Systematic Review
The Demographic Characteristics of High-Level and Recreational Athletes Undergoing Hip Arthroscopy for Femoroacetabular Impingement: A Sports-Specific Analysis Danyal H. Nawabi, M.D., F.R.S.C.(Orth), Asheesh Bedi, M.D., Lisa M. Tibor, M.D., Erin Magennis, J.D., and Bryan T. Kelly, M.D.
Purpose: The purpose of this study was to determine differences in age, gender, and the need for bilateral surgery between high-level athletes grouped by sports with similar mechanical demands on the hip and recreational athletes undergoing hip arthroscopy for femoroacetabular impingement (FAI). Methods: By use of a hip-preservation center registry, a retrospective review of patients undergoing hip arthroscopy for FAI between March 2010 and April 2012 was performed. Athletes were categorized as high level (high school, collegiate, or professional) or recreational. We performed a subgroup analysis for high-level athletes, looking at differences among contact, cutting, impingement, overhead/ asymmetric, endurance, and flexibility sports. Results: The study included 288 high-level athletes and 334 recreational athletes. Being a high-level athlete was associated with a younger age (mean age, 20.2 years v 33.0 years; odds ratio, 0.69; P < .001) and male gender (61.5% v 53.6%; odds ratio, 1.75; P ¼ .03). The percentage of high-level athletes undergoing bilateral surgery was higher than that of recreational athletes (28.4% v 15.9%); however, this association was found to be confounded by age on multivariate analysis. The most common sports for high-level athletes were soccer, hockey, and football. Athletes participating in cutting sports were significantly younger than athletes participating flexibility, contact, or impingement sports. Conclusions: When compared with recreational athletes undergoing arthroscopic treatment for FAI, high-level athletes are more likely to be younger, to be male, and to undergo bilateral surgery. When high-level athletes are grouped by the mechanical demands placed on the hip by their sport, athletes participating in cutting sports are more likely to be younger than those in the other groups. Level of Evidence: Level IV, case series.
F
emoroacetabular impingement (FAI) is increasingly being recognized and treated in athletes. There are many reports of the outcomes of arthroscopic treatment of FAI, in the form of acetabular rim trimming, selective labral debridement and/or labral refixation, and femoral osteochondroplasty, in sports-specific groups of highlevel athletes.1-5 The rates of return to play in highlevel athletes in the recent literature are typically From the Hospital for Special Surgery (D.H.N.), New York; Center for Hip Preservation, Hospital for Special Surgery (E.M., B.T.K.), New York, New York; Department of Orthopaedic Surgery, University of Michigan (A.B.), Ann Arbor, Michigan; and Kaiser Permanente (L.M.T.), South San Francisco, California, U.S.A. The authors report the following potential conflict of interest or source of funding: A.B. receives support from Smith & Nephew as an education consultant. B.T.K. receives support from Smith & Nephew, Pivot, A3 Surgical. Received May 6, 2013; accepted December 16, 2013. Address correspondence to Lisa M. Tibor, M.D., Kaiser Permanente, 1200 El Camino Real, South San Francisco, CA 94080, U.S.A. E-mail: lisa.tibor@ gmail.com Ó 2014 by the Arthroscopy Association of North America 0749-8063/13294/$36.00 http://dx.doi.org/10.1016/j.arthro.2013.12.010
398
high, ranging from 83% to 93% during the first year after surgery.1,3 Although outcome studies on athletes have included recreational-level athletes, there is only 1 study, comprising 80 patients undergoing hip arthroscopy for FAI, that has directly compared return-to-play rates between professional and recreational athletes.6 Furthermore, a demographic comparison between these 2 groups has not been previously performed. Compared with the general population, symptomatic FAI and labral tears are more common in certain types of athletes.7-11 There is also evidence that hip pathoanatomy may be influenced by sports participation during adolescence. Two studies have independently observed a 90% prevalence of decreased head-neck offset in male football and basketball players.9,10 Although decreased head-neck offset is often bilateral, both hips do not always become symptomatic.12-15 This may be related to activity, for example, if the patient participates in a sport that has asymmetric demands on the lower extremities, such as golf or baseball.16 Although there are known gender differences in anatomy around the hip,17 the demographic characteristics in high-level athletes based
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 30, No 3 (March), 2014: pp 398-405
DEMOGRAPHIC DATA OF HIP ARTHROSCOPY PATIENTS
on mechanical demands placed on the hip by certain sports have not been studied. The purpose of this study was to determine differences in age, gender, and the need for bilateral surgery between high-level athletes grouped by sports with similar mechanical demands on the hip and recreational athletes undergoing hip arthroscopy for FAI. We hypothesized that there are statistically significant differences in age, gender, and the need for bilateral surgery for high-level athletes when compared with recreational athletes and that these same demographic characteristics also differ in high-level athletes based on the different mechanical demands that the sports place on the hip joint.
