Content uploaded by Martine Hébert
Author content
All content in this area was uploaded by Martine Hébert on Jan 29, 2018
Content may be subject to copyright.
Gender as a predictor of posttraumatic stress symptoms and
externalizing behavior problems in sexually abused children
Amélie Gauthier-Duchesnea, Martine Héberta, and Marie-Ève Daspea
aDépartement de sexologie, Université du Québec à Montréal, C.P. 8888, succursale Centre-Ville,
Montréal, Québec, Canada, H3C 3P8
Abstract
Despite the proliferation of studies documenting outcomes in sexually abused victims, gender
differences remain understudied. The bulk of studies have relied on retrospective samples of adults
with insufficient representation of male victims to explore gender specificities. This study
examined differential outcomes among boy and girl victims of sexual abuse. A predictive model of
outcomes including abuse characteristics and sense of guilt as mediators was proposed. Path
analysis was conducted with a sample of 447 sexually abused children (319 girls and 128 boys),
aged 6 to 12. Being a girl was a predictor of posttraumatic stress symptoms, while being a boy was
a predictor of externalizing problems. Being a boy was also associated with more severe abuse,
which in turn predicted posttraumatic stress symptoms. Child’s gender was not related to
perpetrator’s relationship to the child or sense of guilt. However, sense of guilt predicted
posttraumatic stress symptoms and externalizing problems while perpetrator’s relationship to the
child predicted externalizing problems. Gender specificities should be further studied among
sexually abused children, as boys and girls appear to manifest different outcomes. Sense of guilt
should be a target in intervention for sexually abused children, as results highlight its link to
heightened negative outcomes.
Keywords
gender differences; child sexual abuse; posttraumatic stress disorder; externalizing problems; sense
of guilt; abuse characteristics
1. Introduction
Sexual abuse (SA) is an important social issue that affects both girls and boys. According to
a worldwide meta-analysis, 1 in 5 women and 1 in 10 men report being sexually victimized
prior to the age of 18 (Stoltenborgh, van Ijzendoorn, Euser, & Bakermans-Kranenburg,
2011). Past studies have clearly demonstrated that childhood SA is a significant risk factor
for depression, anxiety, low self-esteem, suicide attempts, as well as alcohol and drug
dependence (Fergusson, McLeod, & Horwood, 2013). In the short term, children who have
disclosed SA are likely to show posttraumatic stress disorder (PTSD) symptoms (Hébert,
Langevin, & Daigneault, 2016). In addition, relative to their non-abused peers, child victims
Correspondence to: Martine Hébert.
PubMed Central CANADA
Author Manuscript / Manuscrit d'auteur
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
Published in final edited form as:
Child Abuse Negl
. 2017 February ; 64: 79–88. doi:10.1016/j.chiabu.2016.12.008.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
present significant depressive and anxiety symptoms and externalizing behavior problems
such as anger and aggressiveness (Hébert, 2011).
With the aim of orienting treatment priorities and identifying relevant targets for
intervention, studies have explored factors that impact severity of symptoms, including
characteristics of the abuse sustained (severity of the acts, duration of the abuse, the
perpetrator’s relationship to the child), attributions, coping strategies, as well as social
support (Cantón-Cortés, Cantón, Justicia, & Cortés, 2011; Zajac, Ralston, & Smith, 2015).
Besides these potential factors, gender of the child victim may be associated with outcomes
of SA. Yet, few studies have explored gender differences among sexually abused children. In
the vast majority of childhood SA studies, boys are either completely absent from samples
or insufficiently represented, making it difficult to conduct gender-specific analyses
(Maikovich-Fong & Jaffee, 2010; Villeneuve Cyr & Hébert, 2011).
1.1. Gender differences: Findings from past studies
Past studies of adult samples suggest that sexually abused women report more internalizing
problems than sexually abused men, as well as PTSD (Blain, Galovski, & Robinson, 2010).
For women survivors of SA, some studies suggest that they show more anxiety and
depressive symptoms (Banyard, Williams, & Siegel, 2004). However, other studies report no
gender differences between men and women victims of SA regarding depressive symptoms
(Arnow, Blasey, Hunkeler, Lee, & Hayward, 2011) and PTSD (Tolin & Foa, 2006). Adult
victims were also found more likely to have alcohol and drug problems than non-adult
victims, but those risks appear similar for men and women (Dube et al., 2005). From their
30-year longitudinal study on survivors of SA, Fergusson et al. (2013) observed that gender
did not impact adult developmental outcomes. While studies among adult samples are
important to document long-term SA consequences, retrospective studies may include biases
related to memory and may introduce significant measurement error (Hardt & Rutter, 2004).
Gender differences have also been explored in samples of teenagers and children. Sexually
abused teenaged boy victims seem to express more externalizing difficulties, such as
delinquent behaviors, sexual risk behaviors and alcohol and drug abuse compared to
sexually abused girls (Chandy, Blum, & Resnick, 1996; Garnefski & Arends, 1998). In their
study based on examination of judicial and social reports, Soylu et al. (2016) observed that
girl victims under 18 had more psychiatric and major depressive disorder than boys (
n
=
248), yet PTSD was as prevalent in girls and boys. According to Villeneuve Cyr and Hébert
(2011), school-aged SA girls reported more PTSD and anxiety symptoms than boys. Boys (
n
= 33) tended to have more externalizing behaviors than girls whereas no gender difference
was found for internalizing problems (Villeneuve Cyr & Hébert, 2011). These results are in
contrast with those reported by Coohey (2010) with preteens aged 11 to 14. This study
pointed out that boys (
n
= 31) were twice as likely to have internalizing behaviors (52% vs.
24%) than girls (Coohey, 2010). However, Coohey argued that “sexually abused boys may
be more likely to internalize during early adolescence and externalize during later
adolescence, whereas sexually abused girls may be more likely than boys to exhibit
internalizing behavior throughout adolescence” (Coohey, 2010, p. 860). Another study
conducted by Maikovich-Fong and Jaffee (2010) observed no difference between boy (
n
=
Gauthier-Duchesne et al. Page 2
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
117) and girl victims of SA, aged 4 to 16, for internalizing, externalizing and PSTD
symptoms.
