Brunton y Dryer (2022)

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Social Science & Medicine 292 (2022) 114334

Contents lists available at ScienceDirect

Social Science & Medicine


journal homepage: www.elsevier.com/locate/socscimed

Sexual violence and Australian women: A longitudinal analysis of


psychosocial and behavioral outcomes
Robyn Brunton a, *, Rachel Dryer b
a
Charles Sturt University, Bathurst, NSW, 2795, Australia
b
Australian Catholic University, Strathfield, NSW, 2135, Australia

A R T I C L E I N F O A B S T R A C T

Keywords: Rationale: Sexual violence (SV) is associated with adverse psychosocial and behavioural outcomes with revic­
Sexual assault timization likely. However, there are significant gaps in the current literature in regard to (a) whether over time
Sexual violence women’s levels of distress/behaviour change, and (b) whether social support mediates the relationship between
Rape
SV and adverse outcomes.
Child abuse
Intimate partner violence
Methods and results: This study aimed to address these two issues by analysing data from the Australian Longi­
Mental health tudinal Women’s Health Survey, surveys 4 (2006) to 8 (2018). Using repeated-measures data analytic procedures
Substance abuse we found that women who had suffered SV, in comparison to women with no SV history, had greater anxiety,
depression, stress, a lower mental health-related quality of life and less life satisfaction. Moreover, their level of
distress remained higher at all time points, in comparison to the other group. Women with a SV history were also
more likely to suffer re-victimization, consume more cigarettes and illicit drugs than other women. Moreover, SV
predicted all psychosocial outcomes (except life satisfaction) 12 years later, with social support mediating these
relationships. SV predicted drug status; however social support did not mediate this relationship.
Conclusions: These findings suggest that for women who have experienced SV their distress levels remain
elevated. The findings also provide additional insights into the long-term impact of SV in Australian women with
social support being identified as a resource that may assist in reducing some of the negative psychological
outcomes associated with SV.

1. Introduction consequences that many victims suffer (Walsh et al., 2012). The esti­
mated cost of both physical and sexual violence in Australia exceeded
There is growing global recognition of the high cost of violence ten billion dollars in 2016 (KPMG, 2016). Therefore, it is important to
against women and children (WHO, 2017), including sexual violence understand the sequelae of SV for women to develop appropriate and
(SV), which broadly defines any sexual act perpetrated against a per­ effective support resources and intervention/treatment programs that
son’s will (Cox, 2015). SV subsumes a range of behaviors including child have both health-related and economic benefits.
sexual abuse, adult sexual assault and any type of sexual coercion where Child sexual abuse is a form of SV and the associated trauma of this
the perpetrators may be either known or unknown to the victim (AIHW, abuse with adverse psychiatric outcomes is well-established (see Hill­
2019). Globally one in three women are thought to have been affected berg et al., 2011 for a review). Numerous reviews, meta-analyses and
by physical or sexual violence within an intimate relationship (WHO, reviews of reviews confirm that for many survivors of child sexual abuse
2017) however estimates are likely conservative for reasons such as their likelihood of adult psychopathology is high (Maniglio, 2010, 2013;
women not perceiving these acts as violence because they have occurred Putnam, 2003). The longer term influence of adverse childhood expe­
within a relationship, or issues around disclosure such as feeling they riences (i.e., physical, sexual, psychological abuse and neglect) was
may not be believed or supported (Itzin et al., 2010). The occurrence of examined by Coe et al. (2021). These researchers found that mothers
SV in the community carries both a direct financial impact in the form of with experiences of intimate partner violence (IPV) endorsed greater
increased healthcare and social services to address this issue, as well as prenatal depressive symptom and this was moderated by the woman’s
indirect costs that come from the ongoing physical and psychological adverse childhood experiences. Prenatal depression can lead to less

* Corresponding author. School of Psychology, Building 1400, Charles Sturt University, Panorama Avenue, Bathurst, NSW, 2795, Australia.
E-mail address: rbrunton@csu.edu.au (R. Brunton).

https://doi.org/10.1016/j.socscimed.2021.114334
Received 15 December 2020; Received in revised form 27 July 2021; Accepted 19 August 2021
Available online 20 August 2021
0277-9536/© 2021 Elsevier Ltd. All rights reserved.
R. Brunton and R. Dryer Social Science & Medicine 292 (2022) 114334

