Study Design
A cohort study was conducted on elite football players (total n = 111; males n = 63; females n = 48). All players were recruited from a single club: Newcastle Jets A-League FC Academy in Newcastle, Australia. As all participants were under 18 years of age, written informed consent was obtained from both players and their parents/guardians. Ethics approval (protocol number: H-2018-0118) was provided by The University of Newcastle Human Research Ethics Committee. Pre-season data collection occurred in October 2018, consisting of the HAGOS questionnaire, reporting of historical (within their lifetime) hip and groin pain and hip adductor and abductor muscle strength testing. Subsequent in-season monitoring of all episodes of hip and groin pain, irrespective of time-loss, was collected during the ten-month playing season from November 2018 to August 2019.
Participants
To define the population sample, the following participant characteristics were first collected: playing age group, gender, standing height (cm), body mass (kg) and dominant leg length from the lateral malleolus to the greater trochanter (cm). Participants were asked during injury history screening if they had ever experienced hip and groin pain in their lifetime. Hip and groin pain was defined as pain in the anterior hip, pubic region, inguinal regional and proximal adductor insertion (excluding lateral or posterior hip). In addition, participants were also asked to report their dominant limb: “which leg do you prefer to kick a ball with?”.
Pre-season HAGOS Questionnaire
All participants completed the HAGOS, containing 37 questions answered on a Likert scale, grouped into six subscales: Pain, Symptoms, Activities of Daily Living (ADL), Sport & Recreational Activities (Sport/Rec), Participation in Physical Activity (PA) and Quality of Life (QOL) [16]. Each subscale was scored independently and then normalised to a 100-point scale, as per Thorborg et al. [16], with lower scores indicating greater hip and/or groin problems or disabilities. If a participant failed to answer more than one question for PA subscale or more than two questions for all other subscales (Symptoms; Pain; ADL; Sport/Rec; QOL), then the individual subscale was excluded from analysis and treated as ‘missing data’ [16].
Pre-season Strength Testing
Pre-season strength testing took place during a three-week period between playing seasons after squads had been selected and prior to the commencement of training. Unilateral hip adductor and abductor strength was recorded using a handheld dynamometer (HHD) (Lafayette model 01165 manual muscle tester with 7 cm × 3.5 cm pad). All strength tests were undertaken by five physiotherapists with a combined experience of 60 years (range 4–23 years), of whom two were titled musculoskeletal physiotherapists and a third had post-graduate qualifications. The strength testing procedures have previously been described and has been shown to have a high inter-rater intra-class coefficient between 0.86–0.93 for adductor and 0.87–0.98 for abductor strength testing [15, 21]. Participants lay supine with the leg being tested in 0° hip and knee extension and resistance applied 5 cm proximal to the malleolus, while the leg not being tested was in knee flexion with the foot on the plinth (Fig. 1). Participants completed a standardised warm-up of two five-second isometric repetitions against the HHD with ten second rests between repetitions prior to both hip adductor and abductor tests. The tests consisted of 3 ‘break’ repetitions, each held for ≤ 3 s with a 30 s rest period, for both the hip adductors and abductors. A ‘break’ test, defined as an eccentric muscle contraction, was used to determine participant’s maximal strength as it has been shown to be more sensitive in identifying male football players with adductor-related groin pain compared to isometric testing [13]. Examiners instructed participants to “push as hard as you can” and repeated the encouragement “keep going”’ three times for each test. For each test participants were allowed to stabilise themselves by gripping the side of the plinth with their hands and pressing the non-testing leg against the plinth [15, 21].
To attain a reproducible result that represented an individual’s maximal effort, the highest value for the three testing repetitions was used for analysis, unless the highest value was > 10% of the player’s other two values. If this occurred, the highest value was considered an outlier and the second highest of the three values was taken [22]. If a testing repetition was painful, the test was stopped, and this value was excluded from analyses. Hip adductor and abductor strength was measured in Newtons (N), with muscle torque calculated by multiplying the highest accepted strength value by leg length in metres (measured from the lateral malleolus to the greater trochanter of the player’s dominant leg) and then divided by body mass in kilograms [13]. Leg length measurements were taken of the dominant leg only due to time constraints and the limited accuracy of tape measures in detecting subtle between-limb differences. Eccentric hip adductor-to-abductor strength ratio was calculated for each leg [15]. Between limb difference was calculated by subtracting the dominant side from the non-dominant side for force values (N) and torque (N/Kg).
In-Season Monitoring
Players were advised to report any pain or injuries to medical staff at both games and training sessions, irrespective of whether treatment was required. This resulted in both time-loss and non-time-loss hip and groin pain being recorded on a weekly basis [4]. If hip and groin pain was reported by a player, it was then assessed at the first training session of the weekly schedule by a registered physiotherapist with 20 years of musculoskeletal physiotherapy experience. Players with hip and groin pain were identified by the physiotherapist from electronic records kept by physiotherapists and trainers on game days, technical director reports of injuries from coach’s post game reports, players who were not training at this session, and players who presented to the physiotherapist with pain at the training session or the clinic. The physiotherapist assessed all of these identified players at the first weekly training session, with these data entered into an excel spreadsheet that became the final data record used for analysis. Assessment of the hip and groin region included but was not limited to functional movements (e.g. squatting, jumping, kicking football), passive and active joint range of motion, manual muscle strength testing and combined hip-joint movement tests (flexion adduction internal rotation and flexion adduction external rotation). The physiotherapist determined a clinical diagnosis based on their assessment.
Statistical Analysis
Hip and groin pain prevalence in youth elite players was calculated for both in-season hip and groin pain and historical hip and groin pain by dividing the number of players reporting pain by the total number of players included in the study. For analysis, players were categorised as either in-season hip/groin pain (at least one episode) or no in-season hip and groin pain. Descriptive statistics were calculated for each variable with male and female data stratified, as both strength [23] and the experience of pain [24] differs between sexes. For the analysis of strength variables both legs of an individual were averaged as there were no differences between painful and non-painful limbs in players with pain, or between left and right limbs of asymptomatic players, or their left and right as compared to the left and right of pain players matched by weight and height. Data normality was assessed through Shapiro–Wilk tests and visualisation of histograms. Differences between pain and asymptomatic players were calculated using independent t-tests for parametric data or Mann–Whitney-U tests for non-parametric data.
Logistic regression models determined factors associated with having in-season hip and groin pain. Prior to logistic regression, Pearson’s coefficients were used to examine potential correlations between independent variables, with all strength variables found to be correlated, and the majority of HAGOS subscales correlated. Due to the large number of possible candidate variables, univariate modelling was used to determine which variables were to be included in the multivariate model. Variables that underwent univariate modelling were leg length, body mass index (BMI), gender, playing age group at time of testing (ages 11, 12, 13 or 14 years) historical hip and groin pain, hip abductor muscle torque, hip adductor muscle torque, hip adductor/abductor strength ratio and HAGOS subscales Pain, Symptoms, ADL, Sport/Rec and QOL. Variables with p < 0.25 in the univariate models were then included in a multi-variate model analysed using the backwards (Wald) method [25]. A model combining both male and female sexes was used as male and female multi-variate models had similar results to the combined model. The same methodological process was also used for separate analyses that examined differences in hip muscle strength and HAGOS subscale scores for players with either historical hip and groin pain or no historical hip and groin pain. Where variables representing a construct (i.e. strength, HAGOS subscales) were correlated, selection of variables for the multivariate model was based on lowest p-value and/or highest Nagelkerke R2 from the univariate models. SPSS v 27.0 (IBM SPSS Statistics, IBM Corp., Armonk, NY) was used for analyses.