Positive Parenting Matters in the Face of Early Adversity

Abstract

Introduction:
A negative relationship between adverse childhood experiences and both physical and mental health in adulthood is well established, as is the positive impact of parenting on child development and future health. However, few studies have investigated unique influences of adverse childhood experiences and positive parenting together within a large, diverse early childhood sample.

Methods:
The study used data on all children aged 0-5 years (n=29,997) from the National Survey of Children's Health 2011/2012 to examine effects of positive parenting practices and adverse childhood experiences on early childhood social-emotional skills and general development. All analyses were performed in 2017 and 2018.

Results:
More than a third of the sample reported experiencing at least one adverse childhood experience. More than a fourth (26.7%) met study criteria for social-emotional deficits, and 26.2% met criteria for developmental delay risks. The number of adverse childhood experiences exhibited negative marginal associations with social-emotional deficits and developmental delay risks, whereas the number of positive parenting practices showed independent protective effects. Risks associated with an absence of positive parenting were often greater than those of four or more adverse childhood experiences, even among no/low adversity families. The population attributable fractions for social-emotional deficits and developmental delay risks were 17.3% and 13.9% (translating to prevalence reductions of 4.5% and 3.6%) when adopting all positive parenting practices and 4.5% and 7.2% (prevalence reductions of 1.2% and 1.9%) when eliminating adverse childhood experiences.

Conclusions:
The number of adverse childhood experiences was associated with both social-emotional deficits and developmental delay risks in early childhood; however, positive parenting practices demonstrated robust protective effects independent of the number of adverse childhood experiences. This evidence further supports promotion of positive parenting practices at home, especially for children exposed to high levels of adversity.

Full text

RESEARCH ARTICLE
Positive Parenting Matters in the Face
of Early Adversity
D1X XYui Yamaoka, D2X XMD, PhD, D3X XDavid E. Bard, D4X XPhD
Introduction: A negative relationship between adverse childhood experiences and both physical
and mental health in adulthood is well established, as is the positive impact of parenting on child
development and future health. However, few studies have investigated unique influences of adverse
childhood experiences and positive parenting together within a large, diverse early childhood sample.
Methods: The study used data on all children aged 0−5 years (n=29,997) from the National Sur-
vey of Children’s Health 2011/2012 to examine effects of positive parenting practices and adverse
childhood experiences on early childhood social−emotional skills and general development. All
analyses were performed in 2017 and 2018.
Results: More than a third of the sample reported experiencing at least one adverse childhood experi-
ence. More than a fourth (26.7%) met study criteria for social−emotional deficits, and 26.2% met crite-
ria for developmental delay risks. The number of adverse childhood experiences exhibited negative
marginal associations with social−emotional deficits and developmental delay risks, whereas the num-
ber of positive parenting practices showed independent protective effects. Risks associated with an
absence of positive parenting were often greater than those of four or more adverse childhood experi-
ences, even among no/low adversity families. The population attributable fractions for social−emo-
tional deficits and developmental delay risks were 17.3% and 13.9% (translating to prevalence
reductions of 4.5% and 3.6%) when adopting all positive parenting practices and 4.5% and 7.2% (prev-
alence reductions of 1.2% and 1.9%) when eliminating adverse childhood experiences.
Conclusions: The number of adverse childhood experiences was associated with both social−emo-
tional deficits and developmental delay risks in early childhood; however, positive parenting practices
demonstrated robust protective effects independent of the number of adverse childhood experiences.
This evidence further supports promotion of positive parenting practices at home, especially for chil-
dren exposed to high levels of adversity.
Am J Prev Med 2018;000(000):1
−
10. © 2018 American Journal of Preventive Medicine. Published by Elsevier Inc.
All rights reserved.
INTRODUCTION
Adverse childhood experiences (ACEs) continue
to garner public attention for their cumulative
negative health consequences in adulthood.
1−4
Many theorize the accumulation of adversities can lead
to excessive or prolonged stress,
5
and in the absence of
sensitive and responsive caregivers, this stress becomes
toxic and can disrupt brain development, which in turn
causes lifelong impairments. Negative ACE effects are
documented for young adults, adolescents, and even
children.
4,6−9
However, only a few studies
10,11
have
examined, at a population level, the isolated effects of
ACEs on health or health precursors occurring during
critical early stages of childhood.
The prevailing view among early childhood profes-
sionals frames development as a synthesized product of
From the Section of Developmental and Behavioral Pediatrics, Depart-
ment of Pediatrics, University of Oklahoma Health Sciences Center, Okla-
homa City, Oklahoma
Address correspondence to: David Bard, PhD, Developmental and
Behavioral Pediatrics, Department of Pediatrics, University of Oklahoma
Health Science Center, 940 Northeast 13th St. Nicholson Tower, Okla-
homa City OK 73104. E-mail: david-bard@ouhsc.edu.
0749-3797/$36.00
https://doi.org/10.1016/j.amepre.2018.11.018
© 2018 American Journal of Preventive Medicine. Published by Elsevier Inc. All rights
reserved.
Am J Prev Med 2018;000(000):1−10 1
ARTICLE IN PRESS

