DEVELOPMENTAL MEDICINE & CHILD NEUROLOGY REVIEW

Brain plasticity and recovery from early cortical injury

BRYAN KOLB | RICHELLE MYCHASIUK | PRESTON WILLIAMS | ROBBIN GIBB

Department of Neuroscience, University of Lethbridge, Alberta, Canada.

Correspondence to Dr Bryan Kolb, Department of Neuroscience, University of Lethbridge, Lethbridge, AB T1K 3M4, Canada. E-mail: Kolb@uleth.ca

PUBLICATION DATA Neocortical development represents more than a simple unfolding of a genetic blueprint: rather, it
Accepted for publication 28th February 2011. represents a complex dance of genetic and environmental events that interact to adapt the brain
to fit a particular environmental context. Most cortical regions are sensitive to a wide range of
ABBREVIATION experiential factors during development and later in life, but the injured cortex appears to be
FGF2 Fibroblast growth factor 2 unusually sensitive to perinatal experiences. This paper reviews the factors that influence how nor-
mal and injured brains (both focal and ischemic injuries) develop and adapt into adulthood. Such
factors include prenatal experiences in utero as well as postnatal experiences throughout life.
Examples include the effects of sensory and motor stimulation, psychoactive drugs (including illi-
cit and prescription drugs), maternal and postnatal stress, neurotrophic factors, and pre- and post-
natal diet. All these factors influence cerebral development and influence recovery from brain
injury during development.

Over the past 25 years we have developed a model of early
cortical injury in the rodent cortex.1 The underlying assump-
tion of our experiments has been that there are critical periods
during development in which the brain is more able to reorga-
nize, and others where it is less able to reorganize, to allow
restoration of function after neocortical injury. The critical
periods are constrained by the ongoing developmental pro-
cesses that are occurring at the time of injury. For example,
when there is proliferation of neurons (such as on embryonic
day 18 in rats) or astrocytes (such as on postnatal day 8), the
brain can recruit these proliferative processes to facilitate pro-
cesses leading to functional recovery. In contrast, if neurons
are migrating or the brain is shedding synapses or neurons,
injury may lead to a disruption (or enhancement) of this pro-
cess, which may exacerbate the effects of injury. The challenge
is to identify pre- and postnatal treatments that will shift the
critical periods to make it possible partly to stimulate func-
tional restoration at times when it would not normally occur.
We began by searching for factors that can influence devel-
opment in the normal brain of laboratory animals. It is now
possible to identify a wide range of pre- and postnatal experi-
ences that can produce significant changes in both brain and
behavioural development (Table I).2–8 Several principles
emerged from these studies. First, both pre- and postnatal fac-
tors could influence brain and behavioural development. Sec-
ond, early experiences led to changes in the organization of
adult neural networks throughout the cerebral cortex. Third,
one clear mechanism of the brain and behavioural changes was
epigenetic. Thus, early experiences could chronically alter
gene expression, which in turn was presumably responsible, at
least in part, for the altered development. Consider one exam-
ple. When pregnant rats were given a moderate stressor
(10 min sitting on a small platform in an open room) twice
daily during the second week of gestation, there was an
increase in global methylation, altered gene expression, and
modified synaptic organization throughout the cerebral cor-
tex.8 Curiously, if the rats were given a more severe stressor
(30 min on the platform) there was a decrease in methylation
and a different pattern of altered gene expression, suggesting
that stress can produce dose-dependent epigenetic changes in
cerebral function.
Taken together, it has become clear that developmental
experiences have profound effects on brain and behavioural
development and that these effects vary exquisitely with the
precise age and type of experience. These experiences are
bound to influence the severity and outcome following cere-
bral lesions during development.
MODULATING RECOVERY FROM EARLY INJURY
Table II summarizes a range of factors that modulate recovery
from early brain injury.9–18 We will consider these factors in
three general categories: behavioural therapies, neurotrophic
factor therapies, and drug therapies.
Behavioural therapies
There have been surprisingly few laboratory studies that have
tried to identify treatments using animals with perinatal corti-
cal injuries, and the few that are available have not reported
any neural correlates. Our studies began by looking at the
effects of the perinatal experiential treatments that we had
found to influence cortical development in intact animals.
There were two general results. First, nearly all treatments

4 DOI: 10.1111/j.1469-8749.2011.04054.x ª The Authors. Developmental Medicine & Child Neurology ª 2011 Mac Keith Press
 Table I: Modification of cerebral development by experience

