The Complex Aetiology of Cerebral Palsy

Reviews
The complex aetiology of

cerebral palsy
Steven J. Korzeniewski1*, Jaime Slaughter2, Madeleine Lenski3, Peterson Haak4 and
Nigel Paneth3,5
Abstract | Cerebral palsy (CP) is the most prevalent, severe and costly motor disability of
childhood. Consequently, CP is a public health priority for prevention, but its aetiology has
proved complex. In this Review, we summarize the evidence for a decline in the birth prevalence
of CP in some high-income nations, describe the epidemiological evidence for risk factors, such
as preterm delivery and fetal growth restriction, genetics, pregnancy infection and other
exposures, and discuss the success achieved so far in prevention through the use of magnesium
sulfate in preterm labour and therapeutic hypothermia for birth-asphyxiated infants. We also
consider the complexities of disentangling prenatal and perinatal influences, and of establishing
subtypes of the disorder, with a view to accelerating the translation of evidence into the
development of strategies for the prevention of CP.
Cerebral palsy (CP) is the most common severe motor the prevalence was 1.8 per 1,000 live births10,11. These
disability in children, and its severity is demonstrated differences in prevalence indicated that defining
by the fact that 40% of children with the condition can- CP, differentiating CP from other motor disabilities
not walk independently1,2, one-third have epilepsy3, up and determining the precise lower limits of severity
to one-third are non-verbal4,5 and about one-half have that delineate cases were all problematic. This difficulty
some degree of cognitive impairment2,6–8. Lifetime costs has been repeatedly wrestled with, and numerous inter-
for a child with CP in the USA have been estimated national meetings on the definition and classification
at just under US$1 million per individual for health of CP have been held since 1958 (refs12–14), each leading
1
Department of Obstetrics &
Gynecology, Wayne State
expenditures, educational needs, social services and to changes in the d efinition of the condition (Fig. 1).
University School of Medicine, lost economic opportunity9. The prevalence, severity Ongoing population surveillance for CP began in the
Detroit, MI, USA. and burden of CP make it a public health priority for Nordic countries15–17. In the second half of the past cen-
2
Department of Health prevention, and recognition that perinatal exposures and tury, most reviews concluded that the prevalence of CP
Systems and Sciences pregnancy complications are strongly linked to the risk (generally expressed in relation to numbers of live births)
Research and Department of
of CP provides opportunities for prevention. However, in industrialized nations was fairly stable at 1.5–2.5 cases
Epidemiology and
Biostatistics, Drexel
the aetiology of CP has proved complex, making p rogress per 1,000 live births18,19, but with a modest increase in
University, Philadelphia, in its prevention difficult. the last two decades of the 20th century owing largely
PA, USA. In this Review, we consider the epidemiological to the greatly increased survival of very premature infants
3
Department of Epidemiology observations that provide evidence for the contribution as a result of the success of the new technology19,20.
and Biostatistics, College of of various developmental pathways to the pathogenesis of Estimates of CP prevalence in the 21st century, how-
Human Medicine, Michigan
CP and for the substantial success in prevention to date. ever, reveal a mixed picture. During the first decade of
State University, East
Lansing, MI, USA.
We also consider the complexities of disentangling the century, estimates of CP prevalence were generally
4
Michigan Department of
prenatal and perinatal influences, with a view to accelera higher than in the 20th century in high-income coun-
Health and Human Services, ting the translation of evidence into clinical approaches tries (HICs). In the USA, prevalence estimates increased
Lansing, MI, USA. to the prevention of CP. from 2 to as high as 3 cases per 1,000 live births between
5
Department of Pediatrics 2002 and 2012 (refs2,21–25), although the most recent of
and Human Development, The prevalence of cerebral palsy these surveys showed a slight decline to 2.9 per 1,000
College of Human Medicine, In the 1940s, in the USA, two voluntary organizations, 8-year-old children in 2010 from 3.5 in the same sur-
Michigan State University,
East Lansing, MI, USA.
the National Society for Crippled Children (later named veillance area in 2006 (ref. 26) . However, studies in
Easter Seals) and United Cerebral Palsy, initiated two Australia3, Europe27–29, Canada30, Sweden31 and Japan32
*e-mail: sKorzeni@
med.wayne.edu population surveys to determine CP prevalence10,11. have provided evidence for a declining prevalence of
https://doi.org/10.1038/ In Schenectady, New York, the prevalence was 5.9 cases CP over time, mostly among low-birthweight and pre-
s41582-018-0043-6 per 1,000 births, whereas in Minneapolis, Minnesota, term infants (Table 1). In China, a decline from 1.6 to
528 | SEPTEMBER 2018 | volume 14 www.nature.com/nrneurol

Reviews
born during 1959–1966. The NCPP provided compel-

Key points
ling evidence that clinical indicators of birth asphyxia
• Several high-income countries have reported a decline in the prevalence of cerebral (for example, fetal bradycardia, a low Apgar score and
palsy (CP); therapeutic hypothermia and magnesium sulfate for neuroprotection delayed time to first breath) occurred in only a minor-
might have played a role, but other factors might be important. ity of children who had CP53 and rarely led to CP when
• The minimum age at which CP can be reliably diagnosed is controversial; evidence they occurred. Approximately 70% of children with CP
that early diagnosis and intervention improves motor outcome is sparse, but there are had Apgar scores of ≥7 at 5 min after birth, and 95% of
hints of benefits for cognitive outcomes.
children with a normal birthweight and an Apgar score
• No consensus exists about CP subtypes; the general term CP is important for public of 0–3 at 5 min after birth were free from major disa-
health and health services planning, but subtypes might have different aetiologies.
bility at early school age. These findings made it clear
• Gestational age and birthweight have been a major focus of CP investigators; to move that birth asphyxia was not the dominant cause of CP
forward, we must seek a deeper understanding of why these factors convey that many clinicians and the lay public had assumed.
information about increased risk.
A convenience survey of clinicians (largely paediatricians
• Preconception factors, including maternal obesity and age, should be considered and obstetricians) shortly after the NCPP publications
because they can modify the relationships between CP and other factors that occur
found that the average estimate of the risk of CP after
later in pregnancy.
a low Apgar score was 40%, eightfold higher than the
• Two-hit and multi-hit models that consider accumulation of risk factors can identify a
5% risk identified by the NCPP54,55.
synergistic increase in the risk of CP, but the time order and clinical relevance of the
model components must be established.
The NCPP reinforced a finding noted by many stu-
dents of birth asphyxia that the risk of CP is markedly
elevated by the presence of abnormal neurological signs
1.25 cases per 1,000 children between 1999 and 2017 has in the newborn period, most notably seizures, an ina-
been reported33,34, although ascertainment methods for bility to suck and breathing difficulties, which together
these reports might have differed from the long-standing indicate the syndrome of neonatal encephalopathy.
registers used in Australia and Europe. This syndrome is often assumed to be the consequence
The reported prevalence of CP in South Korea, Japan of so-called hypoxic–ischaemic brain damage but can
and India is 2–3 cases per 1,000 live births35–37, but fig- occur in the absence of markers of distress during labour
ures greater than 3 cases per 1,000 live births in Taiwan38, and might even have a closer relationship to pre-labour
Egypt39 and Uganda40 have been reported in the past few factors56,57. Low Apgar scores, delayed onset of respira-
years. Some evidence suggests that rates in rural Africa tion, and seizures might be signs of birth asphyxia, but
are much higher41. New CP registers are emerging in they are neurological findings themselves and can reflect
Bangledesh42, Mexico43 and Jordan44, and hospital-based brain damage that was present before birth. Ellenberg
surveillance is being developed in Vietnam45. and Nelson have argued that the attribution of CP to
birth asphyxia often results from the conflation of the
Major aetiological factors consequences of factors that underlie both CP and birth
The first systematic clinical descriptions of CP were asphyxia58,59, and point out that placental pathology and
authored in the 1840s by the orthopaedic practitioner markers of infection often precede abnormal fetal heart
W. J. Little46,47. Little’s assertion that nearly all cases of patterns60 and low Apgar scores61.
what he called spastic rigidity of newborn children Breech vaginal delivery is now avoided in most
resulted from preterm birth or asphyxia at birth has left HICs owing to an assumed risk of birth asphyxia and
an enduring mark on subsequent thinking about the later CP. The NCPP indeed found an elevated risk of
aetiology of CP. Sigmund Freud, in his first career as a CP for normal-weight babies who were in the breech
child neurologist, cautioned against assuming that these position in utero, but the risk was the same whether the
two factors were fully causal48–50, but only in the latter baby was born vaginally or by Caesarian section53,62.
half of the 20th century did research begin to illustrate Breech position at term can reflect fetal abnormalities
the complex nature of these associations. that preceded the onset of labour, including abnor-
Nevertheless, Little’s insight that the perinatal period mal maternal thyroid function63, fetal growth restric-
was important to the pathogenesis of CP has been sup- tion (FGR), oligohydramnios , gestational diabetes
ported by subsequent research. Therefore, in this section, and fetal anomalies64. In randomized trials, mortality and
we begin by considering the two factors identified by short-term morbidity associated with breech presenta-
Little before reviewing factors related to preconception, tion were reduced by Caesarian section, but long-term
Apgar score
pregnancy and the perinatal period. A comprehensive neurodevelopmental outcomes were unaffected65.
A scoring system to predict the review of the risk factors for CP is available elsewhere51.
need of newborn babies for Gestational age and its correlates
medical attention; the overall Birth complications Children born preterm account for one-third to one-half
score is composed of scores for
The obstetric literature from the 20th century is rife with of CP diagnoses in HICs, although this proportion is
heart rate, respiratory effort,
muscle tone, response to examples of physical trauma in labour, sometimes exa much lower in low-income countries (LICs), where
stimulation and skin cerbated by instrumented deliveries or severe asphyxia mortality of very preterm babies remains high40. The
colouration. in labour and delivery, all of which are capable of caus- more preterm the newborn baby, the higher the risk of
ing brain injuries in children that could lead to CP52. CP; the prevalence of CP reaches ~10% among infants
Oligohydramnios
A pregnancy complication
However, these events were not placed in an appropri- who are born before 28 weeks of gestation, a prevalence
characterized by low amniotic ate statistical context until the National Collaborative that is ~50-fold higher than that among children born
fluid levels. Perinatal Project (NCPP) study of ~50,000 children at term66,67.
NaTuRE REviEwS | NeurOLOgY volume 14 | SEPTEMBER 2018 | 529

Reviews
Much research has been devoted to identifying the

“The condition of spastic rigidity of
the limbs of newborn children.” 1861 factors that place preterm infants at the greatest risk of
(Little, 1861) CP; the numerous factors that have been studied include
“Infantile cerebral palsy would thus organ immaturity, lack of hormones and growth factors,
be defined as the general concept
of all cerebral diseases in infancy metabolic factors and environmental toxins, infection
1868 caused by a direct effect of and inflammation, physiological instability, medical
“A paralysis of cerebral origin accidental etiology, occurring either interventions and pregnancy-related complications68.
occurring in the first six years of life. in the fetal period or after birth, and
affecting one or more neuron The most strongly associated indicators of CP risk in
It is always a hemiplegia, and the
cases which affect both sides of the systems.” (Freud, 1868) preterm infants are cranial ultrasonography observa-
body will be found to be no tions in the first few weeks of life69. Lesions that indicate
exception to this rule, but are 1916
white matter injury — usually seen as echolucency and
simultaneous lesions upon the “Cerebral palsy may be defined as a
hemispheres.” (Wilson, 1916) condition characterized by ventricular enlargement — are associated with a sub-
paralysis, paresis, incoordination, stantially increased risk of CP70, whereas isolated germi-
dyskinesia, or any aberration of nal matrix and ventricular haemorrhage, which are more
“Cerebral palsy is a descriptive term 1952 motor function that is due to
involvement of the motor control common than white matter injury in preterm infants, are
applied to a group of motor
disorders of young children, in centres of the brain.” (Peristein, associated with a much lower risk of CP71–76.
whom full function of one or more 1952) Mechanical ventilation is widely used for children
limbs is prevented by paresis, who are born very preterm and has been linked to
involuntary movement, or 1955
incoordination.” (Balf and Ingram, Cerebral palsy “may be defined as CP70,77. The effects of mechanical ventilation on risk are
1955) one component of a group of confounded, as its use is an indicator of illness sever-
childhood neurologic disorders
which reflect cerebral dysfunction ity78, but animal models and observational studies have
rather than damage per se and provided evidence that mechanical ventilation can cause
“Cerebral palsy is a persistent but 1960 which may result from cerebral systemic inflammation and lung damage and can alter
not unchanging disorder of maldevelopment, infection, injury,
movement and posture, appearing or anoxia before or during birth or blood gas levels in ways that might be detrimental to
in the early years of life and due to in the early years of life.” (Denhoff the perinatal brain79–81. Ventilator settings that produce
a non- progressive disorder of the and Robinault, 1960) hypocapnia have caused particular concern because they
brain, the result of interference
1964 have been associated with brain damage at autopsy and
during its development.” (The Little
Club, 1958 (Bax, 1964)) “Cerebral palsy is a group of with an increased risk of CP82–88. Antenatal administra-
disorders characterised by reduced tion of steroids during preterm labour to stimulate lung
“A disorder of movement and ability to make voluntary use of the
muscles, caused by a maturation has been associated with a reduced risk of
posture due to a defect or lesion of
the immature brain.” (Bax, 1964) 1967 non-progressive and non- CP89; by contrast, postnatal administration of steroids
hereditary brain disorder arising has raised concerns about an increase in CP90,91.
before or at delivery or during the
first years of life.” (Christensen and Intermittent or sustained systemic inflammation in
“Cerebral palsy is the result of a Melchior, 1967) which single92 or multiple inflammation-related proteins
lesion or maldevelopment of the are persistently upregulated93 in the newborn period has
brain, non-progressive in character
1969
and existing from earliest “Cerebral palsy is defined as a been associated with a twofold to threefold increase in
childhood. The motor deficit finds nonprogressive disorder of motion the risk of CP in babies born before 28 weeks of gesta-
expression in abnormal patterns of and posture due to brain insult or tion. Studies of infants who are born at a weight ≤1,000 g
posture and movement, in injury occurring in the period of
association with an abnormal 1976 early brain growth (generally under yield similar evidence94. Insufficient blood levels of
postural tone,” (Bobath, 1969) 3 years of age).” (The Kennedy neurotrophic and/or angiogenic proteins also seem to
Institute (Vining et al., 1976)) influence the risk of CP and other neurodevelopmen-
“Cerebral palsy is a descriptive term tal disorders in some settings. For example, transient
for a collection of nonprogressive
1986
“The term cerebral palsy does not hypothyroxinaemia, a common complication of preterm
neuromotor disorders of central designate a disease in any usual
origin that become manifest early medical sense. It is, however, a
delivery and a correlate of low birthweight95–98, occurs
in life and are not the result of a useful administrative term which most often in the smallest and sickest of neonates96–105,
recognized cerebral malformation.” covers individuals who are and some observational studies have found that neonatal
(Paneth, 1986) handicapped by motor disorders
thyroid hormone deficits are linked to CP, behavioural
1988 which are due to nonprogressive
abnormalities of the brain.” problems and lower cognitive performance assessed at
“A group of permanent disorders (Crothers and Paine, 1988) school age or in adolescence106–110, although other studies
of the development of movement
and posture, causing activity have not111,112. For example, a nationwide Dutch cohort
limitation, that are attributed to “An umbrella term covering a group study of infants born very preterm and/or with very low
non-progressive disturbances that 1992
of non-progressive, but often birthweights showed that relative hypothyroxinaemia
occurred in the developing fetal or changing, motor impairment
infant brain. The motor disorders (defined as values >1 s.d. below the mean) was associ-
syndromes secondary to lesions or
of cerebral palsy are often anomalies of the brain arising in the ated with psychomotor developmental delay at 2 years
accompanied by disturbances of early stages of its development.” corrected age109, neurological dysfunction assessed at
sensation, perception, cognition, 2007 (Mutch et al., 1992)
communication, and behaviour, by the age of 5 years, and school failure at age 9 years113.
epilepsy, and by secondary A population-representative study of 1,105 children
musculoskeletal problems.” born with a weight of <2,000 g similarly found a fourfold
(Rosenbaum et al., 2007)
increase in the risk of CP in newborn children with thy-
roid hormone levels >2.6 s.d. below the state norm com-
Fig. 1 | Definitions of cerebral palsy over time. The definition of cerebral palsy has pared with children with normal levels after adjustment
changed several times since 1861, and the most recent definition was described in 2007. for gestational age and multiple prenatal, perinatal and

