Single nucleotide polymorphisms associated with non-contact soft tissue injuries in elite professional soccer players: influence on degree of injury and recovery time

Data were collected on injuries suffered by 73 elite soccer players from Futbol Club Barcelona (Barcelona, Catalunya, Spain) over the course of three consecutive soccer seasons. All the players included in the study, either from the first (n = 49) and the second team (n = 26) lived within 30 km of the training field, and were thus subjected to the same climate and environmental conditions. All the players undertook similar amount of work, followed similar diet (data not available), and took the same ergogenic aids. The training field, the playing fields and the injury prevention protocols were also identical for all the players. The treatment protocol for each type of injury, including medication and physical therapy, was the same for all the players, and all treatment was supervised by the same medical team.

Given the high qualification level of the study population (73 male professional soccer players from the same football team) and based on the sample size (n = 242) of the NCSMTIs encountered, we decided to study the most common injury in each tissue group, i.e. hamstring injuries for muscle injuries, patellar tendon injuries for tendon injuries, and medial collateral ligament injuries for ligament injuries. Data on injuries were collected in accordance with the Union of European Football Associations (UEFA) protocol [30]. Ultrasound and magnetic resonance imaging scans were used to morphologically classify the injuries by anatomic region. Injuries were classified as mild, moderate or severe [31] according to the number of days that a player needed to be absent from training and/or competition [32, 33]: mild, 1–15 days; moderate, 16–30 days; and severe, more than 30 days. A mild lesion presents minimal tissue damage (up to 25%), a moderate injury involves 50% of the tissue and, finally, in a serious injury more than 50% of the tissue is involved. Recovery time was defined as the time from the date of injury until the date the player could return to full training or competition.

Approximately 4 mL of whole blood was collected from each subject into EDTA vacutainer tubes, and stored at 4°C until total DNA extraction. Genomic DNA from whole blood was isolated using QIAmp DNA Blood Minikit (Qiagen, Valencia, CA) following the manufacturer’s instructions. To measure DNA quantity, a NanoDrop ND-1000 Spectrophotometer (Thermo Fisher Scientific INC, Waltham, MA) was used. DNA was stored at −20°C until analysed.

Table 1 shows the characteristics of all the SNPs analyzed. Primers and probes were obtained from Applied Biosystems (AB; Assays-on-Demand SNP genotyping product, Foster City, CA).

SNP analysis was performed using a real-time polymerase chain reaction (PCR) Allelic Discrimination TaqMan Assay (AB) with minor modifications. All PCR reactions were run in duplicate, and contained 50 ng of each individual’s DNA; 6.25 μL TaqMan Universal Master Mix (AB); 0.25 μL primers and probes (AB) and water for a final volume of 12.5 μL. Real-time PCR was performed on an ABIPrism 7500 Sequence Detection System (AB) using the following conditions: 50°C for 2 minutes, 95°C for 10 minutes, and 40 cycles of amplification (95°C for 15 seconds and 60°C for 1 minute). For each cycle, the software determined the fluorescent signal from the VIC- or FAM- labeled probe.

Descriptive statistics of the main demographic variables of the studied population (mean, standard deviation, or median, range, for continuous data, and frequency tables for specific data) was calculated. Frequency tables were used for the distribution of the SNPs for the different genes evaluated.

The association between type and degree of injury and the SNPs (in ELN, TTN, SOX15, IGF2, CCL2, COL1A1, COL5A1 and TNC) was determined with the Chi-square test and Fisher’s exact test when necessary. The association between SNPs and injury recovery time was evaluated using multivariate analysis of variance. All statistical analyses were performed using SPSS version 14.0 for Windows (SPSS Inc., Chicago, IL). Significance was set at P ≤ 0.05. The Benjamini-Hochberg P-value corrective test for multiple comparisons was applied.

Of the 73 soccer players included in the study, 43 (58.9%) were White, 11 (15.1%) were Black Africans, and 19 (26%) were Hispanics. The median age for all players was 26.2 years (range, 19–35), median weight was 75.6 kilos (range, 64–92), median height was 1.79 meters (range, 1.66-1.95), and median work load was 16027 minutes per year (267 hours/season/player) (range, 15301–16544), with no significant differences between ethnic groups for any of these characteristics.

Over the course of the three seasons of the study, a total of 242 NCMSTIs were recorded for all 73 players. Two hundred and three were muscle injuries, of which 129 (63.5%) were mild, 69 (34%) moderate, and 5 (2.5%) severe. Twenty-four were ligament injuries, of which 15 (62.5%) were mild, 3 (12.5%) moderate, and 6 (25%) severe. Fifteen were tendon injuries, of which 7 (46.7%) were mild, 7 (46.7%) moderate, and 1 (6.6%) severe.

Table 2 shows the allele frequencies of the eight genes, both for the present study and according to the NCBI dbSNP. The frequency of the SNPs varied among the three sub-groups in the present study (p < 0.0001).

Table 3 shows the association between SNPs and injured structure. SNPs in IGF2 and CCL2 were associated with the severity of muscle injuries. The 93 muscle injuries associated with the IGF2 GC genotype were significantly less severe than those associated with the IGF2 CC or GG genotypes (p = 0.034) (Figure 1A, Table 3). The CCL2 genotypes CC and CG were also associated with less severe muscle injuries than the CCL2 GG genotype (p = 0.026) (Figure 1B, Table 3). Finally, the COL5A1 TC genotype showed a tendency towards an association with more severe muscle injuries (p = 0.08)) (Figure 1C, Table 3).

The 10 ligament injuries associated with the ELN AA genotype were more severe than those associated with the ELN AG or GG genotypes (p = 0.009) (Figure 2A, Table 3). SNPs in ELN also showed evidence of a statistically significant association with recovery time. Injuries associated with the ELN AG genotype required a shorter mean recovery time (24.7days) than those associated with the ELN GG (37.5 days) or AA (83.2 days) genotypes (p = 0.043) (Figure 2B, Table 3).

There was no evidence of a statistically significant association between the SNPs considered in the present study and tendon injuries (see Additional file 1).