l cul

Ireity Sc

Premature birthPreterm birthExecutive functionSocial competenceBehavior Rating Inventory of Executive FunctionPreschool

ionofteources required to meet social demands and achieve social goals, in children born

) and within a specicl competence emergesreects a dynamic in-ironment [14]. Social

Early Human Development 90 (2014) 299306

Contents lists available at ScienceDirect

Early Human D

j ourna l homepage: www.e lsevcompetence can be assessed using child interviews, direct observations,or parent and teacher questionnaires [1520]. Social competencehas im-portant implications for public health from infancy to adulthood becauseAbbreviations: BRIEF, Behavior Rating Inventory of Executive Function; BW, birthchildren with histories of prematurity [4,5,7,8], attention-decit/hyperactivity disorder (ADHD) [4,9], aggression [2] or brain injury [10].Poor EF skills are related to difculties in important areas of function, in-cluding cognitive and academic domains [2,11]. It also has been proposed

cial interaction (e.g., parentchild, peer relationscontext (e.g., home, school) ([13], page 13). Sociagradually through childhood and adolescence andterplay between the individual and his/her envtion ([1] page 1). EF is necessary for purposeful, goal-directed activi-ty [2]. The EF skills that are developing in the preschool-aged childinclude inhibitory control, set shifting, planning and working memory[36]. Decits in EF are seen in young child clinical populations, such as

Social competence is one important facet of children's socialemotional development [1]. Social competence can be dened as activeand skillful coordination of multiple processes and resources of the childtomeet social demands and achieve social goals in a particular type of so-weight; EF, executive function; GA, gestational age; GECSCBE, Social Competence and Behavior Evaluation; SD, stanomic status. Corresponding author at: Stanford University, 750 W

CA 94304, United States. Tel.: +1 650 723 5711; fax: +1E-mail address: iloe@stanford.edu (I.M. Loe).

http://dx.doi.org/10.1016/j.earlhumdev.2014.02.0110378-3782/ 2014 Elsevier Ireland Ltd. All rights reserveds higher-order cognitiveought, action, and emo-

However, published studies have focused on general social skills [1,12],and are challenging to interpret as theymeasure awide variety of psycho-social constructs spanning cooperative behavior and social functioning.Executive function (EF) has been dened aprocesses involved in the self-regulation of th1. Introductionparent-rated EF, and poorer scores on the two social competence scales. In hierarchical multiple regressionmodels, EF battery z-score and parent-rated EFmade independent contributions to bothmeasures of social com-petence. Preterm birth explained additional variance for one measure of social competence.Conclusions: Standard assessment of EF skills and social competence in young preschool children, including chil-dren born preterm,may identify at-risk children for long-term social difculties andmay also provide targets forintervention.

2014 Elsevier Ireland Ltd. All rights reserved.

that poor EF has a role in social problems, from early childhood onward.Aims: To characterize social competence and assess the relationship between EF and social competence inpreschool-aged children born preterm or full term.Study design: Cross-sectional study.Subjects: Study subjects had a history of preterm birth (34 weeks of gestation) and birth weight b2500 g (n=70). Controls were born full term (37 weeks) (n = 79).Outcome measures: Children completed a battery of EF tasks; a mean age-adjusted z-score for the battery wasgenerated for each child. Parents rated child EF on one scale and child social competence on two standardizedscales.Results: Compared to full term children, preterm children showed a lower mean EF battery z-score, poorerKeywords:well characterized.Accepted 25 February 2014 preterm. The relations between EF and social competence in preterm and full term preschoolers have not beenReceived in revised form 22 February 2014 tence, the processes and resExecutive function is associated with sociapreschool-aged children born preterm or f

Nidia Alduncin, Lynne C. Huffman, Heidi M. Feldman,Division of Neonatal and Developmental Medicine, Department of Pediatrics, Stanford Univers

a b s t r a c ta r t i c l e i n f o

Article history:Received 2 December 2013

Background: Executive functing actions and emotions, is, Global Executive Composite;ndard deviation; SES, socioeco-

elch Road, Suite 315, Palo Alto,650 725 8351.

.ompetence inl term

ne M. Loe hool of Medicine, Stanford, CA 94305, United States

(EF), dened as higher-order cognitive processes used in planning and organiz-n impaired in children born preterm. Few studies have assessed social compe-

evelopment

i e r .com/ locate /ear lhumdevit is associated with mental health, academic performance, and workplace functioning [19,21].

Studies of children born preterm have shown increased likeli-hood of adverse developmental, cognitive, and behavioral outcomes[4,5,7,8,2224]. Further, researchers have shown that, compared to

300 N. Alduncin et al. / Early Human Development 90 (2014) 299306peers born at term, youth born preterm have greater difculty withsocial relationships [22,25,26]. However, very few studies have spe-cically assessed the social competence of children born preterm[27,28], particularly at young ages. Studies have shown that EF con-tributes to social outcomes in older preterm children [29,30]. Studiesof typically developing young children suggest that emerging EFskills may facilitate children's social competence [1,3,11]. An under-standing of the associations of EF and social competence in youngpreschool children born preterm might inform intervention strate-gies to improve their EF and social competence early in life. The over-all goal of this study was to determine if EF skills play a role in thesocial competence of preschool-aged children born preterm.

Sociodemographic factors, such as socioeconomic status andage, may also affect the development of both social competence andEF. Socioeconomic status (SES) has been associated with a wide arrayof socioemotional, cognitive, and health outcomes in childhood [31],including preterm birth [32]. Poor EF skills have been associated witheconomic disadvantage in children [33,34]. A longitudinal study of pre-term and full term children followed from birth to thirty-six monthsshowed that higher SESwas predictive of better cognitive and social de-velopment for all children [35]. Age is another important factor in thedevelopment of social competence and EF as both typically improvefrom childhood through adolescence [36].

