Child Neuropsychology, 13: 494509, 2007http://www.psypress.com/childneuropsychISSN: 0929-7049 print / 1744-4136 onlineDOI: 10.1080/09297040601114878

2007 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business

LANGUAGE OUTCOME AFTER PERINATAL STROKE: DOES SIDE MATTER?

Angela O. Ballantyne,1 Amy M. Spilkin,1 and Doris A. Trauner21University of California, San Diego, School of Medicine, Departments ofNeurosciences, La Jolla, CA, 2University of California, San Diego, School ofMedicine, Department of Neurosciences and Pediatrics, La Jolla, CA, USA

The goal of this study was to examine structured language skills in children with perinatalstrokes. Participants were 28 school-age children with early focal brain lesions (17 with lefthemisphere [LH] damage, 11 with right hemisphere [RH] damage), and 57 controls. Astandardized test of language (Clinical Evaluation of Language FundamentalsRevised)was administered. Receptive, Expressive, and Total Language scores, as well as subtestscores, were analyzed. Control participants scored within the normal range, whereas theLH and RH groups scored significantly more poorly than did controls. There were no dif-ferences between the LH and RH groups on any of the language scores, and all scores werebelow the 14th percentile. Within the lesion group as a whole, scores were not related tolesion laterality, site, or severity. Results also were not accounted for by socioeconomic sta-tus or IQ. However, children who experienced seizures demonstrated significantly poorerperformance than did children who did not experience seizures. Damage to either the LH orRH early in development adversely affects later language abilities, particularly on taskswith structured and complex linguistic demands. Although lesion side has little effect, thepresence or absence of seizures is a major contributor to language outcome.

INTRODUCTION

Since Brocas seminal observation more than 100 years ago that aphasia was associ-ated with a large left hemisphere (LH) stroke (Broca, 1865), it has been well documentedthat LH strokes in adults typically result in aphasia, whereas right hemisphere (RH)strokes are likely to spare propositional language (Benson, 1986, 1993; Broca, 1865;Shallice, 1988). During the period of active brain development, however, the results ofunilateral brain injury may be quite different. Children with perinatal (28 weeks gestationto 28 days after birth) focal brain damage may demonstrate language delays early indevelopment, regardless of the hemisphere involved (Bates, 1997, 1999; Bates, Dale, &Thal, 1995; Marchman, Miller, & Bates, 1991; Nass, 1997; Reilly, Bates, & Marchman,1998; Stiles, 1995; Thal et al., 1991). Under the age of 5 years, children with perinatal LH

Our sincerest appreciation is extended to all of the children and their parents whose participation wasvital to this research. This project was funded by the National Institutes of Health (NINDS 5-P50-NS22343).

Address correspondence to Angela Ballantyne, Ph.D., University of California, San Diego, Department ofNeurosciences, 9500 Gilman Dr., #0935, La Jolla, CA 92093-0935, USA. E-mail: aballant@crl.ucsd.edu

LANGUAGE OUTCOME AFTER PERINATAL STROKE 495

lesions are likely to have particular expressive language deficits, whereas children withRH lesions are more likely to have comprehension or receptive language deficits (Bates,1992). In general, however, children with LH or RH lesions go on to attain good func-tional language abilities (i.e., these children use language in the context of everyday com-munication without marked impairment) by school age (Aram, 1988; Aram & Ekelman,1987; Bates, Thal et al., 1997; DallOglio, Bates, Volterra, DiCapua, & Pezzini, 1994;Kiessling, Denckla, & Carlton, 1983; Riva & Cazzaniga, 1986; Riva, Cazzaniga, Pantaleoni,Milani, & Fredrizzi, 1986; Stiles, 1995). These findings are consistent with the concept ofneural plasticity, the ability of the brain to reorganize in the face of damage (Bates, 1998,1999; Lennenberg, 1967; Stiles, 1995; Thal et al., 1991).

Despite reasonable functional recovery of language by the school-age years, is therea linguistic price to be paid for damage to part of the brain early in development? Whenchildren with focal brain damage are challenged with more complex language tasks (e.g.,structured question-and-answer formats; formal aspects of language form and content,including word meanings, word and sentence structure, and recall and retrieval), are theyable to perform at an age-appropriate level? There have been relatively few systematicstudies of complex language in the school-age years. Aram and Ekelman (1987) foundpersisting language deficits in school-age children with either LH or RH lesions on a testof auditory comprehension, though they speculated that impairments may also have beenrelated to memory or attentional problems. Kiessling et al. (1983) found preserved basiclanguage skills in children with hemiplegic cerebral palsy, but impairments in high-level language in children with right hemiparesis, inferring that left hemisphere braindamage produced these findings. Studies of narrative discourse in children with early-onset stroke have found evidence of less competent language skills into the school yearsfor children with either LH or RH lesions, as compared with controls (Chapman, Max,Gamino, McGlothlin, & Cliff, 2003; Reilly et al., 1998). As with other studies, however,these differences were relatively mild and the children were able to engage in narrativediscourse to some extent.

The relationship between functional language abilities and the potentially interven-ing variables of lesion laterality, site, or severity are not well established in children withearly onset stroke (Aram, Ekelman, Rose, & Whitaker, 1985; Bates, Reilly et al., 2001;Bates & Roe, 2001; Stiles, 1995). For example, some studies that examine lesion lateralityin school-age children indicate that there are no LH versus RH differences on any lan-guage measures (Bates et al., 2001; Dick et al., 1999; Kempler, van Lancker, Marchman, &Bates, 1999), whereas other studies have found hemisphere-specific differences in lan-guage performance (Riva et al., 1986; Woods & Teuber, 1978). There have also beenmixed findings regarding the role of lesion severity on cognitive outcome, with some stud-ies finding a relationship between larger lesion size and greater impairment (Montour-Proulx et al., 2004; Vargha-Khadem, Isaacs, & Muter, 1994; Vargha-Khadem, Isaacs, VanDer Werf, Robb, & Wilson, 1992), another hypothesizing a U-shaped curve with thesmallest and largest lesions leading to better outcome than intermediate size lesions(Bates, 1997), and yet other studies finding no relationship between lesion size and level ofimpairment (Ballantyne, Scarvie, & Trauner, 1994; DallOglio et al., 1994; Vargha-Khadem, OGorman, & Watters, 1985). In light of the inconsistency of previous findings,there is as yet no clear-cut relationship between language after early stroke and the predictorvariables of lesion laterality, site, or severity. Certainly, differences in subject-selectioncriteria, numbers of children included in the study, and the language or cognitive measuresused may contribute to this lack of clarity.

