1
The importance of early experience for brain and
behavioral development Implications for
childhood policy and practice

• Nathan A. Fox
• University of Maryland
• Key Note Address, Seeds of Change
• ISSA 10th annual conference

2
Critical Questions for the study of Human
Development

• The effects of early experience on development
• The influence of individual differences on
  psychological development
• Translation of research to practice Bench to
  Bedside

3
Primary questions

• What are the effects of early experiences on
  different domains of behavior?
• Are there sensitive or critical periods during
  which the effects of experience have their
  greatest impact?
• What are the mechanisms or processes by which
  experience and environment interact?

4
Central tenants of developmental psychology

• Early experiences in the first months of life
  have a significant impact upon later
  psychological functioning
• There are critical or sensitive periods during
  which certain types of stimulation must be
  received for the normal development of certain
  domains of functioning

5
What is the evidence for the first tenant? (Early
experience matters)

• Most of the empirical research is with non-human
  primates and rodents
• It includes the work on early handling and stress
  reactivity (Dennenberg, Levine, Plotsky and
  Meaney)
• And the work on early maternal separation
  (Harlow, Sackett)

6
Infant monkey with cloth surrogate mother
7
What is the evidence for the second tenant?
(Critical periods in development)

• Again, much of this work has been completed in
  non-human primates and rodents
• It includes the work on imprinting (e.g. Lorenz)
• And the work on early sensory stimulation (e.g.
  binocular deprivationHubel and Wiesel)

8
Lorenz and imprinted ducklings walking after him
9
Experience Shapes Brain Architecture by
Over-Production Followed by Pruning (700 synapses
formed per second in the early years)
birth 6 years
14 years
10
Neural Circuits are Wired in a Bottom-Up
Sequence
Language
Higher Cognitive Function
Sensory Pathways (Vision, Hearing)
FIRST YEAR
-8 -7 -6 -5 -4 -3 -2 -1 1 2 3 4 5 6 7 8 9 10 11
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Birth
(Months)
(Years)
Source C.A. Nelson (2000)
11
General Conclusions about Plasticity

• Collectively, in most cases sensory/perceptual
  development proceed normally if such systems are
  set correctly during a sensitive period of
  developmentHubel and Weisel
• Not clear what aspects of cognitive or social and
  emotional development require experience at
  particular (e.g., sensitive) points in time.
  Inferences drawn from intervention studies
  suggest some advantage to early experience.

12
Bench to Bedside

• Children in institutions represent a natural
  experiment---one in which the effects of early
  experience can be examined
• Children in institutions represent a world wide
  problem

13
Proportion of all children under 3 years who are
in institutional care, 2003 (blue lines are
estimates).
14
Placement of children under 3 taken into care,
2002-2003 Browne et al. (2006)
15
Reason for institutionalisation, 2003
16
Average cost per placement per year for a child
under 3 years (in Euros, 2002/2003).
17
Children adopted from institutions demonstrate a
number of emotional and behavioral problems, such
as

• disturbances and delays in social/emotional
  development (particularly attachment).
• aggressive behavior problems.
• inattention/hyperactivity.
• syndrome that mimics autism.

18

• The Bucharest Early Intervention Project seeks
  to
• Examine the effects of institutionalization on
  brain and behavioral development of young
  children
• Determine if these effects can be remediated
  through intervention, in this case foster care
• Improve the welfare of children in Romania by
  establishing foster care as an alternative to
  institutionalization

19
Project Background
20
CEAUSESCU LEFT BEHIND A SOCIETY UNABLE AND
UNWILLING TO TAKE CARE OF ITS CHILDREN

• Communist Policy1966 decree
• Raise production by increasing population
• belief that greater population greater power
• Establishment of the MENSTRUAL POLICE - state
  gynecologists who conducted monthly checks of
  women of childbearing age who had not borne at
  least 5 children
• Establishment of CELIBACY TAX - families received
  a stipend for having more than 2 children were
  levied tax for having fewer than 5 children
• OUTLAWED all contraception and abortion

21
THE RESULTS OF CEAUSESCUS 1966 POLICY

• Child abandonment became a national disaster, as
  families could not afford to keep their children,
  and were encouraged to turn them over to the
  state. This destroyed the family unit and led to
  thousands of children being raised in
  institutions.

