Sports Participation in Patients After Total Hip Arthroplasty vs Periacetabular Osteotomy: A Propensity Score-Matched Asian Cohort Study

Accepted Manuscript Sports Participation in Patients After Total Hip Arthroplasty Versus Periacetabular Osteotomy: A Propensity Score-Matched Asian Cohort Study Daisuke Hara, M.D., Ph.D., Satoshi Hamai, M.D., Ph.D., Keisuke Komiyama, M.D., Goro Motomura, M.D., Ph.D., Kyohei Shiomoto, M.D., Yasuharu Nakashima, M.D., Ph.D. PII:

S0883-5403(17)30758-1

DOI:

10.1016/j.arth.2017.08.035

Reference:

YARTH 56065

To appear in:

The Journal of Arthroplasty

Received Date: 10 August 2017 Revised Date:

21 August 2017

Accepted Date: 26 August 2017

Please cite this article as: Hara D, Hamai S, Komiyama K, Motomura G, Shiomoto K, Nakashima Y, Sports Participation in Patients After Total Hip Arthroplasty Versus Periacetabular Osteotomy: A Propensity Score-Matched Asian Cohort Study, The Journal of Arthroplasty (2017), doi: 10.1016/ j.arth.2017.08.035. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Original research article:

2

Sports Participation in Patients After Total Hip Arthroplasty Versus Periacetabular

3

Osteotomy: A Propensity Score-Matched Asian Cohort Study

4

RI PT

1

Daisuke Hara, M.D., Ph.D.1: [email protected]

6

Satoshi Hamai, M.D., Ph.D.1: [email protected]

7

Keisuke Komiyama, M.D.1: [email protected]

8

Goro Motomura, M.D., Ph.D.1: [email protected]

9

Kyohei Shiomoto, M.D.1: [email protected]

M AN U

10

SC

5

Yasuharu Nakashima, M.D., Ph.D.1: [email protected]

11 12

1

13

3-1-1 Maidashi, Higashi-ku, Fukuoka, 812-8582, Japan

TE D

14

Department of Orthopaedic Surgery, Graduate School of Medical Sciences, Kyushu University,

Corresponding author: Satoshi Hamai, M.D., Ph.D.

16

Department of Orthopaedic Surgery, Graduate School of Medical Sciences, Kyushu University

17

3-1-1 Maidashi, Higashi-ku, Fukuoka, 812-8582, Japan

18

TEL: +81-92-642-5488, FAX: +81-92-642-5507

19

E-mail: [email protected]

AC C

EP

15

1

ACCEPTED MANUSCRIPT 1

Sports Participation in Patients After Total Hip Arthroplasty Versus Periacetabular Osteotomy:

2

A Propensity Score-Matched Asian Cohort Study

3 Abstract

5

Background: No studies have compared sports participation between total hip arthroplasty (THA) and

6

periacetabular osteotomy (PAO) in matched Asian cohorts. We investigated sports participation and

7

activity levels in Asian THA cohort, and compared them between THA and PAO cohorts.

SC

RI PT

4

8

Methods: Multivariate analyses were applied to determine which factors were associated with

M AN U

9

postoperative sports participation and University of California-Los Angeles (UCLA) activity score in

11

(1) 524 THA patients and (2) 487 acetabular dysplasia patients (295 THA patients and 192 PAO

12

patients). Additionally, postoperative sports participation and UCLA score were compared between 62

13

THA and 62 PAO patients after adjusting for baseline characteristics with propensity score matching.

14

TE D

10

Results: Sports participation and UCLA score significantly increased after THA (P < .001 in both

16

analyses). Preoperative sports participation was the factor most associated with both postoperative

17

sports participation and UCLA score in both 524 THA patients, and 487 acetabular dysplasia patients (P

18

< .001 in all analyses). Multivariate analysis in 487 acetabular dysplasia patients demonstrated that

19

THA, compared with PAO, was negatively associated with postoperative sports participation (P < .001)

20

but not postoperative UCLA score (P = .22). THA patients showed significantly lower rate of

21

postoperative sports participation (32.3 and 51.6%, respectively, P = .046) but not postoperative UCLA

22

score (5.0 ± 1.6 and 5.2 ± 1.9, respectively, P = .47) compared with matched PAO patients.

AC C

EP

15

23 24

Conclusion: THA significantly increased both sports participation and activity levels. Both -1-

ACCEPTED MANUSCRIPT 25

multivariate and propensity score-matched analyses showed that postoperative activity levels were

26

comparable between THA and PAO cohorts.

27 Keywords: total hip arthroplasty, periacetabular osteotomy, acetabular dysplasia, sports, activity level

RI PT

28

AC C

EP

TE D

M AN U

SC

29

-2-

ACCEPTED MANUSCRIPT 30 31

Introduction Total hip arthroplasty (THA) provides pain relief, improves function, and promotes good longterm outcomes; thus, it is considered one of the most successful procedures in orthopedic surgery [1].

33

As expectations of participation in sports among patients undergoing THA have increased because of

34

the clinical success of THA, sports participation after THA have become the focus of attention by

35

orthopedic surgeons. Previous studies reported favorable outcomes or showed high activity levels after

36

THA[2-6]. However, most of these studies were from Western countries [2,4-6], and few English

37

articles that reported the participation in sports or activity levels of Asian patients who had undergone

38

THA exist. In Asian countries, the major cause of THA is secondary osteoarthritis (OA) due to

39

acetabular dysplasia, which is different from that in Western countries [7-9].

SC

M AN U

40

RI PT

32

Acetabular dysplasia is a common cause of secondary OA [10]. For hips with symptomatic acetabular dysplasia in young- to middle-aged patients, periacetabular osteotomies (PAO) have been

42

developed to delay or prevent the development of OA [11-14]. However, whether THA or PAO should

43

be indicated in middle-aged patients with symptomatic secondary OA due to acetabular dysplasia

44

remains controversial. Thus, in these patients, deciding whether undergoing PAO soon to prevent OA

45

progression or THA after OA progression is difficult. Although several studies have compared THA

46

and PAO with respect to pain relief, function, cost-effectiveness, or complications [15-19], only few

47

studies have included activity levels as an outcome [18,19]. In addition, these studies [18,19] did not

48

examine whether patients actually participated in sports preoperatively or postoperatively and were

49

limited to a relatively smaller number of subjects or unmatched demographic cohorts. To the best of

50

our knowledge, no reports have compared both participation in sports and activity levels between THA

51

and PAO in matched cohorts.

