University of Southern California Dissertations and Theses

Mental and physical health consequences of intimate partner violence in a multi-ethnic sample of women

(USC Thesis Other)

Mental and physical health consequences of intimate partner violence in a multi-ethnic sample of women

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content MENTAL AND PHYSICAL HEALTH CONSEQUENCES OF INTIMATE
PARTNER VIOLENCE IN A MULTI-ETHNIC SAMPLE OF WOMEN
Copyright 2004
by
Kristen Nicole Presley Green
Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
EDUCATION (COUNSELING PSYCHOLOGY)
December 2004
Kristen Nicole Presley Green
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3145206
Copyright 2004 by
Green, Kristen Nicole Presley
All rights reserved.
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
®
UMI
UMI Microform 3145206
Copyright 2004 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Dedication
This dissertation is dedicated in loving memory to my father, Mr. Strickland
Cornelius Presley. His love and support made this document possible.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgements
I would like to thank my husband, mother, grandmother, sister, brother,
godmother, and my family and friends for their constant love, patience and
encouragement throughout this process. I would also like to acknowledge and thank
my advisor Dr. Michael Newcomb for helping me through this process; Dr. Melora
Sundt for her unwavering support and assistance, and Dr. Elaine Bell Kaplan for her
patience and wisdom on this project. Finally, I would like to thank Dr. Gail Wyatt
for her mentorship, guidance, boundless generosity, and unquestioning belief in me.
Without all o f you, this dissertation would not have been possible.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
Dedication..............................................................................................................................ii
Acknowledgements....................................... iii
List o f tables........................................................................................................... v
List o f figures........................................................................................................... vii
Abbreviations..................................................................................................................... viii
A bstract................................................................................................................................. ix
Chapter 1. Introduction.........................................................................................................1
Chapter 2. Literature review ................................................................................................ 8
Chapter 3. Methodology............................... 47
Chapter 4. R esults............................................................................................................... 59
Chapter 5. D iscussion.......................................................................................... 94
Chapter 5. Discussion.........................................................................................................94
References................. 117
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Tables
Table 1 Demographic Description o f the Sample....................... 49
Table 2 Scale Alphas for IPV M easures.......................................................................... 53
Table 3 Scale Alphas for Outcome M easures................................................................ 56
Table 4 Correlations among all Mental and Physical Health Outcome Measures 56
Table 5 Ethnic Differences in Correlation Coefficients for Demographic Variables 61
Table 6 Rates of Sexual Trauma in the Sample.............................................................. 64
Table 7 Relationship between CSA and Demographics in the Sample.......................65
Table 8 Relationship between ASA and Demographics in the Sample.......................65
Table 9 Mean Ethnic Differences in IPV ........................................................................ 66
Table 10 Relationship between IPV and Demographic Variables................................68
Table 11 Correlation Coefficients b/w IPV & Chronic Burden Scale Item s.............. 69
Table 12 Correlation Coefficients between IPV and Health O utcom es......................72
Table 13 Ethnic Differences in Mental and Physical Health Outcomes......................73
Table 14 Mixed ANOVAs of health outcomes.............................................................. 75
Table 15 Correlation Coefficients for Demographics and Health Outcomes 76
Table 16 Correlation Coefficients for Demographics and Health Outcomes for AA 77
Table 17 Correlation Coefficients for Demographics and Health Outcomes for CA 78
Table 18 Correlation Coefficients for Demographics and Health Outcomes for LA 79
Table 19 Regression Results for Mental Health Outcomes among AA .............. 83
Table 20 Regression Results for Physical Health among AA....................................... 83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
Fable 21 Regression Results for Mental Health Outcomes among Caucasians....... 85
Fable 22 Regression Results for Physical Health Outcomes among Caucasians..... 85
Table 23 Regression Results for Mental Health Outcomes among Latinas...............87
Fable 24 Regression Results for Physical Health Outcomes among Latinas............87
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vii
List of Figures
Figure 1 Mixed ANOVA for Negative Health Symptoms......................... 75
Figure 2 IPV Main Effects for Depressive Symptoms at T 2 ..................................88
Figure 3 Ethnicity Interaction for Anxiety Symptoms at T 2 ..................................88
Figure 4 Main Effects for Health Diseases T2............................................................89
Figure 5Main Effects for Health Symptoms T2......................................................... 89
Figure 6 Significant Main Effects for Health Quality O f Life................................90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Abbreviations
A A .............................................................. African Americans
Anxiety S x ............................................... Anxiety Symptoms
ASA........................................................ Adult Sexual Abuse
C A ............................................................ Caucasians
C S A ....................... ................................. Childhood Sexual Abuse
Depressive S x ....................................... Depressive Symptoms
Health D x ................................................. Health Diseases
Health S x ............................................... Negative Health Symptoms
H Q L ........................................................ Health Quality o f Life (MOS)
IP V ............................................................ Intimate Partner Violence
L A ............................................................ Latinas
PIPV......................................................... Partner Intimate Partner Violence
RIPV........................................................ Respondent Intimate Partner Violence
SES...................................................... Socioeconomic Status
T1 ............................................................ Time 1 (baseline data)
T 2 .......................................................... Time 2 (12 Month Follow-Up Data)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
Intimate Partner Violence (IPV) is one of the leading killers of women in the
United States. Four million American women experience an assault by an intimate
partner during an average 12-month period. Women of all ethnicities, income levels
and educational backgrounds suffer from intimate partner violence; however, some
ethnic groups may be disproportionately affected by the experience of IPV. The
National Center for Injury Prevention and Control (2003) suggests the need for more
comprehensive data collection on this problem, including data with regard to
sociodemographic risk factors and the long-term indirect mental and physical health
consequences over time for women who experience IPV. A longitudinal
investigation of the physical health and mental health consequences of IPV in a
multiethnic sample of women was conducted. Two questions were considered: 1)
what are the differential sociodemographic risk factors for IPV among African
Americans, Caucasians and Latinas?; and 2) to what extent does IPV impact the
physical and mental health of African American, Caucasian and Latina women over
time, above and beyond, what can be explained by other life-stressors, such as
poverty or histories of sexual trauma?
