Attention and language as mediators of academic outcomes following early psychosocial deprivation

Abstract

Children adopted from institutions at older ages are at increased risk of persistent attention problems, lower cognition, and academic difficulties. This study examined cognitive and behavior problems as mediators of the association between early psychosocial deprivation and academic functioning. Participants were 8–17-year-old children adopted from psychosocially-depriving Russian institutions after 14 months of age (n = 34) and before 9 months of age (n = 39). Children completed a cognitive assessment, while their parents completed questionnaires on child behavior problems and use of learning support services in school. Children adopted after 14 months were found to have significantly lower vocabulary, higher levels of attention problems, and higher rates of using learning support services relative to children adopted before 9 months after controlling for age at assessment. The two groups did not differ significantly in nonverbal reasoning, anxiety, or oppositional behavior. Attention and vocabulary significantly mediated the association of early psychosocial deprivation with the use of learning support services. These findings suggest that interventions targeting attention regulatory and language skills may be beneficial in terms of improving school performance in this population.

Keywords

attention cognition early deprivation post-institutionalized children service use

Children adopted from institutions are exposed to a circumscribed period of early deprivation followed by placement into usually stable and well-educated middle-class families (Hellerstedt et al., 2008). Following adoption, marked developmental catch-up is generally observed (van IJzendoorn, Juffer, & Poelhuis, 2005). However, post-institutionalized (PI) children adopted at older ages are at increased risk of attention, cognitive, and academic difficulties years after adoption (Castle et al., 2006; Eigsti, Weitzman, Schuh, De Marchena, & Casey, 2011; Stevens et al., 2008). Studies of general development and education suggest that attention and cognitive problems often contribute to poor school performance (Breslau et al., 2010). Although institutions range in quality and level of deprivation, they tend to be characterized by psychosocial deficiencies, including the lack of consistent and responsive caregiving (Gunnar, 2001), and early exposure to psychosocial deprivation may lead to attention and cognitive difficulties. In the current study, we examined attention, cognitive, and academic functioning among school-aged children adopted from “psychosocially depriving” institutions that provided adequate physical resources but failed to provide a consistent set of responsive caregivers (St. Petersburg–USA Orphanage Research Team, 2005). We investigated attention and cognitive functioning as mediators of the association between early psychosocial deprivation and academic difficulties.

Academic difficulties in post-institutionalized children

PI children have lower reading and math achievement scores (Beckett et al., 2007) and a higher rate of using learning support services in school (Beverly, McGuinness, & Blanton, 2008; Castle et al., 2006; Le Mare, Audet, & Kurytnik, 2007; Loman, Wiik, Frenn, Pollak, & Gunnar, 2009) than children reared in their biological homes and children adopted at young ages from non-institutional settings. Academic service use was significantly associated with children’s level of reading and math difficulty, supporting the validity of service use as a measure of academic difficulties (Castle et al., 2006). Understanding the specific set of challenges that increase PI children’s risk of poor academic achievement can inform the design of interventions tailored to this population.

Cognitive and behavior problems in post-institutionalized children

Research has indicated that PI children are susceptible to a number of cognitive and behavioral difficulties that may contribute to their poor school achievement. Cognitive functioning includes verbal (e.g., vocabulary) and nonverbal (e.g., matrix reasoning) skills. As a specific area of cognition, language skills include phonological processing, syntax, and vocabulary. Although many studies have found deficits in both verbal and nonverbal IQ (Fox, Almas, Degnan, Nelson, & Zeanah, 2011; Loman et al., 2009), some results indicate that verbal functioning (e.g., vocabulary) might be more susceptible to deprivation effects than nonverbal reasoning (Eigsti et al., 2011). Similarly, preschool- and school-age PI children exhibit language difficulties compared to children reared in their biological homes and children adopted at young ages from non-institutional settings (Glennen & Bright, 2005; Rutter et al., 2007; Windsor et al., 2011). Some studies indicate language outcomes within the average range on standardized measures, but significantly lower than comparison groups, suggesting that standardized scores alone may not capture language difficulties in this population (Eigsti et al., 2011; Loman et al., 2009). Language difficulties have been found even when PI children were placed in family care in their country of origin and were tested in their language of origin (Windsor et al., 2011).

Emotional and behavior problems have also been found among PI children. Specifically, problems with inattention, hyperactivity, and impulsivity, including elevated rates of attention-deficit/hyperactivity disorder (ADHD), have been documented across multiple samples of PI children (Gunnar & van Dulmen, 2007; Stevens et al., 2008). Attention problems are present at preschool-age and persist into the school-age years (Audet & Le Mare, 2011). Previous research has yielded inconsistent findings for other emotional and behavior problems, such as anxiety and oppositional behavior (Gunnar & van Dulmen, 2007; Rutter et al., 2007).

In never-institutionalized samples, children with poor language skills are less likely to succeed in reading comprehension and mathematics (Ouellette, 2006; Young et al., 2002). In addition, attention problems often result in poor school achievement and increased learning support service use (Breslau et al., 2010; Duncan et al., 2007). In a study of severely deprived PI children, group differences in academic achievement at age 11 were largely mediated by IQ at age 11, and to a lesser extent, teacher-reported inattention and hyperactivity-impulsivity at age 6 (Beckett et al., 2007).