negative outcomes from toxic stress and positive, adap-
tive outcomes from “protective factors.”
11,12
Garner and
Shonkoff
13
capture current thinking, explaining “the
essence of toxic stress is the absence of buffers [i.e., pro-
tective factors] needed to return the physiologic stress
response to baseline.”This emphasis on adversity harms
and protective buffers permeates most modern health
promotion service systems and aligns with the mental
health dual continuum movement.
14
The Centers for
Diseases Control and Prevention, frontrunners of early
child adversity research, widely promote creation of
“safe, stable, and nurturing relationships and environ-
ments”as essential protective factors for all children.
15
Similarly, Strengthening FamiliesTM is a broadly imple-
mented program encouraging the protective factors
approach with high-adversity families.
16
One commonly cited, modifiable protective factor is par-
enting.
11,17
Key parenting practices not only protect chil-
dren from adversity but also stimulate development that
enhances resiliency. Recently, the National Academy of Sci-
ences, Engineering, and Medicine released Parenting Mat-
ters: Supporting Parents of Children Ages 0
−
8,
18
which
underscores the importance of quality parenting for child
development. Despite popularity of this view,
18−20
limited
information exists detailing the combined impact of child-
hood adversities and parenting practices on early develop-
ment,
11
in part because selective samples from clinical trials
usually restrict variability on parenting and/or adversity
outcomes. As such, essential questions remain, and herein,
special attention is drawn to one: Do protective effects of
positive parenting practices (PPPs) persist even in the face
of adversities? To explore this question further, this study
considers two reasons why PPPs may not evidence protec-
tive effects in the presence of ACEs: (1) Because these are
two sides of the same coin, and ACEs confound the rela-
tionship between PPPs and development; or (2) benefits of
PPPs degrade under higher levels of adversity exposure.
Equipped with a large, nationally representative survey
sample, this study aims (1) to examine the relationship
between ACEs and development during early childhood (0
to 5 years) and (2) to examine protective effects of PPPs, in
the presence and absence of ACEs. To quantify public
health benefits of prevention, the study also aims (3) to esti-
mate population attributable fractions (PAFs)
21−23
for
developmental risks among very young children when
eliminating ACEs or universally adopting PPPs.
METHODS
Study Sample
Study data came from the National Survey of Children’s Health
(NSCH) 2011/2012—a U.S. representative, cross-sectional, list-
assisted random-digit-dial telephone survey. This survey was
initiated, designed, led and sponsored by the Health Resources
and Services Administration/Maternal and Child Health Bureau
and administered by the National Center for Health Statistics
under contract by Health Resources and Services Administration/
Maternal and Child Health Bureau.
24
The authors obtained the
dataset and codebook for the 2011−2012 NSCH from the Child
and Adolescent Health Measurement Initiative Data Resource
Center for Child and Adolescent Health (www.childhealthdata.
org; also sponsored by Health Resources and Services Administra-
tion/Maternal and Child Health Bureau). Participating parents
responded to questions about a single randomly selected child.
The current study included children aged 0−5 years to evaluate
social−emotional skills, development, ACEs, and parenting prac-
tices (n=29,997; 31.4% of total NSCH sample). NSCH item word-
ing for study variables appears in Appendix Table 1. The response
rate for this survey was 23.0%.
24
Additional methodology details
are available elsewhere.
25
Measures
Parent-reported developmental concerns for children aged
4months−5 years were elicited using an NSCH-version of the
Parents’Evaluation of Developmental Status (PEDS). The clinical
PEDS is a standardized, screening instrument assessing parental con-
cerns about developmental delay of children aged <8years.
26
The
NSCH-version includes nine questions from the clinical PEDS but
omits all open-ended comments. Comparable with past
work,
19,20,27,28
this study uses the NSCH codebook criteria for PEDS
scoring to create a binary indicator for developmental delay risk
(DDR) that differentiates low or no risk from moderate or high risk.
The NSCH 2011/2012 included flourishing items developed by
a subgroup of the Child and Adolescent Health Measurement Ini-
tiative−led, Technical Expert Panel that set forth a framework,
domains, and candidate items from which a condensed set was
later selected after vetting combined input from public commen-
tary and subject matter experts.
27,29,30
Although the construct is
multifaceted,
31,32
the authors of this article contend the NSCH
flourishing items for young children (aged 6 months−5 years) pri-
marily assess expected positive health outcomes linked to essential
social−emotional skills. Three items address content areas that
strongly overlap with other social−emotional assessments for this
age group (e.g., Ages and Stages Questionnaire−Social-Emo-
tional
33
) and are conceptually linked to (1) caregiver-child attach-
ment (“tender/affectionate”), (2) self-regulation and resiliency
(“bounces back”), and (3) positive affect (“laughs a lot”).
34
The
fourth item assesses child’s aspiration level (“interest/curiosity in
learning new things”) and closely maps to social−emotional
learning skills which are conceptually linked to the Openness to
Experiences personality factor.
35
Following similar scoring rou-
tines for the older-child flourishing items,
36−39
responses to ques-
tions are dichotomized into 1 for sometimes/rarely/never and 0
for always/usually. Summed scores are collapsed into a binary,
social−emotional deficit (SED) outcome which differentiates
scores above zero (i.e., any sometimes/rarely/never response) and
at zero (i.e., all rated always/usually). Because few have used youn-
ger-age flourishing items, the Appendix describes psychometric
analyses that demonstrate a single factor captures inter-item cor-
relations reasonably well at an adequate level of internal reliability
and that sensitivities and specificities for SED are, to varying
degrees, comparable to those of the clinical PEDS.
40,41
ARTICLE IN PRESS
2Yamaoka and Bard / Am J Prev Med 2018;000(000):1
−
10
www.ajpmonline.org

