These documents were presented by Dr. NATHAN A. FOX, UNIVERSITY OF MARYLAND at the session sponsored by The Nippon Foundation, Tokyo, Japan, in the international society for the prevention of child abuse and neglect conference in Nagoya, Japan, on September 14, 2014.

1. TIMING MATTERS: HOW TO THINK ABOUT THE EFFECTS OF EARLY EXPERIENCE ON BRAIN AND BEHAVIORAL DEVELOPMENT
NATHAN A. FOX
UNIVERSITY OF MARYLAND
NIPPON FOUNDATION
INTERNATIONAL SOCIETY FOR THE PREVENTION OF CHILD ABUSE AND NEGLECT CONFERENCE
NAGOYA, JAPAN
SEPTEMBER 14-17, 2014
WHITEW

2. OUTLINE OF TALK
Principles of brain and behavioral development
Sensitive periods
Using the Bucharest Early Intervention Project to illustrate these principles
Implications for policy

3. 3
Brains are built over time,
starting in the earliest years
of life. Simple skills come
first; more complex skills
build on top of them.
Brain Architecture Supports Lifelong Learning, Behavior, and Health
Cognitive, emotional, and
social capabilities are
inextricably intertwined throughout the life course.
A strong foundation in the early years improves the odds for positive outcomes and a weak foundation increases the odds of later difficulties.

4. The Ability to Change Brains
Decreases Over Time
Source: Levitt (2009)
Birth
10
20
30
Physiological “Effort”Required to Enhance Neural Connections
Normal Brain Plasticity Influenced by Experience
Age (Years)
40
50
60
70

5. Neural Circuits are Wired in a
Bottom-Up Sequence
FIRST YEAR
-8-7-6-5-4-3-2-1123456789101112345678910111213141516171819
Birth
(Months)
(Years)
Sensory Pathways
(Vision, Hearing)
Language
Higher Cognitive Function
Source: C.A. Nelson (2000)

6. birth 6 years 14 yearsExperience Shapes Brain Architecture by Over-Production Followed by Pruning(700 synapses formed per second in the early years)

7. Interaction shapes brain circuitry

8. 8
Young children naturally reach out for interaction through babbling, facial expressions, and gestures, and adults respond in kind.
Serve & Return Builds Brains and Skills
These “serve and return" interactions are essential for the development of healthy brain circuits.
Therefore, systems that support the quality of relationships in early care settings, communities, and homes also support the development of sturdy brain architecture.

9. SENSITIVE PERIODS
Are limited time periods during which the effect of experience on the brain is particularly strong
Allow experience to instruct neural circuits to process information in an adaptive way
Provide information that is essential for normal development and may alter performance permanently

10. Mechanisms of architectural change in the brain that underlie sensitive periods
Knudsen, 2004

11. Lorenz and imprinted ducklings walking after him

12. Hubel and Weisel: Classic studies on early experience and sensitive periods
Monocular deprivation in early infancy led to deficits in brain organization in visual cortex

13. Effects of unilateral and bilateral deprivation of patterned vision (Lewis & Maurer, 2005)
Studies of infants born with bilateral cataracts—timing of surgical removal
Daphne Maurer

14. Early experience establishes the neural substrate for capabilities that emerge at a much later point in development
14 years later children whose cataracts were removed late in infancy were deficient in face processing
Sleeper effects as a function of sensitive period input

15. Janet Werker
Effects of experience on perception of language (Werker & Tees, 2005)
The timing and nature of experience with language affects perception of different languages
Before 9-10 months of age, infants can discriminate the sounds of all languages. By the end of the first year of life, they are only able to discriminate the language(s) they hear in their environment

16. Werker & Tees (2005)
Multiple sensitive periods across development for different domains
Most probably there are different sensitive periods for different skills across the first years of life

17. Sensitive Periods for Integration of Auditory and Visual Information
Eric Knudsen studies Barn Owls manipulating their early visual or auditory experience identifying sensitive periods for these inputs

18. People blind from birth hear language in their visual cortex
A sensitive period for language in the visual cortex: Distinct patterns of plasticity in congenitally versus late blind adults
Marina Bedny

19. (Hensch, 2005)
Can we reopen sensitive periods in the visual system?
The answer appears to be yes!!
Takao Hensch

20. GENERAL CONCLUSIONS ABOUT SENSITIVE PERIODS
Collectively, in most cases sensory/perceptual development proceed normally if such systems are “set”correctly during a sensitive period of development—(e.g. Hubel and Wiesel)
There is also evidence for sensitive periods in specific domains of language and perception
The human brain “expects” certain types of input at particular times in development.
It is 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.

