“The Caregiving Context in Institution-Reared and Family-Reared Infants and Toddlers in Romania.”

Tulane University, LA, USA.
Journal of Child Psychology and Psychiatry (Impact Factor: 6.46). 03/2007; 48(2):210-8. DOI: 10.1111/j.1469-7610.2006.01694.x
Source: PubMed
We assess individual differences in the caregiving environments of young children being raised in institutions in Romania in relation to developmental characteristics such as physical growth, cognitive development, emotional expression, and problem and competence behaviors.
Videotaped observations of the child and favorite caregiver in their 'home' environment were coded for caregiving quality, and this was related to child characteristics. Child emotional reactivity was assessed during responses to interactional tasks. Cognitive development was assessed from child responses to the Bayley Scales of Infant Development. Data regarding problem behaviors and competence were obtained from caregiver report. Children reared in institutions were compared on all of these measures to never institutionalized children to assist gauging degree of impairment.
Children raised in institutions demonstrated marked delays in cognitive development, poorer physical growth, and marked deficits in competence. Individual differences in caregiving environment were associated with cognitive development, competence, and negative behavior among these young children being reared in institutions.
These data confirm previous findings regarding deficits associated with institutional care and extend our understanding of the impact of individual differences in caregiving quality on the development of young children in institutions.


Available from: Charles H Zeanah
The caregiving context in institution -reared and
family-reared infants and toddlers in Romania
Anna T. Smyke,
Sebastian F. Koga,
Dana E. Johnson,
Nathan A. Fox,
J. Marshall,
Charles A. Nelson,
Charles H. Zeanah,
and the BEIP Core Group
Tulane University, USA;
University of Minnesota, USA;
University of Maryland, USA;
Temple University, USA;
Harvard University, USA
Background: We assess individual differences in the caregiving environments of young children
being raised in institutions in Romania in relation to developmental characteristics such as physical
growth, cognitive development, emotional expression, and problem and competence
behaviors. Method: Videotaped observations of the child and favorite caregiver in their ‘home’ envir-
onment were coded for caregiving quality, and this was related to child characteristics. Child emotional
reactivity was assessed during responses to interactional tasks. Cognitive development was assessed
from child responses to the Bayley Scales of Infant Development. Data regarding problem behaviors and
competence were obtained from caregiver report. Children reared in institutions were compared on all
of these measures to never institutionalized children to assist gauging degree of impair-
ment. Results: Children raised in institutions demonstrated marked delays in cognitive development,
poorer physical growth, and marked deficits in competence. Individual differences in caregiving envir-
onment were associated with cognitive development, competence, and negative behavior among these
young children being reared in institutions. Conclusions: These data confirm previous findings
regarding deficits associated with institutional care and extend our understanding of the impact of
individual differences in caregiving quality on the development of young children in institu-
tions. Keywords: Infancy, orphans, institutionalization, caregiving quality, cognitive deficits, physical
growth, emotional expression, competence. Abbreviations: BEIP, Bucharest Early Intervention Pro-
ject; BSID-II, Bayley Scales of Infant Development II; DQ, Developmental Quotient; IG, institution
group; ITSEA, Infant–Toddler Social Emotional Assessment; MDI, Mental Development Index; NIG,
never institutionalized group; ORCE, Observational Record of the Caregiving Environment.
Historically, institutions have been the most com-
mon form of care for orphaned and abandoned
children throughout the world, and they remain so
today. Because institutional rearing often involves
social and even material deprivation, disturbances of
growth, cognitive development, and language as well
as greater levels of problem behaviors have been
noted for more than 50 years among young children
raised in institutions (MacLean, 2003; Zeanah,
Smyke, & Settles, 2006).
Studies published in the past 15 years have
demonstrated impairments in young children being
raised in Romanian institutions and in those adop-
ted out of institutions. Children being raised in
institutions have demonstrated cognitive delays,
serious impairments in social behavior (Smyke,
Dumitrescu, & Zeanah, 2002; Zeanah, Smyke, Koga,
& Carlson, 2005), and abnormalities of cortisol
regulation compatible with high levels of stress
(Carlson & Earls, 1997). Young children adopted out
of institutional care often have persisting cognitive,
socioemotional, and health problems (Castle et al.,
1999; Fisher, Ames, Chisholm, & Savoie, 1997;
Gunnar, 2000; Johnson, 2000; Rutter et al., 1999).
Despite this, Rutter and colleagues (2001) noted
surprising variability in the degree to which children
were affected by their institutional experience and
significant, if not full, recovery has been observed
among many children adopted out of Romania
(Fisher et al., 1997).
Interestingly, although studies of children adopted
out of institutions have shown that many early
developmental deficits improve, they have not shown
consistently that individual differences in outcomes
are related to individual differences in adoptive
family environments. This raises the question of
whether differences noted in children adopted out of
institutions may relate to differences in the care-
giving environments of the institutions from which
they were adopted. Variability in quantity and qual-
ity of care has been noted across and within insti-
tutions in previous research, but too few studies
have included direct observations of quantity and
quality of care in institutions for young children. In
fact, better quality institutions have been associated
with less cognitive impairment in children raised
there (Roy, Rutter, & Pickles, 2000; Tizard & Rees,
1974; Vorria et al., 1998). Still, demonstrating
associations between caregiving characteristics in
institutions and individual characteristics of the
children being raised there has not been reported.
Institutional rearing in Romania often has
Conflict of interest statement: No conflicts declared.
Journal of Child Psychology and Psychiatry 48:2 (2007), pp 210–218 doi:10.1111/j.1469-7610.2006.01694.x
2007 The Authors
Journal compilation 2007 Association for Child and Adolescent Mental Health.
Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
Page 1
been characterized by rotating shifts of multiple
caregivers, large child to caregiver ratios, and limited
social, cognitive, and language stimulation (Carlson
& Earls, 1997; Kaler & Freeman, 1994; Smyke et al.,
2002; Zeanah et al., 2003). Staff often have inad-
equate education and face challenging working
conditions. Nevertheless, quality of care varies
among and within these settings, and it is important
to know if these differences are associated with dif-
ferences in young children’s development.