AC C

EP

TE D

41

-3-

ACCEPTED MANUSCRIPT The purpose of this study, therefore, was to investigate Asian patients’ participation in sports

53

and activity levels before and after undergoing THA and compare these characteristics in patients with

54

secondary OA due to acetabular dysplasia between THA and PAO using a propensity score-matching

55

method. Specifically, the following questions were addressed: (1) what is the percentage of Asian

56

patients participating in sports before and after THA, (2) what factors are associated with participation

57

in sports and activity levels after THA, (3) what factors are associated with participation in sports and

58

activity levels in patients with secondary OA due to acetabular dysplasia after THA or PAO, and (4)

59

how different are participation in sports and activity levels between patients after THA and PAO in the

60

propensity score-matched cohorts?

M AN U

61 62

Material and methods

63

THA cohort and sports

After obtaining approval from our local institutional review board, we retrospectively reviewed

TE D

64

SC

RI PT

52

data on 904 patients (1125 hips) who underwent primary cementless THA between January 1998 and

66

December 2011. Among these patients, 782 patients (983 hips) satisfied all the following inclusion

67

criteria: (1) alive at the time of survey, (2) last surgery performed more than a year previously, (3)

68

evaluation by a surgeon within the past one year, and (4) absence of any condition due to physical or

69

mental disorders unrelated to the hip joint that may lead to bed rest. We mailed a questionnaire to these

70

782 patients, of which 524 patients (626 hips, 67.0%) completed and returned the questionnaire with

71

written informed consent (Fig. 1). Data were handled in accordance with the Ethical Standards of the

72

Helsinki Declaration.

AC C

EP

65

73

The questionnaire asked about (1) whether patients participated in sports at least once a month

74

before and after THA; (2) the sport(s) they participated in; (3) all patient-derived scores, including the

75

Oxford Hip Score (OHS) [20,21]; (4) their physical activity levels, determined by the University of -4-

ACCEPTED MANUSCRIPT California-Los Angeles (UCLA) activity scale score [21,22], before and after THA; and (5) whether

77

they had undergone treatment for any other condition related to their hip disorder [23]. If patients

78

selected “yes” for participation in sports, multiple answers based on 13 items were allowed (12 specific

79

sports and “others”) (Fig. 2). The name of the sport had to be specified if the patient selected “others.”

80

The OHS and UCLA score are validated, reliable, and self-reported metric assessments for patients

RI PT

76

with hip OA [20-22]. The OHS assesses the pain and function of the hip during daily activities, while

82

the UCLA score measures physical activity levels. Patients treated with bracing or surgery for infant or

83

childhood hip dysplasia were included in the group with a history of treatment for developmental hip

84

dislocation. Only one hip was randomly selected from each of the 102 patients with bilateral THA to

85

fulfill the statistical assumption of independent observation. Consequently, 524 THA patients were

86

reviewed (Fig. 1). All demographic factors, including age at the time of surgery, sex, body mass index

87

(BMI), diagnosis, follow-up duration, and history of hip osteotomy prior to THA, such as proximal

88

femoral osteotomy or PAO, were obtained from the patients’ medical records.

89 PAO cohort as control data

EP

90

TE D

M AN U

SC

81

For the control group, we retrospectively reviewed data on 342 patients (372 hips) who underwent

92

PAO for symptomatic acetabular dysplasia between January 1998 and December 2011. Among them,

93

269 patients (295 hips) satisfied all of the following inclusion criteria: (1) surgery performed more than

94

a year previously, (2) evaluation by a surgeon within the past one year, (3) absence of any condition

95

due to physical or mental disorders unrelated to the hip joint that may lead to bed rest, and (4) no THA

96

for the contralateral hip. We mailed the same questionnaire sent to the THA patients to the 269 PAO

97

patients, of which 203 patients (232 hips, 75.5%) completed and returned the questionnaire with

98

written informed consent [23]. Twelve hips (11 patients) were excluded because of conversion to THA.

AC C

91

-5-

ACCEPTED MANUSCRIPT Only one hip was randomly selected from each of the 28 patients with bilateral PAO to fulfill the

100

statistical assumption of independent observation. Thus, 192 PAO patients were reviewed (Fig. 1). All

101

demographic factors, including age at the time of surgery, sex, BMI, and follow-up duration, were

102

obtained from the patients’ medical records.

RI PT

99

103

105

Surgical techniques and rehabilitations

SC

104

The same group of nine surgeons performed all THAs using a posterolateral approach. The cup was inserted in a press-fit fashion; the diameter of the implant was 1-mm larger than that of the last reamer

107

used to prepare the acetabulum [24]. Two or three supplemental 6.5-mm-diameter screws were used in

108

most cases. After reduction was achieved using the trial stem and head, an intraoperative radiograph

109

was obtained to confirm the appropriate placement and determine the size of cup and stem, and neck

110

length. Patients were allowed full weight-bearing as tolerated with crutches or a walker for a period of

111

3–4 weeks postoperatively and advised to progress without ambulatory aids when they were pain free

112

and without a limp.

TE D

The indication for PAO was based on radiological evidence of acetabular dysplasia with a center-

EP

113

M AN U

106

edge angle of <25° and pain in the hip that interfered with daily activities. The same group of eight

115

surgeons performed all PAOs using a lateral approach. The PAO procedure has been reported in detail

116

[13,14,25]. Briefly, the patient was placed in the lateral decubitus position. After circumferential

117

exposure of the articular capsule, spherical osteotomy was performed using a curved osteotome,

118

starting at 20 mm proximal to the superior acetabular edge and passing through the midpoint between

119

the greater sacral fossa, the posterior edge of the acetabulum, and the innominate sulcus of the ischium.

120

Pubic osteotomy was performed lateral to the iliopubic tubercle. Active range of motion exercises and

121

partial weight-bearing on two crutches was started 1 and 2 weeks after surgery, respectively. Weight-6-

AC C

114

ACCEPTED MANUSCRIPT 122

bearing was gradually increased during postoperative rehabilitation according to the extent of the

123

patient’s pain. Walking with full weight was allowed 5 to 8 weeks after surgery; the patients used a

124

cane or crutches for a mean of 36 weeks.