496 A. O. BALLANTYNE ET AL.

To better delineate the potentially adverse consequences of early unilateral brain dam-age on language outcome, the present study examined expressive and receptive languageabilities in a large group of school-age children who had experienced perinatal unilateralstrokes. We used a standardized test of complex language (Clinical Evaluation of LanguageFundamentalsRevised; CELF-R; Semel, Wiig, & Secord, 1987) and analyzed the results inrelation to other cognitive variables (Wechsler IQ, single-word vocabulary), lesion character-istics (side, size, site, severity), neurological characteristics (seizure status), and age at testing.

MATERIALS AND METHODS

Subjects

Participants were 28 children with focal brain lesions (FL) (17 with LH damage, 11with RH damage) and 57 controls. Participants ranged in age from 5 years 0 months to 16years 6 months, with the mean ages of the LH group, RH group, and controls being 6 years10 months, 9 years 1 month, and 7 years 8 months, respectively. See Table 1 for demo-graphic information on the FL and control groups.

Only participants with single, unilateral lesions that were the result of perinatalstrokes (occurrence between 28 weeks gestation and 28 days after birth; infarct or hemor-rhage) were included in the lesion group. Children with bilateral, generalized, or evolving(e.g., tumor) brain damage, as well as those with traumatic brain injuries, were excludedfrom the study. Each child was identified as having a perinatal, unilateral focal lesion byreview of medical history and neuroimaging results, including Magnetic Resonance Imag-ing (MRI) or Computed Tomography (CT). A clinical neuroradiologist, blinded to subjectstatus, performed a clinical assessment of each neuroimaging study, providing documen-tation of lesion location and lesion severity. Severity was rated qualitatively on a 5-pointscale (adapted from Vargha-Khadem et al. [1985]), with 1 being the smallest lesion and 5being a lesion involving multiple lobes (see Table 2). Control children were recruited from

Table 1 Summary of Demographic Variables for the Focal Lesion and Control Groups.

Focal Lesion (n=28) Control (n=57)

Mean Age at Testing 7 yrs. 9 mos. 2 yrs. 4 mos. 7 yrs. 8 mos. 2 yrs. 5 mos.Mean Socioeconomic Status* 1.79 .92 1.65 .86Sex 16 males, 12 females 24 males, 33 femalesHandedness 15 left, 13 right 7 left, 50 right

*Based on the Hollingshead Four Factor Index of Social Status (Hollingshead, 1975), with 1 being the highestand 5 being the lowest socioeconomic status.

Table 2 Grading System for Rating Severity of the Focal Brain Lesion.

Rating Lesion Description

1 Focal ventricular dilation or atrophy seen on 3 cuts on CT or MRI3 Focal porencephaly involving one lobe only, 3 cuts on CT or MRI5 Porencephaly or cortical atrophy involving multiple lobes

Note: CT, computed tomography; MRI, magnetic resonance imaging.

LANGUAGE OUTCOME AFTER PERINATAL STROKE 497

the community through advertisements and from the University of California, San Diego(UCSD) subject pool. Complete medical and family histories were obtained for all chil-dren. All controls included in the study had normal medical and developmental historiesand scored within normal limits on measures of cognition, language, and academic skills.For all participants, information on socioeconomic status (SES) was obtained using theHollingshead Four Factor Index of Social Status (Hollingshead, 1975).

Procedure and Measures

Participants were tested as part of a larger, 20-year longitudinal study of cognitive,spatial, and linguistic development in normal and neurologically impaired children. TheClinical Evaluation of Language FundamentalsRevised (CELF-R) (Semel et al., 1987)was administered to all participants during a 10-year period (1991 to 2001) for continuityof longitudinal data collection. The CELF-R is a standardized test that comprehensivelyassesses receptive and expressive language skills, including competency in word mean-ings (semantics), word and sentence structure (morphology and syntax), and recall andretrieval of spoken language (memory), in children ages 5 through 16 years. The test hasnine core subtests across the entire age range, for which scaled scores (Mean=10, SD=3)are calculated. The receptive subtests are: Linguistic Concepts, Sentence Structure, OralDirections, Word Classes, and Semantic Relationships. The expressive subtests are: WordStructure, Formulated Sentences, Recalling Sentences, and Sentence Assembly. ReceptiveLanguage, Expressive Language, and Total Language Standard Scores (Mean=100,SD=15) are derived. See Table 3 for a description of each subtest and the language skillarea assessed, as described by the CELF-R Technical Manual (Semel et al., 1987).

Table 3 CELF-R Core Subtests, Task Requirements, and Language Skill Areas (Semel et al., 1987).

Scale/Subtest Task Requirement Language Skill Area

ReceptiveLinguistic Concepts Interpret oral directions which contain

linguistic concepts requiring logical oper-ations such as and, eitheror, and ifthen.

Content (Semantics)

Word Classes Perceive the associative relationships between word concepts.

Content (Semantics)

Sentence Structure Knowledge of structural rules at the sentence level.

Form (Syntax and Morphology)

Semantic Relationships Interpret different semantic relationships in sentences.

Content and Form Integrated (Semantics, Syntax, and Morphology)

Oral Directions Interpret, recall, and execute oral commands of increasing length and complexity.

Memory

ExpressiveWord Structure Knowledge of morphological rules. Form (Syntax and Morphology)Formulated Sentences Formulate compound and complex

sentences.Content and Form Integrated (Semantics, Syntax, and Morphology)

Sentence Assembly Assemble syntactic structures into grammatically and semantically accept-able sentences.

Content and Form Integrated (Semantics, Syntax, and Morphology)

Recalling Sentences Recall and reproduce sentence surface structure of varying length and syntactic complexity.