In previous studies of children raised in institu-
tions in Romania and elsewhere, children raised in
institutions are compared to children raised in fam-
ily settings (foster care or with their birth parents),
without determining whether child characteristics
were related to caregiving characteristics. The
association between quality of care and children’s
development is important to examine because dif-
ferences in children raised in institutions and in
families may be due to variables other than rearing
conditions. Since young children are not placed in
institutions at random, the reasons why some chil-
dren are raised in institutions may contribute
importantly to any outcomes assessed. In Romania,
for example, the chief reason given for child aban-
donment is poverty (Zeanah et al., 2003), which is
associated with limited prenatal care, maternal
malnutrition, and prenatal exposure to alcohol and
other substances, all of which are known to increase
the risk of compromised postnatal development
(Nelson & Bosquet, 2000).
In this study, we explored whether individual dif-
ferences in the caregiving environment were asso-
ciated with individual differences in young children’s
development. Because the children in this study
were recruited from all six institutions for young
children in Bucharest, they are broadly representa-
tive of institutional care in Romania. Although
direction of effects cannot be determined from cross-
sectional designs, associations between individual
differences in the caregiving environment and chil-
dren’s development are compatible with the relative
importance of early experiences of caregiving in
institutions. A major purpose of the present invest-
igation is to determine if we can demonstrate such
an association.
A second purpose of the current investigation is to
extend the examination of child characteristics to
areas other than cognitive and physical develop-
ment, which have been the most frequently assessed
characteristics. For example, there is good reason to
expect that young children being raised in institu-
tions will demonstrate significant behavioral prob-
lems. In addition to risk factors that precede
placement, studies of children adopted out of insti-
tutions have found aggression, temper tantrums,
hyperactivity, difficulty with attention and anxiety
problems in young children (Ellis, Fisher, & Zaharie,
2004; Fisher et al., 1997; Gunnar, 2000; Kreppner,
O’Connor, & Rutter, 2001; Rutter et al., 2001; Rutter
et al., 1999). Another purpose of the current study is
to examine behavior problems in children receiving
institutional care in relation to caregiving quality.
Curiously, there have been no previous attempts
to examine either competence or emotional expres-
sion among young children being raised in institu-
tions, despite the presumed importance of these
characteristics in explaining variability in outcomes
(Rutter et al., 2001). Young children in institutions
who are more competent and who display more
positive affect may be better able to recruit support
from caregivers, to withstand deprivation more
heartily, and/or to recover from early setbacks more
readily than less competent and more negative chil-
dren. Another purpose of the present investigation is
to determine if caregiving quality is related to these
Because the measures we employed had not been
used in published reports of Romanian children, our
initial analyses compared outcomes in children
reared in institutions and community children to
assess developmental characteristics in these two
groups of children. In keeping with the literature on
institutional care, we predicted substantial differ-
ences in every domain of development. In this report,
we were specifically interested in individual differ-
ences among children being reared in institutions.
Therefore, we predicted that child characteristics,
caregiving quality, and amount of adverse rearing
environment would be associated with develop-
mental outcomes in these young children across a
range of different domains. We also had specific
hypotheses regarding caregiving quality and amount
of adverse rearing environment.
We predicted that better caregiving quality
in young children being raised in institutions in
Romania would be related to better outcomes in
cognitive development, problem behaviors, emotion
expression, competence, and physical development,
after controlling for child age, ethnicity, gender, and
birthweight. In keeping with previous research, we
also predicted that developmental differences would
be greater among children who had experienced a
greater percentage of their lives in institutions. We
expected that a higher percentage of a child’s life
spent in an institution would be associated with
poorer cognitive development, greater numbers of
behavior problems, reduced levels of competence,
and less positive expression of emotion.
Participants were 208 infants and toddlers ranging in
age from 5–31 months (M ¼ 20.65, SD ¼ 7.26) at the
time of baseline evaluation, which is the subject of this
report (see Zeanah et al., 2003). Institutional caregivers
or parents of community children provided some of
the data regarding children’s developmental status.
Caregiving and development in institutions and families 211
2007 The Authors
Journal compilation 2007 Association for Child and Adolescent Mental Health.
Page 2
Participants were drawn from all of the placement
centers (institutions for young children) in the 6 sectors
of Bucharest. All children less than 31 months of age in
April 2001 living in institutions in Bucharest were
included, with the exception of children scheduled for
imminent adoption and children with serious handi-
capping conditions (e.g., fetal alcohol syndrome, severe
cerebral palsy). Children were placed in the institu-
tional setting at a variety of ages. Of the 125 subjects in
the Institution (IG) group, 56 had resided in an insti-
tution since birth, usually after abandonment in the
maternity hospital. While the average percentage of life
institutionalized was 85.94 (SD ¼ 21.14), the median
percentage of life instititutionalized was 98.16. The
variation in percent of life institutionalized was con-
ceptualized as a means to begin to deconstruct the age/
exposure to institutionalization inter-relationship.
Children living in institutions were in the custody of
local governmental officials and informed consent was
provided by legally responsible personnel. In addition,
we obtained consent from all birth parents whom we
could locate (n ¼ 29) for children to be randomized into
foster care. Eleven children originally cleared for parti-
cipation in the study (6 boys, 5 girls; 7 Romanian, 1
Roma, 3 unknown; Mean age ¼ 18.09 months), later
were determined to have conditions that met exclusion
criteria. These children were excluded from analyses,
although their inclusion did not alter any results. In
addition, never institutionalized infants and toddlers
were recruited from a pediatric community health cen-
ter in Bucharest. Seventy-two community families be-
gan the study at baseline. Shortly thereafter, 7 families
chose to discontinue participation (7 girls; 7 Romanian;
Mean age ¼ 21.92 months). Child characteristics may
be found in Table 1. For other details of the study
sample see Zeanah et al. (2003).
After the study was approved by the Institutional Re-
view Boards at the investigators’ universities and by the
National Authority of Child Protection and the Ministry
of Health in Romania, baseline assessments of physical
development and growth, cognitive level, problem be-
haviors and competence, and emotional expression
were obtained on participants from the Institution
(IG) and Never Institutionalized (NIG) groups. Data
collection involved 3 visits to the BEIP laboratory and a
1½-hour home visit either at the child’s home (NIG) or
at the placement center (IG). All measures were
administered by Romanian research assistants who
had received specialized training from the US team.
Cognitive development.
The Bayley Scales of Infant
Development II (BSID-II; Bayley, 1993) were used to
assess developmental status. The Mental Development
Index (MDI), a scaled score, ranged from 50 to 150.