Reviews
Table 1 | Trends in the prevalence of cerebral palsy overall and by birthweight or gestational age at delivery
Location (study) Prevalence group Years Observed prevalence (per 1,000 95% CI
in specified prevalence group)
Northern Alberta, Canada All live births 2008 2.38 2.19–2.57
(Robertson et al.30) 2009 2.26 2.10–2.44
2010 1.89 1.70–2.10
Birthweight <1,000 g 1985–1988 98.4 59.5–137.4
2008–2010 40.1 34.5–45.7
Birthweight 1,000–1,499 g 1985–1988 70.7 43.8–97.6
2008–2010 22.7 19.3–26.0
Birthweight 1,500–2,499 g 1985–1988 11.0 7.9–14.1
2008–2010 6.7 5.9–7.5
Birthweight ≥2,500 g 1985–1988 1.7 1.6–1.8
2008–2010 1.3 1.2–1.4
Europe Birthweight <1,000 g 1980 40.9 12.1–97.3
(Sellier et al.27) 2005 38.2 26.0–53.8
Birthweight 1,000–1,499 g 1980 70.9 41.7–110.9
2005 35.9 26.6–47.2
Birthweight 1,500–2,499 g 1980 8.5 5.4–12.7
2005 6.2 4.9–7.8
Birthweight ≥2,500 g 1980 1.2 0.9–1.5
2005 0.9 0.8–1.0
Japan All live births 1988–1997 1.8 1.6–2.0
(Touyama et al.32) 1998–2007 0.97 0.94–0.99
Birthweight <1,000 g 1988–1997 65.5 46.1–90.3
1998–2007 108.7 83.9–138.5
Birthweight 1,000–1,499 g 1988–1997 89.9 72.2–110.6
1998–2007 74.6 59.6–92.3
Birthweight 1,500–1,999 g 1988–1997 28.6 22.0–36.6
1998–2007 21.0 15.7–27.3
Birthweight 2,000–2,499 g 1988–1997 2.4 1.6–3.5
1998–2007 3.0 2.1–4.0
Birthweight ≥2,500 g 1988–1997 0.6 0.5–0.7
1998–2007 0.5 0.4–0.7
USA 8-year-old children 2006 3.5 3.2–3.9
(Durkin et al. ) 26
2008 3.2 2.9–3.5
2010 2.9 2.6–3.2
8-year-old children, excluding those with 2006 3.2 2.9–3.5
a documented post-neonatal cause 2008 3.0 2.7–3.3
2010 2.6 2.3–2.9
Victoria, Australia Gestational age ≥37 weeks 1983–1991 1.2 Not available
(Reid et al.3) 1992–2000 1.2 Not available
2001–2009 1.0 Not available
Gestational age 32–36 weeks 1983–1991 5.3 Not available
Gestational age 28–31 weeks 1983–1991 41.5 Not available
Gestational age <28 weeks 1983–1991 92.1 Not available

Reviews
early and late neonatal variables110. In a similarly large CP with these genes or any others studied after correc-
study of infants born with very low birthweight, tran- tion for multiple comparisons131. A nested case–control
sient hypothyroxinaemia was associated with a twofold study conducted in California, USA, that was intended
greater risk of white matter damage, a key antecedent to replicate the previously identified links between sev-
of CP, after adjustment for gestational age and various eral polymorphisms and CP in 127 affected children
measures of illness severity114. Whether the association indicated an association of CP with the inducible nitric
between neonatal thyroid hormone deficits and CP is oxide synthase (iNOS)-231 T allele (which is involved
truly causal can be determined only with a randomized in inflammation) and the APOE*ε4 allele. Both associa-
trial of thyroid hormone supplementation95,115. tions were statistically significant at the 0.04–0.05 level,
Similarly, severe maternal iodine deficiency, which but neither remained significant after adjustment for
presumably leads to maternal and thus fetal hypothyrox- multiple comparisons125.
inaemia, is associated with cognitive impairments and
neurological deficits that resemble CP116. Iodine defi- Multiple births. The prevalence of CP among twins is
ciency remains an important remediable cause of cog- fourfold higher than among singletons134,135, and this
nitive, neurosensory and motor impairment in several excess is greater for higher-order multiples136,137. Nearly
parts of the world117. all of this excess risk is accounted for by the lower gesta-
tional age and lower birthweights associated with multi-
Preconception and early gestation ple births138,139, but the risk of CP is slightly higher even
Genetics. Familial clustering of CP has been described: for full-term multiple births than for singletons born at
three studies since 2007 have indicated that the relative term140–142. CP in one twin is associated with a greatly
risk of CP for the sibling of a child with CP is four118, increased risk of CP in their co-twin: in one analysis of
five119 and nine120, respectively. However, even these >20,000 twin sets in Norway, the twin of a child with
relative risks translate to a small absolute risk of CP CP had a 15-fold higher risk of CP than singletons120.
of 1–2%. Copy number variants and mutations in sin- Despite this elevated risk, fewer than 12% of twin sets are
gle genes have been implicated in CP, but these find- concordant for CP143. In the only two published series of
ings are limited by small numbers of patients, genetic identical twins, concordance for CP was 18–40%144,145.
heterogeneity and a paucity of validation studies121. Little is known about the factors that contribute
Whole-exome sequencing has revealed several poten- to disparate susceptibility to CP between co-twins,
tially disease-causing gene variants, but functional and although factors such as birth order, birthweight dis-
pathway studies are needed to validate these findings122. cordance, gender and chorionicity have been exam-
Among the most studied genetic risk factors is APOE ined145–148. Monochorionic placentation is thought to
genotype. APOE encodes apolipoprotein E (ApoE), a convey a higher risk than dichorionic placentation149, as
lipid transport protein that is abundant in the brain; for the risk of congenital anomalies150, but information
the APOE*ε4 allele has been associated with several on zygosity and chorionicity is rarely known for children
neurological conditions, most notably Alzheimer dis- with CP, making this thesis difficult to investigate in
ease. Studies of APOE genotype and the risk of CP have most databases. Genetics could have a role, but other fac-
produced a variety of results. The APOE*ε4 allele has tors are more strongly associated; for example, the death
been associated with an elevated relative risk of CP of a co-twin in utero is associated with a s ubstantially
in one Brazilian study123 and two studies conducted in higher risk of CP in the surviving twin139,151,152.
the USA124,125, one of which also indicated an elevated
risk among carriers of the APOE*ε2 allele. In a large, Socio-economic status and correlates. Several studies
population-based study conducted in Norway, the have demonstrated that children who are socially dis-
APOE*ε4 allele was associated with an increased sever- advantaged are at higher risk of CP than those who are
ity of CP126, whereas the APOE*ε2 or APOE*ε3 alleles not138,139,153–157. For example, a higher prevalence of CP
and the s59007384 polymorphism in the TOMM40 has been identified among African-American children
gene (which is located on chromosome 19 close to than among other children138,139, but this observation
APOE) were associated with reduced severity of CP127. was only partially explained by differences in level of
However, in a second Brazilian study, an elevated risk maternal education and was largely a function of higher
of CP was associated only with the APOE*ε2 allele128, rates of preterm birth among socially disadvantaged
and a Norwegian family study identified APOE*ε3 women26,154. However, although multiple indicators of
to be the APOE allele most closely linked with CP129. social disadvantage are associated with an increased
Furthermore, three much larger studies — two con- risk of CP, these relationships seem to be moderated or
ducted in Australia130,131 and one in China132 — found mediated by differences in gestational age, birthweight
no association of any APOE allele with the risk of CP. and/or their correlates (for example, maternal obe-
Hypocapnia
In a review published in 2009 (ref.133), an associ- sity139,153,154, which is more common among women of
A condition characterized by
low blood levels of carbon ation between certain thrombophilia-related genes lower socio-economic status158–161).
dioxide. and CP caused by intrauterine strokes was suggested, Analysis of a database of 6 million births in California,
but the literature on this area is sparse. A subse- USA, revealed a dose–response relationship between
Chorionicity quent Australian study of candidate maternal or fetal pre-pregnancy obesity and CP: the relative risk of CP for
Whether twins share the same
placenta (monochorionic) or
thrombophilia-related genes that included 587 children the children of mothers classed as morbidly obese (2.7)
each have their own placenta with CP and their mothers and 1,154 healthy mother was significantly higher than for those whose mothers
(dichorionic). and child pairs identified no significant association of were classed as obese (1.3)162. Large studies conducted

Reviews
in South Carolina (USA)163, Norway and Denmark164, chikungunya virus infection192. Zika virus infection
and Sweden25 produced similar results. In the Swedish in utero can damage the fetal brain193, but the magnitude
study, the association was limited to infants born at term. of the contribution of Zika virus to CP is not yet under-
Putative mechanisms for this consistent finding include stood. To date, seven case-series and one cohort study
the excess inflammation seen in obesity165,166, placental have examined whether children with clinical evidence
dysfunction167,168 and thyroid hormone deficits169,170. of congenital Zika virus infection exhibit early indica-
However, maternal obesity does not seem to be associ- tors of motor dysfunction and epilepsy; in total, 54% of
ated with an increased risk of CP among children who these children had seizures and all of them were judged
are born before the 28th week of gestation171. to have abnormal motor development after follow-up
periods of 3–12 months194.
Pregnancy and congenital anomalies The presence of several nonspecific indicators of infec-
Fetal growth restriction and maternal pre-eclampsia. tion, such as maternal fever195–197, maternal receipt of anti-
Several studies have identified associations of CP with biotics198 and chorioamnionitis197,199,200, close to the time
FGR in infants born at term or near term 172–174, in of delivery have been linked to an increased risk of CP.
infants born late or moderately preterm174 or in infants Infections earlier in pregnancy have often been a ssociated
born at any gestational age175. The most definitive data with CP as well201–204, but not all studies agree205.
come from the pooling of several national CP registers
referred to as Surveillance of Cerebral Palsy in Europe, Birth defects. Two lines of evidence indicate that devel-
which revealed a significant fourfold to sixfold excess opmental aberrations similar to those that cause birth
risk of CP among infants with FGR who were born defects are involved in an appreciable fraction of CP cases.
at 32–42 weeks of gestation176. This finding is consist- First, imaging studies of children with CP have shown
ent with evidence from an Australian reconstructed that cerebral malformations, which are often unsuspected
population cohort study in which being small for before CT or MRI, are not infrequent. One systematic
gestational age and pregnancy-induced hypertension review of the topic showed that such malformations were
were associated with a twofold to ninefold increase in found in 10% of children with CP, mostly in children who
the risk of CP among all infants born after 27 weeks were born at term206. Several different brain develop-
of gestation175. mental defects seem capable of producing CP, especially
FGR and maternal pre-eclampsia have both been neuronal migration disorders207, such as schizencephaly208
associated with a reduced risk of CP in studies of children and polymicrogyria209. Interestingly, cytomegalovirus
born with low birthweight177,178. However, birthweight infection might underlie these neuronal migration
is inextricably linked with gestational maturity, which is disorders in some instances191,210. In general, the brain
strongly associated with CP; therefore, studies of chil- malformations involved seem to arise from a wide variety
dren with a birthweight below a specified threshold will of genetic and environmental insults, are not rare in
include growth-restricted newborn babies from more CP and can be congenital or acquired211. Intrauterine
mature gestational age strata, who have a lower risk of CP infections are associated with congenital anomalies; the
than infants of lower gestational ages179. Population-wide most recently identified association is with Zika virus
studies that include infants of all gestational ages do not infection212,213. A detailed review of the development
reproduce this anomaly118,176,180. of these abnormalities is available elsewhere51.
This issue particularly affects studies of pre-eclampsia The second line of evidence is the frequent presence
and CP. A study conducted in Australia showed that of malformations outside the nervous system in children
pre-eclampsia seems to provide a strong protective with CP. The NCPP showed that such malformations are
effect against CP in infants with low birthweights, but observed threefold more often in children with CP than
that no association with CP is observed in a sample in healthy children, a finding that has been confirmed in
defined by truncation of gestational age rather than of subsequent studies214–216. Hypertensive disorders of preg-
birthweight181. Confounding of FGR with gestational nancy are associated with congenital malformations, in
age seems to be reduced, if not completely avoided182, particular congenital heart defects217,218, as are indicators
by selecting samples of preterm infants on the basis of of thyroid dysfunction219–221; each of these factors are
truncated gestational age rather than truncated birth- associated with FGR, CP and other neurodevelopmen-
weight, a practice that is becoming more common183 but tal disorders (possibly reflecting mitochondrial dysfunc-
is still not universal184. tion222). A study conducted in Australia found that 53%
of children with CP and severe FGR who were born at
Infection. Maternal infections can lead to CP by transmis- or near term had a major birth defect180.
sion of pathogens to the fetus (even without a detectable
maternal inflammatory response185) and by induction Perinatal risk factors
of persistent systemic inflammation that can sensitize Perinatal stroke. Perinatal stroke, which occurs between
Schizencephaly
A birth defect characterized by the brain to subsequent insults186–188. Infections such late gestation and 28 days after birth, might account
abnormal clefts in the cerebral as toxoplasmosis, rubella, cytomegalovirus and her- for as much as half of hemiplegic CP in infants born
hemispheres of the brain. pes simplex virus during pregnancy have been associ- at term223,224, whereas children with other CP subtypes
ated with increased risks of CP189–191, but these agents often have multifocal or more diffuse injury. The most
Polymicrogyria
A birth defect characterized by
account for only a small fraction of CP cases in HICs. common form of perinatal stroke is thrombosis in
abnormal ridges and folds in A subtype of CP that is associated with microcephaly has the arterial distribution, usually in the middle cere
the brain. been reported as a result of perinatal mother-to-child bral artery, but periventricular venous infarction can

