Hip osteoarthritis (OA) is one of the most common and most disabling conditions affecting the older population; most adult patients with hip OA in Japan have hip dysplasia, a disease concept that indicates radiographic findings of mechanical instability1. Shifting the efforts of hip OA research and public health intervention to primary prevention2 may greatly improve the current management paradigm. However, among the risk factors for hip OA in adults, only a few are known as modifiable factors, such as body weight3. A reevaluation of hip OA as having a sequential pathology from newborns to older adults might effectively reveal modifiable behavioral changes in the earlier life stage4.

Developmental dysplasia of the hip (DDH) is a disease characterized by hip dislocation or subluxation and acetabular morphological abnormalities observed in infants and children5. The incidence of DDH is racially different and is reportedly high in Japanese6, Singaporean5, Native American7, and Turkish populations8, and in Sami populations in Scandinavia9. In the Japanese population prior to 1965, dislocation was present in approximately 1.1% to 3.5% of all newborns6, and severe subluxation was present among adults10 (Fig. 1).

Fig. 1:

Radiograph demonstrating a once-typical case of hip osteoarthritis (OA) in Japan. The radiograph shows the hips of a female patient born in 1964 who had a history of DDH treatment in childhood. She began experiencing hip pain in her mid-40s and first visited a core hospital at 58 years old. We diagnosed both hips as having secondary OA due to hip dysplasia; the left hip showed Crowe type-II subluxation.

Hip dislocation was common worldwide half a century ago11,12. The Japanese people had a tradition of forcibly maintaining the legs of a newborn in an extended position with use of a diaper6,13, as is done through swaddling in some regions of China14; this position was confirmed to cause dislocated hips in animal models15. However, around 1972 to 1973, Ishida and Yamamuro6,13 initiated a program to educate obstetricians, midwives, health nurses, and pregnant women that the hips of newborns should be in the naturally flexed-leg position, as the primary prevention of DDH, and that this position promotes acetabular development. Manufacturers of diapers, diaper covers, and baby clothes were advised on the proper design of clothing that allows newborns’ hips and knees to move freely. This campaign, which utilized a population approach16, was extended nationwide in Japan6,17, and reports have shown that DDH has since drastically reduced6,13,18.

The perinatal education campaign for preventing DDH via a population approach might be associated with the epidemiology of hip OA in adolescence and adulthood; however, to our knowledge, no studies have reported this perspective. The purpose of the present study was to describe the life course epidemiology of hip OA in adolescents and adults and to investigate its association with past exposure to the perinatal education campaign for the primary prevention of DDH.

Materials and Methods Study Design

Following approval from our institutional review board, patient recruitment for this multicenter, cross-sectional study was conducted from January 1, 2022, to December 31, 2022. We utilized a life course approach, which investigates the long-term effects of physical and social exposures in the fetal period, childhood, adolescence, and early adulthood on adult disease risk4. An opt-out patient consent procedure was utilized to obtain the routine medical care information of the patients. This study was conducted per the tenets of the Medical Research Involving Human Subjects Act and the principles of the Declaration of Helsinki. This paper was structured according to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement19. We enrolled participants from 12 core hospitals (8 special-function hospitals and 4 regional medical care support hospitals) located in various prefectures of Japan, subsequently recruiting >1,000 patients with hip OA who first reported to the core hospitals in 2022 (Fig. 2).

Fig. 2:

Flow diagram showing the patient-selection process for the analysis. The prefectures in gray indicate the location of the 12 core hospitals (8 special-function hospitals and 4 regional medical care support hospitals) that participated in the study.

Patient Selection

New patients ≥15 years old who visited any of the 12 core hospitals with a chief concern related to hip pain were assessed for eligibility. Following prior studies1,20, we included patients diagnosed with hip OA on the basis of clinical and radiographic findings of hip pain, no history of inflammation, and joint space narrowing. We also included patients with hip pain without joint space narrowing who had abnormal joint morphology such as hip dysplasia21–23. Patients with a history of previous hip surgery on the affected side were excluded because of difficulties in morphological evaluation.

Radiographic Assessment of Hip OA

The stage of OA was assessed with use of the Kellgren-Lawrence grade24. The minimum joint space width25 was measured and utilized as supportive information because it could be increased in conditions such as hip OA following Legg-Calvé-Perthes disease26.

Primary Outcome: Descriptive Epidemiology Prevalence of Hip OA Subclassifications by Patient Age

The hip OA subclassification was determined with use of an algorithm based on the progress of the arthropathy, patient history, radiographs (supine hip anteroposterior and lateral or Lauenstein views), and magnetic resonance imaging for specific patients (Fig. 3). Hip OA etiology was subclassified into 9 types: rapidly destructive coxopathy, Legg-Calvé-Perthes disease, slipped capital femoral epiphysis, trauma, skeletal dysplasia, hip dysplasia, femoroacetabular impingement (FAI), subchondral insufficiency fracture (SIF) of the femoral head, and primary OA.

Fig. 3:

Diagnostic criteria of the hip OA subclassification. This algorithm was shared with the 12 core hospitals before January 1, 2022. The final diagnoses were determined by a conference at each facility; therefore, a reliability assessment was not performed. If the diagnosis was inconclusive, a consensus meeting was held among members of the collaborating institutions. OA = osteoarthritis, RDC = rapidly destructive coxopathy, LCPD = Legg-Calvé-Perthes disease, SCFE = slipped capital femoral epiphysis, DDH = developmental dysplasia of the hip, LCEA = lateral center-edge angle, ARO = acetabular roof obliquity, FAI = femoroacetabular impingement, MRI = magnetic resonance imaging, SIF = subchondral insufficiency fracture.

Hip dysplasia was diagnosed according to the criteria by Nakamura et al.27: a Sharp angle of ≥45°, a lateral center-edge angle (LCEA) of ≤20°, or an acetabular roof obliquity (ARO) of ≥15°. These 3 angles were measured using the lateral acetabular rim as described by Wiberg12,28, with the reference line passing through the bilateral acetabular teardrops on the anteroposterior view.

FAI29 was diagnosed according to the Japanese Hip Society criteria30. Pincer-type FAI was diagnosed in patients with an LCEA of ≥40°, an LCEA of ≥30° and an ARO of ≤0°, or an LCEA of ≥25° and a positive crossover sign. Cam-type FAI was diagnosed in patients with an LCEA of ≥25° and 1 of the following: an alpha angle of ≥55°, a head-neck offset ratio of <0.14, a positive pistol grip deformity, or a positive herniation pit. SIF of the femoral head, caused by osteoporosis and resultant femoral head collapse, has also been proposed as an etiology of secondary OA31.

Secondary Outcomes Patients with a History of DDH Treatment in Childhood

If primary prevention successfully prevents DDH, a patient would not need to receive any treatment for DDH in childhood. All patients included in the primary analysis were surveyed for their history of DDH treatment in childhood. We utilized information from the patient’s self-reported history of DDH treatment before 15 years of age (i.e., before the closure of the epiphyseal line). Given the likelihood of recall bias and the fact that orthotic treatment is usually applied to both sides, we considered a patient to have undergone treatment if there was a history of treatment on either side.

Severe Subluxation Among Patients with Hip Dysplasia

We assumed that severe subluxation is a result of the failure of both primary and secondary prevention. Therefore, we utilized the Crowe classification10, the most commonly utilized classification system for dysplastic hips in adult patients32, to evaluate subluxation in patients with secondary hip OA due to hip dysplasia. We defined severe subluxation as Crowe type II, III, or IV. The measurements were made on the radiographic anteroposterior view and were classified on the basis of the amount of subluxation: Crowe type I, <50% subluxation; Crowe type II, 50% to 75%; Crowe type III, 75% to 100%; and Crowe type IV, >100%.

Statistical Analysis

We determined the prevalence of secondary OA due to hip dysplasia among new patients with hip OA at the core hospitals. Patients were grouped according to their age at the time of the visit (15 to 29, 30 to 39, 40 to 49, 50 to 59, 60 to 69, 70 to 79, and ≥80 years old) based on their year of birth4. The number of patients in the assessment groups and the percentage of the total study population were then described for each group. A Fisher exact test was performed for categorical variables, and a Kruskal-Wallis test was performed for continuous variables compared among multiple groups. The intra- and interobserver reliability of the radiographic measurements are presented in Appendix Supplementary Table 1.

To smooth the data and to estimate the trend regarding the percentage of hips with a history of DDH treatment in childhood, we utilized a centered moving average33 (see Appendix Supplementary Table 2), a method that is commonly utilized for data smoothing, especially for time series data.

Patients with secondary hip OA due to hip dysplasia were divided into those born in or before 1972 and those born in or after 1973, and the prevalence of severe subluxation, defined as Crowe type II, III, or IV, was investigated. The prevalence of severe subluxation was compared between groups with use of a Fisher exact test. The odds ratio and 95% confidence interval were calculated. All analyses were conducted with use of R (version 4.0.5; The R Foundation) and RStudio (version 2023.06.0). The level of significance was set at p < 0.05.

Results

Overall, 1,157 patients (1,458 hips) were assessed for eligibility. A total of 1,095 patients (1,381 hips) were included in the analysis after excluding patients with a history of previous surgery on the same side (57 patients [69 hips]), patients with rheumatoid arthritis (2 patients [2 hips]), patients with an infectious disease (1 patient [1 hip]), and patients with insufficient data (5 patients [5 hips]). The mean age (and standard deviation) was 63.5 ± 14.7 years (range, 15 to 95 years), and the most common age range at the time of the initial visit was 70 to 79 years (406 hips; 29.4%), followed by 60 to 69 years (401 hips; 29.0%; Table I). Overall, 795 patients (1,019 hips; 73.8% of hips) were diagnosed with secondary OA due to hip dysplasia. However, this percentage varied among age groups (p < 0.001). When the cohort was stratified by age and year of birth, the prevalence of hip dysplasia was highest (87.4%) in the group 40 to 49 years old and the prevalences of primary OA and SIF were both highest in the group ≥80 years old. The prevalence of hip dysplasia was 71.5% (894 of 1,251 hips) among hips with OA without a history of DDH treatment in childhood and 96.2% (125 of 130 hips) among those with such a history (Table I).

TABLE I - Descriptive Epidemiology of New Patients with Hip OA Among 12 Core Hospitals in 2022* Overall Patient Age in 2022 (Patient Birth Year) P Value ≤29 Years Old (1993-2007) 30-39 Years Old (1983-1992) 40-49 Years Old (1973-1982) 50-59 Years Old (1963-1972) 60-69 Years Old (1953-1962) 70-79 Years Old (1943-1952) ≥80 Years Old (1927-1942) No. of hips 1,381 60 35 103 247 401 406 129 Age at visit†(yr) 63.5 ± 14.7 21.5 ± 4.7 35.0 ± 2.9 45.4 ± 2.9 55.2 ± 2.8 64.9 ± 2.8 73.8 ± 2.7 84.2 ± 3.3 <0.001‡ Age at onset†(yr) 57.0 ± 17.6 17.9 ± 6.1 30.8 ± 6.8 38.9 ± 10.3 47.0 ± 11.5 58.6 ± 10.0 67.6 ± 10.6 78.3 ± 11.3 <0.001‡ Sex (no. [%] of hips) 0.100  Male 236 (17.1) 17 (28.3) 5 (14.3) 12 (11.7) 34 (13.8) 70 (17.5) 72 (17.7) 26 (20.2)  Female 1,145 (82.9) 43 (71.7) 30 (85.7) 91 (88.3) 213 (86.2) 331 (82.5) 334 (82.3) 103 (79.8) BMI†(kg/m 2 ) 23.8 ± 4.3 20.3 ± 2.5 23.1 ± 4.2 23.7 ± 5.0 25.2 ± 5.4 24.2 ± 4.1 23.5 ± 3.6 22.9 ± 3.7 <0.001‡ Side (no. [%] of hips) 0.175  Right 784 (56.8) 26 (43.3) 20 (57.1) 64 (62.1) 132 (53.4) 225 (56.1) 245 (60.3) 72 (55.8)  Left 597 (43.2) 34 (56.7) 15 (42.9) 39 (37.9) 115 (46.6) 176 (43.9) 161 (39.7) 57 (44.2) Hemilateral (no. [%] of hips) 809 (58.6) 26 (43.3) 17 (48.6) 51 (49.5) 125 (50.6) 235 (58.6) 263 (64.8) 92 (71.3) <0.001‡ History of DDH treatment in childhood (no. [%] of hips) <0.001‡  Yes 130 (9.4) 6 (10.0) 3 (8.6) 16 (15.5) 40 (16.2) 36 (9.0) 28 (6.9) 1 (0.8)  No 1,251 (90.6) 54 (90.0) 32 (91.4) 87 (84.5) 207 (83.8) 365 (91.0) 378 (93.1) 128 (99.2) Subclassification (no. [%] of hips) <0.001‡  Secondary OA due to hip dysplasia 1,019 (73.8) 50 (83.3) 26 (74.3) 90 (87.4) 207 (83.8) 316 (78.8) 276 (68.0) 54 (41.9)  Primary OA 187 (13.5) 0 (0.0) 2 (5.7) 3 (2.9) 13 (5.3) 39 (9.7) 82 (20.2) 48 (37.2)  Secondary OA due to FAI 67 (4.9) 7 (11.7) 7 (20.0) 6 (5.8) 11 (4.5) 16 (4.0) 15 (3.7) 5 (3.9)  Secondary OA due to SIF 51 (3.7) 0 (0.0) 0 (0.0) 2 (1.9) 3 (1.2) 16 (4.0) 19 (4.7) 11 (8.5)  Other§ 47 (3.4) 3 (5.0) 0 (0.0) 2 (1.9) 11 (4.5) 12 (3.0) 12 (3.0) 7 (5.4)  Unclassifiable 10 (0.7) 0 (0.0) 0 (0.0) 0 (0.0) 2 (0.8) 2 (0.5) 2 (0.5) 4 (3.1)

*OA = osteoarthritis, BMI = body mass index, DDH = developmental dysplasia of the hip, FAI = femoroacetabular impingement, SIF = subchondral insufficiency fracture.

†Values are given as the mean ± standard deviation.

‡Significant, p < 0.05.

§Included rapid destructive coxopathy, trauma, Legg-Calvé-Perthes disease, slipped capital femoral epiphysis, and skeletal dysplasia.

Overall, 9.4% of hips (130 of 1,381) or 9.5% of patients (104 of 1,095) had a history of DDH treatment in childhood. According to the 20-year centered moving average, approximately 13% to 15% of hips among patients born from 1963 to 1972 (aged 50 to 59 years at the time of the survey) had a history of DDH treatment in childhood, representing a plateau; however, around the boundary birth years of 1972 and 1973, the trend changed and the percentage of hips with a history of DDH treatment decreased (Fig. 4). A similar trend was observed when the width of the centered moving average was set to 10 years (see Appendix Supplementary Figure 1).

Fig. 4:

Histogram showing the number of hips with new hip OA (gray bars) and the number of such hips with a history of DDH treatment in childhood (black bars) by birth year. A line for the 20-year centered moving average was utilized to estimate the trend in the percentage of hips with a history of DDH treatment in childhood across birth years.

In total, 795 patients (1,019 hips) with secondary OA due to hip dysplasia were included in the final analysis (Table II; see also Appendix Supplementary Table 3). The prevalence of severe subluxation (i.e., Crowe type II, III, or IV) was 2.4% (4 of 166 hips) among patients born in or after 1973, which was significantly less than the 11.1% (95 of 853 hips) among patients born in or before 1972 (odds ratio, 0.20; p < 0.001; Table III). Histograms of the number of hips by patient birth year and Crowe classification showed that there were only 4 hips with Crowe type II among patients born in or after 1973 and no cases of Crowe type III or IV (Fig. 5).

TABLE II - Descriptive Epidemiology of New Patients with Secondary Hip OA Due to Hip Dysplasia* Overall Patient Age in 2022 (Patient Birth Year) P Value ≤29 Years Old (1993-2007) 30-39 Years Old (1983-1992) 40-49 Years Old (1973-1982) 50-59 Years Old (1963-1972) 60-69 Years Old (1953-1962) 70-79 Years Old (1943-1952) ≥80 Years Old (1931-1942) No. of hips 1,019 50 26 90 207 316 276 54 Age at visit†(yr) 61.6 ± 14.3 21.5 ± 4.7 34.9 ± 3.1 45.4 ± 2.9 55.1 ± 2.8 64.8 ± 2.8 73.7 ± 2.7 83.4 ± 2.9 <0.001‡ Age at onset†(yr) 54.6 ± 16.9 18.0 ± 6.2 30.0 ± 7.6 40.1 ± 8.1 46.5 ± 11.9 57.9 ± 10.0 67.0 ± 9.8 75.5 ± 14.5 <0.001‡ Sex (no. [%] of hips) <0.001‡  Male 130 (12.8) 10 (20.0) 1 (3.8) 6 (6.7) 14 (6.8) 39 (12.3) 46 (16.7) 14 (25.9)  Female 889 (87.2) 40 (80.0) 25 (96.2) 84 (93.3) 193 (93.2) 277 (87.7) 230 (83.3) 40 (74.1) BMI†(kg/m 2 ) 23.9 ± 4.4 20.4 ± 2.6 23.6 ± 4.5 24.0 ± 5.2 25.1 ± 5.3 24.3 ± 4.3 23.4 ± 3.5 22.9 ± 3.8 <0.001‡ Side (no. [%] of hips) 0.356  Right 579 (56.8) 22 (44.0) 14 (53.8) 56 (62.2) 113 (54.6) 176 (55.7) 167 (60.5) 31 (57.4)  Left 440 (43.2) 28 (56.0) 12 (46.2) 34 (37.8) 94 (45.4) 140 (44.3) 109 (39.5) 23 (42.6) Hemilateral (no. [%] of hips) 552 (54.2) 22 (44.0) 11 (42.3) 43 (47.8) 97 (46.9) 167 (52.8) 172 (62.3) 40 (74.1) <0.001‡ History of DDH treatment in childhood (no. [%] of hips) <0.004‡  Yes 125 (12.3) 6 (12.0) 2 (7.7) 16 (17.8) 39 (18.8) 35 (11.1) 26 (9.4) 1 (1.9)  No 894 (87.7) 44 (88.0) 24 (92.3) 74 (82.2) 168 (81.2) 281 (88.9) 250 (90.6) 53 (98.1) KL grade (no. [%] of hips) <0.001‡  Grade 1 134 (13.2) 44 (88.0) 15 (57.7) 32 (35.6) 19 (9.2) 13 (4.1) 11 (4.0) 0 (0.0)  Grade 2 111 (10.9) 5 (10.0) 4 (15.4) 30 (33.3) 38 (18.4) 21 (6.6) 11 (4.0) 2 (3.7)  Grade 3 206 (20.2) 1 (2.0) 4 (15.4) 13 (14.4) 44 (21.3) 83 (26.3) 52 (18.8) 9 (16.7)  Grade 4 568 (55.7) 0 (0.0) 3 (11.5) 15 (16.7) 106 (51.2) 199 (63.0) 202 (73.2) 43 (79.6) LCEA†(deg) 16.1 ± 11.8 14.4 ± 6.8 12.2 ± 8.0 11.8 ± 14.0 16.1 ± 13.6 16.0 ± 11.8 17.9 ± 10.8 18.8 ± 7.9 <0.001‡ ARO†(deg) 20.4 ± 8.4 16.2 ± 5.5 19.0 ± 8.6 20.2 ± 9.9 21.2 ± 8.3 21.4 ± 8.6