Content uploaded by Maria Carmen Viana
Author content
All content in this area was uploaded by Maria Carmen Viana on Dec 01, 2017
Content may be subject to copyright.
Pathways from witnessing parental violence during childhood to
involvement in intimate partner violence in adult life: The roles of
depression and substance use
CLARICE S. MADRUGA
1,2
, MARIA CARMEN VIANA
3
, RENATA RIGACCI ABDALLA
1,2
,
RAUL CAETANO
1,4
& RONALDO LARANJEIRA
1,2
1
National Research Institute on Alcohol and Other Drugs, Sao Paulo, Brazil,
2
Department of Psychiatry, Federal University
of São Paulo, Sao Paulo, Brazil,
3
Departament of Social Medicine and Post-Graduate Program in Collective Heath, Federal
University of Espírito Santo, Vitória, Brazil, and
4
Prevention Research Center, Oakland, USA
Abstract
Introduction and Aims. The aims of this study were to determine the prevalence of witnessing parental violence (WPV) during
childhood and of current intimate partner violence (IPV) victimisation and aggression in a Brazilian sample, in order to verify
pathways between WPV and involvement in IPV as an adult. Design and Methods. The mediating roles of substance use
and depression were investigated. Data came from the Second Brazilian National Alcohol and Drugs Survey, a multi-cluster
probabilistic household survey, which gathered information on the use of psychoactive substances, current depressive disorder, history
of childhood direct and indirect exposure to domestic violence and IPV in a nationally representative sample. A subsample of 2120
individuals aged 14 years or older was analysed. Weighted prevalence rates, adjusted odds ratio and conditional path models were
performed. Results. Being a victim of IPV was reported by 6% of the sample. Thus being, 4.1% reported being IPV perpetrators;
these rates were 16.6% and 7.3%, respectively, among those who reported WPV (13%). WPV was associated with being a victim of
IPV in adult life, but not with becoming a perpetrator, regardless of being a victim of physical violence during childhood. There was a
direct effect of WPV on IPV mediated by depressive symptoms. Alcohol and cocaine consumption and age of drinking initiation me-
diated only when combined with depressive symptoms. Discussion and Conclusions. Intergenerational transmission models of
IPV through exposure during childhood can help to explain the high rates of domestic violence in Brazil. Our findings provide
evidence to implement targeted prevention strategies where they are needed most: the victims of premature adverse experiences.
[Madruga CS, Viana MC, Abdalla RR, Caetano R, Laranjeira R. Pathways from witnessing parental violence during
childhood to involvement in intimate partner violence in adult life: The roles of depression and substance use. Drug
Alcohol Rev 2017;36:107-114]
Key words: intimate partner violence, witnessing interparental violence, adverse childhood experiences,
epidemiology, Brazil.
Introduction
Intimate partner violence (IPV) is a major public health
issue across the globe and in developing countries alike,
with worldwide rates ranging from 15% in Japan to
71% in Ethiopia [1]. It is estimated that over one-third
of women around the world have experienced either
intimate partner violence or non-partner sexual violence
in their lifetime. The consequences of IPV go beyond
harming the partner’s physical and/or psychological
well-being. As IPV occurs within the household, it often
affects children and adolescents, either as direct victims
who suffer psychological, physical and/or sexual abuse,
or indirect, when witnessing parental violence (WPV)
[2–4].
Witnessing the perpetration of physical aggression
within the family can be as damaging to children as
personally suffering violence, and both are associated
with mood and anxiety disorders. This includes post-
Clarice S. Madruga MSc, PhD, Research Associate, Maria Carmen Viana MD, PhD, Professor, Renata Rigacci Abdalla MD, Phd Student, Raul
Caetano MD, PhD, Deputy Director, Ronaldo Laranjeira MD, PhD, Deputy Director. Correspondence to Dr Clarice S Madruga,
INPAD/UNIFESP, Psychiatry Department, Federal University of São Paulo, 570/82 Borges Lagoa, Sao Paulo, SP 04038000, Brazil.
Tel: +55 11 999700973; E-mail: clarice@uniad.org.br
Received 2 December 2015; accepted for publication 18 October 2016.
© 2017 Australasian Professional Society on Alcohol and other Drugs
Drug and Alcohol Review (January 2017), 36, 107–114
DOI: 10.1111/dar.12514
bs_bs_banner
REVIEW
traumatic stress disorder [3], and can also trigger
conduct disorders that may progress to the perpetration
of IPV in later stages of adult life, during which
cohabiting relationships are established [2,5–7].
There is a large body of evidence linking exposure to
adverse childhood events (ACE) with premature use of
psychotropic substances during adolescence and their
misuse in adult life [8–12]. Previous studies have also
shown the association between the consumption of alco-
hol and illegal drugs with the occurrence of IPV [13–17].