Treatment Result Basic reference

Complex housing at weaning
Complex housing in adulthood
Complex housing prenatally
Postnatal ‘handling’
Postnatal tactile stimulation
Prenatal tactile stimulation
Play behaviour
Prenatal valium
Prenatal fluoxetine
Perinatal antipsychotics
Prenatal stress
Increased dendritic growth; decreased spine density Kolb et al.2
Increased dendritic growth; increased spine density Kolb et al.2
Decreased dendritic length; increased spine density R Gibb, C Gonzalez, B Kolb
Decreased dendritic length; spines unchanged Gibb and Kolb3
Decreased dendritic length; decreased spine density Kolb and Gibb4
Decreased dendritic length; increased spine density R Gibb, B Kolb3
Increased dendritic length in medial prefrontal cortex with more play partners Bell et al.5
Increased dendritic length Kolb et al.6
Decreased dendritic length Kolb et al.6
Decreased dendritic length and spine density Frost et al.7
Altered gene methylation, gene expression, and synaptic organization Mychasiuk et al.8

All experience-dependent changes were measured in adulthood with the exception of the effects prenatal stress, which were measured at
postnatal day 21. References without bibliography numbers refer to unpublished observations by the authors.

Table II: Modification of the effects of early cortical injury

Treatment Result Basic reference

Complex housing at weaning Enhanced recovery after P1–7 frontal injury Kolb & Elliott9
Complex housing in adulthood No functional recovery after P1–5 frontal injury Comeau et al.10
Complex housing prenatally Enhanced recovery after P4 frontal injury R Gibb, C Gonzalez, B Kolb
Postinjury tactile stimulation Enhanced recovery after P4 frontal, parietal, or motor injury Kolb and Gibb4
Prenatal tactile stimulation Enhanced recovery after P4 frontal injury R Gibb, B Kolb3
Postinjury FGF2 Enhanced recovery after P4 frontal or parietal lesions Comeau et al.10
Prenatal FGF2 Enhanced recovery after P4 frontal or parietal lesions Comeau et al.11
Postinjury FGF2 Enhanced recovery after P10 or P35 motor cortex lesions Monfils et al.12–14; Nemati and Kolb15
Postinjury FGF2 Enhanced recovery after P7 hypoxic–ischemic lesions Williams16
Prenatal diazepam Enhanced recovery after P7 frontal injury Kolb et al.6
Prenatal fluoxetine Blocked recovery after P7 frontal injury Kolb et al.6
Perinatal nicotine Facilitates recovery after P4 injury McKenna et al.17
Postinjury nicotine Facilitates recovery after P7 hypoxia–ischemia Williams16
Enhanced diet Enhanced recovery after P7 frontal injury Halliwell et al.18

All functional changes were measured in adulthood. References without bibliography numbers refer to unpublished observations by the authors.
P, postnatal day.

enhanced cognitive, and to a lesser extent motor, outcome
after perinatal cortical injuries. Second, there were neural cor-
relates of the enhanced recovery in every case.
Perhaps the biggest behavioural effect is seen in perina-
tally brain-injured animals given complex housing either at
weaning or prenatally (and thus pre-injury). In both cases
there is a reduction in the loss of adult brain weight and a
parallel increase in spine density in pyramidal cortical neu-
rons. In contrast, placing infant brain-injured rats into
complex environments as adults failed to reverse the
behavioural deficits.
There is also a marked behavioural advantage to animals
that receive tactile stimulation with a soft brush postinjury.
For example, when infant rats receive either medial prefrontal
or posterior parietal lesions followed by 10 days of tactile
stimulation, they show enhanced performance in adulthood
on tests of both spatial navigation and skilled reaching com-
pared with untreated lesion animals.4 The neural correlates in
this study were unexpected, however, for two reasons (Fig. 1).
First, whereas the tactile-stimulated rats with parietal lesions
showed enhanced dendritic length, the rats with frontal lesions
did not. In contrast, the tactile-stimulated frontal animals
showed increased spine density but the parietals did not. Thus,
the neural response varied with lesion location. Second, the
increased spine density in the tactile-stimulated frontal ani-
mals was opposite to what was observed in the tactile-stimu-
lated control rats. Thus, the same treatment had opposite
effects in the control and frontal-injured rats but both showed
enhanced behaviour.
Neurotrophic therapies
Fibroblast growth factor 2 (FGF2) is known to influence neu-
ral development both via its effects on neural mitosis and dif-
ferentiation. It also shows enhanced expression when animals
are placed in complex environments or given psychomotor
stimulants.19 FGF2 levels also are increased in skin and brain
and the FGF2 receptor is increased in brain after tactile stimu-
lation in infant rats. FGF2 also shows enhanced expression in
the second week of life in rats,20 which is the time of best func-
tional recovery from focal injuries. These findings led us to
administer FGF2 directly as a therapy. It has the advantage
that it passes the blood–brain barrier and thus can be given
subcutaneously. In our first series of studies we gave animals
with perinatal medial frontal or posterior parietal lesions
FGF2 either prenatally, post injury, or both.10,11 All three
treatment regimens proved to be beneficial, with the com-