22
1989 The fall of the Ceausescu regimeThe
aftermath.
100,000 children warehoused in state
institutions

• Poverty 1 reason for child abandonment
• International media brought the plight of these
  children to the attention of the world
• Large numbers of children adopted
  internationally, often by Western families
  unprepared for challenges that lay ahead
• And then, Romania banned international adoption

23
Immigrant visas issued to orphans going to the
USA from European accession countries,1989 to
2002.
24
The Challenge

• In a given week, children come into contact with
  a large number of professionals and
  paraprofessionals

17 caregivers, working rotating 8 hr. shifts 3
housekeepers 4 nurses 2 educators 1
psychologist 1 physical therapist 1 physician
Although children may become familiar with
caregivers, the opportunity to form
relationships with them is limited
25
The Study
26
Investigative Team

• Principle Investigators
• Charles Zeanah, MD, Tulane University
• Nathan Fox, Ph.D., University of Maryland
• Charles A. Nelson, Ph.D., Harvard Medical
  School/Childrens Hospital Boston
• Co-Investigators
• Anna Smyke, Ph.D., Tulane University
• Dana Johnson, MD, Ph.D., Jennifer Windsor, Ph.D.,
  University of Minnesota
• Peter Marshall, Ph.D., Temple University
• Helen Link-Egger, MD, Duke University
• Stacy Drury, MD, Ph.D., Tulane University

27
Study Design

• First ever randomized trial of foster care as
  intervention for social deprivation associated
  with institutionalization
• 136 institutionalized children between 6 and 31
  months initially assessed at baseline (Mean
  Age20 months)
• 68 randomly assigned to remain in institution
  (IG) 68 randomly assigned to foster care (FCG)
• 72 never-institutionalized children (NIG) matched
  on age and gender serve as controls
• Following baseline assessment, children assessed
  comprehensively at 9, 18, 30, 42 and 54 months

28
Domains of Assessment

• Physical Development
• Language
• Cognition
• Brain Function
• Emotional reactivity
• Caregiving Environment
• Attachment
• Mental Health Problems
• Genetics

Data derived from measures listed in bold will
be discussed in this talk
29
General Hypotheses

• Institutional rearing will have profound effects
  upon childrens cognitive and socio-emotional
  development
• Removing children from the institution and
  placing them in family environments will
  remediate some of these deficits.
• The age or timing of placement into foster care
  will be a significant factor explaining
  intervention effects (thought this may vary by
  domain)

30
The Intervention
31
BEIP Foster Care

• Goal was to have foster care that was
• Effective
• Affordable
• Replicable
• Culturally sensitive
• Informed by latest clinical and research findings

32
The Foster Care Team BucharestSupport Services
to Foster Parents

• Caseload of 18-20 families
• Visits to foster parents every 10 15 days
• Intensive phone contact
• Inquiry regarding child behavior/adjustment
• Foster Parent Support Group
• Education/Support
• Supplies
• Cribs, car seats, diapers, clothing

33
Findings to be Discussed

• Cognitive Development (DQ/IQ)
• Brain Development (EEG power)
• Attachment (Strange Situation)
• Mental Health Problems (Psychiatric Screen)

34
Cognitive Development

• What are the effects of institutionalization on
  IQ/DQ?
• At baseline, Mean DQ
• 64 (Institutionalized Group)
• 103 (Never Institutionalized Group)

35
Bayley Scales of Infant Development (MDI)(at
baseline)
36
Change in Group Assignment Over Time/Subject
Attrition (as of 54 months)
37
IQ Scores of Foster Care and Institutionalized
Groups at Follow-up
38
How does DQ/IQ differ for children in foster
care as a function of age of entry? 42
Months (Bayley) 54 Months (WPPSI) Age
at placement N Mean Std Dev Std Err
N Mean Std Dev Std Err 0-18 months 14
94.4 11.9 3.2 14 84.8
16.0 4.3 18-24 months 16 89.0
11.3 2.8 15 86.7
14.8 3.8 24-30 months 22 80.1 13.3
2.8 22 78.1 19.5
4.2 30 months 9 79.7 17.1
5.7 8 71.5 23.8 8.4

39
IQ Scores of Foster Care and Institutionalized
Groups at Follow-Up Broken Down by age of entry
into Foster Care
40
Change in Group Assignment Over Time/Subject
Attrition (as of 96 months)
Enrollment
Assessed for Eligibility (n187)
Excluded (n51)
Randomized (n136)
Allocation
Allocated to Foster Care (n68)
Allocated to Care as Usual (n68)

• Placement at 96 months (n60)
• - 31 MacArthur Foster Care
• - 7 adopted
• - 8 Government Foster Care
• - 12 Returned to Bio Family
• 2 Social Apartments
• Discontinued Participation (n8)

• Placement at 96 months (n56)
• - 15 Institutional Care
• - 4 adopted
• - 18 Government Foster Care
• - 18 Returned to Bio Family
• 1 Family Placement
• Discontinued Participation (n12)