The NSCH 2011/2012 includes nine items addressing a child’s
lifetime experience with the following adversities: (1) hard to get
by on current income, (2) divorce/separation of parent, (3) lived
with someone with alcohol or drug problem, (4) victim or wit-
nessed neighborhood violence, (5) lived with someone who was
mentally ill or suicidal family member, (6) witnessed domestic
violence, (7) parent served time in jail, (8) treated or judged
unfairly because of race/ethnicity, and (9) death of parent. Item 1
is recoded as binary (collapsing very and somewhat often catego-
ries) to match the other item scales, and ACE counts are catego-
rized into four levels: 0, 1, 2-3, and ≥4. The NSCH-ACE items
have been studied extensively, and support exists for cumulative
score usage.
42
Respondents reported the number of days in the past week
when caregivers engage the child in reading stories, storytelling/
singing, eating meals together, playing with similar-age children
(playing with peer), and family outings. Caregivers also provide
the number of hours or minutes the child spends watching TV.
Adapting previous scoring procedures,
19,20
a PPP binary indicator
is constructed for each activity and operationalized as positive
whenever the frequency of the first five activities is >3 days (more
than half a week) and whenever TV watching is ≤2 hours. All
indicators were summed to produce a PPP count variable that
mimicked ACEs score construction. Items selected represent
behaviors all parents of young children could practice daily. This
set partially overlaps a 2007 NSCH home environment measure
that excluded peer play and family outings and included items not
assessing daily participation (smoking status; breastfeeding
history).
43
Characteristics of children (sex, age, race/ethnicity) and house-
hold (highest education level, income) appear in statistical models
as control variables. Race/ethnicity is categorized as non-Hispanic
white, non-Hispanic black, Hispanic, or other race/multi-race.
Education is coded as more than high school, high school gradu-
ate, and less than high school, and household income is catego-
rized as below federal poverty level (FPL), 100%−199% FPL,
200%−399% FPL, or ≥400% FPL. Unfortunately, sex of surveyed
caregiver is not available; however, NSCH documentation states
69% of respondents are female guardians, 24% male guardians,
and 5% grandparents.
Statistical Analysis
Population proportions for all variables are estimated for the full
sample as are prevalences of individual ACEs and PPPs among
children aged 0−2 and 3−5 years. Hierarchical regression analysis
(not to be confused with Hierarchical Models
44,45
) is used to
quantify effects of ACEs and PPPs on social−emotional deficits
(SEDs) and developmental delay risks (DDRs) and evaluate
potential confounding. This multiple logistic regression procedure
sequentially introduced variable sets starting with an unadjusted
ACEs model (Model 1), then adding demographic controls
(Model 2), and finally adding PPPs (Model 3). An alternate sec-
ond model (Model 2b), which replaced ACEs with PPPs, was also
run to compare PPP effects with (Model 3) and without (Model
2b) ACEs adjustments. PAFs
21
represent the predicted propor-
tional reduction in cases (e.g., children with DDR) when either
risk factor is eliminated (e.g., reducing ACEs) or protective factors
are elevated (e.g., increasing PPPs). Confounder-adjusted PAF
results are presented for ideal alternatives where either all six
PPPs are adopted or all ACEs are eliminated. (The PAF formula is
in Appendix Table 4.) Analyses adjust for complex survey design
variables (sampling weights, clusters, strata) using SVY proce-
dures of Stata, version 14.1.
46
R, version 3.5.0 is used to produce
figures.
47
Analyses were performed in 2017 and 2018. The Univer-
sity of Oklahoma Health Sciences Center IRB reviewed and
approved this study.
RESULTS
Table 1 provides variable proportions for all children
and indicates more than one third (36.7%) experience at
least 1 ACE, most (89.2%) experience ≥3 PPPs, and
roughly one quarter meet study criteria for SED (26.7%)
and DDR (26.2%). Income hardship is the most frequent
ACE reported (Table 2) with comparable prevalence
(24.5% and 26.1%, respectively) in strata of children
aged 0−2 and 3−5 years. All other ACEs affect <15% of
the sample but are 2−3 times more frequent in the older
age group. For younger children, the most frequently
endorsed PPPs are limited TV watching (87.9%), family
meal (84.4%), and storytelling/singing (83.9%). Family
meals were the most popular practice among older chil-
dren (83.8%), and except for family outings (51.7%), the
other PPPs were also highly prevalent (ffi≥75%).
The correlation between ACEs and PPPs was signifi-
cant but small (r=−0.07, p<0.001) and would not typi-
cally signal severe confounding. The smallest raw
frequency for any ACE by PPP combination was 72;
85% of combinations involve ≥200 children. The joint
distribution is characterized in Appendix Figure 1 and
Appendix Table 2.Table 3 shows the effects of ACEs
and PPPs on SED and DDR. ACEs OR, comparing 1+
ACE categories to zero ACEs, displays a significant posi-
tive gradient with SED and DDR (Model 1). All but two
ORs, the 2−3 ACEs effect for SED (p=0.39) and the 1
ACE effect for DDR (p=0.051), remain significant after
adjusting for demographic covariates and PPPs in Model
3. The ORs of Model 3 increase from 1.10 to 1.36 for
SED and from 1.17 to 2.04 for DDR. PPPs show signifi-
cant protective effects for both outcomes after control-
ling for ACEs. Relative to the lowest PPP category
(count <3), those providing all PPPs were attributed half
the odds of meeting criteria for SED (OR=0.49) or DDR
(OR=0.53). Finally, inclusion of interaction terms
between ACEs XPPPs results in, at best, weak evidence
for effect modification. None of the simple effects for
ACE group differences in PPP trend reach statistical sig-
nificance (all p>0.10), but there is a visible difference in
the DDR prediction curve for the 4+ ACEs group. This dif-
ference suggested little or no protective PPP advantage for
this outcome and may be underpowered because of low
numbers of 4+ ACEs participants (Appendix Figures 2 and
3). Per recommendations of recent NSCH work,
42
analyses
ARTICLE IN PRESS
Yamaoka and Bard / Am J Prev Med 2018;000(000):1
−
10 3
&
2018

were repeated using a new ACE measure that dropped the
income hardship item. As shown in Appendix Table 3,this
change does not affect overall model conclusions (and
ACE XPPP interactions remain nonsignificant, p>0.08)
but does result in lower ACEs effects.
It is worth comparing ACEs effects of Model 2 (ACEs
and covariates) and Model 3 (ACEs, PPPs, and covari-
ates) to evaluate the impact of a parenting confounder.
Conversely, comparing PPP main effects in Model 2b
(PPPs and covariates) and Model 3 (ACEs, PPPs, and
Table 1. Description of Child and Household
Characteristics
Unweighted,
a
n(%)
Weighted,
% (95% CI)
Child’s characteristics
Age, years, mean (SD/SE) 2.55 (1.7) 2.53 (0.02) / (2.48, 2.57)
Sex
Male 15,233 (50.8) 51.0 (49.8, 52.3)
Female 14,742 (49.1) 49.0 (47.7, 50.2)
Race
White, non-Hispanic 18,228 (62.3) 50.1 (48.8, 51.3)
Hispanic 4,609 (15.8) 26.4 (25.2, 27.7)
Black, non-Hispanic 2,698 (9.2) 12.2 (11.4, 13.0)
Other, multirace 3,734 (12.8) 11.4 (10.6, 12.2)
Household characteristics
Highest education in household
>High school 22,995 (78.1) 67.7 (66.4, 68.9)
High school graduate 4,584 (15.6) 20.5 (19.4, 21.6)
<High school 1,863 (6.3) 11.8 (10.9, 12.9)
Household income
≥400% FPL 9,875 (32.9) 25.3 (24.3, 26.3)
200%−399% FPL 8,595 (28.7) 26.8 (25.7, 27.9)
100%−199% FPL 5,734 (19.1) 21.8 (20.7, 22.8)
0−99% FPL 5,793 (19.3) 26.1 (25.0, 27.3)
PPPs
PPP counts
0−2 2,420 (8.1) 10.8 (9.9, 11.6)
3 3,983 (13.3) 15.3 (14.4, 16.3)
4 7,553 (25.2) 25.2 (24.2, 26.3)
5 10,134 (33.8) 31.4 (30.3, 32.6)
6 5,907 (19.7) 17.2 (16.4, 18.1)
ACEs
ACE score
0 19,810 (66.8) 63.3 (62.1, 64.6)
1 6,351 (21.4) 24.1 (23.0, 25.3)
2−3 2,676 (9.0) 9.9 (9.2, 10.7)
≥4 804 (2.7) 2.6 (2.3, 3.0)
Social−emotional skill and general development
Social−emotional deficit
No 21,413 (77.6) 73.3 (72.1, 74.5)
Yes 6,199 (22.5) 26.7 (25.5, 27.9)
Developmental delay risk
No/Low risk 21,722 (76.1) 73.8 (72.7, 75.0)
Moderate/High risk 6,818 (23.9) 26.2 (25.0, 27.3)
Note: Weighted % was calculated using design variables (sampling weights and strata indicators).
a
n=29,997.
ACE, adverse childhood experience; FPL, federal poverty level; PPP, protective parenting practice.
ARTICLE IN PRESS
4Yamaoka and Bard / Am J Prev Med 2018;000(000):1
−
10
www.ajpmonline.org