Secondary data analysis probed these issues among 47 African American, 58
Caucasian and 52 Latina women. Results indicate that, while no ethnic differences in
rates of IPV were found, socioeconomic status and sexual trauma history were
related to the experience of IPV among ethnically diverse women. Additionally,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
X
ethnicity and IPV appear to differentially impact mental and physical health over
time; and ethnicity and IPV appear to interact, disproportionately decreasing mental
health in women of color. These findings suggest the need to move beyond ethnic
categorizations and move toward a comprehensive understanding of how
sociodemographic background, other trauma and IPV interact to decrease health in
women. Clinical implications and areas of future study are discussed. Overall, IPV
should be considered a clear and present danger to women’s mental and physical
health over time; and is related to a consistent and chronic burden that impacts their
mind, body and spirit. This phenomenon, gone unaddressed, will continue to plague
future generations of men, women and children worldwide.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1. INTRODUCTION
Statement o f the Problem
Domestic Violence (DV) or Intimate Partner Violence (IPV) is one of
the leading killers of women in the United States (Erickson & Hart, 1998).
Seventeen hundred women are killed each year by their partners and one in four
women will be beaten by a man in her lifetime (Kemic, Wolf, & Holt, 2000).
Four million American women experience an assault by an intimate partner
during an average 12-month period accounting for almost two million injuries,
more than 550,000 of which require medical attention ( CDC National Center
Injury Prevention and Control, 2003; Fee, Brown, Lavearaus & Therman, 2002).
Women of all ethnicities, income levels and educational backgrounds suffer
from intimate partner violence; however, some ethnic groups may be
disproportionately affected by the experience of IPV (Aciemo, Resnick, &
Kilpatrick, 1997; Office on Women’s Health, 2003).
In addition to the devastating cost to human life, IPV has economic and
public resource costs as well. According to the CDC National Center for Injury
Prevention and Control (2003), intimate partner violence (including rape,
physical assault, and stalking) cost estimates range from $3.4 billion to $7.6
billion annually, almost $4.1 billion due to direct medical and mental health care
services and $1.8 billion in lost productivity from paid work and household
chores for victims. Additionally, physical violence against women accounts for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
40,000 physician visits and over 100,000 hospital days per year (Lee,
! hompson, & Mechanic, 2002). This is compared to the annual $16 billion cost
of cardiovascular disease which affects four times as many (16 million) men and
women and the annual $6 billion cost of all other infectious diseases combined
(Health Behavior News Service, 2003). Intimate partner violence is a major
personal, community and environmental health issue in the United States.
Alarmingly, these reported IPV costs may be underestimated as several
cost components, including criminal justice costs and the cost o f the wide range
of physical effects that are indirectly related to abuse such as headaches,
reproductive health problems, chronic pain, and digestive problems, are often
excluded from reports (CDC National Center for Injury Prevention and Control,
2003). It is also not known whether or not these costs are consistent across
ethnic groups. The Office on Women’s Health (2003) reports socioeconomic
differences with regard to reported physical assaults, rape and stalking among
women; for instance, it is known that lower income is related to higher physical
assault rates. However, not much is known about how specific race, ethnicity
and specific sociodemographic variables (i.e. income) interact with IPV to
impact women’s lives.
The CDC National Center for Injury Prevention and Control (2003)
suggests the need for more comprehensive data collection on this problem
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
including data with regard to indirect health effects of IPV, sociodemographic
risk factors and the long-term indirect physical and mental health consequences
over time for women who experience IPV (Sutherland, Sullivan & Bybee, 2001;
Tjaden & Thoennes, 2000).