Reviews
contribute225,226. Most perinatal strokes are ischaemic, guidelines that recommend MgSO4 for fetal neuro-
but haemorrhagic strokes can occur, sometimes as a protection in the setting of imminent preterm birth at
complication of ischaemic injury227. <32–34 weeks of gestation252. Implementation of these
The causes of perinatal stroke are largely unknown. guidelines might be a contributor to the reduction in CP
Perinatal stroke can be a complication of congenital heart prevalence among children who have low birthweights
disease228 and bacterial or viral meningitis229. Placental in many HICs3. Nevertheless, MgSO4 can cause maternal
factors such as chorioamnionitis199, prolonged rupture of adverse effects248, such as respiratory depression, hypo-
membranes230 and placental thrombosis231 have also been tension and confusion, and questions remain about the
implicated. Pre-eclampsia and FGR are also risk factors223. optimal timing of administration253, dosage253,254 and
Genetic predispositions to t hrombophilia have been duration255,256, and modifiers of therapeutic efficacy
described above. (for example, maternal obesity)257,258.
Kernicterus. Neonatal jaundice is the most common Therapeutic hypothermia

complication of the newborn period, usually caused by A series of randomized trials with consistent results have
unconjugated hyperbilirubinaemia232. Unconjugated shown that lowering of body and/or head temperature
bilirubin crosses the blood–brain barrier in the first days by 2°C for 48 h, beginning in the first few hours after
of life and can damage several parts of the brain, particu- birth, can reduce the risk of and mortality from CP in
larly the basal ganglia and acoustic nuclei. The yellow children who are born after 36 weeks of gestation and
staining seen at autopsy in this condition is the origin who have hypoxic–ischaemic encephalopathy259,260.
of the neuropathological term kernicterus. Symptoms of All trials enrolled infants with signs of encephalopathy
bilirubin-associated encephalopathy in newborn babies and indicators of fetal and immediate neonatal dis-
include lethargy, impaired tone, cyanosis, vomiting and tress that are thought to reflect asphyxia. Therapeutic
absence of the suck reflex233. Chronic bilirubin encepha hypothermia is now used internationally261,262 and has
lopathy manifests as aberrant processing disorders become the standard of care in many centres in HICs.
(particularly hearing impairment) and CP. Trials are underway to determine the safety and efficacy
CP that results from kernicterus is typically choreo- of cooling therapy in low-income and middle-income
athetotic or dystonic, indicating damage to extrapyram- countries (LMICs)263 and in infants whose treatment was
idal structures233. In preterm infants, kernicterus can delayed beyond the first few hours. Whether head cool-
occur with bilirubin levels that would not pose risks to ing would be effective in infants with neonatal enceph-
infants born at term. Kernicterus was once a common alopathy that is not associated with evidence of hypoxia
cause of CP234,235, but in HICs, the number of cases of or ischaemia is an open question.
CP associated with kernicterus has declined substan- Among infants who are born at term or late preterm
tially owing to prevention and better management of and who have moderate to severe encephalopathy, hypo-
newborn hyperbilirubinaemia. Kernicterus remains thermia improves survival and neurodevelopmental out-
a notable cause of CP in LICs236,237. comes, including the development of CP, at 18 months of
age257, but CP is not eliminated altogether by this treat-
Prevention of cerebral palsy ment. Children with severe newborn encephalopathy
Magnesium sulfate remain at increased risk of severe neurodevelopmental
Case–control studies conducted in the 1990s suggested impairment despite the treatment264. Preterm birth and
that magnesium sulfate (MgSO4), which is adminis- comorbid exposure to clinical or histological chorioam-
tered in pregnancy to treat pre-eclampsia and in pre- nionitis (or the correlates of each) are associated with a
term labour to slow contractions, reduces the risk of reduced response to hypothermia265.
CP in infants born preterm238–240. Modest support for Attempts to combine cooling with adjunctive ther-
this hypothesis was provided by some cohort studies240 apies to provide additional benefits are underway266,267.
but not others241,242. A concern in all such observational For example, 1,000 U/kg erythropoietin administered
studies was confounding by indication78 — the possibil- intravenously in three doses (immediately after birth, at
ity that recipients of MgSO4 were at an inherently lower 24 h and 1 week later) in conjunction with hypothermia
risk of CP, given that maternal pre-eclampsia has been is well tolerated and produces plasma concentrations
associated with a lower risk of CP in studies of infants that are neuroprotective in animals. However, a large
with truncated birthweights178,243–245. However, the bene efficacy trial is needed to determine whether this add-on
fit of MgSO4 was clearly demonstrated by the BEAM therapy improves outcomes in infants u ndergoing
trial, in which the risk of CP in infants born before 32 hypothermia268.
weeks of gestation and who were randomly assigned
to receive MgSO4 during labour was nearly 40% lower Investigational therapies
than among infants of the same gestational age who Various other therapies for CP have been tested and are
received placebo246,247. Two meta-analyses of randomized under investigation. Caffeine is one of the most com-
trials converge on an estimated 30% reduction in the monly prescribed medications in neonatal intensive care
risk of CP as a result of MgSO4 administration248–250. units as a treatment for apnoea269. A Canadian multi-
A similar reduction was reported in a meta-analysis of centre trial of caffeine in premature infants with apnoeic
observational studies251. episodes indicated a reduction in the risk of motor and
Canadian, British, American and Australian obstet- cognitive impairment with the treatment. The finding was
ric societies and bodies have issued clinical practice statistically significant at 18 months270 but not at 5 years271.

Reviews
Melatonin treatment in combination with hypo- Claims that accurate identification of children who
thermia has been compared with hypothermia alone are at high risk of CP is possible at just a few months of
in one small trial (n = 15 in each arm) in infants with age are increasing in the literature. However, clinical pre-
birth asphyxia272. The results hinted that melatonin diction studies have often assessed CP in combination
improved survival during 6 months of follow-up and with other disorders or used other proxies for a diag-
lowered the risk of seizures and indicators of white nosis of CP, such as an interim clinical diagnosis of a
matter injury assessed at 2 weeks after birth. A trial of high risk of CP before a true diagnosis is confirmed287,288.
melatonin in preterm birth before 28 weeks of gestation A review published in 2017 concluded that CP could
is ongoing273. be diagnosed on the basis of the absence of so-called
Erythropoietin has been studied in four randomized jittery movements in infancy287, but the conclusions were
trials in infants born very preterm or who are very small. based largely on expert opinion, most of the reviewed
The treatment seemed to reduce the proportion of chil- studies defined CP in combination with other motor
dren with Bayley Mental Development Index scores <70 problems, and none included data beyond the age of
at follow-up, but the treatment did not affect the risk 2 years (see the appendix in the review287). The use of var-
of CP274. ious definitions of ‘high risk of CP’ has hampered direct
Finally, stem cell therapies for CP are also under comparisons between studies. For example, of 12 studies
investigation. These studies are not yet at a sufficiently that focused on motor dysfunction (see a review of these
advanced stage to discern the effects of the treatment on studies288), the outcome assessed in 6 was CP ‘or other
CP275,276, but evidence from small studies hints at benefits disorders’, and in 5 the clinical decision relied on clini-
for some children277. cal impressions, intelligence or development quotients;
in only 1 of the 12 was CP assessed alone. In addition,
Early diagnosis and intervention the majority of these studies do not include follow-up of
The age at which a CP diagnosis is validly, reliably and children at or beyond 2 years of age, and most are not
fully ascertained by ruling out transient or progressive population-based or geographically representative of the
motor problems is controversial. Long-standing evi- general population or a targeted high-risk group.
dence indicates that a proportion of motor problems In some studies in Australia and Europe, use of the
detected before the age of 1 year (for example, develop- General Movements Assessment in high-risk patients
mental delay, coordination problems and transient dys- has indicated a high sensitivity and specificity for the
tonia) resolve by school age without intervention278–282 prediction of a high risk of CP. These findings have
and that a small fraction of motor disability in children prompted the Australian investigators to recommend
of school age is the result of progressive motor patho widespread adoption into clinical practice289. However,
logy (for example, in metabolic disorders283), neither of given that decisions about diagnostic tests should be
which fits the current model of CP. Clinical prediction made on the basis of patient benefits289, we feel that
models and neuroimaging have been used to diag- additional research is needed, as do others290. Indeed,
nose CP before the age of 2 years, but further research although some evidence indicates that early detection
is necessary. and intervention can improve some cognitive out-
comes291, the improvements observed in motor dysfunc-
Clinical prediction studies tion are much more modest287,291, and the evidence for
Few studies have repeatedly assessed CP status over a benefit is limited by a dearth of high-quality trials292.
the first several years of childhood among the general
population. Evidence from two studies suggests that Neuroimaging studies
a diagnosis of CP at or before a child’s first birthday We have discussed the use of neuroimaging in detail in
is unreliable280,281. In the largest study to date, which a previous review293, in which we concluded that neu-
included 37,000 children, 45% of those who were diag- roimaging is not required for diagnosis of CP because
nosed with definite CP at their first birthday ‘outgrew’ the disorder is based on clinical findings, and the prin-
their motor problems by the age of 7 years280, and fewer cipal contribution of imaging is to the understanding of
than 3% of infants who were thought to have probable CP aetiology and pathogenesis. We also concluded that
CP at age 1 year had CP at age 7 years. It is impor- imaging studies were less informative than they could
tant to distinguish between CP diagnosed at an early be, largely because study designs were rarely based on
age solely on the basis of neurological findings and CP generalizable samples.
accompanied by clear evidence of a disability. Thus, a Since this previous review, several population-based
diagnosis made after ~2 years of age is more reliable284, neuroimaging studies have provided evidence that the
particularly when the motor problems are disabling occurrence of white matter injury has declined and that
(for example, an inability to walk five steps unaided of grey matter injury has increased, whereas the preva-
or a need for physical assistive devices). However, in a lence of malformations in children with CP remained
study of children who were born preterm and weighed about the same. However, the authors of a review of
<2,000 g at birth, diagnosis of non-disabling CP was these studies294 concluded that a dearth of standardized
not a stable diagnosis until after 6 years of age285. This protocols and terminology were probably responsible for
instability of diagnosis at the milder end of CP probably the heterogeneity of these findings (for example, there
explains why some, but not all, CP registers conduct was no minimum or standard age at assessment).
regular clinical reassessments at the age of 5 years to The well-designed Generation R study of a geo-
ensure diagnostic accuracy286. graphically representative sample from the Netherlands,

Reviews
published in 2017, highlighted a diverse array of research or four events increased the risk of CP to 17%, 27%, 40%
questions that can be addressed by merging neuro and 55%, respectively305.
imaging with developmental neuroscience and epi-
demiology 295; an example is an ongoing effort to Hidden from view
develop growth curves of optimal brain development. Although a small fraction of CP is caused by obvious
Nevertheless, studies of selected populations continue postnatal events, such as meningitis or head trauma, at
to be more common, and these types of studies were least 90% of the factors that predispose infants to CP in
considered in a 2018 systematic review of early MRI HICs are no longer operative and are often no longer
for detection of motor outcomes at term-corrected age detectable after the perinatal period. Furthermore, many
in children born preterm296. This review identified a prenatal causes are often hidden from view because the
high sensitivity and specificity of MRI for detection inflammatory stimuli that seem to contribute to preterm
of CP or other motor problems, but the findings were birth, the silent strokes and the hormonal changes oper-
difficult to generalize because not all participants had ate under the cover of amnion and uterus. However, a
CP, and few studies were of unselected, sequentially firm diagnosis of CP cannot usually be made until the
recruited, representative infants. Instead, most were child’s nervous system is mature enough to display
carried out in tertiary centres with high-risk patients, the motor deficits that describe CP, usually at about the
and recruitment rates of eligible infants were as low as age of 2 years. By the time a child is diagnosed with CP,
17% in some settings. Consequently, further research the history of the pregnancy and the perinatal period
is needed to establish the effect of sample compo- can be reconstructed only from medical records and
sition 297 and the clinical relevance of MRI in early maternal recall, which are useful but have limitations
identification of CP. (for example, recall bias and reliance on clinical notes).
Prospective data collection can improve the precision of
Complexities in cerebral palsy epidemiology data about exposures, but to date, few prospective pre-
Accumulation and interaction of insults conception or pregnancy cohort studies have included
CP can often be the result of combined insults, but not follow-up of children to school age or beyond to assess
all insults are equally influential and some probably have neurodevelopmental outcomes. Consequently, delivery
context-specific effects; the importance of timing and and neonatal factors have received much consideration,
co-occurrence of different insults is unclear, although partly because of the dearth of quality evidence about
several examples have been studied. The combination earlier antecedents of CP. This bias has important impli-
of antenatal inflammation with postnatal systemic cations because pre-pregnancy and early pregnancy risk
inflammation is associated with an increased risk of CP factors (such as body weight) and their correlates can
in children born before 28 weeks of gestation298, as are influence the capacity for delivery-related and postna-
combinations of postnatal systemic inflammation and tal factors to provide information about the risk of CP
high or low levels of proteins that are related to angio- (for example, maternal age)306.
genesis and thyroid dysfunction299,300. FGR and/or very
preterm birth seem to make infants particularly suscep- One disorder or many?
tible to multiple-hit phenomena that involve white mat- The relationships between clinical entities that are used as
ter injury and its correlates because these infants tend to diagnoses and established pathophysiological pathways
have more vigorous postnatal inflammatory responses vary greatly. Some disorders are defined by abnormal-
than do other infants301–303. Chronic placental inflam- ities at the molecular level, such as sickle cell anaemia,
mation followed by acute fetal inflammation followed but many are defined by their clinical appearance307
by neonatal illness is similarly associated with cerebral and often represent the common end point of various
white matter injury in infants who are born before pathophysiological changes. Use of such disease entities
32 weeks of gestation304. often continues despite recognition of their heterogeneity
Additional evidence that accumulation of insults is because the entity is useful in clinical management; for
involved in CP comes from a population-based study of example, whether a stroke was ischaemic or thrombotic
infants with birthweights <2 kg with risk factors associ- matters little for the rehabilitation of a patient.
ated with ventilator use. This study showed that the risk of As discussed above, the term CP includes entities that
CP increased incrementally with addition of three venti- are apparently caused by multiple pathways, yet the sin-
latory risk factors: hyperoxia, hypocapnia and ventilation gle term persists. Some epidemiologists have referred to
for longer than expected for gestational age. The pres- ‘cerebral palsies’308 and the current definition refers to a
ence of one ventilatory risk factor was associated with an group of disorders to indicate the heterogeneity309,310, but
11% risk of CP, two risk factors were associated with a this usage has not taken hold in clinical settings, largely
35% risk and three risk factors were associated with because the traditional phenomenological entity and its
a 57% risk88. In a study of over 200,000 pregnancies, observable subdivisions are clinically useful for prog-
pre-eclampsia, neonatal infection, presumed birth nosis and management. Novel classification schemes
asphyxia and neonatal illness were cumulatively asso- based on topography have not consistently improved
ciated with an increased risk of CP213. Similarly, a study reliability in describing CP subtypes311; therefore, the
of nearly 6,000 infants with very low birthweights in field has moved away from classification according
Taiwan indicated a strong interaction between sepsis and to the underlying impairment and towards a focus on
postnatal hypoxic–ischaemic events. Infants with sepsis functioning312–315. What this trend means for aetiological
alone had a 10% risk of CP, but adding one, two, three research remains to be seen.

