Download Report

Child Abuse & Neglect 37 (2013) 14–21
Contents lists available at SciVerse ScienceDirect
Child Abuse & Neglect
Child physical and sexual abuse in a community sample of young
adults: Results from the Ontario Child Health Study夽,夽夽
Harriet L. MacMillan a,b,∗ , Masako Tanaka a , Eric Duku a , Tracy Vaillancourt a,c ,
Michael H. Boyle a
a
Department of Psychiatry and Behavioural Neurosciences, McMaster University, 1280 Main Street West, Hamilton, Ontario, Canada L8S 4K1
Department of Pediatrics, McMaster University, 1280 Main Street West, Hamilton, Ontario, Canada L8S 4K1
Faculty of Education and School of Psychology, Faculty of Social Sciences, University of Ottawa, Lamoureux Hall (LMX) 145, Jean-Jacques-Lussier Private,
Ottawa, Ontario, Canada K1N 6N5
b
c
a r t i c l e
i n f o
Article history:
Received 5 May 2012
Received in revised form 16 June 2012
Accepted 21 June 2012
Available online 3 January 2013
Keywords:
Child physical abuse
Child sexual abuse
Prevalence
Risk
a b s t r a c t
Objectives: Exposure to child maltreatment is associated with physical, emotional, and
social impairment, yet in Canada there is a paucity of community-based information about
the extent of this problem and its determinants. We examined the prevalence of child
physical and sexual abuse and the associations of child abuse with early contextual, family,
and individual factors using a community-based sample in Ontario.
Methods: The Ontario Child Health Study is a province-wide health survey of children aged
4 through 16 years. Conducted in 1983, a second wave was undertaken in 1987 and a third
in 2000–2001. The third wave (N = 1,928) included questions about exposure to physical
and sexual abuse in childhood.
Results: Males reported signiﬁcantly more child physical abuse (33.7%), but not severe
physical abuse (21.5%), than females (28.2% and 18.3%, respectively). Females reported signiﬁcantly more child sexual abuse (22.1%) than males (8.3%). Growing up in an urban area,
young maternal age at the time of the ﬁrst child’s birth, and living in poverty, predicted child
physical abuse (and the severe category), and sexual abuse. Childhood psychiatric disorder
was associated with child physical abuse (and the severe category), while parental adversity was associated with child sexual abuse and severe physical abuse. Siblings of those
who experienced either physical abuse or sexual abuse in childhood were at increased risk
for the same abuse exposure; the risk was highest for physical abuse.
Conclusions: These ﬁndings highlight important similarities and differences in risk factors for physical and sexual abuse in childhood. Such information is useful in considering
approaches to prevention and early detection of child maltreatment. Clinicians who identify physical abuse or sexual abuse in children should be alert to the need to assess whether
siblings have experienced similar exposures. This has important implications for assessment of other children in the home at the time of identiﬁcation with the overall goal of
reducing further occurrence of abuse.
© 2012 Elsevier Ltd. All rights reserved.
夽 This work was carried out at the Offord Centre for Child Studies, McMaster University.
夽夽 This study was funded by a grant from the Canadian Institutes of Health Research (CIHR). The research was also supported by the CIHR Institutes of
Gender and Health; Aging; Human Development, Child and Youth Health; Neurosciences, Mental Health and Addiction; and Population and Public Health.
Harriet MacMillan is supported by the David R. (Dan) Offord Chair in Child Studies. Michael Boyle is supported by a Canada Research Chair in the Social
Determinants of Child Health.
∗ Corresponding author at: Offord Centre for Child Studies, Department of Psychiatry and Behavioural Neurosciences, McMaster University, 1280 Main
Street West, Hamilton, Ontario, Canada L8S 4K1.
0145-2134/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.chiabu.2012.06.005
H.L. MacMillan et al. / Child Abuse & Neglect 37 (2013) 14–21
15
Introduction
Despite increasing recognition that child maltreatment is a major public health problem, there continues to be relatively little community-based information available about the distribution and determinants of child abuse in Canada. The
Canadian Incidence Study (CIS) provides important data about ofﬁcial reports of maltreatment (PHAC, 2010; Trocmé et al.,
2005), but only a small proportion of child abuse victims come to the attention of child protection agencies (MacMillan,
Jamieson, & Walsh, 2003). The Ontario Health Supplement, a 1990–1991 province-wide survey, examined the prevalence
and correlates of mental health in a probability sample of residents 15 years of age and older and included questions about
exposure to child physical and sexual abuse (Boyle et al., 1996; MacMillan et al., 1997; Offord et al., 1996). Although this
information served to quantify the extent of the problem, it was collected retrospectively from a predominantly adult population at a single point in time. Furthermore, the Supplement was conducted more than 20 years ago. It is important to
examine more recent information about the prevalence of maltreatment, and predictors of its occurrence from longitudinal
studies.