Review 5
 Male synapse number
25
Prenatal stress
Average synapses (E7)
20 Control
15
10
5 5
Cg3 AID CA1
Figure 1: Synaptic changes in medial prefrontal (Cg3), orbital frontal (AID), and hippocampus (CA1) after prenatal stress. The total number of synapses was
estimated by combining stereological measurement of neuronal number and Golgi analysis of dendritic length and spine density. Stress reduced synapse
number both in males and females in CA1 and produced a small change in male but not female Cg3. The effects of stress were sexually dimorphic in AID,
with a large decrease in males compared with a large increase in females. (Modified from Mychasiuk8.)
bined prenatal and postnatal treatments being the most advan-
tageous; furthermore, cognitive functions showed better
recovery than motor behaviours.
In a second set of studies, rats with hypoxic–ischemic lesions
on postnatal day 7 received FGF2 and again showed func-
tional benefits. In addition to showing more normal motor
maps and less cortical atrophy than untreated lesion animals,
their reaching performance with the contra-lesion forelimb
was restored.16
In a final set of studies we administered FGF2 to rats with
focal motor cortex lesions on postnatal day 10. Although such
animals normally show partial recovery of function, the recov-
ery is incomplete. We therefore wondered if the FGF might
further enhance the recovery. To our surprise, the FGF2
treatment not only improved motor recovery but that recov-
ery was correlated with a regeneration of the lost tissue
(Fig. 2).13–15 This tissue does not appear normal in lamination
but it does have functional connections with the spinal cord
and the cells have fairly normal patterns of spontaneous
discharge. Removing the regenerated tissue produces immedi-
ate loss of the recovered functions. We subsequently made
lesions in other, more posterior cortical regions and adminis-
tered FGF2, but did not see any regeneration of lost tissue.
Perhaps the distance from the subventricular zone where the
cells are generated is too far or there is some obstacle to their
migration to more posterior regions.
In sum, FGF2 appears to play an important role both in
spontaneous recovery and as an agent to enhance recovery.
One minor worry in translating the use of neurotrophic factors
to the clinic is their powerful action on the mitotic process,
which may produce risk of unwanted spontaneous tissue
generation (i.e. tumours). However, we have not seen any
evidence of this in infant rats or in adult rats stimulated to
regenerate neuronal tissue after stroke.21
6 Developmental Medicine & Child Neurology 2011, 53 (Suppl. 4): 4–8
Female synapse number
35
30
Prenatal stress
Synapse number (E7)
Control
25
20
15
10
Cg3 AID CA1
Drug therapies
We have administered prescription drugs (such as diazepam
and fluoxetine) prenatally to animals that then received medial
prefrontal lesions on postnatal day 7.6,17 Valium stimulates the
GABA (c-aminobutyric acid) receptor, which in turn plays an
important role in cerebral plasticity,22,23 and is able to shift the
critical period for ocular dominance plasticity (T Hensch, per-
sonal communication). We hypothesized that valium might
facilitate recovery from early injury. It did, and this was corre-
lated with increased dendritic length in pyramidal cells in per-
ilesional regions. In contrast to diazepam, fluoxetine
completely blocked any spontaneous recovery, which was cor-
related by a parallel blockade of compensatory dendritic
changes.
We also did a series of studies with nicotine because we
have found dramatic increases in dendritic length and spine
density in pyramidal cells in the prefrontal and motor cortex
of rats given nicotine as adults.24 We reasoned that nicotine
might be beneficial when given either prenatally or postinju-
ry. It was. Although we have not yet completed our analysis
of neural correlates, we showed that rats given nicotine after
day 7 hypoxia–ischemia had both behavioural improvement
and more normal motor maps compared with untreated
animals.16 Nicotine was not effective in the same way after
day 14 hypoxia–ischemia, which implies that age-at-injury
interacts with the treatment. In addition, we also showed that
nicotine did not have to be given immediately after the injury
but was effective in animals with day 3 frontal injuries that
received the drug in adulthood. Curiously, however, animals
given the prenatal nicotine but not given the later lesions
fared worse than saline controls.17 Once again, it appears that
events can have different effects on the intact and injured
brains.
 a Basilar dendritic length ence induces selective epigenetic effects in the injured brain
2200 remains a mystery. We do have preliminary evidence that
Non stroking
infant frontal lesions can increase (postnatal day 10) or
Stroking decrease (postnatal day 3) global gene methylation in the cor-
2000
tex,8,25 but we are still ignorant as to how gene methylation
might be translated into changes in spine density. This is
Mean dendritic length (µm)