126 127

RI PT

125 Statistical analysis

First, analyses for postoperative participation in sports and postoperative UCLA score were performed in 524 THA patients (Fig. 1). Wilcoxon signed-rank test and McNemar’s test were used to

129

compare preoperative and postoperative participation in sports and UCLA score, as deemed

130

appropriate. Univariate logistic regression analysis was performed to compare patients’ demographics

131

whether patients participated in sports. A multivariate binary logistic regression analysis and multiple

132

linear regression analysis were employed to determine the factors associated with postoperative

133

participation in sports and postoperative UCLA score, respectively. After excluding correlated factors,

134

the following six factors that could affect postoperative participation in sports and postoperative UCLA

135

score were selected for the multivariate analyses: age at the time of surgery, sex, BMI, cause of THA,

136

prior hip osteotomy, and preoperative participation in sports.

M AN U

TE D

Second, after excluding the patients (1) who had undergone hip osteotomy prior to THA and (2)

EP

137

SC

128

those whose THA was not caused by secondary OA due to acetabular dysplasia from the THA cohort,

139

we performed multivariate analyses in a total 487 acetabular dysplasia patients treated with either THA

140

or PAO (295 THA patients and 192 PAO patients, Fig. 1). Wilcoxon signed-rank test and McNemar’s

141

test were used to compare participation in sports and UCLA score before and after PAO, as deemed

142

appropriate. A multivariate binary logistic regression analysis and multiple linear regression analysis

143

were employed to determine the factors associated with postoperative participation in sports and

144

postoperative UCLA score, respectively, in 487 acetabular dysplasia patients treated with either THA or

AC C

138

-7-

ACCEPTED MANUSCRIPT PAO. After excluding correlated factors, the following six factors that could affect postoperative

146

participation in sports were selected for the multivariate analyses: age at the time of surgery, sex, BMI,

147

history of treatment for developmental hip dislocation, preoperative participation in sports, and type of

148

surgery (THA or PAO).

149

RI PT

145

Finally, propensity score matching was performed between 295 THA patients and 192 PAO patients (Fig. 1). Based on the propensity score, we generated 1:1 matched cohorts to facilitate comparison

151

between THA and PAO patients [26-28]. When calculating the propensity score by multivariate logistic

152

regression analysis for each patient, we included confounders consisting of age at the time of surgery,

153

sex, BMI, the time of surgery (i.e. follow-up duration), history of treatment for developmental hip

154

dislocation, and preoperative UCLA score. We matched the patients using the nearest neighbor

155

technique, with a predefined caliper width equal to 0.2 of the standard deviation of the logit of the

156

propensity score.[28] After propensity score matching, all variables were successfully matched. A total

157

of 62 patients were included for the propensity score-matched analysis in each cohort (Fig. 1). After

158

matching, Wilcoxon signed-rank test and McNemar’s test were used, as deemed appropriate, to

159

compare the characteristics between the THA and PAO cohorts (matched analysis; Fig. 1).

TE D

M AN U

SC

150

Values are expressed as mean ± standard deviation. The significance level was set at <0.05 for all

161

tests. All statistical analyses were performed using JMP Pro® Version 13.0 (SAS Institute Inc., Cary,

162

NC, USA).

AC C

163

EP

160

164

Results

165

Patients’ demographics in the THA cohort

166

In the THA cohort, 440 women and 84 men were included. The mean age at the time of surgery,

167

follow-up duration, and BMI were 62.9 ± 10.1 years (range, 22–86), 68.2 months (range, 12–199), and

168

22.9 ± 3.3 kg/m2 (range, 15.2–41.2), respectively. At the final follow-up, the average OHS was 41.0 ± -8-

ACCEPTED MANUSCRIPT 169

8.5 (range, 2–48). The causes of THA were primary OA in 88 patients, secondary OA due to acetabular

170

dysplasia in 366 patients, osteonecrosis of the femoral head in 56 hips, and rheumatoid arthritis in 14

171

hips. Prior to THA, 91 hips had osteotomies. Two cementless stems, PerFix and PerFix-910 (Kyocera, Osaka, Japan), were mainly used. PerFix

RI PT

172

stem was used in 109 hips in 2004, and PerFix-910 was used in 401 hips thereafter. Versys (Zimmer,

174

Warsaw, IN) and SROM (Depuy, Warsaw, IN) cementless stems were employed in eight and six hips,

175

respectively. Cementless acetabular cups were used in all primary THAs, including AMS cup

176

(Kyocera) in 516 hips and Trilogy cup (Zimmer) in 8 hips. Head size increased from 22 to 26 mm in

177

2002 and then to 32 mm in April 2010: 22 mm in 60 hips, 26 mm in 346 hips, and 32 mm in 118 hips.

178

Highly cross-linked polyethylene liner was used in 504 hips, while conventional polyethylene liner was

179

used in 20 hips.

180 Sports and activity levels in the THA cohort

TE D

181

M AN U

SC

173

The number of patients who participated in sports significantly increased from 81 (15.5%)

183

preoperatively to 160 (30.5%) postoperatively (P < .001). Overall, the mean UCLA score significantly

184

increased from 3.8 ± 2.1 preoperatively to 4.6 ± 1.7 at follow-up (P < .001). According to an American

185

Association of Hip and Knee Surgeons/Hip Society survey [29], most patients participated in low-

186

impact sports both preoperatively and postoperatively (Fig. 2). Only 33 (6.3%) patients participated in

187

high-impact sports preoperatively and 13 (2.5%) patients postoperatively. Additionally, no significant

188

differences in age at the time of surgery were found, regardless of patient participation in high-impact

189

sports either preoperatively (60.3 ± 8.8 vs. 63.1 ± 10.2, P = .13) or postoperatively (62.3 ± 9.6 vs. 62.9

190

± 10., P = .83).

AC C

EP

182

191

The preoperative and postoperative UCLA scores were significantly higher in patients who

192

participated in sports postoperatively than in those who did not (4.3 ± 2.3 vs. 3.5 ± 2.0, respectively, P -9-

ACCEPTED MANUSCRIPT < .001 and 5.7 ± 1.8 vs. 4.1 ± 1.5, respectively, P < .001; Table 1). In addition, among patients who

194

participated in sports postoperatively, there were significantly more patients who had participated in

195

sports preoperatively than those who had not (66 patients [41.2%] vs. 15 patients [4.1%], P < .001;

196

Table 1). Moreover, there were significantly more patients with male sex, and patients who participated

197

in sports postoperatively had higher postoperative OHS than those who did not (P < .001 and < .001,

198

respectively; Table 1). However, no significant differences were found in age at the time of surgery,

199

BMI, postoperative follow-up duration, cause of THA, and history of prior osteotomy between patients

200

who participated and those who did not participate in sports postoperatively (Table 1). Multivariate

201

binary logistic regression analysis demonstrated that preoperative participation in sports (odds ratio

202

[OR], 16.4; 95% confidence interval [CI], 8.85–30.5; P < .001) and male sex (OR, 2.60; 95% CI, 1.44–

203

4.69; P = .002) were significantly associated with postoperative participation in sports (Table 2). In a

204

multiple linear regression analysis, younger age at the time of surgery, male sex, lower BMI, and

205

preoperative participation in sports were identified as independent factors positively associated with

206

postoperative UCLA score for the THA cohort (P = .002, .02, .004, and < .001, respectively; Table 2).