Children who obtained scaled scores below 50 were
assigned a numeric MDI score of 49. Raw scores were
assigned an age equivalent score to enable analyses
when scaled scores <50 were obtained (Lindsey &
Brouwers, 1999). Developmental Quotients (DQ) were
computed for each child (i.e., [age equivalent score/
chronological age] · 100), allowing analyses to be car-
ried out on the entire sample. For later institution-only
regression analyses, we generated a dummy variable
consisting of children who had MDI scores of 50 or
greater (50 ¼ 0) or below 50 (<50 ¼ 1).
Physical growth. Measurements of weight (gm),
height (cm), weight for height, and occipitofrontal cir-
cumference (cm) were converted to z scores (Centers for
Disease Control, 2002). (See Table 2.) Birthweight was
obtained from record review (IG) or from parent report
(NIG). Birthweight information was available for 86% of
the children from the IG and all but 3 of the children
from the NIG. For purposes of data reduction, factor
analysis was conducted, resulting in a single size fac-
tor. This variable was used in more in-depth analyses.
Emotional expression. The temperament/emotion
expression episodes were modified from the puppets
and peek-a-boo episodes in the Laboratory Assessment
of Temperament Battery (LAB-TAB; Goldsmith & Roth-
bart, 1999) to measure emotion expression in response
to social situations. During the episode, a female ex-
aminer introduced a pair of puppets to the child and
acted out a conversation between them. The caregiver
was seated behind the child. For coding, each 2-minute
episode was divided into 30-second epochs to identify
discrete emotion behaviors as well as intensities of
facial and bodily expressions of positive and negative
emotions. Data reduction was based on Kochanska,
Coy, Tjebkes, and Husarek (1998) and resulted in
standardized aggregates for both positive and negative
affect. The negative affect aggregate was composed of
Table 1 Demographic characteristics of institution-reared and family-reared children
Child characteristics Institution group (n ¼ 124) Community group (n ¼ 66) Significance
Age (months) (SD) 21.07 (7.29) 19.32 (7.16) ns
Romanian 56.8% 90.9% ***
Roma (Gypsy) 28.8% 6.1%
Unknown/Other 14.4% 3.0%
Female (%) 50.4% 53.0% ns
Percent of life institutionalized 86.06 (21.09)
(n ¼ 112)
(n ¼ 63)
Birthweight (gm.) (SD) 2834 (596) 3331 (454) ***
Note: ***p < .001, **p < .01, *p < .05, ns ¼ nonsignificant.
212 Anna T. Smyke et al.
2007 The Authors
Journal compilation 2007 Association for Child and Adolescent Mental Health.
Page 3
variables reflecting the expression of fear, anger, and
Problem behavior and competence. The Infant Tod-
dler Social Emotional Assessment (ITSEA; Carter,
Briggs-Gowan, Jones, & Little, 2003), a 195-item
questionnaire, was administered to caregivers/parents
to assess problem behaviors and competencies in chil-
dren 12 months and older. Externalizing (activity/
impulsivity, aggression/defiance, peer aggression);
Internalizing (depression/withdrawal, anxiety, separ-
ation distress, inhibition to novelty); Dysregulation
(negative emotionality, sleep, eating, sensory sensitiv-
ity); and Competence (compliance, imitation/play,
attention, mastery motivation, empathy, prosocial peer
relations) scores were obtained. Scores also were de-
rived for maladaptive behaviors, social relatedness, and
atypical behaviors. Because this measure had been
normed by age group and gender (Carter et al., 2003), a
given child’s score could be compared to scores for
children of similar age and gender.
Principal components factor analysis was used for
data reduction and produced a competence score
(competence, social relatedness) and a negative behav-
ior score (externalizing, internalizing, dysregulated,
maladaptive, atypical) which were used in more
in-depth analyses.
Caregiving environment. The Observational Record
of the Caregiving Environment (ORCE; NICHD Early
Child Care Research Network, 1996) was adapted and
used to assess the child’s caregiving experience in ei-
ther the institution or family setting. Children were
videotaped with their preferred caregiver for hours.
Caregivers were given no special instructions prior to
the videotaping. Coding consisted of both qualitative
ratings and quantitative ratings for caregivers and
children. The former included such items as intrusive-
ness and stimulation of development. Quantitative
items, 55 in all, included: caregiver available, caregiver
interacting, and child unoccupied/watching and were
rated during a series of 30 episodes. Coders first
watched 10 minutes of videotape and then began to
code for the presence or absence of the 55 quantitative
items as they watched for 30 sec and then coded for
30 sec. This was accomplished in 10-minute blocks,
followed by a 2-minute period of observation only. This
pattern was repeated twice. Qualitative ratings were
completed after 10 more minutes of observation.
The training process for coders included orientation
to and study of the quantitative and qualitative items
from the original ORCE manual as well as those ad-
ded for purposes of the BEIP study. Practice tapes
were coded and discussed and then reliability tapes
were coded. In addition, a portion of tapes were
double coded, to ensure ongoing fidelity of the coding
The Caregiving Quality score was obtained by
averaging five qualitative scores (i.e., sensitivity, sti-
mulation of development, positive regard for child, flat
affect [reversed], detachment [reversed]), each of which
received a rating from 1 (not at all characteristic) to 4
(highly characteristic). Caregiving Quality scores ran-
ged from 1.0 to 3.8 (M ¼ 2.39, SD ¼ .65). Scale reliab-
ility was excellent (Cronbach’s alpha ¼ .86). Inter-rater
reliability ranged from .88 to .99 (M ¼ .95). Scores for
caregiver available, caregiver interacting, and child
unoccupied/watching were selected from among the
quantitative variables to examine not only how con-
sistently caregivers were available but also the number
of episodes in which caregivers actually interacted with
the child. For most children living in families, a care-
giver was present in 30/30 of episodes, which repre-
sented a ceiling for this item, thus reducing variability.
Children living in the institution, however, frequently
did not have a caregiver present and available to them.
In fact, there was a wide range of number of episodes in
which this was true, reflecting a range in quality among
the institutions. Children who were unoccupied/
watching were not participating in meaningful activities
with their caregivers or with their peers. This may be a
marker for the degree to which caregivers organized and
supported the participation of the children in mean-
ingful activities. Where appropriate, analyses were
conducted using t-tests for unequal variances.