However, there are far fewer studies investigating the
association between WPV as a child and becoming
exposed to IPV as a victim and/or perpetrator in adult life
[2,18], regardless of having been a victim of physical
abuse from a parent in childhood. Finally, to the authors’
knowledge, no previous studies have yet explored the
mediating effects of depression in the relationship
between WPV as a child and being involved with IPV
later in life, and how the use of substances might affect
this relationship. Given the literature gap on this topic,
the aim of this work was to describe national rates on
exposure to IPV during childhood and in adult life.
Further, we investigate the predictive value of WPV
during childhood and the occurrence of IPV in adult life,
controlling for being a victim of direct physical violence
as a child. The pathways between WPV and IPV were
estimated in a conditional model that considers the
mediating effects of depressive disorder, alcohol
consumption, age of drinking initiation and frequency
of cocaine use in the previous year (cocaine was the only
illicit substance assessed in the conditional model as it
was the most commonly used illicit substance among this
population, and based on the fact that Brazil is among the
countries with the highest rates of cocaine use in the
world [19]). A better understanding of the pathways
leading to IPV could indicate priorities in the
development of more efficient prevention strategies and
management procedures.
Methods
This research protocol was approved by the Ethics
Committee of the Federal University of São Paulo and
by the National Commission of Ethics in Research. All
subjects provided written informed consent prior to the
interview.
Sampling and procedures
The Second Brazilian National Alcohol and Drugs
Survey was conducted between November 2011 and
March 2012 [20]. A multistage cluster sampling
procedure was used to select 4607 individuals aged
14 years and older from the Brazilian household
population, including an oversampling of 11 57
adolescents (14 to 18 years old). The overall response
rate was 77% and 79% for the adolescent sample. The
sampling process was conducted in three steps: (i) selec-
tion of 149 counties using probability proportional to size
methods; (ii) selection of two census sectors for each
county, totalling 375 census sectors, also using probabil-
ity proportional to size methods; and (iii) within each
census sector, eight households were selected by simple
random sampling, followed by the selection of a house-
hold member to be interviewed using the ‘the closest
future birthday’technique. One-hour, face-to-face
interviews were conducted in the respondents’home by
trained interviewers using a standardised fully structured
questionnaire. This study analysed a subsample of
2120 individuals (46% of the total sample of survey
respondents) who were married or living with their
partners.
Measurements
Adverse childhood experiences. Having been a victim of
physical violence perpetrated by a parent during
childhood was measured using seven items from the
parent–child version of the Conflict Tactics Scale
[21,22]. The items covered the following types of
physical aggression: ‘During your childhood or adolescence,
were you ever 1) Insulted or humiliated publicly? 2) Hurt with
an object? 3) Pushed, scratched, pinched or knocked over? 4)
Burned or scalded with boiling water? 5) Threatened with a
knife or a gun? 6) Shot with a gun or attacked with a knife?
7) Hit until bruised?’. Responses were scored 0 (never), 1
(sometimes), 2 (often)and3(very often). Scores were
summed up to create a violence index with a possible
range of 0 to 21.
Witnessing parental violence. WPV was based on the fol-
lowing items of the Conflict Tactics Scale: ‘During your
childhood or adolescence, how frequently did you see your
parents threatening to harm each other or others?’and ‘During
your childhood or adolescence, how frequently did you see your
parents physically harm each other or others?’with the same
scores for responses ranging from ‘never’to ‘very often’(0
to 3) as mentioned previously for the assessment of
adverse childhood events. The variable WPV was created
through the sum of the two items with a possible range of
0to6.
Intimate partner violence—victimisation and perpetration. All
questions were from the Conflict Tactics Scale
[21,23], Form R. Respondents were asked a total of nine
questions about the occurrence of different types of
violent behaviours in the last 12 months, including less
severe physical violence (throwing something; pushing,
108 C. S. Madruga et al.
© 2017 Australasian Professional Society on Alcohol and other Drugs
grabbing or shoving; slapping) and severe physical and/or
sexual violence (kicking, biting or hitting; trying to hit
with something; burning or scalding; forced sex; threat-
ening with a knife or gun; using a knife or gun). First,
the respondents were asked if they had perpetrated
any of these acts against their partner (perpetration),
and then were asked to report if their partner had per-
petrated any of these acts against them (victimisation).
Cocaine use assessment. Lifetime and previous year use
and frequency of cocaine consumption were assessed.
Frequency responses were scored as 0 (never), 1 (one to
two times in the week), 2 (three to five times in the week)or
3(six times a week or every day). In order to guarantee
confidentiality, all illicit drug related questions were not
asked face-to-face, but self-reported separately by the
participant alone, using a standard form, which was
collected at the end of the interview in sealed envelopes
and immediately placed in sealed bags in front of the
respondent.