TE D

M AN U

SC

RI PT

193

207 208

Differences in parameters between the THA and PAO cohorts In the PAO cohort, 177 women and 15 men were included. The mean age at the time of surgery,

210

follow-up duration, and BMI were 43.2 ± 10.4 years (range, 12–62), 97.4 months (range, 13–195), and

211

22.3 ± 3.4 kg/m2 (range, 15.4–33.9), respectively. At the final follow-up, the average OHS was 43.0 ±

212

6.2 (range, 23–48). The number of patients who participated in sports significantly increased from 61

213

(31.8%) preoperatively to 106 (55.2%) postoperatively (P < .001). Overall, the mean UCLA score

214

significantly increased from 4.7 ± 2.2 preoperatively to 5.5 ± 2.0 at follow-up (P < .001). Similar to the

215

THA cohort, most PAO patients participated in low-impact sports both preoperatively and

216

postoperatively (Fig. 2).

AC C

EP

209

-10-

ACCEPTED MANUSCRIPT Multivariate binary logistic regression analysis demonstrated that in 487 acetabular dysplasia

218

patients treated with either THA or PAO, preoperative participation in sports (OR, 13.1; 95% CI, 7.29–

219

23.6; P < .001), male sex (OR, 2.36; 95% CI, 1.16–4.76; P = .002), and PAO (OR, 2.29; 95% CI, 1.27–

220

4.14; P = .006) were significantly associated with postoperative participation in sports (Table 3).

221

Multiple linear regression analysis showed that younger age at the time of surgery, lower BMI, and

222

preoperative participation in sports were identified as independent factors positively associated with

223

postoperative UCLA score (P = .001, .004, and < .001, respectively), whereas the type of surgery (THA

224

or PAO) was not (P = .22; Table 3).

SC

RI PT

217

After propensity score matching, no significant differences in all matched variables and

226

preoperative participation in sports between THA and PAO cohorts were noted (Table 4). The number

227

of patients who participated in sports in the THA and PAO cohorts significantly increased, respectively,

228

from 8 and 16 (12.9 and 25.8%, respectively) preoperatively to 20 and 30 (32.3 and 51.6%,

229

respectively) postoperatively (P < .001 and < .001, respectively). Similarly, the mean UCLA scores in

230

the THA and PAO cohorts significantly increased, respectively, from 3.8 ± 2.3 and 3.8 ± 1.5

231

preoperatively to 5.0 ± 1.6 and 5.2 ± 1.9 at follow-up (P < .001 and < .001, respectively). In the

232

matched analysis, the THA cohort showed significantly lower rate of postoperative participation in

233

sports than the PAO cohort (Table 4). However, no significant differences in postoperative UCLA score

234

and postoperative OHS between the two cohorts were found.

236

TE D

EP

AC C

235

M AN U

225

Discussion

237

This study had three major findings. First, both the rate of participation in sports and UCLA

238

score significantly increased with favorable clinical outcomes in Asian cohort after THA. Second,

239

preoperative participation in sports was the factor that is most associated with both participation in

240

sports and higher UCLA score after THA. Third, no significant difference in postoperative UCLA score -11-

ACCEPTED MANUSCRIPT 241

between the two matched cohorts was found, which was similar to the result of our multivariate

242

analysis in 487 acetabular dysplasia patients treated with either THA or PAO.

243

THA almost doubled the rate of participation in sports and significantly increased UCLA score in this study, which is consistent with previous studies [2,4-6]. In previous studies, 35–80% of patients

245

participated in sports preoperatively and 64–83% postoperatively [2,6,30], and the UCLA score

246

increased from 3.3–5.0 preoperatively to 6.2–8.0 postoperatively [4,5,18,19]. Therefore, both the rate

247

of participation in sports and UCLA score were lower in this study than those in previous studies. The

248

differences in the rate of participation in sports and UCLA score between our study and previous

249

studies could be attributed to the difference in patients’ backgrounds, such as age (mean, 62.9 years vs.

250

32.7–70.6 years), distribution of the stage of OA, the type of sports the patients were involved in (low-

251

impact, intermediate-impact, or high-impact sports), wishes to avoid wear or dislocation, or advice

252

from their surgeons. Additionally, the previous studies were reported from Western countries [2,4-

253

6,18,19], whereas all patients in this study were Asians. Participation in sports or activity levels may

254

differ between Western and Asian countries probably because more THAs instead of PAOs are

255

performed for patients with early stage of OA in Western countries than in Asian countries;

256

nevertheless, further studies are needed. Regardless of these differences, our study demonstrated that

257

THA for Asian patients provided a significant increase in both the rate of participation in sports and

258

UCLA score with favorable clinical results, which is consistent with the findings of previous studies

259

from Western countries [2,4-6,18,19].

SC

M AN U

TE D

EP

AC C

260

RI PT

244

Our analysis in 524 THA patients demonstrated that preoperative participation in sports was the

261

factor that is most associated with both postoperative participation in sports and higher postoperative

262

UCLA score. Several studies [3,31,32] reported that the number of THA patients who participated in

263

sports or showed higher activity levels postoperatively was higher among those patients who had

264

participated in sports or showed higher activity level preoperatively than those who had not, which is -12-

ACCEPTED MANUSCRIPT consistent with our results. Similar to previous, male sex was also associated with both postoperative

266

participation in sports and higher postoperative UCLA score in this study [3,31,32]. Younger age at the

267

time of surgery and smaller lower BMI was associated with higher postoperative UCLA score, but not

268

postoperative participation in sports, which could be associated with the finding that most patients

269

participated mainly in low-impact sports; low-impact sports is relatively easier for older or obese

270

patients compared with high-impact sports and does not need higher activity levels. These findings may

271

be essential in advising patients on participation in sports after THA.

