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