Reviews
The temptation, which is consistent with the current born at term, among whom rates of congenital birth
emphasis on precision medicine, is to divide the con- defects are very high180. Some evidence indicates that
dition into narrower subdivisions that are more likely patterns of childhood growth differ according to CP
to be aetiologically homogeneous. However, when a subtype326. Finally, the antecedents of the severest forms
clinical entity is fairly uncommon, as is the case for CP, of CP seem to differ from those of less severe motor dis-
creating finer and finer subdivisions makes investiga- orders327, and in some studies the profile of comorbid
tion of aetiology very difficult. Even in the collaborative neurodevelopmental disorders differs by CP subtype
perinatal project, which included 50,000 participants, (for example, the co-occurrence of intellectual disa-
analysis did not consider much beyond the diagnosis bility and epilepsy seems to be more likely to develop
of CP itself; the only difference emphasized was that among children who develop hemiplegia than among
between children with birthweights above and below children who develop other types of CP)328,329.
2,500 g (refs53,316,317). Two initiatives could help to over- Despite this evidence, without agreement on relia-
come this difficulty: one is the newly initiated integra- ble definitions of CP subtypes, the findings are difficult
tion of Danish and Norwegian birth cohorts (referred to interpret and hard to replicate across populations
to as MOBAND) 318, which aggregates the medical with different background prevalence of contributing
records from 200,000 well-studied pregnancies and factors. Integration of findings from hypothesis-driven
births and links them to the national CP registers of clinical studies and preclinical studies is most likely
those countries, and the other is the Global Pregnancy to inform about the appropriate subdivision of
‘CoLaboratory’319 (CoLab; see Related links), a network neurodevelopmental syndromes that are rooted in
of investigators from multiple population-based register pregnancy.
and cohort studies who seek to facilitate standardiza-
tion of data collection instruments and study designs Future directions
and to enable data sharing to solve complex ques- Large public health interventions (for example, improved
tions about major pregnancy complications and their access to clean water, hygiene and other interventions
sequelae. However, no consensus has been reached that focus on common infections before, during and
on the kinds of subdivisions (for example, gestational between pregnancies) combined with rigorous empha-
age at birth, type of motor abnormality or distribu- sis on hypothesis-driven data collection in LMICs could
tion of affected limbs) that would lead to the greatest probably teach us a lot about CP aetiology. A similar
aetiological insight318. focus on periconception modifiable risk factors that
Specificity is problematic for all symptom-based influence the early gestation milieu that contributes
case definitions320, and mis-specification of heteroge- to increased risk of CP would be informative in LICs
neous syndromes as singular entities can mask impor- and HICs — for example, a focus on alleviating social
tant findings about aetiology321, including evidence of disadvantage. Overall, there is a pressing need to under-
therapeutic benefits322. Subtypes of a syndrome can be stand how the environment influences brain develop-
established in three ways. The first is to demonstrate ment, how it affects placentation and later placental
different risk profiles. The second is to establish that the functioning while the brain grows, and why specific
underlying morbid anatomy or pathophysiology differs. environmental factors seem to raise the risk of CP and
The third is to demonstrate that a subset of a disorder precipitating pregnancy disorders.
can be prevented (as in the case of kernicteric brain Current knowledge of the pathogenesis of CP dur-
damage prevented by control of bilirubin blood levels). ing embryogenesis and placentation rests largely on
In reality, none of these strategies alone will provide a evidence from models in cell lines and embryos and
definitive answer about subtypes of CP because some on downstream placental histopathology and related
similarities and differences in the antecedents, neu- clinical syndromes330–332. However, we do not know
roanatomy and efficacy of interventions are observed whether these methods represent in vivo pathogene-
across subtypes. sis or how to translate emerging preclinical informa-
Several lines of evidence indicate that environ- tion into population-level benefits. Maternal and fetal
mental and maternal characteristics are differentially blood-based profiling and imaging techniques are much
associated with different subtypes of CP. Some studies less informative about pregnancy disorders before the
have provided evidence that social status323 and mater- 15th gestational week333,334, which seems to be a key
nal body weight324 are associated with increased risk of aetiological window333,335–337. Emerging technologies,
hemiplegia, but not other CP subtypes. Consideration such as trophoblast retrieval and isolation from the
of maternal age also influences which factors are most cervix338,339, might provide relevant information about
strongly associated with increased risk of different early placental development and function in ongoing
types of CP306. Stratification of infants according to pregnancies, but large follow-up studies are needed to
postnatal occurrence of persistent systemic inflam- discern the relevance to CP and its correlates. Single bio-
mation seems to provide different information about markers are unlikely to be very useful clinically for large
the risk factors associated with quadriplegia299,300, but sections of the general population, but molecular pro-
not other CP subtypes, among children who are born filing to establish the timing and order of processes that
before the 28th week of gestation. The CP risk factor disrupt physiological homeostasis during pregnancy
profiles for different subtypes among infants with FGR might prompt better thinking about aetiology and pre-
also differ from those for infants with normal growth325, vention. For example, consideration of key concepts
especially for children with CP who have FGR and are from physiology, such as homeostasis, regulated systems

Reviews
and redundancy, as major intellectual tools to under- Conclusions

stand aetiology might improve our capacity to formulate We know that the aetiological factors involved in most
and test hypotheses340 (details of specific examples are cases of CP operate between conception and a few weeks
available elsewere341–343). after birth, yet identifying preventable causes of CP has
Infection and other inflammation-inducing expo- proved difficult. Several risk factors have been identi-
sures before and after very preterm birth are asso- fied, but they overlap and interact with each other in
ciated with an increased risk of CP via complex ways that are not easy to dissect. Nevertheless, impor-
pathways92–94,298,344. Some evidence suggests that the tant to keep in mind is that the young 21st century has
risk of CP and comorbid brain disorders is decreased witnessed the development of two preventive measures
by administration of exogenous proteins that reduce for CP: MgSO4 treatment for mothers at risk of preterm
neuroinflammation, possibly by controlling the regu delivery, and head and/or body cooling for infants born
lation of systemic inflammation and neuropoietin at term with neonatal encephalopathy.
signalling 116,345,346; however, meticulously designed, The ongoing formation of large national databases
hypothesis-driven preclinical and clinical studies of data from unselected pregnancies in combination
will be needed to translate these observations into with multidisciplinary collaboration to integrate these
population-wide benefits322,347. resources for the study of complex diseases in preg-
We cannot overstate the importance of preterm deliv- nancy gives hope for new discoveries; these data include
ery and FGR348, which are themselves heterogeneous real-time surveys of exposures in pregnant women, and
syndromes349,350, in the CP risk profile, but large studies biological specimens such as archived serum, urine
of infants born at term, which account for approximately and newborn blood spots. This information is invalu
half of all CP diagnoses, might enable easier identifi- able when accompanied by follow-up information about
cation of antecedents, in part because the number of the presence or absence of CP in the offspring, or when
indicators of different pathological processes increases linked to records of CP. If the modest information about
as the gestational age at delivery decreases and, conse- pregnancy that is currently available from maternal recall
quently, so does the complexity351–353. Population-based and medical records years later can be supplemented
studies of pregnancies at all gestational ages will con- with data sources that provide clear information about
tinue to be the most helpful. CP registers and emerging exposures to infections, nutrients, environmental toxins,
consortiums that focus on standardized data collection, allergens and many other phenomena during pregnancy,
analysis and sharing to study complex pregnancy disor- our limited knowledge could give way to an era in which
ders (for example, CoLab) are beginning to enable the the widespread prevention of CP is a feasible goal.
kind of research necessary to examine these possibilities
more closely354,355. Published online 13 August 2018
1. Kirby, R. S. et al. Prevalence and functioning of 11. History of the study of cerebralpalsy. Acta Pædiatr. loss, intellectual disability, and vision impairment,
children with cerebral palsy in four areas of the United 46, 2–3 (1958). metropolitan atlanta, 1991–2010. PLOS One 10,
States in 2006: a report from the Autism and 12. Polani, P. 1. Classification of cerebral palsy: yesterday, e0124120 (2015).
Developmental Disabilities Monitoring Network. and today. Cerebral Palsy Bull. 2, 36–39 (1959). 25. Maenner, M. J. et al. Prevalence of cerebral palsy and
Res. Dev. Disabil. 32, 462–469 (2011). 13. Bax, M. et al. Proposed definition and classification of intellectual disability among children identified in two
2. Christensen, D. et al. Prevalence of cerebral palsy, cerebral palsy, April 2005. Dev. Med. Child Neurol. U. S. National Surveys, 2011–2013. Ann. Epidemiol.
co-occurring autism spectrum disorders, and motor 47, 571–576 (2005). 26, 222–226 (2016).
functioning – Autism and Developmental Disabilities 14. Rosenbaum, P. et al. A report: the definition 26. Durkin, M. S. et al. Prevalence of cerebral palsy
Monitoring Network, USA, 2008. Dev. Med. and classification of cerebral palsy April 2006. among 8-year-old children in 2010 and preliminary
Child Neurol. 56, 59–65 (2014). Dev. Med. Child Neurol. Suppl. 109, 8–14 (2007). evidence of trends in its relationship to low
3. Reid, S. M. et al. Temporal trends in cerebral palsy 15. Andersen, B. Incidence of cerebral palsy in norway and birthweight. Paediatr. Perinat. Epidemiol. 30,
by impairment severity and birth gestation. some C. P. problems. Acta Paediatr. 44, 22–22 496–510 (2016).
Dev. Med. Child Neurol. 58 (Suppl. 2), 25–35 (1955). 27. Sellier, E. et al. Decreasing prevalence in cerebral
(2016). 16. Pharoah, P. O. Epidemiology of cerebral palsy: palsy: a multi-site European population-based study,
4. Zhang, J. Y., Oskoui, M. & Shevell, M. A a review. J. R. Soc. Med. 74, 516–520 (1981). 1980 to 2003. Dev. Med. Child Neurol. 58, 85–92
population-based study of communication impairment 17. Goldsmith, S. et al. An international survey of cerebral (2016).
in cerebral palsy. J. Child Neurol. 30, 277–284 palsy registers and surveillance systems. Dev. Med. 28. Andersen, G. L. et al. Cerebral palsy among children
(2015). Child Neurol. 58 (Suppl. 2), 11–17 (2016). born moderately preterm or at moderately low
5. Mei, C. et al. Language outcomes of children with 18. Paneth, N. Etiologic factors in cerebral palsy. birthweight between 1980 and 1998: a European
cerebral palsy aged 5 years and 6 years: a population- Pediatr. Ann. 15, 191 (1986). register-based study. Dev. Med. Child Neurol. 53,
based study. Dev. Med. Child Neurol. 58, 605–611 19. Paneth, N., Hong, T. & Korzeniewski, S. 913–919 (2011).
(2016). The descriptive epidemiology of cerebral 29. Glinianaia, S. V., Rankin, J., Colver, A. & North of
6. Levy, S. E. et al. Autism spectrum disorder and palsy. Clin. Perinatol. 33, 251–267 (2006). England Collaborative Cerebral Palsy Survey. Cerebral
co-occurring developmental, psychiatric, and medical 20. Bhushan, V., Paneth, N. & Kiely, J. L. Impact of palsy rates by birth weight, gestation and severity
conditions among children in multiple populations improved survival of very low birth weight infants on in North of England, 1991–2000 singleton births.
of the United States. J. Dev. Behav. Pediatr. 31, recent secular trends in the prevalence of cerebral Arch. Dis. Child 96, 180–185 (2011).
267–275 (2010). palsy. Pediatrics 91, 1094–1100 (1993). 30. Robertson, C. M. T. et al. Prevalence estimate
7. Pakula, A. T., Van Naarden Braun, K. & 21. Yeargin-Allsopp, M. et al. Prevalence of cerebral palsy of cerebral palsy in Northern Alberta: births,
Yeargin-Allsopp, M. Cerebral palsy: classification and in 8-year-old children in three areas of the United 2008–2010. Can. J. Neurol. Sci. 44, 366–374
epidemiology. Phys. Med. Rehabil. Clin. N. Am. 20, States in 2002: a multisite collaboration. Pediatrics (2017).
425–452 (2009). 121, 547–554 (2008). 31. Hollung, S. J., Vik, T., Lydersen, S., Bakken, I. J. &
8. Delobel-Ayoub, M. et al. Prevalence and 22. Arneson, C. L. et al. Prevalence of cerebral palsy: Andersen, G. L. Decreasing prevalence and severity
characteristics of autism spectrum disorders in Autism and Developmental Disabilities Monitoring of cerebral palsy in Norway among children born
children with cerebral palsy. Dev. Med. Child Neurol. Network, three sites, United States, 2004. 1999 to 2010 concomitant with improvements in
59, 738–742 (2017). Disabil. Health J. 2, 45–48 (2009). perinatal health. Eur. J. Paediatr. Neurol. https://doi.
9. Kancherla, V., Amendah, D. D., Grosse, S. D., 23. Kirby, R. S. et al. Prevalence and functioning of org/10.1016/j.ejpn.2018.05.001 (2018).
Yeargin-Allsopp, M. & Van Naarden Braun, K. Medical children with cerebral palsy in four areas of the United 32. Touyama, M., Touyama, J., Toyokawa, S. & Kobayashi, Y.
expenditures attributable to cerebral palsy and States in 2006: a report from the Autism and Trends in the prevalence of cerebral palsy in children
intellectual disability among Medicaid-enrolled Developmental Disabilities Monitoring Network. born between 1988 and 2007 in Okinawa, Japan.
children. Res. Dev. Disabil. 33, 832–840 (2012). Res. Dev. Disabil. 32, 462–469 (2011). Brain Dev. 38, 792–799 (2016).
10. Incidence of infantile cerebral palsy. Acta Pædiatr. 46, 24. Van Naarden Braun, K. et al. Trends in the prevalence 33. He, P., Chen, G., Wang, Z., Guo, C. & Zheng, X.
12–21 (1957). of autism spectrum disorder, cerebral palsy, hearing Children with motor impairment related to cerebral