Purpose o f the Study
A longitudinal investigation of the physical health and mental health
consequences of IPV in a multiethnic sample of women was conducted. The two
questions considered were: 1) what are the differential sociodemographic risk
factors for IPV among African Americans, Caucasians and Latinas?; and 2) to
what extent does IPV impact the physical and mental health of Caucasian,
African-American and Latina women over time above and beyond what can be
explained by other life-stressors, such as poverty, or histories o f sexual trauma?
Ethnic and socioeconomic differences in IPV were hypothesized, as well as a
differential health impact over time among ethnic women experiencing IPV.
Secondary data analysis was conducted to investigate these issues. The
questions were investigated using partial data from The UCLA-Drew Women
& Family Project (WFP). This NIMH-funded 7-year (1993-2001) longitudinal
study examined the impact of HIV on women’s lives, including life stresses,
relationships, sexual functioning, coping mechanisms and disease progression.
For the purposes o f this study, only the HIV negative sample was used in order
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
to eliminate confounds of decreased health due to HIV. Intimate partner
violence, physical health, and mental health were assessed utilizing baseline and
12-month follow-up data on 47 African Americans, 58 Caucasians, and 52
Latinas.
The Importance o f the Study
This study addresses several important issues. Ethnicity, income, SES,
and sexual trauma variables were statistically examined to better understand the
factors that place women at risk for IPV. Furthermore, the mental and physical
health consequences of IPV in women were carefully assessed utilizing
longitudinal data that detail the length and extent of the health consequences. At
the end of the study, information was gained with regard to socioeconomic risks
for IPV, how experiences with violence impact health over time and the
contribution o f IPV to health disparities among women of color.
Limitations o f the Study
While this data set allowed for the exploration of trends and changes
over time, the data and measures may be low in construct validity for purposes
of the current research. Additionally, the lack of control during the data
collection phase created the potential for a host of confounding variables. Next,
some data may be obsolete, as some of the data were over 10 years old.. Finally,
due to the longitudinal nature of the study, some participants could not be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
retained, creating a reduced sample size. Despite these potential limitations,
however, IPV was investigated in a culturally competent way using this diverse
data set.
Definitions
One of the important distinctions within the violence literature is the
definition of domestic violence (DV), now referred to as intimate partner
violence (IPV). According to Tjaden and Thoennes (2000), there is currently no
consensus among researchers on how to define the term "intimate partner
violence" (IPV). As a result, definitions of the term vary widely from study to
study, making comparisons difficult. The Bureau of Justice Statistics (2000)
reports that intimate partner violence includes murder, sexual assault, robbery,
aggravated assault and simple assault. All of these forms of violence must be
perpetrated by a current or former date, spouse, or cohabiting partner to fall
within the IPV continuum. Within this continuum, cohabiting refers to partners
who live together at least some of the time as a couple. Both same-sex and
opposite-sex cohabitants are included in the definition (Rennison & Welchans,
2000). Unlike the Bureau of Justice Statistics, however, the Center for Disease
Control includes stalking, the willful, malicious and repeated following and
harassing of another person, in its definition of IPV (Meloy, 1998; Saltzman,
Fanslow, McMahon & Shelley, 1999).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
In addition to physical assaults, IPV is often accompanied by
emotionally abusive and controlling behavior. In response, Erickson and Hart
(1998) define IPV as a pattern of coercive control consisting of physical, sexual,
and/or psychological assaults against current or former intimate partners.
Definition Limitations
A few controversies exist within the literature as a result of the range of
definitions of Intimate Partner Violence. The first concern includes whether or
not to limit the definition of IPV to acts carried out with the intention of, or
perceived intention of, causing another person physical pain or injury.
Although this approach presents a definition of IPV that can be readily
operationalized in research, it ignores a multitude of behaviors that individuals
may use to control, intimidate, and otherwise dominate another person in the
context of an intimate relationship (Tjaden & Thoennes, 2000).
A second difference includes limiting the definition of the term to
violence occurring between two persons who are married to each other or living
together as a couple. This definition would narrow the scope o f IPV and would
exclude persons who are dating or coupled but who live in separate housing
(Tjaden & Thoennes, 2000). Finally, most o f the research on IPV focuses
exclusively on violence occurring in marital relationships, heterosexual
cohabiting relationships and heterosexual dating relationships. Few studies have
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
examined violence in same-sex cohabiting or dating relationships (Greenwood
et al., 2002).