sensitive parenting behavior and concerningly this is an early indicator importantly, the pattern of substance use is less considered, which is
of child maltreatment, suggesting child abuse may be a risk factor for important in assessing changes over time.
IPV and maltreatment of the children of mothers with experiences of Increased alcohol consumption can be a consequence of SV (see
abuse. SV experienced as an adult (either by known or unknown per­ Devries et al., 2014 for a review). Messman-Moore and Long (2003)
petrators) has also been associated with adverse outcomes. One propose that alcohol and drugs are used to reduce negative feelings,
Australian study of 230 women from a general medical practice (M age forget the abuse, or avoid abuse-specific memories and affective re­
= 51.1, SD = 16.4), found that women who experienced adult SV (e.g., sponses. Longitudinal studies have confirmed this cause-and-effect
unwanted sexual acts and coercion experienced in the last year) had relationship. A two-year study of 4009 women examined violent as­
higher levels of anxiety and depression than other women. This finding sault (physical and sexual) and alcohol and illicit drug abuse, and found
remained even after adjusting for age and the occurrence of child sexual that assault led to increased odds for alcohol abuse and drug use (Kil­
abuse (Tarzia et al., 2017). Similarly, Pico-Alfonso et al. (2006) exam­ patrick et al., 1997). Hedtke et al. (2008) examined substance use and
ined a community sample of 182 Spanish women and found that those interpersonal violence by interviewing 4008 women annually, over 3
who suffered intimate partner sexual abuse had higher depression scores years about their lifetime history of interpersonal violence (sexual and
and more suicide attempts than women with no history of this type of physical assault and witnessing violence), and mental health and sub­
abuse. stance use. For women who with a SV history, past year substance abuse
In contrast, literature reviews that have sought to synthesize research problems (alcohol and non-experimental drugs), were predictive of
findings for adult SV have resulted in inconsistent conclusions regarding substance abuse problems during the first year of the study (OR = 1.64)
the longer-term sequelae of SV. One early review by Steketee and Foa but not predictive of substance abuse in the third year of the study,
(1987), reported that while social and sexual functioning were signifi­ relative to no history of SV. While these findings suggest that the risk of
cantly impacted immediately after SV (i.e., victims of rape), most substance use problems may not remain stable over time, few studies
women return to pre-assault levels of anxiety and depression within 12 have directly examined whether the risk for substance use increases or
months. In contrast, a more recent systematic and meta-analytic review decreases over time.
conducted by Dworkin et al. (2017) found the effect of SV (search terms While the pattern of risk for substance use is unclear, what is well
included sexual assault, rape, sexual victimization) on psychopathology established is that the risk of revictimization of women with past sexual
to be relatively durable across time. Inconsistencies in these findings abuse experience is high. Up to 70% of child sexual abuse survivors are
may be related to the reviews containing both cross-sectional and lon­ sexually victimized in adulthood (see Classen et al., 2005 for a review;
gitudinal studies. While causality is implied based on the temporal order Elliott et al., 2004). Concerningly, the risk for revictimization is also
of the SV and these outcomes, longitudinal data are needed for this present for women who have no past child sexual abuse but who expe­
conclusion to be stated with certainty. rience sexual victimization in adulthood (Classen et al., 2005). Several
The findings from subsequent studies on this issue have done little to theories seek to explain this vulnerability (e.g., situational and person­
clarify the issue of stability and durability of psychopathology after SV. ality variables, see Breitenbecher (2001) for a review) with social
However, they do identify a number of factors that may contribute to the learning theories having good empirical support for explaining child
maintenance of distress and poorer psychosocial functioning. Shin et al. sexual abuse and later revictimization (Cochran et al., 2011). For sur­
(2020) examined PTSD symptomology of 94 female survivors of sexual vivors of child sexual abuse, learning theories propose that abuse
assault (M = 24.78, SD = 6.89) perpetrated by either a stranger or modelled in childhood may normalize the behavior leading to increased
partner. Women were assessed directly after the assault and one month acceptance of abuse in later adult relationships. Evidence around adult
later and based on their PTSD scores were classified as high distress or victimization and revictimization is less compelling likely due to the
low distress. While the highly distressed women showed a decrease in complexity of this issue. That is, the risk of revictimization may be
symptom severity, the low distress women did not show significant related to various factors such as when the assault occurred or rela­
changes in their symptoms. Negative thoughts and beliefs about the SV tionship to the perpetrator (see Classen et al., 2005 for a review).
were found to be strongly correlated with PTSD symptoms in both Notwithstanding, learning theories may also account for why some
groups which led these researchers to suggest that post-trauma cogni­ women do not perceive sexual violence within an intimate partner
tions may play a significant role in the maintenance of psychopathology. relationship as such, which further increases their risk of revictimization
By comparison, Chang et al. (2020) examined negative psychological (Cox, 2015).
outcomes in 151 female college students (M = 20.23, SD = 1.63). For However not all studies are unanimous in their conclusions
these women sexual assault victimization (i.e., verbal, unwanted regarding SV and deleterious outcomes. Bensley et al. (2003) surveyed
touching, attempted and completed penetration) was a stronger pre­ 3527 women (18–46 years) but did not find any relationship between
dictor of adverse outcomes (e.g., anxiety and alcohol use) whereas child sexual abuse and recent IPV. Others have similarly noted that some
negative life events (e.g., financial problems, death of a loved one), was adults sexually abused in childhood do not exhibit later psychopathol­
a stronger predictor of negative psychological outcomes such as lower ogy. In one study, children around 10 years of age were interviewed and
life satisfaction. Similarly, Jamison et al. (2021) found that more again as adults. Approximately 50% of children who had been previ­
frequent IPV experiences were associated with depressive symptomol­ ously abused in childhood reported no psychopathology. Moreover,
ogy in their cross sectional study of 112 women (M = 32.26, SD = 5.84). these survivors had significantly better psychosocial functioning than
In this sample, 60% of the women reported sexual IPV in the last 6 other adults who had not been abused and had no self-reported psy­
months. chiatric problems (Collishaw et al., 2007). The authors proposed that
Not surprisingly, survivors of SV have also been reported to engage some of these individuals may have a certain resilience that is linked to
in greater substance use/abuse (i.e., alcohol, cigarettes, and drugs) and quality relationships with parents, friends, and partners. In other words,
have a higher incidence of dependence or substance abuse disorders social support may mediate the relationship between sexual abuse and
than other individuals. Support for this relationship comes from studies psychological well-being.
that have focused on specific populations such as Indigenous peoples, This proposition is supported by studies such as the one conducted by
incarcerated women or certain socio-economic groups (e.g., Bohn, 2002 Littleton (2010) who noted that lower levels of social support predicted
examined lifetime and current sexual abuse; Libby, 2005 examined post-sexual-assault psychopathology such as PTSD and depressive
childhood sexual abuse). While a consensus exists that child sexual symptomology. Littleton argued that social support may play an
abuse (a form of SV) is a risk factor for substance abuse (see Halpern important role in post-assault recovery by contributing to preserving
et al., 2018; Simpson and Miller, 2002 for reviews), few studies have self-worth. Consistent with this finding, Brinker and Cheruvu (2017)
examined the longitudinal effects of any form of SV. Also, and using data from a large telephone survey conducted in 2010 across five

2
R. Brunton and R. Dryer Social Science & Medicine 292 (2022) 114334

US states (N = 12,487), found that both perceived emotional and social Table 2
support (support characterized by empathy and love) mediated the Sample demographics, N = 9145.
relationship between adverse childhood experiences and adult Ageb M = 30.61 (1.46), range = 26–35a
depression. c
Marital status
Never married 2174 (23.77%)
1.1. The current study Married/Defacto (opposite-sex) 4870 (53.25%)
Defacto (same sex) 1648 (18.02%)
Separated/divorced/widowed 390 (4.26%)
The current study addressed some of the gaps in the literature by Household income d
examining the relationship between SV and psychosocial wellbeing (i.e., $1 - $699 ($1-$36,999 pa) 662 (7.24%)
anxiety, depression, stress, life satisfaction, mental health, and revic­ $700 - $1499 pw ($37,000 - $77,999 pa) 2595 (28.38%)
$1500 or more pw ($78,000 or more pa) 3628 (39.67%)
timization), and behavioral variables (alcohol consumption, smoking,
Other 1604 (17.54%)
illicit drug use) in a longitudinal data set involving five date collection ARIAþ e
timepoints over a period of 12 years (i.e., 2006 to 2018). We examined Major cities 6150 (67.25%)
the following three hypotheses: 1) women with a history of SV were Inner regional 1687 (18.45%)
expected to have more adverse psychosocial outcomes than women with Outer regional 905 (9.90%)
Remote 170 (1.86%)
no history of SV, 2) women with a history of SV would have a higher Very remote 71 (0.78%)
alcohol consumption, consume more cigarettes, and take more illicit SEIFA f 1016.85 (86.46), range =
drugs than women with no SV history, and 3) a history of SV would 556.51–1251.87
predict adverse psychological and behavioral outcomes mediated by Education g
No formal qualifications 70 (0.77%)
perceived social support. Lastly, no specific hypotheses are offered for
School certificate (year 10 or equivalent) 581 (6.36%)
the level of distress or behavior over time as these will be exploratory. Higher school certificate (year 12 or 1329 (14.53%)
equivalent)
2. Methods Trade/apprenticeship (e.g., hairdresser, 248 (2.71%)
chef)
Certificate diploma (e.g., technician) 2139 (23.39%)
2.1. Participants University degree 3134 (34.27%)
Higher university degree (Masters PhD) 1590 (17.39%)
Data for this study are from the 1973–78 cohort from the Australian
Note. Unless otherwise noted, demographics are from survey 4. b = age range
Longitudinal Women’s Health Survey (ALWHS). The ALWHS initially differs to Table 1 due to missing data. Missing data: b = 21 (0.23%), c = 63
sampled women from the Medicare insurance database from urban, (0.69%), d = 656 (7.17%), e = 140 (1.53%) f = 291 (3.18%), g = 54 (0.59%).
rural and remote zones (see Brown et al. (1998) for recruitment, attri­ Household income is the gross income of families including partners or parents
tion and participant details). This study utilized surveys obtained at in a household; Other = those who did not wish to answer or had no income/
timepoints 4 (2006) to 8 (2018) as they used the same measures in each lives alone.
survey. Table 1 provides details the response rates for these surveys.
Subsequent analysis has confirmed that despite attrition the sample is
still broadly representative of the same demographic in the Australian Table 3
population for making inferences despite some over representation of Women who experienced sexual violence, surveys 4 to 8.
women with tertiary education (Brown et al., 1998; Powers and Loxton, Survey Last 12 months More than 12 months ago
2010). Yes No Yes No
Table 2 shows the sample demographics from survey 4. In this
4 (2006)a 73 (0.8) 9055 (99.0) 581(6.4) 8547 (93.5)
sample most women were partnered, employed, and had a good edu­ 5 (2009) b 71 (0.9) 7626 (93.0) 484 (5.9) 7213 (88.0)
cation. The Accessibility and Remoteness Index of Australia (ARIA+) 6 (2012) c 55 (0.7) 7698 (96.1) 559 (7.0) 7194 (89.8)
and the Socio-Economic Indexes for Areas (SEIFA) indicated that women 7 (2015) d 54 (0.8) 7008 (97.5) 471 (6.6) 6591 (91.7)
were mostly from major cities and on average had good economic 8 (2018) e 72 (1.0) 6910 (97.0) 551 (7.7) 6431 (92.1)

resources. Note. Missing a = 17, b = 120, c = 256, d = 124, e = 139. Percentages in