Role of early psychosocial deprivation

Under typical family rearing conditions, rudimentary attention and language emerge in the first year of life, and these skills develop rapidly over the course of early childhood (Rueda, Posner, & Rothbart, 2005). Caregiver responsiveness and scaffolding have been found to facilitate children’s attention (NICHD Early Child Care Research Network, 2003; Olson, Bates, Sandy, & Schilling, 2002), cognitive, and language development (Landry, Smith, & Swank, 2006; Tamis-LeMonda, Bornstein, & Baumwell, 2001). When these caregiving resources are absent, such as in cases of child maltreatment, multiple foster placements, and caregiver disruptions, children often exhibit poor attention and cognitive functioning (Pears, Fisher, Bruce, Kim, & Yoerger, 2010).

Based on these studies, early exposure to the psychosocial deficiencies of institutions, such as the lack of a consistent set of responsive caregivers, may interfere with development of attention and cognitive processes. While it is difficult to precisely specify the level of early deprivation (Gunnar, 2001), several Russian institutions were classified as psychosocially depriving after they were found to provide adequate physical resources, such as nutrition and medical care, but expose children to frequent changes in caregivers and a lack of sensitive, responsive caregiving. Specifically, in these institutions, nine or more different caregivers often worked with a group of 12–14 children in a given week, and children were periodically “graduated” to new groups with different caregivers. By 19+ months of age, children had potentially experienced 60 to 100 caregivers (St. Petersburg–USA Orphanage Research Team, 2005). Caregivers rarely initiated social interactions, responded to infants’ social bids, responded promptly to emotional distress, or provided warmth and affection (Muhamedrahimov, 1999; see also Tirella et al., 2008). School-aged children adopted from these psychosocially-depriving institutions at older ages had higher rates of parent-reported attention difficulties than those adopted at younger ages and the normative sample (Merz & McCall, 2010). But, children exposed to early psychosocial deprivation have not been examined with regard to attention and cognitive predictors of academic difficulties.

Age at adoption

Prior research on PI children has indicated a step function or threshold effect whereby risk of attention and cognitive difficulties is increased only after a certain age at adoption. Specifically, 11-year-old children adopted from severely depriving Romanian institutions before the age of 6 months did not differ significantly from children adopted within the United Kingdom at young ages; children adopted after the age of 6 months were at increased risk, but within this group, there was no association between age at adoption and attention or cognitive outcomes (Rutter et al., 2007). Studies of children adopted from less severely depriving institutions have suggested a step-function at older ages of adoption, such as after 18–24 months (Gunnar & van Dulmen, 2007; Merz & McCall, 2010; Fox et al., 2011). In some studies, linear or dose-response associations have been found with age at adoption for some developmental outcomes (Sheridan, Drury, McLaughlin, & Almas, 2010). Taken together, findings across studies consistently indicate that PI children adopted before 6–9 months do not differ from typically-developing children, whereas PI children adopted after 14–18 months have greater attention and cognitive problems. Thus, younger-adopted children represent a useful comparison group for older-adopted children.

The current study

The current study examined cognition, behavior problems, and learning support service use among school-age children adopted from psychosocially depriving institutions and investigated mediators of the association between early psychosocial deprivation and learning support service use. Children adopted before 9 months and after 14 months of age were compared in terms of attention problems, vocabulary, nonverbal reasoning, anxiety, and oppositional behavior. These age-at-adoption groups were chosen to be consistent with prior research while also preserving within-group sample size. Older-adopted children were expected to have poorer attention and vocabulary and a higher rate of learning support service use relative to younger-adopted children. Predictions were not made for nonverbal reasoning, anxiety, or oppositional behavior, although these problems are less consistently found in studies of PI children. Based on prior studies, we expected attention problems and vocabulary to mediate the association between early psychosocial deprivation and the use of learning support services in school. Analyzing group differences in a range of potential mediators addressed the specificity of attention and vocabulary as mediators of academic difficulties over other cognitive and behavior problems.

Method

Participants

Participants were 73 8–17-year-old children adopted from psychosocially depriving institutions primarily in St. Petersburg, Russia (68%); 54% came from the specific institutions described in the introduction (St. Petersburg–USA Orphanage Research Team, 2005). There were 39 children adopted ≤ 9 months (range: 4–9 months) and 34 children adopted ≥ 14 months (range: 14–48 months). Although some children had spent time with their birth families prior to adoption, all children in the sample had spent half or more of their pre-adoption time in institutions. Children were excluded from analyses if they had marked functional deficits (e.g., autism spectrum disorder with cognitive impairment) or fetal alcohol spectrum disorder (n = 5). As shown in Table 1, there were no differences between the PI groups in terms of child gender, age at assessment, or birth circumstances. The adoptive families of the two PI groups did not differ significantly in marital/partnership status, parental education, or income. Of the 19 children taking stimulant medications, 11 children took their medication on the day of testing. The percentage of children taking stimulant medication at the time of testing did not differ between groups (see Table 1).

Table 1.

Descriptive statistics for child and family characteristics.

	Age at adoption (months)
	≤9 (n = 39)	≥14 (n = 34)	Statistical test
Age (years)	13.04 (2.57)	12.69 (3.46)	t(60) = .48, ns
Male	33%	38%	χ² (1, N = 73) = .19, ns
Country of origin	Russia (35) Ukraine (1) Belarus (1) Uzbekistan (2)	Russia (29) Belarus (4) Uzbekistan (1)
Months in an institution	7.21 (1.45)	19.76 (8.14)	t(35) = 8.88, p < .001
Years in adoptive home	12.39 (2.57)	10.71 (3.49)	t(60) = 2.32, p < .05
Took stimulant meds on day of testing	15%	18%	χ² (1, N = 73) = .07, ns
Born prematurely	37%	41%	χ² (1, N = 57) = .08, ns
Birth weight (lbs)	6.36 (1.36)	5.61 (1.59)	t(49) = 1.79, p = .08
Low birth weight	28%	42%	χ² (1, N = 51) = 1.05, ns
Parent married/partnered	90%	82%	χ² (1, N = 73) = .84, ns
Parent with 4-year college degree	95%	94%	χ² (1, N = 73) = .01, ns
Household income >US$100,000	79%	74%	χ² (1, N = 73) = .22, ns

Note. The parent education measure indicated families in which at least one parent had a 4-year college degree or higher education. For PI children adopted ≤ 9 months, time in an institution ranged from 4 to 9 months. For PI children adopted ≥ 14 months, time in an institution ranged from 7 to 47 months.