For the purposes of the study, IPV was defined as a pattern of coercive
control consisting of physical and/or psychological assaults occurring between
current intimate partners either married (living together or not) or living together
(unmarried). Rates of respondents’ violence toward partners were reported;
however, given this project’s focus on the health impact of violence against
women, only partner’s violence against respondents was analyzed in depth.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
CHAPTER 2. LITERATURE REVIEW
Prior to the 1980s, little was known about the prevalence or incidence of
intimate partner violence in the United States. Although vast evidence suggests
that IPV. is neither a recent nor modem phenomenon, very little research was
conducted on IPV in the United States before this time (Fee, E., Brown, T. M.,
Lazarus, J., & Theerman, 2002). During the 1980s a national strategic plan for
health promotion was released by the United States government significantly
increasing the amount of attention placed on IPV (Lee, Thompson, & Mechanic,
2002). One example of this increase was the passage of the Violence Against
Women Act in 1994. A second example was the establishment of a National
Advisory Council on Violence Against Women (Lee, Thompson, & Mechanic,
2002). As a result of these actions, the national consciousness has been raised
around issues o f violence, especially violence against women. Many theories
have been developed in an attempt to understand the experience o f intimate
partner violence including hypotheses about individual and systemic causes.
Theories of Intimate Partner Violence
Psychosocial Theoretical Perspectives
Several psychosocial theories have described the phenomenon of
violence against women. Bandura’s (1977) social learning theory has been used
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
to explain violence in intimate relationships. This theory argues that gender
related norms are patterned, learned and transmitted from one generation to the
next through the process of socialization. This socialization teaches violence
through reinforcement and modeling by peers, family, societal groups and media
(Gelles & Cornell, 1990; Sharma, 1997). Some research supports this
explanation for intimate partner violence. Lenton (1995a) notes that witnessing
violence between one’s parents has consistently been reported as a risk marker
for abusive men and abused women. Rodgers (1994) found that women in
violent marriages were three times more likely than women in non-violent
marriages to report that their fathers-in-laws were violent towards their spouses
and twice as likely to have witnessed their own fathers assaulting their mothers.
Likewise, Hotaling and Sugarman (1986) found that in 73 percent of 52 studies
reviewed, abused women were more likely to have witnessed violence between
their parents than those who were not abused. Eighty-eight percent of the studies
showed that abusive men were more likely to have witnessed violence between
their parents than husbands who did not abuse their partners (as reported in
Lenton, 1995a).
Other psychosocial arguments suggest a status, power and/or resources
theory o f violence. Straus and Gelles (1990) hypothesize that broader
sociological factors including male dominance in society and in the family cause
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
io
intimate partner violence. Straus and Gelles’ (1990) power theory contends that
when a culture justifies physical violence against women as an acceptable mode
of dominance, and when stress and intra-family conflict are present in the
relationship, men tend to be violent against female partners (Sharma, 1997).
Similarly, Goode’s Resource Theory argues that violence against women
is an attempt to maintain power and control, especially when the perpetrator
does not have other requisite resources. This theory concludes that when a
husband perceives his status as inconsistent with social norms relative to his
wife, and when his wife challenges his male role, he may be more prone to
engage in intimate partner violence (Anderson, 1997).
According to Lenton (1995a), there is support in the literature for various
aspects of the power theory o f violence, including the Straus, Gelles and
Steinmetz (1980) finding that egalitarian marriages have the lowest rate of
abuse, followed by wife-dominant marriages, and husband-dominant marriages
reporting the highest rate of abuse (as reported by Lenton, 1995). Findings that
violence is more common among lower income couples and couples
experiencing bouts of unemployment are also consistent with this theoretical
framework ( Fox, Benson, Demaris, and Van Wyk, 2002; Rennison &
Welchans, 2000).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
However, Fox, Benson, Demaris and Van Wyk (2002) found in their
study comparing Family Stress and Resource Theories (utilizing data from
National Survey of Families and Households and the 1990 U.S. Census), that
data “failed to support either the family stress or relative resource model of
violence, in that neither employment status of the man or the women, singly or
jointly, was directly linked to the odds of violence” (p. 804).
Feminist Theories
Feminist sociologists and psychologists argue that violence is part of a
system of coercive controls through which men maintain social dominance over
women. Male violence is allowed and encouraged by patriarchal societal
structures that sanction women’s subordination to men (Sharma, 1997).
Very little research has been conducted to empirically support feminist
theories of violence (Lenton, 1995a). However, a few studies have found
significant results. For example, Smith (1990) conducted a survey of 604
women and found that men who adhered to an ideology o f patriarchy were the
most likely to abuse their wives. In addition, husbands whose friends and peers
approved of slapping their wives were also significantly more likely to strike
their wives. Other studies have shown that when there is a wide discrepancy
between the husband’s and the wife’s acceptance of patriarchal values, wives are
at greater risk of abuse (Lenton, 1995b).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
There is also much criticism of the Feminist Theory of violence against
women. Johnson (1999) argues that feminist scholars employ single-variable
analyses that concentrate on patriarchy and ignore the impact of factors such as
income, unemployment, and age. These may affect the perpetration of domestic
assaults. Lenton (1995a) argues that any theory based entirely on gender
inequality is “bound to be inadequate because it is not able to deal effectively
with other forms of inequality such as class” (p. 310).