1.1. Study objective and hypotheses

The objective of our study was to characterize the relationshipbetween EF and social competence in preschool-aged children whowere born preterm or full term. We previously have demonstrated EFdifferences between full term and preterm children [37]. In this study,we focused on children born preterm as a model of biomedical risk forboth impairments in EF and problems in social competence. We hadthree main hypotheses. Hypothesis 1 addresses social competence dif-ferences between full term and preterm children: Compared to fullterm children, preterm children show poorer parent-rated social com-petence. Hypothesis 2 addresses links between EF and social compe-tence in both preterm and full term children: EF, whether assessedusing performance-based measures or parent rating, is associated withsocial competence. These associations persist after controlling for ma-ternal education and child age. Hypothesis 3 addresses contributionsof preterm birth to social competence: Given the increased risk of ad-verse outcomes in children born preterm, gestational group contributesadditional variance, beyond the effect of EF, to social competence.

2. Methods

2.1. Participants

Participants were 3- to 5-year-old children recruited from Palo Alto,CA, and surrounding counties. Gestational age (GA), birth weight (BW),andmedical complicationswere gathered from parent report andmed-ical records. Study subjects had a history of preterm birth (34 weeksof gestation) and BW b 2500 g (n = 70). Controls were born full term(37 weeks) and had nomajormedical complications (n= 79). Exclu-sion criteria for all participants included sensory impairments (i.e., blindor deaf), identied genetic syndrome or congenital heart disease, andinability to comprehend task instructions.

The study population consisted of a convenience sample of childrenborn from 2004 to 2009. Preterm subjects were recruited by letters sentto families of children who were evaluated at High Risk Infant Follow-up Services at Lucile Packard Children's Hospital Stanford in Palo Alto,CA, and by postings on local parent message boards. Control childrenwere recruited by postings on parent message boards, by yers in gen-eral pediatric clinics, and by word of mouth. Controls were group-matched to children born preterm for age, gender, ethnicity, and race.

The study was approved by the Stanford Institutional Review Board. Aparent or legal guardian provided informed consent. Participants werecompensated for participation.

2.2. Procedure

Parents completed a demographic questionnaire and standardizedEF rating scales. Children completed a battery of EF tasks that were ad-ministered by trained research assistants. Parents also completed twostandardized measures of social competence.

2.3. Measures and variables

2.3.1. Demographics and health informationA demographic and health questionnaire collected information

addressing child age, gender, ethnicity, race, and school participation(attending daycare, preschool, or kindergarten). Maternal education(b4 years in college, 4 years in college, Master's degree) was usedas an indicator of SES for this relatively high SES population. Healthinformation included gestational age at time of delivery (preterm34 weeks of gestation; full term 37 weeks of gestation), birthweight, and medical conditions. Information concerning receipt ofearly intervention services also was collected.

2.3.2. Performance-based EFTasks were selected to represent core EF constructs based on the de-

velopmental literature on EF in typical and preterm preschoolers[4,5,38]. Children completed a behavioral battery of EF tasks thatincluded:

1. Three- and six-box tasks, a measure of working memory and plan-ning. Six stationary boxes with different color lids are baited with atreat in view of the child. The child's task is to efciently nd all treatsby keeping track of the boxes that have already been searched. Thedependent variable was the total number of reaches to nd all treats.

2. Day/night task, a measure of complex response inhibition. The childshould respond night when shown a picture of the sun and daywhen shown a picture of moon and stars. The task requires the inhi-bition of a prepotent verbal response in the presence of competing vi-sual information. The dependent variable was the number of correctresponses reverse scored by subtracting the number of practice trials.

3. Bird/dragon task, a modied Simon says task, a measure of complexresponse inhibition andworkingmemory. The child must hold a rulein mind and respond while inhibiting a prepotent response. The de-pendent variablewas thenumber of correct responses reverse scoredby subtracting the number of practice trials.

4. Dimensional change card sort, a measure of cognitive exibility/taskswitching and attention shifting. The child is shown cards depictingcolored shapes that can be sorted by color or shape. The child mustsort according to one dimension and then shift to sort according tothe other dimension. The dependent variable was the number of cor-rect responses on the post-switch phase.

5. Verbal uency task, a measure of verbal productivity with languageand organizational components. The child is asked to generate asmany words as possible within specic categories (i.e., animals,foods) in one minute. The dependent variable was the total numberof words that were unique and correct (belongs to the category).

We have reported on the differences between preterm and fullterm groups on each of these tests, with a detailed description of thetasks [37]. We converted continuous outcome measures from theperformance-based EF tasks to z-scores derived from performance(mean, SD) of full term subjects in the three age groups of 3, 4, and5 years. As outlined in the original study, we found signicant small tomoderate correlations among all performance-based EF measures [37].Given the correlations among the tasks, for our analyses, we used anaverage z-score for the ve EF tasks as a summary measure of child EF

performance.

associated with several of the factors measured in the study, includingpreterm birth, impaired EF skills, and social outcomes [33]. For the rst

participant age was 4.5 years (SD= .79), 50% male, 60%White, and 9%

301N. Alduncin et al. / Early Human Development 90 (2014) 2993062.3.3. Parent-rated EFParents completed the Behavior Rating Inventory of Executive

Function-Preschool version (BRIEF-P), a standardized parent-ratingscale of behavioral manifestations of EF in children 2.05.11 years ofage within the context of everyday environments [39,40]. Sixty-threeitems address ve EF constructs: inhibit, shift, emotional control, work-ing memory, and plan/organize. Items are rated on a 3-point scale ofnever, sometimes, or often. Three broad indices (inhibitory self-control, exibility, emergent metacognition) and a composite score(Global Executive Composite, GEC) are generated. Index scores arereported as T-scores (mean of 50, SD of 10); higher scores indicategreater executive dysfunction. Scores at or above 65 are consideredclinically signicant. The measure is normed for gender and ageby two groupings: 2:0 to 3:11 and 4:0 to 5:11. Content validity isbased on factor analysis of clinical and normative samples, conver-gence/discriminance with preschool rating scale measures, and onthe ability to detect EF decits in children with risk factors or disor-der [40]. Examples of items from the BRIEF-P that tap specic EF con-structs are:

Inhibitacts too wild or out of control. Workingmemorycannot stay on the same topic when talking andhas trouble remembering something, even after a brief period oftime.