Participation
Analyzed (n53) Excluded from analysis (n7) -
WISC data not available
Analyzed (n50) Excluded from analysis (n6) -
WISC data not available
Analysis
41
WISC Data Comparing FCG Children Currently Living
in MacArthur FC to IG Children Living in the
Institution Government Foster Care
Note Groups differ at p lt .01, p lt .05, p lt
.10
42
Latent Profile Analysis of IG FCG Three
Classes
IQ Score
---- 24 FCG, 6 IG ---- 12 FCG, 13 IG
---- 17 FCG, 31 IG
Age at Assessment
43
Predictors of Profile Membership

• Typical Profile
• Secure attachment classification (42 months)
• Organized attachment classification (42 months)
• Positive relationship with primary caregiver (42
  months)
• Declining Profile
• Disorganized attachment classification
• For FCG, later age at placement in foster care
• Very Low Profile
• Low birthweight
• For FCG, later age at placement in foster care
• Insecure attachment classification
• Disorganized attachment classification
• Less positive relationship with primary caregiver

44
Summary of Cognitive Development

• Institutionalization has a very detrimental
  effect on cognitive function
• Foster care appears to be effective in improving
  cognitive function for those children placed
  before age 2
• Duration of time in foster care does not
  influence timing effects.
• Effects continue through to age 8

45
Brain DevelopmentElectroencephalogram (EEG)

• The EEG reflects the electrical activity
  generated by the entire brain, and provides a
  general measure of brain development
• The EEG is recorded by placing sensors on the
  head, which detect the electrical activity
  generated by the brain.

46
Procedures

• - 3 minutes of EEG data were collected during
  spinning of a bingo wheel.

LEFT
RIGHT
Gnd
F4
F3
Fz
T8
T7
C4
Cz
C3
M1
M2
Pz
P4
P3
O1
O2
47
Alpha Power (6-9 HZ) Across Different Brain
Regions

Ln Power (pW Ohms)


48
Beta RP (10-18 Hz)

Ln Power (pW Ohms)
p lt.001
49
3-5 Hz
6-9 Hz
10-18 Hz
IG
NIG
4.06
50
Summary of Baseline EEG Findings

• Compared to community controls, institutionalized
  children have lower levels of brain activity
• Across different brain regions
• Across different frequency bands

51
EEG Alpha Power for Foster Care Children,
collapsed across 30 and 42 month assessments, for
children entering foster care before or after 24
months of age

Marshall et al 2008, Development and
Psychopathology
52
Scatterplots for Foster Care Children displaying
age of placement and alpha power at each of two
assessment ages
30-Mo (FCG only)
42-Mo (FCG only)
Marshall et al, 2008 Development and
Psychopathology
53
Follow up at age 8

• Participants
• 101 institutionalized children
• 53 (28 male) randomly assigned to a Foster Care
  Intervention.
• 25 (10 male) placed into care before 24mo
• 28 (18 male) placed into care after 24mo
• 48 (25 male) received care as usual.
• 42 (25 male) age and gender matched
  never-institutionalized children.
• Task
• 6 minutes of EEG, alternating 1 minute epochs of
  eyes open and eyes closed.

54
Absolute Alpha Power



µV2
Group F(3,139) 6.467, plt.001 Site x Group
F(12,556)1.975, plt.05
Bonferroni Adjusted plt.05
55
Absolute Beta Power

µV2
Group F(3,139) 3.311, plt.05
Bonferroni Adjusted plt.05
56
Summary of EEG Power Findings

• Theta power decreases and Alpha power increases
  in children placed in foster care relative to
  those who remain in the institution
• These effects are more pronounced for children
  who were placed early in foster care (before they
  were 24 months of age)
• Length of intervention is correlated with
  increases in alpha power in the foster care group
• These effects persist through 8 years of age.