These contradictory results may relate to methodological differences, including the age of
the participants in the different samples, the definition of SA that varied across these studies,
and the underrepresentation of boys. Indeed, studies on children and adolescents who have
disclosed SA usually included around 30 boys only (Feiring, Taska, & Lewis, 1999), which
limits the power to detect significant differences. Analyses of gender differences among SA
victims are often limited to descriptive analyses and few studies have examined possible
explanatory mechanisms. The present study will attempt to overcome these limitations by
testing mediators of the relation between gender and outcomes in a large school-aged sample
of SA victims.
1.2. Gender differences: Potential interpretations
A number of interpretations can be considered to account for gender differences in SA
outcomes. First, gender role in socialization may contribute to gender differences. For
example, boys may be less often reprimanded than girls for displaying aggressive behaviors,
making them more likely to develop externalizing symptoms, and making the latter more
prone to develop internalized symptoms.
Second, the SA experienced by boys and girls might be different, which could impact the
type and intensity of outcomes. For example, some studies have revealed that SA perpetrated
toward boys is more likely to involve severe or intrusive gestures (Edinburgh, Saewyc, &
Levitt, 2006; Soylu et al., 2016; Villeneuve Cyr & Hébert, 2011). The greater severity of SA
experienced by boys may explain the higher level of externalizing problems observed among
male victims (Banyard et al., 2004). Being a girl seems to be associated with longer duration
of SA and a closer perpetrator (Coohey, 2010; Soylu et al., 2016; Villeneuve Cyr & Hébert,
2011). These characteristics may negatively influence SA outcomes (Hébert, Tremblay,
Parent, Daignault, & Piché, 2006; Yancey & Hansen, 2010).
Third, boys may experience more guilt because of the internalized stigma related to same
gender perpetrator (Banyard et al., 2004). The vast majority of reported child abusers are
male (Dube et al., 2005; Soylu et al., 2016), which means that boys, compared to girls, are
often abused by a same gender person. This might create an additional issue, unique to boys,
about masculinity and sexual orientation (Banyard et al., 2004). Boys may report a greater
sense of guilt because they may perceive that they were not able to protect themselves,
which is a prescribed role for men. In fact, these gender norms may reinforce guilt felt by
boys, which may influence outcomes and delay disclosure (Gagnier & Collin-Vézina, 2016).
Sense of guilt and self-blame are correlates that have been shown to mediate SA outcomes
(Feiring & Cleland, 2007), such as PTSD symptoms (Cantón-Cortés et al., 2011). According
to the traumagenic dynamics theory of Finkelhor and Browne (1985), stigmatization, which
encompasses guilt and shame, contributes to the apparition of externalizing behavior
problems (drug and alcohol abuse, criminal activity, suicide attempts). If boys do indeed
have a higher sense of guilt than girls, they may consequently develop more externalized
behavior problems following SA.
Gauthier-Duchesne et al. Page 3
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
1.3. The present study
The aim of this study is to examine the role of gender in SA outcomes across 6 to 12 year
old child victims of SA. A predictive model of SA outcomes including SA characteristics
and sense of guilt as mediators (see Figure 1) is proposed. We hypothesize that being a girl
will predict more PTSD symptoms and that this association will be mediated by the
frequency of SA and the relationship with the perpetrator. We also hypothesize that being a
boy will be associated with more behavior problems and that this relationship will be
mediated by the severity of the abuse and sense of guilt.
2. Method
2.1. Participants
The sample consisted of 447 sexually abused children (319 girls and 128 boys), aged 6 to 12
(
M
= 8.99,
SD
= 2.05) and one of their non-offending parental figures (347 mothers, 51
fathers, 45 other significant parental figures and 4 youth center educators). Participants were
recruited during the initial evaluation at different centers located in the province of Quebec
offering specialized services for sexually abused children. All families of children aged 6 to
12 consulting the agencies were invited to participate and during the study, 447 families
accepted to participate while 42 out of 489 eligible families declined to participate. A total
of 86.3% of the participants were French Canadians. Table 1 shows socio-demographic
characteristics for the sample. No significant difference was observed between girls and
boys for any of these characteristics.
2.2. Measures
2.2.1. Socio-demographic Characteristics—Parental figures completed a
questionnaire on socio-demographic regarding family structure, family income, education
level, child’s age and child’s gender. Child’s gender was coded as follows: 0 = girl, 1 = boy.
2.2.2. Characteristics of SA—An adaptation of the
History of Victimization Form
(HVF; Parent & Hébert, 2006) was used to codify SA characteristics based on information
from the child’s medical or clinical record by trained research assistants. Prior analyses of
inter-rater reliability were based on 30 records and indicated high agreement; the median
intraclass correlation was 0.86 (Hébert et al., 2006). When information regarding SA history
and characteristics was missing from the medical or clinical record, we inquired from other
sources (parental figures). Otherwise, the lacking information was treated as missing data.
The severity of the acts involved was coded as 1 = less severe (exhibitionism, voyeurism,
kisses, exposure to pornographic material, physical contact over clothing), 2 = severe
(physical contact under clothing, touching of the genitals), and 3 = very severe (oral sex,
vaginal or anal penetration or attempted penetration). The frequency of the SA was
categorized as 1 = single episode, 2 = some events (less than 6 months), and 3 = repetitive or
chronic (more than 6 months). Perpetrator’s relationship to the child included four
categories: 1 = immediate family (parent, stepparent, sibling and stepparent’s child), 2 =
extended family (uncle, aunt, cousin and grandparent), 3 = family acquaintance (such as
foster parent, daycare provider, child’s friend, neighbor) and 4 = stranger. For the few
situations that involved more than one perpetrator (
n
= 27), the variable was coded for the
Gauthier-Duchesne et al. Page 4
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
perpetrator who had the closest relationship to the child. Gender and age of the perpetrator
were also collected in the HVF.