Based on the ecological approach, risk factors for abuse can be divided into four categories: demographic, familial,
parental, and child (Belsky & Vondra, 1989). Of the longitudinal studies examining these factors, three involve general
population samples. In two, exposure to child maltreatment was measured through retrospective self-reports and ofﬁcial
agency reports, and in one study it was measured by retrospective self-report. The ﬁrst study involved 644 families who were
part of a larger sample of families randomly selected in 1975 from two upstate New York counties who were re-interviewed
in 1983, 1986, and 1991–1993. Brown, Cohen, Johnson, and Salzinger (1998) identiﬁed variables linked to physical abuse,
including low religious attendance, poor marital quality, single parent status, welfare, low parental involvement and other
maternal factors such as serious illness, low education, youth and perinatal complications. The factors associated with child
sexual abuse included being female, presence of disability in a child, harsh punishment, negative life events experienced by
a child, parental death, presence of a stepfather, maternal youth, maternal sociopathy, and unwanted pregnancy. The second
study, the Christchurch Health and Development Study, involved a birth cohort of 1,265 children in the Christchurch, New
Zealand region recruited during mid-1977 and followed at four months and then annually until age 16 and then at 18 years
followed by further assessments in adulthood. The following risk factors for child sexual abuse were identiﬁed: being female,
marital conﬂict, low parental attachment, a high level of overprotection, and parental alcohol problems (Fergusson, Lynskey,
& Horwood, 1996). The third study, the National Longitudinal Study of Adolescent Health (Add Health) followed by a US
national sample of 20,745 adolescents (grades 7 through 12) into young adulthood (ages 18–26 years). Hussey, Chang, and
Kotch (2006) identiﬁed the associations between lower parental education and physical neglect and physical assault, and
the associations between lower family income and neglect and contact sexual abuse. In this study, child maltreatment experienced by the start of the 6th grade was measured. In general, risk factors for physical abuse include those associated with
psychosocial disadvantage while risk factors for sexual abuse in childhood included family dysfunction, parental absence,
and/or economic disadvantage.
The Ontario Child Health Study (OCHS), a province-wide longitudinal study, included self-report questions regarding
childhood abuse in the third wave when the original participants were young adults (Boyle et al., 1987; Boyle, Georgiades,
Racine, & Mustard, 2007). The OCHS provides data about exposure to child abuse within a time frame closer to its occurrence
than the earlier Ontario Health Supplement for most participants (more than 70% of Supplement respondents were over age
35 at the time of the survey), and prospective information on correlates. Additionally, the OCHS offers the opportunity to
study childhood abuse within families. This study summarizes results from the OCHS regarding the prevalence of and risk
factors for physical and sexual abuse in childhood.
Methods
The OCHS was conducted in 1983 to investigate the distribution and determinants of health status in children aged 4–16
years (Offord et al., 1987); a second wave in 1987 assessed continuity and change in health status (Offord et al., 1992); and
a third wave in 2000–2001 investigated functional outcomes in young adulthood. Information was collected about children
who were 4–16 years of age in 1983 for the ﬁrst wave, and about the same children when they were four years older in 1987
for the second wave. By the time of the third wave, the children from the original sample were adults between the ages of
21–35 years. The target population in 1983 included all children born from January 1966 through January 1979 who resided in
a household dwelling in the province of Ontario. The sampling frame was the 1981 Census; selection was made by stratiﬁed,
clustered and random sampling. In 1983 and 1987, the “person most knowledgeable” (usually the mother) reported on the
child. In 2000–2001, the children, now young adults, were respondents, and attempts were made to include all children who
participated in 1983. Tracing, enlistment and data collection were completed by Statistics Canada. Participants gave verbal
consent prior to the interview, and written consent to share the data with researchers at McMaster University. The study
was approved by the Hamilton Health Sciences/McMaster University Research Ethics Board.
Participants
In the 1983 OCHS, there were 2,052 households with eligible children; 1,869 (91.1%) of these households including 3,294
children participated in the original study. Of these 3,294 children, 2,355 (71.5%) provided some information to the study
16
H.L. MacMillan et al. / Child Abuse & Neglect 37 (2013) 14–21
Items
Minimum required frequency for:
Physical abuse
1. How many times before age 16 did an adult slap you
Severe physical abuse
3-5 times
>10 times
3-5 times
>10 times
1-2 times
1-2 times
on the face, head or ears or hit or spank you with
something like a belt, wooden spoon or something hard?
2. Before age 16 did an adult push, grab, shove or throw
something at you to hurt you?
3. Before age 16 how many times did an adult kick, bite,
punch, choke, burn you, or physically attack you in some
way?
Sexual abuse
1.Before age 16 when you were growing up, did anyone
1-2 times
ever do any of the following things when you didn't want
them to: touch the private parts of your body or make
you touch their private parts, threaten or try to have sex
with you or sexually force themselves on you?
Note. Endorsing one or more items at or above the minimum required frequency was deemed abuse.
Fig. 1. Deﬁnition of physical and severe physical abuse, and sexual abuse for the short form of the Childhood Experiences of Violence Questionnaire.
in 2000–2001, while a sub-group of 1,928 (58.5%) were full participants. Fourteen key variables from 1983 were used to
model the effects of attrition in both groups: these variables represented child health status, health service use and child
functioning, parental health, and family structure, functioning and disadvantage. Several of these variables were signiﬁcantly
related to attrition; with the two strongest being rental housing and the child’s mental health/social service use. Attrition
weights were developed separately for each group. When these attrition weights were applied to each group (N = 2,355 and
N = 1,928), the original 1983 sample characteristics (N = 3,294) were re-captured accurately (Boyle et al., 2006, 2007).