Reviews
palsy: Prevalence, severity and concurrent 61. Rouse, D. J. et al. The Maternal-Fetal Medicine Units 86. Greisen, G. & Vannucci, R. C. Is periventricular
impairments in China. J. Paediatr. Child Health 53, cesarean registry: chorioamnionitis at term and its leucomalacia a result of hypoxic-ischaemic injury?
480–484 (2017). duration-relationship to outcomes. Am. J. Obstet. Hypocapnia and the preterm brain. Biol. Neonate 79,
34. Liu, J. M., Li, S., Lin, Q. & Li, Z. Prevalence of cerebral Gynecol. 191, 211–216 (2004). 194–200 (2001).
palsy in China. Int. J. Epidemiol. 28, 949–954 (1999). 62. O’Callaghan, M. et al. Epidemiologic associations with 87. Volpe, J. J. Neurobiology of periventricular
35. Toyokawa, S., Maeda, E. & Kobayashi, Y. Estimation of cerebral palsy. Obstet. Gynecol. 118, 576–582 leukomalacia in the premature infant. Pediatr. Res.
the number of children with cerebral palsy using (2011). 50, 553–562 (2001).
nationwide health insurance claims data in Japan. 63. Vissenberg, R. et al. Abnormal thyroid function 88. Collins, M. P., Lorenz, J. M., Jetton, J. R. & Paneth, N.
Dev. Med. Child Neurol. 59, 317–321 (2017). parameters in the second trimester of pregnancy are Hypocapnia and other ventilation-related risk
36. Park, M. S. et al. Prevalence and lifetime healthcare associated with breech presentation at term: a nested factors for cerebral palsy in low birth weight infants.
cost of cerebral palsy in South Korea. Health Policy cohort study. Eur. J. Obstet. Gynecol. Reprod. Biol. Pediatr. Res. 50, 712–719 (2001).
100, 234–238 (2011). 199, 169–174 (2016). 89. Chawla, S. et al. Association of neurodevelopmental
37. Gladstone, M. A review of the incidence and 64. Macharey, G. et al. Breech presentation at term and outcomes and neonatal morbidities of extremely
prevalence, types and aetiology of childhood cerebral associated obstetric risks factors-a nationwide premature infants with differential exposure to
palsy in resource-poor settings. Ann. Trop. Paediatr. population based cohort study. Arch. Gynecol. Obstet. antenatal steroids. JAMA Pediatr. 170, 1164–1172
30, 181–196 (2010). 295, 833–838 (2017). (2016).
38. Chang, M. J., Ma, H. I. & Lu, T. H. Estimating the 65. Hofmeyr, G. J. & Hannah, M. E. Planned caesarean 90. Halliday, H. L., Ehrenkranz, R. A. & Doyle, L. W. Late
prevalence of cerebral palsy in Taiwan: a comparison section for term breech delivery. Cochrane Database (>7 days) postnatal corticosteroids for chronic lung
of different case definitions. Res. Dev. Disabil. 36C, Syst. Rev. 7, CD000166 (2003). disease in preterm infants. Cochrane Database
207–212 (2014). 66. Himmelmann, K., Hagberg, G. & Uvebrant, P. Syst. Rev. 10, CD001145 (2009).
39. El-Tallawy, H. N. et al. Cerebral palsy in Al-Quseir City, The changing panorama of cerebral palsy in Sweden. 91. Linsell, L., Malouf, R., Morris, J., Kurinczuk, J. J. &
Egypt: prevalence, subtypes, and risk factors. X. Prevalence and origin in the birth-year period Marlow, N. Prognostic factors for cerebral palsy
Neuropsychiatr. Dis. Treat. 10, 1267–1272 (2014). 1999–2002. Acta Paediatr. 99, 1337–1343 and motor impairment in children born very
40. Kakooza-Mwesige, A. et al. Prevalence of cerebral (2010). preterm or very low birthweight: a systematic review.
palsy in Uganda: a population-based study. 67. Schieve, L. A. et al. Population impact of preterm birth Dev. Med. Child Neurol. 58, 554–569 (2016).
Lancet. Global Health 5, e1275–e1282 (2017). and low birth weight on developmental disabilities in 92. Kuban, K. C. et al. Systemic inflammation and
41. Couper, J. Prevalence of childhood disability in US children. Ann. Epidemiol. 26, 267–274 (2016). cerebral palsy risk in extremely preterm infants.
rural KwaZulu-Natal. S. Afr. Med. J. 92, 549–552 68. Dammann, O. & Leviton, A. Perinatal brain damage J. Child Neurol. 29, 1692–1698 (2014).
(2002). causation. Dev. Neurosci. 29, 280–288 (2007). 93. Kuban, K. C. et al. The breadth and type of systemic
42. Khandaker, G. et al. Bangladesh Cerebral Palsy 69. Tran, U., Gray, P. H. & O’Callaghan, M. J. Neonatal inflammation and the risk of adverse neurological
Register (BCPR): a pilot study to develop a national antecedents for cerebral palsy in extremely preterm outcomes in extremely low gestation newborns.
cerebral palsy (CP) register with surveillance of babies and interaction with maternal factors. Pediatr. Neurol. 52, 42–48 (2015).
children for CP. BMC Neurol. 15, 173 (2015). Early Hum. Dev. 81, 555–561 (2005). 94. Carlo, W. A. et al. Cytokines and neurodevelopmental
43. Delgado, M. R. The Mexican Academy for Cerebral 70. Pinto-Martin, J. A. et al. Cranial ultrasound prediction outcomes in extremely low birth weight infants.
Palsy and Neurodevelopmental Disorders: new kid on of disabling and nondisabling cerebral palsy at age J. Pediatr. 159, 919–925.e3 (2011).
the block. Dev. Med. Child Neurol. 58, 109 (2016). two in a low birth weight population. Pediatrics 95, 95. La Gamma, E. F., Korzeniewski, S. J., Ballabh, P. &
44. Almasri, N. A., Saleh, M., Abu-Dahab, S., Malkawi, S. H. 249–254 (1995). Paneth, N. Transient Hypothyroxinemia of
& Nordmark, E. Development of a cerebral palsy 71. Skovgaard, A. L. & Zachariassen, G. Cranial ultrasound Prematurity. NeoReviews 17, e394–e402 (2016).
follow-up registry in Jordan (CPUP-Jordan). findings in preterm infants predict the development of 96. Shih, J. L. & Agus, M. S. Thyroid function in the
Child Care Health Dev. 44, 131–139 (2018). cerebral palsy. Dan. Med. J. 64, A5330 (2017). critically ill newborn and child. Curr. Opin. Pediatr. 21,
45. Khandaker, G. et al. Protocol for hospital 72. Reubsaet, P. et al. The impact of low-grade germinal 536–540 (2009).
based-surveillance of cerebral palsy (CP) in Hanoi matrix-intraventricular hemorrhage on neuro 97. Kaluarachchi, D. C., Colaizy, T. T., Pesce, L. M.,
using the Paediatric Active Enhanced Disease developmental outcome of very preterm infants. Tansey, M. & Klein, J. M. Congenital hypothyroidism
Surveillance mechanism (PAEDS-Vietnam): a study Neonatology 112, 203–210 (2017). with delayed thyroid-stimulating hormone elevation in
towards developing hospital-based disease 73. Radic, J. A., Vincer, M. & McNeely, P. D. Outcomes of premature infants born at less than 30 weeks
surveillance in Vietnam. BMJ Open 7, e017742 intraventricular hemorrhage and posthemorrhagic gestation. J. Perinatol. 37, 277–282 (2017).
(2017). hydrocephalus in a population-based cohort of very 98. Cavarzere, P. et al. Congenital hypothyroidism
46. Little, W. J. The classic: hospital for the cure of preterm infants born to residents of Nova Scotia from with delayed TSH elevation in low-birth-weight
deformities: course of lectures on the deformities of 1993 to 2010. J. Neurosurg. Pediatr. 15, 580–588 infants: incidence, diagnosis and management.
the human frame. 1843. Clin. Orthop. Relat. Res. (2015). Eur. J. Endocrinol. 175, 395–402 (2016).
470, 1252–1256 (2012). 74. Mukerji, A., Shah, V. & Shah, P. S. Periventricular/ 99. Silva, M. H., Araujo, M. C., Diniz, E. M., Ceccon, M. E.
47. Little, W. J. in Transactions of the Obstetrical Society intraventricular hemorrhage and neurodevelopmental & Carvalho, W. B. Nonthyroidal illnesses syndrome in
of London. 293–344 (Longmans, Green and Co, outcomes: a meta-analysis. Pediatrics 136, full-term newborns with sepsis. Arch. Endocrinol.
1862). 1132–1143 (2015). Metab. 59, 528–534 (2015).
48. Freud, S. Die Infantilen Cerebrallahmungen 75. Sola-Visner, M. Prognostic significance of low-grade 100. Marks, S. D. Nonthyroidal illness syndrome in
(KuK Hoflieferant, 1897). intraventricular hemorrhage in the current era of children. Endocrine 36, 355–367 (2009).
49. Evans, P. R. Antecedents of infantile cerebral palsy. neonatology. JAMA Pediatr, 1–2 (2013). 101. De Groot, L. J. Dangerous dogmas in medicine:
Arch. Dis. Child 23, 213–219 (1948). 76. Bolisetty, S. et al. Intraventricular hemorrhage and the nonthyroidal illness syndrome. J. Clin.
50. Freud, S. Infantile cerebral paralysis (Univ. of Miami neurodevelopmental outcomes in extreme preterm Endocrinol. Metab. 84, 151–164 (1999).
Press, 1968). infants. Pediatrics 133, 55–62 (2014). 102. McIver, B. & Gorman, C. A. Euthyroid sick syndrome:
51. Dan, B., Rosenbloom, L. & Paneth, N. Cerebral palsy 77. Tsai, W. H. et al. Association between mechanical an overview. Thyroid 7, 125–132 (1997).
(Mac Keith Press, 2014). ventilation and neurodevelopmental disorders in a 103. Chowdhry, P., Auerbach, R., Scanlon, J. & Abbassi, V.
52. Yannet, H. The etiology of congenital cerebral palsy. nationwide cohort of extremely low birth weight The nature of hypothyroxinemia in sick preterm
J. Pediatr. 24, 38–45 (1944). infants. Res. Dev. Disabil. 35, 1544–1550 (2014). infants. Pediatr. Res. 15, 505–505 (1981).
53. Nelson, K. B. & Ellenberg, J. H. Antecedents of 78. Bosco, J. L. et al. A most stubborn bias: 104. Diamond, F. B., Parks, J. S., Tenore, A., Marino, J. M.
cerebral palsy. Multivariate analysis of risk. no adjustment method fully resolves confounding by & Bongiovanni, A. M. Hypothyroxinemia in sick and
N. Engl. J. Med. 315, 81–86 (1986). indication in observational studies. J. Clin. Epidemiol. well preterm infants. Clin. Pediatr. 18, 559–561
54. Paneth, N. & Fox, H. E. The relationship of Apgar 63, 64–74 (2010). (1979).
score to neurologic handicap: a survey of clinicians. 79. Murase, M. & Ishida, A. Early hypocarbia of preterm 105. Cuestas, R. A., Lindall, A. & Engel, R. R. Low thyroid
Obstet. Gynecol. 61, 547–550 (1983). infants: its relationship to periventricular leukomalacia hormones and respiratory-distress syndrome of the
55. Paneth, N. & Stark, R. I. Cerebral palsy and mental and cerebral palsy, and its perinatal risk factors. newborn. Studies on cord blood. N. Engl. J. Med. 295,
retardation in relation to indicators of perinatal Acta Paediatr. 94, 85–91 (2005). 297–302 (1976).
asphyxia. An epidemiologic overview. Am. J. Obstet. 80. Graziani, L. J. et al. Mechanical ventilation in preterm 106. Ares, S., Quero, J., Diez, J. & Morreale de Escobar, G.
Gynecol. 147, 960–966 (1983). infants: neurosonographic and developmental studies. Neurodevelopment of preterm infants born at 28 to
56. Badawi, N. et al. Cerebral palsy following term Pediatrics 90, 515–522 (1992). 36 weeks of gestational age: the role of hypothy
newborn encephalopathy: a population-based study. 81. Leviton, A. et al. Early blood gas abnormalities and roxinemia and long-term outcome at 4 years.
Dev. Med. Child Neurol. 47, 293–298 (2005). the preterm brain. Am. J. Epidemiol. 172, 907–916 J. Pediatr. Endocrinol. Metab. 24, 897–902
57. Badawi, N. et al. Intrapartum risk factors for newborn (2010). (2011).
encephalopathy: the Western Australian case-control 82. Okumura, A. et al. Hypocarbia in preterm infants with 107. Hollanders, J. J. et al. Transient hypothyroxinemia
study. BMJ 317, 1554–1558 (1998). periventricular leukomalacia: the relation between of prematurity and problem behavior in young
58. Ellenberg, J. H. & Nelson, K. B. The association of hypocarbia and mechanical ventilation. Pediatrics adulthood. Psychoneuroendocrinology 72, 40–46
cerebral palsy with birth asphyxia: a definitional 107, 469–475 (2001). (2016).
quagmire. Dev. Med. Child Neurol. 55, 210–216 83. Wiswell, T. E. et al. High-frequency jet ventilation in the 108. Den Ouden, A. L., Kok, J. H., Verkerk, P. H., Brand, R.
(2013). early management of respiratory distress syndrome is & Verloove-Vanhorick, S. P. The relation between
59. Nelson, K. B., Sartwelle, T. P. & Rouse, D. J. Electronic associated with a greater risk for adverse outcomes. neonatal thyroxine levels and neurodevelopmental
fetal monitoring, cerebral palsy, and caesarean Pediatrics 98, 1035–1043 (1996). outcome at age 5 and 9 years in a national cohort of
section: assumptions versus evidence. BMJ 355, 84. Giannakopoulou, C. et al. Significance of hypocarbia in very preterm and/or very low birth weight infants.
i6405 (2016). the development of periventricular leukomalacia in Pediatr. Res. 39, 142–145 (1996).
60. Salafia, C. M., Mangam, H. E., Weigl, C. A., Foye, G. J. preterm infants. Pediatr. Int. 46, 268–273 (2004). 109. Meijer, W. J., Verloovevanhorick, S. P., Brand, R. &
& Silberman, L. Abnormal fetal heart rate patterns 85. Greisen, G. & Borch, K. White matter injury in the Vandenbrande, J. L. Transient hypothyroxinemia
and placental inflammation. Am. J. Obstet. Gynecol. preterm neonate: the role of perfusion. Dev. Neurosci. associated with developmental delay in very preterm
160, 140–147 (1989). 23, 209–212 (2001). infants. Arch. Dis. Childhood 67, 944–947 (1992).