Alcohol consumption. The age of initiation of alcohol
consumption was assessed using the question ‘At what
age did you start drinking alcohol (do not consider when you
only tried one or two sips)’. The number of drinks
consumed in a typical day (alcohol intake) was also
measured with the assistance of a unit/drinks demonstra-
tion chart [24,25].
Diagnostic and Statistical Manual of Mental Disorders, Fifth
Edition (DSM-5) Alcohol Use Disorder. This was assessed with
the Brazilian version of the Composite International
Diagnostic Interview (CIDI 2.1) [26]. Although this
survey pre-dates DSM-5, the questionnaire included
questions about craving, which allowed for the creation
of a diagnosis based on DSM-5 criteria, covering the 11
criteria included in the DSM-5. In the analysis herein,
the presence of two or more criteria in the past 12 months
was considered a positive diagnosis of alcohol use
disorder.
Depressive symptoms. Assessed using the Brazilian vali-
dated version of the 20-item Center for Epidemiological
Studies Depression Scale (CES-D). The CES-D Scale
measures the experience of depressive symptomatology
during the previous two week. The measure was
developed from items appearing on longer, well-validated
depression scales. The items assess cognitive, affective,
behavioural and somatic symptoms of depression, and
positive affect. Each item is rated on a 4-point scale
ranging from 0 = rarely or none of the time (less than
1 day) to 3 = most or all of the time (5–7 days). A total
score is then calculated by summing the responses after
reversing the positive affect items. Higher scores reflect
greater levels of depressive symptomatology. Radloff
[27] reported good internal consistency for the measure,
with Cronbach’s alpha coefficients of 0.84–0.85 in White
community samples and 0.90 in clinical samples. There
also was strong evidence for validity. The CES-D
discriminates between psychiatric inpatient and general
population samples, and among levels of severity within
patient groups; and is associated with other measures of
depressive symptomatology. A score of 16 or above was
considered as the cutoff point [28,29] for case indica-
tion of depressive disorder [28,29], and was used in
the multivariate analysis. The total score (ranging from
0 to 80) was considered an index accounting for the
presence and severity of depressive symptoms, and
was included as a mediating factor in the conditional
model.
Control variables. Included demographic variables (sex,
age, education/number of years attending school and
monthly income) and the score for suffering physical
violence during childhood within the household
(obtained through the adverse childhood events
assessment).
Statistical analysis
In an attempt to account for the complex sampling
design, data was weighed for the inverse probability of
respondents’selection, including the oversampling of
adolescents, and a post-stratification weight was applied
to correct for non-response and to adjust both samples
to known population distributions on demographic
variables (education, age, gender and region of the
country) according to the Brazilian Census of 2010.
Weighted prevalences and bivariate associations were
estimated using STATA 13.0 [30]. The conditional
analysis was performed using SPSS21/PROCESS
(‘processmacro.org’macro v2.14.), which is a computa-
tional procedure that implements moderation or
mediation analysis, as well as their combination in an
integrated conditional process model (i.e. mediated
moderation and moderated mediation). It uses a path
analysis framework similar to the approach described by
Edwards and Lambert [31,32]. Cross-tabulations were
used to examine prevalence rates of WPV by the main
risk factors (depressive and alcohol disorders, cocaine
use and IPV perpetration and victimisation). A multivar-
iate analysis using Logistic regression models was used to
assess the independent associations between WPV and
IPV, controlling for sociodemographic factors, cocaine
use and depressive symptoms.
Conditional modelling. The hypothesis was to determine
whether the direct association between WPV and IPV in
adult life could be mediated by: (i) depressive symptoms
score (measured by the occurrence and severity of
Pathways of witnessing parental violence 109
© 2017 Australasian Professional Society on Alcohol and other Drugs
depressive symptoms from the CES-D scale); (ii) alcohol
intake (measured by amount of drinks consumed in a
typical day); (iii) age of drinking onset; and (iv) frequency
of cocaine use. Several hypotheses were tested to define
the conditional pathways between WPV as a child and
being involved with IPV, as a victim or a perpetrator in
adult life. The Parallel Multiple Mediator Model was
chosen, as it allows many causal effects operating through
multiple mechanisms simultaneously [33,34], presenting
coefficients for all direct and indirect paths tested with
low Pvalues, even though the R square value was
considered low (30%) [35]. WPV was considered as the
predictor (X), and the two forms of IPV involvement as
the outcome (Y)—as a victim (IPV-V) and as a
perpetrator (IPV-P). These associations were tested with
the four mediators in thepath. All models were calculated
as weighted linear composites of scale items and
controlled by the covariates sex, age, income and being
a victim of parental/household physical aggression during
childhood. The mediations were conducted to estimate
the effect of the four mediators in the relation between
WPV and IPV-V and IPV-P, using the product of coeffi-
cients method [36], which involves the multiplication of
regression coefficients for the regression of the mediator
on the independent variable (a-path) and for the regres-
sion of the outcome on the mediator (b-path) with the
independent variable included in the model (c-path),
and with a*b considered the mediated effect. All
mediation effects were estimated in Process using a
maximum likelihood estimator and 10 000 bootstrap
draws to obtain confidence intervals for the indirect
effect. All mediation models were evaluated using multi-
ple indices of model fit: a non-significant χ²-statistic,
comparative fit index values greater than 0.95 and
standardised root mean square residual values less than
0.08 [37].
