Biologic Embedding of Neighborhood- Level Stress in Children · Genetic variation Environment. Telomeres ???? Early life ... • Biology Adverse birth outcomes, ... TL shorter in

Katherine P. Theall, PhDTulane UniversitySchool of Public Health and Tropical Medicine

The Pennington Biomedical Research Center Childhood Obesity & Public Health Conference

October, 2015

Biologic Embedding ofNeighborhood-Level Stress in Children

Outline

• Adverse childhood experiences (ACE), Life Course and Toxic Stress▫ Neighborhood context

• Biological ‘embedding’ and empirical evidence▫ Telomere length

• Implications for obesity and chronic disease risk

Disclosures: no financial conflicts

• Funding: R01ES020447 (NIEHS)* K01SH000002 (CDC) R01 MH101533 (NIMH & NIMHD supplement, Drury)* Tulane Oliver Fund* Frost Foundation* R21 MH094688 (NIMH, Drury) Tulane Research Enhancement Fund

* current

Parents are important Trauma is not good Lots of things effect

kids Kids are different from

each other Kids are different

from adults

A couple of obvious points….

Conceptual Model – Neighborhoods & Obesity

Neighborhood Conditions

Family/Household Conditions

Biologic Stress / Telomere Length

Obesity / Overweight

Biologic Stress / Telomere Length

Individual Characteristics

FamilyMicrosystem

PeerMicrosystem

NeighborhoodMicrosystem

SchoolMicrosystem

Macrosystem

Exosystem

Bronfenbrenner’s Ecological Theory

Abuse

Household Dysfunction

(e.g., substance abuse, mental illness, IPV)

Neglect Support and Praise

Tis but a scratch

Why Children?

Critical or Sensitive Periods (Fine and Kotelchuck, 2010)

Impact of adverse experience may be greatest at specific critical or sensitive periods of development (e.g. during fetal development, in early childhood, during adolescence) Sensitive – more able to reverse or change vs. critical

period

Biological embedding

Accumulation Models (Kuh and Shlomo, 2004)

The accumulation of adverse experiences matter

Life Course Concepts ‘T2E2’

• Timeline – health is cumulative and longitudinal, i.e., developed over a lifetime.

• Timing – health and health trajectories are particularly affected during critical and sensitive periods.

• Environment – the broader environment (social, physical, economic) affects health and development.

• Equity – health inequality (inequity = unjust) reflects more than genetics and personal choice.

– Fine and Kotelchuck

Three levels of stress:

• Positive stress ▫ Brief increases in heart rate, mild

elevations in stress hormones• Tolerable stress▫ Serious, temporary stress response

buffered by supportive relationships• Toxic stress▫ Prolonged activation of the stress

response systems in the absence of protective relationships

11

Center on the Developing Child Working Paper: Excessive Stress Disrupts the Architecture of the Developing Brain

Adverse early experience

Altered stress Response systems

Altered neurodevelopmentalTrajectories/circuits/synapses

Altered behavioral,psychological,

cognitive and healthoutcomes

Genetic variation

Environment

Telomeres ????

Early life adversity

DNAHealth

Telomere basics – Beginning of the End

• Links with diseases:▫ Obesity▫ Cardiovascular▫ Dementia▫ Diabetes▫ Mental illness

• Links with environment:▫ Caregiving stress▫ Violence▫ Abuse▫ Lead and other toxins▫ Radiation

…similar to the negative health outcomes associated with adverse childhood events

Telomeres as sensors and antennas

Systemic Level: Paths to Cell Aging

Biochemical Stressors

Cortisol

Cytokines

Insulin

Oxidative Stress

Adiposity

Telomere and Telomerase

Maintenance System

Threat perception and brain response

Epel, Hormones, 2009

Mitochondrial Damage

Oxidative Stress

Youth studies and telomere length

• Diseases▫ Obesity▫ Diabetes▫ HIV ▫ Autism

• Environment▫ Institutional care▫ Abuse and maltreatment▫ Community disorder▫ Family instability▫ Lead▫ Parental education▫ Prenatal smoking▫ Neighborhood violence*

• Biological ▫ HPA/ANS reactivity▫ Testosterone▫ DHEA*

• Transgenerational▫ Parental alcoholism*▫ Maternal obesity*

*in preparation/under review

Why Neighborhoods?• Diseases▫ Obesity▫ Cardiovascular disease▫ Mental health▫ Mortality

• Environment▫ Domestic violence▫ Health behaviors (e.g. smoking, physical activity)

• Biology▫ Adverse birth outcomes, low birth weight ▫ Biological stress, e.g., telomere length, cortisol

PUBLIC HEALTH IMPACT!

National Adolescent Sample

• Research Question:▫ Is the neighborhood crime risk environment

and cumulative neighborhood risk (e.g., crime, alcohol outlet density, concentrated disadvantage) associated with physical activity (PA), obesity/overweight, and allostatic load (AL)?

▫ Is the relation moderated by sex?

PA classified as engaging in vigorous or moderate physical activity in last week; obesity/overweight based on age-for-sex CDC growth charts; AL based on 9 biomarkers across multiple systems.