Memory

498 A. O. BALLANTYNE ET AL.

In addition, all FL and control children received the age-appropriate Wechsler Intel-ligence Scale (WPPSI-R, WISC-R, or WISC-III; Wechsler, 1974, 1989, 1991), all FLchildren received the Peabody Picture Vocabulary TestRevised (PPVT-R; Dunn &Dunn, 1981), and 24 FL children received the Expressive One-Word Picture VocabularyTestRevised or Upper-Extension (EOWPVT-R or EOWPVT-UE; Gardener, 1983, 1990).

Informed consent was obtained prior to testing each participant in accordance withInstitutional Review Board procedures at the University of California, San Diego.

Statistical Analyses

Potential group differences (FL versus control) on the demographic variables of ageat testing, socioeconomic status, and sex were analyzed using t tests. An Analysis of Vari-ance framework (ANOVA and ANCOVA) was used to analyze the CELF-R data. IQ wasassessed as a relevant covariate in all analyses. In the overall between-group analyses, thedependent variables were CELF-R test scores and the independent variables were groupmembership (FL versus control; LH versus RH). Profile analyses were used to examinesubtest profiles (in terms of parallelism, levels, and flatness) for the LH, RH, and controlgroups. Follow-up subtest analyses were conducted using t tests. In an attempt to furthercharacterize the strengths and weakness in the performance of the FL group, the subtestswere examined to determine the ranges in which mean scores fell (e.g., average, belowaverage, etc.). To assess potentially intervening variables, correlational analyses wereused to examine relationships between age at testing and CELF-R Receptive, Expressive,and Total Language Scores; and t tests were used to examine the effect of lesion severity(i.e., single lobe versus multiple lobe involvement) on these CELF-R scores. The effect ofseizures (seizures versus no seizures), lesion site (frontal, temporal, parietal, occipital,subcortical), and lesion side and site combined (e.g., left temporal versus all other lesions)on CELF-R scores were analyzed in a between-subjects ANOVA framework. The rela-tionship between seizure status and lesion severity was assessed using a chi-square analy-sis. Within group analyses were used to compare scores within the LH and RH groupson the CELF-R and measures of single-word vocabulary. For all quantitative analyses, astatistical significance value of p < .05 was used.

RESULTS

Table 4 summarizes lesion and demographic information, as well as language(CELF-R Receptive, Expressive, and Total Language scores) and IQ (Full Scale IQ) per-formance for all participants in the Focal Lesion group.

Comparison of FL and Control Groups on the CELF-R

The focal lesion and control participants were not significantly different on thedemographic variables of age at testing, socioeconomic status, or sex (see Table 1). Asshown in Table 5, the control group was within the normal range for Receptive, Expres-sive, and Total Language scores. The FL group as a whole, as well as the LH and RHlesion subgroups, performed in the low average to below average range, and their meanscores were below the 14th percentile on all three indices. All of the lesion groups (FL,LH, RH) performed significantly more poorly than did controls and there were no significantdifferences between the LH and RH groups on any of the CELF-R indices (see Table 5).

499

Tabl

e 4

Foca

l Les

ion

Dem

ogra

phic

Info

rmat

ion,

Le

sion

/Neu

rolo

gica

l Inf

orm

atio

n,

and

Lang

uag

e an

d IQ

Te

st S

core

s.

Subje

ct/S

ex

Age

at

Te

stin

g (ye

ars)

Ethn

icity

SES

Han

d Pr

efer

ence

Ges

tatio

nal

Age

(w

eeks

)Le

sion

Late

ralit

yLe

sion

Loca

tion

Lesio

n

type

Seve

rity

Seiz

ures

Hem

ipar

esis

RLS

ELS

TLS

FSIQ

1F5

Cauca

sian

1L

40L

SB

asal

gan

glia

in

farc

t2

NR

101

9598

108

2M5

Cauca

sian

2L

40L

T-P-

OPo

ren

ceph

aly

5Y

R83

7376

633M

5Ca

uca

sian

1L

36L

P-S

Glio

sis3

NR

115

9510

512

34M

5Ca

uca

sian

2L

34L

F-T-

P-O

-SIn

farc

t5

NR

8376

7794

5F6

Cauca

sian

1L

39L

F-T-

P-O

-SH

emi-a

troph

y5

NR

9184

8682

6M6

Cauca

sian

2L

40L

F-T-

PPo

rence

phal

y5

YR

7262

6581

7M6

Cauca

sian

1L

30L

F-S

Cyst

ic

ence

phal

omal

acia

3N

N91

8486

115

8F7

Cauca

sian

1L

39L

F-T-

P-O

Hem

i-atro

phy

5N

R89

8686

102

9F7

Cauca

sian

2R

40L

P-O

Infa

rct

5N

N80

5467

6710

F7

Cauc

asia

n1

L40

LT

Foca

l vol

ume

loss

1N

R10

186

9310

111

F7

Hisp

anic

4R

40L

F-S

Infa

rct

4Y

N72

5061

6812

F7

Asia

n1

L37

LF-

T-P-

O-S

Infa

rct

5Y

R87

7378

8213

M7

Hisp

anic

3L

42L

F-T-

SIn

farc

t5

NR

5050

5040

14M

7Ca

uca

sian

1L

40L

TPo

rence

phal

y4

NR

7673

7310

515

M8

Cauca

sian

2R

42L

F-P-

O-S

Pore

nce

phal

y5

NN

9380

8596

16F

9Ca

uca

sian

4L

40L

P-S

Infa

rct

4Y

R50

5450

5917

M9

Cauca

sian

1L

40L

T-P-

SPo

rence

phal

y5

NN

8067

7211

318

M5

Cauca

sian

1R

38R

F-T-

P-O

-SIn

farc

t5

YL

9170

7886

19M

5Ca

uca

sian

2R

39R

F-T-

P-O

-SIn

farc

t5

YL

7462

6674

20F

6Ca

uca

sian

1R

41R

F-P-

OPo

rence

phal

y5

YL

7262

6557

21F

7Ca

ucas

ian

2R

39R

T-P-

SIn

farc

t5

NL

122

130

131

123

(Co

ntin

ued

)