Measures
Childhood physical and sexual abuse. Exposure to childhood physical abuse and sexual abuse before the age of 16 years was
measured retrospectively in 2000-2001 with a modiﬁed version (Tanaka et al., 2012) of the Childhood Experiences of Violence
Questionnaire (CEVQ; Walsh, MacMillan, Trocmé, Jamieson, & Boyle, 2008). In a sample of 179 adolescents aged 12–18 years
from clinical and community settings, the original CEVQ showed good two-week test–retest reliability, with kappas of 0.85,
0.77, 0.92 and 0.87 for physical abuse, severe physical abuse, sexual abuse, and severe sexual abuse, respectively, and good
criterion validity when measured against clinical assessment, with kappas of 0.67 and 0.70 for physical and sexual abuse,
respectively. The CEVQ has demonstrated content and construct validity (Walsh et al., 2008).
To reduce respondent burden, the CEVQ was shortened for the OCHS by the authors (HM, MB), mainly by combining items
from the original; for example, the two items “kicked, bit, or punched you to hurt you?” and “choked, burned or physically
attacked you in some other way?” became one: “kick, bite, punch, choke, burn you, or physically attack you in some way?”.
Similarly, the one sexual abuse item is three items in the original CEVQ. As in the original, the response categories were
“never”, “1 or 2 times”, “3 to 5 times”, “6 to 10 times” and “more than 10 times”. Fig. 1 shows the items in the short version
and the frequency thresholds for severe and non-severe physical abuse and sexual abuse (with no severity classiﬁcation
based on one item that includes several abusive acts varying in severity). The shorter form of the CEVQ is comparable to
the original CEVQ in the two-week test–retest reliability, criterion validity, and construct validity (Tanaka et al., 2012). The
CEVQ was self-completed; most of the survey was administered face-to-face.
H.L. MacMillan et al. / Child Abuse & Neglect 37 (2013) 14–21
17
Sociodemographic status. The socioeconomic status (SES) of each child’s family was determined from 1983 data, based on a
composite measure derived from ﬁve variables which were standardized, added together and transformed to have a mean
of zero and a standard deviation of one. These variables were the natural log of household income, years of education of the
mother and her spouse/partner (if present in the home), and the prestige associated with the occupation of the mother and
her spouse/partner (if present in the home and working) measured by the Blishen scale (Blishen, Carroll, & Moore, 1987).
A variable classifying family income below Statistics Canada low income cut-offs (taking into account family size and area
of residence) was included to represent extreme economic disadvantage. Urban/rural residence was also taken from 1983
data; urban was deﬁned as a census area with a population over 25,000.
Family characteristics. Other 1983 variables examined included parental and family characteristics: young mother status
(20 years or younger) at time of ﬁrst child’s birth; parental adversity which included being hospitalized for nerves and/or
ever being arrested; family dysfunction, measured by maternal report on a twelve-item general dysfunction subscale of the
McMaster Family Assessment Device (FAD; Byles, Byrne, Boyle, & Offord, 1988; Epstein, Baldwin, & Bishop, 1983). The scale
includes 12 statements that describe family behavior and relationships in each of the dimensions, with response options
that include the following: strongly agree (=1), agree (=2), disagree (=3), strongly disagree (=4). After recoding negativelyoriented items, a total score was obtained, which could range from 12 to 48, with lower scores indicating better functioning.
For this analysis, responses were dichotomized as follows: at risk for family dysfunction (score of 27 and above) and not at
risk (value below 27).
Child characteristics. Child characteristics included presence of psychiatric disorder (deﬁned as one or more of conduct
disorder, emotional disorder and attention deﬁcit disorder), based on the OCHS scales parent report (Boyle et al., 1987,
1993) and functional limitation as indicated by the child’s mother. A functional limitation was deﬁned as a condition lasting
six months or more and consisting of one or more restrictions in physical activity, mobility or self care due to an illness,
injury or medical condition and/or a limitation in role performance (kind or amount of ordinary play or schoolwork) due to
physical, emotional or learning problems.
Analyses
Because of the relatively high overlap of exposure to different types of abuse and small sample sizes associated with ‘pure’
categories, studies of child maltreatment rarely subdivide abuse into mutually exclusive types. Accordingly, respondents
who reported both physical and sexual abuse in this study were included in both categories.
OCHS data were weighted to compensate for differential sample loss (Boyle et al., 2007). Respondents with missing
covariates were assigned mean (quantitative variables) or modal (qualitative variables) values, were classiﬁed as missing
some information, and retained in the analysis. Prevalence estimates and between-sibling associations for child abuse came
from weighted data using SPSS version 11.0.1. To estimate between-sibling associations for different types of abuse, we
created a separate ﬁle of households that had information on the experience of child abuse from two or more siblings.
Households with three or more siblings were subdivided so that the information collected on all combinations of two
siblings could be analyzed. The resulting 2 × 2 tables were used to estimate conditional risk (i.e., chance of one sibling being
exposed to abuse conditional on another sibling being exposed to the same type of abuse) as well as the relative odds and
agreement on exposure to the same type of abuse within families. Sample weights were rescaled to sum to the total number
of households to avoid generating unrealistically precise conﬁdence intervals. To test the accuracy of these estimates, we
modeled the strength of association between siblings in the occurrence of abuse using logistic regression, controlling for the
number of paired comparisons in families with 3 or more siblings. The estimated odds ratios were almost identical to the
estimates obtained in the simple analyses described above.
