2.2.1. Body Composition

Body height (Martin metal anthropometer, GPM Anthropological Instruments, Zurich, Switzerland) and body mass (medical electronic scale, A&D Instruments Ltd., Abingdon, UK) were measured to the nearest 0.1 cm and 0.05 kg, and body mass index (BMI) was also calculated (kg/m2). Body composition was measured by dual-energy X-ray absorptiometry (DXA) using the DPX-IQ densitometer (Lunar Corporation, Madison, WI; USA) Participants were scanned in light clothing while lying flat on their backs with arms on their sides. Whole body fat percent (body fat %), FM, and lean body mass (LBM) values were obtained. All DXA measurements and results were evaluated by the same examiner. The coefficient of variations (CVs) for the obtained results was less than 2% [25].

#### 2.2.2. Resting Energy Expenditure and Aerobic Performance

Resting energy expenditure (REE) was measured in the morning after an overnight fast. Participants were instructed to avoid any intense physical activity for the 24 h period before REE measurement. After voiding, subjects laid down for 15 min before the measurement of oxygen consumption (VO2) and carbon dioxide (VCO2) production over 30 min. The first 5 min and last 5 min of the measurement were discarded to ensure adequate measurement [26]. A portable open circuit spirometry system (MetaMax 3B, Cortex Biophysic GmbH, Leipzig, Germany) was used, data were stored at 10 s intervals, and the mean of the 20 min was used to calculate REE according to Weir's equation [27]: Basal metabolic rate (BMR) (kcal/min) = 3.9 [VO2 (l/min)] + 1.1 [VCO2 (l/min)], and REE (kcal/day) = BMR × 1440 min.

Maximal aerobic performance was determined by a stepwise incremental exercise test until volitional exhaustion using an electrically braked bicycle ergometer (Corival V3; Lode, Netherlands) [28]. The initial work rate was 40 W, and the stage increment was 35 W every 3 min until the maximal voluntary exhaustion was reached. The test was designed to elicit maximal power output at approximately 15–18 min for each subject [28]. Pedaling frequency was set to 60–70 rpm. Participants were strongly encouraged to produce the maximal effort. Respiratory gas exchange variables were measured throughout the test using breath-by-breath mode with data being recorded at 10 s intervals. Subjects breathed through a facemask. Oxygen consumption, carbon dioxide output and minute ventilation were continuously measured using a portable open-air spirometry system (MetaMax 3B, Cortex Biophysic GmbH, Leipzig, Germany). The analyzer was calibrated with gases of known concentration before the test according to the manufacturer's guidelines. All data were calculated by means of computer analysis using standard software (MetaMax-Analysis 3.21, Cortex, Leipzig, Germany). Peak oxygen consumption was measured, and maximal aerobic performance was defined as described previously [28].

#### 2.2.3. Blood Analysis

Venous blood samples were drawn between 8:00 and 9:00 a.m. after an overnight fast from an antecubital vein with the participants sitting in an upright position. Blood serum was separated and frozen at −80 ◦C for further analyses. Irisin was determined using an enzyme-linked immunosorbent assay (ELISA) kit using a specific Irisin/FDNC5 monoclonal antibody (R&D Systems Inc., Minneapolis, MN, USA) [29]. This assay had intra- and inter-assay CVs of 2.5% and 8.7%, respectively, and the least detection limit was 0.25 ng/mL. Fibroblast growth factor-21 (FGF-21) was assessed by a commercially available ELISA kit (R&D Systems Inc., Minneapolis, MN, USA) with a minimum detectable level of 1.61 pg/mL, and intra-assay CV 3.5% and inter-assay CV 5.2%. Leptin was determined by Evidence® Biochip Technology (Randox Laboratories Ltd., Crumlin, UK) with the intraand inter-assay CVs of 4.6% and 6.0%. Resistin was also measured by Evidence® Biochip Technology (Randox Laboratories Ltd., Crumlin, UK) with the intra- and inter-assay CVs of 5.2% and 9.1%.

### *2.3. Statistical Analysis*

Data analysis was performed using the SPSS software version 21.0 package for Windows (Chicago, IL, USA). Standard statistical methods were used to calculate means and standard deviations (±SD). Evaluation of data normality was performed with the Kolmogorov-Smirnov method. Data that were not normally distributed were logarithmically transformed prior to analyses to approximate a normal distribution. This was necessary for body FM and serum leptin values. Statistical comparisons between the groups were made using an independent t-test. In addition, effect size (ES, eta squared) thresholds of 0.01, 0.06, and 0.14 were used to identify small, moderate, and large differences, respectively, to define the magnitude of the difference [30]. Pearson correlation coefficients were calculated to assess linear relationships. In addition, partial correlation analyses controlling for age, body FM, and LBM were used to control for confounders [19]. The level of significance was set at *p* < 0.05