Methods Data Source: Hip Outcomes Registry Since March 1, 2010, the senior author (B.T.K.) has participated in a hip outcomes registry. Demographic and clinical data are entered into the registry by a research assistant at the time of the office visit. After giving informed consent, all patients who are candidates for open or arthroscopic hip-preservation procedures are prospectively enrolled in the registry. The registry and projects using registry data have approval from the institutional review board. Data Collection The hip outcomes registry database was queried by a research assistant responsible for the registry for all patients undergoing at least 1 hip arthroscopy for clinically and radiographically (based on plain radiographs, computed tomography, and magnetic resonance imaging) proven FAI performed by the senior author between March 1, 2010, and April 1, 2012. Patients were excluded if they were aged younger than 15 years or older than 45 years, if they did not participate in any sport or active pursuit before or after surgery, or if there was no available information about sports participation either in the registry or after review of the electronic medical record. The following demographic data were collected: age at surgery, gender, bilateral or unilateral surgery, side of surgery for patients undergoing unilateral surgery, primary sport, and level of participation/competition. After data extraction, the patient information was reviewed
399
for consistency and completeness. A review of the electronic medical record was performed to obtain missing or conflicting information from the registry (e.g., to clarify the primary sport and level of participation). In those patients with a preoperative computed tomography scan, the maximum alpha angle as measured by a fellowshiptrained musculoskeletal radiologist was recorded for comparison of proximal femoral morphology between groups. The radiologist measured the alpha angle from a series of 2-dimensional radial reconstructions across the femoral neck axis; the alpha angle was defined as the angle subtended between the midline of the femoral neck and a line connecting the center of the femoral head to the point along the head-neck junction that first deviates from the sphericity of the femoral head. Classification Criteria Patients were considered high-level athletes if they were participating in a single primary sport at the high school, collegiate, national, international, or professional level at the time of presentation. All other patients were considered recreational athletes. High-level athletes were further grouped into 6 distinct categories based on the mechanical load placed on the hip by the sporting activity (Table 1). These categories were created by consensus agreement between the 2 senior authors (A.B. and B.T.K.), who are experienced in assessing and treating athletes with FAI. Cutting sports were characterized as running sports that require repeated changes of direction and cutting, leading to high rotational loads across the hip joint. Impingement sports were characterized as sports requiring high flexion, adduction, and internal rotation motions that re-create the classic “anterior impingement” position. Flexibility sports were characterized as sports that require supraphysiological ranges of motion such that impingement may occur in the presence of relatively minor impingement morphology or even “normal” bony anatomy. Asymmetric/overhead sports were characterized as sports that have clear side-to-side differences in mechanical/ functional hip motion requirements based on the unique asymmetry of the primary sporting motion. Contact sports were characterized as sports that had an increased likelihood for impingement-induced instability or traumatic lateral impact mechanisms. Endurance sports were characterized as sports that require extended periods of
Table 1. Sports Categories for High-Level Athletes Cutting Soccer Basketball Lacrosse Field hockey
Flexibility Contact Impingement Asymmetric/Overhead Endurance Dance Football Ice hockey Baseball Track, cross-country, other running Gymnastics Rugby Crew/rowing Softball Cycling Yoga Wrestling Baseball Catcher Tennis Swimming (not breaststroke) Cheer Water polo Golf Figure skating Equestrian polo Volleyball Martial arts Breaststroke swimmer Field events (javelin, discus, shot put, hammer throw)