covariates) allows for evaluation of ACEs confounding.
Figure 1 provides a plot of predicted probabilities from
Models 2, 2b, and 3 for this purpose. The ACE and PPP
absolute risk differences do not change much in Model 3
compared with Models 2 and 2b, which suggests little
confounding of either effect. Notably, when comparing
0-2 PPP and 6 PPP families who also report zero ACEs,
models predicted an 11.6% and 10.6% reduction in SED
and DDR (fixing control covariates at their mean or
mode). This same comparison among families with 4+
ACEs reveals risk reductions of 13.6% and 14.4%,
respectively. Flipping the scenario and comparing zero
ACE and 4+ ACE families with 6 reported PPPs, risk
reductions of 4.4% (SED) and 12.4% (DDR) were pre-
dicted. This same ACE comparison among 0-2 PPP fam-
ilies results in 6.4% and 16.1% risk reductions. When
contrasting these simple effects across models, absolute
risk reductions for PPP were similar for both outcomes,
whereas extreme ACE differences produced greater risk
reduction for the DDR outcome.
Under the condition that all families provide all 6 PPPs,
the estimated PAFs for SED and DDR are 17.3% and
13.9%, which represent reductions of 4.5% and 3.6% in
risk prevalence. This translates to an outcome reversal/
benefit (i.e., moving from at risk to not at risk) for roughly
1.1 million children aged <6 years at risk for SED and
0.9 million children at risk for DDR nationwide (Appen-
dix Table 4). Under the condition that all families have
zero ACEs, PAFs for SED and DDR are 4.5% and 7.2%,
which infer prevalence reductions of 1.2% and 1.9%.
Among U.S. children aged <6 years, this equates to an
SED reversal/benefitfor≥282,000 children and a DDR
reversal/benefitfor≥454,000 children.
Surprisingly, the full SED model predicted higher proba-
bility of risk for families reporting low PPPs (0-2) and zero
ACEs than for families reporting all 6 PPPs and 4+ ACEs
(27.1% vs 19.9%). Similarly, Model 3 for DDR predicted
comparable risks for these types of families (27.4% vs
29.2%). Ergo, in some instances, absence of positive parent-
ing among the lowest ACE families can be viewed as
roughly equivalent to the impact of 4+ ACEs.
DISCUSSION
This study finds that, before the age of 6 years and as
early as 4 months, accumulated ACEs already manifest
signs of negative impact on social−emotional skills and
general development. More than one third of children
aged less than 6 years had already experienced at least
one of nine NSCH adversities. Given ACEs prevalence
and associated long-term societal costs,
48
the increased
attention and importance placed on childhood adversi-
ties seems well justified.
Fortunately, PPPs appear to mitigate negative effects of
adversities on these same outcomes and over this same
period of early development. The evidence presented sug-
gests the absence of PPPs can be viewed, itself, as another
adversity that at the extremes is equivalent to the addition
of four or more ACE score units. This finding, coupled
with the lack of evidence for effect modification, seems
Table 2. Proportions of Adverse Childhood Experiences (ACEs) and Positive Parenting Practices (PPPs) Among Young Children
Variable
Aged 0−2 years,
Weighted % (95% CI)
Aged 3−5 years,
Weighted % (95% CI)
Childhood adversity experiences
Hard to get by on current income 24.5 (22.9, 26.1) 26.1 (24.6, 27.8)
Parent divorced or separated 5.4 (4.6, 6.2) 14.0 (12.8, 15.4)
Lived with someone with drug or alcohol problem 3.5 (3.0, 4.2) 7.4 (6.5, 8.4)
Witnessed or was victim of neighborhood violence 1.4 (1.1, 1.9) 4.0 (3.4, 4.6)
Lived with someone who was mentally ill or suicidal 3.9 (3.2, 4.6) 7.1 (6.3, 8.0)
Witnessed domestic violence 2.2 (1.8, 2.7) 5.8 (5.0, 6.8)
Parent served time in jail 2.9 (2.4, 3.6) 5.9 (5.2, 6.7)
Targeted or judged unfairly due to race/ethnicity 0.6 (0.3, 1.2) 1.2 (0.9, 1.6)
Death of parent 0.6 (0.3, 1.2) 1.1 (0.9, 1.5)
Positive parenting practices (≥4 days/week)
Reading a book 65.7 (63.9, 67.5) 77.5 (75.9, 79.0)
Storytelling/Singing 83.9 (82.3, 85.3) 74.9 (73.3, 76.4)
Playing with peer 39.4 (37.6, 41.2) 75.1 (73.5, 76.5)
Family outing 52.9 (51.0, 54.7) 51.7 (49.9, 53.4)
Family meal 84.4 (83.0, 85.7) 83.8 (82.4, 85.1)
TV watching (≤2 hours/day) 87.9 (86.7, 89.0) 76.9 (75.4, 78.3)
Note: Weighted % was calculated using design variables (sampling weights and strata indicators).
ARTICLE IN PRESS
Yamaoka and Bard / Am J Prev Med 2018;000(000):1
−
10 5
&
2018

