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, 2014WHIT
EW
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
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.
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
Neural Circuits are Wired in a
Bottom-Up Sequence
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)
Sensory Pathways(Vision, Hearing)
LanguageHigher Cognitive Function
Source: C.A. Nelson (2000)
birth 6 years 14 years
Experience Shapes Brain Architecture by Over-Production Followed by Pruning
(700 synapses formed per second in the early years)
Interaction shapes brain circuitry
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.
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
Mechanisms of architectural
change in the brain that underlie
sensitive periods
Knudsen, 2004
Lorenz and imprinted ducklings walking
after him
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
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
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
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
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
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
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
(Hensch, 2005)
Can we reopen sensitive periods in the visual
system?
The answer appears to be yes!!
Takao Hensch
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.
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
Project Background
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
• 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
• 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 regime
The aftermath….
100,000 children “warehoused” in state institutions
The Study
12
63 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
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 bold and underlined
will be discussed in this talk
THE INTERVENTION
BEIP FOSTER CARE
Goal was to have foster care that was:
• Effective
• Affordable
• Replicable
• Culturally sensitive
• Informed by latest clinical and research findings
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
THE FOSTER CARE TEAM
– BUCHAREST
SUPPORT 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
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)
COGNITION,
LANGUAGE, AND
EXECUTIVE
FUNCTIONS
COGNITIVE DEVELOPMENT
(AT BASELINE)
Smyke et al (2007)
IQ Scores of Foster Care and
Institutionalized Groups at Follow-up
Nelson et al (2007)
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
N0-18 1418-24 1624-30 2230+ 9
*
Nelson, et al (2007)
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
Pa
rtic
ipa
tio
nA
na
lysis
Allo
ca
tio
nE
nro
llme
nt
WISC DATA AT 96 MONTHS OF AGE
7072747678808284868890
Verbal Perceptual Working Memory
Processing Speed
Full Scale IQ
CAU (n=41)
FCG (n=48)
*
Note: *p = .05
Fox et al (2011) JCPP
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
0
20
40
60
80
100
120
VCI PRI WMI PSI Full IQ
CAUG
FCG
12-YEAR IQ COMPOSITES
BY GROUP
***
* p<.05, **p<.01
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
LANGUAGE
AGE OF PLACEMENT SUBGROUPS42 MONTHS
UTTERANCE LENGTH (42 MONTHS)
0
2
4
6
8
10
IG FCG>24 FCG<24 NIG
Mean
len
gth
of
uttera
nce
words
morphemes
-2
-1.5
-1
-0.5
0
z s
core
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.
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.
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
INSTITUTIONALIZATION & BRAIN ACTIVITY
Institution Community Proportion
0.58
0.46
Theta, 3-5 Hz
McLaughlin et al, 2010, Biological PsychiatryMarshall, Fox, and BEIP (2007)
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)
2.44μV2
3.80μV2
Brain Activity (EEG) Changes as a function of intervention and timing
(8 years of age)
CAUGFCG > 24
NIG
FCG < 24
Vanderwert et al (2010)
BRAIN EFFECTS OF EARLY
PSYCHOSOCIAL DEPRIVATION
Initial reports on post institutionalized children 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.
Magnetic Resonance Imaging
(MRI)
STRUCTURAL MRI
Performed in Bucharest on 1.5T Siemens
machine
DTI also obtained on 80% of the children
STRUCTURAL MRI
• FreeSurfer used to automatically detect
grey/white matter borders and identify:
– Total grey matter
– Total white matter
– Hippocampus
– Amygdala
– Basal ganglia : caudate, putamen,
globus pallidus, nucleus accumbens
– Corpus callosum andsubdivisions
– Thalamus
Sheridan et al, (2012) PNAS
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
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)
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
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
MRI SUMMARY
Dramatic reduction in grey and white matter among
ever institutionalized children
No intervention effect for grey matter
Possible (modest) intervention effect for white
matter
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
MENTAL HEALTH PROBLEMS
WHIT
EW
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
PREVALENCE OF ADHD IN THE BEIP
STUDY
0
5
10
15
20
25
Institution Foster Care Community
Pre
vale
nce o
f A
DH
D
(Zeanah et al 2009)
Intervention Effects on Internalizing Disorders
*
0
10
20
30
40
50
60
CAU FC CAU FC
Girls Boys
Pre
vale
nce
of
Inte
rnali
zin
g D
iso
rders
WHITEWWHITEW
WHIT
EW
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
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
ATTACHMENT
BASELINE DIFFERENCES:
11-31 MONTHS
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
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
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)
Intervention Effects
at 42 Months
EFFECTS OF INTERVENTION ON
SECURITY OF ATTACHMENT42 MONTHS
Smyke et al (2010)
Security of attachment
as a function of age of entry
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
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.
STRANGER AT THE DOOR
54 MONTHS
Gleason et al (2013)
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
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
DOMAINS SEEMINGLY
UNAFFECTED BY
EARLY ADVERSITY
Face processing
Emotion recognition
Face recognition
DOMAINS
UNAFFECTED BY
INTERVENTION
Externalizing
problems
Executive function
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
OVERALL
CONCLUSIONS
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)
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
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
Top Related