500

Tabl

e 4

(Con

tinued

)

Subje

ct/S

ex

Age

at

Te

stin

g (ye

ars)

Ethn

icity

SES

Han

d Pr

efer

ence

Ges

tatio

nal

Age

(w

eeks

)Le

sion

Late

ralit

yLe

sion

Loca

tion

Lesio

n

type

Seve

rity

Seiz

ures

Hem

ipar

esis

RLS

ELS

TLS

FSIQ

22M

7Ca

ucas

ian/

Hisp

anic

2R

34R

T-P

Infa

rct

5Y

N74

7372

80

23F

9Ca

uca

sian

2R

40R

F-T-

P-O

Infa

rct,

cysti

c en

ceph

alom

alac

ia5

YL

6364

6176

24F

11Ca

uca

sian

3L

40R

T-P-

SH

emor

rhag

e5

NN

9584

8912

125

M11

Cauca

sian

1R

38R

F-P

Pore

nce

phal

y5

NL

8372

7510

926

M11

Cau

casia

n2

R40

RS

Inte

rnal

cap

sule

gl

iosis

2Y

N74

7673

81

27M

12Ca

uca

sian

3R

42R

F-P-

SIn

farc

t5

NL

9576

8410

228

M13

Cauca

sian

1R

28R

F-S

Hem

orrh

age

4N

N54

5452

94

Abb

revi

atio

ns. F

=Fe

mal

e; M

=M

ale;

SES

=So

cio

econom

ic st

atus

(1=

high

est,

5=lo

wes

t); R

=R

ight

; L=

Left;

F=

Fronta

l; T

=Te

mpo

ral;

P=Pa

rieta

l; O

=O

ccip

ital;

S=Su

bcort

ical

;Se

verit

y (1

=sm

alle

st, 5

=la

rges

t); Y

=Y

es; N

=N

o; R

LS=

CELF

-R

Rec

eptiv

e La

ngu

age

Stan

dard

Sco

re; E

LS=

CELF

-R

Exp

ress

ive

Lang

uage

Sta

nda

rd Sc

ore

; TLS

=CE

LF-R

To

tal

Langu

age

Stan

dard

Sco

re; F

SIQ=

Full

Scal

e IQ

.

LANGUAGE OUTCOME AFTER PERINATAL STROKE 501

Full Scale IQ (FSIQ) was assessed as a covariate for all between-group analyses, and itcontributed a significant portion of variance to CELF-R scores (FSIQ: LH=88.2 23,RH=91.2 21, Control=118.3 12). Nonetheless, when FSIQ was partialled out all sig-nificant differences between the lesion groups (FL, LH, or RH) and control groupremained. Due to the fact that the analysis of CELF-R scores with IQ as a covariate issomewhat artificial (i.e., the child is cognitively and linguistically a unitary whole), andsignificant differences remained even after covarying for IQ, the results presented belowrepresent performance without IQ as a covariate.

The profiles of CELF-R subtest scores for the LH, RH, and control groups areshown in Figure 1. Profile analyses indicated that the subtest profiles for the controls, LH,and RH groups were statistically parallel. Moreover, the LH and RH groups performed ata significantly lower level than did controls (p < .0001), and the LH and RH groups didnot differ in level of performance. Follow-up analyses indicated that the LH and RHgroups did not significantly differ on any of the receptive or expressive subtests. Lastly,although the profiles were not flat, indicating some variability across the subtests withineach group, the small magnitude of the differences between subtests was not significant

Table 5 Mean CELF-R Receptive, Expressive, and Total Scores for the Focal Lesion (FL), Left Hemisphere(LH), Right Hemisphere (RH), and Control Groups.

CELF-R Index

FL* (n=28) LH (n=17) RH (n=11) Controls (n=57)

M SD p** M SD p** M SD p** M SD

Receptive 82.54 17.12

502 A. O. BALLANTYNE ET AL.

(i.e., for each group [LH, RH] each subtest mean was not significantly different from themean of all subtest scores for the group [Semel et al., 1987]).

Although each subtest mean was not significantly different from the mean of allsubtest scores for each group, each subtest mean for the FL group was qualitatively exam-ined to determine on which subtests the FL group performed below average (i.e., below ascaled score of 8). The means revealed that the FL group performed below average on thefollowing subtestsReceptive: Word Classes, Semantic Relationships, and Oral Direc-tions; Expressive: Word Structure, Formulated Sentences, Sentence Assembly, andRecalling Sentences. The FL group scored within the average range on only two subtestsof the CELF-R and both were on the Receptive Language Scale (Linguistic Concepts andSentence Structure).

Examination of Potential Intervening Variables

Although this study was cross-sectional, there were no significant correlationsbetween test performance and age (i.e., scores did not systematically increase or decreaseover age). There were no differences on any of the CELF-R indices as a result of severityof the lesion (single lobe versus multiple lobes). These results were found in both the FLgroup as a whole, as well as in the LH and RH groups individually.

When the FL group was divided into seizure (those participants who had a history ofseizures beyond the neonatal period) and non-seizure (those participants who had neverhad a seizure or who had seizures in the neonatal period only) subgroups, a salient patternemerged. Figure 2 presents the effect of seizure status on performance within the focallesion group. Children in the seizure group demonstrated significantly poorer performancethan did children in the non-seizure group (Receptive p=.027, Expressive p=.031, TotalLanguage p=.021). However, children in the non-seizure group still performed signifi-cantly more poorly (p < .0001 for all indices) than did controls. Chi-square analysis

Figure 2 CELF-R Receptive, Expressive, and Total scores for the Seizure, Non-seizure, and Control groups.Note: On all CELF-R indices, the Seizure group performed significantly more poorly than the Non-seizure groupand the Control group; the Non-seizure group performed significantly more poorly than the Control group andsignificantly better than the Seizure group.

LANGUAGE OUTCOME AFTER PERINATAL STROKE 503

revealed no significant relationship between seizure status and lesion severity (single lobeversus multiple lobes), thus the variable of seizure status is not merely a proxy for lesionseverity.