Results are organized to present initially compari-
sons of IG and NIG. Following this, we turn to pre-
dictors of developmental characteristics in the
Institution vs. community
For purposes of gauging both areas and degree of
developmental impairment among children experi-
Table 2 Mean scores (SD) for measures of cognitive develop-
ment, growth, competence, negative behavior, caregiving
environment, and emotional expression by group
Cognitive development (n ¼ 124) (n ¼ 66)
Mental Development
66.16 (15.43) 102.87*** (12.98)
74.26 (13.26) 103.43*** (10.88)
Growth (n ¼ 123) (n ¼ 62)
Height (z) ).89 (.90) .06*** (.98)
Weight (z) )1.25 (1.07) ).06** (1.02)
Head circumference (z) ).77 (.97) .17*** (.79)
Weight for height (z) ).79 (1.03) .002*** (.99)
Size ).93 (.77) .044*** (.89)
Competence and
negative behavior
(n ¼ 101) (n ¼ 52)
Competence )1.71 (2.45) .00*** (1.00)
Negative behavior .63 (1.18) .00** (1.00)
Caregiving environment (n ¼ 122) (n ¼ 64)
Caregiving quality 2.18 (.60) 2.80*** (.54)
Caregiver available 24.80 (6.87) 29.34*** (1.81)
Caregiver interacting 15.79 (8.53) 24.53*** (5.50)
Child unoccupied/
11.58 (6.13) 6.52*** (4.87)
Emotional expression (n ¼ 125) (n ¼ 66)
Positive ).46 (.65) ).09** (.77)
Negative ).13 (.44) ).28* (.24)
Note: ***p < .001, **p < .01, *p < .05.
Caregiving and development in institutions and families
2007 The Authors
Journal compilation 2007 Association for Child and Adolescent Mental Health.
Page 4
encing institutional care, we compared children who
had been raised in institutions to those raised in
Control analyses. We examined the distribution of
gender, ethnicity, age, and birthweight for the IG and
NIG (see Table 1) as well as the percent of life insti-
tutionalized for the IG. The ethnic distribution of the
two groups differed: v
(3) ¼ 26.25, p < .001. Chil-
dren in the NIG were primarily ethnic Romanians. A
substantial minority of the IG was of Roma ethnicity.
Birthweight data were missing for 13 of the children
in the institution group and 3 children in the com-
munity group. We could not discern any pattern for
the missing data regarding birthweight, and we
considered them randomly distributed. On average,
birthweight was lower for children in the institution
than for those raised in families, t(168) ¼ 5.51,
p < .001. In the initial analyses, rearing environ-
ments (Institution vs. Never Institutionalized) were
Caregiving environment. As expected, Caregiving
Quality scores were higher for children in the NIG
than for those in the IG (Table 2). Also, the number
of episodes during which a caregiver was available
to the child, and during which a caregiver inter-
acted with the child, was markedly lower in the IG
(Table 2), and children were observed to be unoc-
cupied or watching, rather than participating,
during more episodes in the children being reared
in institutions than in the never institutionalized
Physical growth. Children reared in the institu-
tional setting had poorer growth when compared to
their community age mates. When birthweight was
entered as a covariate, findings were similar, with
the exception of weight for height which was no
longer significantly different.
Cognitive development. The Mental Development
Index (MDI) scores for children being raised in
institutions were markedly below those of never
institutionalized children, as were Developmental
Quotient scores (see Table 2). These data also were
examined by displaying the obtained scaled scores
for each group by age (see Figure 1). As expected,
MDI scaled scores for the NIG were consistent across
the age range (M ¼ 103) at baseline and chronologi-
cal age was not related to obtained scaled score, r ¼
).09. For the IG, the MDI scaled scores declined as
chronological age increased, r ¼ ).56, p < .001.
There were no children in the NIG with MDI scaled
scores less than 50, whereas there were 32 such
scores in the IG.
Emotional expression. Children from the IG dis-
played positive affect less frequently, and negative
affect more frequently, than community children.
Problem behaviors and competence. Children
reared in the institutional environment were repor-
ted to have more maladaptive and more atypical
behaviors than their community counterparts. On
the other hand, there were no overall differences
between groups on the externalizing, internalizing,
and dysregulation scales. Children raised in the
institution were rated lower on the competence scale,
and lower on the social relatedness scale than com-
munity children.
Summary. As expected, young children raised in
institutions showed significantly more developmen-
tal difficulties across most domains when compared
to young children raised in their families.
Predictors of developmental status in
institutionalized children
Having established a range of differences between
children raised in institutions and children raised in
families, our next step was to examine the specific
factors that influence the development of children in
institutions. Our goal was to determine the contri-
bution of caregiving quality, above and beyond other
factors. Therefore, we selected two sets of variables
as predictors of developmental characteristics of
children raised in institutions. In the first set, we
chose child characteristics such as age, gender, and
ethnicity [ethnic Romanian ¼ 0 vs. Roma ethni-
city ¼ 1]. In addition, we included birthweight in the
analysis for physical size. The second set consisted
of caregiving quality, percent of life institutionalized,
and BSID-II MDI score 50 vs. <50 [coded 0 and 1,
respectively]. Using ordinary least squares regres-
sion (OLS) with these predictor variables, we were
able to account for significant variance in the fol-
lowing developmental variables: (1) cognitive abilities
(Developmental Quotient), F(6,115) ¼ 25.16,
p < .001, R
¼ .545; (2) physical size, F(7,100) ¼
ical A
e (months)
Mental Development Index
Figure 1 Mental development index for family-reared
and institution-reared children by age
214 Anna T. Smyke et al.
2007 The Authors
Journal compilation 2007 Association for Child and Adolescent Mental Health.
Page 5
4.451, p < .001, R
¼ .184; (3) competence,
F(6,92) ¼ 6.326, p < .001, R
¼ .246; and (4) negat-
ive behavior, F(6,92) ¼ 4.613, p < .001, R
¼ .181
(see Table 3). The regression equations for positive
and negative emotional expression were not signi-
With regard to cognitive abilities, we found that
observed caregiving quality was positively related to
DQ, while percent of life institutionalized and
having had a Bayley MDI score <50 were both ne-
gatively associated with DQ. As caregiving quality
increased and percent of the child’s life spent in the
institution decreased, DQ was higher, indicating
that caregiving quality was an important con-
tributor to DQ even with other contributors
accounted for.
We also examined physical size and found that it
was associated (positively) only with birthweight.