Multilevel logistic regressions using weighted data with MLwiN (Rasbash et al., 2000) tested models for three outcomes:
child physical abuse, sexual abuse, and severe physical abuse (see Fig. 1 for deﬁnitions). This study uses a two-level logistic
regression with simultaneous consideration of i children, nested in j families. In models with two or more levels, each level is
associated with its own, unexplained residual error. Residual error at the individual level is constrained to 1; each successive
level is associated with its own error term e −N (0, ˚2 ) where ˚2 estimates the residual between-group variation (Snijders
& Bosker, 1999). Final models are presented in which all of the covariates under consideration are included.
Results
Twenty-one (1.1%) respondents out of 1,928 did not have sufﬁcient data to make a determination of child physical abuse;
a further 14 (0.7%) did not answer the child sexual abuse item; 69 respondents were missing covariate data and were
retained in the analysis. The ﬁnal unweighted analysis sample was 1,893 (98.2% of 1,928). These 1,893 respondents came
from 1,253 families; 757 individuals (40.0%) were the sole respondent in their family and 748 respondents (39.5%) were
from two-respondent families. The remaining respondents were from three, four and ﬁve-respondent families.
Respondent characteristics are described in Table 1. Of particular note, more than one-ﬁfth of respondents (21.4%) were
living in poverty as a child in 1983. Table 2 shows the prevalence of child abuse by sex for the OCHS sample. More males
18
H.L. MacMillan et al. / Child Abuse & Neglect 37 (2013) 14–21
Table 1
Characteristics of OCHS respondents by sex.
Characteristic
Males
% (SE)
N = 924a
Married/cohabiting
Childhood urban residence
Living in poverty in childhood
Maternal young age at ﬁrst birth
Family dysfunction
Parental adversity
Childhood psychiatric disorder
Childhood functional limitation
49.1 (.02)
67.7 (.01)
21.9 (.01)
9.2 (.01)
11.1 (.01)
8.4 (.01)
14.4 (.01)
5.0 (.01)
Characteristic
Total
% (SE)
N = 1893a
58.6 (.02)
67.6 (.01)
21.0 (.01)
11.2 (.01)
10.6 (.10)
10.0 (.01)
8.2 (.01)
3.9 (.01)
53.7 (.01)
67.7 (.01)
21.4 (.01)
10.2 (.01)
10.8 (.01)
9.1 (.01)
11.4 (.01)
4.4 (.00)
Females
Mean (SE)
Males
Mean (SE)
Childhood SES
Age
a
Females
% (SE)
N = 969a
0.0 (1.00)
27.1 (.12)
Total
Mean (SE)
0.0 (1.00)
27.1 (.12)
0.0 (1.00)
27.1 (.09)
Unweighted N; estimates based on weighted data.
Table 2
Prevalence of self-reported childhood abuse by sex.
Physical abuse
Sexual abuse
Severe physical abuse
Physical and/or sexual abuse
Both physical and sexual abuse
a
Total
N = 1,893a
% (SE)
Males
N = 924a
% (SE)
Females
N = 969a
% (SE)
Rao-Scott ␹2
statistic
p-value
31.0 (.01)
15.0 (.01)
19.9 (.01)
37.9 (.01)
8.1 (.01)
33.7 (.01)
8.3 (.01)
21.5 (.01)
37.1 (.01)
4.9 (.01)
28.2 (.01)
22.1 (.01)
18.3 (.01)
38.8 (.02)
11.6 (.01)
4.81
50.68
2.16
0.46
19.18
.03
<.0001
.14
.50
<.0001
Unweighted N; estimates based on weighted data.
Table 3
Physical and sexual abuse occurring within families.
Physical abuse
Percentage of sibling pairs where both reported same type of
abuse among families where at least one respondent
reported abusea
Risk: OR (95% CI)
Agreement: Kappa (SE)
Severe physical abuse
Sexual abuse
50.0, 54.6
32.3, 32.6
23.4, 24.3
3.8 (2.5–5.7)
.30 (.05)
2.5 (1.5–4.1)
.16 (.05)
2.0 (1.1–3.6)
.10 (.05)
Number (unweighted) of 2+ respondent families N = 496.
Number (unweighted) of within family pairs N = 809 (from 374 in 2 respondent households; 103 in 3 respondent households; 16 in 4 respondent households;
3 in 5 respondent households).
a
Two percentages for each abuse type represent conditional risk related to each sibling in the pair.
than females reported exposure to physical abuse in childhood, while child sexual abuse was substantially higher among
females than males.
Within multiple-child households, there was increased risk of another child being exposed to physical abuse (odds ratio
(OR) = 3.8, 95% conﬁdence interval (CI) 2.5–5.7) as well as severe physical abuse (OR = 2.5, 95% CI = 1.5–4.1) and sexual abuse
(OR = 2.0, 95% CI = 1.1–3.6) if one child reported such experiences; the risk was highest for physical abuse (Table 3). Also
shown in Table 3 are the percentages of sibling pairs where both respondents reported exposure to the same type of abuse
among families in which at least one reported abuse as well as the level of agreement based on Cohen’s Kappa.
