Adverse Childhood Exp
Child Health
Emalee G. Flaherty, MD; Richard Thom
Adam J. Zolotor, MD, MPH; H
Desmond K. Runyan, MD, DrPH; Diana J
Objective.—The relationship between adverse childhood expo-
sures and poor health, illness, and somatic complaints at age 12
was examined.
Methods.—LONGSCAN (Consortium for Longitudinal Studies
of Child Abuse and Neglect) tracks a group of children with vari-
able risk for maltreatment. Of the participating child-caregiver
sample of children found a link between adversities andversities occurring in the first 6 years of life (early child-
hood) and those occurring in the second 6 years (middle
childhood) on the emergence of health outcomes in early
adolescence. We hypothesized that: (1) as children age, ad-
verse experiences would have a greater impact on their
health than was found in the earlier research on younger
children; (2) there are significant links between earlier
childhood adversities and reported health, serious illness,
and somatic complaints at age 12; and (3) adversities oc-
curring in the second 6 years of life more strongly predict
these reported health outcomes.
and Joint Doctoral Program in Clinical Psychology, San Diego State
University and University of California at San Diego, San Diego, Calif
(Dr Litrownik); Department of Family Medicine, University of North
Carolina, Chapel Hill, NC (Dr Zolotor); Department of Pediatrics,
University of Maryland, Baltimore, Md (Dr Dubowitz); Department of
Social Medicine, Pediatrics, University of North Carolina, Chapel Hill,
NC (Dr Runyan); School of Social Work, University of Washington,
Seattle, Wash (Dr English); and Department of Psychiatry, University of
North Carolina, Chapel Hill, NC (Dr Everson).
Address correspondence to Emalee G. Flaherty, MD, Children’s Memo-
rial Hospital, 2300 Children’s Plaza, Box 16, Chicago, Illinois 60614
(e-mail: e-flaherty@northwestern.edu).
Received for publication August 7, 2008; accepted November 7, 2008.Tadults who experienced adverse childhood experi-
ences were more likely to rate their health as poor and to
have health problems in adulthood such as ischemic heart
disease, cancer, chronic lung disease, skeletal fracture,
obesity, and liver disease.
1–4
Several studies have also
shown a relationship between child maltreatment and adult
somatic complaints.
5,6
The postulated link between childhood adverse experi-
ences and health outcomes has not been examined longitu-
dinally, however. One previous study examining a high-risk
From the Department of Pediatrics, Children’s Memorial Hospital and
Northwestern University’s Feinberg School of Medicine, Chicago,
Ill (Dr Flaherty); Department of Research, Juvenile Protective Association,
Chicago, Ill (Dr Thompson); Department of Psychology, San Diego StateACADEMIC PEDIATRICS
Copyright  2009 by Academic Pediatric Association 150adult studies.
7
These findings, along with the limitations
of the earlier research, suggested the need to examine the
impact of adversities on health over time as children
mature.
The timing of adversities may be important. Research
on maltreatment has suggested that maltreatment occur-
ring across several different developmental periods has
a much more profound impact on child behavioral
outcomes, particularly predictors of socialization adapta-
tion, than does maltreatment occurring in only one
period.
8
In addition, the timing of childhood maltreat-
ment has been shown to influence the appearance of ag-
gression in later childhood and mental health outcomes
in later adulthood.
9,10
The current study examines the impact of childhood ad-and other recent research has demonstrated that poor health at age 6, but without the dose effect found indyads, 805 completed an interview when the child was age 4 or
age 6, as well as interviews at age 8 and 12. The relationships
between 8 categories of childhood adversity (psychological mal-
treatment, physical abuse, sexual abuse, child neglect, caregiver’s
substance/alcohol use, caregiver’s depressive symptoms, care-
giver’s being treated violently, and criminal behavior in the
household) and child health at age 12 were analyzed. The impact
of adversity in the first 6 years of life and adversity in the second 6
years of life on child health were compared.
Results.—Only 10% of the children had experienced no adversity,
while more than 20% had experienced 5 or more types of child-
he Adverse Childhood Experiences (ACE) Studyosures and Reported
at Age 12
pson, PhD; Alan J. Litrownik, PhD;
oward Dubowitz, MD, MS;
. English, PhD; Mark D. Everson, PhD
hood adversity. At age 12, 37% of the children sampled had
some health complaint. Exposure to 5 or more adversities, partic-
ularly exposure in the second 6 years of life, was significantly as-
sociated with increased risks of any health complaint (odds ratio
[OR] 2.24, 95% confidence interval [95% CI] 1.02–4.96), an
illness requiring a doctor (OR 3.69, 95% CI 1.02–15.1), and
caregivers’ reports of child’s somatic complaints (OR 3.37, 95%
CI 1.14–1.0). There was no association between adverse expo-
sures and self-rated poor health or self-rated somatic complaints.
Conclusions.—A comprehensive assessment of children’s health
should include a careful history of their past exposure to adverse
conditions and maltreatment. Interventions aimed at reducing
these exposures may result in better child health.