Integrate Theories o f Gender, Race & Class
Some theories have attempted to incorporate the feminist perspective,
social learning, and resource theories with those that apply specifically to
women of color (Sharma, 1997). Anderson (1997) examined the integration of
gender theory with family violence and resource views. In response, she
suggests several alternative hypotheses to intimate partner violence. First, she
predicted that men who have fewer economic resources compared to their
female partners would be more likely to perpetrate domestic assaults. She also
noted that this pattern would not hold true for women, for whom relative
resources would not be associated with violent acts. She predicted domestic
assaults would mediate the effects of ethnicity, age, and cohabitation and
socioeconomic status on intimate partner violence for male respondents. To test
these hypotheses, Anderson (1997) analyzed a data sub-sample of coupled men
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
(n = 2459) and women (n = 2489) of African American, Hispanic, and
Caucasian descent from the National Survey of Families and Households. She
found that all theories received partial support, indicating that sociodemographic
factors play an important role in domestic violence. In particular age, race, and
cohabiting status were correlates o f domestic assaults. Additionally, the United
States Bureau of Justice Statistics has integrated these sociodemographic
variables and found higher rates of violence among younger, poorer, less
educated and unmarried African American, Hispanic and urban couples
(Rennison & Welchans, 2000).
In sum, several theories have been hypothesized to understand the
experience of intimate partner violence. Social Learning Theory,
Resource/Power Theories, Feminist Theory and most recently “Integrated”
Theories have all found at least partial empirical support. However, as argued by
Bograd (1999), “domestic violence is not a monolithic phenomenon” (p. 276)
and intersectionalities must be considered to more comprehensively understand
the experience o f intimate partner violence, particularly when considering the
experiences o f ethnically and socioeconomically diverse women.
Besides theories of why intimate partner violence occurs, there are also
theories about how it occurs in relationships.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
intimate Partner Violence Distinctions
Johnson & Ferraro (2000), in their review o f the family literature on
Intimate partner violence, argue the importance of making distinctions regarding
types or patterns of IPV. They suggest that at the relationship level, there are
different patterns of violence that exist within these violent relationships. In fact,
one can distinguish four major patterns o f partner violence. They include
Common Couple Violence, Intimate Terrorism, Violent Resistance and Mutual
Violent Control.
Johnson and Ferarro (2000) argue that Common Couple Violence arises
in the context of a specific argument in which one or both of the partners lash
out physically at the other. This type of violence is considered the least
dangerous o f the types of IPV (Johnson, 1999, 2000). Intimate Terrorism, in
contrast, is a way to exert control over one’s partner. It involves more per-
couple incidents of violence than does Common Couple Violence and is more
likely to escalate over time, involve serious injury and involve both violent and
nonviolent behaviors. Violent Resistance is a term preferably used over self-
defense. Violent resistance is perpetrated almost entirely by women and may be
one important indicator that a woman will soon leave her abusive partner
(Johnson & Ferraro, 2000). Mutual Violent Control identifies a couple pattern in
which both partner are controlling and violent. Mutual Violent Control involves
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
two intimate terrorists. While this couple pattern is presumed to be rare, it
remains under-researched, as do the other types of intimate partner violence.
While it is suggested that these patterns of violence exist, very little research has
investigated the frequency with which these patterns occur. Johnson & Ferraro
(2000) argue that IPV be further described in terms of the patterns that exist
within these violent relationships; however most studies do not distinguish
patterns of IPV.
Intimate P artn er Violence Demographics
Gender
It has been well established that women represent the majority o f victims
of IPV (Office on Women’s Health, 2003). Approximately 90% of reported
intimate partner violence cases involve women. Women are 10-13 times more
likely than men to be victimized by intimate partners and more likely to sustain
severe injuries as a result of the epidemic. Even worse, women are 30% more
likely than men to be killed as a result o f IPV (Aciemo, Resnick, & Kilpatrick,
1997, Kemic, Wolf, & Holt, 2000).
Men also experience IPV. Rennison and Welchans (2000) report that 3%
o f IPV was committed against men between 1993 and 1998. However, men are
more likely to perpetrate IPV than to be victimized by it. Straus and Gelles
(1990) in their national study o f 8,145 families found that almost 13% of men
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
have been physically aggressive toward their partners in the past year. More
than 33% of those aggressive acts were classified as severe, e.g., punching,
beating up, or threatening with a knife or gun (Kalmuss & Straus, 1990). Gay
men and lesbian women experience domestic violence as well (Greenwood et
al., 2002). However, given the overrepresentation of women in heterosexual
relationships as victims of intimate partner violence, this literature review will
focus on the experiences of women.
The above statistics represent general aggregates of a nation-wide
problem. Most studies include only a predominantly Caucasian sample and do
not include ethnic, racial or economic differences (Anderson, 1997). It is not
known based on the above data whether or not women o f specific
socioeconomic or racial backgrounds are overrepresented in the data, nor is it
known if IPV affects women from varying backgrounds in different ways.