21. The Bucharest Early Intervention Project seeks to:
•Examine the effects of psychosocial deprivation on brain and behavioral development of young children
•Determine if these effects can be remediated through intervention, in this case: foster care
•Examine the issue of timing of intervention or duration of deprivation and its effects on brain and behavior with an eye towards identifying sensitive periods in cognitive and social development

22. Project Background

23. CEAUSESCU’S LEGACY TO ROMANIA
Communist Policy:1966 decree
•Raise production by increasing population
•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

24. •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.
THE RESULTS OF CEAUSESCU’S 1966 POLICY

25. •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
1989: The fall of the Ceausescu regimeThe aftermath….
100,000 children “warehoused”in state institutions

28. The Study

29. 1
2
6
3
4
5
Institutional Group
FCG
n=68
NIG
n=72
CAUG
n=68
After baseline assessment (pre-group assignment), comprehensive
follow up performed at 30, 42, 54 months, 8 and 12 years
BEIP Study design

30. DOMAINS OF ASSESSMENT IN BEIP
•Physical Development
•Language•Cognition•Language•Brain Function
•Emotional reactivity
•Caregiving Environment•Attachment
•Social competence
•Mental Health Problems
•Genetics
*Data derived from measures listed in boldand underlined
will be discussed in this talk

31. THE INTERVENTION

32. BEIP FOSTER CARE
Goal was to have foster care that was:
•Effective
•Affordable
•Replicable
•Culturally sensitive
•Informed by latest clinical and research findings

33. THE FOSTER CARE TEAM –BEIP
Comprised 3 social workers
Match families and children
Preparation of applications for foster care
Appearances before the Commission
Supervision of visits with biological family
Management of challenging child behavior
Permanency planning

34. –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

35. GENERAL HYPOTHESES OF THE BEIP STUDY
•Institutional rearing will have profound effects upon children’s 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)

36. COGNITION, LANGUAGE, AND EXECUTIVEFUNCTIONS

37. COGNITIVE DEVELOPMENT (AT BASELINE)
Smyke et al (2007)

38. IQ Scores of Foster Care and Institutionalized Groups at Follow-up
Nelson et al (2007)

39. How does IQ differ for children in foster care as a function of age of placement?
Age at placement
70
75
80
85
90
95
0-18
18-24
24-30
30+
DQ/IQ
N
0-18 14
18-24 16
24-30 22
30+ 9
*
Nelson, et al (2007)

40. CHANGE IN GROUP ASSIGNMENT OVER TIME/SUBJECT ATTRITION (AS OF 96 MONTHS)
Randomized (n=136)
Assessed for Eligibility (n=187)
Excluded (n=51)
Allocated to Foster Care (n=68)
Allocated to Care as Usual (n=68)
Placement at 96 months (n=60)
-31 MacArthur Foster Care
-7 adopted
-8 Government Foster Care
-12 Returned to Bio Family
-2 Social Apartments
Discontinued Participation (n=8)
Placement at 96 months (n=56)
-15 Institutional Care
-4 adopted
-18 Government Foster Care
-18 Returned to Bio Family
-1 Family Placement
Discontinued Participation (n=12)
Analyzed (n=53)
Excluded from analysis (n=7)
-WISC data not available
Analyzed (n=50)
Excluded from analysis (n=6)
-WISC data not available
Participation
Analysis
Allocation
Enrollment

41. WISC DATA AT 96 MONTHS OF AGE
70
72
74
76
78
80
82
84
86
88
90
Verbal Perceptual Working
Memory
Processing
Speed
Full Scale
IQ
CAU (n=41)
FCG (n=48)
*
Note: *p = .05
Fox et al (2011) JCPP

42. COMPARING CURRENT PLACEMENT IN INSTITUTIONS,
GOVERNMENT FOSTER CARE AND MACARTHUR FOSTER CARE
0
20
40
60
80
100
V PR WM PS FSIQ
Institution (n=12)
Gov't FC (n=23)
Mac FC (n=28)
Note: V = Verbal Comprehension, P = Perceptual, WM = Working Memory, PS = Processing Speed,
FSIQ = Full Scale IQ; *p < .05, **p < .01.
*
*
** *
Fox et al (2011) JCPP

43. 0
20
40
60
80
100
120
VCI
PRI
WMI
PSI
Full IQ
CAUG
FCG
12-YEAR IQ COMPOSITESBY GROUP
*
**
* p<.05, **p<.01

44. INTERIM SUMMARY: IQ
Young children living in institutions show significant delays in IQ
Removal from institutions, particularly prior to 24 months of age, and placement into families remediates IQ deficits
Stability of family placement over age is an important factor in IQ outcomes
Remarkably, 10 years after the intervention began there are still positive effects on IQ