Plan/organizehas trouble thinking of a different way to solvea problem or complete an activity when stuck and wheninstructed to clean up, puts things away in a disorganized, randomway.

Similar to the EF battery, we found signicant correlations amongthe BRIEF GEC and all indices [37], therefore, we used the GEC as a sum-mary measure of parent-rated EF skills.

2.3.4. Social competenceParents completed the Vineland-II and the Social Competence and

Behavior Evaluationshort form (SCBE-30):

1. Vineland-II is a standardized parent-rating scale of adaptive behaviorfrom birth to adulthood [41]. Items are rated on a 3-point Likert scale(0 = never, 1 = sometimes or partially, 2 = usually). It generatesstandard scores (M = 100, SD = 15) reecting the followingdomains: communication, daily living skills, socialization, motorskills and maladaptive behavior. The socialization scale includesthe following subscales: interpersonal relationships, play andleisure time, and coping skills. Examples of Vineland items forthe socialization scale are: interpersonal relationships showsinterest in children the same age; play and leisure time plays cooperatively with one or more children for 5 minutes;and coping skills responds appropriately to reasonable chang-es in routine. The questions are different from those in the BRIEF-P. We used the socialization scale standard score as a summaryrating of social competence, a decision in keepingwith other stud-ies [20].

2. SCBE-30, research edition, is a well-validated 30-item questionnairethat assesses emerging pro-social behaviors and interactions in thepreschool period [1619]. Items are rated on a 6-point Likert scale(1 = never, 2 or 3 = sometimes, 4 or 5 = often, 6 = always) andyield three scales: social competence, anxious/withdrawn, andanger/aggression. The social competence scale items reect emotion-al expressions, social interaction with peers, and social interactionwith teachers; exemplar items include: comforts or assists anotherchild in difculty, cooperates with other children in group activi-ties, and helps with everyday tasks. The questions in the SCBE-30 are distinct from the questions in both the BRIEF-P and theVineland-II. We used the social competence scale raw score as a sec-

ond measure of social competence.Latino. The two groups differed, by study design, with regard to birthweight and gestational age. Medical complications at birth in the pre-term group included: 11 had abnormal ndings on head ultrasound orMRI (at least grade 2 intraventricular hemorrhage or IVH, echodensities,or cystic lesions) and 10 had mildly abnormal ndings (grade 1 IVH orchoroid plexus cyst); 40 had RDS and 7 developed chronic lung disease;5 had necrotizing enterocolitis; and 8 were small for gestational age(dened as lying at or below the 3rd percentile in birth weight for ges-tational age). The proportion of mothers with an educational level lessthan a college degree was larger for the preterm group compared tothe full term group (30% vs. 9%, p b .003). There were no differencesbetween groups in the proportion of children attending daycare,preschool, or kindergarten. As expected, the proportion of participantswith a history of receiving early intervention services was largerin the preterm group compared to the full term group (41% vs. 0%,model predicting Vineland socialization score, child agewas not entered;the Vineland covers the lifespan and is normed for age [41]. For thesecond model predicting SCBE social competence score, child age wasentered in the rst step, as the SCBE-30 is designed for use across a nar-row age range (30 to 78 months) and is not normed for age [16,18]. EFscores were entered into the regression models after demographic vari-ables. We chose to enter the performance-based EF score (EF Tasksz-score) before parent-rated EF score due to the potential of sharedmethod variance contributing to the outcome (i.e., BRIEF, Vineland andSCBE-30 are based on parent report). Gestational group status (pretermvs. full term) was entered in the nal step, after maternal education andEF measures, to capture additional variance attributable to groupdifferences.

3. Results

3.1. Descriptive analyses and group differences

3.1.1. Demographics and healthOnehundred forty-nineparticipantswere enrolled: 70born preterm

and 79 born full-term (see Table 1). Across gestational groups, child age,gender, race, and ethnicity distributionswere similar. Overall, the mean2.4. Statistical analyses

2.4.1. Descriptive analyses and group differencesMeans and proportions were used to describe the sample with re-

gard to demographic variables. To compare full term and pretermgroups on demographic variables, EF scores, and social competence var-iables, we used t-tests for continuous variables and Pearson chi-squarefor dichotomous outcomes. Data from the BRIEF was missing for 4 pre-term children. Data from the Vineland wasmissing for 5 full term and 9preterm subjects. Data from the SCBE-30wasmissing for 2 full term and4 preterm subjects. Data were missing from the EF battery as follows:one preterm child for the bird/dragon task (would not play); one pre-term and one full term for the day/night task (technical failure); and14 preterm and 6 full term for the verbal uency task (change in testingprotocol).

2.4.2. Regression analysesHierarchical multiple regression was conducted on the entire sample

to determinewhich demographic (maternal education; child age) and EF(performance-based EF EF tasks z-score; parent-reported EFBRIEFGEC) variableswere related to social competence (Vineland socializationand SCBE social competence).We enteredmaternal education in therststep to account for potential contributions of SES, as lower SES has beenp b .001).

3.2. Models predicting to social competence

Table 1Demographic characteristics of preterm and full term participants.