Some researchers suggest that sociodemographic factors influence intimate
partner violence through the greater stress or social isolation experienced by
individuals of lower socioeconomic status or non-white ethnicities (Anderson,
1997). Other scholars emphasize the effects of sociodemographic characteristics
on couple dynamics. Goode’s Resource Theory suggests that individuals lacking
other sources o f power, such as income or educational status, will be more likely
to rely on violence to achieve power within the relationship (Anderson, 1997).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
income and Employment
In order to gain a better understanding of the problem of IPV, it is
accessary to understand both the victims and their partners, particularly from
economic and ethnic standpoints. There is much empirical evidence to suggest
that IPV is found across ethnicities and socioeconomic classes (Aciemo,
Resnick, & Kilpatrick, 1997). Therefore, some researchers choose to de-
emphasize issues such as poverty in favor of a “universality” o f violence
perspective. This perspective argues that IPV occurs in all communities
universally and that economic, racial, or ethnic factors may not be the only
factors, or even most important contributors, to intimate violence (Riger, &
Krieglstein, 2000).
Income
On the other hand, there has been criticism among advocates for battered
women about the under-emphasis on poverty in IPV research. Crenshaw (1991)
criticizes the “universality of intimate violence” as an attempt to focus attention
away from poor and/or minority women and instead focus on white middle-class
victims. Substantial research has shown that although IPV crosses
socioeconomic lines, it is much more frequent among those with low incomes.
One example includes the high rates of violence against women on welfare
(Rennison & Welchans, 2000; Riger & Krieglstein, 2000; Schecter, 1999).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
Another example includes the National Crime Victimization Survey which
consistently reports that the rates of "abusive violence" for poor (under $10,000
annual income) women is more than 3.5 times that o f rates for middle-class
women (over $50,000 annual income) (Bachman & Saltzman, 1995; Straus &
Gelles, 1990). Tolman and Rosen (2001) found in their sample of 753 Michigan
women receiving welfare that 51% had experienced severe abuse in their
lifetime and 14.9% had experienced abuse in the last year. Aciemo, Resnick and
Kilpatrick (1997) report that the “propensity for intimate partner violence has
been found to be inversely related to the male partner’s income and age” (p.55).
Lloyd and Taluc (1999) also examined the connection between income
and IPV. They studied 824 women ages 19-91 in one low-income neighborhood
in Chicago and found that 33% of welfare recipients and 25% o f low-income
non-welfare recipients had experienced "severe aggression" by a partner.
Further, 19% of welfare recipients and 8% of non-recipients had experienced
serious aggression within the past 12 months. Looking at 216 low-income
housed and 220 homeless mothers in Worcester, MA, Browne and Bassuk
(1997) reported that 32% of the women experienced severe physical violence
during the previous two years.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
While these are important findings, it is unclear whether or not these data
are in part due to other variables associated with lower income, including
employment and/or education.
Employment
Rodriguez, Lasche, Chandra and Lee (2001) found that persons receiving
welfare were at significantly higher risk for IPV. In addition to income, they
examined the contribution of employment status, and other individual and
contextual factors to physical aggression during marital conflict. They analyzed
data of 4780 married or cohabitating individuals, 151 of which reported
engaging in IPV. They found that unemployed respondents were at greater risk
of family violence than employed respondents, even after controlling for
income, education, age and other factors. This research indicates that
employment and economics may have to be carefully assessed. However, one
design flaw o f this study was that women and men were not analyzed separately
and ethnicity was not discussed.
Finally, Fox, Benson, DeMaris, VanWyk (2002) attempted to
disentangle the relationship between financial status, employment status and
IPV. They examined the employment status of men and women, the kinds of
work they did, and their feelings about their work as possible predictors of
violence. The data were obtained from the National Survey o f Families and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
Households and the 1990 U.S. Census. Their findings suggested that neither
male nor female employment status had an effect on risk of violence. However,
characteristics of the work one does, the nature of the job, its effects on the
worker and the sense that working is a necessity contributed to the risk of
violence in the home. They found that the feeling one has with respect to work,
whether one is employed in a low-status job and whether the work promotes
irritability, all increased the odds o f man-to-woman violence.
Overall, the data suggest that family financial status is a significant
predictor of IPV. However, the data also suggest that employment and economic
status may be separate issues that may increase the risk for IPV in different
ways. More research is needed to differentiate between factors of employment,
income, education and the relationship to IPV. In addition, ethnicity must be
considered when investigating risks for IPV.
Ethnicity
According to APA (2003), race is the category to which others assign an
individual on the basis of physical characteristics. Ethnicity is defined as the
acceptance o f group’s mores and practices o f one’s culture of origin and the
concomitant sense o f belonging. While many studies have addressed race and/or
ethnicity in the IPV literature, these terms are difficult to differentiate, as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
researchers have yet to make this distinction or address how race or ethnicity
data was assessed in their work (APA, 2003).