Results
Descriptive analysis
Among the subsample of individuals who were mar-
ried or cohabiting (N= 2120), 52% were female, with
most (71%) aged between 26 and 59 years (mean age
41 years; SD = 15.0) and earned an average monthly
income of £125.00 (R$766.00, equivalent to 1.3 times
the minimum monthly wage in Brazil at the time of
the survey—2012). More than half (53.2%) reported
having completed primary school and 10.2% had a
college or a university degree (data available upon
request).
Thirteen percent reported WPV at some point during
their childhood (Table 1), and over half of the sample
were victims of physical aggression within the house-
hold during their childhood (57.6%). Among individ-
uals who reported being victims of physical aggression
as a child, about one-third also witnessed parental
physical violence (32.2%—data not shown); those who
WPV were six times more likely to have been victims
of childhood physical violence. The prevalence rate of
being a victim of IPV was over three times higher
among those who witnessed parental violence during
childhood compared to the total sample (16.6 vs. 6%,
Table 1). Those who reported WPV were nearly four
times as likely to become a victim of IPV in adulthood
(odds ratio 3.9; 95% confidence interval 2.4–6.2). The
prevalence of being a perpetrator of IPV increased from
4.1% to 7.3% among those who WPV compared to the
Table 1. Descriptive and multivariate analysis of IPV, alcohol abuse and depressive disorder in the whole sample and among individuals exposed
to WPV during childhood, in the general household population living in Brazil (N =2120)
Whole sample % [95% CI] Witnessed parental violence %; OR [95% CI]
a
100 13.0
Childhood maltreatment 20.9 [18.2–23.8] 57.6; 6.0 [4.7–7.6]
IPV victim 6.0 [4.9–7.3] 16.6; 3.9 [2.4–6.2]
IPV perpetrator 4.1 [3.1–5.4] 7.3; 1.6 [0.8–3.3]
Depressive disorder (CES-D) 23.0 [19.9–26.5] 35.5; 2.4 [1.9–3.1]
Cocaine use 1.1 [0.6–1.7] 3.4; 3.9 [1.9–7.7]
Alcohol use disorder (DSM-V) 9.9 [8.2–11.8] 14.3; 2.0 [1.5–2.9]
Mean ± SE Mean ± SE; IRR [95%CI]
b
Age of alcohol use onset 17.4 ± 5.5 16.2 ± 5.7; 0.94 [0.90–0.97]
a
Logistic regression, adjusted for sociodemographic characteristics (sex, age, income, education).
b
Poisson regression, adjusted for
sociodemographic characteristics (sex, age, income, education). Bold indicates statistically significant associations. CES-D, Center
for Epidemiological Studies Depression Scale; CI, confidence interval; DSM-V, Diagnostic and Statistical Manual of Mental Disor-
ders, Fifth Edition; IPV, intimate partner violence; OR, odds ratio; WPV, witnessing parental violence.
110 C. S. Madruga et al.
© 2017 Australasian Professional Society on Alcohol and other Drugs
whole sample, but this difference was not significant in
any of these multivariate analyses (Table 1). All
variables associated with being involved in IPV in adult
life were also associated with WPV during childhood
(depressive disorder, cocaine use and alcohol use
disorder). Over one-third of individuals who reported
WPV were currently depressed (odds ratio 2.4; 95%
confidence interval 1.9–3.1). Those who WPV were
3.9 times as likely to use cocaine and twice as likely to
have a 12-month DSM-5 alcohol use disorder
compared to the general population, and although they
started drinking at a younger age, this difference was
not statistically significant.