400
D. H. NAWABI ET AL.
repetitive hip motion and hip flexor firing across the anterior aspect of the hip joint. Arthroscopic Interventions The senior author (B.T.K.) performed all arthroscopic procedures. All patients underwent hip arthroscopy in the supine position under regional anesthesia with fluoroscopic guidance and traction. The central compartment was addressed first, followed by the peripheral compartment, upon release of the traction. The detailed surgical technique has been previously described.3 The interventions performed in the central and peripheral compartments were based on treatment algorithms formulated during preoperative assessment and intraoperative diagnostic findings at arthroscopy. The interventions included, but were not limited to, acetabular rim trimming, labral debridement, labral refixation, partial psoas release, ligamentum teres debridement, femoral osteochondroplasty, and combined rim trimming and femoral osteochondroplasty. Statistical Analysis Data from Excel spreadsheets (Microsoft, Redmond, WA) were imported into SAS statistical software (version 9.3; SAS Institute, Cary, NC) for statistical analysis. Frequency measures were computed for all data; number and percent were calculated for each categorical variable, and mean and standard deviation were calculated for each continuous variable. We used c2 tests of each categorical predictor variable or Student t tests of each continuous predictor variable to determine significance before logistic regression modeling. The Wilcoxon rank sum was used for the none parametrically distributed variables. The post hoc Bonferroni technique was used for all multiple comparisons. For predictor variables showing a significant association with the outcome variable, logistic regression was conducted by use of both univariable and multivariable models. Univariable logistic regression was performed for each predictor variable on the outcome variable “high-level athlete,” with no adjustments. A full multivariable model was constructed using all predictor variables, followed by stepwise deletion of the variable with the lowest P value to obtain an optimal model. Odds ratios (ORs) and 95% confidence intervals (CIs) were produced in all analyses. An a level of .05 was considered significant for all tests.
Results A total of 724 patients aged between 15 and 45 years were identified. Of these patients, 102 were excluded because there was no information available about sports participation or they did not participate in sports at any level. This left 622 patients, 288 (46.3%) of whom met our criteria for high-level athletes and 334 (53.7%) who were classified as recreational athletes. There were 356
Fig 1. Age distribution for high-level and recreational athletes.
male patients (57.2%) and 266 female patients (42.8%). There were 135 patients (21.7%) who underwent bilateral surgeries, 284 right-sided surgeries (45.7%), and 203 left-sided surgeries (32.7%). The mean age at surgery was 27.1 years (range, 15.0 to 44.7 years) (Fig 1). There was a significant difference in the gender distribution between the high-level and recreational athletes, with 61.5% male patients among the high-level athletes, as compared with 53.6% male patients among the recreational athletes (P ¼ .05) (Table 2). This difference remained significant in a multivariable model controlling for age at surgery (OR, 1.81; 95% CI, 1.08 to 3.03) (Table 3). The high-level athletes were significantly younger than the recreational athletes (mean age at surgery, 20.2 years v 33.0 years; P < .001) (Table 2). Age at surgery between the 2 groups remained significantly different in the multivariable model including gender (OR, 0.69; 95% CI, 0.65 to 0.73; P < .0001) (Table 3). The maximum alpha angle on preoperative computed tomography scan was available for 360 hips and was reported by type of sporting activity (Table 4). The Table 2. Frequencies by Athlete Type High-Level Athletes
Recreational Athletes
Male Female Injury side* [n (%)]
177 (61.5) 111 (38.5)
179 (53.6) 155 (46.4)
Right Left Injury type [n (%)]
116 (40.3) 90 (31.3)
168 (50.3) 113 (33.8)
Unilateral Bilateral Age at surgery (yr)
206 (71.5) 82 (28.4) 20.2 4.0
281 (84.1) 53 (15.9) 33.0 7.2
Gender [n (%)]
*For patients with unilateral surgery.
Statistical Data 3.91 (c2 test) (P ¼ .05) 0.59 (c2 test) (P ¼ .44) 14.46 (c2 test) (P < .001)
28.09 (t test) (P < .001)
401
DEMOGRAPHIC DATA OF HIP ARTHROSCOPY PATIENTS Table 3. Analysis of Age, Gender, and Bilateral Surgery for High-Level Athletes (Logistic Regression) Multivariable: Fully
Univariate* Gender (male) Age at surgery Bilateral surgery
OR 1.38 0.69 2.11
95% CI 1.00-1.90 0.65-0.73 1.43-3.12
P Value .05
The Demographic Characteristics of High-Level and Recreational Athletes Undergoing Hip Arthroscopy for Femoroacetabular Impingement: A Sports-Specific Analysis Danyal H. Nawabi, M.D., F.R.S.C.(Orth), Asheesh Bedi, M.D., Lisa M. Tibor, M.D., Erin Magennis, J.D., and Bryan T. Kelly, M.D.