2.2.3. PTSD Symptoms and Sense of Guilt—Children completed the
Children’s
Impact of Traumatic Events Scale II
(CITES-II; Wolfe, 2002). The scale includes 46 items
evaluating re-experiencing symptoms, avoidant behaviors and hyperarousal problems.
Participants answered questions by referring to the SA symptoms experienced in the last
month. A subscale of the CITES-II, which includes 3 items, measured the sense of guilt
revealed by the child specific to the SA experienced. The scale for each item of the CITES-II
is 0 = not true, 1 = sometimes true, and 2 = very true. Scores are calculated by adding the
results of each item and ranged from 0 to 92 for the global scale and from 0 to 6 for the
sense of guilt subscale. Internal consistency was excellent for the global PTSD symptoms
subscale (α = 0.92) and acceptable for the sense of guilt subscale (α = 0.71).
2.2.4. Internalizing and Externalizing Behavior Problems—The
Child Behavior
Checklist
(CBCL; Achenbach & Rescorla, 2001) was completed by the parental figure. This
instrument, consisting of 113 items, covers behavioral problems observed in the last two
months in children aged 6 to 18. Internalizing problems include anxious/depressed
symptoms, withdrawal and somatic complaints. Externalizing problems refer to rule-
breaking and aggressive behaviors. Each item of the CBCL is ranked using the following
scale: 0 = not true, 1 = somewhat or sometimes true, and 2 = very true or often true.
T
-scores
based on normalization samples were used for this scale.
T-
scores above 63 on the
internalizing and externalizing subscales are considered clinically significant (Achenbach &
Rescorla, 2001). A total of 39.8% of the sample reached the clinical threshold for
internalizing problems and 46.8% for externalizing problems. For this study, internal
consistency was good for internalizing problems (α = 0.88) and excellent for externalizing
problems (α = 0.93).
2.3. Procedure
Families were invited to participate in the research project on children victims of SA and
their parents during their first visit to the specialized center. For those who agreed, the
consent form was explained and signed. Confidentiality was assured to participants. The
child completed the questionnaire with the assistance of a trained research assistant in
psychology or sexology. Meanwhile, the parent was asked to complete a questionnaire alone
or with the assistance of a research assistant if needed. This study was approved by the
Human Research Review Committee of Ste-Justine Hospital and the Human Research
Review Committee of the Université du Québec à Montréal.
2.4. Data Analysis
T
-tests and chi-square tests were first conducted to examine gender differences in
characteristics and symptoms of SA. In addition, correlational analyses allowed the
identification of variables that were significantly associated with the victim’s gender as well
as examination of associations between the studied variables. Results were used to identify
the relevant variables to be included in the model. Path analysis was conducted to test the
predictive model of PTSD symptoms and behavior problems with victim’s gender as the
Gauthier-Duchesne et al. Page 5
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
exogenous variable and abuse characteristics as well as guilt as mediators. Various indices
were used to determine whether the specified model adequately fitted the observed data.
Good fit was indicated by a non-significant chi-square or a ratio of chi-square to degrees of
freedom (χ2/
df
) less than 3 (Jöreskog & Sörbom, 1993), a comparative fit index (CFI)
higher than 0.95 (Hu & Bentler, 1999), and root mean square error of approximation
(RMSEA) less than 0.06 with a confidence interval (CI) ranging between 0.00 and 0.08 (Hu
& Bentler, 1999). Considering the small proportion of missing data (between 0.4% and
5.8%) for the studied variables and a non-significant result for the Little test (χ2(73) =
71.606,
p
= 0.524), data was assumed to be missing completely at random. Mplus 7.0
(Muthén & Muthén, 1998–2015) was used to perform path analysis.
3. Results
3.1. Gender differences in characteristics of SA
Chi-square tests showed some gender differences in characteristics of SA experienced (see
Table 2). First, boys (74.6%) were more likely than girls (56.4%) to have suffered intrusive
acts (χ2(2) = 12.64;
p
= 0.002). Second, regarding duration of the SA, half (51.3%) of the
boys had experienced some events of SA, compared to 1 in 3 (36.4%) girls (χ2(2) = 7.87,
p
=
0.020). Girls (38.1%) were more likely to report repetitive or chronic events than boys
(30.2%), but this difference was not statistically significant. Third, just over half (52.2%) of
the girls were abused by an adult aged between 20 and 59 years old, compared to 37.1% of
the boys (χ2(3) = 14.45;
p
= 0.002). The boys (43.6%) were in turn more often the victims of
a juvenile perpetrator under 15 years old than girls (27.6%). No gender difference was found
regarding perpetrator’s relationship to the child (χ2(3) = 1.36;
p
= 0.716) and perpetrator’s
gender (χ2(1) = 0.39;
p
= 0.532), which was, for the majority of the sample, a male family
member.
3.2. Gender differences in symptoms of SA
T
-tests (see Table 3) revealed that girls (
M
= 46.09;
SD
= 17.15) showed higher global
PTSD scores (
t
(425) = 3.23;
p
= 0.001; Cohen’s
d
= 0.35) than boys (
M
= 39.92;
SD
=
19.38). As shown in Table 3, gender differences were observed for the three subscales of the
CITES-II. No significant difference between girls and boys was observed for the sense of
guilt specific to SA (
t
(425) = 1.05;
p
= 0.297; Cohen’s
d
= 0.11). Boys (
M
= 64.71;
SD
=
10.86) were reported by parents to display greater externalizing behavior problems (
t
(425) =
−3.76;
p
< 0.001; Cohen’s
d
= 0.40) than girls (
M
= 60.09;
SD
= 11.84). No significant
difference was found for internalizing problems (
t
(425) = −1.58;
p
= 0.114; Cohen’s
d
= 0.17)
except for the withdrawal subscale (
t
(425) = −3.21;
p
= 0.001; Cohen’s
d
= 0.34), where boys
(
M
= 63.23;
SD
= 10.18) had higher scores than girls (
M
= 60.05;
SD
= 8.99).