Gestational group Total (n = 149) Preterm (n = 70) Full term (n = 79)t or 2 p

Age (years), mean (SD) 4.5 (.79) 4.6 (.76) 4.4 (.82) 1.5 .135Birth characteristics, mean (SD)Gestational age (weeks) 34.7 (5.2) 29.6 (2.6) 39.3 (1.4) 27.4 b .001Birth weight (grams) 2409 (1097) 1365 (455) 3331 (507) 24.8 b .001

Race, n (%) .064 .867White 91 (39) 42 (60) 49 (61)Non-White 58 (61) 28 (40) 31 (39)

Ethnicity, n (%) 1.6 .453Non-Hispanic/Non-Latino 108 (72) 47 (67) 61 (76)Hispanic/Latino 14 (9) 8 (12) 6 (8)Mixed 28 (19) 15 (21) 13 (16)

Gender, n (%) 2.4 .141Male 75 (50) 40 (57) 35 (44)Female 74 (50) 30 (43) 44 (56)

Maternal education, n (%) 12 .003b4 years in college 28 (19) 21 (30) 7 (9)

3)7)

6)4)0)

9)1)

302 N. Alduncin et al. / Early Human Development 90 (2014) 2993063.1.2. Executive functionAs previously reported, we found signicant differences on all tasks

in the EF battery and the BRIEF composite, indices, and scales, with thepreterm group having poorer performance or higher scores (more EFproblems) than the full term group [37]. For this study, we include thedifferences between groups on summary EF measures (see Table 2).On average, preterm children had signicantly lower EF tasks z-scoresand higher BRIEF GEC scores, indicating more EF problems, than didfull term children. The correlation between average EF tasks z-scoreand parent-rated GEC was modestly but statistically signicant (r = .176, p = .034).4 years in college 50 (33) 23 (3Master's degree 71 (48) 26 (3

Child care/school, n (%)Daycare 23 (16) 11 (1Preschool 113 (87) 57 (8Kindergarten 17 (13) 7 (1

Early intervention, n (%)Yes 41 (27) 41 (5No 108 (73) 29 (43.1.3. Social competenceThemeanVineland-II socialization score, and all Vineland social sub-

scale scores were signicantly lower, indicating poorer social compe-tence, for the preterm group compared to the full term group (seeTable 2). The mean Vineland-II socialization score for the full termgroupwas at the high end of the average range and almost one standarddeviation higher than the preterm group. The mean SCBE social

Table 2Executive function and social competence.

Gestational group Preterm (n = 70)

Mean (SD) or n (%)

Executive function measuresPerformance-based EF tasksz-Score .81 (1.1)

BRIEFGEC 54.3 (15)

Social competence measuresVineland-IISocialization 97.0 (13)Interpersonal relationships 14.2 (3)Play and leisure time 13.3 (3)Coping skills 16.3 (3)

SCBESocial competence 37.6 (8)

Abbreviations: BRIEFBehavior Rating Inventory of Executive Function; GECGlobal Executive CIn the rst set of models (Table 3), parent-rated social competenceas measured by the Vineland-II Socialization Scale was examined as afunction of maternal education, EF, and gestational group variables.Maternal educationwas entered in step 1 and did notmake a signicantcompetence scorewas also signicantly lower, corroborating poorer so-cial competence, for the pretermgroup compared to the full term group.The correlation between SCBE social competence scores and Vinelandsocialization scores was statistically signicant (r = .509, p b .001).

27 (34)45 (57)

12 (15) .009 .92656 (91) 1.8 .17710 (13) .23 .635

65 b .0010 (0)

79 (100)contribution to the model (R-squared change = .002). EF variableswere entered in steps 2 and 3. Results demonstrated that theperformance-based EF skills variable (EF tasks z-score) accounted for10.2% of the variance in social function. The parent-rated EF variable(BRIEF GEC) added 7.8% to the total explained variance. In step 4, gesta-tional group accounted for another 9.1% of the variance in socialfunction; the BRIEF GEC remained signicantly associated with the so-cialization score, but the EF tasks z-score did not. This model accounted

Full term (n = 79)

Mean (SD) or n (%) t or 2 p

.01 (.6) 5.5 b .001

43.6 (8) 5.2 b .001

110.3 (13) 5.8 b .00116.8 (3) 5.7 b .00115.9 (2) 5.5 b .00118.1 (3) 3.4 .001

42.3 (7) 3.6 b .001

omposite; EFexecutive function; and SCBESocial Competence and Behavior Evaluation.

for a total of 27.3% of the variance in parent-rated social competence, asmeasured by the Vineland-II Socialization Scale.

In the second set of models (Table 4), parent-rated social compe-

gestational group did not add signicantly to the total variance(R-squared change = .002); both performance-based and parent-rated EF remained signicantly associated with social competence.

Table 3Hierarchical multiple regression model predicting to social competence as measured by the Vineland-II Socialization Scale.

Outcome Predictors R2 p (model) B SE B p (predictor) R2

Vineland-II socializationa .002 .663Maternal education .76

1.7 .039 .663 .002Vineland-II socializationb .104 .001

Maternal education .90 1.7 .046 .586EF Tasks z-score 4.78 1.3 .320 b .001 .102

Vineland-II socializationc .182 b .001Maternal education 1.59 1.6 .081 .320EF Tasks z-score 3.57 1.3 .239 .005BRIEF GEC .33 .096 .294 .001 .078

Vineland-II socializationd .273 b .001Maternal education 2.77 1.5 .142 .074EF tasks z-score 1.90 1.26 .127 .135BRIEF GEC .21 .096 .188 .028Gestational group 10.4 2.64 .356 b .001 .091

Outcome is the Vineland-II socialization score. B indicates unstandardized coefcient; SE B, standard error of B;, standardized coefcient; EF, executive function; and BRIEF GEC, BehaviorRating Inventory of Executive Function Global Executive Composite.