According to Tjaden and Thoennes (2000), previous studies have
produced contradictory findings regarding how race and ethnicity affect one's
risk o f intimate partner violence. The Office on Women’s Health (2003) asserts
that violence is not linked specifically to racial or ethnic factors, but is linked to
socioeconomic status. The Bureau of Justice Statistics (1995) also revealed that
white and black women showed equivalent rates of violence committed by
intimates (Lee, Thompson, & Mechanic, 2002). Lane (2003) compared white,
black and Latina women and found that women who had been physically
attacked were only “marginally” (not statistically significant) less likely to be
white.
Other researchers have found a connection between ethnicity and risk of
IPV. When data on African-American, Asian/Pacific Islander, American
Indian/Alaska Native, and mixed-race respondents were combined, nonwhite
women and men reported significantly more intimate partner violence than their
white counterparts (Tjaden & Thoennes, 2000). In addition, the National
Violence against Women Survey (2002) found that rates o f IPV vary
significantly among women of diverse ethnic backgrounds. The survey found
that Asian/Pacific Islander women and men tend to report lower rates o f IPV
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
than do women and men from other communities of color. Furthermore,
African-Americans and American Indian/Alaska Natives report significantly
higher rates o f IPV than other ethnicities (Lee, Thompson, & Mechanic, 2002).
However, some of these differences among minority groups diminished when
the researcher controlled other sociodemographic and relationship variables.
African American Women
It has been found that 52% o f African American women are subject to
physical assault at some point in their lifetimes (primarily by intimate partners),
compared to 53% of Hispanic/Latino women, and 51% of white women (The
Office on Women’s Health, 2003). Rennison and Welchans (2000), in their
Bureau of Justice Statistics special report on IPV, address other
sociodemographic variables that may confound the relationship between
ethnicity and IPV among African Americans. Although they found that between
1993 and 1998, blacks were victimized at significantly higher rates than persons
of any other race, they also found that being young, divorced/separated, earning
lower incomes, living in rental housing and living in urban areas were all
associated with higher rates o f IPV as well.
Latina Women
The Office on Women’s Health (2003) reports that Latina women were
more likely than non-Hispanic women to be victims of physical assaults.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
However, immigration may have been an important confounding factor.
Anderson (1993) reports that, in a sample 157 undocumented Latinas in the D.C.
metropolitan area, 60% were battered by their spouses. In addition, a survey
conducted by the Immigrant Women’s Task Force of the Northern California
Coalition for Immigrant Rights revealed that 34% of Latinas surveyed had
experienced intimate partner violence either in their country of origin, in the
U.S. or in both; and for those who had immigrated, the violence increased since
their move to the United States (Anderson, 1993).
These findings support the need for research on intimate partner violence
among specific ethnic groups. Many studies compare victimization rates of only
one minority group with those of whites (i.e. white versus black victimization
rates), while others compare the victimization rates of all minority groups
(women of color) with those o f whites. Very few studies compared victimization
rates o f several diverse racial groups, as well as other sociodemographic factors
(Tjaden & Thoennes, 2000). Combining data on different ethnicities may
exaggerate differences between whites and nonwhites but, at the same time,
obscure very large differences among ethnic persons.
Some research indicates that there are not only interethnic differences
among women, but specific intraethnic differences as well, including
sociodemographic differences, geographic differences, and even significant
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
tribal differences in rates of violence in the Native American community.
Differences between ethnic women bom in this country and those who have
recently immigrated have been found (Lee, Thomspon & Mechanic, 2002,
Yoshihama, 2002). Lown & Vega (2001) investigated the lifetime prevalence of
physical abuse among women of Mexican origin and found that physical abuse
among Mexican women in America was indeed high; however, being bom in
the United States increased risk for abuse.
Finally, there may be ethnic differences by type of violence. Natera,
Tiburcio, and Villatoro (1997) found high correlations among certain forms of
violence in Mexican American women. Specifically, Hispanic women
experienced IPV in the forms of control over their daily activities, beatings and
violent jealousy. It was also found that those women who reported forced sex
also reported violent jealousy and partner violence toward children. Tjaden and
Thoennes (1998) report Hispanic women were significantly more likely than
non-Hispanic women to report that they were raped by a current or former
intimate partner at some time in their lifetime.
Within these findings, reporting differences should also be considered. It
has been suggested that perhaps significant IPV differences in ethnic
communities may be explained by differences in willingness to report
victimization to interviewers. Actual prevalence of victimization may require
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
further study. According to the National Crime Victimization Surveys, only
about half of IPV is reported to police and it may be that ethnic minority women
such as African American women are more likely to report IPV than are white
women (Lee, Thompson, & Mechanic, 2002).
In sum, research has shown that women from ethnic minority groups
(women of color) may be at more risk for violence than white women.
However, these findings may be confounded by other sociodemographic
realities such as income, employment and/or education that may better detail
differences. There may also be within-ethnic group differences, such as
sociodemographic differences, immigration status, and/or type of violence.