45. LANGUAGE

46. AGE OF PLACEMENT SUBGROUPS42 MONTHS

47. UTTERANCE LENGTH (42 MONTHS)
0
2
4
6
8
10
IG
FCG>24
FCG<24
NIG
Mean length of utterance
words
morphemes

48. -2
-1.5
-1
-0.5
0
z score
Reading (8years)
(Single word identification)
IG (N=47)
FCG>24 (N=31)
FCG<24 (N=24)
This figure includes all children, assigning a zero score to (mainly IG) children who could not read.

49. INTERIM SUMMARY: LANGUAGE
Young children living in institutions have significant delays in expression and receptive language
Removal from institutions, particularly before 24 months of age, is associated with remediation of language
These effects continue through age 8.

50. BRAIN ACTIVITY-EEG
Beta 10-18 Hz
Alpha 6-9 Hz
Theta 3-5 Hz
Delta <3 Hz
CAT CAT CAT
Cat
cat
CAT CAT CAT
Cat

51. INSTITUTIONALIZATION & BRAIN ACTIVITY
Institution
Community
Proportion
0.58
0.46
Theta, 3-5 Hz
McLaughlin et al, 2010, Biological Psychiatry
Marshall, Fox, and BEIP (2007)

52. INSTITUTIONALIZATION & BRAIN ACTIVITY
Institution
Community
Alpha, 6-9 Hz
McLaughlin et al, 2010, Biological Psychiatry
Proportion
0.58
0.46
Marshall, Fox & BEIP group (2007)

53. 2.44μV2
3.80μV2
Brain Activity (EEG) Changes as a function of intervention and timing
(8 years of age)
CAUG
FCG > 24
NIG
FCG < 24
Vanderwert et al (2010)

54. BRAINEFFECTS OF EARLY PSYCHOSOCIAL DEPRIVATION
Initial reports on post institutionalizedchildren reported reduced metabolic activity in temporal and frontal cortex;
In one study Tottenham et al. show enlarged amygdala volumes, whereas Hanson/Pollak et al. show smaller amygdala volume. In both cases, however, must keep in mind potential for sample bias.

55. Magnetic Resonance Imaging (MRI)

56. STRUCTURAL MRI
Performed in Bucharest on 1.5T Siemens machine
DTI also obtained on 80% of the children

57. STRUCTURAL MRI
•FreeSurferused to automatically detect grey/white matter borders and identify:
–Total grey matter
–Total white matter
–Hippocampus
–Amygdala
–Basal ganglia : caudate, putamen, globuspallidus, nucleus accumbens
–Corpus callosum andsubdivisions
–Thalamus
Sheridan et al, (2012) PNAS

58. TOTAL CORTICAL GREY MATTER
*
IG B= -39.9, t= -3.01, p= .004
FCG B= -38.5, t= -2.79, p= .007
Regression controlling for age and gender
*
Sheridan et al (2012), PNAS

59. TOTAL CORTICAL WHITE MATTER
*
IG B= -24.1, t= -2.17, p= .03
FCG B= -18.1, t= -1.5, p= .12
Regression controlling for age and gender
n.s.
Sheridan et al (2012), PNAS)

60. a
+
t = 1.8, p = .06
b
*
t = 2.9, p = .005
d
+
t = 1.9, p=.06
e
n.s.
t = 1.4, p = 15
c
+
t = 1.9, p = .06
Regression Analysis: Group, Age, Gender

61. INSTITUTIONALIZATION & NEURAL STRUCTURE
•Institutionalization associated with:
•Total grey matter *
•Total white matter*
•Hippocampus
•Amygdala
•Basal ganglia: caudate, putamen, globus pallidus, nucleus accumbens
•Corpus callosum*
•Thalamus

62. MRISUMMARY
Dramatic reduction in grey and white matter among ever institutionalized children
No intervention effect for grey matter
Possible (modest) intervention effect for white matter

64. STRUCTURE AND FUNCTION
•Associations between institutionalization and reductions in alpha power are fully mediated by white matter integrity
•White matter variation in typical adults accounts for individual differences in EEG alpha power1
•Damage to white matter results in decreases in EEG alpha power2
1. Valdés-Hernández et al., 2010; 2. Claudio Babiloni et al., 2006

65. MENTAL HEALTH PROBLEMS
WHITEW

66. 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

67. PREVALENCE OF ADHD IN THE BEIP STUDY
0
5
10
15
20
25
Institution
Foster Care
Community
Prevalence of ADHD
(Zeanah et al 2009)