SC

272

RI PT

265

Similar to the result of our analysis in 524 THA patients, preoperative participation in sports was the factor that is most associated with both postoperative participation in sports and higher

274

postoperative UCLA score in 487 acetabular dysplasia patients treated with either THA or PAO, which

275

is also consistent with previous studies of PAO cohort [23,33]. Younger age at the time of surgery,

276

male sex, and lower BMI were also associated with postoperative participation in sports or higher

277

postoperative UCLA score in the multivariate analysis. These results were comparable with previous

278

studies of THA cohort [3,31,32]; however, similar to our analysis in 524 THA patients, several

279

discrepancies between participation in sports and UCLA score were observed. Additionally, multiple

280

linear regression analysis demonstrated that the type of surgery was not independently associated with

281

postoperative UCLA score. Previous studies [18,19] showed that no significant differences in

282

postoperative activity levels including UCLA score between THA and PAO cohorts were found (6.6–

283

6.9 and 7.4–7.5 for UCLA score, respectively), which is consistent with our result.

TE D

EP

AC C

284

M AN U

273

By contrast, our multivariate binary logistic regression analysis in 487 acetabular dysplasia

285

patients treated with either THA or PAO showed that PAO was significantly associated with

286

postoperative participation in sports. Generally, the PAO cohort showed higher preoperative clinical

287

scores than the THA cohort because the type of surgery depends on the stage of OA (PAO for early

288

stage; THA for advanced stage) [15,18,19], which may lead to more atrophy of the attendant muscle -13-

ACCEPTED MANUSCRIPT and cautious behavior in THA cohort than in PAO cohort. Moreover, the number of patients who had

290

participated in sports preoperatively was higher in the PAO cohort than in the THA cohort in this study,

291

which could be attributed to the difference in baseline patient characteristics. As demonstrated in our

292

study and previous studies [3,23,31,32], preoperative participation in sports significantly affects

293

postoperative participation in sports in both the THA and PAO cohorts. Thus, the analysis, which

294

matched not only patients’ demographics but also preoperative participation in sports or activity levels,

295

was useful to compare more appropriately postoperative participation in sports and activity levels

296

between the two cohorts.

SC

The propensity score-matching method enabled the comparison of postoperative outcomes

M AN U

297

RI PT

289

between the THA and PAO cohorts with similar baseline characteristics, including preoperative

299

participation in sports and preoperative UCLA score. The matched analysis showed no significant

300

differences in postoperative UCLA score and postoperative OHS between the two cohorts. This result

301

is comparable with that of previous studies [18,19] and reconfirmed that both THA and PAO are

302

effective treatments to improve activity levels in middle-aged patients with secondary OA due to

303

acetabular dysplasia. Nevertheless, the THA cohort showed significantly lower rate of postoperative

304

participation in sports despite the absence of a significant difference in postoperative UCLA score

305

compared with the matched PAO cohort, which is consistent with our multivariate analyses in 487

306

acetabular dysplasia patients treated with either THA or PAO. This lower rate of postoperative

307

participation in sports in the THA cohort could be attributed to specific reasons, such as fear of

308

damaging the prosthetic joint. Moreover, several studies reported that patients do not participate in

309

sports after THA not only because of pain and inability to do the required movements but also because

310

of medical advice and fear of damaging the prosthetic joint [2,3,30,34]. Surgeons know that revision

311

THA will be more difficult than primary THA while the failure of PAO means the conversion to THA

312

which can be successful as with primary THA without prior PAO [35,36]. In addition, similar to THA

AC C

EP

TE D

298

-14-

ACCEPTED MANUSCRIPT patients, most PAO patients participated in low-impact sports, which could explain the comparable

314

postoperative UCLA score between the THA and PAO cohorts despite the lower rate of postoperative

315

participation in sports in the THA cohort. These findings may be helpful for patients with secondary

316

OA due to acetabular dysplasia in deciding whether to undergo PAO or THA.

317

RI PT

313

This study has several limitations. First, this retrospective study, which used a self-administered questionnaire, has a potential recall and selection bias. A further prospective study with a longer

319

follow-up duration is needed. Second, the questionnaire for sports did not include the specific

320

competitiveness level for each sport and has not been comprehensively validated. However, we used

321

the UCLA score to assess physical activity levels. UCLA score and OHS have been validated [20-22],

322

and most of the evaluation items were objective. Third, there were more female patients than male

323

patients in both the THA and PAO cohorts, which may not accurately reflect the general patient

324

population, especially for the THA cohort in Western countries. However, approximately 90% of OA is

325

secondary OA due to acetabular dysplasia in Asian countries, and most of the patients with acetabular

326

dysplasia are female [3,37]. We believe that the samples in this study accurately reflect the patients’

327

population in Asian countries. Fourth, this retrospective study has inherent heterogeneousness of some

328

variables especially in THA cohort such as the difference in the stage of OA, and different type of

329

implants. Fifth, this study lacked quantified polyethylene wear. However, most of the polyethylene

330

liners were highly cross-linked polyethylene liners (96.2%) whose wear-resistant performance has been

331

established [38]. Finally, our findings need to be re-examined especially if a new confounding factor is

332

discovered given that the propensity score-matching method cannot control for unknown confounding

333

factors.

AC C

EP

TE D

M AN U

SC

318

334 335

Conclusion -15-

ACCEPTED MANUSCRIPT After THA, both participation in sports and UCLA score significantly increased with favorable

337

clinical outcomes in an Asian cohort. Overall, 30.5% of patients participated in sports postoperatively,

338

and preoperative participation in sports was the factor that is most associated with both participation in

339

sports and higher activity levels after THA. Although the THA cohort showed lower rate of

340

postoperative participation in sports than the propensity score-matched PAO cohort, no significant

341

difference in postoperative UCLA score between the two matched cohorts was found, which is similar

342

to the result of our multivariate analysis in 487 acetabular dysplasia patients treated with either THA or

343

PAO. These results may be useful in the decision-making process for patients with symptomatic hip

344

disorder.

M AN U

SC

RI PT

336

AC C

EP

TE D

345

-16-

ACCEPTED MANUSCRIPT 346

References

347

[1]

Learmonth ID, Young C, Rorabeck C. The operation of the century: total hip replacement. Young C, Rorabeck C. The operation of the century: total hip replacement. Lancet

349

2007;370:1508–19. doi:10.1016/S0140-6736(07)60457-7.