PI participants were recruited through an adoption agency specializing in adoption from Russia. Most of the participants had also participated at one or more time points in a larger main study which examines behavior and development of all children adopted through this agency via parent-report questionnaires (Merz & McCall, 2010). In the main study, adoptive parents were sent four waves of questionnaires which asked about child and family background, child health, diagnoses, and service use, and child behavior and development. During recruitment for the current study, adoptive parents received a letter of invitation and a follow-up phone call.

The participation rate was about 54%. Primary reasons for declining participation included lack of interest and time demands. According to a selective responding check conducted on the main study sample, parents of children with more difficulties were no more likely to respond to a single questionnaire than parents of children with fewer adjustment difficulties (Hawk et al., 2013). In addition, participants in the current study (n = 65) and in the main study (n = 143) did not differ significantly on parent-rated inhibitory control or working memory on the Behavior Rating Inventory of Executive Functioning (Gioia, Isquith, Guy, & Kenworthy, 2000) after covarying age at assessment and age at adoption, Wilks’ λ = .99, F(2,203) = .55, ns.

Procedure

PI children and their parents were visited at their homes in Pittsburgh, Philadelphia, or Washington, DC, one time for approximately 2 hours by a researcher. After parent and child consent was obtained, the child completed a task battery, which included two subtests from the Wechsler Intelligence Scales for Children-Fourth Edition (WISC-IV; Wechsler, 2003). The parent completed a questionnaire that asked about family characteristics and the child’s background, early development, medical history, and service use and included the Conners’ Parent Rating Scale-Revised (CPRS; Conners, 2000). Parents and children were compensated US$30 each for their participation.

Measures

Cognitive functioning

The WISC-IV Vocabulary subtest (Wechsler, 2003) measures semantic language development, word knowledge, concept formation, and receptive and expressive language (Sattler & Dumont, 2004) and is similar to the Word Definitions subtest of the Clinical Evaluations of Language Fundamentals-4 (Semel, Wiig, & Seccord, 2003). The WISC-IV Matrix Reasoning subtest (Wechsler, 2003) measures nonverbal perceptual reasoning. These subtests yield scaled scores (M = 10, SD = 3) with lower scores indicating lower functioning. These subtests had high reliability and validity (Sattler & Dumont, 2004; Wechsler, 2003).

Behavior problems

The long, 80-item version of the CPRS (Conners, 2000) requires parents to rate their child’s behavior in the past month on a 4-point scale from 0 (not true at all) to 3 (very much true). The Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) Total ADHD Symptoms scale (18 items) assesses inattention (e.g., “difficulty sustaining attention,” “distractibility”) and hyperactive-impulsive behavior (e.g., “fidgetiness,” “responding out of turn”). The other CPRS scales used in this study were Anxious–Shy (8 items; e.g. “has a lot of fears,” “timid”) and Oppositional (10 items; e.g. “fights,” “defies adult requests”).

CPRS norms are based on a nationally representative standardization sample (N = 2,482). Separate norms are given for boys and girls, in 3-year intervals, for ages 3–17. The CPRS yields T scores (higher scores indicate greater problems) which have a population mean of 50 (SD = 10), and scores 1.5 SD above the mean are considered to be in the clinical range. Adequate reliability and validity coefficients have been reported for the CPRS (Conners, 2000). The parents of one PI child did not complete the CPRS.

Parents reported on whether their child was ever given a professional diagnosis of various psychiatric disorders. As expected, ADHD diagnosis was significantly associated with scoring in the clinical range on the Total ADHD Symptoms scale, χ² (1, N = 73) = 20.19, p < .001, supporting the validity of this measure. While the rate of ADHD diagnosis was high in the total PI sample (34%), rates were low (≤ 6%) for other psychiatric disorders.

Learning support service use

Parents provided whether their child received learning support services through an Individualized Education Plan (IEP), 504 service agreement, or a learning support/resource classroom. Public schools in the United States are mandated to provide supportive instructional services to children who qualify; these services are most often associated with either an IEP or Section 504. The IEP outlines and tracks a child’s specialized instruction and related educational services, goals, and achievements. The 504 plan outlines child access to accommodations that will ensure his/her academic success and access to the learning environment. We also asked a general question about child participation in a learning support/resource classroom. If the parent endorsed the child’s current use of any of these services, the child was coded as using learning support services in school.

Overall, 32% (23/73) of PI children were currently receiving learning support services (see Table 2). Ten children were reported to be using more than one service (IEP, Section 504, or resource classroom). Parents were also asked about the child’s primary educational disability as listed on the IEP and the diagnosis associated with the 504 plan. Children with IEPs had the following primary educational disabilities: specific learning disability (n = 7), speech/language impairment (n = 4), emotional/behavioral disability (n = 2), and developmental delay (n = 1). Six out of seven 504 plans (see Table 2) were associated with an ADHD diagnosis.