Table 4 shows the results of multilevel logistic regressions for the three abuse outcomes. Child physical abuse and severe
physical abuse were both associated with living in poverty, urban residence, child psychiatric disorder and young maternal
age at the time of the ﬁrst child’s birth. Age (older) and sex (male) were signiﬁcant predictors of child physical abuse only
and not the severe category. Of particular note, parental adversity was a signiﬁcant predictor of severe child physical abuse.
The signiﬁcant predictors for child sexual abuse included being female, presence of parental adversity, age (older), urban
residence, living in poverty and young maternal age at the time of the ﬁrst child’s birth.
Discussion
The OCHS provides population-based information about the prevalence of exposure to child physical and sexual abuse.
Although the measures and timing of retrospective reports are somewhat different, some comparison is warranted with the
H.L. MacMillan et al. / Child Abuse & Neglect 37 (2013) 14–21
19
Table 4
Models of physical and sexual abuse: estimated odds ratios (ORs) and conﬁdence intervals.
Variable
Demographic
Male
Age
Childhood urban residence
Childhood SES
Living in poverty in childhood
Parent/family
Young maternal age at ﬁrst birth
Parent adversity (hospitalized for nerves/arrested)
Family dysfunction
Child
Psychiatric disorder
Functional limitation
Missing data
Random effects
Level 2 (family) (SE)
Level 1 (individual)
*
Physical abuse
OR (95% CI)
Severe physical
abuse
OR (95% CI)
Sexual abuse
OR (95% CI)
1.3 (1.0–1.5)*
1.0 (1.0–1.1)*
1.3 (1.0–1.6)*
1.0 (0.8–1.1)
1.4 (1.1–2.0)*
1.1 (0.9–1.4)
1.0 (1.0–1.0)
1.3 (1.0–1.7)*
1.0 (0.8–1.1)
1.6 (1.1–2.3)*
0.3 (0.2–0.4)*
1.1 (1.0–1.1)*
1.6 (1.2–2.1)*
0.9 (0.8–1.1)
1.5 (1.0–2.1)*
1.6 (1.1–2.3)*
1.3 (0.9–1.9)
1.2 (0.8–1.7)
1.6 (1.1–2.3)*
1.6 (1.1–2.4)*
1.4 (1.0–2.1)
2.0 (1.3–2.8)*
2.2 (1.5–3.3)*
1.4 (0.9–2.2)
1.9 (1.4–2.6)*
1.5 (0.9–2.5)
1.1 (0.6–2.0)
1.7 (1.1–2.4)*
1.4 (0.8–2.5)
0.7 (0.3–1.4)
1.4 (0.9–2.1)
1.6 (0.9–2.9)
0.7 (0.3–1.7)
0.52 (0.12)*
1.00
0.51 (0.21)*
1.00
0.16 (0.26)
1.00
p < .05.
earlier Ontario community data from the Ontario Health Supplement discussed in the Introduction. In the Supplement, the
Child Maltreatment History Self-Report (CMHSR) assessed physical abuse (6 items) and sexual abuse (4 items) by an adult
while the respondent was “growing up”; the OCHS CEVQ physical abuse item stems are similar “before age 16, did an adult. . .”.
Comparing the sample aged 21–35 from the Supplement, rates of physical abuse are similar: 30.0% of Supplement males and
22.9% Supplement females report child physical abuse. However, the OCHS shows higher rates of child sexual abuse for males
and females: 22.1% of OCHS females compared with 14.1% in the Supplement females; 8.3% of OCHS males compared with
3.3% of Supplement males. Both questionnaires ask about events occurring to the respondent during childhood; however,
the OCHS CEVQ sexual abuse stem inquires about “anyone” whereas the CMHSR speciﬁed “an adult”. This could account for
the higher rates of child sexual abuse in the OCHS sample as it is now recognized that the perpetrators of sexual abuse in
childhood are often younger than 18 (Finkelhor, 1994; Finkelhor, Ormrod, Turner, & Hamby, 2005). The OCHS ﬁndings with
regard to sex differences are consistent with the Supplement ﬁndings (MacMillan et al., 1997).
The OCHS has some important advantages over the Supplement; information on risk factors for abuse was gathered
prospectively from informants other than the respondents. The study by Brown et al. (1998) and the Christchurch Health
and Development Study (Fergusson et al., 1996) have the advantage of access to ofﬁcial child protection agency reports
available for comparison with self-reports; this was not possible for the OCHS because of concerns about conﬁdentiality.
The child maltreatment questions asked in the study by Brown and colleagues were quite limited with one question for each
of physical abuse, sexual abuse, and neglect. In the case of the Christchurch Health and Development Study, respondents
were asked about a series of 15 sexual acts. In the study by Hussey et al. (2006), maltreatment measures were modiﬁed
versions of previous studies with one question for each of supervision neglect, physical neglect, physical assault, and contact
sexual abuse. The OCHS has the advantage of using a child maltreatment measure that has been evaluated for its reliability
and validity. Despite such wide variability in the measures used to ask retrospectively about exposure to maltreatment, it is
useful to compare the risk factors identiﬁed by the OCHS with the US and New Zealand-based studies.