Statistics on sexual victimization appear in Table 3. In each year parentheses.
nearly 1% of the sample reported recent SV and around 6–8% reported
SV more than 12 months prior to the survey. These figures are below possible and further assesses the psychometrics of the scales within the
national estimates of 18% for women who experienced SV since the age study. For further details on the scales and treatment of missing data see
of 15 years (Cox, 2015). Our lower prevalence is likely due to our survey https://www.alswh.org.au/for-researchers/data. All measures in this
only considering occurrence whereas national estimates include occur­ study were obtained in Survey 4 to Survey 8.
rence, attempt, and threat.
2.2.1. Psychosocial variables
2.2. Materials Anxiety. Anxiety was assessed by the 9-item subscale from the
Goldberg Anxiety and Depression Inventory (GADS, Goldberg et al.,
The ALWHS uses scales with published psychometric data where 1988). Items such as “Have you been irritable,” are scored dichoto­
mously 1 (yes) and 0 (no). Scores range from 0 to 9 with higher scored
indicating greater anxiety.
Table 1
Response and attrition surveys 4 to 8. Depression. Depression was assessed by the 10-item Center for
Epidemiologic Studies Depression Scale–Shortened Version (CES-D,
Survey Year Age N Response rate
Andresen et al., 1994). Participants rate their feelings over the past week
4 2006 28–33 9145 64% from 0 (rarely or none of the time) to 3 (most or all of the time) to items such
5 2009 31–36 8199 58%
as “I felt fearful.” Scores range from 0 to 30 with higher scores indicating
6 2012 34–39 8009 56%
7 2015 37–42 7186 50% greater distress.
8 2018 40–45 7121 50% Stress. Stress was assessed by the 14-item Perceived Stress Scale
Stress (PSS, Bell and Lee, 2002) which covers domains such as health,
Note. Survey 1 (1996) N = 14,247.

3
R. Brunton and R. Dryer Social Science & Medicine 292 (2022) 114334

work/employment, money, and relationships. Psychometric analysis non-drinker, 2) rarely drinks, 3) low long-term risk, ≤ 2 drinks per day,
after survey 1 (1996) resulted in redundant items such as “relationships 4) risky, 3–4 drinks per day, 5), high risk, > 5 drinks per day.
with boyfriends” being removed. Other items were revised for inclu­ Smoking. Smoking status was determined by two questions related
siveness (i.e., ‘boyfriends’ changed to ‘friends’). For all surveys, re­ to frequency (e.g., “How often do you currently smoke cigarettes or any
sponses were rated as 0 (not applicable, not at all stressed) to 4 (extremely tobacco products”). Responses ranged from daily to not at all. Using the
stressed). Scores range from 0 to 56 with higher scores indicative of Australian Institute of Health and Welfare guidelines, five categories of
greater stress. smoking status were derived, 1) never smoker, 2) ex-smoker, 3) irregular
Mental Health Index. The SF-36 (Ware et al., 1993) is a 36-item smoker, 4) weekly smoker, 5) daily smoker.
survey measuring health-related quality-of-life. This study utilized the Drug use. A history of the pattern of non-medical drug usage was
mental health index which includes items such as “Have you been a very determined by questions related to ten commonly used drugs (i.e.,
nervous person?” rated from, 1 (all of the time) to 6 (none of the time). marijuana, amphetamines, LSD, natural hallucinogens, tranquilizers,
Scores are transformed into an index score ranging from 0 to 100 with cocaine, ecstasy, inhalants, heroin, or barbiturates). For this study they
higher scores representing greater mental health-related quality of life. were categorized as 1) Never used, 2) only ever used Marijuana but not in
Intimate Partner Violence. The 28-item revised Composite Abuse last 12mths, 3) used multiple/single drug other than Marijuana but not in the
Scale assessed IPV (Loxton et al., 2013). The scale has three factors: last 12mths, 4) only used Marijuana in the last 12mths, and 5) used mul­
emotional (13-items, e.g., told me I was stupid), physical (10-items, e.g., tiple/single drug other than Marijuana more than 12mths.
beat me up), harassment (4-items, e.g., harassed me at work) and a
single item that indicates SV (“forced me to take part in unwanted sexual 2.3. Statistical analyses
activity”) that was not used in this study. For survey 4 the three sub­
scales were scored on 6 response options, reflecting the frequency of the This repeated-measures study examined SV group status over five
experience (never to daily), higher scores indicate greater frequency. For time periods. Women were included in the SV group if they indicated
survey 8, responses were scored dichotomously as 1 = yes and 0 = no. they had suffered SV either in the last 12 months or more than 12
For both surveys an additional response of 8 indicated never had a months ago in any of the five surveys. Women were allocated to the no
partner/spouse and these respondents were excluded from any analyses SV group if they indicated they had never suffered SV in any survey. All
for this variable. The scale has good reliability and validity as reported in analyses were weighted by area to account for the deliberate over­
Loxton et al. (2013). sampling of rural and remote areas.
Sexual Violence. One question assessed a history of sexual violence Using data from Survey 4, Spearman’s Rho correlations were
and was asked under the Life Events section (i.e., being forced to take calculated between all psychosocial (i.e., anxiety, depression, stress,
part in any unwanted sexual activity). This question was scored mental-health index, IPV [emotional, physical and harassment], life
dichotomously (1 = yes, 2 = no) for two time periods (in the last 12 satisfaction, social support), behavioral (i.e., alcohol, smoking, drugs)
months and more than 12 months ago). and demographic (i.e., marital status, education, income) variables to
Life Satisfaction. This 8-item scale was custom developed for the identify covariates and assess multicollinearity between the dependent
ALWHS. A single question asks participants how satisfied they are in variables.
their achievements in certain aspects of their life (e.g. work, career, SV and no SV group differences for the psychosocial and behavioral
study). Responses are scored from 1 (very satisfied) to 4 (very dissatisfied). variables were determined using Repeat Measures Factorial Multivariate
Scores range from 8 to 24 with higher scores representing greater life Analysis of Variance (MANOVA). All assumptions pertaining to
satisfaction. normality, homogeneity of covariance, homoscedasticity of residuals,
Social Support. The abbreviated 6-item version of the Medical Out­ linearity, multivariate outliers and multicollinearity were assessed using
comes Study Social Support Index (MOS, Sherbourne and Stewart, 1991) multiple measures as prescribed by Field (2013).
was used. The MOS provides a measure of social support (i.e., the degree For the Repeat Measures Factorial MANOVA all individual scores for
to which interpersonal relationships serve functions) with the brief the relevant variables were transformed into Percent of Maximum
version comprised of 2 items from each MOS subscale (emotiona­ Possible (POMP) scores. POMP scores provide a standardized measure­
l/informational support, tangible support, affectionate support, positive ment of the different scales and enables more direct comparisons to be
social interaction). Responses are coded from 1 (none of the time) to 5 (all evident. Given that the dependent variables all use different scoring
of the time) to items following a stem question “How often is the conventions this transformation provided a meaningful way to
following support available to you if you need it?” Items are summed to communicate the magnitude of the observed relationships. Using POMP,
obtain a total score that can range from 6 to 30. Higher scores indicate individual scores are converted to a percentage of the maximum possible
greater social support. score of the scale. Given that this is a linear transformation of the scores
it does not change the output (i.e., test statistics). Cohen et al. (1999)
2.2.2. Behavioral variables provides a detailed explanation of POMP.
Demographics. Variables from survey 4 provided an overall picture Mediation analysis was conducted using Hayes (2013) PROCESS
of the sample including age, employment, marital status, rurality, in­ macro (v3.4). The sample size satisfied the required ratio of cases to
come, education, and country of birth. The ARIA+ variable provided an predictors. Eleven mediation models (see Fig. 1) examined SV history at
indication of location with categories of, major cities, inner regional, survey 4 as a predictor (X) of the eight psychosocial and three behavioral
outer regional, remote, and very remote. The SEIFA index of economic outcomes (Y) at survey 8. Social support (M) (average support across the
resources ranges from 245 to 1281 and is an index of financial aspects of five time periods) was included as a mediator in all models. All relevant
relative socio-economic advantage and disadvantage including high/ outcome variables from Survey 4 and any identified covariates (C) were
low income and variables correlating with high/low wealth. Areas with controlled for in all mediation models (e.g., when examining anxiety as
higher scores have relatively greater access to economic resources than an outcome from survey 8, anxiety from survey 4 was controlled). Given
areas with lower scores. the design of the models (examining SV from survey 4 and outcomes
Alcohol. Alcohol consumption was determined using three fre­ from survey 8), it was not possible to examine any revictimization that
quency questions (e.g., “How often do you usually drink alcohol?”) rated occurred in surveys 5–7.
from I never drink alcohol to every day, and one quantity question (“…
how many standard drinks do you usually have?”). Responses range from 1 3. Results
or 2 drinks per day to 9 or more. This study used the National Health and
Medical Research Council guidelines and assigned ordinal categories, 1) Spearman’s Rho correlations between the demographic (age