Table 2.

Cognitive functioning, behavior problems, and learning support service use in PI children adopted ≤ 9 and ≥ 14 months.

	Age at adoption (months)
	≤9 (n = 39)	≥14 (n = 34)
	M (SD)	M (SD)	F(1,71)	p
WISC-IV Vocabulary	11.62 (1.44)	10.03 (2.22)	14.69	.000
WISC-IV Matrix reasoning	10.77 (2.77)	9.82 (3.11)	2.00	.16
CPRS Total ADHD symptoms	52.61 (10.24)	62.26 (12.58)	15.00	.000
CPRS Anxious–Shy	49.79 (10.16)	52.68 (12.79)	1.37	.25
CPRS Oppositional	52.29 (11.94)	56.82 (12.12)	3.14	.08
	% (n)	% (n)	χ² (1, N = 73)	p
Learning support service use composite^a	21% (8)	44% (15)	4.69	.03
IEP	13% (5)	27% (9)	2.18	.14
504 service agreement	5% (2)	15% (5)	1.92	.17
Learning support/resource classroom use	10% (4)	29% (10)	4.30	.04

Note. Bolded means are significantly higher than the CPRS standardization sample mean of 50 (SD = 10); analyses of CPRS and WISC-IV control for age at assessment; CPRS = Conners’ Parent Rating Scale; WISC-IV = Wechsler Intelligence Scale for Children, 4th edition; IEP = individualized education program.

^aIEP, 504 service agreement, or use of a learning support/resource classroom.

Pre-adoption history

Parent-reported age at adoption was strongly correlated with time in an institution, r = .94, p < .001, reflecting that most children were placed in institutions in the first few months of life. Age at adoption was used in analyses rather than time in an institution because it was more frequently and likely more accurately reported.

Parents were asked about their child’s birth weight and whether their child’s birth was premature. Birth weight was reported for 70% of the sample, and prematurity data were given for 79% of the sample. Of the children with birth-weight data, 35% (18/51) were coded as low birth weight (LBW; < 5.5 pounds). As expected, prematurity was significantly associated with LBW, χ² (1, N = 51) = 31.63, p < .001.

Post-adoption history

Parents indicated their highest level of education completed, family income before taxes, and marital/partnership status. These responses were dichotomized as shown in Table 1.

Statistical analysis

We investigated whether attention problems and lower cognitive functioning associated with early institutionalization were responsible for elevated rates of using learning support services among PI children adopted ≥ 14 months relative to PI children adopted ≤ 9 months using standard tests of statistical mediation. Four criteria must be met to establish evidence of mediation (Baron & Kenny, 1986; MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002). First, an association between the exposure and outcome must be demonstrated. Here, we examined group differences in learning support service use using chi-square tests and logistic regression.

Second, the exposure must be associated with the mediator. We examined differences in cognitive functioning, attention problems, anxiety, and oppositional behavior between PI children adopted ≥ 14 months and PI children adopted ≤ 9 months using univariate analyses of variance (ANOVAs) with group (adopted ≥ 14 vs. ≤ 9 months) as a between-subjects factor. Third, the mediator must be associated with the outcome after partialling out the effect of the exposure. Here, we examined the associations of attention problems and vocabulary with the use of learning support services using logistic regression. Age at assessment was included as a covariate in all analyses, and statistical significance was evaluated using .05-level, two-sided tests.

Mediation is then tested by examining the degree of attenuation in the association between the exposure and outcome in a model that includes the mediator. Significant attenuation of this association provides evidence of a significant indirect effect of the exposure on the outcome through the mediator (Baron & Kenny, 1986; MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002). Here, we tested the significance of the mediator using a bootstrapping approach that provides bias-corrected confidence intervals (Preacher & Hayes, 2008). Confidence intervals that do not include zero indicate significant mediation.

Results

Age at assessment was significantly correlated with attention problems, r = .24, p = .04, and oppositional behavior, r = .24, p = .04, but not anxiety, r = .17, ns, cognitive functioning, r = −.08 to −.18, ns, or learning support service use, r = .18, ns. Time in adoptive home was not significantly correlated with outcome or mediator variables, r = .04–.16, ns. There were no gender differences in behavior problems, t(70) = .26–.78, ns, cognitive functioning, t(71) = .27–.37, ns, or learning support service use, χ² (1, N = 73) = .91, ns. These preliminary analyses supported including age at assessment as a covariate in the main analyses.

Analyses were conducted to examine the validity of the age at adoption groups. There were no significant correlations between age at adoption and outcome or mediators within either the PI group adopted ≤ 9 months (r = .02–.22, ns) or the PI group adopted ≥ 14 months (r = .01–.29, ns). In addition, we examined these associations across the total PI sample with age at adoption as a continuous variable. Age at adoption was correlated significantly with attention problems, r = .30, p = .01, and oppositional behavior, r = .28, p = .02, but not with anxiety, r = .21, p = .08, vocabulary, r = −.19, p = .11, matrix reasoning, r = −.13, ns, or academic difficulties, r = .13, ns. These analyses were limited in terms of finding a linear association due to the recruitment of two separate age at adoption groups (i.e., adopted ≤ 9 [range: 4–9 months] and adopted ≥ 14 months [range: 14–48 months]). Nonetheless, together these results support the analysis of age at adoption in terms of two groups.