350

[2]

RI PT

348

Ollivier M, Frey S, Parratte S, Flecher X, Argenson JN. Pre-operative function, motivation and duration of symptoms predict sporting participation after total hip replacement. Bone Joint J

352

2014;96-B:1041–6. doi:10.1302/0301-620X.96B8.32813. [3]

354 355

Abe H, Sakai T, Nishii T, Takao M, Nakamura N, Sugano N. Jogging After Total Hip Arthroplasty. Am J Sports Med 2013;42:131–7. doi:10.1177/0363546513506866.

[4]

M AN U

353

SC

351

Heiberg KE, Figved W. Physical Functioning and Prediction of Physical Activity After Total

356

Hip Arthroplasty: Five-Year Followup of a Randomized Controlled Trial. Arthritis Care Res

357

2016;68:454–62. doi:10.1002/acr.22679.

358

[5]

Innmann MM, Weiss S, Andreas F, Merle C, Streit MR. Sports and physical activity after cementless total hip arthroplasty with a minimum follow-up of 10 years. Scand J Med Sci

360

Sports 2016;26:550–6. doi:10.1111/sms.12482.

361

[6]

TE D

359

Chatterji U, Ashworth MJ, Lewis PL, Dobson PJ. Effect of total hip arthroplasty on recreational and sporting activity. ANZ J Surg 2004;74:446–9. doi:10.1111/j.1445-

363

1433.2004.03028.x. [7]

AC C

364

EP

362

Jingushi S, Ohfuji S, Sofue M, Hirota Y, Itoman M, Matsumoto T, et al. Multiinstitutional

365

epidemiological study regarding osteoarthritis of the hip in Japan. J Orthop Sci 2010;15:626–

366

31. doi:10.1007/s00776-010-1507-8.

367

[8]

Dudda M, Kim Y-J, Zhang Y, Nevitt MC, Xu L, Niu J, et al. Morphologic differences between

368

the hips of Chinese women and white women: could they account for the ethnic difference in

369

the prevalence of hip osteoarthritis? Arthritis Rheum 2011;63:2992–9. doi:10.1002/art.30472. -17-

ACCEPTED MANUSCRIPT 370

[9]

Hoaglund FT, Oishi CS, Gialamas GG. Extreme variations in racial rates of total hip

371

arthroplasty for primary coxarthrosis: a population-based study in San Francisco. Ann Rheum

372

Dis 1995;54:107–10. doi:10.1136/ard.54.2.107. [10]

374 375

Harris WH. Etiology of osteoarthritis of the hip. Clin Orthop Relat Res 1986:20–33.

RI PT

373

doi:10.1097/00003086-198612000-00004. [11]

Kaneuji A, Sugimori T, Ichiseki T, Fukui K, Takahashi E, Matsumoto T. Rotational

Acetabular Osteotomy for Osteoarthritis with Acetabular Dysplasia: Conversion Rate to Total

377

Hip Arthroplasty within Twenty Years and Osteoarthritis Progression After a Minimum of

378

Twenty Years. J Bone Joint Surg Am 2015;97:726–32. doi:10.2106/JBJS.N.00667. [12]

380 381

M AN U

379

SC

376

Delimar D, Cicak N, Klobucar H, Pećina M, Korzinek K. Acetabular roof reconstruction with pedicled iliac graft. Int Orthop 2002;26:344–8. doi:10.1007/s00264-002-0381-1.

[13]

Fujii M, Nakashima Y, Noguchi Y, Yamamoto T, Mawatari T, Motomura G, et al. Effect of intra-articular lesions on the outcome of periacetabular osteotomy in patients with symptomatic

383

hip dysplasia. J Bone Joint Surg Br 2011;93:1449–56. doi:10.1302/0301-620X.93B11.27314.

384

[14]

TE D

382

Matsuo A, Jingushi S, Nakashima Y, Yamamoto T, Mawatari T, Noguchi Y, et al. Transposition osteotomy of the acetabulum for advanced-stage osteoarthritis of the hips. J

386

Orthop Sci 2009;14:266–73. doi:10.1007/s00776-009-1327-x. [15]

Hsieh P-H, Huang K-C, Lee PC, Chang Y-H. Comparison of periacetabular osteotomy and

AC C

387

EP

385

388

total hip replacement in the same patient: a two- to ten-year follow-up study. J Bone Joint Surg

389

Br 2009;91:883–8. doi:10.1302/0301-620X.91B7.22183.

390

[16]

391 392 393

Garbuz DS, Awwad MA, Duncan CP. Periacetabular osteotomy and total hip arthroplasty in patients older than 40 years. J Arthroplasty 2008;23:960–3. doi:10.1016/j.arth.2007.08.015.

[17]

Sharifi E, Sharifi H, Morshed S, Bozic K, Diab M. Cost-effectiveness analysis of periacetabular osteotomy. J Bone Joint Surg Am 2008;90:1447–56. doi:10.2106/JBJS.G.00730. -18-

ACCEPTED MANUSCRIPT 394

[18]

Gray BL, Stambough JB, Baca GR, Schoenecker PL, Clohisy JC. Comparison of

395

contemporary periacetabular osteotomy for hip dysplasia with total hip arthroplasty for hip

396

osteoarthritis. Bone Joint J 2015;97-B:1322–7. doi:10.1302/0301-620X.97B10.35741. [19]

Kalore NV, Cheppalli SPR, Daner WE, Jiranek WA. Acetabular Dysplasia in Middle-Aged

RI PT

397 398

Patients: Periacetabular Osteotomy or Total Hip Arthroplasty? J Arthroplasty 2016;31:1894–8.

399

doi:10.1016/j.arth.2016.02.032. [20]

Uesugi Y, Makimoto K, Fujita K, Nishii T, Sakai T, Sugano N. Validity and responsiveness of

SC

400

the Oxford hip score in a prospective study with Japanese total hip arthroplasty patients.

402

Orthop Sci 2009;Volume 14:35–9. doi:10.1007/s00776-008-1292-9.

403

[21]

M AN U

401

Naal FD, Impellizzeri FM, Eisenhart-Rothe von R, Mannion AF, Leunig M. Reproducibility,

404

validity, and responsiveness of the hip outcome score in patients with end-stage hip

405

osteoarthritis. Arthritis Care Res 2012;64:1770–5. doi:10.1002/acr.21746.

406

[22]

Amstutz HC, Thomas BJ, Jinnah R, Kim W, Grogan T, Yale C. Treatment of primary osteoarthritis of the hip. A comparison of total joint and surface replacement arthroplasty. J

408

Bone Joint Surg Am 1984;66:228–41.

