1.1 Inclusion criteria

This review included peer-reviewed observational, cohort, case–control, and population-based registry studies reporting primary data on the incidence or prevalence of Perthes disease in paediatric populations. Eligible studies addressed temporal trends, diagnostic practices, or environmental and sociodemographic factors influencing incidence. Only English-language articles published within the past 10 years were considered to ensure contemporary relevance and methodological quality.

1.2 Exclusion criteria

Studies were excluded if they lacked primary incidence data, including case reports, small case series, narrative reviews, editorials, and commentaries. Research focused solely on treatment outcomes, adult populations, non-human subjects, or lacking clear diagnostic criteria was also excluded. Additionally, non-English publications, conference abstracts, and unpublished dissertations without sufficient methodological detail were omitted.

3.1 Epidemiology and trends in incidence

Several studies have reported a marked decline in the incidence of LCPD over recent decades. Mullan et al., 2017 conducted a retrospective cohort study comparing incidence trends between 1992–1998 and 2007–2013, reporting a 61 % decrease in LCPD cases over the 15-year period.² The decline was attributed to improvements in public health factors, including increased breastfeeding rates, higher vaccination coverage, reduced lead emissions, and decreased smoking rates among women.² Similarly, Laine et al., 2021, in a prospective multicentre cohort study, observed an 18.9 % reduction in LCPD incidence, which was strongly associated with a decline in smoking prevalence.³ Supporting this trend, Zverev et al., 2024, highlighted in a review study that reduced paediatric exposure to cigarette smoke in the UK corresponded with a decline in LCPD cases.⁴ Furthermore, Beni et al., 2024, through a scoping review, reported a decrease in LCPD incidence from 12.2 per 100,000 in 1990 to 2.48 per 100,000 in 2022, attributing this change largely to improved living conditions and overall socioeconomic development.⁵

3.2 Aetiology

1Genetic and biochemical markers

Rodríguez-Olivas et al., 2022 suggested a possible autosomal dominant inheritance pattern in the development of LCPD.¹ This genetic link was not supported by Johansson et al., 2017, whose nationwide cohort study found no evidence to substantiate such an association.⁶

Several studies have implicated endothelial nitric oxide synthase (eNOS) polymorphisms in the pathogenesis of LCPD.^1,7,8 Elevated interleukin-6 (IL-6) levels were also identified as a contributing factor.^1,9,10 Spasovski et al., 2023 specifically noted that elevated IL-6 may lead to bone resorption and ischemic osteonecrosis, a phenomenon observed in the Indian paediatric cohort examined in their study.⁸

Another genetic factor frequently cited is the COL2A1 polymorphism.^8,10–12 However, this association was contested by Lindblad 2024, who found insufficient evidence to support a correlation with LCPD.²⁷ The Factor V Leiden mutation has also been linked to LCPD, with supporting evidence from Hailer & Hailer 2018, Pavone et al., 2019, Wilkinson & Zeggini 2021, and Basit & Khoshhal 2022.^7,10,12,13 Additionally, decreased levels of protein C and S were associated with the condition in studies by Hailer & Hailer 2018 and Basit & Khoshhal 2022.^7,12

Asadollahi et al., 2021 highlighted ethnic variability in genetic markers associated with LCPD.¹¹ Factor V Leiden mutations were observed in German, Jewish, and Brazilian populations, TRPS1 and eNOS polymorphisms in Germans, elevated IL-6 in Iranians and Serbians, and COL2A1 polymorphisms in Japanese and Chinese populations.¹¹ Conversely, Metcalfe et al., 2016, in a twin study, concluded that genetic predisposition may not be a primary driver of LCPD, pointing instead to intrauterine and early childhood environmental exposures as likely contributors.¹⁴

Berthaume et al., 2016 provided biomechanical insight, suggesting that structural features such as skeletal immaturity, rostral sparing, and age-specific vascular patterns in children aged 5–8 years increase susceptibility of the femoral head to ischemia and collapse.¹⁵ This may explain the higher incidence of LCPD in young boys and is further compounded by risk factors such as short stature, low birth weight, and lower socioeconomic status.¹⁵2Birth and perinatal factors

Multiple studies have identified maternal smoking as a contributing factor in the development and progression of LCPD.^8,12,16–18 Passive smoking has been particularly associated with fibrinogen gene polymorphism, which may alter fibrinogen levels, leading to hypofibrinolysis and an elevated risk of thrombotic venous occlusion, ultimately impairing blood flow to the femoral head. This mechanism was reported by Gao et al., 2020 and Joseph et al. 2023.^19,20

