Professional golfers’ hips: prevalence and predictors of hip pain with clinical and MR examinations
The title of this study does not refer to femoroacetabular impingement (FAI), despite this pathology appearing to be the underlying concern, particularly as the authors focus on measuring alpha angles and assessing for pincer morphology. The title should be more reflective of the results.
The main aims are clearly set out:
- To determine the prevalence of pain as opposed to diagnosis of FAI or associated hip pathology, simply what is the prevalence of pain in golfers.
- Whether there was a difference between the lead leg and trail leg with regards the hips. It would be assumed that the lead leg would be more affected throughout a professional golfers career, particularly reduction in hip internal rotation, however this study does not suggest that.
- To establish demographic, clinical and morphological characteristics predictive of hip pain.
The authors should be commended for their initial method of data collection in the field as opposed to asking participants to attend a particular study site. While there are reduced subject numbers for the subsequent clinical examinations and MR examinations this is clearly stated for the reader.
Although 70% of the total field completed the questionnaire the authors do not provide enough information to determine whether there may have been bias in the recruitment of these cases. For example (as we discussed), if participants were told that if they completed this questionnaire they could then access a scan of their hip there may have been a higher likelihood that those who completed the questionnaire did have hip pain. Uncertainty arises therefore whether the 30% who did not complete the questionnaire were completely free of hip pain. The 70% therefore may not have been a random sample. The authors also do not provide demographic details of ethnicity.
QUESTIONNAIRE: The primary question used to determine hip pain prevalence (“In the past month have you had any pain in the hip or groin lasting 1 day or longer”) is not specific enough and perhaps too subjective. Additionally, the inclusion of groin pain also in their question which can potentially result from adductor problems, pubic symphysis problem, inguinal-related problems etc. is not in keeping with the main focus of the paper and may capture cases of non-hip-related pathology. Further this specific question is not in keeping with FAI which is more of a chronic pain. The specifics of the question (lasting 1 day or longer) may have potentially captured an acute, traumatic, isolated injury in this study cohort, however this is not clearly evaluated. The outcomes measure used (iHOT 12) is appropriate for the study population.
CLINICAL EXAMINATION: The testing they used to assess for hip pain (FADIR and FABER examinations, in addition to measures of internal rotation and flexion) would suggest the authors were looking for FAI pathology specifically, potentially ruling out other sources of hip pain. Based on the references supplied by the authors for the technique of ROM measurement (15) it appears they have taken appropriate steps to accurately measure these using multiple examiners to facilitate movement, measurement and recording and is therefore robust.
MR EXAMINATION: This seems appropriate overall. The authors reference standard cut-offs for the alpha angle, 55 degrees. The presence of pincer abnormality was incompletely assessed in this study; using only a measure of acetabular depth was insufficient and the most likely the main reason no pincer abnormalities were observed. Within the literature there is an increased focus on femoral neck antetorsion which considers how anteverted the femoral neck is with respect to the axis of the knee (normal approx. 15 degrees). As this angle approaches 0 degrees (or progressing beyond to negative degrees ‘retroversion’) implies a more neutral alignment of the femoral neck to the pelvic plane more likely causing impingement of the hip. The authors considered anything <0 as retrotorsion. As the focus appears to be on bony morphology, MRI is perhaps not the most appropriate means which to assess this. The additional use of x-rays would have been more appropriate for assessing true bony morphology, although ethical approval would have to be granted to facilitate this which may have been an influencing factor in the authors decision not to use x-ray imaging.
The authors report a “significantly lower” score in the iHOT as their main finding for the lead hip as compared to the trail hip. However this difference is just 1 point between the two hips which is not clinically relevant but the authors do not provide further statistics (such as measures of effect size) to explain this, which is misleading. We acknowledge that the authors do refer to the potential lack of clinical meaningful of this difference later in the discussion (an MCID for iHOT 12 is 6.1 points).
The report of mean 101o flexion for both hips seems very reduced and question arises again whether only those with symptomatic hip pain were included for this analysis considering less than half (47%) of competitors were only included.
For the MR examination results, the authors report the greatest difference between lead and trail hip alpha angles occurs between the 12 and 3 o’clock positions however the results reported in Table 2 would contradict this, whereby statistical significance was only observed between the 2 and 3 o’clock positions. It is unclear therefore why the authors incorporated the 12 and 1 o’clock position when these was not statistically significant. The authors further utilised the results of these non-significant 12 and 1 o’clock positions in their regression model to form the basis of another of their main findings – that the mean alpha angle between 12 and 3 o’clock positions predict quality of life.
The authors also report that “No player had pincer morphology” however the presence of true pincer morphology (which is typically assessed using x-ray imaging by measuring lateral or anterior centre-edge angle, crossover sign, etc.) was not incorporated in the assessment of these players. MRI is a less appropriate assessment for pincer morphology.
Expert Review Group Summary:
- There may have been some selection bias in the recruitment of cases for each line of analysis.
- The number of players who had a subsequent MR examination was relatively low (35%) however the authors do not provide any information on how many of these had hip pain to begin with or what proportion of players with hip pain were included/not included. There is no indication as to the difference in demographics, clinical examination findings or prevalence of pain between those who were scanned and those who were not.
- The assessment of pain using a single question was not specific enough to address or differentiate between hip pain and groin pain. The conclusion therefore by the authors that “Hip pain affects 19% of professional golfers..” may not be specific and representative.
- The authors state in the conclusion that “the lead hip was more frequently affected that the trail”, however this is not true as no statistical significance was found (p=0.378).
- For alpha angles there was small variation between trail and leading hip and the authors have included measurement at 12 and 1 o’clock positions despite no significant difference for these measurements between lead and trail hip.
- Variation in the iHOT was very small (1 point) which could not represent any clinical significance, irrespective of what their alpha angle or age was.
The conclusions made by the authors in this paper are not supported by their results.
01 October 2020
Expert Review Group:
- Mr Patrick Carton, Consultant Orthopaedic Surgeon, The Hip and Groin Clinic
- Mr David Filan, Senior Research Assistant, UPMC Whitfield
- Dr Karen Mullins, Research Assistant, UPMC Whitfield
- Mr Derek O’Neill, Senior Physiotherapist, UPMC Whitfield
- Mr Dualtach Mac Colgain, Sports Medicine Doctor, UPMC Whitfield
- Dr Michael Hanlon, Lecturer, Waterford Institute of Technology
- Mr Shane Lawlor, Sports Chiropractor