3.3. Correlations between the studied variables
Table 4 shows, for the total sample, correlation coefficients, means and standard deviations
for characteristics and symptoms of SA. Severity was the only characteristic of SA
associated with higher scores of PTSD symptoms. Furthermore, externalizing problems
were associated with only one characteristic of SA, which is perpetrator’s relationship to the
child: a closer relationship to the perpetrator was associated with lower externalizing
Gauthier-Duchesne et al. Page 6
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
problems. PTSD symptoms and externalizing problems were both positively associated with
child’s sense of guilt. Because the frequency of SA was not associated with any symptom of
SA, it was not included into the predictive model. As internalizing problems were not
associated with the child’s gender, sense of guilt or abuse characteristics, the variable was
excluded from the path model.
3.4. Mediation model
Path analysis was conducted to study gender as a predictor of SA symptoms, with sense of
guilt and abuse characteristics as mediators. The Maximum Likelihood method of estimation
was used and missing values were handled using Full Information Maximum Likelihood.
The predictive model (see fig. 2) indicated a good fit to the data (χ2(1) = 0.07;
p
= 0.785;
χ2/
df
= 0.07; CFI = 1.00; RMSEA = 0.00 with 90% CI [0.00 − 0.08]). Being a girl predicted
more PTSD symptoms (β = −0.15;
p
< 0.001), whereas being a boy predicted more
externalizing problems (β = 0.18;
p
< 0.001). An indirect effect of gender on PTSD
symptoms was also observed through severity of the abusive acts involved (
b
= 0.61 with
95% CI [0.11 − 1.37]), with a proportion of 9.9% of this effect going through abuse severity.
Guilt and perpetrator’s relationship to the child were not associated with child’s gender (β =
−0.05;
p
= 0.297 and β = 0.04;
p
= 0.454 respectively) and therefore did not mediate the
relationship between gender, PTSD symptoms and externalizing problems. Results however
suggest that sense of guilt was the most important predictor of PTSD symptoms (β = 0.39;
p
< 0.001), and also a predictor of externalizing problems (β = 0.13;
p
= 0.008). Moreover,
having a distant relationship to the perpetrator predicted more externalizing problems (β =
0.14;
p
= 0.003), but was not associated with PTSD symptoms (β = 0.07;
p
= 0.095). The
model explains 19% of the variance in PTSD symptoms and 7% of the variance in
externalizing problems.
4. Discussion
The aim of this study was to examine the role of gender in SA outcomes in school-aged
victims of SA by testing a predictive model including abuse characteristics and sense of guilt
as mediators. Results indicated that boy and girl victims of SA seem to manifest different
symptoms. Indeed, boys were more likely to display externalizing behavior problems, which
is consistent with the literature (Banyard et al., 2004). The association between gender and
PTSD seemed to be more complex. Results suggested that being a girl was directly
associated with PTSD symptoms. For boys however, this relationship was mediated by the
severity of abuse, as defined by the degree of intrusiveness of the sexual acts involved.
Therefore, being a boy was related to a greater severity of abuse, which in turn predicted
higher PTSD symptoms. Our results make an important contribution to better understanding
the trajectories of children victims of SA, as few studies have included such a large number
of boys in their samples. Girls seem more prone to experience traumatic symptoms (re-
experiencing the trauma, avoidant behaviors and hyperarousal problems) following SA. For
boys, the link between SA and PTSD symptoms seems less straightforward and more
dependent on the type of abuse experienced. The current results suggest that boys are
victims of more intrusive acts and that the more severe the abuse is, the more likely they are
to experience PTSD symptoms.
Gauthier-Duchesne et al. Page 7
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
Regarding externalizing symptoms, the current results suggest that compared to girls, boys
express more aggressiveness and delinquency following SA. This is consistent with an
interpretation that suggests that prescribed gender roles may influence SA outcomes
(Banyard et al., 2004). Indeed, it is possible that boys express their negative emotions
through these types of behavior problems, because it is more accepted for boys to show
externalizing behaviors (Kim, Arnold, Fisher, & Zeljo, 2005). Another hypothesis that could
explain why boys express more aggressiveness is the same gender-perpetrator issue
(Banyard et al., 2004). Boys, who fear being stigmatized as homosexuals, may show more
externalizing problems and less PTSD symptoms to correspond to gender norms (Connell,
2014). This interpretation is consistent with the idea that boys are confronted with additional
issues impeding disclosure, such as fear of homophobic stigma (Collin-Vézina, De La
Sablonnière-Griffin, Palmer, & Milne, 2015), and may be less likely to disclose abuse than
girls (Hébert, Tourigny, Cyr, McDuff, & Joly, 2009; O’Leary & Barber, 2008).
The variable that was the most important predictor of SA outcomes in the model is sense of
guilt. Yet, no gender difference was apparent as boys and girls revealed similar means of
sense of guilt. This absence of gender difference may be explained by the young age of
participants, who may not have yet internalized social attributions of blame. Male victims
may have greater guilt regarding the SA during adolescence or adulthood. In our study,
children that revealed higher sense of guilt had more PTSD symptoms and more
externalizing problems. Sense of guilt as a predictor of PTSD symptoms has been observed
among adults victims of SA (Cantón-Cortés et al., 2011; Feiring & Cleland, 2007), but to
our knowledge had not been explored with sexually abused children. As Finkelhor and
Browne (1985) conceptualized, even for children, believing that the abuse is partly one’s
fault exacerbates SA outcomes. Our results suggest that the sense of guilt expressed by child
victims is a better predictor of SA outcomes than SA characteristics.
While it was not a mediator, perpetrator’s relationship to the child predicted externalizing
behavior problems. Our results showed that the more distant the relationship between the
child and the perpetrator, the more likely the child is to present externalizing problems. This
result was unexpected and contrary to what is reported in the literature (Yancey & Hansen,
2010). It is important to remain critical about this result, considering the distribution of the
variable (only 2% of the sample had been abused by a stranger and 73% by a family
member). One hypothesis could be that the more distant perpetrators (e.g. stranger) are more
likely to use force to perpetrate the acts of abuse (Fischer & McDonald, 1998) and that this
violence is associated with greater externalizing symptoms (Yancey & Hansen, 2010).