Table 3. Effects of Adverse Childhood Experiences (ACEs) and Positive Parenting Practices (PPPs) on Social−Emotional Defi-
cits and Developmental Delay Risks
Variable
Model 1,
OR (95% CI)
Model 2,
OR (95% CI)
Model 2b,
OR (95% CI)
Model 3,
OR (95% CI)
Social−emotional deficits
ACE score
ref: 0
11.52 (1.31, 1.77) 1.19 (1.01, 1.40) —1.18 (1.00, 1 .39)
2−31.60 (1.32, 1.94) 1.11 (0.90, 1.38) —1.10 (0.89, 1.36)
≥41.82 (1.37, 2.42) 1.35 (1.01, 1.80) —1.36 (1.02, 1.81)
Covariates
Age, years —1.08 (1.04, 1.12) 1.09 (1.05, 1.13) 1.09 (1.05, 1 .13)
Sex (ref: female)
Male —1.25 (1.10, 1 .42) 1.25 (1.10, 1 .42) 1.24 (1.09, 1 .42)
Race (ref: white, non-Hispanic)
Hispanic —1.07 (0.89, 1.29) 0.98 (0.81, 1.19) 0.99 (0.82, 1.20)
Black, non-Hispanic —1 .77 (1.48, 2.11) 1.65 (1.38, 1.97) 1.66 (1.38, 1 .98)
Other, multiracial —1.60 (1.31 , 1.96) 1.50 (1.23, 1.84) 1.51 (1.24, 1.85)
Parental education (ref: >high school)
High school graduate —1.34 (1.13, 1.59) 1.27 (1.07, 1.51) 1.27 (1.07, 1.50)
<High school —1.70 (1.33, 2.18) 1 .53 (1.18, 1 .98) 1.54 (1.19, 1.99)
Poverty status (ref: >400% FPL)
200%−399% FPL —1.17 (0.97, 1.40) 1.14 (0.95, 1.37) 1.12 (0.93, 1.34)
100%−199% FPL —1.46 (1.19, 1.79) 1.45 (1 .20, 1.77) 1.39 (1 .13, 1.70)
0−99% FPL —2.13 (1.71, 2.64) 2.11 (1.73, 2.59) 1.98 (1.60, 2.45)
No. of parenting practices (ref: 0−2)
3——0.74 (0.57, 0.97) 0.74 (0.57, 0.97)
4——0.70 (0.55, 0.90) 0.70 (0.55, 0.90)
5——0.52 (0.41, 0.67) 0.53 (0.41, 0.67)
6——0.49 (0.37, 0.64) 0.49 (0.37, 0.65)
Developmental delay risks
ACE score (ref: 0)
11.41 (1.22, 1 .63) 1.18 (1.01, 1.37) —1.17 (0.99, 1.36)
2−31.92 (1.58, 2.33) 1.44 (1.16, 1.78) —1.42 (1.15, 1.76)
≥42.64 (1.99, 3.51) 2.01 (1.48, 2.73) —2.04 (1.49, 2.80)
Covariates
Age, years —1.20 (1.16, 1.25) 1.23 (1 .19, 1.28) 1 .22 (1.17, 1.26)
Sex (ref: female)
Male —1.43 (1.26, 1 .62) 1.43 (1.26, 1.62) 1.43 (1 .26, 1.62)
Race (ref: white, non-Hispanic)
Hispanic —1.40 (1.18, 1.67) 1.27 (1.07, 1.51) 1.30 (1.09, 1.54)
Black, non-Hispanic —1 .31 (1.09, 1.58) 1.23 (1.02, 1.48) 1.23 (1.02, 1.48)
Other, multiracial —1.54 (1.27, 1.87) 1.45 (1.19, 1.76) 1.45 (1.19, 1.77)
Parental education (ref: >High school)
High school graduate —1.12 (0.94, 1.33) 1.06 (0.89, 1.26) 1.06 (0.89, 1.26)
<High school —1.48 (1.17, 1.87) 1.30 (1.03, 1.65) 1.33 (1.05, 1.69)
Poverty status (ref: >400% FPL)
200%−399% FPL —0.91 (0.77, 1.09) 0.92 (0.77, 1.10) 0.88 (0.73, 1.05)
100%−199% FPL —1.06 (0.87, 1.28) 1.10 (0.91, 1.33) 1.00 (0.82, 1.22)
0−99% FPL —1.39 (1.12, 1.73) 1.49 (1.21 , 1.83) 1.30 (1 .04, 1.61)
No. of parenting practices (ref: 0−2)
3——0.87 (0.67, 1.14) 0.87 (0.67, 1.13)
4——0.65 (0.51, 0.83) 0.65 (0.51, 0.82)
(continued on next page)
ARTICLE IN PRESS
6Yamaoka and Bard / Am J Prev Med 2018;000(000):1
−
10
www.ajpmonline.org

promising for prevention professionals. As the recent
National Academy of Sciences, Engineering, and Medicine
report Parenting Matters: Supporting Parents of Children
Ages 0
−
8
18
noted, “High-quality ‘serve and return’parent-
ing skills do not always develop spontaneously,”especially
among families living with adversities. So to promote
PPPs, policies that strengthen and fund evidence-based
parent training (e.g., home visiting) and parent resource
(e.g., Reach Out and Read) programs ought to remain at
the forefront of early childhood prevention efforts.
Table 3. Effects of Adverse Childhood Experiences (ACEs) and Positive Parenting Practices (PPPs) on Social−Emotional Defi-
cits and Developmental Delay Risks (continued)
Variable
Model 1,
OR (95% CI)
Model 2,
OR (95% CI)
Model 2b,
OR (95% CI)
Model 3,
OR (95% CI)
5——0.55 (0.43, 0.70) 0.55 (0.43, 0.69)
6——0.54 (0.41, 0.69) 0.53 (0.41, 0.69)
Note: Boldface indicates statistical significance (p<0.05); Model 1: ACEs, Model 2: ACEs + covariates, Model 2b: PPPs + covariates, Model 3:
ACEs + PPPs + covariates; Ref = referent value for OR calculations.
ACE, Adverse childhood experience; FPL, federal poverty level; PPP, positive parenting practices.
Social−Emotional Deficit Developmental Delay Risk
Model 2
Parenting Excluded
0−23456 Model 2
Parenting Excluded
0−23456
0.15
0.20
0.25
0.30
0.35
0.40
Positive Parenting Practices
Predicted Probability
ACE Score
0
1
2−3
4+
Model 2b
ACEs Excluded
Figure 1. Predicted probability comparisons across models.
Note: Symbols are proportional to population size. Solid circles reflect (Model 2b) predicted probabilities WITHOUT adjustment for ACEs. Disconnected
dots reflect (Model 2) predicted probabilities WITHOUT adjustment for positive parenting. Predictions were produced with covariates fixed to modal
(male, white not Hispanic, household educated beyond high school) or mean (2.53 years) values.
ACE, adverse childhood experience.
ARTICLE IN PRESS
Yamaoka and Bard / Am J Prev Med 2018;000(000):1
−
10 7
&
2018