It is difﬁcult to interpret the ﬁnding of increased risk for exposure to physical (and severe physical) and sexual abuse
for those who resided during childhood in communities with populations of over 25,000 residents. In the Supplement,
child physical abuse was reported more commonly among females raised in rural areas compared to those who grew up
in urban areas; no such association was found for males (MacMillan et al., 1997). Rural for the Supplement was deﬁned as
a community with fewer than three thousand residents, so a direct comparison is not possible. To our knowledge, urban
residence in childhood has not been identiﬁed previously as a risk indicator for maltreatment.
Of particular note is the OCHS ﬁnding that a general measure of SES was not a signiﬁcant predictor of either child physical
or severe physical abuse, but living in poverty in childhood – a measure of extreme disadvantage – was associated with an
increased risk of physical (and severe physical) and sexual abuse. The association between living in poverty and child sexual
abuse differs from some reports in the literature (Finkelhor et al., 2005; Finkelhor, 1994; MacMillan et al., 1997). Although
ofﬁcial reports of victimization have shown an association between low SES and child sexual abuse, this has been considered
a reporting bias, whereby suspicions of maltreatment in low-income families are more likely to be reported (Hampton &
Newberger, 1985). In the study by Brown et al. (1998), child physical abuse but not sexual abuse was predicted by welfare
dependence. Similarly in the Christchurch Health and Development Study, there were no signiﬁcant associations between
child sexual abuse and family socioeconomic status, although lower levels of maternal education were associated with
increasing rates of child sexual abuse. Although Hussey et al. (2006) found a signiﬁcant association between family income
20
H.L. MacMillan et al. / Child Abuse & Neglect 37 (2013) 14–21
below the federal poverty level and child sexual abuse, their measurement of child abuse (by 6th grade) does not capture
child sexual abuse that occurred during adolescence.
One of the most important ﬁndings in the OCHS is the consistency with which the ﬁrst child being born to a young mother
predicted child physical abuse (and the severe category), and sexual abuse. This risk factor was associated with abuse after
controlling for family SES and poverty. Maternal youth was associated with both child physical and sexual abuse in the
New York study, but does not appear to have been measured in the Christchurch Health and Development Study and Add
Health. Parental adversity was a predictor of child sexual abuse in the OCHS; this was also the case for both the New York
(maternal sociopathy) and New Zealand-based studies, especially parental alcoholism in the latter. Finkelhor et al. (2005)
have suggested that the unavailability of a parental ﬁgure may place a child at risk for exposure to child sexual abuse. These
studies support this hypothesis.
The ﬁndings about risk of exposure to abuse among siblings are particularly notable: this information highlights the
importance of considering risk to other siblings in a household where one child is identiﬁed as a victim of physical abuse
(including the severe category), as well as sexual abuse. Given that parents are most commonly the perpetrators of child
physical abuse, but this is not the case for child sexual abuse (Gilbert et al., 2009), it is not surprising that the risk for the
same exposure among siblings is higher for physical than sexual abuse. However, a two-fold risk of sexual abuse in siblings
is important; in this sample, among sibling pairs where one respondent reported exposure to sexual abuse in childhood,
in almost 25% of those pairs, both siblings reported such exposure, emphasizing the need to consider the risk to siblings in
clinical evaluations.
Strengths and limitations
The validity of the ﬁndings of the present study is strengthened by the fact that information about predictors was collected
in 1983 and from informants other than those providing information about exposure to maltreatment. It is possible to identify
young maternal age at time of ﬁrst birth as a risk factor, since it clearly preceded the child’s exposure to maltreatment, and
the potential confounding effects of low socioeconomic status at the point of assessment in 1983 are taken into account.
Other predictors can, at most, be described as risk indicators, since it is unknown whether the other signiﬁcant parent/family
variables preceded or followed abuse. This is particularly the case for child variables such as presence of child psychiatric
disorder, which was associated with child physical and severe physical abuse, but not exposure to sexual abuse.
Some authors criticize the retrospective approach to asking about exposure to child maltreatment (Widom, Raphael, &
Dumont, 2004). But as others emphasize, to measure this contemporaneously requires identiﬁcation of maltreated children
in the community with all the legal and ethical challenges inherent in doing so (Kendall-Tackett & Becker-Blease, 2004).
Focusing exclusively on maltreated children involved with the child protection system is also not the answer: prospective
designs with such samples would likely miss a large proportion of those abused. Data from the Supplement suggest that only
5% of those who reported either child physical or sexual abuse or both, reported contact with the child protection system
(MacMillan et al., 2003). In a study by Kendall-Tackett (1991), 82% of those who disclosed sexual abuse indicated that they
had never reported this to a child protection or legal agency. She suggests that participants in prospective and retrospective
studies of child maltreatment may be two distinct sub-groups, and it is important to have information from both samples.
The OCHS did not include questions about exposure to neglect because of the challenges in measuring its occurrence,
especially retrospectively, in large community-based surveys. It is one of the most common and serious forms of child
maltreatment (Trocmé et al., 2005), and developing reliable and accurate approaches to measure it should be a priority
(Sullivan, 2000). As progress is made in deﬁning neglect (Dubowitz et al., 2005), we hope it will be possible to incorporate
psychometrically sound questions in future Canadian surveys. Much less research is available to inform approaches to
measuring emotional abuse within the context of a community-based survey (Tonmyr, Draca, Crain, & MacMillan, 2011).
Ideally a survey would include questions about all major categories: child physical, sexual, emotional abuse, neglect and
exposure to intimate partner violence.