None of the other variables contained in the regres-
sion equation were associated with birthweight,
either negatively or positively. Because substantial
numbers of data points for the birthweight variable
were missing for the children being raised in insti-
tutions, we chose to include birthweight in the con-
trol block only for the size variable.
Next, we examined caregiver reported competence
for children receiving institutional care. We found
that both age and observed caregiving quality were
positively related to caregiver reported competence,
while obtaining an MDI score of <50 was associated
with lower levels of competence. Again, caregiving
quality made significant contributions to children’s
competence, over and above what was accounted for
by other factors.
Finally, we examined caregiver report of children’s
negative behavior. We found that negative behavior
was associated with observed poorer caregiving
quality as well as having obtained a Bayley MDI
score <50 (see Table 3). Again, caregiving quality was
related to negative behavior even taking into account
contributions from other factors.
This study represents the most ambitious attempt to
date to explore the effect of individual differences in
the caregiving environment and developmental status
in young children being raised in institutions. The
most important finding was that, even after control-
ling for a number of child characteristics and percent
of life raised in an institution, observed caregiving
quality was associated with cognitive development
and competence in young children. This was a par-
ticularly rigorous test of the importance of caregiving
quality because it was assessed in the context of sev-
eral other potential risk factors. Nevertheless, be-
cause of the cross-sectional nature of the data, it is not
possible to determine whether more compromised
children elicited poorer quality care or whether poor
caregiving led to more developmental delay and less
competence. Both directions may plausibly occur over
time: children who receive better quality caregiving
may use their interactions with their caregivers to gain
a more complex knowledge of the environment and
their place within it, and children who explore actively
and initiate interactions with their caregivers may
elicit caregiving which more readily promotes
Caregiving quality was related to three of six
developmental outcomes whereas percentage of time
institutionalized was related only to one of six
developmental outcomes, demonstrating that the
mere fact of institutionalization is less powerful than
the microcaregiving environment within which each
child develops. Even within problematic institutional
environments, individual differences in caregiving
appear to be importantly related to young children’s
Not surprisingly, older children showed a higher
level of cognitive impairment than younger children
(Figure 1). This is compatible with the idea that
cognitive ability deteriorates as institutional care
continues. Length of institutional care also was lin-
Table 3 Regression equations showing influence of caregiving quality, percent of child’s life institutionalized, and Bayley MDI < 50
on Developmental Quotient, size, competence, and negative behavior after controlling for demographic variables institution group
DQ (n ¼ 122) Size (n ¼ 108) Competence (n ¼ 99)
Negative behavior
(n ¼ 99)
b(SE b) DR
b(SE b) DR
b(SE b) DR
b(SE b) R
Control block .144*** .213*** .178*** .102*
Age ).245 (.13) .021 (.01) .234*** (0.05) .042 (.02)
Gender 1.931 (1.66) ).081 (.14) .403 (.43) ).146 (.22)
Ethnicity ).272 (1.81) ).128 (.15) ).22 (.47) ).313 (.23)
Birthweight .567*** (.12)
Predictors .424*** .024 .114** .129**
Caregiving quality 3.487* (1.49) ).153 (.13) .863* (.40) ).428* (.20)
Percent life in institution ).089* (.04) ).003 (.004) ).005 (.01) .001 (.006)
Bayley score <50 )18.907*** (2.10) ).239 (.17) )1.222* (.51) .675** (.26)
Total R
.54*** .18*** .25*** .18***
Note: ***p < .001, **p < .01, *p < .05.
Caregiving and development in institutions and families
2007 The Authors
Journal compilation 2007 Association for Child and Adolescent Mental Health.
Page 6
early related to cognitive compromise in studies of
children adopted out of Romanian institutions
(O’Connor et al., 2000).
The second set of findings, concerning compar-
isons between Romanian never institutionalized
children and institutionalized children, were largely
unsurprising. Nevertheless, the magnitude of the
differences between these two groups across mul-
tiple developmental domains confirmed previous
findings regarding the association between institu-
tionalization and serious developmental compro-
mise, with large differences in growth, cognitive level,
competence, and emotional expression all favoring
never institutionalized children. Though Romanian
institutions have received a good deal of negative
international attention for their poor quality, the
pattern of findings in this study is similar to what
has been reported in other contemporary studies
that have compared children residing in institutions
and home-reared children in Greece (Vorria et al.,
2003), the United Kingdom (Tizard & Rees, 1974),
and the United States (Harden, 2002).
One surprising finding was that internalizing and
externalizing behavior problems were no more com-
mon among children living in institutional care than
among children reared in their families. Studies of
internationally adopted post-institutionalized chil-
dren have reported mixed findings on the question of
whether externalizing problems are increased. Rut-
ter and colleagues (2001) noted that Romanian
children adopted into the UK had no greater in-
cidence of behavior problems, including conduct and
emotional problems and difficulties in peer interac-
tion, than children adopted within the UK. Higher
total CBCL scores, although scores were not within
the clinical range, were obtained for a group of Ro-
manian children adopted into Canada, when com-
pared to a sample of Canadian children who were not
adopted (Marcovitch et al., 1997). In contrast, Ellis
and colleagues (2003) reported that 23.5% of their
sample of 2–6-year-olds residing in Romanian insti-
tutions met the screening cut off for oppositional
defiant disorder according to caregiver-completed
questionnaires. In a sample of 9-year-old children
residing in institutions, behavioral difficulties were
noted for boys (emotional difficulties, conduct prob-
lems) and for girls (emotional disturbance) (Vorria
et al., 1998). Abnormal behaviors, such as stereo-
typies, were noted in over 50% of a small sample of
children adopted into Manitoba, Canada (Benoit,
Joycelyn, Moddemann, & Embree, 1996). Similar
abnormalities of behavior were noted by Beckett and
her colleagues (2002) and are consistent with our
findings that atypical behaviors are more common
among infants and toddlers reared in institutions
than among family-reared children.
Age differences may account for some of the dis-
crepancy in findings. Harden (2002) noted that a
group of infants and toddlers in congregate care in
the US did not differ on measures of behavioral dif-
ficulties but did differ on measures of child compet-
ence, findings more similar to those obtained in the
current study. In other words, internalizing and
externalizing problems may emerge as a later con-
sequence of early institutional rearing.