a First model. Step 1, maternal education; maternal education was not signicantly associated with the socialization score.b Second model. Step 1, maternal education; step 2, EF Tasks z-score. EF Tasks z-score was signicantly associated with socialization score.c Thirdmodel. Step 1,maternal education; step 2, EF tasks z-score; step 3, BRIEF GEC. Both EFmeasures (BRIEF GEC and EF Tasks z-score)were signicantly associatedwith socialization

score.d Fourthmodel. Step1,maternal education; step 2, EF tasks z-score; step 3, BRIEFGEC; step 4, gestational group. After gestational group is added to themodel, one EFmeasure (i.e., BRIEF

GEC) remains signicantly associated with socialization score; EF Tasks z-score no longer contributes signicantly. There is an independent contribution of gestational group to socializa-tion score.

e So

303N. Alduncin et al. / Early Human Development 90 (2014) 299306tence as measured by the SCBE social competence scale was examinedas a function of maternal education, child age, EF, and gestationalgroup variables. Maternal education and child age, entered in steps 1and 2, did not make signicant contributions to the model (R-squaredchanges = .021 and .017, respectively). EF variables were entered insteps 3 and 4. Results demonstrated that the performance-based EFskills variable (EF tasks z-score) signicantly accounted for 10.7% ofthe variance in social competence. The parent-rated EF variable (BRIEFGEC) added another 14.2% to the total explained variance. In step 5,

Table 4Hierarchical multiple regression model predicting to social competence as measured by thOutcome Predictors R2 p (model)

SCBE social competencea .021 .085Maternal education

SCBE social competenceb .038 .068Maternal educationChild age

SCBE social competencec .145 b .001Maternal educationChild ageEF Tasks z-score

SCBE social competenced .287 b .001Maternal educationChild ageEF Tasks z-scoreBRIEF GEC

SCBE social competencee .290 b .001Maternal educationChild ageEF tasks z-scoreBRIEF GECGestational group

Outcome is the Social Competence and Behavior Evaluation (SCBE) social competence score. BEF, executive function; and BRIEF GEC, Behavior Rating Inventory of Executive Function Global

a First model. Step 1, maternal education; maternal education was not signicantly associateb Second model. Step 1, maternal education; step 2, child age. Child age was not signicantlc Third model. Step 1, maternal education; step 2, child age; step 3, EF tasks z-score. EF taskd Fourthmodel. Step 1,maternal education; step 2, child age; step 3, EF tasks z-score; step 4 B

with social competence score. There is an independent contribution of child age to social compe Fifthmodel. Step 1,maternal education; step 2, child age; step 3, EF tasks z-score; step 4, BR

signicantly associated with social competence score. There is an independent contribution ofThis model accounted for a total of 29.0% of the variance in parent-rated social competence as measured by the SCBE Socialization Scale.

4. Discussion

This study of three- to ve-year old preterm and full term preschoolchildren found that compared to full term children, preterm childrenshowed poorer social competence on two parent rating measures thatassess different components of social competence. As previously

cial Competence and Behavior Evaluation (SCBE).B SE B p (predictor) R2

1.52 .882 .145 .085 .021

1.76 .89 .167 .0491.33 .85 .132 .120 .017

1.61 .843 .153 .0571.53 .805 .152 .0592.71 .653 .329 b .001 .107

.96 .782 .091 .2211.73 .739 .173 .0201.83 .621 .222 .004.25 .048 .399 b .001 .142

.86 .800 .081 .2811.75 .741 .174 .0201.68 .665 .203 .011.24 .050 .383 b .001.91 1.39 .056 .642 .002

indicates unstandardized coefcient; SE B, standard error of B; , standardized coefcient;Executive Composite.d with social competence score.y associated with social competence score.s z-score was signicantly associated with social competence score.RIEF GEC. Both EFmeasures (EF tasks z-score and BRIEF GEC)were signicantly associatedetence score.IEF GEC; step 5, gestational group. Both EFmeasures (EF tasks z-score and BRIEF GEC)werechild age to socialization and no contribution of gestational group.

304 N. Alduncin et al. / Early Human Development 90 (2014) 299306reported, preterm children compared to full term children had poorerEF skills on both performance-based and parent-rated EF measures[37]. Our investigation of the association between EF and social compe-tence showed that both performance-based and parent-rated measuresof EF made independent contributions to social competence. Gestation-al group contributed additional variance to social competence mea-sured by the Vineland-II, but not the SCBE.

The social competence results are consistent with previous studiesof preterm children that found that preterm compared to full termgroups have poor social outcomes [27,28]. One major contribution ofthis study is its use of multiple measures of social competence. Thestudy of social competence is important because problems in the socialdomain contribute to psychological distress, social isolation, and re-duced self-esteem and may impact greatly on quality of life [14]. Poorsocial competence is associated with behavior problems, conduct disor-der, internalizing disorders (such as depression), cognitive impair-ments, and developmental and neurological conditions (e. g., ADHD,schizophrenia, brain injury) [14,42,43].

The EF results are consistent with previous studies on children bornpreterm that used an administered EF battery or parent report of EFskills [2,5,7,8]. This study adds to the literature by examining the associ-ation between EF and social competence using both performance-basedand parent-rated EF measures. Parent report and direct observation ap-pear to capture overlapping as well as distinctive aspects of EF [37];these variables were signicantly but only modestly correlated (r = .18). In addition, as discussed below, these two variables made inde-pendent contributions to social competence, suggesting that they assessdistinct subcomponents of EF.

A second contribution is the focus on preschoolers. A large longitudi-nal follow-up study of adolescents born preterm showed persistence ofsocial difculties [25], suggesting that identifying childrenwith poor so-cial competence during preschool and intervening early may improvelong-term outcomes.