57
Attachment
58
Strange Situation Classifications

• Infancy
• Secure
• Avoidant
• Resistant/Ambivalent
• Disorganized
• Unclassified

• Preschool
• Secure
• Avoidant
• Resistant/Dependent
• Disorganized
• Controlling
• Insecure/Other

59
Strange Situation Classifications

• Infancy
• Secure
• Avoidant
• Resistant/Ambivalent
• Disorganized
• Unclassified

• Preschool
• Secure
• Avoidant
• Resistant/Dependent
• Disorganized
• Controlling
• Insecure/Other

Not organized
60
Baseline Differences11-31 months
61
BEIP SSP Classifications

• Community
• 76.7 secure
• 3.6 avoidant
• 0.0 resistant
• 19.7 disorganized
• 0.0 unclassifiable

• Institution
• 16.8 secure
• 4.7 avoidant
• 0.0 resistant
• 65.4 disorganized
• 13.1 unclassifiable

Zeanah, et al 2005 Child Development
62
Continuum of Attachment

• 5 -- ABCD patterns of attachment
• 4 -- Patterns of attachment with behavioral
  anomalies
• 3 -- Clear preference but passive
• 2 -- Preference discernible
• 1 -- No attachment behaviors evident

63
Degree to Which Attachment Has Formed Bucharest
Early Intervention Project
Zeanah et al 2005 Child Development
64

• Intervention Effects
• at 42 Months

65
Secure vs. Insecure 42 months
Smyke et al (in press)
66
Organized vs. Not 42 months
Smyke et al (in press)
67
Secure vs. Insecure
ns
? 2 10.58 p .001
68
Organized vs. Not Organized
? 2 11.96, p .001
ns
ns
69
Summary of SSP Findings

• Large differences at baseline IG vs. NIG
• Security
• Organization
• Large intervention effects, but incomplete
  recovery
• Timing effects on security and organization
• More children secure if placed before 22-24
  months
• More children organized earlier they were placed

70
Attachment Behavioral Assessment of
Indiscriminate Behavior
71
Mental Health Problems
72
Incidence of Mental Health Problems

• The caregiver was administered the Romanian
  version of the Preschool Age Psychiatric
  Assessment (Romanian PAPA) when the children in
  the BEIP were 54 months old
• The PAPA is a comprehensive structured
  psychiatric interview assessing mental health
  symptoms and disorders in children ages 2 through
  5 years old

73
DSM-IV Disorders

• Emotional disorders
• Depression (major depression, dysthymia,
  depression NOS)
• Anxiety disorder separation anxiety disorder
  (SAD), generalized anxiety disorder (GAD),
  specific phobia, social phobia, selective mutism,
  and post traumatic stress disorder (PTSD)
• Behavior disorders
• Oppositional defiant disorder (ODD)
• Conduct disorder (CD)
• Attention deficit hyperactivity disorder (ADHD)
• Any disorder any emotional or behavioral disorder

74
Prevalence of disorders, overall
IG FCG NIG
Any disorder 54.7 (N29) 41.7 (N25) 17.0 (N10)
Any emotional disorder 45.3 (N24) 21.7 (N13) 13.6 (N8)
Any behavioral disorder 30.2 (N16) 25.0 (N15) 6.8 (N4)
75
Comparison of children ever institutionalized and
community children
IG or FCG NIG OR
Any disorder 47.8 (N60) 17.0 (N10) 4.5 (2.1, 9.7) p0.0001
Any emotional disorder 32.8 (N37) 13.6 (N8) 3.1 (1.3, 7.2) p0.008
Any behavioral disorder 27.7 (N31) 6.8 (N4) 5.2 (1.7, 16) p0.003
76
Comparison of institutionalized and foster care
children
IG FCG OR
Any disorder 54.7 (N29) 41.7 (N25) 1.7 (0.8, 3.6) p0.12
Any emotional disorder 45.3 (N24) 21.7 (N13) 3.0 (1.3, 6.8) p0.009
Any behavioral disorder 30.2 (N16) 25.0 (N15) 1.3 (0.6, 3.0) p0.54
77
Summary of findings on psychiatric disorders

• In general, children raised in institutions
  display a greater incidence of both internalizing
  and externalizing disorders compared to community
  controls
• But, children placed into foster care displayed
  fewer internalizing problems at 54 months
  compared to children in the institutionalized
  group
• And, there was no difference between the foster
  care and institutionalized children in the
  incidence of externalizing problems
• And, age of placement and time in foster
  care/institution did not appear to affect these
  outcomes

78
Timing effects

• IQ (better if placed before 22 months)
• Attachment (better if placed before 24-26 months)
• Brain activity (better if placed before 24 months)

• Psychiatric disorders---timing of placement or
  length of time in institution/foster care did not
  matter

79
In conclusion

• Children exposed early in life to severe social
  deprivation display deficits in IQ, emotional
  reactivity, attachment behaviors, and brain
  activity
• Removing children from institutions and the
  accompanying deprivation enhances childrens
  lives across all domains
• For some domains (IQ, EEG, attachment) earlier
  placement is better than later
• For other domains, timing of intervention did not
  affect outcome (behavior problems)
• Finally, the lives of children in Romania are
  better now as a result of government policy
  changes towards early institutionalization