Purpose: The purpose of this study was to determine differences in age, gender, and the need for bilateral surgery between high-level athletes grouped by sports with similar mechanical demands on the hip and recreational athletes undergoing hip arthroscopy for femoroacetabular impingement (FAI). Methods: By use of a hip-preservation center registry, a retrospective review of patients undergoing hip arthroscopy for FAI between March 2010 and April 2012 was performed. Athletes were categorized as high level (high school, collegiate, or professional) or recreational. We performed a subgroup analysis for high-level athletes, looking at differences among contact, cutting, impingement, overhead/ asymmetric, endurance, and flexibility sports. Results: The study included 288 high-level athletes and 334 recreational athletes. Being a high-level athlete was associated with a younger age (mean age, 20.2 years v 33.0 years; odds ratio, 0.69; P < .001) and male gender (61.5% v 53.6%; odds ratio, 1.75; P ¼ .03). The percentage of high-level athletes undergoing bilateral surgery was higher than that of recreational athletes (28.4% v 15.9%); however, this association was found to be confounded by age on multivariate analysis. The most common sports for high-level athletes were soccer, hockey, and football. Athletes participating in cutting sports were significantly younger than athletes participating flexibility, contact, or impingement sports. Conclusions: When compared with recreational athletes undergoing arthroscopic treatment for FAI, high-level athletes are more likely to be younger, to be male, and to undergo bilateral surgery. When high-level athletes are grouped by the mechanical demands placed on the hip by their sport, athletes participating in cutting sports are more likely to be younger than those in the other groups. Level of Evidence: Level IV, case series.
F
emoroacetabular impingement (FAI) is increasingly being recognized and treated in athletes. There are many reports of the outcomes of arthroscopic treatment of FAI, in the form of acetabular rim trimming, selective labral debridement and/or labral refixation, and femoral osteochondroplasty, in sports-specific groups of highlevel athletes.1-5 The rates of return to play in highlevel athletes in the recent literature are typically From the Hospital for Special Surgery (D.H.N.), New York; Center for Hip Preservation, Hospital for Special Surgery (E.M., B.T.K.), New York, New York; Department of Orthopaedic Surgery, University of Michigan (A.B.), Ann Arbor, Michigan; and Kaiser Permanente (L.M.T.), South San Francisco, California, U.S.A. The authors report the following potential conflict of interest or source of funding: A.B. receives support from Smith & Nephew as an education consultant. B.T.K. receives support from Smith & Nephew, Pivot, A3 Surgical. Received May 6, 2013; accepted December 16, 2013. Address correspondence to Lisa M. Tibor, M.D., Kaiser Permanente, 1200 El Camino Real, South San Francisco, CA 94080, U.S.A. E-mail: lisa.tibor@ gmail.com Ó 2014 by the Arthroscopy Association of North America 0749-8063/13294/$36.00 http://dx.doi.org/10.1016/j.arthro.2013.12.010
398
high, ranging from 83% to 93% during the first year after surgery.1,3 Although outcome studies on athletes have included recreational-level athletes, there is only 1 study, comprising 80 patients undergoing hip arthroscopy for FAI, that has directly compared return-to-play rates between professional and recreational athletes.6 Furthermore, a demographic comparison between these 2 groups has not been previously performed. Compared with the general population, symptomatic FAI and labral tears are more common in certain types of athletes.7-11 There is also evidence that hip pathoanatomy may be influenced by sports participation during adolescence. Two studies have independently observed a 90% prevalence of decreased head-neck offset in male football and basketball players.9,10 Although decreased head-neck offset is often bilateral, both hips do not always become symptomatic.12-15 This may be related to activity, for example, if the patient participates in a sport that has asymmetric demands on the lower extremities, such as golf or baseball.16 Although there are known gender differences in anatomy around the hip,17 the demographic characteristics in high-level athletes based
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 30, No 3 (March), 2014: pp 398-405
DEMOGRAPHIC DATA OF HIP ARTHROSCOPY PATIENTS
on mechanical demands placed on the hip by certain sports have not been studied. The purpose of this study was to determine differences in age, gender, and the need for bilateral surgery between high-level athletes grouped by sports with similar mechanical demands on the hip and recreational athletes undergoing hip arthroscopy for FAI. We hypothesized that there are statistically significant differences in age, gender, and the need for bilateral surgery for high-level athletes when compared with recreational athletes and that these same demographic characteristics also differ in high-level athletes based on the different mechanical demands that the sports place on the hip joint.
