Participants
Participants were recruited by convenience from schools and two soccer clubs. Two groups of asymptomatic male adolescents were included: soccer players and adolescents not performing similar impact physical activity (control group), aged 13 to 18 years. The inclusion criteria for the asymptomatic adolescent soccer players were to have practiced the sport for at least two consecutive years with a minimum frequency of three times per week for a period of one to three consecutive hours. All soccer players were enrolled in the State Football Federation. The control group contained asymptomatic adolescents and was matched for age, gender, height, and weight. The inclusion criterion for the controls were an absence of soccer or similar sport more than once a week for more than one hour.
The exclusion criteria for both groups were a history of lesions, surgery or osteoarticular alterations, malformations, history of chronic disease (diabetes; hypertension; rheumatic, cardiac, renal, respiratory or neurological disease; chronic hepatitis disease; or a BMI Z-score greater than + 3. Thirty individuals in each group were invited to participate in the study. Five participants were excluded from the athletes due to previous surgery, while on control group 2 participants were excluded due to previous injuries, 5 for soccer-related sports participants and 3 due to symptoms.
The anthropometric evaluation measured weight and height for further body mass index calculations. The patients were weighed in light clothing, barefoot, and positioned standing in the center of the scale. Height was measured by positioning the barefoot patient in the center of the equipment with his head free of props while standing upright with heels attached and arms extended along the body according to the Frankfort horizontal plane. The heels, shoulders, and gluteal muscles were in contact with the stadiometer. The nutritional status of children and adolescents was classified according to the BMI Z-score [23]. The presence or absence of symptoms was evaluated via an interview with the researcher, who asked about the presence of pain, functional limitations, mechanical symptoms, and any discomfort in the lumbar column in the prior six months. In this interview, the level of physical activity was evaluated using the physical activity questionnaire entitled Questionnaire on Physical Activity for Adolescents (QAFA). This questionnaire is composed of 24 questions and has been validated for use by young Brazilians [24].
Spine evaluation
The images were collected on 1.5 T devices with multi-channel coils, e.g., Achieva model equipped with 8-channel coils (Koninklijke Philips, Best, The Netherlands) and Magnetom Aera model (Siemens GmbH, Erlangen, Germany) equipped with 24-channel coil. Weighted fast spin-echo T1 sagittal sequences were obtained (TR:400 ms to 600 ms; TE: 6.3 ms to 15 ms), T2-weighted fast spin-echo sagittal (TR: 2200to 4500 ms/TE: 60 ms to 110 ms), fast sagittal in T2 with fat saturation (TR: 2200 to 4500 ms/TE: 60 ms to 110 ms) with the following minimum image parameters: matrix of 320 × 256; cutting thickness of 4.0 mm with a 1.0 mm interval; 30 cm field-of-view. Oblique fast spin-echo axial images were also acquired with the following parameters: 320 × 224 matrix; cutting thickness of 4.0 mm with a 1.0 mm interval; and 20 cm field-of-view.
Two radiologists with the title of specialist from the Brazilian College of Radiology with more than 5 years of experience in the evaluation of vertebral column images made an independent analysis of the images. The tests were randomly interpreted via double blinded analysis. Each MRI test was analyzed for the presence or absence of plateau edema, protrusions, and disc extrusions (sequestered or not). The evaluation also included stress reactions in the pedicles, spondylolysis, spondylolisthesis, hypertrophic alterations in the interfaces, ligament edema, and muscle edema. Individual analysis was performed by each radiologist (inter-observer). An intra-observer analysis was performed by one of the evaluators with an interval greater than two months for intra-observer variation calculations.
Muscular and bone ligament edema was characterized by a pre-lesion, i.e., an alert sign of the development of a structured lesion of the lumbar spine. Edema evaluation was restricted to the presence of signal alterations including the type of edema pattern in the projection of interspinous spaces, vertebral plateau, and paraspinal musculature. The diagnosis of displacements due to disc hernias was based on the criteria proposed by the working group from the North American Spine Society, the American Society of Spine Radiology, and the American Society of Neuroradiology [25].
Intervertebral discs are physiologically prominent in the age group of our subjects; thus, only cases in which there was some degree of disc degeneration associated with altered disc contours were considered. This avoids an overdiagnosis of bulging. Considering the age of the patients, any hypertrophy, sclerosis, or irregularity of the interface joints was also considered to be a positive finding. The presence of edema or fracture of the pedicles was considered positive for the diagnosis of a stress reaction. The occurrence of sclerosis was not considered. The presence of arthritis of the interfaces was inferred via the presence of synovitis in the facets or joint effusion.
The groups of lesions were classified as follows: Warning signs included edemamuscular and bone ligament edema (vertebral plateau, pedicle) and instituted lesions. Disc hernia include protrusion, extruded, or sequestered discs as well as spondylolysis, spondylolisthesis, and facet arthrosis/arthritis. Disc hernia of anterior lesions included protrusion, extruded, or sequestered discs. Posterior lesions include spondylolysis, spondylolisthesis, and facet arthrosis/arthritis.
Statistical analysis
The age variable was symmetric, and the between-group comparisons used a t-test for independent samples. The BMI z-score was measured via asymmetric behavior, and the groups were compared via the Mann-Whitney test. We compared the percentage of lesions between the groups via Pearson’s chi-squared test except for anterior lesions that used Fischer's exact test. The inter and intra-observer analysis (evaluators 1 and 2) were compared via the Kappa test. Data analysis used SPSS software® version 19.0.
Considering the fixed sample size of 25 subjects in the group of athletes and 20 in the control group, the proportion of lesions in each of the groups was 76% and 35%, respectively. The power of the test to compare the groups was 90% at a significance level of 5%.