Childhood hip disease remains medically mysterious despite genetic clues

<p>Perthes’ disease was first described by Arthur Legg, Jacques Calvé and Georg Perthes in 1910. It is defined as osteonecrosis of the femoral head in children. The necrosis leads to a deformation of the femoral head. Symptoms debut between 2-15 years of age, most often between 4-8 years of age. Symptoms most commonly include pain in the hip or groin and limping. The necrosis is self-healing to varying degrees, but causes elevated risk for hip osteoarthritis in adult age. Diagnosis is today only possible once skeletal changes are already present, and there are currently no methods of screening or prevention. There is also no definitive gold standard treatment.</p><p>The cause of Perthes’ disease is not known, but there are several theories around the etiology. Environmental factors such as mechanical load, structural deviations, sex and geographic location have been suggested as risk factors. Previous smaller studies have shown that the incidence of Perthes’ disease is higher at more northerly latitudes, and possibly in a Caucasian population. Boys are 3-4 times more often affected than girls. The presence of various genetic mutations, mostly in COL2A1, has been found in family clusters and case studies. Newer studies have explored a possible connection to other genetic factors such as IL-6 and eNOS. All agree that Perthes’ disease seems to have a multifactorial etiology as we understand it today. Despite all these smaller studies there is a lack of large or nationwide studies within this area. Sweden is one of the few countries that have a long history of registers with a coverage of almost 100%. The primary aim of this dissertation was to study etiological factors for Perthes’ disease in a large nationwide study population. The second aim was to explore possible genetic differences in a Nordic population compared to existing knowledge.</p><p>Utilizing data linked from several Swedish national registers on children born 1973-1993, the incidence of Perthes’ disease in Sweden was calculated, and the association with socioeconomic factors was studied (Paper I). The correlation to risk factors during pregnancy and childbirth (paper II) and comorbidity with collagen synthesis disorders (Paper III) were studied, also based on data from Swe-dish national registers. DNA analyses on 10 ethnically Swedish patients with Perthes’ disease were conducted (Paper IV).</p><p>The incidence of Perthes’ disease in Sweden was 9.3/100 000. There was an association with a lower educational level but not to the level of income. The risk was also higher in children with both parents of Nordic ethnicity (Paper I). Children born in breech position vaginally or by emergency caesarean section had a 25% higher risk for developing Perthes’ disease. There was a link between the development of Perthes’ disease and low birth weight, small size for gestational age and prematurity. These factors are likely related, as babies born preterm are also often small (Paper 2). No correlation to the included collagen synthesis dis-orders was found, this lack of overlap is interesting (Paper 3). In DNA analysis, no COL2A1 mutations were found. We did find differences in the presentation of eNOS G894T and IL-6 -597 G&gt;A compared to the control material (Paper IV).</p><p>This dissertation increases the knowledge on several of the previously suggested risk factors for Perthes’ disease. The connection to a lower level of education most likely indicates the presence of confounding factors such as differences in health-related behaviour and lifestyle choices. Our findings on risk factors during pregnancy and childbirth are in line with previous small scale studies, which strengthens the theory that these factors in some ways contribute to the debut of Perthes’ disease. No correlation between Perthes’ disease and the included collagen synthesis disorders was found. In regards to the DNA analysis, it is interesting to note that COL2A1 mutations most commonly found in previous studies were not present at all in our group. This could indicate different genetic components depending on ethnicity. The material of the genetic study is however too small to draw any conclusions, and the study should be regarded as a pilot for larger studies.</p>