Multivariate analysis
Four multivariate analysis models were performed to ex-
plore the association between WPV as a child and being
involved in IPV as a victim or as a perpetrator in the last
12 months (Table 2). The first model was adjusted by
the sociodemographic variables and history of being a vic-
tim of parental physical violence during childhood. This
model found that WPV was associated with being a vic-
tim of IPV in adult life, but not with becoming a perpetra-
tor of IPV. The same was found in the other three
models, including adjustment for depressive symptoms
(CES-D total score), alcohol consumption (number of
drinks in a typical day) and cocaine use in the previous
year. WPV was a predictor of being a victim of IPV in
all models tested, with odds ratios ranging from 2.9 to
3.7. WPV was not associated with becoming a perpetra-
tor in marital/cohabiting relationships in adult life. All
models were also adjusted for being a victim of physical
violence as a child.
Conditional model
The Parallel Multiple Mediator Model proposed was
used to test only the outcome of being a victim of IPV
(IPV-V), because being a perpetrator of IPV was not sta-
tistically associated (P>0.05) with WPV in the
multivariate analysis. The model tested four possible me-
diators in the association between the predictor X (WPV)
and the outcome Y (IPV-V): depressive symptoms (CES-
D total score, accounting for the presence and severity of
symptoms—DSI), alcohol consumption (number of
drinks consumed in a typical day—AC), age of drinking
initiation (AI) and frequency of cocaine consumption
(number of days in a month—CC).
Being a victim of IPV in adult life. As seen in Figure 1, the
conditional analysis demonstrated that being a victim of
IPV is directly predicted by WPV during childhood
(P= 0.0071). Among all the combinations of possible
indirect effects of WPV on IPV-V, considering the
effects of the four mediators tested, only three pathways
were significant (P<0.05). The model indicates a direct
effect of WPV on IPV, as well as an indirect effect on IPV
via DSI. WPV did not affect age of alcohol initiation nor
alcohol consumption, even though the latter did affect
IPV. WPV affected cocaine consumption; however, it
was not a mediator of its association with IPV. Interest-
ingly, all valid indirect paths had the depressive
symptoms index (DSI) in their route, mediating the
effect of WPV on IPV. The total indirect effect of
0.0306 was significant (Boot LLCI: 0.0073 and Boot
ULCI: 0.0651). All valid paths are described below and
illustrated in the model shown in Figure 1:
Path 1: WPV –DSI –IPV-V.
Path 3: WPV –DSI –AC –IPV-V;
Path 6: WPV –DSI –AC - CC –IPV-V;
Path 7: WPV –DSI –AI –CC –IPV-V.
Discussion
Based on a nationally representative sample, our findings
demonstrated that 13% of Brazilians witnessed parental
violence during childhood and, among those, almost
60% were also victims of direct physical violence as a
child within the household. On the other hand, among
those who reported being victims of physical violence in
childhood (21% of the population), less than one-third
Table 2. Multivariate logistic regression analysis between witnessing parental violence as a child and involvement with IPV in adult life, in the
general household population living in Brazil (N = 2120).
Model 1 OR [95%CI] Model 2 OR [95%CI] Model 3 OR [95%CI] Model 4 OR [95%CI]
IPV victim 3.1 [2.3–6.0] 2.9 [1.7–4.9] 3.3 [1.8–6.3] 3.7 [2.3–6.0]
P=0.00 P=0.00 P=0.00 P=0.00
IPV perpetrator 1.6 [0.8–3.2] 1.3 [0.6–2.8] 1.7 [0.7–4.6] 1.5 [0.7–3.2]
P=0.23 P=0.46 P=0.26 P=0.24
M1: Adjustment for sociodemographics (sex, age, income, education) + childhood physical violence. M2: Adjustment for
sociodemographics + childhood physical violence + depressive symptoms. M3: Adjustment for sociodemographics + childhood
physical violence + alcohol consumption. M4: Adjustment for sociodemographics + childhood physical violence + cocaine use. CI,
confidence interval; IPV, intimate partner violence; OR, odds ratio.
Pathways of witnessing parental violence 111
© 2017 Australasian Professional Society on Alcohol and other Drugs
had also witnessed parental violence. Similar results were
also reported by Song and colleagues [2], assessing a
representative sample of household residents in South
Korea, China, the Philippines and other countries
[38,39]. The present study arises from the premise that,
even though being a victim of physical violence and
witnessing parental violence as a child are highly associ-
ated, and probably share the same underlying factors,
the victims’profile might be distinct, and therefore, its
associated factors and long-term consequences should
be studied separately.
There is a large body of evidence describing the long-
term consequences of early life exposure to violence
[40,41], its association with substance use and mental
health disorders [9,12], and IPV later in life, which has
been referred to as intergenerational transmission of
IPV [2,7,14,16,42]. However, the assumption that the
consequences of direct childhood exposure to personal
violence superimpose the long-term effect of being a wit-
ness of parental violence is not entirely true. We tested the
hypothesis that witnessing parental violence can indepen-
dently lead to harmful consequences regardless of having
suffered violence directly. Our results suggest that being a
victim of direct violence (physical violence, as assessed
here) and witnessing parental violence do not necessarily
overlap, and that witnessing parental violence is indepen-
dently associated with being a victim of IPV later in life.