A third contribution of this study is that it demonstrates that EF,whether assessed using performance-based measures or parent rating,is associated with social competence. Using regression models, wefound that these relations persist after controlling for maternal educa-tion and child age. With the addition of gestational age group in thenal steps of the models, performance-based EF was no longer signi-cant in predicting toVineland-II socialization score, but remained signif-icant when predicting to SCBE social competence score. In the nalmodel predicting to the Vineland socialization score, gestational agegroup alsomade an independent contribution to the outcomemeasure.In the nal model predicting to SCBE social competence score, gesta-tional age group made no independent contribution to the outcome.These ndings may indicate that prematurity is associated with thedevelopment of skills in social competence, as measured by a develop-mental measure, such as the Vineland-II, whereas the contributions ofEF to social and emotional functioning, as measured by the SCBE arevery similar in preterm and full term preschoolers.

The differences in the contribution of gestational age group to socialcompetence outcomes may also relate to (1) how the two measures ofsocial competence were validated and (2) subtle differences in the un-derlying theoretical constructs being measured under the broader um-brella of social function. Standardization of the Vineland-II included anormative sample consisting of 20 age groups covering the age rangefrom birth through 90 years [41]. In addition, eleven clinical groups(i.e., attention-decit/hyperactivity disorder; autism; specic learningdisability; intellectual disability) were studied to validate identicationof adaptive decits in those populations. Criterion-related validity stud-ies compared Vineland-II scores to other instruments assessing behav-ior and ability, such as the Adaptive Behavior Assessment System,Second Edition; the Behavior Assessment System for Children, SecondEdition; and the Wechsler Intelligence Scale for Children, Third Edition[41]. In contrast, the SCBE is used for children across a narrow age

range (3078 months) [16,18]. It was compared to the Child BehaviorChecklist/Teacher Report Form to establish convergent and discrimi-nant validity but not the Vineland. Criterion validity for the SCBE wasestablished using peer sociometrics and direct observation of socialparticipation; the short form of the SCBE utilized a structured, directlyobserved compliance task [16,18]. The social competence factor of theSCBE taps a range of items designed to assess the positive qualities ofthe child's adaptation, rather than specic behavioral competencies. So-cial competence in the context of the SCBE refers to a well-adjusted,exible, and generally prosocial pattern of adaptation [16]. Giventhese differences in the two measures, the Vineland may have beenmore sensitive to group differences in a clinical population of pretermchildren compared to full term children, whereas the SCBE may bemore sensitive to prosocial patterns of behaviors observed across allchildren.

Our ndings are consistent with a recently published study address-ing social competence in preschool-aged children born very preterm[28]. In their study, Jones et al. found that cognitive ability, as measuredby IQ,mediated the effects of brain injury on social competence; howev-er, the study did not examine EF [28]. Loe et al. found that, compared to9- to 16-year-old youth born at term, same-aged youth born pretermhad lower performance-based EF and lower parent-rated child function,a composite comprised of child activities, school and social function; EFcontributed to parent-rated child function, which included aspects ofsocial competence [30]. Taylor et al. found that 8 year old preterm chil-drenhad lower scores than termchildren on academic achievement andadaptive functioning, which included socialization measured on theVineland [28]. In that study, neuropsychological skills assessed using astandardizedmeasure of EF skills, including planning andmemory, me-diated the effects of neonatal risk on academic and functional outcomes[29]. Other studies of clinical populations such as childrenwith traumat-ic brain injury [10] and children with fetal alcohol exposure [44] alsofound relationships between parent-rated EF using the same measuresas our study (BRIEF) and social outcomes. The study of children withtraumatic brain injury also examined performance-based measures ofEF, but did not nd signicant associations with social outcomes [10].

The clinical assessment of EF and its association with functional out-comes such as social competence are important areas of applied re-search focus, particularly in at-risk populations such as children bornpreterm. Given the links between EF skills and social competence, wesuggest that EF skills and social competence may serve as targets forintervention to improve overall outcomes, including social competencein children born preterm. A few studies have explored the potentialof neuropsychological interventions in preterm children, includingcomputer-based working memory intervention that targets EF skills[45]. Preschool curricula targeting multiple EF skills have shown im-provements in EF; the Head Start REDI (Research Based, Developmen-tally Informed) intervention was tested in a randomized-controlledtrial of 356 preschoolers and demonstrated EF gains [46]. The possibilityof interventions in the reverse direction also could be considered: thatis, effecting change in social competence in order to improve EF. Inthis vein, a study of 27 students (1114 years) with high functioningautism showed signicant improvement on parent reports of socialskills and EF after administration of a group-based social competenceintervention that used cognitive behavioral principles [47]. Whilestudies with larger samples and application in naturalistic settings arewarranted, social competence intervention appears promising andmay be useful in other at-risk populations.

Our study had limitations. The mean maternal education of thesample was high, consistent with the geographic area, and may not berepresentative of the preterm and full term populations with lowerSES. High maternal education may have limited the ability to identifycontributions of SES to outcomes. In addition, the full term group hada mean socialization score on the Vineland at the high end of the aver-age range compared to the preterm group, which alsomay have limitedthe ability to identify contributions of SES. Nonetheless, even with the

resources and access to services accompanying higher maternal

305N. Alduncin et al. / Early Human Development 90 (2014) 299306education, the preterm children in our study showed impairment in EFand social competence compared to full term children. We did notcollect teacher reports of EF skills or social competence, as not all chil-dren were in school programs. Future work that includes teacher ornon-parent caregiver reports would decrease the shared response biasattributable to parent ratings of both EF and social competence. Themeasurement of social competence through direct assessment and/orstandardized observation rather than parent report would also allowfurther examination of the associations with reduced response biasamong predictor and outcome variables. Longitudinal study of the de-velopment of EF and social outcomes would allow for a better under-standing of the developmental trajectory of skills in preterm children.

In summary, this study found that parent-rated and performance-based measures of EF were associated with two different measures ofsocial competence. These results contribute to our understanding ofthe links between EF, as a cognitive domain of function, and social com-petence, another important but less well studied domain of function inchildren born preterm. We propose a standard assessment of EF skillsand social competence in young children in the preschool period, espe-cially in children born preterm, in order to identify children at-risk forlong-term social difculties. EF and social competence measures mayalso provide targets for intervention and opportunities to assess changeafter treatment.