68. Intervention Effects on Internalizing Disorders
*
0
10
20
30
40
50
60
CAU
FC
CAU
FC
Girls
Boys
Prevalence of Internalizing Disorders
WHITEW
WHITEW

69. SUMMARY OF FINDINGS ON PSYCHIATRIC DISORDERS
In general, children raised in institutions display greater internalizing and externalizing disorders compared to community controls.
But children removed from institutions and placed into foster care homes display significantly less internalizing problems compared to those randomized to remain in the institution.
•This was particularly true for girls
There was no intervention effect for externalizing problems— specifically ADHD

70. INTERIM SUMMARY
Early exposure to severe psychosocial deprivation is associated with higher slow wave and lower high frequency EEG activity and heightened internalizing and externalizing problems
Foster care intervention (particularly if it occurs early in life) remediates this pattern of EEG activity and internalizing problems. It does not remediate externalizing problems— specifically ADHD
EEG activity early in life is associated with lower internalizing problems and mediates the relation between early experience and later problems with ADHD

71. ATTACHMENT

72. BASELINE DIFFERENCES: 11-31 MONTHS

73. 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

74. 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

75. DEGREE TO WHICH CHILDREN LIVING IN INSTITUTIONS HAVE
FORMED ATTACHMENTS TO THEIR CAREGIVERS
Romanian
Community
Romanian
Institution
1=No attachment
0%
9.5%
2=Some
differentiation
0%
25.3%
3=Some preference
0%
30.5%
4=Attachment with
anomalies
0%
31.6%
5=Clearly recognizable
attachment patterns
100%
3.2%
Zeanah et al (2005)

76. Intervention Effects
at 42 Months

77. EFFECTS OF INTERVENTION ON SECURITY OF ATTACHMENT42 MONTHS
Smyke et al (2010)

78. Security of attachment
as a function of age of entry

79. 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

80. STRANGER AT THE DOOR
Caregiver/mother and child answer door (pre-arranged).
RA: “Come with me, I have something to show you.”
Walk out the door and around the corner to find RA from previous home visit.

81. STRANGER AT THE DOOR 54 MONTHS
Gleason et al (2013)

82. EFFECTS ON TIMING OF PLACEMENT ON INDISCRIMINATE BEHAVIOR
Children placed into families before 24 months of age showed absence of indiscriminate behavior across follow-up

83. DOMAINS WITH SENSITIVE PERIOD
IQ
Attachment
EEG power
Social skills (teacher report)
Indiscriminate behavior
DOMAINS WITH NO SENSITIVE PERIOD (BUT INTERVENTION EFFECTS)
Psychiatric outcomes (anxiety)
Positive emotional reactivity
Peer social competence

84. DOMAINS SEEMINGLY UNAFFECTED BY EARLY ADVERSITY
Face processing
Emotion recognition
Face recognition
DOMAINS UNAFFECTED BY INTERVENTION
Externalizing problems
Executive function

85. SENSITIVE PERIODS AND BEIP
Both language and EEG activity show sensitive periods in development
Cognition (IQ) does not but does show an intervention effect even 10 years after the start of the intervention.
Early experience affects different domains and intervention has differential effects depending on timing

86. OVERALLCONCLUSIONS
Children raised in institutions during early development demonstrate significantly impaired physical, cognitive, language, social-emotional, and brain development compared to community children.
•Here we demonstrate specific cognitive deficits and neural neural compromises that may be associated with these deficits
Insofar as we have been able to look at our data, our model of foster care as an intervention appears to effectively ameliorate many of the negative sequelae of institutionalization…
Although not all…for example, we are seeing only small effects on executive functions….and only at 12 years of age
Some aspects of brain development (i.e. certain white matter tracts) are remediated in children placed into foster care, whereas others are not (i.e. total gray matter, certain white matter tracts)

87. IMPLICATIONS FOR POLICY
Brains develop as a function of expected environmental experience
In the absence of expected experience brain and behavioral develop are derailed
Children living in conditions of neglect are likely to suffer from the absence of stimulation (warm, responsive, contingent interaction)
Removing children from conditions of neglect (and the earlier the better) is good for their brain development and good for their behavioral development

88. INVESTIGATIVE TEAM
Principal Investigators
Charles Zeanah, MD, Tulane University
Nathan A. Fox, Ph.D., University of Maryland
Charles A. Nelson, Ph.D., Harvard Medical School/Children’s Hospital Boston
Funding:
•John D. and Catherine T. Macarthur Foundation
•NIMH MH091363-01