The third wave of the OCHS was conducted when the respondents were young adults; it would be preferable for information about exposure to maltreatment to be ﬁrst collected in adolescence. Such information collected from population-based
samples of youths, combined with surveillance data about the incidence of reported child maltreatment using the CIS would
provide the most comprehensive knowledge about child maltreatment, one of the major public health problems affecting
Canadian children and youths.
Conclusions
The OCHS provides important information about the high prevalence of exposure to child physical and sexual abuse.
Noteworthy is the strong evidence for the association between young maternal age at the time of ﬁrst birth, living in poverty
in childhood and urban residence, and both types of abuse. This is new information for health care providers and child
protection workers, as previous studies have suggested that these factors were associated primarily with child physical
abuse, but not sexual abuse. Our ﬁndings suggest that children with these characteristics should be considered at risk of
experiencing both child physical and sexual abuse. Furthermore, identiﬁcation of these factors lends additional support
to the recommendation by Olds and Kitzman (1990) to target interventions to communities where there are high rates
of teenage parenthood and poverty. Given the commonalities in risk indicators for both child physical and sexual abuse,
H.L. MacMillan et al. / Child Abuse & Neglect 37 (2013) 14–21
21
it would be important to determine if implementation of effective programs such as the Nurse Family Partnership (Olds,
Sadler, & Kitzman, 2007), which has been shown to reduce ofﬁcial reports of child maltreatment, results in reductions in
speciﬁc subtypes of child maltreatment, such as physical and sexual abuse. Our results also emphasize the importance of
child protection workers and health care providers assessing the siblings of children exposed to child physical and sexual
abuse. The fact that the co-occurrence of sexual abuse and even more so physical abuse (as well as the severe category) in
childhood is so high among siblings, gives strong rationale to implementing policies that ensure all children in a household
where one child is identiﬁed as physically or sexually abused, are also assessed for such exposure.
References
Belsky, J., & Vondra, J. (1989). Lessons from child abuse: The determinants of parenting. In D. Cicchetti, & V. Carlson (Eds.), Current research and theoretical
advances in child maltreatment (pp. 153–202). Cambridge, England: Cambridge University Press.
Blishen, B. R., Carroll, W. K., & Moore, C. (1987). The 1981 socioeconomic index for occupations in Canada. Canadian Review of Sociology and Anthropology,
24, 465–488.
Boyle, M. H., Georgiades, K., Racine, Y., & Mustard, C. (2007). Neighborhood and family inﬂuences on educational attainment: Results from the Ontario Child
Health Study follow-up 2001. Child Development, 78, 168–189.
Boyle, M. H., Hong, S., Georgiades, K., Duku, E., Racine, Y., & Mustard, C. (2006). Ontario Child Health Study Follow-up 2001: Evaluation of sample loss. Hamilton,
Ontario: Offord Centre for Child Studies.
Boyle, M. H., Offord, D. R., Campbell, D., Catlin, G., Goering, P., Lin, E., & Racine, Y. A. (1996). Mental health supplement to the Ontario Health Survey:
Methodology. Canadian Journal of Psychiatry, 41, 549–558.
Boyle, M. H., Offord, D. R., Hofmann, H. G., Catlin, G. P., Byles, J. A., Cadman, D. T., Crawford, J. W., Links, P. S., Rae-Grant, N. I., & Szatmari, P. (1987). Ontario
Child Health Study. I. Methodology. Archives of General Psychiatry, 44, 826–831.
Boyle, M. H., Offord, D. R., Racine, Y., Sanford, M., Szatmari, P., & Fleming, J. E. (1993). Evaluation of the original Ontario Child Health Study scales. Canadian
Journal of Psychiatry, 38, 397–405.
Brown, J., Cohen, P., Johnson, J. G., & Salzinger, S. (1998). A longitudinal analysis of risk factors for child maltreatment: Findings of a 17-year prospective
study of ofﬁcially recorded and self-reported child abuse and neglect. Child Abuse & Neglect, 22, 1065–1078.
Byles, J., Byrne, C., Boyle, M. H., & Offord, D. R. (1988). Ontario Child Health Study: Reliability and validity of the general functioning subscale of the McMaster
Family Assessment Device. Family Process, 27, 97–104.
Dubowitz, H., Newton, R. R., Litrownik, A. J., Lewis, T., Briggs, E. C., Thompson, R., English, D., Lee, L. C., & Feerick, M. M. (2005). Examination of a conceptual
model of child neglect. Child Maltreatment, 10, 173–189.
Epstein, N. B., Baldwin, L. M., & Bishop, D. S. (1983). The McMaster Family Assessment Device. Journal of Marital and Family Therapy, 9, 171–180.
Fergusson, D. M., Lynskey, M. T., & Horwood, L. J. (1996). Childhood sexual abuse and psychiatric disorder in young adulthood: I. Prevalence of sexual abuse
and factors associated with sexual abuse. Journal of the American Academy of Child and Adolescent Psychiatry, 35, 1355–1364.
Finkelhor, D. (1994). Current information on the scope and nature of child sexual abuse. The Future of Children, 4, 31–53.
Finkelhor, D., Ormrod, R., Turner, H., & Hamby, S. L. (2005). The victimization of children and youth: A comprehensive, national survey. Child Maltreatment,
10, 5–25.