Financial disclosure

The authors have no nancial relationships relevant to this article todisclose.

Conict of interest

The authors have no conict of interest relevant to this article todisclose.

Acknowledgments

This work was supported by the Lucile Packard Foundation forChildren's Health under a Pilot Early Career Grant, the Society forDevelopmentalBehavioral Pediatrics under the Young InvestigatorAward, the Eunice Kennedy Shriver National Institute of Child Healthand Human Development, National Institutes of Health, under theMentored Patient-oriented Research Career Development AwardGrant K23HD071971 to Irene M. Loe, and the National Center for Re-search Resources, National Institutes of Health under the Stanford Clin-ical and Translational Science Award 1UL1 RR025744.

We thank the children and families who participated in our study.Special thanks to Maya Chatav and Walter S. Chang for their assistancewith the data collection.

References

[1] Ciairano S, Visu-Petra L, Settanni M. Executive inhibitory control and cooperativebehavior during early school years: a follow-up study. J Abnorm Child Psychol Jun2007;35(3):33545.

[2] Raaijmakers MA, Smidts DP, Sergeant JA, Maassen GH, Posthumus JA, van EngelandH, et al. Executive functions in preschool children with aggressive behavior: impair-ments in inhibitory control. J Abnorm Child Psychol Oct 2008;36(7):1097107.

[3] RiggsNR, Jahromi LB, Razza RP, Dillworth-Bart JE,Mueller U. Executive function and thepromotion of socialemotional competence. J Appl Dev Psychol 2006;27(4):3009.

[4] Mulder H, Pitchford NJ, Hagger MS, Marlow N. Development of executive functionand attention in preterm children: a systematic review. Dev Neuropsychol2009;34(4):393421.

[5] Aarnoudse-Moens CS, Smidts DP, Oosterlaan J, Duivenvoorden HJ,Weisglas-KuperusN. Executive function in very preterm children at early school age. J Abnorm ChildPsychol Oct 2009;37(7):98193.

[6] Zelazo P, Carter A, Reznick S, Frye D. Early development of executive function: aproblem-solving framework. Rev Gen Psychol 1997;L(2):198226.

[7] Burnett AC, Scratch SE, Anderson PJ. Executive function outcome in preterm adoles-cents. Early Hum Dev Apr 2013;89(4):21520.[8] Aarnoudse-Moens CS, Duivenvoorden HJ, Weisglas-Kuperus N, Van Goudoever JB,Oosterlaan J. The prole of executive function in very preterm children at 4 to12 years. Dev Med Child Neurol Mar 2012;54(3):24753.

[9] Araujo Jimnez EA, Jan Ballabriga MC, Bonillo Martin A, Javier Arrufat F, SerraGiacobo R. Executive functioning in children and adolescents with symptoms ofsluggish cognitive tempo and ADHD. J Atten Disord 2013:18.

[10] Ganesalingam K, Yeates KO, Taylor HG, Walz NC, Stancin T, Wade S. Executive func-tions and social competence in young children 6 months following traumatic braininjury. Neuropsychology Jul 2011;25(4):46676.

[11] Jacobson L, Williford A, Pianta RC. The role of executive function in children's com-petent adjustment to middle school. Child Neuropsychol 2011;17(3):25580.

[12] MillerM, Hinshaw SP. Does childhood executive function predict adolescent function-al outcomes in girls with ADHD? J Abnorm Child Psychol Apr 2010;38(3):31526.

[13] Iarocci G, Yager J, Elfers T.What geneenvironment interactions can tell us about socialcompetence in typical and atypical populations. Brain Cogn Oct 2007;65(1):11227.

[14] Beauchamp MH, Anderson V. SOCIAL: an integrative framework for the develop-ment of social skills. Psychol Bull Jan 2010;136(1):3964.

[15] Martinez W, Carter JS, Legato LJ. Social competence in children with chronic illness:a meta-analytic review. J Pediatr Psychol Sep 2011;36(8):87890.

[16] LaFreniere PJ, Dumas JE. Social Competence and Behavior Evaluation PreschoolEdition (SCBE). 2nd printing ed. Los Angeles: Western Psychological Services;1995.

[17] LaFreniere PJ, Dumas JE. Social Competence and Behavior Evaluation in children ages3 to 6 years: the short form (SCBE-30). Psychol Assess 1996;8(4):36977.

[18] Kotler JC, McMahon RJ. Differentiating anxious, aggressive, and socially competentpreschool children: validation of the Social Competence and Behavior Evaluation-30 (parent version). Behav Res Ther 2002;40:94759.

[19] Denham SA, Wyatt TM, Bassett HH, Echeverria D, Knox SS. Assessing socialemotional development in children from a longitudinal perspective. J EpidemiolCommunity Health Jan 2009;63(Suppl. 1):i3752.

[20] Bornstein MH, Hahn CS, Haynes OM. Social competence, externalizing, and internal-izing behavioral adjustment from early childhood through early adolescence: devel-opmental cascades. Dev Psychopathol Nov 2010;22(4):71735.

[21] Hawkins JD, Kosterman R, Catalano RF, Hill KG, Abbott RD. Promoting positive adultfunctioning through social development intervention in childhood: long-termeffects from the Seattle Social Development Project. Arch Pediatr Adolesc Med2005;159:2531.

[22] Winchester SB, Sullivan MC, Marks AK, Doyle T, DePalma J, McGrath MM. Academic,social, and behavioral outcomes at age 12 of infants born preterm. West J Nurs ResNov 2009;31(7):85371.

[23] Talge NM, Holzman C, Wang J, Lucia V, Gardiner J, Breslau N. Late-preterm birth andits association with cognitive and socioemotional outcomes at 6 years of age.Pediatrics Dec 2010;126(6):112431.

