Design
A prospective, parallel-group randomized control trial was designed. Mustafa Kemal University (MKU) Ethical Board approved the study (B.30.2.MKU.0.01.01.00/255). The exact nature of the study was explained to participants and they were asked to sign a written consent. The tests were carried out at Middle East Technical University Medical Center, Ankara, Turkey between January and June 2008. The independent variables were groups (n=4) and time (n=2). The dependent variable was serum cartilage oligomeric matrix protein (COMP) levels measured at three phases of exercise: before (Recovery), immediately after (Fatigue), and 30-min after the exercise (Regeneration). Participants were subjected to the following tests prior to and at the end of the intervention period: body mass index (BMI), isokinetic leg muscle strength, maximum oxygen consumption (VO2max) and serum COMP levels.
Participants
115 applicants volunteered to take part in this study. Eligibility screening was performed according to inclusion criteria. From 104 eligible applicants, 48 healthy, sedentary male university students aged between 18 to 25 years old (Mage = 21.8±1.9) were selected by a simple drawing. However, study completed with 44 participants, because of 4 dropouts. Participants were randomly assigned in 1:1:1:1 ratio to parallel groups, three exercise groups (swimming, cycling and running) and one control. In order to have a random allocation, each participant received a number unknown to researchers which was written on a piece of paper and then drawn from a box that allowed to assign individuals to the control and exercise groups. Baseline blood samples were stored in -80°C and they were not analyzed during participants' allocation. Exclusion criteria, surveyed by a questionnaire, were: osteoarthritis, rheumatoid arthritis or other inflammatory joint disease, intra-articular steroid injection, mal-alignment of the knees (varus/valgus) larger than 15°, and recent (within six months) fracture of lower extremity. Active athletes and volunteers with previous sports background were not included. During the intervention period, one participant was excluded from the study because of sudden kidney disease and another one decided to withdraw from further participation. In addition, two participants were excluded from the study before biochemical analyses since one subject from running group didn't meet minimum requirement of 75% of exercise session attendance, and one subject from control group didn't complete the posttests. Descriptive data of participants are presented in.
Interventions
After the first measurements, participants were randomly and equally assigned to the swimming, the running, the cycling and the control groups. All exercise groups participated in sessions of 40-min per day, three days per week, for a period of 12 weeks. Each session began with a five minutes warm-up, continued with a main set of 30-min exercise at their individual target heart rate zone (60-70% of heart rate reserve), and finished with a five minute cool-down period. The individual target heart rate zone was determined according to the Guidelines for Exercise Testing and Prescription of American College of Sport Medicines \[1\]. Swimming exercises composed of front crawl swimming and kicking drills. During the main sets, the participants performed front-crawl swimming at their individual heart rate zones which were determined by the Karvonen formula \[2\]. Cyclers exercised between 60-80 repetition per minute (RPM) on an ergometer (Monark E 834, Varberg, Sweden), which was adjusted for each individual in order to maintain the target heart rate zone. The running group participants exercised on a treadmill with a 1.5 % incline. The speed of the treadmill was determined according to individual heart rate zone. Throughout the 12 weeks of training, Heart Rate Reserves (HRR) was held constant (60-70%) while the speed increased according to individual progression. Individuals of the control group were instructed not to participate in any organized or structured exercise during the 12 weeks of intervention.
Physical and Physiological Measurements
Body Mass Index was calculated as the ratio of body weight to height square (Heyward \& Stolarczyk 1996). VO2max was measured by the standard Bruce Protocol (Bruce, Kusumi \& Hosmer 1973) on a Jager LE 200 CE (Hochberg, Germany) treadmill. Measurements were performed with the VIASYS Healthcare ergospirometry line using the MasterScreen CPX (Wuerzburg, Germany) device. Results were presented in ml/kg/min (milliliters of oxygen per kilogram of body weight per minute). Isokinetic quadriceps strength was recorded with the Biodex System Dynamometer (Biodex Medical Inc, Shirley, NY). Participants were placed in a comfortable position that allowed unrestricted motion of the knee from 90 degrees of flexion to terminal extension on a bicycle ergometer, and asked to warm-up for five minutes at which point they stretched their extremities. Before the test trials, participants were instructed to perform their maximum efforts. Then five isokinetic concentric knee flexion and extension at 60°/sec with their dominant legs were recorded. After, peak torque to body weight ratio (PT/BW) was measured.
Blood Sampling and Enzyme-Linked Immunosorbent Assay (ELISA) Procedures
Blood obtaining procedures were modified from Mundermann et al. \[3\]. Participants were asked to limit their physical activity 48 hours prior to the experiment. On the day of the experiment, participants made their breakfast within one hour after waking up and the experiment started two to three hours after breakfast. In addition, participants were seated on a chair for 15-min immediately before the experiment. Five-milliliter blood samples were obtained by a certified research nurse from the same antecubital vein immediately before, immediately after, and 0.5 hours after a 30-min walking exercise. During the walking exercise of the test protocol, participants walked at a pace of 5 km/h on a treadmill with 1.5 % incline. After the walking exercise, participants sat on an office chair for 30-min and were asked to rest.
Blood sample collection and storage procedures were described previously \[4\]. Venous blood samples were obtained from vena mediana cubitii. After clotting for 60 min at room temperature, samples were centrifuged at 5000 rpm for 10-min. The serum samples were stored at -80°C until analysis.
Serum COMP concentrations were analyzed with a commercially available enzyme-linked immunosorbent assay based on two monoclonal antibodies: COMP ELISA and AnaMar Medical (Sweden). ELISA analysis was performed in a private institution and assay procedures were followed according to the manufacturer's guidelines. The order of blood samples was mixed and biochemist was blind to allocation of samples. All blood samples were analyzed simultaneously to maintain uniform specificity across individuals.
Statistical Procedures
The effects of 12 weeks interventions on general fitness and serum COMP concentrations at pre-tests and post-tests were analyzed using 4x2 (groups and time) mixed repeated measures ANOVA for each variable. Paired-samples t-tests were employed for comparison of the serum COMP concentrations during rest, fatigue and recovery states. In order to decrease the probability of committing type-1 error, a method of multiple comparisons by Bonferroni confidence interval adjustment was performed.
Sample size calculation was performed to eliminate type II error. Sample size was calculated using the G\*Power 3.1.3 (Heinrich-Heine-University) software \[5\]. Significance level was set at 0.05 and a medium effect size (0.25) with two-sided calculation was used. Effect size contains the magnitude of difference between groups and values of 0.25, indicates medium effect of treatment \[6\]. By these assumptions, to achieve a power of 0.80, a sample size of 36 participants was necessary. Conservatively, 48 subjects were recruited since we expected a 20% dropout rate due to the demanding nature of the intervention program and our previous experience.