4
R. Brunton and R. Dryer Social Science & Medicine 292 (2022) 114334

Table 4
Psychosocial variables, between group differences.
Sexual violence history No sexual violence history

Measure Mean (SE) 95% CI Mean (SE) 95% CI

Anxiety 54.86 (.79) 53.31–56.41 43.52 (.57) 42.41–44.63


**** ****
Depression 26.52 (.47) 25.60–27.43 18.82 (.34) 18.16–19.48
**** ****
Stress 25.15 (.35) 24.46–25.84 19.35 (.25) 18.86–19.85
**** ****
Mental health 68.33 (.43) 67.47–69.17 74.01 (.31) 73.40–74.62
Index **** ****
Life satisfaction 72.88 (.42) 72.05–73.71 77.68 (.30) 77.08–78.28
**** ****
Fig. 1. Mediation model.
Note. POMP scores used. SE = standard error, CI = confidence intervals. Sig­
nificance denoted as ****p < .0001.
excluded given the restricted range) and the study variables (survey 4)
were all small with most between r = -0.15 and 0.15 and statistically
significant (likely due to the sample size). One exception was IPV and with a subsequent improvement.
marital status (rs 0.38-0.50, p < .001) but this was to be expected
therefore no covariates were included in the analyses. Multicollinearity 3.1.2. Behavioral variables
was not evident as no dependent variables had correlations > r = 0.80 Using the same equal groups, both between and within-subject ef­
(Field, 2013). fects were examined over the same five time periods for the categorical
outcomes (examined as ordinal scales) of alcohol consumption, smoking
and drug use. The overall model was significant for both between sub­
3.1. Repeat Measures Factorial MANOVA
ject, F(3, 2527) = 45.26, p < .0001, partial ή2 = 0.05 and within subject
F(12, 2518) = 31.08, p < .0001, partial ή2 = 0.13 differences indicating
3.1.1. Psychological variables
that both groups differed on these outcomes and also over time. Pair-
Both between and within-subject effects were examined by SV status
wise comparisons are reported in Table 5 for between groups. As
and time (2006–2018), for the continuous outcome measures of anxiety,
shown only alcohol consumption did not significantly differ between
depression, stress, mental-health index, and life satisfaction. A signifi­
groups (p = .21).
cant Box’s M test (p < .0001) indicated that the assumption of homo­
There was a significant within-subjects interaction observed between
geneity of covariance could not be assumed and this was further
time and SV status F(12, 2518) = 2.34, p = .006, ή2 = 0.01. The
confirmed by the observable variances in the covariance matrix
assumption of sphericity was again violated, and the Bonferroni
(Tabachnick and Fidell, 2001). Therefore, two equal SV/no SV groups
adjustment was applied and Greenhouse-Geisser statistic used to assess
were created. This was done by allocating all 1740 participants who had
the interactions. All three dependent variables showed a significant time
a history of SV in survey 4 into one group and randomly sampling 1740
by SV status interaction with alcohol consumption showing a cross over
participants from the 3853 women who indicated they had not suffered
interaction in the absence of a main effect (alcohol, F(3.70) 2.64, p =
any SV into the second group. As recommended by Field (2013),
.04, partial ή2 = 0.01; smoking F(3.16) 2.77, p = .04, partial ή2 = 0.01;
Hotelling’s Trace statistic was used given that it is robust to violations of
drugs F(3.49), 3.29, p = .01. partial ή2 = 0.01). Supplementary Table S2
this assumption when equal groups are used. This approach also satis­
contains the pairwise comparisons.
fied the assumptions of independence and random sampling.
Fig. 3 depicts the three behavioral variables across the five time
The overall model was significant for both between-subject, F(5,
periods. Of note is the increase in alcohol consumption for the women in
2418) = 45.61, p < .0001, partial ή2 = 0.09 and within-subject F(20,
the no-SV group from 2006 (Time 3) onwards with a cross over inter­
2403) = 35.78, p < .0001, partial ή2 = 0.23 differences indicating that
action between time 4 and 5. For both groups their consumption of
women with an SV history had greater anxiety, depression, stress, and
cigarettes decreased, and illicit drug use followed a similar downward
lower mental health index and life satisfaction. Within group difference
trend over time.
indicated that the levels of these outcomes measures differed over time
but there was no significant time by SV status interaction, F(20, 2403) =
0.95, p = .52 (note that the differences in group sizes are due to missing 3.2. Mediation models
data on some variables with pair wise deletion used). A significant
Mauchly’s Test for all dependent variables indicated that the assumption Table 6 presents the results from the eleven mediation analyses.
of sphericity was not met therefore a Bonferroni adjusted significance Apart from life satisfaction, a history of SV significantly predicted all
level of p = 01 was used for all group comparisons (Field, 2013). anxiety, depression, stress, mental health and revictimization
Comparisons are presented in Table 4. (emotional, physical and harassment) with social support mediating
There was also significant variability across time. Supplementary these relationships (consistent with Hayes (2013) we make no distinc­
Table S1 presents the within groups differences between time 1 (2006) tion between full and partial mediation). For the behavioral variables, a
and each other time point (2009–2018) and between group differences history of SV significantly predicted drug use, but social support did not
for each time point and variable. Despite no significant interaction, the mediate this relationship. For alcohol and smoking the results were not
means for each group across time are reported to show trends. statistically significant.
Fig. 2 graphically depicts the SV groups and psychosocial variables
across the five time periods. For all variables, both groups had a marked 4. Discussion
decrease in depression, anxiety, and stress from 2012 (time 3) and an
upward trend thereafter. Similarly, the mental health index was rela­ This study examined the psychosocial and behavioral correlates of
tively stable until 2012 (time 3) and then showed a downward trajectory SV and their sequelae over time. As expected, women who had experi­
indicating that for both groups their mental health-related quality of life, enced SV had greater anxiety, depression, stress, and a lower mental
while stable from 2006 to 2012 then decreased. There was also a sig­ health index and less life satisfaction. As expected, women with a history
nificant decrease in life satisfaction for both groups from 2006 to 2009 of SV had a higher consumption of alcohol, consumed more cigarettes,