Group differences in learning support service use

PI children adopted ≥ 14 months had a significantly higher rate of using learning support services compared to those adopted ≤ 9 months after covarying age at assessment (β = 1.11, SE = .54, p = .04, odds ratio = 3.02). When broken down into the type of service received, the groups differed significantly in terms of percentage in a learning support/resource classroom but not in terms of IEP or 504 agreement services (see Table 2).

Group differences in potential mediators

Cognitive functioning

As shown in Table 2, PI children adopted ≥ 14 months had significantly lower Vocabulary scores compared to those adopted ≤ 9 months. Neither group had significantly lower Vocabulary scores than the WISC-IV normative sample (M = 10, SD = 3). The older- and younger-adopted PI groups did not differ significantly in Matrix Reasoning scores.

Behavior problems

As shown in Table 2, PI children adopted ≥ 14 months had significantly higher T scores on the Total ADHD Symptoms scale than those adopted ≤ 9 months. PI children adopted ≥ 14 months had higher attention problems than the CPRS normative sample, one-sample t(33) = 5.69, p < .001. PI children adopted ≤ 9 months did not differ significantly from the CPRS normative sample on attention problems, one-sample t(37) = 1.57, ns. PI children adopted ≥ 14 and ≤ 9 months did not differ significantly on the Anxious-Shy or Oppositional scales.

Associations among attention, vocabulary, and learning support service use

Examination of bivariate correlations revealed that Total ADHD Symptoms T score was significantly correlated with Vocabulary scaled score, r = −.48, p < .001, and with the use of learning support services, r = .56, p < .001. Vocabulary scaled score was significantly correlated with learning support service use, r = −.48, p < .001.

Logistic regression indicated that PI children with attention problems were significantly more likely to use learning support services in school after accounting for age at adoption group and age at assessment (β = .11, SE = .03, p = .000, odds ratio = 1.11). Similarly, in a separate analysis, PI children with lower vocabulary were significantly more likely to use learning support services in school after accounting for age at adoption group and age at assessment (β = −.60, SE = .19, p = .002, odds ratio = .55).¹

Mediation model

We present the total effect of early institutionalization on learning support service use as well as the indirect effect of early institutionalization once the mediators are included in the mediation model (Preacher & Hayes, 2008). The association between early institutionalization and the use of learning support services (β = 1.08, SE = .54, p = .04) is attenuated when attention problems and vocabulary are added to the model (β = −.18, SE = .72, ns; see Figure 1). The indirect effects of attention problems (point estimate = .76, SE = .57, 95% CI .13–1.99) and vocabulary (point estimate = .69, SE = .58, 95% CI .04–2.01) are statistically significant. These results indicate that the difference in the use of learning support services between PI children adopted ≥ 14 months compared with those adopted ≤ 9 months is explained by differences in attention problems and vocabulary between these groups.

Figure 1.

Attention problems and vocabulary mediate the association of age at adoption with the use of learning support services in school. Paths represent unstandardized beta coefficients (standard errors in parentheses). Path b denotes the association of attention problems (or vocabulary) with learning support service use, controlling for age at adoption. Direct effect c’ denotes the association of age at adoption with learning support service use, controlling for the mediation effect. Age at adoption was coded 0 = ≤ 9 months, 1 = ≥ 14 months. Learning support service use was coded 0 = not currently using services; 1 = currently using services. *p < .05; **p < .01; ***p < .001.

Although bootstrapping has many advantages over other methods of testing mediation, we also ran Sobel tests (MacKinnon et al., 2002). The indirect effect of attention problems was significant (z score = 2.20, SE = .36, p = .03), and the indirect effect of vocabulary approached significance (z score = 1.89, SE = .37, p = .06). Note that one limitation of the Sobel test is that it is underpowered in smaller samples (Preacher & Hayes, 2008).

Discussion

The goals of this study were to examine attention, cognitive, and academic functioning in school-age children adopted from psychosocially depriving institutions and to explore attention and cognition as mediators of the association between early psychosocial deprivation and academic outcomes. Results indicated that post-institutionalized (PI) children adopted after the age of 14 months were more likely to use learning support services than PI children adopted before 9 months of age (44% vs. 21%, respectively). While this result suggests that prolonged early psychosocial deprivation may lead to academic difficulties, it should be noted that most of the PI children were not receiving learning support services.

It should be noted that learning support service use is an indirect measure of academic functioning that may reflect other influences. Given the similar demographic characteristics of the PI groups, the increased service use for older-adopted PI children over younger-adopted PI children is not likely due to differential service availability. However, service use may depend on how much parents advocate for their children, and adoptive parents may have a lower threshold for obtaining services (Miller et al., 2000). Therefore, it is possible that service use may overestimate the actual rate of academic difficulties in the PI sample as a whole.

PI children adopted after 14 months had greater attention problems and lower vocabulary than PI children adopted before 9 months of age. In comparison to standardized test norms, PI children adopted after 14 months had significantly greater attention problems but not lower vocabulary scores; PI children adopted before 9 months did not differ from the standardization sample on either attention or vocabulary. Results for PI children’s vocabulary are consistent with previous research showing poorer performance relative to comparison groups but scores within the average range (Loman et al., 2009). Older-adopted PI children’s average score on the attention problems measure (T = 62) approached the clinical range (T ≥ 65; Conners, 2000).

Older-adopted PI children spent a large proportion of their early lives in psychosocially depriving institutions that provided adequate physical resources but inadequate caregiver responsiveness and supportive relationships with caregivers. Prolonged early psychosocial deprivation may have led to attention and language difficulties years after adoption. Studies of never-institutionalized children suggest that aspects of caregiver sensitive responsiveness and scaffolding facilitate children’s early attention and language development (Tamis-LeMonda et al., 2001). Moreover, early childhood may be a sensitive period during which expected environmental inputs are needed for typical brain development to proceed (Marshall & Kenney, 2009).