Methods Data Source: Hip Outcomes Registry Since March 1, 2010, the senior author (B.T.K.) has participated in a hip outcomes registry. Demographic and clinical data are entered into the registry by a research assistant at the time of the office visit. After giving informed consent, all patients who are candidates for open or arthroscopic hip-preservation procedures are prospectively enrolled in the registry. The registry and projects using registry data have approval from the institutional review board. Data Collection The hip outcomes registry database was queried by a research assistant responsible for the registry for all patients undergoing at least 1 hip arthroscopy for clinically and radiographically (based on plain radiographs, computed tomography, and magnetic resonance imaging) proven FAI performed by the senior author between March 1, 2010, and April 1, 2012. Patients were excluded if they were aged younger than 15 years or older than 45 years, if they did not participate in any sport or active pursuit before or after surgery, or if there was no available information about sports participation either in the registry or after review of the electronic medical record. The following demographic data were collected: age at surgery, gender, bilateral or unilateral surgery, side of surgery for patients undergoing unilateral surgery, primary sport, and level of participation/competition. After data extraction, the patient information was reviewed
399
for consistency and completeness. A review of the electronic medical record was performed to obtain missing or conflicting information from the registry (e.g., to clarify the primary sport and level of participation). In those patients with a preoperative computed tomography scan, the maximum alpha angle as measured by a fellowshiptrained musculoskeletal radiologist was recorded for comparison of proximal femoral morphology between groups. The radiologist measured the alpha angle from a series of 2-dimensional radial reconstructions across the femoral neck axis; the alpha angle was defined as the angle subtended between the midline of the femoral neck and a line connecting the center of the femoral head to the point along the head-neck junction that first deviates from the sphericity of the femoral head. Classification Criteria Patients were considered high-level athletes if they were participating in a single primary sport at the high school, collegiate, national, international, or professional level at the time of presentation. All other patients were considered recreational athletes. High-level athletes were further grouped into 6 distinct categories based on the mechanical load placed on the hip by the sporting activity (Table 1). These categories were created by consensus agreement between the 2 senior authors (A.B. and B.T.K.), who are experienced in assessing and treating athletes with FAI. Cutting sports were characterized as running sports that require repeated changes of direction and cutting, leading to high rotational loads across the hip joint. Impingement sports were characterized as sports requiring high flexion, adduction, and internal rotation motions that re-create the classic “anterior impingement” position. Flexibility sports were characterized as sports that require supraphysiological ranges of motion such that impingement may occur in the presence of relatively minor impingement morphology or even “normal” bony anatomy. Asymmetric/overhead sports were characterized as sports that have clear side-to-side differences in mechanical/ functional hip motion requirements based on the unique asymmetry of the primary sporting motion. Contact sports were characterized as sports that had an increased likelihood for impingement-induced instability or traumatic lateral impact mechanisms. Endurance sports were characterized as sports that require extended periods of
Table 1. Sports Categories for High-Level Athletes Cutting Soccer Basketball Lacrosse Field hockey
Flexibility Contact Impingement Asymmetric/Overhead Endurance Dance Football Ice hockey Baseball Track, cross-country, other running Gymnastics Rugby Crew/rowing Softball Cycling Yoga Wrestling Baseball Catcher Tennis Swimming (not breaststroke) Cheer Water polo Golf Figure skating Equestrian polo Volleyball Martial arts Breaststroke swimmer Field events (javelin, discus, shot put, hammer throw)