In addition to smoking, Lindblad et al., 2020 identified breech birth and maternal diabetes as potential risk factors for LCPD.²⁰ Low birth weight has also been consistently reported as a contributing factor, as evidenced by studies conducted by Pavone et al., 2019, Joseph et al., 2023, and Lindblad et al. 2020.^10,17,20

Furthermore, Perry et al., 2017 conducted a case-control study using questionnaires to evaluate maternal smoking habits during pregnancy.²⁶ Their findings revealed a significantly increased odds ratio for the development of LCPD in children with in utero tobacco exposure.²⁶3Obesity and metabolic factors

Metabolic factors, particularly obesity, have been consistently associated with the development of LCPD across several studies. Hailer & Hailer 2018, Laine et al., 2021, and Basit & Khoshhal 2022, all reported a significant correlation between obesity and LCPD.^3,7,12

Neal et al., 2016 further demonstrated that the prevalence of LCPD was twice as high in obese populations compared to the general paediatric population.²¹ Beckish et al., 2024 suggested that obesity may contribute to LCPD through metabolic and vascular disturbances, particularly by impairing blood flow to the hip joint, thereby playing a key role in the disease's pathogenesis.⁹

Pavone et al., 2019 and Nowicki et al., 2019 highlighted the role of elevated leptin levels in obesity, which may exacerbate LCPD by inhibiting osteoblast activity and reducing cortical bone thickness.^10,22 Notably, Nowicki et al., 2019 reported an odds ratio of 3.4, indicating that obese children are over three times more likely to develop LCPD than those of normal weight.²²

In addition, both Basit & Khoshhal 2022 and Mörlin & Hailer 2021 identified elevated blood pressure as a common feature in paediatric LCPD patients, suggesting a potential vascular link between hypertension and disease development.^12,23 Tisano et al., 2022 also found a higher prevalence of obesity among children diagnosed with LCPD.²⁴ Their study noted that obesity-related LCPD cases often presented later and were less likely to be managed surgically, potentially due to delayed diagnosis or increased treatment complexity.²⁴4Environment and socioeconomic status

Rodríguez-Olivas et al., 2022 reported a higher incidence of LCPD among Caucasian populations, with notably lower rates observed in Asian and Black populations.¹ Low socioeconomic status was also identified as a significant risk factor, attributed to nutritional deficiencies.¹ However, Neal et al., 2016 challenged this explanation, suggesting instead that the increased incidence in lower socioeconomic groups may result from excessive caloric intake leading to obesity and subsequent disruption of bone growth.²¹

Several studies support the association between socioeconomic deprivation and LCPD development.^{8,10,17,18,20} Notably, Joseph et al., 2023 and Beni et al., 2024 both reported a fourfold increase in risk for children from deprived backgrounds.^5,20

Johansson et al., 2017 and Wilkinson & Zeggini 2021 linked low parental educational attainment with increased LCPD incidence, particularly among individuals of Nordic ancestry.^6,13

Geographic and environmental factors have also been implicated. Mörlin & Hailer 2021 identified residence at higher latitudes as a contributing factor, while Beni et al., 2024 quantified a 2.35-fold increase in incidence for every 10-degree increase in latitude.^5,235Mechanical and lifestyle factors

Rodríguez-Olivas et al., 2022, Pavone et al., 2019, and Karski et al., 2021 identified participation in high-impact sports involving mechanical overload, such as gymnastics, as a potential risk factor for the development of LCPD, attributing this to repetitive microtrauma and overuse of the hip joint.^1,10,25

Similarly, Rodríguez-Olivas et al., 2022, Pavone et al., 2019, and Berman et al., 2016 associated hyperactivity, particularly in children with attention-deficit/hyperactivity disorder (ADHD), with an increased risk of LCPD.^1,10,28 This was attributed to the repetitive joint loading and excessive activity levels common among hyperactive children, potentially contributing to mechanical stress on the developing femoral head.^1,10,286.Medication induced

Zverev et al., 2024 reported a potential association between growth hormone (GH) therapy and the development of LCPD.⁴ In their review of three patients undergoing GH treatment, two developed LCPD shortly after initiating therapy.⁴ The third patient had already been diagnosed with LCPD prior to starting GH.⁴ Interestingly, after one year of GH therapy, the progression of LCPD in this patient appeared to have stabilised, suggesting a possible modifying effect of GH on disease progression.⁴

Additionally, Chen et al., 2021 described a case report of a male patient who developed LCPD following prolonged use of steroids and cyclosporin.²⁶ While cyclosporin was excluded as a causative factor due to insufficient evidence linking it to LCPD, the role of long-term steroid use remained uncertain and could not be conclusively ruled out.²⁶