However, the use of force did not correlate with any of the studied variables. Also, Kendall-
Tackett et al. (1993) have underscored that the label of the perpetrator does not necessarily
reflect the affective relationship between the child and the perpetrator. For example, a close
friend of the family who is known by the child since his birth, may be more significant for
him than his mother’s new boyfriend. When the child is less emotionally attached to his
perpetrator, it is possible that the young victim would feel more comfortable to express
externalizing symptoms to show his difficulties. These symptoms would be more likely to be
muted when the perpetrator is more proximal so as not to hurt or disturb the family,
including the abuser. Similarly to results found by Bal, De Bourdeaudhuij, Crombez, and
Gauthier-Duchesne et al. Page 8
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
Van Oost (2004) in an adolescent sample, the relationship between the perpetrator and the
child was not associated with internalizing problems and PTSD symptoms in our study.
Although duration of the SA is frequently associated with SA symptoms (Yancey & Hansen,
2010), it was not correlated to SA outcomes in this study. This absence of correlation may be
explained by the categorization of the frequency (single episode, less than 6 months, and
more than 6 months). A different categorization could have led to different results.
Otherwise, it is possible that frequency of SA has a low impact on SA symptoms in some
samples. Regardless of the number of SA episodes, being victim of SA is an intrusive and
traumatic event that may influence the child’s behavior.
The descriptive analyses on characteristics of SA showed some differences between boys
and girls. A total of 3 of 4 boys reported experiencing penetration or attempted penetration
compared with 1 of 2 girls. Moreover, most boys were abused by a juvenile, while most girls
were abused by an adult. Boys and girls seem to have experienced the same kind of SA in
regard to duration, perpetrator’s gender and the degree of proximity or relationship to the
perpetrator. Coohey (2010) and Villeneuve Cyr and Hébert (2011) observed that girls were
more often victims over a long period and by a perpetrator in a closer relationship. The
discrepancy in results might be explained by the fact that these studies included fewer than
35 boys.
While boys and girls differed in scores of PTSD symptoms and externalizing problems,
gender was not a predictor of internalizing problems. Studies conducted by Maikovich-Fong
and Jaffee (2010) and Villeneuve Cyr and Hébert (2011) also observed no gender difference
in internalizing problems among child victims of SA. In the current study, no difference was
found for anxiety, depression and somatic complaints. However, parents of boys reported
more withdrawal in their child than parents of girls. It is possible that soon after disclosure
of the SA, boys and girls may show the same level of internalizing problems while gender
differences might only appear at later developmental stages.
4.1. Implications of the study
Results suggest that boys and girls appear to reveal different SA outcomes. The fact that
boys express more externalizing behavior problems compared to girls could allow better
detection of signs associated with a situation of SA. For example, a young school-aged boy
that shows aggressive behaviors during class may be trying to express his distress related to
a traumatic event. Particular attention should also be given to boys who have experienced
more intrusive SA, because they are more likely to display PTSD symptoms. It remains
essential to conduct detailed assessments to identify intervention targets for each child
victim.
Since sense of guilt was the most important predictor of SA outcomes, intervention
strategies for sexually abused children should target this issue, as proposed in
Trauma-
Focused Cognitive Behavior Therapy
(TF-CBT; Cohen, Mannarino, & Deblinger, 2006).
During the therapy sessions, the child is encouraged to recognize that his thoughts have an
impact on how he feels. With the support of the therapist, children reporting feeling guilty
Gauthier-Duchesne et al. Page 9
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
about the SA are helped to realize that they are not responsible for the abuse, which in turn
may reduce symptoms associated with SA.
TF-CBT was tested in a 16 sessions format without the trauma narrative. This version
appeared particularly well suited for reducing externalizing problems (Deblinger,
Mannarino, Cohen, Runyon, & Steer, 2011). The time normally accorded to the narrative
can, when needed, be reinvested in other objectives, such as adequate parenting practices.
The improvement of parental practices may explain the significant reduction of children’s
externalizing problems. This treatment modality could be more adapted for some of the
young boys who show severe externalizing problems and few PTSD symptoms.
4.2. Strengths and limitations of the study
This study makes an important contribution to the literature on child victims of SA by
overcoming some of past studies’ limitations. The sample consisted of children who had
recently disclosed the SA with a significant number of boys. Moreover, mediating variables
have been included in an attempt to explain the differences between boy and girl victims of
SA.
Although this study provides relevant information about gender differences among young
child victims, it has some limitations. First, this cross-sectional study cannot establish a
causal relationship between gender and SA outcomes, nor verify whether these gender
differences are maintained over time. Second, only one mediation effect was validated in the
predictive model, which could explain the low percentage of variance accounted for. Some
important variables that can impact outcomes in SA children were not included in the
present model and as such, future studies should examine coping and parental support as
mediators of the relationship between gender and SA outcomes. Indeed, parental reactions
following disclosure may be different according to the child’s gender, and have an influence
on the child’s symptoms (Ullman & Filipas, 2005). Adding these variables could improve
the understanding of the complex situations experienced by boy and girl victims, and thus
increase the percentage of variance explained. Third, the study did not identify any predictor
of internalizing problems among sexually abused children. In addition, the present analyses
did not consider the possible impact of other forms of maltreatment (physical abuse, neglect,
exposure to interparental violence) in the model of outcomes. Fourth, the scale used to
measure sense of guilt contains a small number of items. To collect more accurate data,
future studies should rely on a more comprehensive scale than can evaluate different aspects
of guilt with greater sensitivity.
Future studies should include a second measurement time to verify if the gender difficulties
persist over time and how trajectories of recovery may be gender-specific. The few
longitudinal studies available suggest that girls report fewer difficulties in the long term but
the difficulties reported by boys are maintained over time (Bernier, Hébert, & Collin-Vézina,
2013). If clinical interventions are focused only on PTSD symptoms, the externalizing
behavior problems of boys may crystallize and accentuate over time.