However, perpetration of IPV in adulthood was not pre-
dicted by WPV, when controlling for having suffered di-
rect violence as a child, although this is a significant
predictor of both, victimisation and perpetration [2,7].
The odds of being a victim of IPV as an adult increased
four-fold among those who had experienced WPV during
childhood, with nearly two in 10 individuals reporting
IPV, compared to 6% among the general population.
The multivariate analysis suggested that these individuals
are also twice as likely to report depressive symptoms and
over three times as likely to use cocaine as adults. Being a
victim of physical violence also doubled the chances of
developing alcohol use disorders and reduced the age
of drinking initiation. WPV predicted being a victim of
IPV after adjusting for sex, age, education and
socioeconomic status. This association remained
significant even when further adjusted separately for
depressive symptoms, cocaine use and alcohol use. More
importantly, all associations remained significant
regardless of having suffered direct physical violence as
a child. However, WPV did not predict becoming a
perpetrator of intimate partner violence, even when
adjusting for the presence of depressive symptoms and
substance use.
Based on the findings from the multivariate analysis,
the conditional model was performed to estimate the
direct and indirect effects of WPV on IPV-V, testing
the role of depressive symptoms, alcohol consumption,
age of drinking initiation and cocaine consumption as
possible mediators. A parallel multiple mediator model
demonstrated that witnessing parental violence as a
child had a direct effect on becoming a victim of IPV
later in life. This effect was shown to be mediated by
symptoms of depression (DSI), with three other possible
indirect paths, involving DSI and alcohol consumption;
DSI, alcohol and cocaine consumption; and DSI, early
drinking initiation and cocaine consumption. All these
relationships occur independently from having suffered
direct physical violence as a child. There is an extensive
body of evidence showing that being exposed to adverse
events during childhood is linked with the development
of mood disorders [12,40,43]. Interestingly, all
significant indirect paths involving earlier drinking
initiation, alcohol and cocaine consumption did not
mediate the association between WPV and IPV-V in
Figure 1. Conditional model illustration for the pathways of witnessing parental violence (WPV) during childhood predicting intimate partner violence
victimisation (IPV-V) in adult life.
112 C. S. Madruga et al.
© 2017 Australasian Professional Society on Alcohol and other Drugs
pathways that did not consider depressive symptoms in
its route. This finding suggests that, differing from
exposure to direct physical violence as a child, the
experience of witnessing parental violence might not
activate the HPA stress axis at levels that can lead to
permanent physiological changes, increasing vulnerabil-
ity to developing drug use disorders regardless of a
depressive disorder being present [43,44]. We suggest
that the experience of witnessing parental violence could
increase the chances of being a victim of IPV possibly
via social and role learning [45] than via its impact in
the physiological mechanisms involved in addiction.
However, all the indirect pathways between the
predictor and the outcome that involved the combina-
tion between one of the drug use variables (alcohol
consumption, age of drinking initiation and cocaine
use) and the variable depressive symptoms were
significant. This result is consistent with the extensive
body of evidence acknowledging the high association
between misuse of substances and mood disorders
[46,47]. Because of the cross-sectional nature of this
study, we are unable to establish a temporal order for
depressive symptoms and substance consumption.
Nevertheless, this comorbidity may play an important
role in the establishment and/or maintenance of violence
within intimate relationships.
Conclusions
Our findings suggest that witnessing parental violence
during childhood can increase the chances of being a
victim of intimate partner violence in adult life, regardless
of having been directly exposed to domestic violence as a
child. However, WPV does not predict perpetrating IPV
in adult life. The association between witnessing parental
violence and being a victim of IPV is mediated by
depressive symptoms, and when they are present, the
consumption of alcohol and cocaine can also play a role
mediating this relationship.
Intergenerational transmission of IPV can, in part,
explain the staggeringly high rates of domestic violence
in Brazil, childhood maltreatment and IPV alike.
Immediate prevention actions are of a foremost
importance. Our findings provide the evidence needed
to focus prevention strategies where they are needed
most: the victims of adverse experiences in childhood.
Role of funding source
This study was supported by grants from CNPq during
the design and conduct of the survey and from CAPES
during the stages of data analyses and interpretation.
References
[1] WHO. Global and regional estimates of violence against women: preva-
lence and health effects of intimate partner violence and non-partner sex-
ual violence. Geneve: World Health Organization, London School of
Hygiene and Tropical Medicine, South African Medical Research
Council; 2013.