1800 clearly grist for future studies.

1600
APPLYING LABORATORY-BASED KNOWLEDGE
TO THE CLINIC
Animal models of early brain injury have allowed us to identify
1400 some basic principles underlying recovery from early injury.
The application of rehabilitation strategies based upon the
1200 laboratory animal studies has yet to be realized, but the
laboratory animal studies should give rehabilitation specialists
and families cause for optimism. There appears to be far more
1000
capacity for functional recovery than had previously been
Control Frontal Parietal
recognized. Non-invasive treatments such as tactile stimula-
b Apical spine density tion would be easy to implement and without any obvious
7
risks. In fact, there is a general consensus that tactile stimula-
tion is important in normal human development and especially
Non stroking
in cases in which the infants are compromised such as from
6 Stroking premature birth. The application of complex housing in rats
can be translated into the use of a wide range of games and
Mean spines per 10 µm

activities to stimulate the infant’s brain, and especially the use

5 of reading to infants and young children. Older children
may benefit from Web-based treatment programmes. The
advantage of reading is that it can provide tactile, cognitive,
4 and social stimulation simultaneously.
One of the problems with generalizing from animal studies
to children relates to the validity of rodent models for perina-
3 tal injury in children. There are several issues here including
rodent ⁄ human brain comparisons, injury etiology, and the
practicality of using invasive or pharmacological treatments in
2 children. We consider each in turn.
Control Frontal Parietal Corrected for body size, the human brain is about 12 times
larger than the rat brain. Although the rat brain is not a minia-
Figure 2: Dendritic measures on layer III pyramidal cells from Zilles' Par ture version of the human brain, its organization is remarkably
1. (a) Both frontal and parietal lesions reduced dendritic length. Tactile similar, with the exception of language-related regions (see, for
stroking significantly reversed this effect in parietal but not frontal rats. (b) example, the review by Krubitzer and Kaas26). Perhaps the
Both frontal and parietal lesions significantly reduced spine density. Tac- most important difference is in the ratio of grey to white matter:
tile stroking significantly decreased spine density in neurons in control the human brain has more connections and thus more white
animals but increased spine density in cells frontal lesion animals. Data matter. This difference presents a ‘double-edged sword’. Thus,
are mean € SE the mean (Kolb and Gibb4). the additional connections may provide the human brain with
more plastic potential; however, at the same time, injury to
POTENTIAL MECHANISMS OF PLASTICITY IN THE white matter may lead to more disconnection syndromes in
INJURED BRAIN humans than in rodents. It is our view that the grey ⁄ white mat-
Our demonstration of plastic changes in the normal and ter difference does not impugn the usefulness of rodent models
injured brain clearly show the remarkable capacity of the but it is a factor that we should be mindful of. Furthermore, the
developing brain to form novel neural networks, either chang- principles that underlie brain plasticity appear surprisingly con-
ing existing circuits or making new ones, which can be stant across diverse animal nervous systems.27
inferred from changes in dendritic morphology. A key
question, however, is how this happens. At this point our Injury etiology
understanding of how morphological changes are produced is There are many ways to injure the perinatal brain, including
clearly only in its infancy. It seems likely that the common hypoxia–ischemia, teratological agents, and trauma. Most ani-
molecular event will prove to be a change in gene expression, mal models of perinatal injury use either hypoxic–ischemic
either directly or through epigenetics, but exactly how experi- injury, teratological agents such alcohol, or surgical removal of