Within the FL group, analyses of lesion site (frontal, temporal, parietal, occipital,subcortical) indicated no significant relationship between site and CELF-R performance.Due to the fact that site specific deficits may also be related to the side of the lesion, anal-yses of side and site combined (e.g., left temporal lesion versus all other lesions, left tem-poral lesion versus left nontemporal lesions) were performed; these analyses also did notindicate any relationship with CELF-R scores.

Supplemental Analysis of Single-Word Vocabulary in the Lesion Group

The lesion group performed in the average range on tests of single-word receptiveand expressive vocabulary (PPVT-R Mean=94.6 22.9 and EOWPVT-R/UE Mean=93.7 23.7). The PPVT-R (receptive test) was compared with the CELF-R Receptive Languagescore, and the EOWPVT-R/UE (expressive test) was compared with the CELF-R Expres-sive Language score in the lesion group. Within the lesion group, performance on theCELF-R was significantly poorer than performance on the single-word language tests, asshown in Figure 3 (Receptive p=.002, Expressive p < .0001).

DISCUSSION

The present study examined complex language skills in school-age children withperinatal unilateral brain damage using a structured, standardized language task. Resultsindicate that regardless of the side of the lesion, deficits in complex language abilities arepresent in the school-age years. Children with focal lesions performed in the low averageto below average range on a standardized test of language, the CELF-R, with meanReceptive, Expressive, and Total Language scores all below the 14th percentile (i.e., below

Figure 3 A comparison of single-word vocabulary versus complex language in children with focal lesions (FL).Note: Scores above dotted line represent performance within normal limits.

504 A. O. BALLANTYNE ET AL.

normal limits). Previous seminal studies of Bates and others (Bates et al., 2001; Bateset al., 1997) demonstrated that in free speech situations, children with FL had delays in the veryearly stages of language development, but normal or near-normal language by school-age(Bates et al., 2001; Bates & Roe, 2001; Stiles, 1995). The results of our study indicate thatschool-age children with FL may have intact basic abilities (e.g., single-word vocabulary),but they are notably impaired when required to meet the demands of more structured lan-guage tasks in both the receptive and expressive domains. These results have implicationsfor the ability of the child to function in everyday life, particularly since linguisticdemands increase over the school-age years. Children with early focal brain damage mayface challenges in situations such as remembering or carrying out multi-step verbal instruc-tions, interpreting nuances in conversation, and being able to adequately comprehend and/or express complex thoughts and ideas. Importantly, due to intact basic language skillsand functional abilities in free-speech situations, these deficits may go unrecognized or beattributed to laziness, oppositional behavior, and/or attention deficits.

The similarity in language profiles between the LH and RH groups is in contrast tothe laterality of language abilities that is evident after unilateral strokes in adults. A strik-ing finding of the present study is the extent to which the LH and RH lesion groups aresimilar on structured language tasks. Typical development leads to the neuroanatomicaldistribution of language systems predominantly in the left hemisphere (Benson, 1986,1993; Broca, 1865; Shallice, 1988). Therefore, it could be hypothesized that early LHdamage would produce greater language impairment than RH damage. However, wefound no differences in complex language performance based on the hemispheric side ofthe lesion. Similar performance between the LH and RH groups was seen across all indi-ces studied (CELF-R Receptive, Expressive, and Total Language Standard Scores, as wellas all individual subtest scores). This is consistent with other research that indicates left-right differences may be evident very early in language development but disappear in sub-sequent years (Vicari et al., 2000). In terms of subtest profiles, the LH and RH groupsdemonstrated statistically equivalent subtest performances. Although the profiles were notflat, indicating some variability across the subtests within each group, care must be takento not overinterpret small subtest differences. All children demonstrate variability acrossCELF-R subtests and, moreover, these subtle differences may be due to measurementerror (Semel et al., 1987). More striking than any specific subtest differences in the FLgroup is that the group consistently performed between 1 and 2 standard deviations belowthe mean on all subtests of the CELF-R (with the exception of the Linguistic Concepts andSentence Structure subtests).

Examination of the performance of the FL group across the CELF-R subtests indi-cates that they performed below average on all expressive language tasks, encompassingthe skill areas of language content (semantics), form (syntax and morphology), and mem-ory. In the receptive domain, the FL group performed below average on tasks encompass-ing the skill areas of content (semantics), content and form integrated (semantics, syntax,and morphology), and memory. The only two receptive subtests on which the FL groupperformed within normal limits were Linguistic Concepts, which assesses content in avery concrete format (in contrast to Word Classes, which assesses content in a much moreabstract format), and Sentence Structure, which assesses form. Therefore, the only twosubtests on which the FL group performed within normal limits were among the most con-crete and uncomplicated receptive tasks. When the receptive tasks became more compli-cated and integrated aspects of form, content, and memory, as well as for all expressivetasks, the FL group demonstrated difficulty.

LANGUAGE OUTCOME AFTER PERINATAL STROKE 505

Our current findings indicate that when faced with challenging linguistic tasks, chil-dren with focal brain damage (regardless of hemisphere) demonstrate performance that isbelow normal limits. When language is assessed with a simple, single-word vocabularytest (receptive or expressive), the focal lesion group performs in the average range. Whatare the neural implications of these findings? Children with focal brain damage have beennoted to demonstrate some degree of neural plasticity (Bates et al., 2001; Stiles, 2000;Stiles, Reilly, Paul, & Moses, 2005) and brain reorganization (Mueller et al., 1998; Rutten,Ramsey, van Rijen, Franssen, & van Veelen, 2002; Staudt, Krageloh-Mann, Holthausen,Gerloff, & Grodd, 2004; Stiles et al., 2003). Our findings, however, suggest there are lim-its to the capacity of the brain to recover after early damage. Perhaps when damaged earlyin development, the brain is able to reorganize and find alternate pathways for language toemerge, but these alternate pathways may not be as efficient as the original brain plan.Therefore deficits may emerge when the system is challenged by complex linguistic tasks.Our findings are consistent with other studies that have found school-age LH and RHlesion subjects to have language deficits, particularly on challenging or complex tasks(e.g., Chapman et al., 2003; Kiessling et al., 1983; MacWhinney, Feldman, Sacco, &Valdes-Perez, 2000; Reilly et al., 1998).