Reviews
110. Reuss, M. L., Paneth, N., Pinto-Martin, J. A., 134. Vohr, B. R. et al. Maternal age, multiple birth, and socio-economic score and integration score:
Lorenz, J. M. & Susser, M. The relation of transient extremely low birth weight infants. J. Pediatr. 154, a population-based study. Food Nutr. Res. 57,
hypothyroxinemia in preterm infants to neurologic 498–503.e2 (2009). 21164 (2013).
development at two years of age. N. Engl. J. Med. 135. Luu, T. M. & Vohr, B. Twinning on the brain: the effect 161. Mohsena, M., Goto, R. & Mascie-Taylor, C. N.
334, 821–827 (1996). on neurodevelopmental outcomes. Am. J. Med. Genet Maternal nutritional status (as measured by height,
111. Hollanders, J. J. et al. No association between C Semin. Med. Genet. 151C, 142–147 (2009). weight and BMI) in Bangladesh: trends and
transient hypothyroxinemia of prematurity and 136. Yokoyama, Y., Shimizu, T. & Hayakawa, K. Prevalence socio-economic association over the period 1996 to
neurodevelopmental outcome in young adulthood. of cerebral palsy in twins, triplets and quadruplets. 2007. Publ. Health Nutr. 19, 1438–1445 (2016).
J. Clin. Endocrinol. Metab. 100, 4648–4653 (2015). Int. J. Epidemiol. 24, 943–948 (1995). 162. Crisham Janik, M. D. et al. Maternal diagnosis
112. Trumpff, C. et al. Thyroid-stimulating hormone (TSH) 137. Pharoah, P. O. & Cooke, T. Cerebral palsy and of obesity and risk of cerebral palsy in the child.
concentration at birth in belgian neonates and multiple births. Arch. Dis. Child Fetal Neonatal Ed. 75, J. Pediatr. 163, 1307–1312 (2013).
cognitive development at preschool age. Nutrients 7, F174–F177 (1996). 163. Pan, C., Deroche, C. B., Mann, J. R., McDermott, S. &
9018–9032 (2015). 138. Topp, M. et al. Multiple birth and cerebral palsy in Hardin, J. W. Is prepregnancy obesity associated
113. DenOuden, A. L., Kok, J. H., Verkerk, P. H., Brand, R. Europe: a multicenter study. Acta Obstet. Gynecol. with risk of cerebral palsy and epilepsy in children?
& VerlooveVanhorick, S. P. The relation between Scand. 83, 548–553 (2004). J. Child Neurol. 29, NP196–NP201 (2014).
neonatal thyroxine levels and neurodevelopmental 139. Maenner, M. J. et al. Children with cerebral palsy: 164. Forthun, I. et al. Maternal prepregnancy BMI and risk
outcome at age 5 and 9 years in a national cohort of racial disparities in functional limitations. of cerebral palsy in offspring. Pediatrics 138,
very preterm and/or very low birth weight infants. Epidemiology 23, 35–43 (2012). e20160874 (2016).
Pediatr. Res. 39, 142–145 (1996). 140. Wu, Y. W. et al. Racial, ethnic, and socioeconomic 165. Zhu, M. J., Du, M. & Ford, S. P. Cell Biology
114. Leviton, A. et al. Hypothyroxinemia of prematurity and disparities in the prevalence of cerebral palsy. Symposium: impacts of maternal obesity on placental
the risk of cerebral white matter damage. J. Pediatr. Pediatrics 127, e674–e681 (2011). and gut inflammation and health. J. Anim. Sci. 92,
134, 706–711 (1999). 141. Petterson, B., Nelson, K. B., Watson, L. & Stanley, F. 1840–1849 (2014).
115. La Gamma, E. F. et al. Phase 1 trial of 4 thyroid Twins, triplets, and cerebral palsy in births in Western 166. Huang, L. et al. Maternal prepregnancy obesity is
hormone regimens for transient hypothyroxinemia in Australia in the 1980s. BMJ 307, 1239–1243 associated with higher risk of placental pathological
neonates of <28 weeks’ gestation. Pediatrics 124, (1993). lesions. Placenta 35, 563–569 (2014).
e258–268 (2009). 142. Lorenz, J. M. Neurodevelopmental outcomes of twins. 167. Bronson, S. L. & Bale, T. L. The placenta as a mediator
116. Hong, T. & Paneth, N. Maternal and infant thyroid Semin. Perinatol 36, 201–212 (2012). of stress effects on neurodevelopmental
disorders and cerebral palsy. Semin. Perinatol. 32, 143. Scher, A. I. et al. The risk of mortality or cerebral palsy reprogramming. Neuropsychopharmacology 41,
438–445 (2008). in twins: a collaborative population-based study. 207–218 (2016).
117. Rogan, W. J. et al. Iodine deficiency, pollutant Pediatr. Res. 52, 671–681 (2002). 168. Myatt, L. & Maloyan, A. Obesity and placental
chemicals, and the thyroid: new information on 144. Petterson, B., Stanley, F. & Henderson, D. Cerebral function. Semin. Reprod. Med. 34, 42–49 (2016).
an old problem. Pediatrics 133, 1163–1166 palsy in multiple births in Western Australia: genetic 169. Mannisto, T. et al. Early pregnancy reference intervals
(2014). aspects. Am. J. Med. Genet. 37, 346–351 (1990). of thyroid hormone concentrations in a thyroid
118. O’Callaghan, M. E. et al. Epidemiologic associations 145. Steingass, K. J., Taylor, H. G., Wilson-Costello, D., antibody-negative pregnant population. Thyroid 21,
with cerebral palsy. Obstet. Gynecol. 118, 576–582 Minich, N. & Hack, M. Discordance in neonatal risk 291–298 (2011).
(2011). factors and early childhood outcomes of very low birth 170. Mosso, L. et al. Early pregnancy thyroid hormone
119. Hemminki, K., Li, X., Sundquist, K. & Sundquist, J. weight (<1.5 kg) twins. J. Perinatol. 33, 388–393 reference ranges in Chilean women: the influence of
High familial risks for cerebral palsy implicate partial (2013). body mass index. Clin. Endocrinol. 85, 942–948
heritable aetiology. Paediatr. Perinat. Epidemiol. 21, 146. Hack, K. E. et al. Long-term neurodevelopmental (2016).
235–241 (2007). outcome of monochorionic and matched dichorionic 171. van der Burg, J. W. et al. Are extremely low
120. Tollanes, M. C., Wilcox, A. J., Lie, R. T. & Moster, D. twins. PLOS One 4, e6815 (2009). gestational age newborns born to obese women
Familial risk of cerebral palsy: population based 147. Swamy, R. S. et al. Cognitive outcome in childhood at increased risk of cerebral palsy at 2 years?
cohort study. BMJ 349, g4294 (2014). of birth weight discordant monochorionic twins: J. Child Neurol. 33, 216–224 (2018).
121. Fahey, M. C., Maclennan, A. H., Kretzschmar, D., the long-term effects of fetal growth restriction. 172. Ozturk, A. et al. Antenatal and delivery risk factors
Gecz, J. & Kruer, M. C. The genetic basis of cerebral Arch. Dis. Child Fetal Neonatal Ed. https://doi.org/ and prevalence of cerebral palsy in Duzce (Turkey).
palsy. Dev. Med. Child Neurol. 59, 462–469 (2017). 10.1136/archdischild-2017-313691 (2018). Brain Dev. 29, 39–42 (2007).
122. McMichael, G. et al. Whole-exome sequencing points 148. Halling, C. & Corcoran, J. D. Little and large: 173. Thorngren-Jerneck, K. & Herbst, A. Perinatal factors
to considerable genetic heterogeneity of cerebral the effects of twin growth discordance. Arch. Dis. associated with cerebral palsy in children born in
palsy. Mol. Psychiatry 20, 176–182 (2015). Child Fetal Neonatal Ed. https://doi.org/10.1136/ Sweden. Obstet. Gynecol. 108, 1499–1505 (2006).
123. Meirelles Kalil Pessoa de, B., Rodrigues, C. J., archdischild-2017-314572 (2018). 174. Zhao, M., Dai, H., Deng, Y. & Zhao, L. SGA as a risk
de Barros, T. E. & Bevilacqua, R. G. Presence of 149. Pharoah, P. O. & Dundar, Y. Monozygotic factor for cerebral palsy in moderate to late preterm
apolipoprotein E ε4 allele in cerebral palsy. twinning, cerebral palsy and congenital anomalies. infants: a system review and meta-analysis. Sci. Rep.
J. Pediatr. Orthop. 20, 786–789 (2000). Hum. Reprod. Update 15, 639–648 (2009). 6, 38853 (2016).
124. Kuroda, M. M., Weck, M. E., Sarwark, J. F., 150. Glinianaia, S. V., Rankin, J. & Wright, C. Congenital 175. Blair, E., Watson, L. & Australian Cerebral Palsy
Hamidullah, A. & Wainwright, M. S. Association of anomalies in twins: a register-based study. Register Group. Cerebral palsy and perinatal mortality
apolipoprotein E genotype and cerebral palsy in Hum. Reprod. 23, 1306–1311 (2008). after pregnancy-induced hypertension across the
children. Pediatrics 119, 306–313 (2007). 151. Pharoah, P. O. & Cooke, R. W. A hypothesis for the gestational age spectrum: observations of a
125. Wu, Y. W., Croen, L. A., Vanderwerf, A., Gelfand, A. A. aetiology of spastic cerebral palsy — the vanishing reconstructed total population cohort. Dev. Med.
& Torres, A. R. Candidate genes and risk for CP: twin. Dev. Med. Child Neurol. 39, 292–296 (1997). Child Neurol. 58 (Suppl. 2), 76–81 (2016).
a population-based study. Pediatr. Res. 70, 642–646 152. Grether, J. K., Nelson, K. B. & Cummins, S. K. Twinning 176. Jarvis, S. et al. Cerebral palsy and intrauterine growth
(2011). and cerebral palsy: experience in four northern in single births: European collaborative study. Lancet
126. Lien, E. et al. Apolipoprotein E polymorphisms and California counties, births 1983 through 1985. 362, 1106–1111 (2003).
severity of cerebral palsy: a cross-sectional study in Pediatrics 92, 854–858 (1993). 177. Grether, J. K., Nelson, K. B., Emery, E. S. 3rd &
255 children in Norway. Dev. Med. Child Neurol. 55, 153. Solaski, M., Majnemer, A. & Oskoui, M. Contribution Cummins, S. K. Prenatal and perinatal factors
372–377 (2013). of socio-economic status on the prevalence of cerebral and cerebral palsy in very low birth weight infants.
127. Lien, E. et al. Genes determining the severity of palsy: a systematic search and review. Dev. Med. Child J. Pediatr. 128, 407–414 (1996).
cerebral palsy: the role of single nucleotide Neurol. 56, 1043–1051 (2014). 178. Xiong, X., Saunders, L. D., Wang, F. L., Davidge, S. T. &
polymorphisms on the amount and structure of 154. Oskoui, M., Messerlian, C., Blair, A., Gamache, P. & Buekens, P. Preeclampsia and cerebral palsy in
apolipoprotein E. Acta Paediatr. 104, 701–706 Shevell, M. Variation in cerebral palsy profile by low-birth-weight and preterm infants: implications for
(2015). socio-economic status. Dev. Med. Child Neurol. 58, the current “ischemic model” of preeclampsia.
128. Braga, L. W. et al. Apolipoprotein E genotype and 160–166 (2016). Hypertens. Pregnancy 20, 1–13 (2001).
cerebral palsy. Dev. Med. Child Neurol. 52, 666–671 155. Dolk, H. et al. Socio-economic inequalities in 179. Arnold, C. C., Kramer, M. S., Hobbs, C. A., Mclean, F. H.
(2010). cerebral palsy prevalence in the United Kingdom: & Usher, R. H. Very low birth weight – a problematic
129. Stoknes, M. et al. Child apolipoprotein E gene variants a register-based study. Paediatr. Perinat. Epidemiol. cohort for epidemiologic studies of very small or
and risk of cerebral palsy: estimation from case-parent 24, 149–155 (2010). immature neonates. Am. J. Epidemiol. 134, 604–613
triads. Eur. J. Paediatr. Neurol. 19, 286–291 (2015). 156. Hjern, A. & Thorngren-Jerneck, K. Perinatal (1991).
130. McMichael, G. L. et al. Association between complications and socio-economic differences in 180. Blair, E. M. & Nelson, K. B. Fetal growth restriction
Apolipoprotein E genotype and cerebral palsy is not cerebral palsy in Sweden – a national cohort study. and risk of cerebral palsy in singletons born after at
confirmed in a Caucasian population. Hum. Genet. BMC Pediatr. 8, 49 (2008). least 35 weeks’ gestation. Am. J. Obstet. Gynecol.
124, 411–416 (2008). 157. Tseng, S. H., Lee, J. Y., Chou, Y. L., Sheu, M. L. & 212, 520.e1–520.e7 (2015).
131. O’Callaghan, M. E. et al. Fetal and maternal candidate Lee, Y. W. Association between socioeconomic status 181. Blair, E. The undesirable consequences of controlling
single nucleotide polymorphism associations with and cerebral palsy. PLOS One 13, e0191724 (2018). for birth weight in perinatal epidemiological studies.
cerebral palsy: a case-control study. Pediatrics 129, 158. Heslehurst, N. et al. Trends in maternal obesity J. Epidemiol. Commun. Health 50, 559–563 (1996).
e414–e423 (2012). incidence rates, demographic predictors, and health 182. Kramer, M. S. Gestational age, birthweight, and their
132. Xu, Y. et al. The association of apolipoprotein E gene inequalities in 36,821 women over a 15-year period. influence on neonatal outcome. Acta Paediatr. 104,
polymorphisms with cerebral palsy in Chinese infants. BJOG 114, 187–194 (2007). 5–6 (2015).
Mol. Genet. Genomics 289, 411–416 (2014). 159. Micklesfield, L. K. et al. Socio-cultural, environmental 183. O’Shea, T. M. et al. The ELGAN study of the brain and
133. O’Callaghan, M. E., MacLennan, A. H., Haan, E. A., and behavioural determinants of obesity in black related disorders in extremely low gestational age
Dekker, G. & South Australian Cerebral Palsy Research South African women. Cardiovascular J. Afr. 24, newborns. Early Hum. Dev. 85, 719–725 (2009).
Group. The genomic basis of cerebral palsy: a HuGE 369–375 (2013). 184. Tsimis, M. E. et al. Risk factors for periventricular
systematic literature review. Hum. Genet. 126, 160. Sommer, C. et al. Ethnic differences in maternal white matter injury in very low birthweight neonates.
149–172 (2009). dietary patterns are largely explained by Am. J. Obstet. Gynecol. 214, 380.e1–380.e6 (2016).