Gauthier-Duchesne et al. Page 10
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
5. Conclusion
Our goal was to highlight that boys represent a significant proportion of child victims of SA,
and perhaps boys express their pain differently than girls. Including boys in SA studies and
trying to explain gender differences may help to better understand the reality of these young
victims, and thus promote more effective therapeutic and preventive interventions.
Acknowledgments
This research was funded by a grant from the Canadian Institutes of Health Research (#77614) awarded to Martine
Hébert. This work was submitted as part of the Master’s thesis in sexology of the first author. The first author was
supported by graduate scholarships from the Social Sciences and Humanities Research Council (SSHRC), the
Fonds de recherche du Québec - Société et culture
(FRQSC) and the
Chaire interuniversitaire Marie-Vincent sur les
agressions sexuelles envers les enfants
. We wish to thank the families who participated in this study as well as the
practitioners from the different intervention settings involved in this project.
References
Achenbach, TM., Rescorla, LA. Manual for the ASEBA school-age forms and profiles. An integrated
system of multi-informant assessment. Burlignton, VT: University of Vermont, Research Center for
Children, Youth, & Families; 2001.
Arnow BA, Blasey CM, Hunkeler EM, Lee J, Hayward C. Does gender moderate the relationship
between childhood maltreatment and adult depression? Child Maltreatment. 2011; 16(3):175–183.
DOI: 10.1177/1077559511412067 [PubMed: 21727161]
Bal S, De Bourdeaudhuij I, Crombez G, Van Oost P. Differences in trauma symptoms and family
functioning in intra-and extrafamilial sexually abused adolescents. Journal of Interpersonal
Violence. 2004; 19(1):108–123. DOI: 10.1177/0886260503259053 [PubMed: 14680532]
Banyard VL, Williams LM, Siegel JA. Childhood sexual abuse: a gender perspective on context and
consequences. Child Maltreatment. 2004; 9(3):223–238. DOI: 10.1177/107755904266914
[PubMed: 15245676]
Bernier MJ, Hébert M, Collin-Vézina D. Dissociative symptoms over a year in a sample of sexually
abused children. Journal of Trauma & Dissociation. 2013; 14(4):455–472. DOI:
10.1080/15299732.2013.769478 [PubMed: 23796175]
Blain LM, Galovski TE, Robinson T. Gender differences in recovery from posttraumatic stress
disorder: A critical review. Aggresion and Violent Behavior. 2010; 15(6):463–474. DOI: 10.1016/
j.avb.2010.09.001
Cantón-Cortés D, Cantón J, Justicia F, Cortés MR. Un modelo de los efectos del abuso sexual infantil
sobre el estrés post-traumático: el rol mediador de las atribuciones de culpa y afrontamiento de
evitación. Psicothema. 2011; 23(1):66–73. Retrieved from http://www.psicothema.com/pdf/
3851.pdf. [PubMed: 21266144]
Chandy JM, Blum RW, Resnick MD. Gender-specific outcomes for sexually abused adolescents. Child
Abuse & Neglect. 1996; 20(12):1219–1231. DOI: 10.1016/S0145-2134(96)00117-2 [PubMed:
8985612]
Cohen, JA., Mannarino, AP., Deblinger, E. Treating trauma and traumatic grief in children and
adolescents. New York, NY: Guilford Press; 2006.
Collin-Vézina D, De La Sablonnière-Griffin M, Palmer AM, Milne L. A preliminary mapping of
individual, relational, and social factors that impede disclosure of childhood sexual abuse. Child
Abuse & Neglect. 2015; 43:123–134. DOI: 10.1016/j.chiabu.2015.03.010 [PubMed: 25846196]
Connell, R. Masculinités: Enjeux sociaux de l’hégémonie. Paris, France: Éditions Amsterdam; 2014.
Coohey C. Gender differences in internalizing problems among sexually abused early adolescents.
Child Abuse & Neglect. 2010; 34(11):856–862. DOI: 10.1016/j.chiabu.2010.05.001 [PubMed:
20950857]
Gauthier-Duchesne et al. Page 11
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
Deblinger E, Mannarino AP, Cohen JA, Runyon MK, Steer RA. Trauma-focused cognitive behavioral
therapy for children: impact of the trauma narrative and treatment length. Depression & Anxiety.
2011; 28(1):67–75. DOI: 10.1002/da.20744 [PubMed: 20830695]
Dube SR, Anda RF, Whitfield CL, Brown DW, Felitti VJ, Dong M, Giles WH. Long-term
consequences of childhood sexual abuse by gender of victim. American Journal of Preventive
Medecine. 2005; 28(5):430–438. DOI: 10.1016/j.amepre.2005.01.015
Edinburgh L, Saewyc E, Levitt C. Gender differences in extrafamilial sexual abuse experiences among
young teens. The Journal of School Nursing. 2006; 22(5):278–284. DOI:
10.1177/10598405060220050601 [PubMed: 17172200]
Feiring C, Cleland C. Childhood sexual abuse and abuse-specific attributions of blame over 6 years
following discovery. Child Abuse & Neglect. 2007; 31(11–12):1169–1186. DOI: 10.1016/j.chiabu.