[24] Johnson S, Hollis C, Kochhar P, Hennessy E,Wolke D,MarlowN. Psychiatric disordersin extremely preterm children: longitudinal nding at age 11 years in the EPICureStudy. J Am Acad Child Adolesc Psychiatry May 2010;49(5):45363 [e1].

[25] Hille ETM, Dorrepaal C, Perenboom R, Gravenhorst JB, Brand R, Verloove-VanhorickSP. Social lifestyle, risk-taking behavior, and psychopathology in young adults bornvery preterm or with a very low birth weight. J Pediatr 2008;15(6):793800.

[26] Sullivan MC, Miller RJ, Msall ME. 17-Year outcome of preterm infants with diverseneonatal morbidities: part 2, impact on activities and participation. J Spec PediatrNurs Oct 2012;17(4):27587.

[27] Landry SH, Chapieski ML, Richardson MA, Palmer J, Hall S. The social competence ofchildren born prematurely: effects of medical complications and parent behaviors.Child Dev 1990;61:160516.

[28] Jones KM, Champion PR, Woodward LJ. Social competence of preschool childrenborn very preterm. Early Hum Dev 2013;89(10):795802.

[29] Taylor HG, Klein N, Drotar D, Schluchter M, Hack M. Consequences and risks ofb1000-g birth weight for neuropsychological skills, achievement, and adaptive func-tioning. J Dev Behav Pediatr 2006;27(6):45969.

[30] Loe IM, Lee ES, Luna B, Feldman HM. Executive function skills are associated withreading and parent-rated child function in children born prematurely. Early HumDev Feb 2012;88(2):1118.

[31] Bradley RH, Corwyn RF. Socioeconomic status and child development. Annu RevPsychol 2002;53(1):37199.

[32] Goldenberg RL, Culhane JF, Romero R. Epidemiology and causes of preterm birth.Lancet 2008;371:7584.

[33] Noble KG, McCandliss BD, Farah MJ. Socioeconomic gradients predict individual dif-ferences in neurocognitive abilities. Dev Sci Jul 2007;10(4):46480.

[34] Noble KG, Norman M, Farah MJ. Neurocognitive correlates of socioeconomic statusin kindergarten children. Dev Sci Jan 2005;8(1):7487.

[35] Landry SH, Denson SE, Swank PR. Effects of medical risk and socioeconomic statuson the rate of change in cognitive and social development for low birth weight chil-dren. J Clin Exp Neuropsychol Apr 1997;19(2):26174.

[36] Luna B, Garver KE, Urban TA, Lazar NA, Sweeney JA. Maturation of cognitive process-es from late childhood to adulthood. Child Dev SepOct 2004;75(5):135772.

[37] Loe IM, ChatavM, Alduncin N. Complementary Assessments of Executive Function inPreterm and Full Term Preschoolers. Child Neuropsychol 2014 Accepted March 16,2014 (in press).

[38] Carlson SM. Developmentally sensitive measures of executive function in preschoolchildren. Dev Neuropsychol 2005;28(2):595616.

[39] Gioia GA, Isquith PK, Guy SC, Kenworthy L. Behavior rating inventory of executivefunction. Child Neuropsychol 2000;6(3):2358.

[40] Gioia GA, Isquith PK, Retzlaff PD, Espy KA. Conrmatory factor analysis of the Behav-ior Rating Inventory of Executive Function (BRIEF) in a clinical sample. ChildNeuropsychol 2002;8(4):24957.

[41] Sparrow SS, Balla DA, Cicchetti D. Vineland adaptive behavior scales. Second ed.Bloomington, MN: Pearson Assessments; 2005.

[42] Anderson V, Beauchamp MH, Yeates KO, Crossley L, Hearps SJ, Catroppa C. Socialcompetence at 6 months following childhood traumatic brain injury. J IntNeuropsychol Soc May 2013;19(5):53950.

[43] Yeates KO, Bigler ED, Dennis M, Gerhardt CA, Rubin KH, Stancin T, et al. Social out-comes in childhood brain disorder: a heuristic integration of social neuroscienceand developmental psychology. Psychol Bull May 2007;133(3):53556.

[44] Schonfeld AM, Paley B, Frankel F, O'Connor MJ. Executive functioning predictssocial skills following prenatal alcohol exposure. Child Neuropsychol Dec2006;12(6):43952.

[45] Grunewaldt KH, Lohaugen GC, Austeng D, Brubakk AM, Skranes J. Working memorytraining improves cognitive function in VLBW preschoolers. Pediatrics Mar2013;131(3):e74754.

[46] Bierman KL, Nix RL, Greenberg MT, Blair C, Domitrovich CE. Executive functions andschool readiness intervention: impact, moderation, and mediation in the Head StartREDI program. Dev Psychopathol Summer 2008;20(3):82143.

[47] Stichter JP, HerzogMJ, Visovsky K, Schmidt C, Randolph J, Schultz T, et al. Social com-petence intervention for youth with Asperger syndrome and high-functioningautism: an initial investigation. J Autism Dev Disord Sep 2010;40(9):106779.

306 N. Alduncin et al. / Early Human Development 90 (2014) 299306

Executive function is associated with social competence inpreschool-aged children born preterm or full term1. Introduction1.1. Study objective and hypotheses

2. Methods2.1. Participants2.2. Procedure2.3. Measures and variables2.3.1. Demographics and health information2.3.2. Performance-based EF2.3.3. Parent-rated EF2.3.4. Social competence

2.4. Statistical analyses2.4.1. Descriptive analyses and group differences2.4.2. Regression analyses

3. Results3.1. Descriptive analyses and group differences3.1.1. Demographics and health3.1.2. Executive function3.1.3. Social competence

3.2. Models predicting to social competence

4. DiscussionFinancial disclosureConflict of interestAcknowledgmentsReferences