[2] Song A, Wenzel SL, Kim JY, Nam B. Experience of domestic violence
during childhood, intimate partner violence, and the deterrent effect of
awareness of legal consequences. J Interpers Violence 2015 [Epub ahead
of print].
[3] Cohodes E, Hagan M, Narayan A, Lieberman A. Matched trauma: the role
of parents’and children’s matched experiences of childhood trauma in
parents’report of children’s trauma-related symptomatology. J Trauma
Dissociation 2016;17:81–96.
[4] Warner TD, Swisher RR. The effect of direct and indirect exposure to
violence on youth survival expectations. J Adolesc Health 2014;55:817–822.
[5] Maldonado RC, Watkins LE, DiLillo D. The interplay of trait anger,
childhood physical abuse, and alcohol consumption in predicting intimate
partner aggression. J Interpers Violence 2015;30:1112–1127.
[6] Fonseka RW, Minnis AM, Gomez AM. Impact of adverse childhood
experiences on intimate partner violence perpetration among Sri Lankan
men. PLoS One 2015;10:e0136321.
[7] Widom CS, Czaja S, Dutton MA. Child abuse and neglect and intimate
partner violence victimization and perpetration: a prospective investigation.
Child Abuse Negl 2014;38:650–663.
[8] Monnat SM, Chandler RF. Long term physical health consequences of
adverse childhood experiences. Sociol Q 2015;56:723–752.
[9] K oskenvuo K, Koskenvuo M. Childhood adversities predict strongly the use
of psychotropic drugs in adulthood: a population-based cohort study of
24,284 Finns. J Epidemiol Community Health 2015;69:354–360.
[10] Van Niel C, Pachter LM,Wade R Jr, Felitti VJ, Stein MT. Adverse events in
children: predictors of adult physical and mental conditions. J Dev Behav
Pediatr 2014;35:549–551.
[11] De Venter M, Demyttenaere K, Bruffaerts R. The relationship between
adverse childhood experiences and mental health in adulthood. A systematic
literature review. Tijdschr Psychiatr 2013;55:259–268.
[12] Madruga CS, Laranjeira R, Caetano R, et al. Early life exposure to violence
and substance misuse in adulthood—the first Brazilian national survey.
Addict Behav 2011;36:251–255.
[13] Okuda M, Olfson M, Wang S, Rubio JM, Xu Y, Blanco C. Correlates of
intimate partner violenceperpetration: resultsfrom a national epidemiologic
survey. J Trauma Stress 2015;28:49–56.
[14] Ulloa EC, Hammett JF. The effect of gender and perpetrator–victim role on
mental health outcomes and risk behaviors associated with intimate partner
violence. J Interpers Violence 2016;31:1184–1207.
[15] Stockl H, Penhale B. Intimate partner violence and its association with
physical and mental health symptoms among older women in Germany. J
Interpers Violence 2015;30:3089–3111.
[16] Singh V, Tolman R, Walton M, Chermack S, Cunningham R. Characteris-
tics of men who perpetrate intimate partner violence. J Am Board Fam Med
2014;27:661–668.
[17] AbramskyT, Watts CH, Garcia-MorenoC, et al. What factors are associated
with recent intimate partner violence? Findings from the WHO multi-
country study on women’s health and domestic violence. BMC Public
Health 2011;11:109.
[18] Whitfield CL, Anda RF, DubeSR, Felitti VJ. Violent childhood experiences
and the risk of intimate partner violence in adults: assessment in a large
health maintenance organization. J Interpers Violence 2003;18:166–185.
[19] Abdalla RR, Madruga CS, Ribeiro M, Pinsky I, Caetano R, Laranjeira
R. Prevalence of cocaine use in Brazil: data from the II Brazilian
National Alcohol and Drugs Survey (BNADS). Addict Behav
2014;39:297–301.
[20] Caetano R,Mills B, Madruga C, Pinsky I, Laranjeira R. Discrepant trendsin
income, drinking, and alcohol problems in an emergent economy: Brazil
2006 to 2012. Alcohol Clin Exp Res 2015;39:863–871.
[21] Straus MA, Hamby SL, FinkelhorD, Moore DW, Runyan D. Identification
of child maltreatment with the Parent–Child Conflict Tactics Scales:
development and psychometric data for a national sample of American
parents. Child Abuse Negl 1998;22:249–270.
[22] Reichenheim ME, Moraes CL. Psychometric properties of the Portuguese
version of the Conflict Tactics Scales: Parent–Child Version (CTSPC) used
to identify child abuse. Cad Saude Publica 2006;22:503–515.