The population impacts of PPPs are particularly worth
emphasizing because promotion of PPPs can be a simple,
feasible, and universal intervention. If studied relationships
are (directly or indirectly) causal, adoption of all PPPs
could reduce developmental risks for nearly 1 million chil-
dren aged less than 6 years nationwide. Practitioners and
policy makers would be wise to pay equitable attention to
both ACEs and the absence of positive parenting during
early childhood. Promotion of PPPs, not only as a buffer to
adversity but also as a generally effective intervention for
lowering risks of social−emotional and developmental dis-
abilities, seems to be a worthy public health message that
could be spread in all early childhood service settings.
Limitations
Given the cross-sectional nature of the NSCH, reverse or
reciprocal causation (where development induces ACEs
or PPPs) cannot be ruled out, and follow-up longitudinal
research examining mechanisms of change over time is
warranted. For the sake of interpretive clarity, simple
summed scores of dichotomized items were constructed
for key variables, and these changes could distort vari-
able relationships. To address this concern, sensitivity
analyses without these coarsened scoring approaches
were performed, and the results largely replicated the
general pattern of findings presented. Sensitivity analyses
also explored including individual PPP items (instead of
an aggregate score) and found no statistically significant
difference in AORs, which supports a common summa-
tive PPP effect for these outcomes. (Results available by
contacting corresponding author.) All key measures suf-
fered from limited scope, minimal psychometric sup-
port, or both. Although PEDS is a clinically validated
instrument,
40
the NSCH excluded direct assessment and
open-ended questions about concerns, and these differ-
ences likely affect accuracy (i.e., lower sensitivity/speci-
ficity). Similarly, the PPP measure only addressed
frequency of activities reported by a single caregiver
(whose sex and relationship to child were unavailable),
and thus excluded important aspects of interaction qual-
ity (e.g., caregiver warmth and responsiveness)
12
and
details of multi-caregiver involvement (e.g., value of
father engagement).
49
Although NSCH ACEs were
expanded to include life-course stressors
50
and measure-
ment validity support exists,
42
this measure likely under-
estimates adversity exposure as a result of social
desirability bias and omission of other important adver-
sities (e.g., child maltreatment).
51
Thus, for all measures
used, further research should examine broader construct
coverage and differential impact of construct facets (e.g.,
PPP quality versus quantity, deprivation versus threat
52
adversities). Finally, there was weak evidence of ACEs
moderation of the PPP effect on DDR, which deserves
closer inspection among a larger sample of high ACEs
children.
CONCLUSIONS
ACEs evidence noteworthy negative effects on social−
emotional skill and general development in early child-
hood; however, PPPs exhibit independent and in some
situations (social−emotional skills) larger protective
effects. These data support and champion sustaining
and furthering interventions that promote PPPs at home
for all children, but especially for families experiencing
high levels of adversity.
ACKNOWLEDGMENTS
Dr. Yamaoka is financially supported by the Nippon Foundation
International Fellowship program. Dr. Bard’s support of this
work was partly funded by the Maternal, Infant, and Early Child-
hood Home Visiting Grant Program by the Health Resources and
Services Administration (Grant Numbers: D89MC28275 and
X10MC29496) and the NIH, National Institute of General Medi-
cal Sciences, grant 2U54GM104938-06 (PI Judith James).
No financial disclosures were reported by the authors of this
paper.
SUPPLEMENTAL MATERIAL
Supplemental materials associated with this article can be
found in the online version at https://doi.org/10.1016/j.
amepre.2018.11.018.
REFERENCES
1. Brown DW, Anda RF, Tiemeier H, et al. Adverse childhood experien-
ces and the risk of premature mortality. Am J Prev Med. 2009;37
(5):389–396. https://doi.org/10.1016/j.amepre.2009.06.021.
2. Felitti VJ, Anda RF, Nordenberg D, et al. Relationship of childhood
abuse and household dysfunction to many of the leading causes of
death in adults: the Adverse Childhood Experiences (ACE) Study. Am
J Prev Med. 1998;14(4):245–258. https://doi.org/10.1016/S0749-3797
(98)00017-8.
3. Dube SR, Felitti VJ, Dong M, Chapman DP, Giles WH, Anda RF.
Childhood abuse, neglect, and household dysfunction and the risk of
illicit drug use: the Adverse Childhood Experiences Study. Pediatrics.
2003;111(3):564–572. https://doi.org/10.1542/peds.111.3.564.
4. Bj€
orkenstam E, Hjern A, Vinnerljung B. Adverse childhood experien-
ces and disability pension in early midlife: results from a Swedish
National Cohort Study. Eur J Public Health. 2017;27(3):472–477.
https://doi.org/10.1093/eurpub/ckw233.
5. National Scientific Council on the Developing Child. Excessive
stress disrupts the architecture of the developing brain: Working
Paper No. 3. www.developingchild.harvard.edu.UpdatedEdition.
2005/2014. Accessed 3 May 2017.
6. Lynch BA, Agunwamba A, Wilson PM, et al. Adverse family experien-
ces and obesity in children and adolescents in the United States. Prev
Med. 2016;90:148–154. https://doi.org/10.1016/j.ypmed.2016.06.035.
7. MerskyJP,TopitzesJ,ReynoldsAJ.Impactsofadversechildhoodexperi-
ences on health, mental health, and substance use in early adulthood: a
ARTICLE IN PRESS
8Yamaoka and Bard / Am J Prev Med 2018;000(000):1
−
10
www.ajpmonline.org