5
R. Brunton and R. Dryer Social Science & Medicine 292 (2022) 114334

Fig. 2. Psychosocial variables.


Note. For all figures the dashed line indicates the SV ‘yes’ group and the solid line indicates to SV ‘no’ group. All scores are estimated marginal means.

and used more illicit drugs than other women. We also determined that the causality of SV and these outcomes.
for women who had experienced SV their level of distress remained Our findings that women with a SV history had greater distress than
higher over time than other women however, this distress followed the other women is consistent with studies into the longer term effect of
same pattern as women with no SV history. Our expectation that SV child sexual abuse and sexual violence in adulthood, such as Pico-Al­
would predict adverse psychosocial and behavioral outcomes was sup­ fonso et al. (2006) who examined lifetime sexual IPV (child sexual abuse
ported except for life satisfaction, alcohol consumption and smoking. and adulthood victimization) and Tarzia et al. (2017) study of any type
Further, for anxiety, depression, stress, and the mental health index, of SV in the past 12 months. A significant finding was that a woman’s
social support mediated these relationships. The findings of this study level of distress (i.e., anxiety, depression, and stress) was not influenced
contribute, the literature by firmer conclusions able, be drawn regarding by time. That is, for women who had experienced SV their post-assault

6
R. Brunton and R. Dryer Social Science & Medicine 292 (2022) 114334

Table 5 Of interest was the notable decrease, for both groups, in life satis­
Behavioral variables, between group differences. faction in 2009 (time 2) and the distress measures and mental health
Sexual violence history No sexual violence history index in 2012 (time 3). These ‘dips’ in the trend are consistent with the
cohort’s developmental life stage which would include key milestones
Measure Mean (SE) 95% CI Mean (SE) 95% CI
such as childbirth, and the transition from young, older adulthood. Both
Alcohol 41.04 (.54) 39.98–42.11 40.20 (.39) 39.43–40.97 milestones are linked, mood disorders and lower life satisfaction (Lee,
Smoking 23.11 (.88) **** 21.39–24.83 14.04 (.63) **** 12.80–15.29
Drug use 35.61 (.89) **** 33.85–37.36 23.82 (.65) **** 22.56–25.08
2020; Qu and de Vaus, 2015). While these life stages offer a possible
explanation of these decreases, what is important, note is that while
Note. POMP scores used. SE = standard error, CI = confidence intervals. Sig­ women with a history of SV retain their post-assault levels of distress, it
nificance denoted as, *p < .05, ****p < .0001.
can be influenced by individual variables suggesting that positive in­
terventions may reduce these adverse effects.
distress remained consistently higher than other women. The longitu­ One such intervention is social support. Mediation analyses
dinal data analyzed for this study did not include pre-assault measures of confirmed that SV predicted anxiety, depression, stress, and mental
distress (and therefore it was not possible, directly examine whether health-related quality of life, and that this relationship was mediated by
women with SV experience returned, pre-assault levels of psychosocial social support. Of note was the stronger effects obtained for depression
functioning), however, our findings suggest that these women did not and the mental health index, indicating that for many women who have
return, pre-assault levels within 12 months as has previously been sug­ been victims of SV they will likely have greater psychopathology if they
gested in the literature (Steketee and Foa, 1987). Moreover, despite have lower levels of perceived social support. This is consistent with the
significant differences between the groups in levels of distress, both stress buffering hypothesis, which suggests that for these women,
groups showed a similar pattern of change over time akin, trends in the ensuring they have the necessary support either formally (agencies) or
general population (from 2001, 2018 mood disorders/issues increased informally (friends/family) is integral, their well-being. The non-
by 1%, ABS, 2018). It should be noted that given the issues around significant findings for life satisfaction were unexpected and poten­
self-report (e.g., accuracy of responding Gravetter and Forzano, 2016) tially reflects the diversity of areas which were included in the measure
and that these outcomes were not diagnosed clinically, we make no used, assess their judgement of quality of life. It is possible that for these
distinction about clinical levels of anxiety or depression but rather women while their wellbeing may have been lower in some areas (i.e.,
examine these outcomes as they pertain, significant between group career/study), their global assessment showed more life satisfaction
differences. (Diener et al., 1985).

Fig. 3. Trends in alcohol, smoking and drug status across time.


Note. For all figures the dashed line indicates the SV ‘yes’ group and the solid line indicates to SV ‘no’ group. All scores are estimated marginal means.