4.1 Genetic susceptibility: Persistent but unchanging

As mentioned, several studies have investigated genetic markers associated with LCPD, including COL2A1 and Factor V Leiden polymorphisms.^8,10,12 These genetic risk factors suggest a heritable predisposition to disrupted bone development or coagulation abnormalities that could impair femoral head perfusion. Asadollahi et al., 2021 and Hailer & Hailer 2018 argue that while these markers are frequently identified, their prevalence has not changed over time, indicating that they cannot explain the downward epidemiological trend.^7,11 This is further supported by twin studies showing low concordance for LCPD, reinforcing the idea that genetic risk requires environmental triggers for disease manifestation.¹⁴

4.2 Environmental and perinatal improvements: A Plausible driver

The most compelling argument in favour of a declining incidence relates to improvements in perinatal care and reductions in environmental exposures. Studies by Laine et al., 2021 and Zverev et al., 2024 demonstrate strong associations between declining maternal and passive smoking rates and reduced LCPD incidence.^3,4 Tobacco exposure during pregnancy is believed to impair vascular development in the foetus, potentially compromising femoral head blood supply.^18,26 With public health initiatives targeting smoking cessation and broader prenatal education, these modifiable exposures may be decreasing across many populations.

Mullan et al., 2017 further highlighted the potential role of increased breastfeeding rates and health awareness, which may contribute to improved early childhood bone and vascular development.² Similarly, enhanced vaccination uptake among pregnant women, while not directly linked to LCPD, reflects a general trend of improved maternal health and access to preventive care. These observations collectively support the hypothesis that LCPD's downward trend may be, in part, a consequence of better prenatal and perinatal environments.

4.3 Socioeconomic status: A double-edged influence

Socioeconomic factors are consistently linked to LCPD incidence, with low-income populations showing increased disease rates due to poor nutrition, limited healthcare access, and delayed diagnosis.^5,10,17,20 Joseph et al., 2023 and Beni et al., 2024 found up to a fourfold increased risk in socioeconomically disadvantaged groups.^5,19 Improvements in living standards, particularly in urban centres of high-income countries, may therefore explain part of the observed decline.

However, counterarguments exist. Neal et al., 2016 proposed that low socioeconomic status can also drive high-calorie, low-nutrient diets, leading to childhood obesity, another known risk factor for LCPD.²⁰ This introduces a paradox: if obesity has become more prevalent in recent years, why has LCPD incidence declined? While it is possible that improvements in other determinants e.g. early detection, reduced smoking, better nutrition outweigh the rising obesity burden, the discrepancy underscores the complexity of the disease's aetiology.

4.4 Metabolic and mechanical factors: Contradictory evidence

Obesity is frequently cited as a contributor to LCPD, due to increased mechanical load on the femoral head and metabolic changes that affect bone health.^9,10,22,23 Nowicki et al., 2019 reported that obese children were 3.4 times more likely to develop LCPD than those of normal weight.²²

Yet, the continued rise in childhood obesity globally conflicts with the declining incidence of LCPD. One explanation could be earlier diagnosis and lifestyle interventions, or perhaps that obesity alone is insufficient to trigger disease in the absence of other coexisting risk factors.

Alternatively, as proposed by Mörlin & Hailer 2021, high blood pressure in obese children may be a secondary rather than primary driver, making it harder to isolate the impact of obesity alone.²²

4.5 Less explored contributors: ADHD, Physical activity, and medication

Emerging literature has explored other potential risk factors. Rodríguez-Olivas et al., 2022 and Berman et al., 2016 proposed a link between hyperactivity, particularly in children with ADHD and LCPD, based on the idea that repetitive mechanical overload from hyperactive behaviour may damage the femoral head.^1,28

Likewise, high-impact sports like gymnastics have been associated with increased LCPD risk due to repetitive trauma.^10,24

Medications such as corticosteroids and growth hormone therapy have also been implicated. Zverev et al., 2024 described a potential association between GH therapy and LCPD onset in some patients,⁴ while Chen et al., 2021 documented a case of LCPD following chronic steroid use.²⁵

However, these are isolated findings and currently lack sufficient epidemiological support to influence conclusions about incidence trends.

4.6 Limitations

This review is subject to several limitations. First, heterogeneity across study designs, populations, and diagnostic criteria may limit the comparability of incidence data and affect the reliability of pooled findings. Restricting the analysis to English-language, peer-reviewed publications introduces a risk of selection bias, potentially excluding relevant non-English or grey literature sources.

Many of the included studies relied on retrospective data, which can introduce recall and reporting bias, and limit control over confounding variables. Additionally, inconsistencies in how environmental and sociodemographic factors were reported make it difficult to synthesise findings and draw definitive conclusions about causality.

Lastly, the 10-year publication window may not fully capture long-term trends or the effects of recent changes in diagnostic criteria or public health interventions, potentially underestimating or overestimating shifts in the incidence of Perthes disease.