These results were obtained even though the adoptive families of the older-adopted PI children did not differ in parental education and income compared with those of the younger-adopted PI children. Overall, the PI children were living in affluent homes in which adoptive parents were highly educated. Thus, there is little evidence that older-adopted PI children exhibited poorer functioning because they lacked cognitive and social stimulation or enrichment opportunities after adoption.

In contrast to attention and vocabulary, the PI groups did not differ significantly in nonverbal reasoning, anxiety, or oppositional behavior. Results for anxiety and oppositional behavior are consistent with prior literature showing mixed findings for these behavior outcomes and some suggestion that they may not be associated with early deprivation (Rutter et al., 2007). Results suggest that within the cognitive domain, language may be more affected by early deprivation than spatial reasoning. This particular result is consistent with at least one prior study (Eigsti et al., 2011), but others have demonstrated both spatial reasoning and verbal difficulties among PI children (Fox et al., 2011; Loman et al., 2009).

Attention problems and vocabulary significantly mediated the association between early psychosocial deprivation and learning support service use. These results are fairly consistent with a previous study that examined this question in children adopted from severely depriving institutions (Beckett et al., 2007). Both studies showed significant mediating roles of cognition and attention problems. However, the current study suggests a larger effect for attention problems and highlights a potentially greater role of language over nonverbal reasoning within the cognitive domain. These differences may be due in part to methodological differences between studies.

Attention and language problems could contribute to PI children’s academic difficulties. Many studies of PI children have documented an elevated risk of attention problems from preschool- to school-age (e.g., Audet & Le Mare, 2011). Attention abilities, including sustaining focus and suppressing task-irrelevant information, are needed for many typical school activities. Attention problems are associated with not rehearsing information in mind, poor approaches to homework and assignments, poor planning, and less time on-task in the classroom (Li-Grining, Votruba-Drzal, Maldonado-Carreno, & Haas, 2010). Poorer semantic language skills may also make it challenging to succeed in school, especially in reading (Justice, Bowles, Pence Turnbull, & Skibbe, 2009). Children who enter school with these problems may fall further and further behind over time. Therefore, it is particularly relevant that PI children have been found to demonstrate attention and language difficulties at preschool-age (Audet & Le Mare, 2011; Windsor et al., 2011). It is possible that bidirectional influences between attention and language development over time might explain the onset and persistence of academic difficulties.

These findings suggest that PI children should undergo comprehensive evaluations of their attention and language skills shortly after adoption and prior to school entry whenever possible. PI children who do demonstrate these difficulties and receive interventions might be prevented from doing poorly in school. Future research should examine whether preschool-aged PI children benefit from classroom-based interventions targeting attention and language skills.

This study had a number of limitations. The entire PI sample was adopted from Russia or the former Soviet Union which could limit generalizability to children adopted from institutions in other areas of the world (Gunnar et al., 2007). In addition, like other studies of internationally-adopted PI children, this study could not rule out potentially confounding risk factors, such as genetic history or poor prenatal care. Analyses addressing the potentially confounding role of birth circumstances indicated that the PI groups did not differ significantly in adoptive parent-reported prematurity or birth weight. However, group differences in birth weight approached significance (p = .08; see Table 1), and birth weight was reported for 70% of the PI sample (n = 51). Although multiple studies of PI children have not found associations between birth circumstances and developmental outcomes (Bruce, Tarullo, & Gunnar, 2009; Kreppner et al., 2007; Pollak et al., 2010), future research should continue to explore the effect of birth circumstances in PI samples.

Age at adoption effects are unlikely attributable to selective adoption based on prior studies, such as the Bucharest Early Intervention Project which used a randomized controlled design (Fox et al., 2011), but this cannot be ruled out in the current study. The response rate was fairly low but previous analyses of this sample point to a reduced possibility of response bias (Hawk et al., 2013). Although the PI groups did not differ in the percentage of children taking stimulant medication on the day of testing, it is possible that medication use affected children’s performance on cognitive assessments. This study used a cross-sectional design so we could not examine whether cognitive and behavioral problems preceded academic difficulties.

There were limitations to using service use as a measure of academic functioning in mediation analyses. Children were eligible for learning support services based on specific problems. Therefore, learning support service use may not represent an outcome independent of the educational diagnosis associated with services, which in some cases was either speech or attention difficulties. Finally, although bootstrapping is a preferred method of testing mediation especially with smaller samples, parameter estimates may be more robust with larger samples, for instance, those with 100 or more participants (MacKinnon, Lockwood, & Williams, 2004). Therefore, interpretation of bootstrapping results should take into account the smaller sample size.

This study focused on school-aged children adopted from psychosocially-depriving institutions that provided adequate physical resources but failed to provide a consistent set of responsive caregivers. Results indicated that PI children adopted at older ages had greater attention problems, lower vocabulary, and an increased likelihood of using learning support services compared to those adopted at younger ages. Attention problems and lower vocabulary mediated the association between early psychosocial deprivation and academic difficulties. Nonverbal reasoning, anxiety, and oppositional behavior did not differ significantly between groups. These findings imply that intervention efforts aimed at improving PI children’s attention and language may be beneficial in terms of supporting academic achievement.