7
R. Brunton and R. Dryer Social Science & Medicine 292 (2022) 114334

Table 6 knowledge of health relationships were exploited by perpetrators and


Total and indirect effect of sexual violence on psychosocial and behavioral led, later victimization (Tarzia, 2020). This suggests that perpetrator
variables. variables play a role in adult victimization and revictimization.
Total effect (path c) Indirect effect (path ab) Important, breaking this cycle of violence is perceived social support.
IV →DV R2 IV →DV 95% CI
Our findings confirm that greater social support reduces the risk of
revictimization consistent with Dias et al. (2019). Therefore, the
Psychosocial variables
importance of social support for women who have experienced SV as an
Anxiety 0.50 (.15) *** .23 0.15 (.03) 0.09, 0.21
Depression 1.54 (.30) **** .23 0.43 (.09) 0.26, 0.63 intervention against further revictimization cannot be underestimated.
Stress 0.10 (.03) *** .25 0.03 (.01) 0.02, 0.05 All types of social support (i.e., structural, functional or satisfaction) are
Life satisfaction − 0.05 (.03) .18 − 0.04 (.01) − 0.06, − 0.02 associated with a positive change in IPV victims (Zapor et al., 2018)
Mental health − 4.23 (.92) **** .21 − 1.30 (.24) − 1.75, − 0.83 including improved self-efficacy in help-seeking, leaving the abusive
IPV Emotional 3.55 (.33) **** .16 0.54 (.11) 0.35, 0.76
IPV Physical 2.43 (.22) **** .15 0.25 (.06) 0.15, 0.38
relationship and greater disclosure.
IPV Harassment 1.35 (.13) **** .10 0.16 (.03) 0.10, 0.23 Our findings in relation, the alcohol consumption showed that
Behavioral variables despite no significant differences in consumption between groups, over
Alcohol risk of harm − 0.04 (.04) .29 − 0.03 (.01) − 0.04, − 0.01 time SV status contributes, changes in these behaviors. That is, while
Smoking status 0.05 (.04) .57 0.01 (.01) 0.01, 0.02
both groups of women did not show a significant increase in consump­
Drug status 0.11 (.05) ** .53 0.01 (.01) − 0.01, 0.02
tion over the time periods examined, women with no SV history had a
Note. Predictor (IV) = sexual violence, outcome (DVs) = psychosocial and sharper increase in consumption over time. Notwithstanding this, these
behavioral variables as shown. IPV = intimate partner violence. Analysis based findings should be cautiously interpreted given that both groups of
on 1000 bootstrap resamples. The sample size for each model varied due,
women were in the ‘rarely drinks’, ‘low long-term risk’ categories. What
missing data but ranged from 3771 to 4631. Statistical significance donated as
is important, note is that for these women their level of consumption was
****p < .0001, ***p < .001, **p < .01. Standard error reported in parentheses.
low and for most, this may indicate a level of consumption consistent
with self-medicating.
Social support as an intervention for IPV has good empirical support.
In comparison, for smoking and drug use, both groups decreased
A recent scoping review that examined IPV interventions focused on
their usage over time with the women who had experienced SV showing
improving social support found that these were linked, improved mental
a greater decrease across time. Again, the same caution is warranted in
health outcomes. Social support, in this review, was classified as advo­
interpretation. That is, for smoking, both groups were in the low range
cacy (trained professionals providing supportive relationships and re­
(e.g., non-smoker, ex-smoker) consistent with population trends, ward a
sources), advocacy with psychotherapy, and community-based social
decline in smoking. However, our results for smoking, that show higher
support interventions. These community-based interventions aimed,
consumption, are consistent with extant research which posits that
change community attitudes or involved group therapy. All in­
smoking is a stress reliever for women with a history of SV (Jun et al.,
terventions provided support and those with strong community links
2008). While smoking initiation often occurs in adolescence and these
contributed, improved social support for survivors of violence as well as
women’s pre-abuse smoking was not included here, we cannot say with
more positive mental health outcomes (Ogbe et al., 2020).
certainty that SV was a cause of smoking. Regardless smoking is a public
A significant finding of this study was that women with a history of
health issue and recent campaigns aimed, reduce consumptions likely
SV have a greater risk of being revictimized or continually victimized
explain the decrease.
physically, sexually, or psychologically by their intimate partner. While
By comparison, the drug use of both groups were in the lower ranges
this is consistent with extant research (see Breitenbecher, 2001 for a
(never used - not used in the last 12 months) which is inconsistent with
review), examination of specific IPV subtypes (i.e., emotional, physical
population trends of increased drug use (from 2006, 2019 there was a
and harassment) allowed for greater insights about the long-term IPV-­
2.5% increase, AIHW, 2020). The decreased usage however was
related consequences of SV, particularly in terms of emotional abuse. SV
consistent for the age group of these women; this pattern has trended
significantly predicted all three IPV sub-types with emotional abuse the
downward since 2016 (age 35–44, 2.3% decrease for any illicit drugs).
strongest predictor. This indicates that the negative effect of previous SV
The results for substance use were further elucidated in the media­
may create or increase a woman’s vulnerability for later victimization
tion analyses with SV not significantly predicting alcohol consumption
(Messman-Moore and Long, 2003). The finding that this later victimi­
or smoking. SV history was however a significant predictor of drug use,
zation is more likely in the form of emotional abuse is concerning given
but this relationship was not mediated by social support. For our find­
that this form of abuse is a known precursor for later instances of all
ings relating, smoking and drug use, while SV women had significantly
forms of abuse (Karakurt and Silver, 2013).
higher usage the decrease over time was greater compared, women with
Emotional abuse is considered distinct for other forms of IPV (Jun
no SV history. Further, alcohol consumption did not increase signifi­
et al., 2008). Several theoretical perspectives offer explanations for
cantly over time for women with a history of SV. Therefore, despite their
previously victimized women being more vulnerable, emotional abuse.
distress remaining higher than other women they were not using these
These include cognitive attribution theories (i.e., learned helplessness)
substances at a level, or increasing their level of usage, that would be
or traumatic bonding which proposes that women may form intense
considered harmful, them.
emotional bonds with their abusers motivating them, stay in abusive
The modest effect sizes in the mediation models should not detract
relationships (Bandura, 1986; Dutton, 1988). However as yet no one
from the importance of these findings. A key consideration is that these
theory offers an adequate explanation (Messman-Moore and Long,
effect sizes may be indicative of confounding variables such as women
2003) and we propose that this is due, most theories seeking, explain
exposed, trauma having difficulty noticing, reporting and understanding
revictimization by focusing on victim-related variables only (e.g., low
their emotional states which may be related, limited awareness or an
self-esteem, inability, perceive threats). By focusing on victim only
implicit fear of emotions themselves (Goldsmith et al., 2012). The low
variables, one could argue that responsibility for this revictimization is
effect sizes may also indicate that for some women SV is not perceived as
attributed, them while the perpetrator’s motivations/behavior are
such in intimate relationships, and this is a potential issue in accurate
excluded. Perpetrator characteristics such as seeking women they can
measurement (Loxton, 2013).
control should be considered alongside victim vulnerabilities (see
Messman-Moore and Long, 2003 for a helpful framework). Indeed pre­
4.1. Strengths and limitations
viously abused women (physical, sexual or emotional) felt that conse­
quences of the previous abuse such as low self-esteem or poor
This study had several strengths and some limitations. The use of

