Hip adduction and abduction strength profiles in elite and sub-elite female soccer players according to players level and leg limb-dominance | BMC Sports Science, Medicine and Rehabilitation

Study Design

A descriptive-comparative study was conducted to examine the isometric hip strength profile (i.e., adduction and abduction) and adduction:abduction strength ratio. This comparison was made across players performance level (i.e., elite vs. sub-elite soccer players) and lower-limb dominance (i.e., dominant vs. non-dominant leg).

Participants

A total of 84 female soccer players from a professional soccer club in Spain were potentially eligible from 2019 until 2022. Finally, 82 players participated part in this study. A total of two players (n = 2) were unable to complete the test due to experiencing pain during the administration of the tests. All players were recruited from the same soccer club. Starting from the first year of evaluation (i.e., 2019), only new players who had not been previously tested were included in the study. Participants characteristics measured are presented in Table 1. Before testing procedures, all players provided a written informed consent. Players who had sustained lower limb injuries lasting more than 4 weeks in the last 3 months were excluded. Additionally, players with a history of athletic pubalgia within the last year were specifically excluded from the study. The classification of the group’s performance was based on the participant classification framework developed by McKay et al. [19]. That is, elite players, who compete at the international level (1st division and international championships), were distinguished from sub-elite players (i.e., highly trained/National level).

Procedures

Players underwent testing between July and August across three consecutive pre-season training periods (2019/20, 2020/21, 2021/22). All assessments of hip strength were conducted by the same physiotherapist to minimize potential sources of error. A portable hand-held dynamometer (HHD) (MicroFET 2, Hoggan Scientific, LLC, Salt Lake City, UT) was utilized for the hip strength assessments, with calibration performed before testing. This calibration process included zeroing the device. Furthermore, in each season, the HHD underwent additional calibration by setting it up with a known load to ensure consistent and accurate measurements over time. Maximal voluntary isometric hip adduction and hip abduction force, in both dominant and non-dominant legs were tested. All assessments were done on a massage table. The order in which the tests were conducted was varied systematically among participants. By doing so, we aimed to eliminate or reduce potential biases due to the order of tests and any transference effects that might occur if one test influenced the performance in subsequent tests. Specifically, each player completed a sequence of tests in the order of A, B, B, A, where ‘A’ represents abductor test and ‘B’ denotes adductor test. Subsequently, the average of the two ‘A’ conditions was calculated, and the same process was applied for the ‘B’ conditions. Two sub-maximal familiarization trials were performed to ensure the players were performing the correct action of pushing into the belt and the HHD. Verbal encouragement was provided during the test execution with a standard instruction of “push, push, push”. Prior to each testing period, players performed a standardized warm-up, which consisted of 5 min of stationary bike after 10 repetitions of concentric and eccentric abductor and adductor movements. After a short break of 5 min, players were tested.

Isometric hip adduction and hip abduction were measured in the supine position as introduced by Thorborg et al [9]. The participants were placed in the supine position and were told to stabilize themselves by holding onto the side of the table with their hands. For the adductor measurement, the examiner (E.J.) applied resistance in a fixed position, 2 cm proximal to the edge of the medial malleolus. The abductor measurement was performed using the HHD and a belt-fixation proximal to the edge of the lateral malleolus [20]. The participant being tested exerted a 5 s maximum isometric voluntary contraction against the HHD.

Two trials in the same leg were completed with a 30-s rest period between repetitions. The mean of peak force (measured in Newtons [N]) recorded for each limb across two attempts was used for data analysis. If variability between trials were more than 10%, a new trial was done. A reliability study performed in 10 female soccer players, selected randomly from our sample, showed that the intrarater reliability was found to be excellent according to the index correlation coefficient model 2.k (ICC_2.k) = 0.86 (0.76 to 0.95) and standard error of measurement (SEM) of 0.26 Nm/kg for hip adduction and excellent, ICC_2.k = 0.80 (0.56 to 0.91) and SEM = 0.41 Nm/kg, for hip abduction. In addition, a recent reliability study showed excellent reliability coefficients (ICC = 0.92 to 0.96) and nearly perfect validity scores (r = 0.996) in comparison to fixed-frame dynamometry system [21].

Statistical analysis

All results were expressed as a mean and standard deviation (± SD). Normality and homogeneity of variance assumptions were analyzed using Shapiro–Wilk test and Levene test, respectively. Relative reliability was examined using ICC_2.k, whereas absolute reliability was calculated using SEM [22]. There was a statistically significant association (p < 0.001) between age of participants and their relative hip isometric adduction and abduction strength test scores (with correlation coefficients ranging from 0.30 to 0.45, all of which were statistically significant). As a result, the age factor was included as a covariate. Consequently, to examine the effect of players performance level (i.e., elite vs. sub-elite) and dominance (i.e., dominant vs. non-dominant leg) an analysis of covariance variance (ANCOVA, 2 × 2) was employed. For adduction:abduction ratio, an ANOVA (2 × 2) was performed as the Pearson correlation coefficient between adduction:abduction ratio and age were not statistically significant. Post hoc tests, utilizing the Bonferroni correction, were conducted to address multiple comparison. All post hoc analysis was presented using mean differences (MD) and 95% of confident intervals (CI_95%). Effect size (ES) was calculated according to Cohen formulas [23]. and considered trivial (< 0.20), small (0.20 – 0.59), moderate (0.60 – 1.19), large (1.20 – 1.99), and very large (> 2.00) [24]. The statistically significant level was set at p < 0.05. All calculations were done using a statistical analysis tool (JASP v.0.17.1, the Netherlands).