2007.03.020 [PubMed: 18023871]
Feiring C, Taska L, Lewis M. Age and gender differences in children’s and adolescents’ adaptation to
sexual abuse. Child Abuse & Neglect. 1999; 23(2):115–128. DOI: 10.1016/
S0145-2134(98)00116-1 [PubMed: 10075182]
Fergusson DM, McLeod GF, Horwood LJ. Childhood sexual abuse and adult developmental outcomes:
findings from a 30-year longitudinal study in New Zealand. Child Abuse & Neglect. 2013; 37(9):
664–674. DOI: 10.1016/j.chiabu.2013.03.013 [PubMed: 23623446]
Finkelhor D, Browne A. The traumatic impact of child sexual abuse: A Conceptualization. American
Journal of Orthopsychiatry. 1985; 55(4):530–541. DOI: 10.1111/j.1939-0025.1985.tb02703.x
[PubMed: 4073225]
Fischer DG, McDonald WL. Characteristics of intrafamilial and extrafamilial child sexual abuse. Child
Abuse & Neglect. 1998; 22:915–929. DOI: 10.1016/S0145-2134(98)00063-5 [PubMed: 9777261]
Gagnier C, Collin-Vézina D. The disclosure experiences of male child sexual abuse survivors. Journal
of Child Sexual Abuse. 2016; 25(2):221–241. DOI: 10.1080/10538712.2016.1124308 [PubMed:
26934546]
Garnefski N, Arends E. Sexual abuse and adolescent maladjustment: differences between male and
female victims. Journal of Adolescence. 1998; 21:99–107. DOI: 10.1006/jado.1997.0132
[PubMed: 9503078]
Hardt J, Rutter M. Validity of adult retrospective reports of adverse childhood experiences: Review of
the evidence. Journal of Child Psychology and Psychiatry. 2004; 45(2):260–273. DOI: 10.1111/j.
1469-7610.2004.00218.x [PubMed: 14982240]
Hébert, M. Les profils et l’évaluation des enfants victimes d’agression sexuelle. In: Hébert, M.Cyr, M.,
Tourigny, M., editors. L’agression sexuelle envers les enfants. Vol. 1. Québec, Canada: Les Presses
de l’Université du Québec; 2011. p. 149-204.
Hébert M, Langevin R, Daigneault I. The association between peer victimization, PTSD, and
dissociation in child victims of sexual abuse. Journal of Affective Disorders. 2016; 193:227–232.
DOI: 10.1016/j.jad.2015.12.080 [PubMed: 26773922]
Hébert M, Tourigny M, Cyr M, McDuff P, Joly J. Prevalence of childhood sexual abuse and timing
disclosure in a representative samples of adults from Quebec. The Canadian Journal of Psychiatry.
2009; 54(9):631–636. DOI: 10.1177/070674370905400908 [PubMed: 19751552]
Hébert M, Tremblay C, Parent N, Daignault IV, Piché C. Correlates of behavioral outcomes in sexually
abused children. Journal of Family Violence. 2006; 21(5):287–299. DOI: 10.1007/
s10896-006-9026-2
Hu LT, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional
criteria versus new alternatives. Structural Equation Modeling. 1999; 6(1):1–55. DOI:
10.1080/10705519909540118
Jöreskog, KG., Sörbom, D. LISREL 8: Structural equation modeling with the SIMPLIS command
language. Hillsdale, NJ, England: Lawrence Erlbaum Associates, Inc; 1993.
Kendall-Tackett KA, Williams LM, Finkelhor D. Impact of sexual abuse on children: A review and
synthesis of recent empirical studies. Psychological Bulletin. 1993; 113(1):164–180. DOI:
10.1037/0033-2909.113.1.164 [PubMed: 8426874]
Gauthier-Duchesne et al. Page 12
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
Kim HJ, Arnold DH, Fisher PH, Zeljo A. Parenting and Preschoolers’ Symptoms as a Function of
Child Gender and SES. Child & Family Behavior Therapy. 2005; 27(2):23–41. DOI: 10.1300/
J019v27n02_03
Maikovich-Fong AK, Jaffee SR. Sex differences in childhood sexual abuse characteristics and victims’
emotional and behavioral problems: findings from a national sample of youth. Child Abuse &
Neglect. 2010; 34(6):429–437. DOI: 10.1016/j.chiabu.2009.10.006 [PubMed: 20400178]
Muthén, LK., Muthén, BO. Mplus User’s Guide. Los Angeles, CA: Muthén & Muthén; 1998–2015.
O’Leary PJ, Barber J. Gender differences in silencing following childhood sexual abuse. Journal of
Child Sexual Abuse. 2008; 17(2):133–143. DOI: 10.1080/1053871080916416 [PubMed:
19042242]
Parent, N., Hébert, M. Questionnaire sur la victimisation de l’enfant. Adaptation française du « History
of Victimization Form » par Wolfe, Gentile & Boudreau (1987). Montréal, Canada: Département
de sexologie, Université du Québec à Montréal; 2006.
Soylu N, Ayaz M, Gökten ES, Alpaslan AH, Dönmez YE, Özcan ÖÖ, Ayaz AB, Tufan AE. Gender
differences in sexually abused children and adolescents: A multicenter study in Turkey. Journal of
Child Sexual Abuse. 2016; 25(4):415–427. DOI: 10.1080/10538712.2016.1143073 [PubMed:
27266537]
Stoltenborgh M, van Ijzendoorn MH, Euser EM, Bakermans-Kranenburg MJ. A global perspective on
child sexual abuse: meta-analysis of prevalence around the world. Child Maltreatment. 2011;
16(2):79–101. DOI: 10.1177/1077559511403920 [PubMed: 21511741]
Tolin DF, Foa EB. Sex differences in trauma and posttraumatic stress disorder: a quantitative review of
25 years of research. Psychological Bulletin. 2006; 132(6):959–992. DOI:
10.1037/0033-2909.132.6.959 [PubMed: 17073529]
Ullman SE, Filipas HH. Gender differences in social reactions to abuse disclosures, post-abuse coping,
and PTSD of child sexual abuse survivors. Child Abuse & Neglect. 2005; 29(7):767–782. DOI:
10.1016/j.chiabu.2005.01.005 [PubMed: 16051351]
Villeneuve Cyr M, Hébert M. Analyse comparative des caractéristiques de l’agression sexuelle et des
conséquences associées en fonction du sexe. Service social. 2011; 57(1):15–30. DOI:
10.7202/1006245ar
Wolfe, VV. The Children’s Impact of Traumatic Events Scale II (CITES-II). 2002. Unpublished
assessment instrument. (Available from V. V. Wolfe, Child and Adolescent Centre, London Health
Sciences Centre, 346 South St., London, Ontario, Canada, N6A 4G5.)