8
R. Brunton and R. Dryer Social Science & Medicine 292 (2022) 114334

longitudinal data over a period of 12 years for a large representative Andresen, E.M., Malmgren, J.A., Carter, W.B., Patrick, D.L., 1994. Screening for
depression in well older adults: evaluation of a short form of the CES-D. Am. J. Prev.
sample enabled causal inferences, be drawn concerning SV and is a key
Med. 10 (2), 77–84.
strength of this study. The findings however were limited by self-report Bandura, A., 1986. Social Foundations of Thought and Action. Englewood Cliffs, NJ,
with respondents known, under-report their exposure, IPV (Chan, 2011) p. 1986.
thus our estimates are likely conservative. As with all self-report data Bell, S., Lee, C., 2002. Development of the perceived stress questionnaire for young
women. Psychol. Health Med. 7 (2), 189–201. https://doi.org/10.1080/
caution is needed in interpreting the results. Future studies that assess 13548500120116085.
SV using interview data would be beneficial, further elucidating these Bensley, L., Van Eenwyk, J., Simmons, K.W., 2003. Childhood family violence history
results. The measure used, assess SV was also limited as it did not and women’s risk for intimate partner violence and poor health. Am. J. Prev. Med.
25 (1), 38–44. https://doi.org/10.1016/S0749-3797(03)00094-1.
distinguish between child sexual abuse and adult SV. We also did not Bohn, D.K., 2002. Lifetime and current abuse, pregnancy risks, and outcomes among
control for revictimization of women between the initial survey (4) and native american women. J. Health Care Poor Underserved 13 (2), 184–198. https://
outcome survey (8) nor obtain perpetrator details of the SV. This may doi.org/10.1353/hpu.2010.0624.
Breitenbecher, K.H., 2001. Sexual revictimization among women: a review of the
have limited our findings as repeated victimization and perpetrator literature focusing on empirical investigations. Aggress. Violent Behav. 6 (4),
characteristics (e.g., known/unknown) may influence mental health 415–432. https://doi.org/10.1016/S1359-1789(00)00014-8.
outcomes (see Shin et al., 2020 for example). Also, this study did not Brinker, J., Cheruvu, V.K., 2017. Social and emotional support as a protective factor
against current depression among individuals with adverse childhood experiences.
make a distinction between SV suffered in heterosexual relationships or Preventive medicine reports 5, 127–133. https://doi.org/10.1016/j.
same-sex relationships and further research that delineates these rela­ pmedr.2016.11.018.
tionship types are needed. Brown, W.J., Bryson, L., Byles, J., Dobson, A.J., Lee, C., Mishra, G., Schofield, M., 1998.
Women’s Health Australia: recruitment for a national longitudinal cohort study.
Womens Health 28 (1), 23–40. https://doi.org/10.1300/J013v28n01_03.
5. Conclusion Chan, K.L., 2011. Gender differences in self-reports of intimate partner violence: a
review. Aggress. Violent Behav. 16 (2), 167–175. https://doi.org/10.1016/j.
avb.2011.02.008.
The findings from this study provide additional insights into the
Chang, E.C., Lee, J., Morris, L.E., Lucas, A.G., Chang, O.D., Hirsch, J.K., 2020.
sequelae of SV in Australian women. The greater psychological distress A preliminary examination of negative life events and sexual assault victimization as
women with a history of SV suffer, lower mental-health related quality predictors of psychological functioning in female college students: does one matter
more than the other? J. Interpers Violence 35 (21–22), 5085–5106. https://doi.org/
of life and less life satisfaction, it would seem, remains constant over
10.1177/0886260517719901.
time. Moreover, the experience of SV also places these women at greater Classen, C.C., Palesh, O.G., Aggarwal, R., 2005. Sexual revictimization: a review of the
risk of revictimization but the mechanisms that this occurs through are empirical literature. Trauma Violence Abuse 6 (2), 103–129. https://doi.org/
unclear. Finally, while women with a history of SV partake in more 10.1177/1524838005275087.
Cochran, J.K., Sellers, C.S., Wiesbrock, V., Palacios, W.R., 2011. Repetitive intimate
substance use than other women, their usage is not considered at levels partner victimization: an exploratory application of social learning theory. Deviant
that are categorized as harmful and does not increase over time. For Behav. 32 (9), 790–817.
these women, the role of social support is critical, their wellbeing. Social Coe, J.L., Huffhines, L., Gonzalez, D., Seifer, R., Parade, S.H., 2021. Cascades of risk
linking intimate partner violence and adverse childhood experiences, less sensitive
support provides a positive intervention and for some women may break caregiving during infancy, 0(0), 10775595211000431 Child. Maltreat.. https://doi.
the cycle of intimate partner violence and/or reduce their level of org/10.1177/10775595211000431.
distress. Cohen, P., Cohen, J., Aiken, L.S., West, S.G., 1999. The problem of units and the
circumstance for POMP. Multivariate Behav. Res. 34 (3), 315–346. https://doi.org/
10.1207/S15327906MBR3403_2.
Declaration of competing interest Collishaw, S., Pickles, A., Messer, J., Rutter, M., Shearer, C., Maughan, B., 2007.
Resilience, adult psychopathology following childhood maltreatment: evidence from
a community sample. Child Abuse Neglect 31 (3), 211–229. https://doi.org/
None declared. 10.1016/j.chiabu.2007.02.004.
Cox, P., 2015. Violence against Women in Australia: Additional Analysis of the
Australian Bureau of Statistics. Personal Safety Survey, 2012: ANROWS.
Acknowledgments Devries, K.M., Child, J.C., Bacchus, L.J., Mak, J., Falder, G., Graham, K., Heise, L., 2014.
Intimate partner violence victimization and alcohol consumption in women: a
The research on which this paper is based was conducted as part of systematic review and meta-analysis. Addiction 109 (3), 379–391. https://doi.org/
10.1111/add.12393.
the Australian Longitudinal Study on Women’s Health by the University Dias, N.G., Costa, D., Soares, J., Hatzidimitriadou, E., Ioannidi-Kapolou, E., Lindert, J.,
of Queensland and the University of Newcastle. We are grateful to the Fraga, S., 2019. Social support and the intimate partner violence victimization
Australian Government Department of Health for funding and to the among adults from six European countries. Fam. Pract. 36 (2), 117–124. https://doi.
org/10.1093/fampra/cmy042.
women who provided the survey data. Diener, E., Emmons, R.A., Larsen, R.J., Griffin, S., 1985. The satisfaction with life scale.
J. Pers. Assess. 49 (1), 71–75. https://doi.org/10.1207/s15327752jpa4901_13.
Appendix A. Supplementary data Dutton, D.G., 1988. The Domestic Assault of Women : Psychological and Criminal Justice
Perspectives. Allyn and Bacon.
Dworkin, E.R., Menon, S.V., Bystrynski, J., Allen, N.E., 2017. Sexual assault victimization
Supplementary data to this article can be found online at https://doi. and psychopathology: a review and meta-analysis. Clin. Psychol. Rev. 56, 65–81.
org/10.1016/j.socscimed.2021.114334. https://doi.org/10.1016/j.cpr.2017.06.002.
Elliott, D.M., Mok, D.S., Briere, J., 2004. Adult sexual assault: prevalence,
symptomatology, and sex differences in the general population. J. Trauma Stress 17
Credit roles (3), 203–211. https://doi.org/10.1023/b:Jots.0000029263.11104.23.
Field, A., 2013. Discovering Statistics Using IBM SPSS Statistics, fourth ed. Sage.
Goldberg, D., Bridges, K., Duncan-Jones, P., Grayson, D., 1988. Detecting anxiety and
Robyn Brunton: Conceptualisation, Methodology, data Formal depression in general medical settings. Br. Med. J. 297, 897–899. https://doi.org/
analysis and interpretation and writing reviewing, and editing the 10.1136/bmj.297.6653.897.
manuscript. Rachel Dryer: Conceptualisation, Methodology, data inter­ Goldsmith, R.E., Freyd, J.J., DePrince, A.P., 2012. Betrayal trauma: associations with
psychological and physical symptoms in young adults. J. Interpers Violence 27 (3),
pretation, reviewing and editing the manuscript.
547–567. https://doi.org/10.1177/0886260511421672.
Gravetter, F.J., Forzano, L.B., 2016. Research Methods for the Behavioral Sciences.
References Cengage Learning.
Halpern, S.C., Schuch, F.B., Scherer, J.N., Sordi, A.O., Pachado, M., Dalbosco, C., Von
Diemen, L., 2018. Child maltreatment and illicit substance abuse: a systematic
ABS, 2018. National Health Survey. Retrieved from Canberra. https://www.abs.gov.au/s
review and meta-analysis of longitudinal studies. Child Abuse Rev. 27 (5), 344–360.
tatistics/health/health-conditions-and-risks/national-health-survey-first-results/la
https://doi.org/10.1002/car.2534.
test-release.
Hayes, A.F., 2013. Introduction, Mediation, Moderation and Conditional Process
AIHW, 2019. Family, Domestic and Sexual Violence in Australia: Continuing the National
Analysis: A Regression-Based Approach. Guildford Publications.
Story 2019. Retrieved from Canberra. https://www.aihw.gov.au.
Hedtke, K.A., Ruggiero, K.J., Fitzgerald, M.M., Zinzow, H.M., Saunders, B.E., Resnick, H.
AIHW, 2020. Alcohol,bacco & Other Drugs in Australia. Retrieved from Canberra. htt
S., Kilpatrick, D.G., 2008. A longitudinal investigation of interpersonal violence in
ps://www.aihw.gov.au/reports/alcohol/alcohol-tobacco-other-drugs-australia.