In addition to examining language performance in children with focal lesions, thisstudy aimed to evaluate the possible multi-factorial nature of performance level and out-come. Typical lesion characteristics such as laterality, site, or severity were not predictiveof language performance in our focal lesion participants, nor were results attributable toIQ or socioeconomic status. The fact that language performance in the early-onset focallesion population was not related to these intervening variables is consistent with otherresearch (Bates, 1997; Bates et al., 2001; Chapman et al., 2003; DallOglio et al., 1994;Kempler et al., 1999; Max, 2004; Reilly et al., 1998). Whereas the laterality of the lesionhas traditionally been viewed as paramount (likely due to the significance of lesion lateral-ity in adults with stroke), we found that, in children with early focal brain damage, morerelevant than lesion side was seizure status.

Within the focal lesion group, a notable pattern emerged such that children who hada history of seizures performed more poorly than did children with a history of no seizuresor neonatal seizures only. These data provide strong evidence for the detrimental effects ofseizure activity on cognitive functioning. It is widely known that seizure activity mayadversely affect cognitive functions in children without focal brain damage (Besag, 1988;Corbett, 1989; Masur & Shinnar, 1992; Rodin, Schmaltz, & Twitty, 1986; Seidenberg &Berent, 1992; Strauss et al., 1995; Trimble, 1990). Perhaps seizure activity results indiminished capacity for efficient brain reorganization in children with focal lesions. Theremay be variations in outcome that are associated with seizure focus, type, and medical/pharmacological treatment, and future research can serve to clarify these issues and canserve as the impetus for more effective treatments.

These findings highlight the importance of seizure status in the early FL population,and the necessity of taking this influential variable into account before assuming differ-ences are caused by other lesion characteristics such as lesion side or site. Although sei-zure status is not considered in many FL studies, some researchers have also noted theimportance of taking seizure status into account when interpreting the effect of lesion siteon behavioral outcome (DallOglio et al., 1994).

The present study indicates that IQ plays a significant role in language compe-tence, as would be expected. However, when the effect of IQ was removed, childrenwith focal brain lesions still performed significantly more poorly than did typically

506 A. O. BALLANTYNE ET AL.

developing children. Thus, language problems in the focal lesion population go beyondgeneral cognitive ability and relate specifically to the functional use of language. Hencethe assessment of IQ alone is not an adequate predictor of language abilities in thesechildren.

Interestingly, in this cross-sectional sample, there was no relationship betweenCELF-R language performance and age. Hence there was no evidence for either the LH orRH group systematically catching up or falling further behind over the school-age years.However, longitudinal studies are clearly needed to clarify any trajectories of performanceover age.

Many existing studies of complex language abilities in children with early focal braininjury have used experimental tasks of narrative discourse or online measures of languageprocessing (Bates et al., 2001; Chapman et al., 2003; Dick, Wulfeck, Krupa-Kwiatkowski, &Bates, 2004; Feldman, MacWhinney, & Sacco, 2002; MacWhinney et al., 2000; Reilly,Bates, & Marchman, 1998; Reilly, Losh, Bellugi, & Wulfeck, 2004). A noteworthy aspectto the current study is that our results corroborate previous findings (Chapman et al., 2003;Reilly et al., 1998) using a standardized measure that is easy to administer, to score, and tointerpret. Our findings parallel those of MacWhinney et al. (2000), who used four subtestsof the CELF-R (along with multiple computerized language measures) and found theCELF-R subtests to be sensitive to deficits in language functioning in a heterogeneouspopulation of children with early brain damage, albeit less sensitive than the online com-puter measures for particular subjects. In contrast to the MacWhinney study, however, thepresent study examined performance on the entire CELF-R, including all core subtests aswell as Receptive, Expressive, and Total language indices, and found it to be quite sensi-tive to language problems in the FL group. Moreover, the present study examined seizurestatus, which our results indicate is an important predictor of language outcome in chil-dren with early FL.

In conclusion, insult to either the LH or RH early in development adversely affectsperformance on structured language tasks during the school-age years, and the presence ofseizures results in even greater risk for language impairment. Although these children pos-sess normal single-word vocabulary and language that is basically functional, suggest-ing a significant amount of brain reorganization after early damage, the deficits oncomplex language tasks indicate a limit to neural plasticity. The presence of language dif-ficulties has implications that go beyond language and into the psychosocial and academicdomains. For example, studies of linguistic and affective prosody (Trauner, Ballantyne,Friedland, & Chase, 1996), behavioral functioning (Trauner, Panyard-Davis, & Ballantyne,1996), and academic skills (Ballantyne & Trauner, 1994) indicate that school-age childrenwith focal brain damage are at risk for problems in these domains. The results of thepresent study demonstrate the importance of a thorough language assessment in childrenwith early focal brain damage, in order to reduce the likelihood that potentially subtle butsignificant language deficits may go unrecognized. A lack of recognition could lead tomisplaced attributions of the childs social, academic, or behavioral difficulties to charac-teristics such as laziness, inattention, or uncooperativeness. The present study also hasimplications for the neural specificity of language in that early damage to either hemi-sphere results in similar language outcomes. It also has implications for potential limita-tions on plasticity such that no matter where the site of the lesion, the mere presence of afocal brain lesion early in life results in less than optimal language outcome. Finally, thesefindings indicate that there may be a detrimental impact of seizure activity on brain reor-ganization after early damage. Future FL studies using more in-depth psycholinguistic

LANGUAGE OUTCOME AFTER PERINATAL STROKE 507

measures, as well as volumetric brain analyses, will further our understanding of languagespecialization, and how lesion volume and seizure activity may impact brain plasticity andthe potential for reorganization in the developing brain.

Original manuscript received March 3, 2006Revised manuscript accepted November 5, 2006

First published online February 6, 2007

REFERENCES

Aram, D. (1988). Language sequelae of unilateral brain lesions in children. In F. Plum (Ed.),Language, communication, and the brain (pp. 171197). New York: Raven.

Aram, D., & Ekelman, B. L. (1987). Unilateral brain lesions in childhood: Performance on therevised token test. Brain and Language, 32, 137158.