Footnotes

Acknowledgements

This research was supported by NICHD grants HD39017 and HD050212 to Robert B. McCall and Christina J. Groark. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NICHD or NIH. The authors thank Larisa Mason, Mary Graber, Larry Fish, Johana Rosas, Megan Julian, and Brandi Hawk for assistance in implementing this study. The authors are especially grateful to the parents and children who participated in this study.

Funding

This research was supported by NICHD grants HD39017 and HD050212 to Robert B. McCall and Christina J. Groark

Note

References

Audet

Le Mare

(2011). Mitigating effects of the adoptive caregiving environment on inattention/overactivity in children adopted from Romanian orphanages. International Journal of Behavioral Development, 35(2), 107–115.

Baron

R. M.

Kenny

D. A.

(1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.

Beckett

Maughan

Rutter

Castle

Colvert

Groothues

C.,

Sonuga-Barke

E. J.

(2007). Scholastic attainment following severe early institutional deprivation: A study of children adopted from Romania. Journal of Abnormal Child Psychology, 35(6), 1063–1073.

Beverly

McGuinness

Blanton

(2008). Communication challenges for children adopted from the former Soviet Union. Language, Speech, and Hearing Services in Schools, 39, 1–11.

Breslau

Peterson

Miller

Lucia

V. C.

Bohnert

Nigg

(2010). Change in teachers’ ratings of attention problems and subsequent change in academic achievement: A prospective analysis. Psychological Medicine, 40, 159–166.

Bruce

Tarullo

A. R.

Gunnar

M. R.

(2009). Disinhibited social behavior among internationally adopted children. Development and Psychopathology, 21, 157–171.

Castle

Rutter

Beckett

Colvert

Groothues

Hawkins

A.,

Sonuga-Barke

(2006). Service use by families with children adopted from Romania. Journal of Children’s Services, 1, 5–14.

Conners

K. C.

(2000). Conners’ Parent Rating Scale–Revised(L). New York, NY: Multi-Health Systems.

Duncan

G. J.

Dowsett

C. J.

Claessens

Magnuson

Huston

A. C.

Klebanov

P.,

Japel

(2007). School readiness and later achievement. Developmental Psychology, 43, 1428–1466.

10.

Eigsti

I.-M.

Weitzman

Schuh

De Marchena

Casey

B. J.

(2011). Language and cognitive outcomes in internationally adopted children. Development and Psychopathology, 23, 629–646.

11.

Fox

N. A.

Almas

A. N.

Degnan

K. A.

Nelson

C. A.

Zeanah

C. H.

(2011). The effects of severe psychosocial deprivation and foster care intervention on cognitive development at 8 years of age: Findings from the Bucharest Early Intervention Project. Journal of Child Psychology and Psychiatry, 52(9), 919–928.

12.

Gioia

G. A.

Isquith

P. K.

Guy

S. C.

Kenworthy

(2000). The BRIEF professional manual. Odessa, FL: Psychological Assessment Resources.

13.

Glennen

Bright

B. J.

(2005). Five years later: Language in school-age internationally adopted children. Seminars in Speech and Language, 26(1), 86–101.

14.

Gunnar

M. R.

(2001). Effects of early deprivation: Findings from orphanage-reared infants and children. In Nelson

C. A.

Luciana

(Eds.), Handbook of Developmental Cognitive Neuroscience (pp. 617–629). Cambridge, MA: MIT Press.

15.

Gunnar

M. R.

van Dulmen

M. H. M.

(2007). Behavior problems in postinstitutionalized internationally adopted children. Development and Psychopathology, 19, 129–148.

16.

Hawk

B. N.

Wright

Julian

M. M.

Rosas

J. M.

Merz

E. C.

McCall

R. B.

(2013). Potential selective responding in a parent questionnaire study of post-institutionalized children. Adoption Quarterly, 16(2), 97–107.

17.

Hellerstedt

W. L.

Madsen

N. J.

Gunnar

M. R.

Grotevant

H. D.

Lee

R. M.

Johnson

D. E.

(2008). The International Adoption Project: Population-based surveillance of Minnesota parents who adopted children internationally. Maternal and Child Health Journal, 12(2), 162–171.

18.

Justice

L. M.

Bowles

R. P.

Pence Turnbull

K. L.

Skibbe

L. E.

(2009). School readiness among children with varying histories of language difficulties. Developmental Psychology, 45(2), 460.

19.

Kreppner

J. M.

Rutter

Beckett

Castle

Colvert

Groothues

C.,

Sonuga-Barke

E. J. S.

(2007). Normality and impairment following profound early institutional deprivation: A longitudinal follow-up into early adolescence. Developmental Psychology, 43(4), 931–946.

20.

Landry

S. H.

Smith

K. E.

Swank

P. R.

(2006). Responsive parenting: Establishing early foundations for social, communication, and independent problem-solving skills. Developmental Psychology, 42(4), 627–642.

21.

Le Mare

Audet

Kurytnik

(2007). A longitudinal study of service use in families of children adopted from Romanian orphanages. International Journal of Behavioral Development, 31(3), 242–251.

22.

Li-Grining

C. P.

Votruba-Drzal

Maldonado-Carreño

Haas

(2010). Children’s early approaches to learning and academic trajectories through fifth grade. Developmental Psychology, 46(5), 1062–1077.

23.

Loman

M. M.

Wiik

K. L.

Frenn

K. A.

Pollak

S. D.

Gunnar

M. R.

(2009). Postinstitutionalized children’s development: Growth, cognitive, and language outcomes. Journal of Developmental and Behavioral Pediatrics, 30(5), 426–434.

24.