9
R. Brunton and R. Dryer Social Science & Medicine 292 (2022) 114334

relation, mental health and substance use. J. Consult. Clin. Psychol. 76 (4), 633. reformulation. Clin. Psychol. Rev. 23 (4), 537–571. https://doi.org/10.1016/S0272-
https://doi.org/10.1037/0022-006X.76.4.633. 7358(02)00203-9.
Hillberg, T., Hamilton-Giachritsis, C., Dixon, L., 2011. Review of meta-analyses on the Ogbe, E., Harmon, S., Van den Bergh, R., Degomme, O., 2020. A systematic review of
association between child sexual abuse and adult mental health difficulties: a intimate partner violence interventions focused on improving social support and/
systematic approach. Trauma Violence Abuse 12 (1), 38–49. https://doi.org/ mental health outcomes of survivors. PloS One 156, e0235177. https://doi.org/
10.1177/1524838010386812. 10.1371/journal.pone.0235177.
Itzin, C., Taket, A., Barter-Godfrey, S., 2010. Domestic and Sexual Violence and Abuse. Pico-Alfonso, M.A., Garcia-Linares, M.I., Celda-Navarro, N., Blasco-Ros, C.,
Tackling the Health and Mental Health Effects. Routledge. Echeburúa, E., Martinez, M., 2006. The impact of physical, psychological, and sexual
Jamison, L.E., Howell, K.H., Decker, K.M., Schwartz, L.E., Thurston, I.B., 2021. intimate male partner violence on women’s mental health: depressive symptoms,
Associations between substance use and depressive symptoms among women posttraumatic stress disorder, state anxiety, and suicide. J. Wom. Health 15 (5),
experiencing intimate partner violence. J. Trauma & Dissociation 1–15. https://doi. 599–611. https://doi.org/10.1089/jwh.2006.15.599.
org/10.1080/15299732.2020.1869646. Powers, J., Loxton, D., 2010. The impact of attrition in an 11-year prospective
Jun, H.-J., Rich-Edwards, J.W., Boynton-Jarrett, R., Wright, R.J., 2008. Intimate partner longitudinal study of younger women. Ann. Epidemiol. 20 (4), 318–321. https://doi.
violence and cigarette smoking: association between smoking risk and psychological org/10.1016/j.annepidem.2010.01.002.
abuse with and without co-occurrence of physical and sexual abuse. Am. J. Publ. Putnam, F.W., 2003. Ten-year research update review: child sexual abuse. J. Am. Acad.
Health 98 (3), 527–535. https://doi.org/10.2105/AJPH.2003.037663. Child Adolesc. Psychiatry 42 (3), 269–278. https://doi.org/10.1097/00004583-
Karakurt, G., Silver, K.E., 2013. Emotional abuse in intimate relationships: the role of 200303000-00006.
gender and age. Violence Vict. 28 (5), 804–821. https://doi.org/10.1891/0886- Qu, L., de Vaus, D., 2015. Life satisfaction across life course transitions. Journal of the
6708.vv-d-12-00041. Home Economics Institute of Australia 22 (2), 15. https://search.informit.com.au
Kilpatrick, D., Acierno, R., Resnick, H., Saunders, B., Best, C., 1997. A 2-year longitudinal /documentSummary;dn=596482806333229.
analysis of the relationships between violent assault and substance use in women. Sherbourne, C.D., Stewart, A.L., 1991. The MOS social support survey. Soc. Sci. Med. 32
J. Consult. Clin. Psychol. 65 (5), 834–847. https://doi/10.1037/0022-006X (6), 705–714. https://doi.org/10.1016/0277-9536(91)90150-B.
.65.5.834. Shin, K.M., Kim, Y., Chung, Y.K., Chang, H.Y., 2020. Assault-related factors and trauma-
KPMG, 2016. The Cost of Violence against Women and Their Children in Australia. related cognitions associated with post-traumatic stress symptoms in high-distress
Retrieved from. https://www.dss.gov.au/sites/default/files/documents/08_2016. and low-distress Korean female victims of sexual assault. J. Kor. Med. Sci. 35 (19),
Lee, J., 2020. Trajectories of depression between 30s and 50s: latent growth modeling. e144. https://doi.org/10.3346/jkms.2020.35.e144.
Issues Ment. Health Nurs. 41 (7), 624–636. https://doi.org/10.1080/ Simpson, T.L., Miller, W.R., 2002. Concomitance between childhood sexual and physical
01612840.2019.1688438. abuse and substance use problems: a review. Clin. Psychol. Rev. 22 (1), 27–77.
Libby, A.M., 2005. Childhood physical and sexual abuse and subsequent alcohol and https://doi.org/10.1016/S0272-7358(00)00088-X.
drug use disorders in two American-Indian tribes. J. Stud. Alcohol 65 (1), 74–83 Tabachnick, B.G., Fidell, L.S., 2001. Using Multivariate Statistics, fourth ed. Allyn &
(info:doi/). Bacon.
Littleton, H.L., 2010. The impact of social support and negative disclosure reactions on Tarzia, L., 2020. Ward an ecological understanding of intimate partner sexual violence.
sexual assault victims: a cross-sectional and longitudinal investigation. J. Trauma & J. Interpers Violence, 0886260519900298.
Dissociation 11 (2), 210–227. https://doi.org/10.1080/15299730903502946. Tarzia, L., Maxwell, S., Valpied, J., Novy, K., Quake, R., Hegarty, K., 2017. Sexual
Loxton, D., Powers, J., Fitzgerald, D., Forder, P., Anderson, A., Taft, A., Hegarty, K., violence associated with poor mental health in women attending Australian general
2013. The Community Composite Abuse Scale: reliability and validity of a measure practices. Aust. N. Z. J. Publ. Health 41 (5), 518–523. https://doi.org/10.1111/
of intimate partner violence in a community survey from the ALSWH. J. Wom. 1753-6405.12685.
Health 2 (10.4172), 2325–9795. https://doi.org/10.4172/2325-9795.1000115. Ware, J.E., Snow, K.K., Kosinski, M., Gandek, B., 1993. SF-36 Health Survey. Manual and
Maniglio, R., 2010. Child sexual abuse in the etiology of depression: a systematic review Interpretation Guide. The Health Institute.
of reviews. Depress. Anxiety 27 (7), 631–642. https://doi.org/10.1002/da.20687. Who, 2017. Violence against Women. Retrieved from Geneva Switzerland. https://www.
Maniglio, R., 2013. Child sexual abuse in the etiology of anxiety disorders: a systematic who.int/news-room/fact-sheets/detail/violence-against-women.
review of reviews. Trauma Violence Abuse 14 (2), 96–112. https://doi.org/10.1177/ Zapor, H., Wolford-Clevenger, C., Johnson, D.M., 2018. The association between social
1524838012470032. support and stages of change in survivors of intimate partner violence. J. Interpers
Messman-Moore, T.L., Long, P.J., 2003. The role of childhood sexual abuse sequelae in Violence 33 (7), 1051–1070. https://doi.org/10.1177/0886260515614282.
the sexual revictimization of women: an empirical review and theoretical

10

You might also like