Yancey CT, Hansen DJ. Relationship of personal, familial, and abuse-specific factors with outcome
following childhood sexual abuse. Aggression and Violent Behavior. 2010; 15(6):410–421. DOI:
10.1016/j.avb.2010.07.003
Zajac K, Ralston ME, Smith DW. Maternal support following childhood sexual abuse: associations
with children’s adjustment post-disclosure and at 9-month follow-up. Child Abuse & Neglect.
2015; 44:66–75. DOI: 10.1016/j.chiabu.2015.02.011 [PubMed: 25736053]
Gauthier-Duchesne et al. Page 13
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
Figure 1.
Conceptual mediation model of the correlates of child sexual abuse.
Gauthier-Duchesne et al. Page 14
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
Figure 2.
Mediation model of the correlates of child sexual abuse.
Notes.
The model coefficients are standardized. Only significant paths are illustrated.
Child’s gender has been coded as 0 = girl and 1 = boy.
*
p
< 0.05. **
p
< 0.01. ***
p
< 0.001.
Gauthier-Duchesne et al. Page 15
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author Manuscript
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author
Manuscript
Gauthier-Duchesne et al. Page 16
Table 1
Socio-demographic Characteristics for Girls and Boys
Characteristic Girls (n = 319) Boys (n = 128) Statistical tests p
Mean age (SD) 9.07 (2.02) 8.79 (2.11)
t
(445) = 1.29 .197
Family structure χ2(3) = 6.75 .080
Intact family 18.1% 17.2%
Single-parent family 44.1% 38.3%
Stepfamily 27.9% 25.8%
Foster family 9.9% 18.7%
Family income χ2(3) = 0.84 .839
Less than $20,000 32.8% 33.0%
$20,000 to $39,999 27.6% 28.0%
$40,000 to $59,999 17.8% 14.4%
$60,000 and more 21.8% 24.6%
Mother’s education level χ2(3) = 3.75 .290
Primary school 6.0% 1.7%
High school 45.0% 45.4%
College 36.8% 41.2%
University 12.2% 11.7%
Father’s education level χ2(3) = 2.29 .515
Elementary school 11.1% 7.1%
High school 50.2% 47.5%
College 25.1% 31.3%
University 13.6% 14.1%
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author
Manuscript
Gauthier-Duchesne et al. Page 17
Table 2
SA Characteristics in Percentage (Adjusted Residuals) for Girls and Boys
Characteristic Girls (n = 319)aBoys (n = 128)aStatistical tests p
Severity level of SA χ2(2) = 12.64 .002
Less severe 12.1 (1.5) 7.1 (−1.5)
Severe 31.5 (2.8) 18.3 (−2.8)
Very severe 56.4 (−3.6) 74.6 (3.6)
Frequency of SA χ2(2) = 7.87 .020
Single episode 25.5 (1.5) 18.5 (−1.5)
Some events 36.4 (−2.8) 51.3 (2.8)
Repetitive or chronic 38.1 (1.5) 30.2 (−1.5)
Perp. relationship to the child χ2(3) = 1.36 .716
Immediate family 54.1 (1.0) 48.8 (−1.0)
Extended family 19.2 (−.7) 22.0 (.7)
Family acquaintance 24.5 (−.7) 27.6 (.7)
Stranger 2.2 (.4) 1.6 (−.4)
Perpetrator’s gender χ2(1) = 0.39 .532
Male 93.7 (−.6) 95.2 (.6)
Female 6.3 (.6) 4.8 (−.6)
Perpetrator’s age χ2(3) = 14.45 .002
Less than 15 years old 27.6 (−3.2) 43.6 (3.2)
15–19 years old 11.9 (−1.0) 15.3 (1.0)
20–59 years old 52.2 (2.9) 37.1 (−2.9)
60 years old or more 8.3 (1.6) 4.0 (−1.6)
a
Because of missing data on some variables, number of participants ranges from 302 to 318 for girls and from 119 to 127 for boys.
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author
Manuscript
Gauthier-Duchesne et al. Page 18
Table 3
Mean Differences between Girls and Boys on Outcomes Variables
Variable
Girls Boys
t(425) pCohen’s dM SD M SD
PTSD symptoms 46.09 17.15 39.92 19.38 3.23 .001 0.35
Re-experiencing 12.83 6.68 9.61 7.09 4.41 <.001 0.43
Avoidance 18.30 6.01 16.95 6.81 2.01 .045 0.21
Hyperarousal 14.97 6.86 13.29 7.71 2.20 .028 0.21
Sense of guilt 1.56 1.76 1.36 1.79 1.05 .297 0.11
Internalizing problems 59.55 11.63 61.44 10.29 −1.58 .114 0.17
Anxious/depressed 60.72 9.56 61.82 9.34 −1.09 .276 0.12
Withdrawn 60.05 8.99 63.23 10.18 −3.21 .001 0.34
Somatic complaints 57.84 7.87 57.79 7.20 0.07 .946 0.01
Externalizing problems 60.09 11.84 64.71 10.86 −3.76 <.001 0.40
Rule-breaking 59.94 8.58 62.75 9.16 −3.02 .003 0.32
Aggressive 62.32 10.75 66.98 11.80 −3.97 <.001 0.42
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
PMC Canada Author Manuscript PMC Canada Author Manuscript PMC Canada Author
Manuscript
Gauthier-Duchesne et al. Page 19
Table 4
Summary of Correlations, Means and Standard Deviations for SA Characteristics and Scores on Outcomes
Variables
Variables 1 2 3 4 5 6 M SD
1. Severity of SA - -
2. Frequency of SA .22
***
- -
3. Perp. relationship to the child −.08 −.19
***
- -
4. Sense of guilt .10
*
.01 −.07 . 1.50 1.77
5. PTSD .11
*
−.01 .04 .40
***
44.34 18.00
6. Internalizing problems .04 .06 .08 .03 .16
**
60.11 11.27
7. Externalizing problems .05 .04 .14
**
.11
*
.20
***
.67
***
61.45 11.74
*p
< 0.05.
** p
< 0.01.
*** p
< 0.001.
Child Abuse Negl
. Author manuscript; available in PMC 2018 January 16.
