The Challenge of Identifying Causal Pathways Leading to Cerebral Palsy.

Abstract

Cerebral palsy (CP) is a heterogeneous disorder relating to permanent motor disabilities, neuroimaging findings, subtypes, associated impairments, and etiologies. Although the exact etiology of CP is often impossible to establish, several antenatal, perinatal, and postnatal risk factors have been identified. The birth prevalence of CP has been stable at ∼2 per 1000 for several decades.1 However, recent studies reveal a decreasing trend in prevalence in some high-income countries.2–4 This decrease has not been ascribed to new specific preventive interventions but to general improvements in antenatal, obstetric, and neonatal care.In this issue of Pediatrics, Strøm et al5 report that offspring of mothers with a wide range of chronic maternal conditions have an increased risk of developing CP. With a robust data set, including >1 million live-born children recorded in Norwegian health registries, they found that several autoimmune conditions had twofold to threefold increased adjusted relative risks (aRRs) of developing CP. The highest aRRs were observed for diabetes type 2 (aRR: 3.2), lupus erythematosus (aRR: 2.7), and diabetes type 1 (aRR: 2.2) as compared with mothers without these conditions.The study reveals both strengths and limitations of large population-based registry studies. A major strength of such studies is that they have the potential to identify associations between rare exposures and rare outcomes and generate hypotheses for further etiologic and treatment studies. In contrast, researchers using large population-based registries have limited access to detailed information necessary to study mechanisms in more detail.In many studies, researchers have proposed that a causal pathway or an effect of multiple overlapping risk factors may lead to CP. This includes antenatal disorders and complications (70%–80% of children with CP) and perinatal and postnatal events.6–8 As the number of risk factors increases, the risk of developing CP increases.8,9 Strøm et al5 found the aRR of CP to be 3.4 in pregnancies in which the mothers had ≥2 chronic conditions as opposed to mothers with no chronic condition. Yet, as the authors state, the absolute risk of CP associated with ≥1 chronic maternal condition is low. Among 1000 pregnant women with any chronic and/or autoimmune disorder, >990 will deliver an infant who will not be diagnosed with CP. This low absolute risk is important to keep in mind when discussing with women who have a chronic and/or autoimmune disorder and are planning to become or already are pregnant and their partner.The authors underscore that the observed associations may not reflect a causal relationship and that they lack information on disease activity, duration, and treatment. Thus, the study cannot give clues to any specific preventive treatment. However, if these disorders are part of a causal pathway, optimal treatment might reduce the risk of CP. It is well documented that optimal control of diabetes types 1 and 2 at conception or during pregnancy reduces the risk of congenital malformations and complications at delivery.10–12 Both congenital malformations and complications at delivery are associated with CP.13,14 It is reasonable to speculate that optimal disease control of other autoimmune disorders could also prevent a few cases. Even so, a potential reduction in children with CP due to optimized care of these rare disorders would hardly affect the birth prevalence of CP. Still, as the Royal College of Obstetricians and Gynaecologists in the United Kingdom states, “each baby counts.”15The findings of this study may point to biological pathways, as discussed by the authors. The increased risk associated with autoimmune disorders could initiate further research into inflammatory mechanisms and the role of the placenta, the latter being probably the most important, but at the same time the least explored, organ in the etiology of CP. However, the path from the findings in the study by Strøm et al5 to “biological pathways for which interventions could be possible” is likely to be long.