KEY WORDS: adverse childhood experiences; child abuse;
child neglect; health outcomes; longitudinal studies
Academic Pediatrics 2009;9:150–6Volume 9, Number 3
May–June 2009

giver’s marital status and family income), information col-
lected at the age 12 interview was used. To increase power,
the number of categories of each variable was reduced. Spe-
cifically, child’s race/ethnicity was divided into white, Afri-
can American, or other. Caregiver’s marital status was
divided into married, never married, or formerly married.
Family income was dichotomized into above federal pov-
erty level or at or below federal poverty level.
Adverse Exposures
We examined the pool of variables from the ages 4, 6, 8,
and 12 interviews to identify indicators of adverse child-
hood experiences that corresponded with 8 categories of
childhood adversity analogous to those described in the
ACE Study with the addition of child neglect. Four indices
of maltreatment (psychological maltreatment, physical
abuse, sexual abuse, and neglect) and 4 measures of house-
hold dysfunction (caregiver’s substance use/alcohol abuse,
caregiver’s depressive symptoms, caregiver being treated
violently, and criminal behavior in household) were identi-
fied. Where possible, an indicator for each of the measures
of adversity was specified. These assessment periods wereMETHODS
Participants and Study Design
Data used in these analyses were collected by the Consor-
tium for Longitudinal Studies of Child Abuse and Neglect
(LONGSCAN). LONGSCAN isa consortiumofa coordinat-
ing center and 5 study sites investigating prospectively the
antecedents and consequences of child maltreatment.
11
The study sites represent different geographical regions
and populations with different levels of risk for maltreat-
ment. The Southwest site includes children removed from
their homes by Child Protective Services (CPS) and placed
in foster care, while the Northwest site includes children
reported to CPS judged to be at moderate risk for maltreat-
ment. The Midwest site includes both families reported to
CPS and matched neighborhood controls who were not ini-
tially reported to CPS. The East site recruited low-income
at risk children from their clinics. The Southeast site re-
cruited children at birth considered at risk because of birth
or sociodemographic problems.
12
All sites collect data ac-
cording to the commonly shared protocols at age-specific
data collection points.
These analyses include data on participants who had
completed interviews when the child was either age 4 or
age 6 and who had also completed both the interviews con-
ducted when children were age 8 and 12. Of the 1354 chil-
dren enrolled onto the LONGSCAN studies at baseline (age
4, age 6, or both), 805 (59%) had completed both the age 8
and age 12 interviews. This decrease in numbers was pri-
marily due to attrition; specifically, participating families
moved frequently, and some were not responsive to tele-
phone and mail requests for interviews. In addition, several
children in the Midwest site had not yet aged into the age 12
interview at the time of these analyses. Comparison of de-
mographic characteristics revealed no differences between
those included in the analyses and those not included. The
only significant difference in early adversity between the
2 groups was that those who were included were signifi-
cantly more likely to have been reported as experiencing
psychological abuse (26.6% vs 21.9%, c
2
¼ 3.92, P¼ .04).
The demographic description of the sample is presented
in Table 1. More than 40% of the children were living in
poverty, and approximately one-third had mothers who
had never been married. The biological mother was the pri-
mary caregiver for approximately 75% of the sample, de-
pending on the age assessed. The remainder included
some children in foster care and some being raised by other
relatives.
Human Subjects
Each site and the coordinating center obtained approval
from their local institutional review board. Caregivers pro-
vided informed consent, and children assented to participate.
Variables and Their Measurement
Age-appropriate measures were selected from among
available standard published instruments, as discussed be-
ACADEMIC PEDIATRICSlow. Some measures asked caregivers about events in the
prior year, others in the prior 6 months. Analysis includeddata collected during assessment interviews when the chil-
dren were aged 4, 6, 8, and 12. For several predictor vari-
ables, different measures were used to assess the variable
at different ages. We dichotomized each predictor and out-
come variable to construct a risk profile, unless otherwise
specified as detailed below.
11–14
Demographic Control Variables
Demographic variables were part of the assessment at
each age. Time-invariant demographic variables (child’s
race/ethnicity, gender, and study site) were collected at
age 4 or 6. For variables that could vary over time (eg, care-
Table 1. Demographic Characteristics (N ¼ 805)
Characteristic n (%)*
Gender
Male 399 (50)
Female 406 (50)
Site
Eastern 153 (19)
Southern 147 (18)
Midwestern 113 (14)
Northwestern 175 (22)
Southwestern 217 (27)
Race
White 215 (27)
African American 431 (54)
Other 159 (20)
Poverty
Yes 316 (41)
No 442 (59)
Mother marital status
Never married 269 (34)
Married 314 (39)
Formerly married 218 (27)
*Numbers vary slightly as a result of missing data.
Adverse Childhood Exposures and Health at Age 12 151categorized as adversities occurring in the first 6 years of
the child’s life (assessed at 4 and 6 years) or occurring in