Reviews
185. Leviton, A. et al. Microbiologic and histologic with proven congenital cytomegalovirus infection. countries: evidence from a scoping review.
characteristics of the extremely preterm infant’s Coll. Antropol. 35 (Suppl. 1), 229–234 (2011). World J. Pediatrics 11, 293–299 (2015).
placenta predict white matter damage and later 211. Edebol-Tysk, K. Epidemiology of spastic tetraplegic 238. Kuban, K. C. et al. Maternal toxemia is associated with
cerebral palsy. the ELGAN study. Pediatr. Res. 67, cerebral palsy in Sweden. I. Impairments and reduced incidence of germinal matrix hemorrhage in
95–101 (2010). disabilities. Neuropediatrics 20, 41–45 (1989). premature babies. J. Child Neurol. 7, 70–76 (1992).
186. Dammann, O. Persistent neuro-inflammation in 212. Contreras-Capetillo, S. N. et al. Birth defects 239. Nelson, K. B. & Grether, J. K. Can magnesium sulfate
cerebral palsy: a therapeutic window of opportunity? associated with congenital zika virus infection in reduce the risk of cerebral palsy in very low
Acta Paediatr. 96, 6–7 (2007). mexico. Clin. Pediatr. 57, 927–936 (2017). birthweight infants. Pediatrics 95, 263–269 (1995).
187. Dammann, O. Causality, mosaics, and the health 213. Zheng, X. Y. et al. Intrauterine infections and birth 240. Schendel, D. E., Berg, C. J., Yeargin-Allsopp, M.,
sciences. Theor. Med. Bioeth. 37, 161–168 (2016). defects. Biomed. Environ. Sci. 17, 476–491 (2004). Boyle, C. A. & Decoufle, P. Prenatal magnesium sulfate
188. Dammann, O. & Leviton, A. Inflammatory brain 214. McIntyre, S., Blair, E., Badawi, N., Keogh, J. & exposure and the risk for cerebral palsy or mental
damage in preterm newborns—dry numbers, wet lab, Nelson, K. B. Antecedents of cerebral palsy and retardation among very low-birth-weight children aged
and causal inferences. Early Hum. Dev. 79, 1–15 perinatal death in term and late preterm singletons. 3 to 5 years. JAMA 276, 1805–1810 (1996).
(2004). Obstet. Gynecol. 122, 869–877 (2013). 241. Leviton, A. et al. Maternal receipt of magnesium
189. Schendel, D. E. Infection in pregnancy and cerebral 215. Pharoah, P. O. Prevalence and pathogenesis of sulfate does not seem to reduce the risk of neonatal
palsy. Journal of the American Medical Women’s congenital anomalies in cerebral palsy. Arch. Dis. Child white matter damage. Pediatrics 99, E2 (1997).
Association 56, 105–108 (2001) (1972). Fetal Neonatal Ed. 92, F489–F493 (2007). 242. Paneth, N., Jetton, J., Pinto-Martin, J. & Susser, M.
190. Smithers-Sheedy, H. et al. Congenital cytomegalovirus 216. Rankin, J. et al. Congenital anomalies in children with Magnesium sulfate in labor and risk of neonatal brain
among children with cerebral palsy. J. Pediatr. 181, cerebral palsy: a population-based record linkage lesions and cerebral palsy in low birth weight infants.
267–271 (2017). study. Dev. Med. Child Neurol. 52, 345–351 (2010). The Neonatal Brain Hemorrhage Study Analysis
191. Smithers-Sheedy, H. et al. Congenital cytomegalovirus 217. Nattel, S. N. et al. Congenital heart disease and Group. Pediatrics 99, E1 (1997).
is associated with severe forms of cerebral palsy and neurodevelopment: clinical manifestations, genetics, 243. Schlapbach, L. J. et al. Impact of chorioamnionitis
female sex in a retrospective population-based study. mechanisms, and implications. Can. J. Cardiol. 33, and preeclampsia on neurodevelopmental outcome
Dev. Med. Child Neurol. 56, 846–852 (2014). 1543–1555 (2017). in preterm infants below 32 weeks gestational age.
192. Gerardin, P. et al. Neurocognitive outcome of children 218. Bellizzi, S. et al. Are hypertensive disorders in Acta Paediatr. 99, 1504–1509 (2010).
exposed to perinatal mother-to-child Chikungunya pregnancy associated with congenital malformations 244. Gray, P. H., Jones, P. & O’Callaghan, M. J. Maternal
virus infection: the CHIMERE cohort study on Reunion in offspring? Evidence from the WHO Multicountry antecedents for cerebral palsy in extremely preterm
Island. PLoS Negl. Trop. Dis. 8, e2996 (2014). cross sectional survey on maternal and newborn babies: a case-control study. Dev. Med. Child Neurol.
193. Meneses, J. D. A. et al. Lessons learned at the health. BMC Pregnancy Childbirth 16, 198 (2016). 43, 580–585 (2001).
epicenter of Brazil’s congenital zika epidemic: 219. Howley, M. M. et al. Thyroid medication use and birth 245. Murphy, D. J., Sellers, S., Mackenzie, I. Z., Yudkin, P. L.
evidence from 87 confirmed cases. Clin. Infect. Dis. defects in the National Birth Defects Prevention Study. & Johnson, A. M. Case-control study of antenatal and
64, 1302–1308 (2017). Birth Defects Res. 109, 1471–1481 (2017). intrapartum risk factors for cerebral palsy in very
194. Pessoa, A. et al. Motor abnormalities and epilepsy in 220. Hashmi, S. S. et al. The association between neonatal preterm singleton babies. Lancet 346, 1449–1454
infants and children with evidence of congenital Zika thyroxine and craniosynostosis, Texas, 2004–2007. (1995).
virus infection. Pediatrics 141, S167–S179 (2018). Birth Defects Res. A Clin. Mol. Teratol. 94, 246. Rouse, D. J. & Hirtz, D. What we learned about the
195. Streja, E. et al. Congenital cerebral palsy and prenatal 1004–1009 (2012). role of antenatal magnesium sulfate for the prevention
exposure to self-reported maternal infections, fever, or 221. Stoll, C., Dott, B., Alembik, Y. & Koehl, C. Congenital of cerebral palsy. Semin. Perinatol. 40, 303–306
smoking. Am. J. Obstet. Gynecol. 209, 332.e1–332. anomalies associated with congenital hypothyroidism. (2016).
e10 (2013). Ann. Genet. 42, 17–20 (1999). 247. Rouse, D. J. et al. A randomized, controlled trial of
196. Abdullahi, H., Satti, M., Rayis, D. A., Imam, A. M. & 222. Pei, L. & Wallace, D. C. Mitochondrial etiology of magnesium sulfate for the prevention of cerebral
Adam, I. Intra-partum fever and cerebral palsy in neuropsychiatric disorders. Biol. Psychiatry 83, palsy. N. Engl. J. Med. 359, 895–905 (2008).
Khartoum, Sudan. BMC Res. Notes 6, 163 (2013). 722–730 (2017). 248. Zeng, X., Xue, Y., Tian, Q., Sun, R. & An, R. Effects
197. Wu, Y. W. & Colford, J. M. Chorioamnionitis as a risk 223. Wu, Y. W. et al. Perinatal stroke in children with motor and safety of magnesium sulfate on neuroprotection:
factor for cerebral palsy – a meta-analysis. JAMA 284, impairment: a population-based study. Pediatrics 114, a meta-analysis based on PRISMA Guidelines.
1417–1424 (2000). 612–619 (2004). Medicine 95, e2451 (2016).
198. Meeraus, W. H., Petersen, I. & Gilbert, R. Association 224. Uvebrant, P. Hemiplegic cerebral palsy. Aetiology and 249. Conde-Agudelo, A. & Romero, R. Antenatal
between antibiotic prescribing in pregnancy and outcome. Acta Paediatr. Scand. Suppl. 345, 1–100 magnesium sulfate for the prevention of cerebral
cerebral palsy or epilepsy in children born at term: (1988). palsy in preterm infants less than 34 weeks’ gestation:
a cohort study using the health improvement network. 225. Wu, Y. W., Croen, L. A., Shah, S. J., Newman, T. B. & a systematic review and metaanalysis. Am. J. Obstet.
PLOS One 10, e0122034 (2015). Najjar, D. V. Cerebral palsy in a term population: Gynecol. 200, 595–609 (2009).
199. Wu, Y. W. et al. Chorioamnionitis and cerebral palsy in risk factors and neuroimaging findings. Pediatrics 118, 250. Rouse, D. J. & Gibbins, K. J. Magnesium sulfate for
term and near-term infants. JAMA 290, 2677–2684 690–697 (2006). cerebral palsy prevention. Semin. Perinatol. 37,
(2003). 226. Kirton, A., Deveber, G., Pontigon, A. M., Macgregor, D. 414–416 (2013).
200. Wu, Y. W. Systematic review of chorioamnionitis and & Shroff, M. Presumed perinatal ischemic stroke: 251. Wolf, H. T., Hegaard, H. K., Greisen, G., Huusom, L. &
cerebral palsy. Mental Retard. Dev. Disabilities Res. vascular classification predicts outcomes. Ann. Neurol. Hedegaard, M. Treatment with magnesium sulphate in
Rev. 8, 25–29 (2002). 63, 436–443 (2008). pre-term birth: a systematic review and meta-analysis
201. Bear, J. J. & Wu, Y. W. Maternal infections during 227. Cole, L. et al. Clinical characteristics, risk factors, of observational studies. J. Obstetr. Gynaecol. 32,
pregnancy and cerebral palsy in the child. Pediatr. and outcomes associated with neonatal hemorrhagic 135–140 (2012).
Neurol. 57, 74–79 (2016). stroke: a population-based case-control study. 252. De Silva, D. A. et al. MAGnesium sulphate for fetal
202. Miller, J. E. et al. Maternal infections during JAMA Pediatr. 171, 230–238 (2017). neuroprotection to prevent Cerebral Palsy (MAG-CP)-
pregnancy and cerebral palsy: a population-based 228. Kirton, A. et al. Symptomatic neonatal arterial implementation of a national guideline in Canada.
cohort study. Paediatr. Perinat. Epidemiol. 27, ischemic stroke: the International Pediatric Stroke Implement. Sci. 13, 8 (2018).
542–552 (2013). Study. Pediatrics 128, e1402–1410 (2011). 253. Turitz, A. L., Too, G. T. & Gyamfi-Bannerman, C.
203. Polivka, B. J., Nickel, J. T. & Wilkins, J. R. 3rd. Urinary 229. Fitzgerald, K. C. & Golomb, M. R. Neonatal arterial Proximity of magnesium exposure to delivery and
tract infection during pregnancy: a risk factor for ischemic stroke and sinovenous thrombosis associated neonatal outcomes. Am. J. Obstet. Gynecol. 215,
cerebral palsy? J. Obstet. Gynecol. Neonatal Nurs. with meningitis. J. Child Neurol. 22, 818–822 (2007). 508.e1–508.e6 (2016).
26, 405–413 (1997). 230. Lee, J. et al. Maternal and infant characteristics 254. Ohhashi, M. et al. Magnesium sulphate and perinatal
204. O’Shea, T. M., Klinepeter, K. L., Meis, P. J. & associated with perinatal arterial stroke in the infant. mortality and morbidity in very-low-birthweight infants
Dillard, R. G. Intrauterine infection and the risk of JAMA 293, 723–729 (2005). born between 24 and 32 weeks of gestation in Japan.
cerebral palsy in very low-birthweight infants. 231. Kraus, F. T. & Acheen, V. I. Fetal thrombotic Eur. J. Obstet. Gynecol. Reprod. Biol. 201, 140–145
Paediatr. Perinat. Epidemiol. 12, 72–83 (1998). vasculopathy in the placenta: cerebral thrombi (2016).
205. Brookfield, K. F., Osmundson, S. S., Caughey, A. B. & and infarcts, coagulopathies, and cerebral palsy. 255. McPherson, J. A., Rouse, D. J., Grobman, W. A.,
Snowden, J. M. Does infection during pregnancy Hum. Pathol. 30, 759–769 (1999). Palatnik, A. & Stamilio, D. M. Association of duration
outside of the time of delivery increase the risk of 232. Bhutani, V. K., Vilms, R. J. & Hamerman-Johnson, L. of neuroprotective magnesium sulfate infusion with
cerebral palsy? Am. J. Perinatol. 34, 223–228 Universal bilirubin screening for severe neonatal neonatal and maternal outcomes. Obstet. Gynecol.
(2017). hyperbilirubinemia. J. Perinatol. 30 (Suppl.), S6–S15 124, 749–755 (2014).
206. Korzeniewski, S. J., Birbeck, G., DeLano, M. C., (2010). 256. Gluckman, P. D. et al. Selective head cooling with mild
Potchen, M. J. & Paneth, N. A. Systematic review of 233. Brites, D. & Bhutani, V. Pathways involving bilirubin systemic hypothermia after neonatal encephalopathy:
neuroimaging for cerebral palsy. J. Child Neurol. 23, and other brain-injuring agents in Cerebral Palsy: multicentre randomised trial. Lancet 365, 663–670
216–227 (2008). Science and Clinical Practice 131–150 (Mac Ke3ith (2005).
207. Tsutsui, Y., Nagahama, M. & Mizutani, A. Neuronal Press, 2014). 257. Vilchez, G., Dai, J., Lagos, M. & Sokol, R. J. Maternal
migration disorders in cerebral palsy. Neuropathology 234. Asher, P. & Schonell, F. E. Cerebral palsy. Lancet 257, side effects and fetal neuroprotection according to
19, 14–27 (1999). 1201–1201 (1949). body mass index after magnesium sulfate in a
208. Kulak, W. et al. Schizencephaly as a cause of spastic 235. Asher, P. & Schonell, F. E. A survey of 400 cases of multicenter randomized controlled trial. J. Matern.
cerebral palsy. Adv. Med. Sci. 56, 64–70 (2011). cerebral palsy in childhood. Arch. Dis. Child 25, Fetal Neonatal Med. 31, 178–183 (2018).
209. Clark, M., Carr, L., Reilly, S. & Neville, B. G. 360–379 (1950). 258. McAdams, R. M. & Juul, S. E. Neonatal
Worster-Drought syndrome, a mild tetraplegic 236. Bhutani, V. K. et al. Neonatal hyperbilirubinemia and encephalopathy: update on therapeutic hypothermia
perisylvian cerebral palsy. Review of 47 cases. Brain Rhesus disease of the newborn: incidence and and other novel therapeutics. Clin. Perinatol. 43,
123, 2160–2170 (2000). impairment estimates for 2010 at regional and global 485–500 (2016).
210. Bosnjak, V. M. et al. Malformations of cortical levels. Pediatr. Res. 74 (Suppl. 1), 86–100 (2013). 259. Jacobs, S. E. et al. Cooling for newborns with hypoxic
development in children with congenital 237. Olusanya, B. O. & Slusher, T. M. Infants at risk of ischaemic encephalopathy. Cochrane Database Syst.
cytomegalovirus infection – A study of nine children significant hyperbilirubinemia in poorly-resourced Rev. 1, CD003311 (2013).