Review 7
brain tissue. Most of our experiments have used surgical injuries
because it is easier to control the locus of injury with this tech-
nique. Although a surgical etiology would be uncommon in
perinatal injury in humans, there is little reason to believe that
effectiveness of treatments is etiology dependent. For example,
Williams16 showed that both FGF2 and nicotine are as effective
in treating animals with perinatal hypoxic–ischemic injuries as
they are for surgical injuries. Similarly, Saucier et al.28 and Chou
et al.29 had similar findings with complex housing and tactile
stimulation, respectively, in animals with ischemic injuries.
As noted above, most of the treatments discussed here are
easily implemented with human infants, including especially
tactile stimulation and diet (see Dabydeen et al.30). There is lit-
tle likelihood of negative side effects from such treatments. The
stimulation of endogenous stem cells with neurotrophic factors
or pharmacological agents is not practical, however, given the
uncertainty over the long-term risks with such treatments. It is
our hypothesis, however, that the behavioural treatments are
acting to increase the endogenous production of neurotrophic
factors, which provides a mechanism of action.
Perhaps the major contribution of the animal studies so far
is to demonstrate that a wide range of rehabilitation pro-
grammes can work after early brain injury and that there are
identifiable mechanisms that account for the beneficial effects
of many rehabilitation regimes. The importance of under-
standing the mechanisms underlying recovery from early brain
injury cannot be understated. Understanding how the brain
can compensate for early injury will allow a more rationale
choice of novel rehabilitation strategies that are targeted to
stimulating recovery. It is our hunch that as we continue to
study the molecular underpinnings of rehabilitation treat-
ments we will have a better understanding of how to shift criti-
cal periods after brain injury. This would allow us selectively
to target mechanisms that will stimulate regenerative processes
and reduce the parallel degenerative processes. This may
prove especially important for those infants who may be pre-
sumed to be at risk either because of poor prenatal care,
maternal distress or other maternal health issues, disadvan-
taged socioeconomic status, and so on.
ACKNOWLEDGEMENTS
We thank Cathy Carroll, Dawn Danka, and Kehe Xie for technical
assistance with the anatomical analyses. This work was supported by
Natural Sciences and Engineering Research Council of Canada
grants to BK and RG, and a Canadian Institutes of Health Research
grant to BK.