Competence, despite its importance as a con-
struct, has not been examined previously in children
being raised in institutions. The substantial differ-
ences in competence demonstrated in this study
suggest that the more limited findings with regard to
behavior problems do not derive solely from care-
giver minimization or bias, since behavior problems
and competence were rated by caregivers. An
important question for future exploration is the de-
gree to which competence is impaired, because it
may be a harbinger of later emerging behavioral
problems in children formerly experiencing institu-
tional care.
There are limitations to this report that should be
noted. The results are cross-sectional in nature, and
thus, the direction of effects of caregiving environ-
ment and developmental characteristics cannot be
determined. As we follow these children’s develop-
ment longitudinally, the question of direction of ef-
fects may be more readily determined. Furthermore,
the fact that the individual characteristics that we
assessed were not associated with differences in
developmental status does not mean that other child
characteristics might not be more important.
There is also a question about how much we can
generalize to institutional settings and the effects of
institutional rearing in the rest of Europe, Asia,
Africa, and the Americas from these results, given
the possibility of wide variability in such important
factors as nutrition, instrumental care, and cognitive
and emotional stimulation. Importantly, these find-
ings in the main substantiate the findings from the
previous 50 years about the developmental charac-
teristics of children raised in institutions compared
to children raised in families, albeit in a larger and
more comprehensive examination of these charac-
teristics than has been reported previously (Nelson,
2000; Zeanah et al., 2006).
Having an accurate understanding of the early
rearing environment of institutionally reared young
children can assist adoptive parents and individuals
who provide treatment for internationally adopted
children to do so in a knowledgeable and focused
way. Additionally, countries that understand the ef-
fects of institutional care may be drawn to implement
systems of foster care that permit young abandoned
children to live with families, thus addressing
the children’s need for sensitive, individualized
care and for sufficient cognitive stimulation.
The current study provides a unique perspective
for understanding the institutional ecology and the
effects of institutionalization on the development of
young children. Although children reared in insti-
tutions were markedly different from home-reared
children in a number of developmental spheres,
216 Anna T. Smyke et al.
2007 The Authors
Journal compilation 2007 Association for Child and Adolescent Mental Health.
Page 7
there is evidence that children were not affected
equally by the institutional experience. Further
exploration of individual differences is one important
direction for future research. Analyses of longitud-
inal data will allow us to explore risk and protective
factors that may influence children’s reactions to the
institutional environment as well as their ability to
recover from the effects of institutional care via the
intervention of foster care. Another important direc-
tion for research concerns the timing of remediation
efforts. Though some preliminary efforts have been
made in this direction in studies of children adopted
out of institutions (Chisholm, 1998; Chisholm, Car-
ter, Ames, & Morison, 1995; O’Connor et al., 2000;
Rutter et al., 1998), contributions from these studies
on this question are limited by lack of baseline
measures and non-randomized samples of children.
Follow-up of the current sample may enable us to
examine hypotheses about critical and sensitive
periods within specific developmental domains. Re-
sults from such analyses may clarify important
questions regarding both the effects of deprivation
and the degree of recovery that is possible.
Author note
Additional members of the BEIP Core Group: Susan
W. Parker (Randolph Macon College, USA) and Her-
mi R. Woodward (University of Pittsburgh/MacArt-
hur Research Networks, USA).
The Bucharest Early Intervention Project (BEIP) was
funded by the John D. and Catherine T. MacArthur
Foundation Research Network on Early Experience
and Brain Development (Charles A. Nelson, Network
Chair). The authors wish to acknowledge the many
invaluable contributions of their Romanian partner
institutions, the SERA Romania Foundation, the
Institute for Maternal and Child Health (IOMC), and
the Department for Social Welfare (DGAS), Sector 1,
Bucharest. They are also deeply grateful to their Ro-
manian team whose hard work and dedication have
made this study possible. Thanks to Donald Guthrie
and Scott Keith for assistance with data analysis.
Correspondence to
Anna T. Smyke, Tulane University Health Sciences
Center, Department of Psychiatry/Neurology, 1440
Canal Street TB-52, New Orleans, LA 70112, USA;
Bayley, N. (1993). Bayley Scales of Infant Development
(2nd edn). New York: Psychological Corporation.
Beckett, C., Bredenkamp, D., Castle, J., Groothues, C.,
O’Connor, T.G., Rutter, M., et al. (2002). Behavior
patterns associated with institutional deprivation: A
study of children adopted out of Romania. Develop-
mental and Behavioral Pediatrics, 23, 297–303.
Benoit, T., Joycelyn, L.J., Moddemann, D.M., & Em-
bree, J.E. (1996). Romanian adoption: The Manitoba
experience. Archives of Pediatric and Adolescent
Medicine, 150, 1278–1282.
Carlson, M., & Earls, F. (1997). Psychological and
neuroendocrinological sequelae of early social depri-
vation in institutionalized children in Romania.
Annals of the New York Academy of Sciences, 807,
Carter, A.S., Briggs-Gowan, M., Jones, S.M., & Little,
T.D. (2003). The Infant–Toddler Social and Emotional
Assessment (ITSEA): Factor structure, reliability, and
validity. Journal of Abnormal Child Psychology, 31,
Castle, J., Groothues, C., Bredenkamp, D., Beckett, C.,
O’Connor, T.G., & Rutter, M. (1999). Effects of
qualities of early institutional care on cognitive
attainment. American Journal of Orthopsychiatry,
69, 424–437.
Centers for Disease Control (CDC). (2002). 2000 CDC
Growth Charts.
Chisholm, K. (1998). A three year follow-up of attach-
ment and indiscriminate friendliness in children
adopted from Romanian orphanages. Child Develop-
ment, 69, 1092–1106.
Chisholm, K., Carter, M.., Ames, E., & Morison, S.
(1995). Attachment security and indiscriminately
friendly behaviour in children adopted from Roma-
nian orphanages. Development and Psychopathology,
7, 283–294.
Ellis, B.H., Fisher, P.A., & Zaharie, S. (2004). Predictors
of disruptive behavior, developmental delays, anxiety,
and affective symptomatology among institutionally
reared Romanian children. Journal of the American
Academy of Child and Adolescent Psychiatry, 43,
Fisher, L., Ames, E.W., Chisholm, K., & Savoie, L.
(1997). Problems reported by parents of Romanian
orphans adopted to British Columbia. International
Journal of Behavioral Development, 20, 67–82.
Goldsmith, H.H., & Rothbart, M.K. (1999). Laboratory
Temperament Assessment Battery, Prelocomotor Ver-
sion 3.1. Unpublished manuscript, Madison: Univer-
sity of Wisconsin.