MacKinnon

D. P.

Lockwood

C. M.

Hoffmann

J. M.

West

S. G.

Sheets

(2002). A comparison of methods to test mediation and other intervening variable effects. Psychological Methods, 7, 83–104.

25.

MacKinnon

D. P.

Lockwood

C. M.

Williams

(2004). Confidence limits for the indirect effect: Distribution of the product and resampling methods. Multivariate Behavioral Research, 39(1), 99–128.

26.

Marshall

P. J.

Kenney

J. W.

(2009). Biological perspectives on the effects of early psychosocial experience. Developmental Review, 29, 96–119.

27.

Merz

E. C.

McCall

R. B.

(2010). Behavior problems in children adopted from psychosocially depriving institutions. Journal of Abnormal Child Psychology, 38(4), 459–470.

28.

Miller

Fan

Grotevant

Christensen

Coyl

Van Dulmen

(2000). Adopted adolescents’ overrepresentation in mental health counseling: Adoptees’ problems or parents’ lower threshold for referral? Journal of the American Academy of Child and Adolescent Psychiatry, 39, 1504–1511.

29.

Muhamedrahimov

R. J.

(1999). New attitudes: Infant care facilities in St. Petersburg, Russia. In Osofsky

J. D.

Fitzgerald

H. E.

(Eds.). WAIMH handbook of infant mental health. Vol. 1. Perspectives on infant mental health (pp. 245–294). New York, NY: Wiley.

30.

NICHD Early Child Care Research Network. (2003). Do children’s attention processes mediate the link between family predictors and school readiness? Developmental Psychology, 39, 581–593.

31.

Olson

S. L.

Bates

J. E.

Sandy

J. M.

Schilling

E. M.

(2002). Early developmental precursors of impulsive and inattentive behavior: From infancy to middle childhood. Journal of Child Psychology and Psychiatry, 43(4), 435–448.

32.

Ouellette

G. P.

(2006). What’s meaning got to do with it: The role of vocabulary in word reading and reading comprehension. Journal of Education & Psychology, 98, 554–566.

33.

Pears

K. C.

Fisher

P. A.

Bruce

Kim

H. K.

Yoerger

(2010). Early elementary school adjustment of maltreated children in foster care: The roles of inhibitory control and caregiver involvement. Child Development, 81(5), 1550–1564.

34.

Pollak

S. D.

Nelson

C. A.

Schlaak

M. F.

Roeber

B. J.

Wewerka

S. S.

Wiik

K. L.,

Gunnar

M. R.

(2010). Neurodevelopmental effects of early deprivation in postinstitutionalized children. Child Development, 81(1), 224–236

35.

Preacher

K. J.

Hayes

A. F.

(2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891.

36.

Rueda

M. R.

Posner

M. I.

Rothbart

M. K.

(2005). The development of executive attention: Contributions to the emergence of self-regulation. Developmental Neuropsychology, 28, 573–594.

37.

Rutter

Beckett

Castle

Colvert

Kreppner

Mehta

M.,

Sonuga-Barke

(2007). Effects of profound early institutional deprivation: An overview of findings from a UK longitudinal study of Romanian adoptees. European Journal of Developmental Psychology, 4(3), 332–350.

38.

Sattler

J. M.

Dumont

(2004). Assessment of children: WISC-IV and WPPSI-III supplement. San Diego, CA: Jerome M. Sattler Publisher, Inc.

39.

Semel

Wiig

E. H.

Seccord

W. A.

(2003). Clinical evaluation of language fundamentals (4th ed.). San Antonio, TX: The Psychological Corporation.

40.

Sheridan

Drury

McLaughlin

Almas

(2010). Early institutionalization: Neurobiological consequences and genetic modifiers. Neuropsychology Review, 20(4), 414–429.

41.

St. Petersburg–USA Orphanage Research Team. (2005). Characteristics of children, caregivers, and orphanages for young children in St. Petersburg, Russian Federation. Journal of Applied Developmental Psychology, 26(5), 477–506.

42.

Stevens

S. E.

Sonuga-Barke

E. J.

Kreppner

J. M.

Beckett

Castle

Colvert

E.,

Rutter

(2008). Inattention/overactivity following early severe institutional deprivation: Presentation and associations in early adolescence. Journal of Abnormal Child Psychology, 36(3), 385–398.

43.

Tamis-LeMonda

Bornstein

Baumwell

(2001). Maternal responsiveness and children’s achievement of language milestones. Child Development, 72, 748–767.

44.

Tirella

L. G.

Chan

Cermak

S. A.

Litvinova

Salas

K. C.

Miller

L. C.

(2008). Time use in Russian baby homes. Child: Care, Health and Development, 34(1), 77–86.

45.

van IJzendoorn

M. H.

Juffer

Poelhuis

C. W.

(2005). Adoption and cognitive development: A meta-analytic comparison of adopted and nonadopted children’s IQ and school performance. Psychology Bulletin, 131(2), 301–316.

46.

Wechsler

(2003). Wechsler Intelligence Scale for Children (4th ed.). San Antonio, TX: The Psychological Corporation.

47.

Windsor

Benigno

J. P.

Wing

C. A.

Carroll

P. J.

Koga

S. F.

Nelson

C. A.,

Zeanah

C. H.

(2011). Effect of foster care on young children’s language learning. Child Development, 82, 1040–1046.

48.

Young

A. R.

Beitchman

J. H.

Johnson

Douglas

Atkinson

Escobar

Wilson

(2002). Young adult academic outcomes in a longitudinal sample of early identified language impaired and control children. Journal of Child Psychology and Psychiatry, 43, 635–645.