REFERENCES
1. Kolb B. Brain Plasticity and Behavior. Mahwah, NJ:
Erlbaum, 1995.
2. Kolb B, Gibb R, Gorny G. Experience-dependent changes
in dendritic arbor and spine density in neocortex
vary with age and sex. Neurobiol Learn Mem 2003; 79: 1–10.
3. Gibb R, Kolb B. Neonatal handling alters brain organization
but does not influence recovery from perinatal cortical
injury. Behav Neurosci 2005; 119: 1375–83.
4. Kolb B, Gibb R. Tactile stimulation facilitates functional
recovery and dendritic change after neonatal medial frontal
or posterior parietal lesions in rats. Behav Brain Res 2010;
214: 115–20.
5. Bell HC, Pellis SM, Kolb B. Juvenile peer play experience
and the development of the orbitofrontal and medial
prefrontal cortex. Behav Brain Res 2010; 207: 7–13.
6. Kolb B, Gibb R, Pearce S, Tanguay R. Prenatal exposure to
prescription medication alters recovery from early brain
damage in rats. Soc Neurosci Abstr 2008; 34: 349.
7. Frost DO, Cerceo S, Carroll C, Kolb B. Early exposure to
Haloperidol or Olanzapine induces long-term alterations of
dendritic form. Synapse 2009; 64: 191–9.
8. Mychasiuk R, Ilnytskyy S, Kovalchuk O, Kolb B, Gibb R.
Intensity matters: brain, behaviour and the epigenome of
prenatally stressed rats. Neuroscience 2011; 180: 105–10.
9. Kolb B, Elliott W. Recovery from early cortical damage in
rats. II. Effects of experience on anatomy and behavior fol-
lowing frontal lesions at 1 or 5 days of age. Behav Brain Res
1987; 26: 47–56.
10. Comeau W, Gibb R, Hastings E, Cioe J, Kolb B. Therapeu-
tic effects of complex rearing or bFGF after perinatal frontal
lesions. Dev Psychobiol 2008; 50: 134–46.
11. Comeau W, Hastings E, Kolb B. Differential effect of pre
and postnatal FGF-2 following medial prefrontal cortical
injury. Behav Brain Res 2007; 180: 18–27.
12. Monfils MH, Driscoll I, Vandenberg PM, et al. Basic fibro-
blast growth factor stimulates functional recovery after neo-
natal lesions of motor cortex in rats. Neuroscience 2005; 134:
1–8.
13. Monfils MH, Driscoll I, Kamitakahara H, et al. FGF-2-
induced cell proliferation stimulates anatomical, neurophysi-
ological, and functional recovery from neonatal motor cortex
injury. Eur J Neurosci 2006; 24: 739–49.
14. Monfils M-H, Driscoll I, Vavrek R, Kolb B, Fouad K. FGF-
2 induced functional improvement from neonatal motor cor-
tex injury via corticospinal projections. Exp Brain Res 2008;
185: 453–60.
15. Nemati F, Kolb B. Motor cortex injury has different behav-
ioral and anatomical effects in early and late adolescence.
Behav Neurosci 2010; 124: 612–22.
16. Williams PT. Factors affecting recovery from hypoxic–ische-
mic injury in infant rats. [PhD thesis]. Lethbridge: University
of Lethbridge, 2010.
17. McKenna JE, Brown RW, Kolb B, Gibb R. The effects of
prental nicotine exposure on recovery from perinatal frontal
cortex lesions. Soc Neurosc Abstr 2000; 26: 653.
18. Halliwell C, Comeau W, Gibb R, Frost DO, Kolb B. Fac-
tors influencing frontal cortex development and recovery
from early frontal injury. Dev Neurorehabil 2009; 12: 269–
78.
19. Flores C, Stewart J. Changes in astrocytic basic fibroblast
growth factor expression during and after prolonged expo-
sure to escalating doses of amphetamine. Neuroscience 2000;
98: 287–93.
20. Monfils H-H, Driscoll I, Melvin N, Kolb B. Differential
expression of basic fibroblast growth factor in developing rat
brain. Neuroscience 2006; 141: 213–21.
21. Kolb B, Morshead C, Gonzalez C, Kim N, Shingo T,
Weiss S. Growth factor-stimulated generation of new corti-
cal tissue and functional recovery after stroke damage to
the motor cortex of rats. J Cereb Blood Flow Metab 2007;
27: 983–97.
22. Yazaki-Sugiuama Y, Kang S, Cateau H, Fukai T, Hensch
TK. Bidirectional plasticity in fast-spiking GABA circuits by
visual experience. Nature 2009; 462: 218–21.
23. Morishita H, Hensch TK. Critical period revisited: impact
on vision. Curr Opin Neurobiol 2008; 18: 101–7.
24. Brown RW, Kolb B. Nicotine sensitization increases den-
dritic length and spine density in the nucleus accumbens and
cingulate cortex. Brain Res 2001; 899: 94–100.
25. Gibb RL, Kovalchuk O, Halliwell C, Kolb B. Prenatal stress
affects recovery and DNA methylation patterns after postna-
tal day 10 frontal cortex lesions in rats. Soc Neurosci Abstr
2006; 32: 832.
26. Krubitzer L, Kaas J. The evolution of the neocortex in mam-
mals: how is phenotypic diversity generated? Curr Opin Neu-
robiol 2005; 15: 444–53.
27. Shaw CA, McEachern JC, editors. Toward a Theory of Neu-
roplasticity. Philadelphia: Taylor & Francis, 2001.
28. Saucier DM, Yager JY, Armstrong EA. Housing environ-
ment and sex affect behavioral recovery from ischemic brain
damage. Behav Brain Res 2010; 214: 48–54.
29. Chou IC, Trakht T, Signori C, et al. Behavioral ⁄ environ-
mental intervention improves learning after cerebral
hypoxia–ischemia in rats. Stroke 2001; 32: 192–7.
30. Dabydeen L, Thomas JE, Aston TJ, Hartley H, Sinha SK,
Eyre JA. High-energy and -protein diet increases brain and
corticospinal tract growth in term and preterm infants after
perintal brain injury. Pediatrics 2008; 121: 148–56.

8 Developmental Medicine & Child Neurology 2011, 53 (Suppl. 4): 4–8