Aram, D., Ekelman, B. L., Rose, D. F., & Whitaker, H. A. (1985). Verbal and cognitive sequelaefollowing unilateral lesions acquired in early childhood. Journal of Clinical and ExperimentalNeuropsychology, 7(1), 5578.

Ballantyne, A. O., Scarvie, K. M., & Trauner, D. A. (1994). Verbal and Performance IQ patterns inchildren after perinatal stroke. Developmental Neuropsychology, 10(1), 3950.

Ballantyne, A. O., & Trauner, D. A. (1994). Academic achievement in children after perinatal stroke[abstract]. Annals of Neurology, 36, 297.

Bates, E. (1992). Language development. Current Opinion in Neurobiology, 2, 180185.Bates, E. (1997). Origins of language disorders: A comparative approach. Developmental Neuropsy-

chology, 13, 447476.Bates, E. (1998). Plasticity, localization and language development. In S. Broman & J. M. Fletcher

(Eds.), The changing nervous system: Neurobehavioral consequences of early brain disorders(pp. 214253). New York: Oxford University Press.

Bates, E. (1999). Language and the infant brain. Journal of Communication Disorders, 32, 195205.Bates, E., Dale, P. S., & Thal, D. (1995). Individual differences and their implications for theories of

language development. In P. Fletcher & B. MacWhinney (Eds.), Handbook of child language(pp. 96151). Oxford: Basil Blackwell.

Bates, E., Reilly, J. S., Wulfeck, B., Dronkers, N., Opie, M., Fenson, J., Kriz, S., Jeffries, R., Miller,L., & Herbst, K. (2001). Differential effects of unilateral lesions on language production in chil-dren and adults. Brain and Language, 79, 223265.

Bates, E., & Roe, K. (2001). Language development in children with unilateral brain injury. In C. A.Nelson & M. Luciana (Eds.), Handbook of developmental cognitive neuroscience (pp. 281307).Cambridge, MA: MIT Press.

Bates, E., Thal, D., Trauner, D. A., Fenson, J., Aram, D., Eisele, J., & Nass, R. (1997). From firstwords to grammar in children with focal brain injury. Developmental Neuropsychology, 13,275343.

Benson, D. F. (1986). Aphasia and lateralization of language. Cortex, 22, 7186.Benson, D. F. (1993). Aphasia. In K. M. Heilman & E. Valenstein (Eds.), Clinical neuropsychology

(3rd ed.). New York: Oxford University Press.Besag, F. M. C. (1988). Cognitive deterioration in children with epilepsy. In M. R. Trimble & E. H.

Reynolds (Eds.), Epilepsy, behavior and cognitive function (pp. 113127). New York: JohnWiley and Sons.

Broca, P. (1865). Sur la faculte du langage articule. Bulletins et Memoires de la Societe DAnthro-pologie de Paris, 6, 377393.

Chapman, S. B., Max, J. E., Gamino, J. F., McGlothlin, J. H., & Cliff, S. N. (2003). Discourse plas-ticity in children after stroke: Age at injury and lesion effects. Pediatric Neurology, 29(1), 3441.

Corbett, J. (1989). Intellectual impairment in children with epilepsy. Educational and Child Psy-chology, 6, 6973.

508 A. O. BALLANTYNE ET AL.

DallOglio, A. M., Bates, E., Volterra, V., DiCapua, M., & Pezzini, G. (1994). Early cognition, com-munication and language in children with focal brain injury. Developmental Medicine and ChildNeurology, 36, 10761098.

Dick, F., Wulfeck, B., Bates, E., Saltzman, D., Naucler, N., & Dronkers, N. (1999). Interpretation ofcomplex syntax in aphasic adults and children with focal lesions or specific language impairment[abstract]. Brain and Language, 69, 335336.

Dick, F., Wulfeck, B., Krupa-Kwiatkowski, M., & Bates, E. (2004). The development of complexsentence interpretation in typically developing children compared with children with specific lan-guage impairments or unilateral focal lesions. Developmental Science, 7(3), 360377.

Dunn, L. M., & Dunn, L. M. (1981). Peabody Picture Vocabulary TestRevised (PPVT-R). CirclePines, MN: American Guidance Service.

Feldman, H., MacWhinney, B., & Sacco, K. (2002). Sentence processing in children with unilateralbrain injury. Brain and Language, 83(2), 335352.

Gardener, M. F. (1983). The Expressive One-Word Picture Vocabulary TestUpper Extension(EOWPVT-UE). Novato, CA: Academic Therapy.

Gardener, M. F. (1990). Expressive One-Word Picture Vocabulary TestRevised (EOWVPT-R).Novato, CA: Academic Therapy.

Hollingshead, A. B. (1975). Four Factor Index of Social Status. New Haven, CT: Yale UniversityDepartment of Sociology.

Kempler, D., van Lancker, D., Marchman, V., & Bates, E. (1999). Idiom comprehension in childrenand adults with unilateral brain damage. Developmental Neuropsychology, 15, 327349.

Kiessling, L. S., Denckla, M. B., & Carlton, M. (1983). Evidence for differential hemispheric func-tion in children with hemiplegic cerebral palsy. Developmental Medicine and Child Neurology,25, 727734.

Lennenberg, E. H. (1967). Biological foundations of language. New York: Wiley.MacWhinney, B., Feldman, H., Sacco, K., & Valdes-Perez, R. (2000). Online measures of basic lan-

guage skills in children with early focal brain lesions. Brain and Language, 71, 400431.Marchman, V., Miller, R., & Bates, E. (1991). Babble and first words in children with focal brain

injury. Applied Psycholinguist, 12, 122.Masur, D. M., & Shinnar, S. (1992). The neuropsychology of childhood seizure disorders. In F. Boller &

J. Grafman (Eds.), Handbook of neuropsychology (Vol. 7, pp. 457470). New York: Elsevier.Max, J. E. (2004). Effect of side of lesion on neuropsychological performance in childhood stroke.

Journal of the International Neuropsychological Society, 10, 698708.Montour-Proulx, I., Braun, C. M., Daigneault, S., Roleau, I., Kuehn, S., & Begin, J. (2004). Predic-

tors of intellectual function after a unilateral cortical lesion: Study of 635 patients from infancy toadulthood. Journal of Child Neurology, 19(12), 935943.