Gunnar, M.R. (2000). Early adversity and the develop-
ment of stress reactivity and regulation. In C.A.
Nelson (Ed.), The effects of early adversity on neuro-
behavioral development (vol. 31, pp. 163–200). Mah-
wah, NJ: Lawrence Erlbaum.
Harden, B.J. (2002). Congregate care for infants and
toddlers: Shedding new light on an old question.
Infant Mental Health Journal, 23, 476–495.
Johnson, D.E. (2000). Medical and developmental
sequelae of early childhood institutionalization in
Eastern European adoptees. In C.A. Nelson (Ed.),
The effects of early adversity on neurobehavioral
development (vol. 31, pp. 113–162). Mahwah, NJ:
Lawrence Erlbaum.
Kaler, S., & Freeman, B. (1994). Analysis of environ-
mental deprivation: Cognitive and social development
Caregiving and development in institutions and families 217
2007 The Authors
Journal compilation 2007 Association for Child and Adolescent Mental Health.
Page 8
in Romanian orphans. Journal of Child Psychology
and Psychiatry, 35, 769–781.
Kochanska, G., Coy, K., Tjebkes, T., & Husarek, S.
(1998). Individual differences in emotionality in
infancy. Child Development, 64, 375–390.
Kreppner, J.M., O’Connor, T.G., & Rutter, M. (2001).
Can inattention/hyperactivity be an institutional
deprivation syndrome? Journal of Abnormal Child
Psychology, 29, 513–528.
Lindsey, J.C., & Brouwers, P. (1999). Intrapolation and
extrapolation of age-equivalent scores for the Bayley
II: A comparison of two methods of estimation.
Clinical Neuropharmacology, 22, 44–53.
MacLean, K. (2003). The impact of institutionalization
on child development. Development and Psycho-
pathology, 15, 853–884.
Marcovitch, S., Goldberg, S., Gold, A., Washington, J.,
Wasson, C., Krekewich, K., et al. (1997). Determin-
ants of behavioral problems in Romanian children
adopted in Ontario. International Journal of Behav-
ioral Development, 20, 17–31.
Nelson, C.A., & Bosquet, M. (2000). Neurobiology of
fetal and infant development: Implications for infant
mental health. In C.H. Zeanah (Ed.), Handbook of
infant mental health (2nd edn, pp. 37–59). New York:
Nelson, R.I. (2000). Developmental status of adopted
Chinese children. Holt International Families Maga-
zine, November/December.
NICHD Early Child Care Research Network. (1996).
Characteristics of infant child care: Factors contribu-
ting to positive caregiving. Early Childhood Research
Quarterly, 11, 269–306.
O’Connor, T., Rutter, M., & the English Romanian
Adoption Studies Team. (2000). Attachment disorder
behavior following early severe deprivation: Extension
and longitudinal follow-up. Journal of the American
Academy of Child and Adolescent Psychiatry, 39,
Roy, P., Rutter, M., & Pickles, A. (2000). Institutional
care: Risk from family background or pattern of
rearing. Journal of Child Psychology and Psychiatry,
41, 139–149.
Rutter, M., & the English Romanian Adoptions Study
Team. (1998). Developmental catch-up, and deficit,
following adoption after severe global early privation.
Journal of Child Psychology and Psychiatry, 39, 465–
Rutter, M., Andersen-Wood, L., Beckett, C., Bredenk-
amp, D., Castle, J., Groothues, C., et al. (1999).
Quasi-autistic patterns following severe early global
privation. Journal of Child Psychology and Psychiatry,
40, 537–549.
Rutter, M., Kreppner, J.M., O’Connor, T., & the English
Romanian Adoption Studies Team. (2001). Specificity
and heterogeneity in children’s responses to profound
institutional privation. British Journal of Psychiatry,
179, 97–103.
Smyke, A.T., Dumitrescu, A., & Zeanah, C.H. (2002).
Attachment disturbances in young children. I: The
continuum of caretaking casualty. Journal of the
American Academy of Child and Adolescent Psychi-
atry, 41, 972–982.
Tizard, B., & Rees, J. (1974). A comparison of the effects
of adoption, restoration to the natural mother, and
continued institutionalization on the cognitive devel-
opment of four-year-old children. Child Development,
45, 92–99.
Vorria, P., Papaligoura, Z., Dunn, J., van IJzendoorn,
M.H., Steele, H., Kontopoulou, A., et al. (2003). Early
experiences and attachment relationships of Greek
infants raised in residential group care. Journal of
Child Psychology and Psychiatry, 44, 1208–1220.
Vorria, P., Rutter, M., Pickles, A., Wolkind, S., &
Hobsbaum, A. (1998). A comparative study of Greek
children in long-term residential group care and in
two-parent families: I. Social, emotional, and beha-
vioral differences. Journal of Child Psychology and
Psychiatry, 39, 225–236.
Zeanah, C.H., Nelson, C.A., Fox, N.A., Smyke, A.T.,
Marshall, P., Parker, S.W., & Koga, S.F.M. (2003).
Designing research to study the effects of institutio-
nalization on brain and behavioral development: The
Bucharest Early Intervention Project. Development
and Psychopathology, 15, 885–907.
Zeanah, C.H., Smyke, A.T., Koga, S.F.M., Carlson, E., &
the BEIP Core Group. (2005). Attachment in institu-
tionalized and non-institutionalized Romanian chil-
dren. Child Development, 76, 1015–1028.
Zeanah, C.H., Smyke, A.T., & Settles, L. (2006).
Orphanages as a developmental context for early
childhood. In K. McCartney & D. Phillips (Eds.),
Handbook of early childhood development (pp. 424–
454). New York: Blackwell.
Manuscript accepted 24 July 2006
218 Anna T. Smyke et al.
2007 The Authors
Journal compilation 2007 Association for Child and Adolescent Mental Health.