Gilbert, R., Widom, C. S., Brown, K., Fergusson, D., Webb, E., & Janson, S. (2009). Burden and consequences of child maltreatment in high-income countries.
Lancet, 373, 68–81.
Hampton, R. L., & Newberger, E. H. (1985). Child abuse incidence and reporting by hospitals: Signiﬁcance of severity, class, and race. American Journal of
Public Health, 75, 56–60.
Hussey, J. M., Chang, J. J., & Kotch, J. B. (2006). Child maltreatment in the United States: Prevalence, risk factors, and adolescent health consequences.
Pediatrics, 118, 933–942.
Kendall-Tackett, K. A. (1991). Characteristics of abuse that inﬂuence when adults molested as children seek treatment. Journal of Interpersonal Violence, 6,
486–493.
Kendall-Tackett, K., & Becker-Blease, K. (2004). The importance of retrospective ﬁndings in child maltreatment research. Child Abuse & Neglect, 28,
723–727.
MacMillan, H. L., Fleming, J. E., Trocmé, N., Boyle, M. H., Wong, M., Racine, Y. A., Beardslee, W. R., & Offord, D. R. (1997). Prevalence of child physical and
sexual abuse in the community. Results from the Ontario Health Supplement. The Journal of the American Medical Association, 278, 131–135.
MacMillan, H. L., Jamieson, E., & Walsh, C. A. (2003). Reported contact with child protection services among those reporting child physical and sexual abuse:
Results from a community survey. Child Abuse & Neglect, 27, 1397–1408.
Offord, D. R., Boyle, M. H., Racine, Y. A., Fleming, J. E., Cadman, D. T., Blum, H. M., Byrne, C., Links, P., Lipman, E. L., MacMillan, H. L., Sanford, M. N., Szatmari,
P., Thomas, H., & Woodward, C. A. (1992). Outcome, prognosis and risk in a longitudinal follow-up study. Journal of the American Academy of Child and
Adolescent Psychiatry, 31, 916–923.
Offord, D. R., Boyle, M. H., Campbell, D., Goering, P., Lin, E., Wong, M., & Racine, Y. A. (1996). One-year prevalence of psychiatric disorder in Ontarians 15 to
64 years of age. Canadian Journal of Psychiatry, 41, 559–563.
Offord, D. R., Boyle, M. H., Szatmari, P., Rae-Grant, N. I., Links, P. S., Cadman, D. T., Byles, J. A., Crawford, J. W., Blum, H. M., Byrne, C., Thomas, H., & Woodward,
C. A. (1987). Ontario Child Health Study. II. Six-month prevalence of disorder and rates of service utilization. Archives of General Psychiatry, 44, 832–836.
Olds, D. L., & Kitzman, H. (1990). Can home visitation improve the health of women and children at environmental risk? Pediatrics, 86, 108–116.
Olds, D. L., Sadler, L., & Kitzman, H. (2007). Programs for parents of infants and toddlers: Recent evidence from randomized trials. Journal of Child Psychology
and Psychiatry, 48, 355–391.
Public Health Agency of Canada. (2010). Canadian Incidence Study of Reported Child Abuse and Neglect – 2008: Major Findings. Ottawa.
Rasbash, J., Browne, W. J., Goldstein, H., Yang, M., Plewis, I., Healy, M., Woodhouse, G., Draper, D., Langford, I., & Lewis, T. (2000). A user’s guide to MLwiN,
Version 2.1d. London: Institute of Education, University of London.
Snijders, T., & Bosker, R. (1999). Multilevel analysis: An introduction to basic and advanced multilevel modeling. London: SAGE Publications.
Sullivan, S., & for the Family Violence Prevention Unit, Health Canada. (2000). Child Neglect: Current deﬁnitions and models—A review of child neglect research
1993–1998. Ottawa, Canada: The National Clearinghouse on Family Violence. Cat. H72-21/171-2000E.
Tanaka, M., Wekerle, C., Leung, E., Waechter, R., Gonzalez, A., Jamieson, E., & MacMillan, H. L. (2012). Preliminary Evaluation of the Childhood Experiences
of Violence Questionnaire Short Form. Brief Note. Journal of Interpersonal Violence, 27(2), 396–407.
Tonmyr, L., Draca, J., Crain, J., & MacMillan, H. L. (2011). Measurement of emotional/psychological child maltreatment: A review. Child Abuse and Neglect,
35, 767–782.
Trocmé, N., Fallon, B., MacLaurin, B., Daciuk, J., Felstiner, C., Black, T., Tonmyr, L., Blackstock, C., Barter, K., Turcotte, D., & Cloutier, R. (2005). Canadian
incidence study of reported child abuse and neglect: Major ﬁndings – 2003. Ottawa, Canada: Minister of Public Works and Government Services Canada.
Walsh, C. A., MacMillan, H. L., Trocmé, N., Jamieson, E., & Boyle, M. H. (2008). Measurement of victimization in adolescence: Development and validation of
the Childhood Experiences of Violence Questionnaire. Child Abuse & Neglect, 32, 1037–1057.
Widom, C. S., Raphael, K. G., & DuMont, K. A. (2004). The case for prospective longitudinal studies in child maltreatment research: Commentary on Dube,
Williamson, Thompson, Felitti, and Anda (2004). Child Abuse & Neglect, 28, 715–722.