cohort study of an urban, minority sample in the U.S. Child Abuse Negl.
2013;37(11):917–925. https://doi.org/10.1016/j.chiabu.2013.07.011.
8. Bright MA, Knapp C, Hinojosa MS, Alford S, Bonner B. The comorbid-
ity of physical, mental, and developmental conditions associated with
childhood adversity: a population based study. Matern Child Health J.
2016;20(4):843–853. https://doi.org/10.1007/s10995-015-1915-7.
9. Stevens GD. Gradients in the health status and developmental risks of
young children: the combined influences of multiple social risk factors.
Matern Child Health J. 2006;10(2):187. https://doi.org/10.1007/
s10995-005-0062-y.
10. Bethell CD, Newacheck PW, Fine A, et al. Optimizing health and
health care systems for children with special health care needs using
the life course perspective. Matern Child Health J. 2014;18(2):467–
477. https://doi.org/10.1007/s10995-013-1371-1.
11. Bethell C, Gombojav N, Solloway M, Wissow L. Adverse childhood
experiences, resilience and mindfulness-based approaches: common
denominator issues for children with emotional, mental, or behavioral
problems. Child Adolesc Psychiatr Clin N Am. 2016;25(2):139–156.
https://doi.org/10.1016/j.chc.2015.12.001.
12. National Scientific Council on the Developing Child. Supportive
relationships and active skill-building strengthen the foundations
of resilience: Working Paper 13. www.developingchild.harvard.
edu. Published 2015. Accessed May 3, 2017.
13. Garner AS, Shonkoff JP. Early childhood adversity, toxic stress,
and the role of the pediatrician: translating developmental science
into lifelong health. Pediatrics. 2012;129(1):e224–e231. https://doi.
org/10.1542/peds.2011-2662.
14. Keyes CL. The mental health continuum: from languishing to flourish-
ing in life. J Health Soc Behav. 2002;43(2):207–222. https://doi.org/
10.2307/3090197.
15. Centers for Disease Control and Prevention (CDC). Essentials for
childhood: steps to create safe, stable, nurturing relationships and
environments for all children. www.cdc.gov/violenceprevention/child-
maltreatment/essentials.html. Published 2014. Accessed August 1,
2017.
16. Browne HC. The Strengthening Families Approach and Protective
Factors Framework: Branching out and reaching deeper. https://cssp.
org/resource/the-strengthening-families-approach-and-protective-fac-
tors-framework-branching-out-and-reaching-deeper/. Published 2014.
Accessed August 1, 2017.
17. Traub F, Boynton-Jarrett R. Modifiable resilience factors to childhood
adversity for clinical pediatric practice. Pediatrics. 2017;139(5):
e20162569. https://doi.org/10.1542/peds.2016-2569.
18. National Academies of Sciences, Engineering, and Medicine. Parent-
ing matters: supporting parents of children ages 0−8. Washington,
DC: National Academies Press; 2016. https://doi.org/10.17226/21868.
19. Shah R, Sobotka SA, Chen YF, Msall ME. Positive parenting practices,
health disparities, and developmental progress. Pediatrics. 2015;136
(2):318–326. https://doi.org/10.1542/peds.2014-3390.
20. Cprek SE, Williams CM, Asaolu I, Alexander LA, Vanderpool RC.
Three positive parenting practices and their correlation with risk of
childhood developmental, social, or behavioral delays: an analysis of
the National Survey of Children’s Health. Matern Child Health J.
2015;19(11):2403–2411. https://doi.org/10.1007/s10995-015-1759-1.
21. Greenland S, Drescher K. Maximum likelihood estimation of the
attributable fraction from logistic models. Biometrics. 1993;49(3):865–
872. https://doi.org/10.2307/2532206.
22. Afifi TO,EnnsMW,CoxBJ,AsmundsonGJ,SteinMB,SareenJ.Popula-
tion attributable fractions of psychiatric disorders and suicide ideation
and attempts associated with adverse childhood experiences. Am J Public
Health. 2008;98(5):946–952. https://doi.org/10.2105/AJPH.2007.120253.
23. Roos LE, Mota N, Afifi TO, Katz LY, Distasio J, Sareen J. Relationship
between adverse childhood experiences and homelessness and the
impact of axis I and II disorders. Am J Public Health. 2013;103(suppl
2):S275–S281. https://doi.org/10.2105/AJPH.2013.301323.
24. Bramlett MD, Blumberg SJ, Zablotsky B, et al. Design and operation of
the National Survey of Children’s Health, 2011−2012. Vital Health
Stat. 2017;1(59):1–256.
25. Centers for Disease Control and Prevention (CDC), National Center
for Health Statistics, State and Local Area Integrated Telephone Sur-
vey. 2011−2012 National Survey of Children’s Health Frequently
Asked Questions. www.cdc.gov/nchs/slaits/nsch.htm. Published 2013.
Accessed December 10, 2017.
26. Woolfenden S, Eapen V, Williams K, Hayen A, Spencer N, Kemp L. A sys-
tematic review of the prevalence of parental concerns measured by the
Parents’Evaluation of Developmental Status (PEDS) indicating developmen-
tal risk. BMC Pediatr. 2014;14:231. https://doi.org/10.1186/1471-2431-14-231.
27. Child and Adolescent Health Measurement Initiative (CAHMI), Data
Resource Center for Child and Adolescent Health. 2011−2012
National Survey of Children’s Health (NSCH): Child Health Indicator
and Subgroups SAS Codebook, Version 1.0. www.childhealthdata.org.
Published 2013. Accessed May 4, 2017.
28. Coker TR, Shaikh Y, Chung PJ. Parent-reported quality of preventive
care for children at-risk for developmental delay. Acad Pediatr.
2012;12(5):384–390. https://doi.org/10.1016/j.acap.2012.05.003.
29. Moore K, Bethell C, Murphy D, Martin M, Beltz M. Flourishing
from the start. www.childtrends.org/publications/flourishing-start-
can-measured. Published 2017. Accessed October 10, 2018.
30. Bethell C, Davis M, Gombojav N, Stumbo S, Powers K. Issue Brief: A
national and across state profile on adverse childhood experiences
among children and possibilities to heal and thrive. www.cahmi.org/
wp-content/uploads/2018/05/aces_brief_final.pdf. Published October
2017. Accessed October 10, 2018.
31. Moore KA, Murphey D, Beltz M, Martin MC, Bartlett J, Caal S. Child
well-being: constructs to measure child well-being and risk and pro-
tective factors that affect the development of young children. www.
childtrends.org/publications/child-well-constructs-measure-child-
well-risk-protective-factors-affect-development-young-children. Pub-
lished 2016. Accessed July 18, 2018.
32. Moore KA, Murphey D, Bandy T. Positive child well-being: an index
based on data for individual children. Matern Child Health J. 2012;16
(suppl 1):S119–S128. https://doi.org/10.1007/s10995-012-1001-3.
33. Squires J, Bricker D, Twombly E. The ASQ:SE user’s guide: for the Ages
& Stages Questionnaires: social-emotional. Baltimore, MD: Paul H
Brookes Publishing, 2002.
34. Bai S, Repetti RL. Short-term resilience processes in the family. Fam
Relat. 2015;64(1):108–119. https://doi.org/10.1111/fare.12101.
35. Kautz T, Heckman JJ, Diris R, Weel B, Borghans L. Fostering and
measuring skills: improving cognitive and non-cognitive skills to pro-
mote lifetime success. NBER Work Pap Ser. 2014: 20749. https://www.
nber.org/papers/w20749.pdf.
36. Kandasamy V, Hirai AH, Ghandour RM, Kogan MD. Parental percep-
tion of flourishing in school-aged children: 2011−2012 National Sur-
vey of Children’s Health. J Dev Behav Pediatr. 2018;39(6):497–507.
https://doi.org/10.1097/DBP.0000000000000559.
37. Ruest S, Gjelsvik A, Rubinstein M, Amanullah S. The inverse relationship
between digital media exposure and childhood flourishing. JPediatr.
2018;197:268–274.e2. https://doi.org/10.1016/j.jpeds.2017.12.016.
38. Nabors LA, Meerianos AL, Vidourek RA, et al. Predictors of flourish-
ing for adolescents with asthma. J Asthma. 2016;53(2):146–154.
https://doi.org/10.3109/02770903.2015.1072722.
39. Odar Stough C, Nabors L, Merianos A, Zhang J. Short communica-
tion: flourishing among adolescents with epilepsy: correlates and com-
parison to peers. Epilepsy Res. 2015;117:7–10. https://doi.org/10.1016/
j.eplepsyres.2015.08.004.
40. Glascoe FP. Using parents’concerns to detect and address develop-
mental and behavioral problems. J Soc Pediatr Nurs. 1999;4(1):24–35.
https://doi.org/10.1111/j.1744-6155.1999.tb00077.x.
41. Limbos MM, Joyce DP. Comparison of the ASQ and PEDS in screen-
ing for developmental delay in children presenting for primary care.
&
2018
Yamaoka and Bard / Am J Prev Med 2018;000(000):1
−
10 9
ARTICLE IN PRESS