400
D. H. NAWABI ET AL.
repetitive hip motion and hip flexor firing across the anterior aspect of the hip joint. Arthroscopic Interventions The senior author (B.T.K.) performed all arthroscopic procedures. All patients underwent hip arthroscopy in the supine position under regional anesthesia with fluoroscopic guidance and traction. The central compartment was addressed first, followed by the peripheral compartment, upon release of the traction. The detailed surgical technique has been previously described.3 The interventions performed in the central and peripheral compartments were based on treatment algorithms formulated during preoperative assessment and intraoperative diagnostic findings at arthroscopy. The interventions included, but were not limited to, acetabular rim trimming, labral debridement, labral refixation, partial psoas release, ligamentum teres debridement, femoral osteochondroplasty, and combined rim trimming and femoral osteochondroplasty. Statistical Analysis Data from Excel spreadsheets (Microsoft, Redmond, WA) were imported into SAS statistical software (version 9.3; SAS Institute, Cary, NC) for statistical analysis. Frequency measures were computed for all data; number and percent were calculated for each categorical variable, and mean and standard deviation were calculated for each continuous variable. We used c2 tests of each categorical predictor variable or Student t tests of each continuous predictor variable to determine significance before logistic regression modeling. The Wilcoxon rank sum was used for the none parametrically distributed variables. The post hoc Bonferroni technique was used for all multiple comparisons. For predictor variables showing a significant association with the outcome variable, logistic regression was conducted by use of both univariable and multivariable models. Univariable logistic regression was performed for each predictor variable on the outcome variable “high-level athlete,” with no adjustments. A full multivariable model was constructed using all predictor variables, followed by stepwise deletion of the variable with the lowest P value to obtain an optimal model. Odds ratios (ORs) and 95% confidence intervals (CIs) were produced in all analyses. An a level of .05 was considered significant for all tests.
Results A total of 724 patients aged between 15 and 45 years were identified. Of these patients, 102 were excluded because there was no information available about sports participation or they did not participate in sports at any level. This left 622 patients, 288 (46.3%) of whom met our criteria for high-level athletes and 334 (53.7%) who were classified as recreational athletes. There were 356
Fig 1. Age distribution for high-level and recreational athletes.
male patients (57.2%) and 266 female patients (42.8%). There were 135 patients (21.7%) who underwent bilateral surgeries, 284 right-sided surgeries (45.7%), and 203 left-sided surgeries (32.7%). The mean age at surgery was 27.1 years (range, 15.0 to 44.7 years) (Fig 1). There was a significant difference in the gender distribution between the high-level and recreational athletes, with 61.5% male patients among the high-level athletes, as compared with 53.6% male patients among the recreational athletes (P ¼ .05) (Table 2). This difference remained significant in a multivariable model controlling for age at surgery (OR, 1.81; 95% CI, 1.08 to 3.03) (Table 3). The high-level athletes were significantly younger than the recreational athletes (mean age at surgery, 20.2 years v 33.0 years; P < .001) (Table 2). Age at surgery between the 2 groups remained significantly different in the multivariable model including gender (OR, 0.69; 95% CI, 0.65 to 0.73; P < .0001) (Table 3). The maximum alpha angle on preoperative computed tomography scan was available for 360 hips and was reported by type of sporting activity (Table 4). The Table 2. Frequencies by Athlete Type High-Level Athletes
Recreational Athletes
Male Female Injury side* [n (%)]
177 (61.5) 111 (38.5)
179 (53.6) 155 (46.4)
Right Left Injury type [n (%)]
116 (40.3) 90 (31.3)
168 (50.3) 113 (33.8)
Unilateral Bilateral Age at surgery (yr)
206 (71.5) 82 (28.4) 20.2 4.0
281 (84.1) 53 (15.9) 33.0 7.2
Gender [n (%)]
*For patients with unilateral surgery.
Statistical Data 3.91 (c2 test) (P ¼ .05) 0.59 (c2 test) (P ¼ .44) 14.46 (c2 test) (P < .001)
28.09 (t test) (P < .001)
401
DEMOGRAPHIC DATA OF HIP ARTHROSCOPY PATIENTS Table 3. Analysis of Age, Gender, and Bilateral Surgery for High-Level Athletes (Logistic Regression) Multivariable: Fully
Univariate* Gender (male) Age at surgery Bilateral surgery
OR 1.38 0.69 2.11
95% CI 1.00-1.90 0.65-0.73 1.43-3.12
P Value .05