National Health and Nutrition and Examination Survey (NHANES)

• Continuous Surveys 1999 – 2006▫ Cross-sectional, biennial surveys (4 waves)▫ In-home survey, medical exam, laboratory

• Adolescents age 12 – 20 years (mean=15.6)▫ N=9887 with geographic information▫ Nested within both household and neighborhood

(n=2400 tracts)

Clustering of Physical Activity (PA) and Obesity by Neighborhood

ICC

%

Effect Modification by Sex

Adjusted Odds Ratio (OR) , 95% Confidence Interval (CI)Outcomes: PA, Obesity/Overweight; Exposure: High Crime

Overall Among girls Among boys

Physical Activity0.86 (0.73, 1.00) 0.74 (0.59, 0.92) 0.98 (0.80, 1.21)

Obesity/Overweight1.11 (1.04, 1.31) 1.27 (1.02 , 1.58) 1.11 (0.94, 1.31)

Adjusted for: age, education, race, Poverty Income Ratio, HEI score, survey year, years living in neighborhood, adult PA and BMI, urban-rural status

0

1

2

3

4

5

6

7

Neighborhood Clustering of Allostatic Load and Individual Biomarkers

ICC (%

)

Impact of Neighborhood Cumulative Risk on Allostatic Load (AL)

AL: ≥ 2 biomarkers; > 1 S.D. OR (95% CI) aOR *(95% CI)

Cumulative risk Low to none (0) 1.00 1.00

Medium (1-2) 1.16 (1.16, 1.17) 1.09 (1.08, 1.09)

High (3-4) 1.37 (1.36, 1.37) 1.28 (1.27, 1.30)

Extremely high ( > 4) 1.84 (1.83, 1.85) 1.69 (1.68, 1.70)

ICC (%) 6.23 (empty) to 2.71 2.09Median Odds Ratio (MOR) 2.20 (empty) to 1.47 1.42Explained group level variance 65.0% 15.4%

* Adjusted for sex, age, race/ethnicity, family poverty income ratio (PIR), household adult AL score and education, marital status, years in neighborhood, HEI, physical activity, ICE, urbanicity, survey year.

NSPAC (Neighborhood and Stress Physiology Among Children) Study

• Sample of 120 African American children▫ Ages 5-16▫ Recruited public schools, clinics, local partners in

New Orleans, Louisiana, U.S. ▫ January 2012 – July 2013▫ Age- and Katrina-exposure matched 1:1▫ 32.5% obese or overweight

• Georeferenced to census tract level▫ 51 tracts (“neighborhoods”)▫ Buffers around the home (500, 1000, 2000 meters)

Impact of Witnessed Violence

Adjusted B overall = -0.20, p < 0.05

Adjusted B girls = -0.30, p < 0.05

Adjusted B boys = -0.13, n.s.

Objective Neighborhood Effects

Community Violence and Telomeres, cortisol and behavior Neighborhood exposure - Total Crime per 1000 Domestic Violence Calls

per 1000 Outcomes

Beta (SE) P-value Beta (SE) P-value

Telomere length (T/S ratio)

-1.43 (0.39) 0.0003 -1.11 (0.37) 0.002

Cortisol recovery

0.22 (0.23) 0.081 0.10 (0.004) 0.043

Aggression and impulsivity

15.98 (4.37) 0.032 1.89 (0.90) 0.054

Theall et al. in preparation

What about overweight/obesity? Homicide Rate and BMI and WC

Rho = 0.205; R2 = 5.1% Rho = 0.179; R2 = 3.5%

Conceptual Model – Neighborhoods & Obesity

Neighborhood Conditions

Family/Household Conditions

Biologic Stress / Telomere Length

Obesity / Overweight

Biologic Stress / Telomere Length

Telomere Length and Obesity

• Inconsistent association in children but recent, larger studies:▫ Al-Attas et al. (2010), Arab youth Boys: TL shorter in obese vs non-obese and R2= 86% SBP Girls: No differences by obesity and R2= 24% WC

▫ Buxton et al. (2011), French youth Obese children had a mean LTL 23.9% shorter than non-

obese ▫ García-Calzón et al. (2014), Spanish obese/overweight

youth Significant increase in TL following weight loss; higher

baseline TL higher decrease in body weight in boys

Adolescent Intervention & Impact on TL

Garcia-Calzon et al, 2014

TL and BMI and WC in NSPAC

Rho = -0.371; R2 = 13.8% Rho = -0.459; R2 = 21.1%

Mediation Model – Homicide and BMI

Homicide Rate

Telomere Length

BMI

Direct effect = 0.3939 (0.136); t=2.89, p=0.005

* Indirect effect = -0.062 (Boot SE=0.0.037), Boot CI = -0.1644, -0.0072

* *

Moderation Model – Homicide and BMI

Homicide Rate

Telomere Length

BMI

* Moderating effect TL * Homicide Rate z=1.93, p=0.053

*

Moderation Model – Homicide and BMI

R2 = 20.0%; z=7.39, p = 0.011 R2 = 2.6%; z=2.19, p = 0.143

Among those with higher TL Among those with lower TL

Implications• TL maybe a biological “scar” of early adversity, including

neighborhood exposures▫ Multiple pathways and black boxes!

• Understanding the factors the influence TL may provide new avenues for intervention/prevention

• Monitoring TL may provide insight into the biological, rather than symptomatic, impact of interventions

Collaborators Thank you • Stacy Drury• Elizabeth Shirtcliff• John Carlson

• Richard Scribner• Claire Sterk• Charlie Zeanah • Steve Nelson• Patricia Molina

QUESTIONS?