**2. Aetiology/risk factors for cerebral palsy**

#### **2.1 Aetiology or risk factors?**

The most recent definition of cerebral palsy stipulates that the disorders collectively termed cerebral palsy are attributed to non-progressive disturbances that occurred in the developing foetal or infant brain [6]. The non-progressive disturbances refer to events or processes that within a limited period/duration (static in nature) permanently damage the brain (specifically motor development) or influence the expected patterns of brain maturation [1, 6]. Many different processes or events can result in this damage and so CP is understandably said to be aetiologically heterogeneous (due to multiple aetiologies) and compatible with many aetiologic diagnoses [1, 4]. Indeed, CP on its own is not an aetiologic diagnosis [1]. Furthermore, CP is pathologically heterogeneous since different aetiological factors acting at different stages of development result in different neuropathological substrates even if they lead to the same clinical subtype [4]. For instance, both severe neonatal encephalopathy from perinatal asphyxia and kernicterus from severe unconjugated hyperbilirubinaemia are different aetiological factors for dyskinetic (choreoathetoid/dystonic) CP but they damage different parts of the brain (the ventrolateral nucleus of the thalamus and putamen versus globus pallidus and subthalamic nucleus, respectively) [7, 8].

Direct causal links or relationships are difficult to establish with certainty and explains why the phrase "attributed to" is used in the most recent definition with respect to the causes of CP instead of "are due to" or "caused by" [1, 6]. Therefore, risk factors (correlational) that increase the probability or chances of occurrence of CP are epidemiologically more appropriate than causes. The significance of identification/knowledge of actual causes/risk factors of CP is its implication for prevention. That is, it will help devise prevention strategies for CP.

### **2.2 Identification of the actual causes of CP: the challenge**

The accurate identification of the specific disturbance or specific timing of the event or process that damaged the developing motor system has remained difficult since most (about 75%) of the events occur in the antenatal (prenatal) period [1, 6]. There is also a challenge in designing prospective studies to identify risk factors across populations since only 2–3 per 1000 pregnancies will result in a child with CP [3]. However, neuroimaging has made significant contribution to the understanding of the aetiology and pathology of CP and timing of insults [4 5]. Neuroimaging can be used to identify the neuropathological substrates of the various aetiologic or risk factors of CP, possibly provide information about timing of insults and detect cerebral dysgenesis [2, 4]. Nevertheless, there exists some limitations presently as neuropathologic—aetiological correlations are not yet fully clear and comprehensive [4].

### **2.3 Causal pathways/mechanisms**

Certainly a large number of risk factors are associated with CP and have been identified in numerous earlier studies [9–15] in different parts of the world. These studies have shown that in many cases it is not a single factor but a combination of risk factors (multifactorial) or a cascade of events or disturbances that result in CP [3, 16, 17]. This gave birth to the concept of causal pathways or mechanisms in CP causation and in disorders without a single definitive cause.

According to Stanley et al. [18], a causal pathway refers to a sequence of interdependent events that culminate in disease. This implies that one risk factor leads to another, to another and so on ultimately resulting to the disorder or disease. For instance, prematurity is an important risk factor for CP and can lead to periventricular leukomalacia (white matter injury), poor lung development or respiratory distress at birth, birth asphyxia, increased risk of chronic bilirubin encephalopathy (kernicterus) and so on, ultimately resulting in CP. Nevertheless, prematurity alone is not a sufficient cause for CP since not all children born prematurely have CP. Another example is breech presentation at birth leads to increased risk of cranial trauma or injury and increased risk of CP in places where breech delivery per vaginam is rife.

From the foregoing, it is obvious that the risk factors in a pathway are interconnected and in most cases additively increase the risk of CP. This explains the "two-hit" and "multi-hit" models that consider accumulation of risk factors/insults and a synergistic increase in risk in causation of CP [16, 19]. That is, the brain of a neonate with in-utero exposure to placental inflammation (chorioamnionitis/funisitis/chronic vasculitis) or who had foetal growth restriction ("first-hit") is more vulnerable or conditioned to another injury like sepsis in the neonatal period ("second-hit") ("twohit" model) while exposure to three or more adverse events/risk factors underlies the "multiple-hit model" of CP causation [16, 19]. Another example of interaction of risk factors is the increased cumulative risk of CP in a child with co-occurrence of earlyonset pre-eclampsia, foetal infection/foetal inflammatory response syndrome (FIRS), perinatal asphyxia and neonatal sepsis [19].

One identifiable challenge is that there are so many possible and complex pathways since the risk factors are numerous and each pregnancy presents new possibilities [19]. Some examples of known CP causal pathways are shown in **Figures 1**–**4** below. The significance of identifying and understanding all these complex causal pathways is in formulating preventive strategies as earlier mentioned. Thus, more research is needed to elucidate the combined effect and specific sequence of multiple risk factors on the occurrence of CP.