409

[23]

TE D

407

Hara D, Hamai S, Fukushi J-I, Kawaguchi K-I, Motomura G, Ikemura S, et al. Does Participation in Sports Affect Osteoarthritic Progression After Periacetabular Osteotomy? Am

411

J Sports Med 2017; doi:10.1177/0363546517707942. [24]

AC C

412

EP

410

Nakashima Y, Sato T, Yamamoto T, Motomura G, Ohishi M, Hamai S, et al. Results at a

413

minimum of 10 years of follow-up for AMS and PerFix HA-coated cementless total hip

414

arthroplasty: impact of cross-linked polyethylene on implant longevity. J Orthop Sci

415

2013;18:962–8. doi:10.1007/s00776-013-0456-4.

-19-

ACCEPTED MANUSCRIPT 416

[25]

Hamai S, Nakashima Y, Akiyama M, Kuwashima U, Yamamoto T, Motomura G, et al. Ischio-

417

pubic stress fracture after peri-acetabular osteotomy in patients with hip dysplasia. Int Orthop

418

2014;38:2051–6. doi:10.1007/s00264-014-2375-1. [26]

Hamai S, Nakashima Y, Mashima N, Yamamoto T, Kamada T, Motomura G, et al.

RI PT

419

Comparison of 10-year clinical wear of annealed and remelted highly cross-linked

421

polyethylene: A propensity-matched cohort study. J Mech Behav Biomed Mater 2016;59:99–

422

107. doi:10.1016/j.jmbbm.2015.12.022.

423

[27]

SC

420

Kubo S, Nakayamada S, Nakano K, Hirata S, Fukuyo S, Miyagawa I, et al. Comparison of the efficacies of abatacept and tocilizumab in patients with rheumatoid arthritis by propensity

425

score matching. Ann Rheum Dis 2016;75:1321–7. doi:10.1136/annrheumdis-2015-207784.

426

[28]

M AN U

424

Bangalore S, Guo Y, Samadashvili Z, Blecker S, Xu J, Hannan EL. Everolimus-eluting stents

427

or bypass surgery for multivessel coronary disease. N Engl J Med 2015;372:1213–22.

428

doi:10.1056/NEJMoa1412168. [29]

Klein GR, Levine BR, Hozack WJ, Strauss EJ, D’Antonio JA, Macaulay W, et al. Return to

TE D

429

athletic activity after total hip arthroplasty. Consensus guidelines based on a survey of the Hip

431

Society and American Association of Hip and Knee Surgeons. J Arthroplasty 2007;22:171–5.

432

doi:10.1016/j.arth.2006.09.001. [30]

434 435

Wylde V, Blom A, Dieppe P, Hewlett S, Learmonth I. Return to sport after joint replacement. J

AC C

433

EP

430

Bone Joint Surg Br 2008;90:920–3. doi:10.1302/0301-620X.90B7.20614. [31]

Lübbeke A, Zimmermann-Sloutskis D, Stern R, Roussos C, Bonvin A, Perneger T, et al.

436

Physical activity before and after primary total hip arthroplasty: a registry-based study.

437

Arthritis Care Res 2014;66:277–84. doi:10.1002/acr.22101.

-20-

ACCEPTED MANUSCRIPT 438

[32]

Williams DH, Greidanus NV, Masri BA, Duncan CP, Garbuz DS. Predictors of Participation in

439

Sports After Hip and Knee Arthroplasty. Clin Orthop Relat Res 2012;470:555–61.

440

doi:10.1007/s11999-011-2198-y. [33]

Novais EN, Heyworth B, Murray K, Johnson VM, Kim Y-J, Millis MB. Physical activity level

RI PT

441 442

improves after periacetabular osteotomy for the treatment of symptomatic hip dysplasia. Clin

443

Orthop Relat Res 2013;471:981–8. doi:10.1007/s11999-012-2578-y.

445 446

Huch K. Sports activities 5 years after total knee or hip arthroplasty: the Ulm Osteoarthritis

SC

[34]

Study. Ann Rheum Dis 2005;64:1715–20. doi:10.1136/ard.2004.033266. [35]

Ito H, Takatori Y, Moro T, Oshima H, Oka H, Tanaka S. Total Hip Arthroplasty After

M AN U

444

447

Rotational Acetabular Osteotomy. J Arthroplasty 2015;30:403–6.

448

doi:10.1016/j.arth.2014.10.002.

449

[36]

Amanatullah DF, Stryker L, Schoenecker P, Taunton MJ, Clohisy JC, Trousdale RT, et al. Similar Clinical Outcomes for THAs With and Without Prior Periacetabular Osteotomy. Clin

451

Orthop Relat Res 2015;473:685–91. doi:10.1007/s11999-014-4026-7. [37]

453 454

Sugano N, Nishii T, Takao M. The Asian Hip. In: Callaghan JJ, Rosenberg AG, Rubash HE, editors. The Adult Hip. 2nd ed., Philadelphia: LWW; 2006, p. 1060–77.

[38]

Sato T, Nakashima Y, Akiyama M, Yamamoto T, Mawatari T, Itokawa T, et al. Wear resistant

EP

452

TE D

450

performance of highly cross-linked and annealed ultra-high molecular weight polyethylene

456

against ceramic heads in total hip arthroplasty. J Orthop Res 2012;30:2031–7.

457

doi:10.1002/jor.22148.

AC C

455

-21-

ACCEPTED MANUSCRIPT

1

Acknowledgements The authors would like to thank Junji Kishimoto, a statistician from the Digital Medicine

3

Initiative, Kyushu University, for his valuable comments and suggestions in regard to statistical

4

analysis, the secretaries from Department of Orthopaedic Surgery, Kyushu University, for their

5

assistance with mailing a questionnaire, and Editage (www.editage.jp) for English language

6

editing. The present work was supported by JSPS KAKENHI Grant Number 15K10450.

RI PT

2

AC C

EP

TE D

M AN U

SC

7

1

ACCEPTED MANUSCRIPT

1

Figure legends

2 Fig. 1. Schematic representation of study cohort development.

4

THA: total hip arthroplasty, PAO: periacetabular osteotomy, OA: osteoarthritis.

RI PT

3

5

Fig. 2. Patient participation in sports before and after total hip arthroplasty (THA) and

7

periacetabular osteotomy (PAO).