Pathways of witnessing parental violence 113
© 2017 Australasian Professional Society on Alcohol and other Drugs
[23] Moraes CL, Reichenheim ME. Cross-cultural measurement equivalence of
the Revised Conflict Tactics Scales (CTS2) Portuguese version used to
identify violence within couples. Cad Saude Publica 2002;18:783–796.
[24] Dawson DA, Room R. Towards agreement on ways to measure and report
drinking patterns and alcohol-related problems in adult general population
surveys: the Skarpo conference overview. J Subst Abuse 2000;12:1–21.
[25] Laranjeira R, Pinsky I, Sanches M, Zaleski M, Caetano R. Alcohol use
patterns among Brazilian adults. Rev Bras Psiquiatr 2010;32:231–241.
[26] Quintana MIAS, Jorge MR, Gasta lFL, Miranda CT. The reliability of the
Brazilianversion of the Composite International Diagnostic Interview (CIDI
2.1). Braz J Med Biol Res 2004;37:1739–1745.
[27] Radloff LS. The CES-D Scale: a self-report depression scale for research in
the general population. Appl Psychol Measur 1977;1:385–401.
[28] Batistoni SST, Neri AL, Cupertino APFB. Validity of the Center for
Epidemiological Studies Depression Scale among Brazilian elderly. Rev
Saude Publica 2007;41:598–605.
[29] Bradley KL, Bagnell AL, Brannen CL. Factorial validity of the Center for
Epidemiological Studies Depression 10 in adolescents. Issues Ment Health
Nurs 2010;31:408–412.
[30] Corp S. Stata statistical software. StataCorp LP: College Station, TX, 2013.
[31] Edwards JR, Lambert LS. Methods for integrating moderation and
mediation: a general analytical framework using moderated path analysis.
Psychol Methods 2007;12:1–22.
[32] Hayes A, Preacher KJ. Statistical mediation analysis with a multicategorical
independent variable. Br J Math Stat Psychol 2014;67:451–470.
[33] Preacher KJ. Multilevel SEM strategies for evaluating mediation in
three-level data. Multivariate Behav Res 2011;46:691–731.
[34] Hayes AF, Preacher KJ. Statistical mediation analysis with a multicategorical
independent variable. Br J Math Stat Psychol 2014;67:451–470.
[35] Hayes AF. PROCESS: a versatile computational tool for observed variable
mediation, moderation, and conditional process modeling. 2012.
[36] MackinnonDP, Fairchild AJ. Currentdirections in mediation analysis. Curr
Dir Psychol Sci 2009;18:16.
[37] Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure
analysis: conventional criteria versus new alternatives. Struct Equ Modeling
1999;6:1–55.
[38] Jin X, Eagle M, Yoshioka M. Early exposure to violence in the family of
origin and positive attitudes towards marital violence: Chinese immigrant
male batterers vs. controls. J Fam Violence 2007;22:211–222.
[39] Fehringer JA, Hindin MJ. Like parent, like child: intergenerational
transmission of partner violence in Cebu, the Philippines. J Adolesc Health
2009;44:363–371.
[40] Hovdestad W, Campeau A, Potter D, Tonmyr L. A systematic review of
childhood maltreatment assessments in population-representative surveys
since 1990. PLoS One 2015;10.
[41] Turecki G, Ota VK, Belangero SI, Jackowski A, Kaufman J. Early life
adversity, genomic plasticity, and psychopathology. Lancet Psychiatry
2014;1:461–466.
[42] Bell KM, Higgins L. The impact of childhood emotional abuse and
experiential avoidance on maladaptive problem solving and intimate partner
violence. Behav Sci (Basel) 2015;5:154–175.
[43] Englund MM, Egeland B, Oliva EM, Collins WA. Childhood and
adolescent predictors of heavy drinking and alcohol use disorders in early
adulthood: a longitudinal developmental analysis. Addiction
2008;103:23–35.
[44] Dubow EF, Boxer P, Huesmann LR. Childhood and adolescent predictors
of early and middle adulthood alcohol use and problem drinking: the
Columbia County Longitudinal Study. Addiction 2008;103:36–47.
[45] Bandura. Social learning theory. Englewood Cliffs, NJ: Prentice-Hall, 1977.
[46] Foulds JA, Adamson SJ, Boden JM, Williman JA, Mulder RT. Depression
in patients with alcohol use disorders: systematic review and meta-analysis
of outcomes for independent and substance-induced disorders. J Affect
Disord 2015;185:47–59.
[47] Narvaez JC, Jansen K, Pinheiro RT, et al. Psychiatric and substance-use
comorbidities associated with lifetime crack cocaine use in young adults in
the general population. Compr Psychiatry 2014;55:1369–1376.
114 C. S. Madruga et al.
© 2017 Australasian Professional Society on Alcohol and other Drugs