J Dev Behav Pediatr. 2011;32(7):499–511. https://doi.org/10.1097/
DBP.0b013e31822552e9.
42. Bethell CD, Carle A, Hudziak J, et al. Methods to assess adverse child-
hood experiences of children and families: toward approaches to pro-
mote child well-being in policy and practice. Acad Pediatr. 2017;
17(7 suppl):S51–S69. https://doi.org/10.1016/j.acap.2017.04.161.
43. The Child and Adolescent Health Measurement Initiative, Johns
Hopkins Bloomberg School of Public Health. Experience matters: a
view into the health and wellbeing of US children and families
with Autism Spectrum Disorder (ASD). http://childhealthdata.org/
docs/default-source/cahmi/asdchartbookfinal.pdf?sfvrsn=2. Published
2016. Accessed October 10, 2018.
44. Grace-Martin K. Confusing statistical term #4: Hierarchical regression
vs. hierarchical model. www.theanalysisfactor.com/confusing-statisti-
cal-term-4-hierarchical-regression-vs-hierarchical-model/. Published
2009. Accessed October 10, 2018.
45. Cohen J, Cohen P, West SG, Aiken LS. Applied Multiple Regression/
Correlation Analysis for the Behavioral Sciences. 3rd ed. Mahwah, NJ:
Lawrence Erlbaum Associates Publishers, 2003.
46. StataCorp L. Stata survey data reference manual, release 14. https://
www.surveydesign.com.au/docs/manuals/stata14/svy.pdf. Published
2015.
47. R Core Team. R: A language and environment for statistical comput-
ing. www.R-project.org/. Published 2017. Accessed November 9, 2018.
48. Caspi A, Houts RM, Belsky DW, et al. Childhood forecasting of
a small segment of the population with large economic burden.
Nat Hum Behav. 2016;1:0005. https://doi.org/10.1038/s41562-
016-0005.
49. Sarkadi A, Kristiansson R, Oberklaid F, Bremberg S. Fathers’involve-
ment and children’s developmental outcomes: a systematic review of
longitudinal studies. Acta Paediatr. 2008;97(2):153–158. https://doi.
org/10.1111/j.1651-2227.2007.00572.x.
50. 2011/12 National Survey of Children’s Health, Child and Adolescent
Health Measurement Initiative (CAHMI). 2011−2012 NSCH: Child
Health Indicators and Subgroups SAS Codebook, Version 1.0. www.
childhealthdata.org/docs/nsch-docs/sas-codebook_-2011-2012-nsch-
v1_05-10-13.pdf. Published 2013. Accessed May 4, 2017.
51. Finkelhor D, Shattuck A, Turner H, Hamby S. A revised inventory of
adverse childhood experiences. Child Abuse Negl. 2015;48:13–21.
https://doi.org/10.1016/j.chiabu.2015.07.011.
52. McLaughlin KA, Sheridan MA, Lambert HK. Childhood adversity and
neural development: deprivation and threat as distinct dimensions of
early experience. Neurosci Biobehav Rev. 2014;47:578–591. https://doi.
org/10.1016/j.neubiorev.2014.10.012.
ARTICLE IN PRESS
10 Yamaoka and Bard / Am J Prev Med 2018;000(000):1
−
10
www.ajpmonline.org