent cells were counted.

Statistical analysis

Analyses were performed using SPSS 16.0 (SPSS Inc., Chicago, IL, USA). Paired-sample t-tests were used to compare the effects of an 8-week intervention of exercise and diet on different variables. Pearson’s correlation was calculated to determine associations. Values are reported as mean ± SD. Sample size was calculated based on a previous study that reported a 1.11% change in FMD after each 10 kg weight loss (Joris, Zeegers & Mensink, 2015), with an expected within-subject SD of 1.3% in FMD from our laboratory. The sample size required for the study (with an α of 0.05 and β of 0.20 at 5% level of significance (two-sided) and estimating a refusal rate of 20%) was 20 subjects. Cohen’s d was used to calculated effect sizes on pairwise comparisons, with 0.2, 0.5, and 0.8, representing small, medium, and large effects, respectively (Mullineaux, Bartlett & Bennett, 2001). A P-value <0.05 was used to indicate statistical significance.

Participants

The flow of participants through each stage of the intervention is shown in Fig. 1 according to the TREND statements. Of the 29 initially enrolled subjects, 22 met the criteria for inclusion determined in the study and were selected for the research program. We only analyzed data from subjects that completed the program and provided valid measurements at both baseline and follow-up.

Anthropometry and body composition

To determine how the exercise and diet intervention affected overall health, we measured body composition. As shown in Table 1, body weight significantly decreased from 116.6 ± 24.6 kg to 103.4 ± 22.4 kg (P < 0.001, d =  − 0.56). BMI decreased from 37.8 ± 5.0 kg/m² to 33.5 ± 4.4 kg/m² (P < 0.001, d =  − 0.91) following the 8-weeks of exercise and dietary intervention. We also observed a decrease in body fat mass from 47.4 ± 12.7 kg to 38.2 ± 12.2 kg (P < 0.001, d =  − 0.74) following the intervention (Table 1). Furthermore, body fat percentage decreased from 40.9 ± 4.6% to 36.6 ±5.9% (P < 0.001, d =  − 0.81) after the intervention (Table 1). These results suggest that the intervention was effective in reducing body fat and improving BMI in obese subjects.

Resting heart rate and blood pressure

The intervention of exercise and diet significantly decreased the resting heart rate from 74 ± 9 bpm to 61 ± 9 bpm (P < 0.001, d =  − 1.44) and significantly decreased systolic blood pressure from 127.8 ± 11.8 mmHg to 115.3 ± 7.4 mmHg (P < 0.001, d =  − 1.27) (Table 1). However, no significant change was detected in diastolic blood pressure (before: 85.2 ± 9.4 mmHg vs. after: 79.9 ± 10.8 mmHg, P > 0.05) (Table 1).

Aerobic fitness and maximal strength

We observed an increase in PWC150 from 1.38 ± 0.41 W/kg to 1.74 ± 0.32 W/kg (P < 0.001, d = 0.98) following the intervention. Interestingly, although PWC170 had an increasing trend from 1.99 ± 0.50 W/kg to 2.65 ± 1.03 W/kg after intervention, but this trend did not reach significance (P = 0.06) (Table 1). There was also no significant change in 1RM bench press resulting from the intervention (before: 133 ± 61 lbs vs. after: 140 ± 60 lbs, P > 0.05) (Table 1).

Blood chemistry

Blood lipid levels can indicate a risk of pathological diseases, such as atherosclerosis and diabetes. Thus, we measured the changes in circulating lipids following the intervention. We observed a significant decrease in the serum levels of cholesterol (before: 5.35 ± 0.92 mmol/l vs. after: 4.74 ± 1.34 mmol/l, P < 0.05, d =  − 0.53), triglycerides (before: 2.15 ± 1.08 mmol/l vs. after: 1.32 ± 0.88 mmol/l, P < 0.01, d =  − 0.84), LDL-c (before: 3.44 ± 0.75 mmol/l vs. after: 2.99 ± 0.95 mmol/l, P < 0.05, d =  − 0.53), and fasting insulin (before: 200.8 ± 115.8 pmol/l vs. after: 116.7 ± 81.2 pmol/l, P < 0.01, d =  − 0.84) (Table 2). Analysis of insulin resistance was measured via HOMA-IR. Our results revealed that insulin resistance significantly decreased after the intervention (P < 0.01, d =  − 0.81) (Table 2). Although we observed a decreasing trend in fasting blood glucose from 5.61 ± 0.56 mg/dl to 5.39 ± 0.57 mg/dl after the intervention, this change did not reach significance (P > 0.05). There was no significant change in HDL-c levels resulting from the intervention (before: 0.99 ± 0.14 mmol/l vs. after: 0.99 ± 0.22 mmol/l, P > 0.05) (Table 2). We also measured circulating adiponectin because it is an adipocyte-secreted hormone associated with improved insulin sensitivity and the amelioration of the metabolic syndrome. We found that diet and exercise significantly increased adiponectin levels in obese subjects (P < 0.05, d = 0.80). Two important markers for endothelial function are VEGF and eNOS, and both VEGF and eNOS were significantly improved after the intervention. As shown in Table 2, we also observed a significant increase in circulating levels of irisin after the intervention (before: 43.9 ± 11.0 ng/ml vs. after: 53.9 ± 13.4 ng/ml, P < 0.05, d = 0.82). Moreover, the intervention of diet and exercise markedly improved inflammation and oxidative stress, which was reflected in a reduction of TNF-α and hsCRP levels and an increase in SOD levels (Table 2).

Endothelial function

Figure 2A shows that FMD significantly improved from 7.28 ± 1.89% to 8.33 ± 1.71% (P < 0.05, d = 0.58) after the intervention. The brachial artery peak hyperemic shear rate did not differ in response to the intervention (before: 141.0 ± 84.1 s⁻¹ vs. after: 177.3 ± 128.3 s⁻¹, P > 0.05) (Table 1). Moreover, the intervention had no significant impact on endothelium-independent dilation induced by nitroglycerine (before: 14.16 ± 4.96% vs. after: 13.18 ± 4.87%, P > 0.05) (Fig. 2B).

Circulating EPC levels

As shown in Fig. 3, we detected a significant increase in circulating EPCs after the intervention (before: 0.028 ± 0.018% vs. after: 0.065 ± 0.048%, P < 0.05, d = 1.02).

EPC functions

To assess the functional activity of EPCs, we determined the migration of isolated EPCs in response to VEGF using a modified Boyden chamber. As illustrated in Fig. 4A, there was a significant improvement in the migratory capacity of EPCs after intervention (before: 42.0 ± 4.08 cells/ ×200 field vs. after: 55.67 ± 2.73 cells/ ×200 field, P < 0.001, d = 3.93). Moreover, we assessed the adhesive capacity of EPCs because adhesion to the extracellular matrix is believed to be important during new blood vessel growth. As shown in Fig. 4B, the intervention also significantly improved EPC adhesive function (before: 25.75 ± 4.11 cells/ ×200 field vs. after: 40.50 ± 10.46 cells/ ×200 field, P < 0.05, d = 1.86).

Correlations

A significant correlation between the increase of circulating irisin levels and the increase in the number of EPCs was observed by the intervention (r = 0.52, P < 0.05) (Fig. 5).