Reviews
260. Leng, L. Hypothermia therapy after traumatic ages 2, 6, and 9 years. J. Child Neurol. 31, 461–467 view. Dev. Med. Child Neurol. Suppl. 109, 33–34
brain injury: a systematic review and meta-analysis. (2016). (2007).
Turk. Neurosurg.https://doi.org/10.5137/1019-5149. 285. Zarrinkalam, R. et al. CP or not CP? A review of 310. Alberman, E. D. & Stanley, F. J. The Epidemiology of
JTN.19696-16.2 (2017). diagnoses in a cerebral palsy register. Pediatr. Neurol. the Cerebral Palsies (Blackwell Scientific Publications,
261. Chandrasekaran, M., Swamy, R., Ramji, S., Shankaran, S. 42, 177–180 (2010). 1984).
& Thayyil, S. Therapeutic hypothermia for neonatal 286. Novak, I. et al. Early, accurate diagnosis and early 311. Yeargin-Allsopp, M. Distribution of motor types in
encephalopathy in Indian neonatal units: a survey of intervention in cerebral palsy: advances in diagnosis cerebral palsy: how do registry data compare?
national practices. Indian Pediatr. 54, 969–970 and treatment. JAMA Pediatr. 171, 897–907 (2017). Dev. Med. Child Neurol. 53, 199–200 (2011).
(2017). 287. Spittle, A. J., Doyle, L. W. & Boyd, R. N. A systematic 312. Nguyen, L., Mesterman, R. & Gorter, J. W. Development
262. Thayyil, S. et al. Hypothermia for encephalopathy in review of the clinimetric properties of neuromotor of an inventory of goals using the International
low and middle-income countries (HELIX): study assessments for preterm infants during the first year Classification of Functioning, Disability and Health in a
protocol for a randomised controlled trial. Trials 18, of life. Dev. Med. Child Neurol. 50, 254–266 (2008). population of non-ambulatory children and adolescents
432 (2017). 288. Bossuyt, P. M., Reitsma, J. B., Linnet, K. & Moons, K. G. with cerebral palsy treated with botulinum toxin A.
263. Rao, R. et al. Safety and short-term outcomes of Beyond diagnostic accuracy: the clinical utility of BMC Pediatr. 18, 1 (2018).
therapeutic hypothermia in preterm neonates 34–35 diagnostic tests. Clin. Chem. 58, 1636–1643 (2012). 313. Bartlett, D., Dyszuk, E., Galuppi, B. & Gorter, J. W.
weeks gestational age with hypoxic-ischemic 289. Rosenbloom, L. What is the role of the General Interrelationships of functional status and health
encephalopathy. J. Pediatr. 183, 37–42 (2017). Movements Assessment in clinical practice? Dev. Med. conditions in children with cerebral palsy: a descriptive
264. Johnson, C. T., Burd, I., Raghunathan, R., Child Neurol. 60, 6 (2018). study. Pediatr. Phys. Ther. 30, 10–16 (2018).
Northington, F. J. & Graham, E. M. Perinatal 290. Spittle, A., Orton, J., Anderson, P., Boyd, R. & 314. Paulson, A. & Vargus-Adams, J. Overview of four
inflammation/infection and its association with Doyle, L. W. Early developmental intervention functional classification systems commonly used in
correction of metabolic acidosis in hypoxic-ischemic programmes post-hospital discharge to prevent cerebral palsy. Children 4, E30 (2017).
encephalopathy. J. Perinatol 36, 448–452 (2016). motor and cognitive impairments in preterm infants. 315. Nelson, K. B. & Ellenberg, J. H. Obstetric
265. Lee, J. K. et al. Optimizing cerebral autoregulation Cochrane Database Syst. Rev. 12, Cd005495 complications as risk-factors for cerebral-palsy or
may decrease neonatal regional hypoxic-ischemic (2012). seizure disorders. JAMA 251, 1843–1848 (1984).
brain injury. Dev. Neurosci. 39, 248–256 (2017). 291. Morgan, C. et al. Effectiveness of motor interventions 316. Nelson, K. B. & Ellenberg, J. H. Antecedents of
266. Landucci, E. et al. Neuroprotective effects of in infants with cerebral palsy: a systematic review. cerebral palsy. I. Univariate analysis of risks.
topiramate and memantine in combination with Dev. Med. Child Neurol. 58, 900–909 (2016). Am. J. Dis. Child 139, 1031–1038 (1985).
hypothermia in hypoxic-ischemic brain injury in vitro 292. Korzeniewski, S. J., Birbeck, G., DeLano, M. C., 317. Tollanes, M. C. et al. Cohort profile: cerebral palsy in
and in vivo. Neurosci. Lett. 668, 103–107 (2018). Potchen, M. J. & Paneth, N. A systematic review of the Norwegian and Danish birth cohorts
267. Wu, Y. W. et al. Erythropoietin for neuroprotection in neuroimaging for cerebral palsy. J. Child Neurol. 23, (MOBAND-CP). BMJ Open 6, e012777 (2016).
neonatal encephalopathy: safety and pharma 216–227 (2008). 318. Myatt, L. et al. Strategy for standardization of
cokinetics. Pediatrics 130, 683–691 (2012). 293. Reid, S. M., Dagia, C. D., Ditchfield, M. R., Carlin, J. B. preeclampsia research study design. Hypertension 63,
268. Clark, R. H., Bloom, B. T., Spitzer, A. R. & & Reddihough, D. S. Population-based studies of brain 1293–1301 (2014).
Gerstmann, D. R. Reported medication use in the imaging patterns in cerebral palsy. Dev. Med. Child 319. Hyams, K. C. Developing case definitions for
neonatal intensive care unit: data from a large national Neurol. 56, 222–232 (2013). symptom-based conditions: the problem of specificity.
data set. Pediatrics 117, 1979–1987 (2006). 294. White, T. et al. Paediatric population neuroimaging Epidemiol. Rev. 20, 148–156 (1998).
269. Schmidt, B. et al. Caffeine therapy for apnea of and the Generation R Study: the second wave. 320. Weiss, N. S. & Liff, J. M. Accounting for the
prematurity. N. Engl. J. Med. 354, 2112–2121 Eur. J. Epidemiol. 33, 99–125 (2017). multicausal nature of disease in the design and
(2006). 295. George, J. M. et al. Diagnostic accuracy of early analysis of epidemiologic studies. Am. J. Epidemiol.
270. Schmidt, B. et al. Survival without disability to age 5 magnetic resonance imaging to determine motor 117, 14–18 (1983).
years after neonatal caffeine therapy for apnea of outcomes in infants born preterm: a systematic 321. Schang, A. L., Gressens, P. & Fleiss, B. Revisiting
prematurity. JAMA 307, 275–282 (2012). review and meta-analysis. Dev. Med. Child Neurol. 60, thyroid hormone treatment to prevent brain damage
271. Aly, H. et al. Melatonin use for neuroprotection in 134–146 (2018). of prematurity. J. Neurosci. Res. 92, 1609–1610
perinatal asphyxia: a randomized controlled pilot 296. LeWinn, K. Z., Sheridan, M. A., Keyes, K. M., (2014).
study. J. Perinatol 35, 186–191 (2015). Hamilton, A. & McLaughlin, K. A. Sample composition 322. Dowding, V. M. & Barry, C. Cerebral palsy: social class
272. Wilkinson, D., Shepherd, E. & Wallace, E. M. alters associations between age and brain structure. differences in prevalence in relation to birthweight and
Melatonin for women in pregnancy for neuroprotection Nat. Commun. 8, 874 (2017). severity of disability. J. Epidemiol. Commun. Health
of the fetus. Cochrane Database Syst. Rev. 3, 297. Yanni, D. et al. Both antenatal and postnatal 44, 191–195 (1990).
CD010527 (2016). inflammation contribute information about the risk of 323. Ahlin, K. et al. Non-infectious risk factors for different
273. Fischer, H. S., Reibel, N. J., Buhrer, C. & Dame, C. brain damage in extremely preterm newborns. types of cerebral palsy in term-born babies:
Prophylactic early erythropoietin for neuroprotection Pediatr. Res. 82, 691–696 (2017). a population-based, case-control study. BJOG 120,
in preterm infants: a meta-analysis. Pediatrics 139 298. Korzeniewski, S. J. et al. Are preterm newborns who 724–731 (2013).
(2017). have relative hyperthyrotropinemia at increased risk of 324. Freire, G., Shevell, M. & Oskoui, M. Cerebral palsy:
274. Kulak-Bejda, A., Kulak, P., Bejda, G., Krajewska-Kulak, E. brain damage? J. Pediatr. Endocrinol. Metab. 27, phenotypes and risk factors in term singletons born
& Kulak, W. Stem cells therapy in cerebral palsy: 1077–1088 (2014). small for gestational age. Eur. J. Paediatr. Neurol. 19,
a systematic review. Brain Dev. 38, 699–705 (2016). 299. Korzeniewski, S. J. et al. Elevated endogenous 218–225 (2015).
275. Kiasatdolatabadi, A. et al. The role of stem cells in the erythropoietin concentrations are associated with 325. Stanek, J. L., Emerson, J. A., Murdock, F. A. &
treatment of cerebral palsy: a review. Mol. Neurobiol. increased risk of brain damage in extremely preterm Petroski, G. F. Growth characteristics in cerebral palsy
54, 4963–4972 (2017). neonates. PLOS One 10, e0115083 (2015). subtypes: a comparative assessment. Dev. Med. Child
276. Sun, J. M. et al. Effect of autologous cord blood 300. Leviton, A. et al. Two-hit model of brain damage in the Neurol. 58, 931–935 (2016).
infusion on motor function and brain connectivity in very preterm newborn: small for gestational age and 326. Ahlin, K., Himmelmann, K., Nilsson, S., Sengpiel, V. &
young children with cerebral palsy: a randomized, postnatal systemic inflammation. Pediatr. Res. 73, Jacobsson, B. Antecedents of cerebral palsy according
placebo-controlled trial. Stem Cells Transl Med. 6, 362–370 (2013). to severity of motor impairment. Acta Obstet.
2071–2078 (2017). 301. Barnett, M. L. et al. Exploring the multiple-hit Gynecol. Scand. 95, 793–802 (2016).
277. Solomons, G., Holden, R. H. & Denhoff, E. The hypothesis of preterm white matter damage using 327. Zelnik, N. et al. The role of prematurity in patients
changing picture of cerebral dysfunction in early diffusion MRI. NeuroImage Clinical 17, 596–606 with hemiplegic cerebral palsy. J. Child Neurol. 31,
childhood. J. Pediatr. 63, 113–120 (1963). (2018). 678–682 (2016).
278. Drillien, C. M. Abnormal neurologic signs in the first 302. Wixey, J. A., Chand, K. K., Colditz, P. B. & Bjorkman, S. T. 328. Kitai, Y. et al. Outcome of hemiplegic cerebral palsy
year of life in low-birthweight infants: possible Review: Neuroinflammation in intrauterine growth born at term depends on its etiology. Brain Dev. 38,
prognostic significance. Dev. Med. Child Neurol. 14, restriction. Placenta 54, 117–124 (2017). 267–273 (2016).
575–584 (1972). 303. Korzeniewski, S. J. et al. A “multi-hit” model of 329. Yang, Y. et al. Hypoxic stress forces adaptive and
279. Nelson, K. B. & Ellenberg, J. H. Children who neonatal white matter injury: cumulative contributions maladaptive placental stress responses in early
“outgrew’ cerebral palsy. Pediatrics 69, 529–536 of chronic placental inflammation, acute fetal pregnancy. Birth Defects Res. 109, 1330–1344
(1982). inflammation and postnatal inflammatory events. (2017).
280. Taudorf, K., Hansen, F. J., Melchior, J. C. & Pedersen, H. J. Perinat. Med. 42, 731–743 (2014). 330. Lahti-Pulkkinen, M. et al. Placental morphology is
Spontaneous remission of cerebral palsy. 304. Wang, L. W., Lin, Y. C., Wang, S. T., Yeh, T. F. & associated with maternal depressive symptoms during
Neuropediatrics 17, 19–22 (1986). Huang, C. C. Hypoxic/ischemic and infectious events pregnancy and toddler psychiatric problems. Sci. Rep.
281. Pedersen, S. J., Sommerfelt, K. & Markestad, T. have cumulative effects on the risk of cerebral palsy in 8, 791 (2018).
Early motor development of premature infants with very-low-birth-weight preterm infants. Neonatology 331. Labarrere, C. A. et al. Failure of physiologic
birthweight less than 2000 grams. Acta Paediatr. 89, 106, 209–215 (2014). transformation of spiral arteries, endothelial and
1456–1461 (2000). 305. Schneider, R. E. et al. The association between trophoblast cell activation, and acute atherosis in the
282. Leach, E. L., Shevell, M., Bowden, K., maternal age and cerebral palsy risk factors. basal plate of the placenta. Am. J. Obstet. Gynecol.
Stockler-Ipsiroglu, S. & van Karnebeek, C. D. Treatable Pediatr. Neurol. 82, 25–28 (2018). 216, 287.e1–287.e16 (2017).
inborn errors of metabolism presenting as cerebral 306. Last, J. A. Dictionary of Epidemiology 3rd edn 332. Korzeniewski, S. J. et al. Maternal plasma angiogenic
palsy mimics: systematic literature review. (Oxford Univ. Press, 2002). index-1 (placental growth factor/soluble vascular
Orphanet J. Rare Dis. 9, 197 (2014). 307. Phelps, W. M. The cerebral palsies. Clin. Orthopaed. endothelial growth factor receptor-1) is a biomarker
283. Shevell, M. I., Majnemer, A., Poulin, C. & Law, M. Related Res. 44, 83–88 (1966). for the burden of placental lesions consistent with
Stability of motor impairment in children with 308. Stanley, F. J. & Watson, L. The cerebral palsies uteroplacental underperfusion: a longitudinal
cerebral palsy. Dev. Med. Child Neurol. 50, 211–215 in Western Australia: trends, 1968 to 1981. case-cohort study. Am. J. Obstet. Gynecol. 214,
(2008). Am. J. Obstet. Gynecol. 158, 89–93 (1988). 629.e1–629.e17 (2016).
284. Korzeniewski, S. J. et al. Persistence of cerebral palsy 309. Blair, E., Badawi, N. & Watson, L. Definition and 333. McCowan, L. M. et al. Prediction of small for
diagnosis: assessment of a low-birth-weight cohort at classification of the cerebral palsies: the Australian gestational age infants in healthy nulliparous women

Reviews
using clinical and ultrasound risk factors combined 343. Hagberg, H., Gressens, P. & Mallard, C. Inflammation 353. Roberts, J. M., Redman, C. W. G. & Global Pregnancy, C.
with early pregnancy biomarkers. PLOS One 12, during fetal and neonatal life: implications for Global Pregnancy, C. Global Pregnancy Collaboration
e0169311 (2017). neurologic and neuropsychiatric disease in children symposium: Prepregnancy and very early pregnancy
334. Kumar, M. et al. Adverse fetal outcome: is first and adults. Ann. Neurol. 71, 444–457 (2012). antecedents of adverse pregnancy outcomes: overview
trimester ultrasound and Doppler better predictor 344. Fleiss, B. & Gressens, P. Tertiary mechanisms of brain and recommendations. Placenta 60, 103–109
than biomarkers? J. Matern. Fetal Neonatal Med. 30, damage: a new hope for treatment of cerebral palsy? (2017).
1410–1416 (2017). Lancet Neurol. 11, 556–566 (2012). 354. Roberts, J. M., Mascalzoni, D., Ness, R. B., Poston, L.
335. Levine, R. J. et al. Circulating angiogenic factors 345. Reuss, M. L., Paneth, N. & Susser, M. Does the loss of & Global Pregnancy Collaboration. Collaboration to
and the risk of preeclampsia. N. Engl. J. Med. 350, placental hormones contribute to neurodevelopmental understand complex diseases: preeclampsia and
672–683 (2004). disabilities in preterm infants? Dev. Med. Child Neurol. adverse pregnancy outcomes. Hypertension 67,
336. O’Gorman, N. et al. Accuracy of competing-risks 36, 743–747 (1994). 681–687 (2016).
model in screening for pre-eclampsia by maternal 346. Schang, A. L. et al. Failure of thyroid hormone 355. Paneth, N. & Monk, C. The importance of cohort
factors and biomarkers at 11–13 weeks’ gestation. treatment to prevent inflammation-induced white research starting early in life to understanding child
Ultrasound Obstet. Gynecol. 49, 751–755 (2017). matter injury in the immature brain. Brain Behav. health. Curr. Opin. Pediatr. 30, 292–296 (2018).
337. Bolnick, J. M. et al. Altered biomarkers in trophoblast Immun. 37, 95–102 (2014).
cells obtained noninvasively prior to clinical 347. Stroustrup, A., Teitelbaum, S. L. & Aschner, J. L. The Author contributions
manifestation of perinatal disease. Sci. Rep. 6, 32382 value of preterm infant environmental health cohorts: All authors researched data for the article, wrote the article
(2016). the canary in the coal mine. JAMA Pediatr. 171, and reviewed and/or edited the manuscript before submis-
338. Moser, G., Drewlo, S., Huppertz, B. & Armant, D. R. 1139–1140 (2017). sion. S.J.K. and N.P. made substantial contributions to
Trophoblast retrieval and isolation from the cervix: 348. Myatt, L. & Roberts, J. M. Preeclampsia: syndrome or discussion of the content for the article.
origins of cervical trophoblasts and their potential disease? Curr. Hypertension Rep. 17, 83 (2015).
value for risk assessment of ongoing pregnancies. 349. Goldenberg, R. L. et al. The preterm birth syndrome: Competing interests
Hum. Reprod. Update. https://doi.org/10.1093/ issues to consider in creating a classification system. The authors declare no competing interests.
humupd/dmy008 (2018). Am. J. Obstet. Gynecol. 206, 113–118 (2012).
339. Joyner, M. J. & Pedersen, B. K. Ten questions about 350. Catov, J. M. et al. Neonatal outcomes following Publisher’s note
systems biology. J. Physiol. 589, 1017–1030 (2011). preterm birth classified according to placental Springer Nature remains neutral with regard to jurisdictional
340. Kotas, M. E. & Medzhitov, R. Homeostasis, features. Am. J. Obstet. Gynecol. 216, 411.e1–411. claims made in published maps and institutional affiliations.
inflammation, and disease susceptibility. Cell 160, e14 (2017).
816–827 (2015). 351. Park, E. Y. & Kim, W. H. Prevalence of secondary Reviewer information
341. Lisovska, N. et al. Pathogenesis of cerebral palsy impairments of adults with cerebral palsy according Nature Reviews Neurology thanks R. Boyd, D. Damiano and
through the prism of immune regulation of nervous to gross motor function classification system. K. Graham for their contribution to the peer review of this
tissue homeostasis: literature review. Childs Nerv. J. Phys. Ther. Sci. 29, 266–269 (2017). work.
Syst. 32, 2111–2117 (2016). 352. Kovo, M. et al. Pregnancy outcome and placental
342. Chovatiya, R. & Medzhitov, R. Stress, inflammation, findings in pregnancies complicated by fetal Related links
and defense of homeostasis. Mol. Cell 54, 281–288 growth restriction with and without preeclampsia. CoLab: https://pregnancycolab.tghn.org/
(2014). Reprod. Sci. 22, 316–321 (2015).

The Complex Aetiology of Cerebral Palsy

Uploaded by

Copyright:

Available Formats

The Complex Aetiology of Cerebral Palsy

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Complex Aetiology of Cerebral Palsy

Uploaded by

Copyright:

Available Formats

Reviews

The complex aetiology of

528 | SEPTEMBER 2018 | volume 14 www.nature.com/nrneurol

born during 1959–1966. The NCPP provided compel-

NaTuRE REviEwS | NeurOLOgY volume 14 | SEPTEMBER 2018 | 529

Much research has been devoted to identifying the

530 | SEPTEMBER 2018 | volume 14 www.nature.com/nrneurol

NaTuRE REviEwS | NeurOLOgY volume 14 | SEPTEMBER 2018 | 531

532 | SEPTEMBER 2018 | volume 14 www.nature.com/nrneurol

NaTuRE REviEwS | NeurOLOgY volume 14 | SEPTEMBER 2018 | 533

Kernicterus. Neonatal jaundice is the most common Therapeutic hypothermia

534 | SEPTEMBER 2018 | volume 14 www.nature.com/nrneurol

NaTuRE REviEwS | NeurOLOgY volume 14 | SEPTEMBER 2018 | 535

536 | SEPTEMBER 2018 | volume 14 www.nature.com/nrneurol

NaTuRE REviEwS | NeurOLOgY volume 14 | SEPTEMBER 2018 | 537

and redundancy, as major intellectual tools to under- Conclusions

538 | SEPTEMBER 2018 | volume 14 www.nature.com/nrneurol

NaTuRE REviEwS | NeurOLOgY volume 14 | SEPTEMBER 2018 | 539

540 | SEPTEMBER 2018 | volume 14 www.nature.com/nrneurol

NaTuRE REviEwS | NeurOLOgY volume 14 | SEPTEMBER 2018 | 541

542 | SEPTEMBER 2018 | volume 14 www.nature.com/nrneurol

NaTuRE REviEwS | NeurOLOgY volume 14 | SEPTEMBER 2018 | 543

You might also like