Mueller, R.-A., Rothermel, R. D., Behen, M. E., Muzik, O., Mangner, T. J., Chakraborty, P. K., &Chugani, H. T. (1998). Brain organization of language after early unilateral lesion: A PET study.Brain and Language, 62, 422451.

Nass, R. (1997). Language development in children with congenital strokes. Seminars in PediatricNeurology, 4(2), 109116.

Reilly, J. S., Bates, E. A., & Marchman, V. A. (1998). Narrative discourse in children with earlyfocal brain injury. Brain and Language, 61, 335375.

Reilly, J. S., Losh, M., Bellugi, U., & Wulfeck, B. (2004). Frog, Where Are You? Narratives inchildren with specific language impairment, early focal brain injury, and Williams syndrome.Special Issue, Brain and Language, 88(2), 229247.

Riva, D., & Cazzaniga, L. (1986). Late effects of unilateral brain lesions sustained before and afterage one. Neuropsychologia, 24, 423428.

Riva, D., Cazzaniga, L., Pantaleoni, C., Milani, N., & Fredrizzi, E. (1986). Acute hemiplegia inchildhood: The neuropsychological prognosis. Journal of Pediatric Neuroscience, 2, 239250.

Rodin, E. A., Schmaltz, S., & Twitty, G. (1986). Intellectual functions of patients with childhood-onset epilepsy. Developmental Medicine and Child Neurology, 28, 2533.

LANGUAGE OUTCOME AFTER PERINATAL STROKE 509

Rutten, G. J., Ramsey, N. F., van Rijen, P. C., Franssen, H., & van Veelen, C. W. (2002). Interhemi-spheric reorganization of motor hand function to the primary motor cortex predicted with func-tional magnetic resonance imaging and transcranial magnetic stimulation. Journal of ChildNeurology, 17(4), 292297.

Seidenberg, M., & Berent, S. (1992). Childhood epilepsy and the role of psychology. American Psy-chologist, 47, 11301133.

Semel, E., Wiig, E. H., & Secord, W. (1987). Clinical Evaluation of Language FundamentalsRevised:(CELF-R). San Antonio, TX: Psychological Corporation.

Shallice, T. (1988). From neuropsychology to mental structure. Cambridge: Cambridge University Press.Staudt, M., Krageloh-Mann, I., Holthausen, H., Gerloff, C., & Grodd, W. (2004). Searching for

motor functions in dysgenic cortex: A clinical transcranial magnetic stimulation and functionalmagnetic resonance imaging study. Journal of Neurosurgery, 101(1 Suppl), 6977.

Stiles, J. (1995). Plasticity and development: Evidence from children with early occurring focalbrain injury. In B. Julesz & I. Kovacs (Eds.), Maturational windows and adult cortical plasticity(pp. 217237). Reading, MA: Addison-Wesley.

Stiles, J. (2000). Neural plasticity and cognitive development. Developmental Neuropsychology,18(2), 237272.

Stiles, J., Moses, P., Roe, K., Akshoomoff, N. A., Trauner, D. A., Hesselink, J., Wong, E. C., Frank,L. R, & Buxton, R. B. (2003). Alternative brain organization after prenatal cerebral injury: Conver-gent fMRI and cognitive data. Journal of the International Neuropsychological Society, 9(4), 604622.

Stiles, J., Reilly, J., Paul, B., & Moses, P. (2005). Cognitive development following early braininjury: Evidence for neural adaptation. Trends in Cognitive Sciences, 9(3), 136142.

Strauss, E., Loring, D., Chelune, G., Hunter, M., Hermann, B., Perrin, E., Westerveld, M., Trenerrey,M., & Barr, W. (1995). Predicting cognitive impairment in epilepsy: Findings from the BozemanEpilepsy Consortium. Journal of Clinical and Experimental Neuropsychology, 17, 909917.

Thal, D. J., Marchman, V., Stiles, J., Aram, D., Trauner, D. A., Nass, R., & Bates, E. (1991). Earlylexical development in children with early focal brain injury. Brain and Language, 40, 491527.

Trauner, D. A., Ballantyne, A. O., Friedland, S., & Chase, C. (1996). Disorders of affective and lin-guistic prosody in children after early unilateral brain damage. Annals of Neurology, 39, 361367.

Trauner, D. A., Panyard-Davis, J. L., & Ballantyne, A. O. (1996). Behavioral differences in schoolage children after perinatal stroke. Assessment, 3, 265276.

Trimble, M. R. (1990). Antiepileptic drugs, cognitive function, and behavior in children: Evidencefrom recent studies. Epilepsia, 31, S30S34.

Vargha-Khadem, F., Isaacs, E., & Muter, V. (1994). A review of cognitive outcome after unilaterallesions sustained during childhood. Journal of Child Neurology, 9(2), 6773.

Vargha-Khadem, F., Isaacs, E., Van Der Werf, S., Robb, S., & Wilson, J. (1992). Development ofintelligence and memory in children with hemiplegic cerebral palsy. Brain, 115, 315329.

Vargha-Khadem, F., OGorman, A. M., & Watters, G. V. (1985). Aphasia and handedness in rela-tion to hemispheric side, age at injury and severity of cerebral lesion during childhood. Brain andLanguage, 108, 677696.

Vicari, S., Albertoni, A., Chilosi, A. M., Cipriani, P., Cioni, G., & Bates, E. (2000). Plasticity andreorganization during language development in children with early brain injury. Cortex, 36, 3146.

Wechsler, D. (1974). Wecshler Intelligence Scale for ChildrenRevised (WISC-R). New York: Psy-chological Corporation.

Wechsler, D. (1989). Wechsler Preschool and Primary Scale of IntelligenceRevised (WPPSI-R).San Antonio, TX: Psychological Corporation.

Wechsler, D. (1991). Wechsler Intelligence Scale for ChildrenIII (WISC-III). San Antonio: Psy-chological Corporation.

Woods, B. T., & Teuber, H. L. (1978). Changing patterns of childhood aphasia. Annals of Neurol-ogy, 3, 272280.