8

The data before the propensity score matching are shown. Multiple answers were allowed in the

9

questionnaire. Racket games included tennis, badminton, and table tennis. Ball games included

10

soccer, baseball, basketball, and volleyball. Others included gate-ball, archery, fishing, hiking,

11

skiing, snowboarding, bodyboarding, Judo, the Japanese art of fencing (kendo), and karate. The

12

sports were classified as either low-, intermediate-, or high-impact sports based on an American

13

Association of Hip and Knee Surgeons/Hip Society survey [29].

M AN U

TE D

EP AC C

14

SC

6

1

ACCEPTED MANUSCRIPT Table 1. Results of univariate logistic regression of associated factors for participation in sports after THA Sports (+)

Sports (-)

Crude odds ratio

(N = 160)

(N = 368)

(95% CI)

Age at the time of surgery, y

62.3 ± 9.5

63.1 ± 10.4

0.99 (0.97–1.01)

Male sex, n (%)

40 (25.0)

44 (12.1)

Body mass index, kg/m2

23.0 ± 2.9

22.8 ± 3.4

Follow-up duration, mo

67.4 ± 43.2

68.6 ± 41.8

1.00 (0.995–1.003)

.78

Acetabular dysplasia#, n (%)

116 (72.5)

250 (68.7)

1.20 (0 80–1.81)

.38

Prior hip osteotomy, n (%)

32 (20.0)

59 (16.2)

1.29 (0.80–2.08)

.29

Preoperative participation in sports, n (%)

66 (41.3)

15 (4.1)

16.3 (8.92–29.9)

< .001*

Preoperative UCLA score

4.3 ± 2.3

3.5 ± 2.0

1.20 (1.10–1.31)

< .001*

5.7 ± 1.8

4.1 ± 1.5

1.77 (1.55–2.02)

< .001*

43.1 ± 6.8

40.1 ± 9.0

1.05 (1.02–1.08)

< .001*

P values

< .001*

1.01 (0.96–1.07)

.62

SC

2.42 (1.503.91)

M AN U

EP

Postoperative Oxford Hip Score

TE D

Postoperative UCLA score

.40

RI PT

Parameters

AC C

Continuous values are expressed as mean ± SD. THA: total hip arthroplasty, UCLA: University of California, Los Angeles, 95%CI: 95% confidence intervals. *Significantly different between group (P < .05). #Compared with primary osteoarthritis, osteonecrosis of the femoral head, or rheumatoid arthritis.

1

ACCEPTED MANUSCRIPT Table 2. Results of multivariate logistic regression of associated factors for postoperative participation in sports and multiple linear regression of associated factors for postoperative UCLA score after THA

Adjusted odds ratio

For postoperative UCLA score

RI PT

For postoperative participation in sports

Estimate

Parameters

P values

T values

P values

-0.022 (0.001)

-3.17

.002*

.002*

0.247 (0.106)

2.32

.02*

.98

-0.067 (0.023)

-2.90

.004*

(standard error)

0.99 (0.97–1.01)

Male sex

2.60 (1.44–4.69)

Body mass index,

1.00 (0.93–1.07)

Acetabular dysplasia#

1.46 (0.88–2.44)

.14

0.070 (0.084)

0.83

.41

Prior hip osteotomy

1.21 (0.70–2.09)

.50

0.13 (0.096)

1.39

.17

Preoperative participation in sports

16.4 (8.85–30.5)

< .001*

0.597 (0.101)

5.92

< .001*

AC C

TE D

M AN U

Age at the time of surgery

EP

.23

SC

(95% CI)

UCLA: University of California, Los Angeles, THA: total hip arthroplasty, 95%CI: 95% confidence intervals. *Statistically significant (P < .05). # Compared with primary osteoarthritis, osteonecrosis of the femoral head, or rheumatoid arthritis.

1

ACCEPTED MANUSCRIPT Table 3. Results of multivariate logistic regression of associated factors for postoperative participation in sports and multiple liner regression of associated factors for postoperative UCLA score in 487 acetabular dysplasia patients treated with either THA or PAO

Adjusted odds ratio

For postoperative UCLA score Estimate

P values

(95% CI)

1.00 (0.98–1.02)

Male sex

P values

-0.021 (0.008)

-2.53

.001*

2.36 (1.16–4.76)

.002*

0.217 (0.135)

1.68

.09

1.01 (0.94–1.07)

.86

-0.036 (0.024)

-1.49

.004*

TE D

Body mass index

T values (standard error)

.87

M AN U

Age at the time of surgery

SC

Parameters

RI PT

For postoperative participation in sports

0.92 (0.58–1.46)

.72

-0.018 (0.086)

-0.21

.84

Preoperative participation in sports

13.1 (7.29–23.6)

< .001*

0.582 (0.097)

6.00

< .001*

2.29 (1.27–4.14)

.006*

0.138 (0.112)

1.23

.22

AC C

PAO

EP

History of treatment for developmental hip dislocations

UCLA: University of California, Los Angeles, THA: total hip arthroplasty, PAO: periacetabular osteotomy, 95%CI: 95% confidence intervals. *Statistically significant (P < .05).

1

ACCEPTED MANUSCRIPT

Table 4. Demographics of the patients with secondary osteoarthritis due to acetabular dysplasia after propensity score matching PAO (N = 62)

P values

51.3 ± 6.8

51.4 ± 6.2

.88

4 (6.5)

5 (8.1)

.54

Body mass index, kg/m2

22.7 ± 3.4

22.7 ± 2.9

.79

Follow-up duration, mo

85.1 ± 43.0

90.3 ± 45.1

.22

History of treatment for developmental hip dislocations, n (%)

39 (62.9)

41 (66.1)

.85

Preoperative UCLA score

3.8 ± 2.3

3.8 ± 1.5

.89

8 (12.9)

16 (25.8)

.96

20 (32.3)

30 (51.6)

.046*

5.0 ± 1.6

5.2 ± 1.9

.47

43.7 ± 5.8

42.4 ± 6.3

.27

Age at the time of surgery, y

Postoperative UCLA score Postoperative Oxford Hip Score

TE D

AC C

Postoperative participation in sports, n (%)

EP

Preoperative participation in sports, n (%)

M AN U

Male sex, n (%)

RI PT

THA (N = 62)

SC

Parameters

Continuous values are expressed as mean ± SD. Age at the time of surgery, sex, body mass index, follow-up duration, history of treatment for developmental hip dislocation, and preoperative UCLA score were included to the propensity-score matching as confounders. THA: total hip arthroplasty, PAO: periacetabular osteotomy, UCLA: University of California, Los Angeles, 95%CI: 95% confidence intervals. *Significantly different between group (P < .05) 1

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT