The purpose of this study was to determine the annual incidence of Perthes’ disease in Korea and compare this with other populations. A survey identified all newly diagnosed children with Perthes’ disease aged 14 years or younger in South Honam, Korea, between January 1999 and December 2001. A total of 84 children were included: 29 in 1999, 28 in 2000 and 27 in 2001. The mean annual incidence was 3.8 per 100 000. This is similar to that reported in other Asian countries, but higher than in black populations and lower than in Caucasians.
The reported incidence of Perthes’ disease ranges from 0.2 per 100 000 children in Hong Kong to 29.4 per 100 000 in the Faroe Islands.1,2 Barker and Hall3 reported that the incidence of Perthes’ disease varies geographically, both between and within countries. However, these differences among countries, regions and races have not been clarified. To investigate these issues, we undertook an epidemiological study of Perthes’ disease in South Honam, South Korea.
Materials and Methods
We conducted a survey of patients aged 14 years or under who were diagnosed with Perthes’ disease in South Honam between January 1, 1999 and December 31, 2001.
In order to calculate the annual incidence during the study period, total population figures and numbers of children aged 14 years or under in South Honam were obtained from the Korean National Statistical Office. The population of South Honam was 3 516 253 in 1999, 3 513 067 in 2000 and 3 497 904 in 2001, of which there were 734 645 children aged 14 years or under in 1999, 728 901 in 2000, and 719 504 in 2001.
We used two approaches to identify all newly-detected children with Perthes’ disease during this period. First, we approached the Health Insurance Review Agency (HIRA) for the names of patients registered between 1999 and 2002 as having Perthes’ or similar conditions such as transient synovitis, spondyloepiphyseal dysplasia, multiple epiphyseal dysplasia, hypoparathyroidism, Meyer’s dysplasia, and post-traumatic or post-infectious avascular necrosis. We excluded patients aged 15 years or older and those registered and diagnosed in 2002. We included three children whose condition was diagnosed in 2002 but occurred in 2001.
Secondly, as HIRA lists did not include all children with Perthes’ disease, we visited all potential hospitals to which patients suspected of Perthes’ disease might have been referred. A total of 125 were visited and we reviewed appropriate medical records and radiographs. In order to confirm the diagnosis, all the records and radiographs were reviewed by two paediatric orthopaedic surgeons (SMR, STJ) and two senior orthopaedic residents (BHB, KDK).
A total of 224 children were identified from both sources. Of these, we excluded 140 children; 51 because they developed the condition before 1999 and 89 because they had other diagnoses (67 with transient synovitis, eight with avascular necrosis secondary to trauma and infection, seven with Meyer’s dysplasia, three with multiple epiphyseal dysplasia, two with spondyloepiphyseal dysplasia and two with hypoparathyroidism).
A total of 84 children (89 hips) aged 14 years or younger were newly diagnosed with Perthes’ disease; 29 in 1999, 28 in 2000 and 27 in 2001 giving a mean annual incidence of 3.8 per 100 000 (Table I⇓).
Of the 84 children, 69 were boys and 15 were girls. The incidence was higher in boys (6.1 per 100 000) compared with girls (1.5 per 100 000). It was bilateral in five children. It involved 46 right (52%) and 43 left hips (48%). The mean age at onset was six years (2 years 5 months to 13 years 2 months) and the onset was earlier in boys (mean of 5 years 11 months) than girls (mean of 6 years 7 months). The mean delay in bone age versus chronological age in children with Perthes’ disease was 10.4 months in boys and 4.6 months in girls.4
With regard to regional variation within South Honam, we found the incidence of Perthes’ disease was lower in Gwanju Metropolitan City (3.2 per 100 000) than rural Chonnam Province (4.3 per 100 000).
We identified 15 papers which detailed the incidence of Perthes’ disease (Table II⇓).1,2,5–17 By combining our results with these findings, we were able to compare incidences among countries, races and regions within countries.
Three races are described in the literature; Asian, African and Caucasian. The incidence in Asians is reported from four countries (including our study) as 0.2 per 100 000 in Hong Kong,2 0.4 to 4.4 in India,5 1.73 in South Africa (Asians only)6 and 3.8 in Korea. The incidence in the black population has been reported from two countries, 0.45 per 100 000 in South Africa6 and 2.0 per 100 000 in Nigeria.7 The incidence in Caucasians has been reported from five countries, 4.4 to 16.9 per 100 000 in England,3,8–13 5.1 in Canada,14 5.7 to 10 in the United States,15,16 8.5 in Sweden,17 8.5 to 29.4 in the Faroe Islands1,17 and 10.8 in South Africa (Caucasians only).6 Thus the incidence of Perthes’ disease is highest in Caucasian, lowest in black populations with Asians in between.
Also, regional differences were noted within countries.5,8,9,13,18 we found that Perthes’ disease was less common in a city (Gwangju Metropolitan City; 3.2 per 100 000) than in a rural area of South Honam (Chonnam Province; 4.3 per 100 000). Joseph et al5 similarly reported a higher incidence in rural areas of southwest India. However, these results contrast with studies from England. Hall et al10 and Hall and Barker19 suggested a higher incidence in urban areas and lower social classes, and Barker et al,8 in a study of three regions in England also found the disease more common in conurbations.
The reasons for differences among countries, races, and regions are not clearly understood. However, several theories are presented. First, differences may be due to anatomical differences in the arterial supply to the femoral head. Chung20 reported that there are fewer ascending arteries in the femoral necks of white children than in black children and Purry6 also suggested that this may explain different incidences to Perthes’ disease.
Secondly, socio-economic factors may play a role. Wynne-Davies and Gormley21 reported that a child about to develop Perthes’ disease is already constitutionally and socially disadvantaged. They added that during the perinatal period and early years of life Perthes’ disease-prone children are probably more susceptible to trauma than normal children. Margetts et al13 suggested that environmental influences may come into play several years before a child presents with hip pain. Some authors8,19,22 also indicated the risk of Perthes’ disease increases as socio-economic levels decrease and when environmental conditions are less favourable.
The third reason concerns differences in skeletal maturation. Barker and Hall3 suggested that lower incidences of Perthes’ disease among populations of African and Chinese descent may reflect ethnic differences in skeletal maturation. Thompson et al2 showed evidence that normal children in Southern China have a different pattern of bone ageing than European and American children. Caffey23 suggested that the black race is resistant to Perthes’ disease because of their generally advanced skeletal maturation at birth and greater cortical bony thickness than Caucasians. Also, skeletal maturation in European and American children with Perthes’ disease is delayed by about two years in both boys and girls.24,25 However, we found that delay of maturation in Korean children with Perthes’ disease is only 10.4 months in boys and 4.6 months in girls as determined by the Greulich Pyle method.4
The fourth reason for these differences involves genetic factors. Hall and Barker9 indicated that the large geographic differences observed in Perthes’ disease cannot be explained by socio-economic differences between urban and rural areas. The high incidence of Perthes’ disease on the Faroe Islands can be attributed to a high level of intermarriage within a relatively small population.1,17 Ebong7 also suggested that genetic factors might account for low incidences in black populations.
No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
- Received June 6, 2005.
- Accepted June 30, 2005.
- © 2005 British Editorial Society of Bone and Joint Surgery