Page 9
  • Source
    • "In particular, environmental inputs that are absent or atypical among children raised in institutional settings, as well as among children raised in poverty, are promising candidates. Institutional rearing is associated with an absence of environmental inputs of numerous kinds, including the presence of an attachment figure, variation in daily routines and activities, access to age-appropriate enriching cognitive stimulation from books, toys, and interactions with adults, and complex language exposure (Smyke et al., 2007; Zeanah et al., 2003). Some of these dimensions of environmental experience have also been shown to be deprived among children raised in poverty, including access to cognitively enriching activities—including access to books, toys, and puzzles; learning opportunities outside the home (e.g., museums) and within the context of the parent–child relationship (e.g., parental encouragement of learning colors, words, and numbers, reading to the child); and variation in environmental complexity and stimulation— as well as the amount and complexity of language input (Bradley, Convyn, Burchinal, McAdoo, & Coll, 2001; Dubowitz et al., 2002; Garrett, Ng'andu, & Ferron, 1994; Hart & Risley, 1995; Hoff, 2003; Linver, Brooks-Gunn, & Kohen, 2002). "
    [Show abstract] [Hide abstract] ABSTRACT: Despite long-standing interest in the influence of adverse early experiences on mental health, systematic scientific inquiry into childhood adversity and developmental outcomes has emerged only recently. Existing research has amply demonstrated that exposure to childhood adversity is associated with elevated risk for multiple forms of youth psychopathology. In contrast, knowledge of developmental mechanisms linking childhood adversity to the onset of psychopathology—and whether those mechanisms are general or specific to particular kinds of adversity—remains cursory. Greater understanding of these pathways and identification of protective factors that buffer children from developmental disruptions following exposure to adversity is essential to guide the development of interventions to prevent the onset of psychopathology following adverse childhood experiences. This article provides recommendations for future research in this area. In particular, use of a consistent definition of childhood adversity, integration of studies of typical development with those focused on childhood adversity, and identification of distinct dimensions of environmental experience that differentially influence development are required to uncover mechanisms that explain how childhood adversity is associated with numerous psychopathology outcomes (i.e., multifinality) and identify moderators that shape divergent trajectories following adverse childhood experiences. A transdiagnostic model that highlights disruptions in emotional processing and poor executive functioning as key mechanisms linking childhood adversity with multiple forms of psychopathology is presented as a starting point in this endeavour. Distinguishing between general and specific mechanisms linking childhood adversity with psychopathology is needed to generate empirically informed interventions to prevent the long-term consequences of adverse early environments on children’s development.
    Full-text · Article · Feb 2016 · Journal of Clinical Child & Adolescent Psychology
  • Source
    • "Children in foster care also appeared to develop fewer attachment issues (Smyke et al., 2012), have better psychological adjustment (Nowacki & Schoelmerich, 2010), and accrue fewer criminal convictions in adulthood, even after adjusting for number and duration of placements (Dregan & Gulliford, 2012; Johnson, Browne, & HamilitonGiachristis, 2006). However, a sizeable proportion of evidence that illustrates better outcomes for foster care focuses on infants and younger children (Smyke, Dumitrescu, & Zeanah, 2002; Smyke et al., 2007; van IJzendoorn, Luijk, & Juffer, 2008; Zeanah, Smyke, Koga, & Carlson, 2005). On the other hand, several studies that included older children above 4 years of age demonstrated good outcomes in institutional care (Knorth, Harder, Zandberg, & Kendrick, 2008; McKenzie, 1997 McKenzie, , 1999 van IJzendoorn et al., 2008; Wolff & Fesseha, 2005) in the presence of positive caregiving and holistic organizational structures. "
    [Show abstract] [Hide abstract] ABSTRACT: The negative impact of childhood maltreatment, which can often extend well into adulthood, consistently appears to be ameliorated if victimized children possess several resiliencies or strengths. However, little is known about how vulnerable children's outcomes are affected by different levels of strengths across different out-of-home placement settings. Hence, this study examined the association of two factors — children's strengths and placement type, with outcomes at two time-points during out-of-home care. The Child and Adolescent Needs and Strengths (CANS) tool was used to assess the outcomes of 285 children placed in residential homes or foster care in Singapore. Multiple regressions were conducted on CANS domain scores to evaluate whether level of baseline strengths and placement type predicted outcomes at two time-points after controlling for race, prior placements, age, gender, interpersonal trauma, and baseline needs scores. Results indicate that relative to residential care, foster care children are reported to be younger, with lower baseline needs, more prior placements, fewer baseline strengths and suffered fewer types of interpersonal trauma. After controlling for covariates, higher baseline strengths significantly predicted lower baseline needs of children across 3 of 4 CANS domains, regardless of placement settings. However, at reassessment 1 year later, there were significant interactions between strengths and placement type, whereby baseline strengths significantly predicted lower life functioning needs only in foster care. To conclude, in both residential and foster care, the protective effects of high strengths against child maltreatment were similarly apparent at baseline, despite clear differences in children's profiles across placement types. Over time, these initial benefits appeared to persist somewhat for children in foster care but seemed to diminish in more restrictive, residential settings and this warrants further investigation on children with more similar profiles. Nonetheless, it is clear that the continual development of children's strengths should be prioritized in case planning.
    Full-text · Article · Nov 2015 · Children and Youth Services Review
  • Source
    • "При этом период нахождения в доме ребенка, при котором наблюдается негативная динамика развития, составляет по данным нашего исследования 7–9 месяцев. В научной литературе по изучению детей, усыновленных из домов ребенка, известно, что превышение шестимесячного периода пребывания в учреждении в тяжелых депривационных условиях влияет на последующее развитие детей1314151617181920. Поскольку, по данным исследований [1, 22], изменения социального окружения в доме ребенка приводят к значительному улучшению развития детей уже при 3-х месячном пребы-вании в новых условиях, высока вероятность, что изменения развития детей в доме ребенка могут быть обнаружены и при меньшем периоде пребывания (до 6 месяцев). "
    [Show description] [Hide description] DESCRIPTION: Development of Infants in Association with Timing of Institutionalization This longitudinal study analyzes infant’s development according to the institutional life timing. Two groups of young orphan children (preterm children, n=14, gestational age=34,9; and full-term children, n=21, gestational age=39,1) were assessed at age 3 and 9 months by the Battelle Development Inventory (LINC Associates, 1988). The comparative analysis of the Total Score of general development between groups at 3 and 9 months and the analysis of the Total Score with increasing of the institutional timing from 1-3to 7-9 months were conducted. Preterm children had significantly lower developmental quotients (DQ=81,07) in comparison with their full-term peers (DQ=88,76) at age 3 month but not at age 9 month. 7-9 months of the institutional timing has a negative impact on the general mental development of children in both groups.
    Full-text · Research · Apr 2015
Show more