Overall, more research should be conducted to fully understand the experience
of IPV for women of color.
Health Consequences of Intimate Partner Violence
In addition to understanding the prevalence rates and demographic risk
factors for IPV, it is also important to understand the effects of IPV on women.
Physical Health Effects o f Intimate Partner Violence
Heise, Pitanugy, and Germain (1994) estimate that in industrialized
countries, rape and intimate partner violence take an estimated five life years
away from women ages 15-44. In the U.S., violence inflicted by intimate
partners is estimated to be the most comrm cause of women’s injuries. These
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
injuries result in emergency room visits, hospitalizations and even death (Abbott
arid Williamson, 1999). The National Crime Victimization Survey estimated that
52% o f women victimized by an intimate partner sustained injuries, with 30%-
4(5% of these women requiring medical care and 15% requiring hospitalizations
(Kemic, Wolf, & Holt, 2000). According to Lee, Thomspon and Mechanic
{ 2002), between 22% and 35% of emergency room visits by women are
responses to partner violence. Even more disconcerting, approximately 53% of
intimate partner violence victims visit physicians repeatedly (six or more times)
with trauma -related injuries.
Physical Injury
Straus and Gelles (1990) found that intimate partner violence is the
source o f a multitude of physical injuries sustained by women including black
eyes, broken bones, concussions, abdominal injuries and miscarriages. The
most common locations o f injuries among battered women are the face, neck,
upper torso, breast and abdomen (Abbott and Williamson, 1999). Aciemo,
Resnick and Kilpatrick (1997) reported that 33% to 50% o f women who were
physically assaulted by their partners also suffered sexual assaults. These are
the direct consequences of battering that most health care professionals associate
with IPV (Campbell, Jones, Dienemann, & Kub, 2002).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
Hospital Admissions
Kemic, Wolf, and Holt (2000) assessed the history of hospitalizations
among women involved in violent intimate relationships. They compared the
rates of the number of hospital admissions (for diagnoses previously found to be
associated with IPV) among women who had been victims of violence in
intimate relationships to those who reported no abuse. Females (30, 842 white
and 5,690 non-white) ages 18-44 with admissions of injuries, mental disorders,
suicide attempts, assaults, and gastrointestinal disorders were compared. Results
revealed that women known to be exposed to a violent intimate relationship
were significantly more likely to be hospitalized with diagnoses ranging from
psychiatric diagnosis, injury, and poisoning to digestive system diseases and
diagnoses o f assault. These victims were also at risk of attempting suicide in the
year before filing a protection order. The authors concluded an increased relative
risk for overall and diagnosis-specific hospitalizations among abused women.
However, this data combines women of color and compares them to white
women. Not much is known about how these health risks differentially impact
“non-white” women by specific ethnic group.
Indirect Health Consequences
In addition to direct injuries and hospital visits, indirect health
consequences have also been studied (Sutherland, Sullivan & Bybee, 2001).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
Sutherland, Bybee and Sullivan (2002) describe the relationship between IPV,
physical injury and stress, and indirect health consequences. They found
decades of rigorous research that demonstrates that stress as a result of IPV can
seriously compromise physical and psychological well-being.
Negative Physical Symptoms
High levels of stress and/or depression have been linked to various
physical health conditions (Campbell, 2002). The consequences of these health
problems can last far longer than the violent relationship. They include pain or
discomfort from recurring central nervous system symptoms, and illnesses
associated with chronic fear and stress such as functional gastrointestinal
disorders, appetite loss, viral infections such as cold and flu, and cardiac
problems such as hypertension and chest pain (Campbell, Jones, Dienemann,
& Kub 2002). Physical ailments including headaches, reproductive health
problems, and chronic pain have been found as well (CDC National Center for
Injury Prevention and Control, 2003).
Campbell, Jones, Dienemann, and Kub (2002) also compared the
physical health consequences o f abused women versus those who had not
experienced abuse, but who had similar access to health care. African
American, Caucasian and “other” women’s ages 21-55 were compared.
Ethnicity was statistically controlled for in the analyses using direct weights to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
standardize the comparisons. Final results showed that abused women had more
headaches, back pain, sexually transmitted diseases, vaginal bleeding, vaginal
infections, pelvic pain, painful intercourse, urinary tract infections, appetite loss,
abdominal pain and digestive problems. Abused women also had more
gynecological, chronic stress related, central nervous system and total health
issues. Again, ethnicity was statistically controlled for rather than fully
considered in the data.
Finally, Sutherland, Sullivan and Bybee (2001) examined whether
intimate partner violence has a significant effect on women’s health beyond that
which can be explained by poverty (their sample included primarily
white/Caucasians). This study indicated that all women, regardless of income
level who had experienced abuse in the past six months had more significant
health problems than did women with no IPV. This study also demonstrated
that women’s experiences of abuse, and not necessarily their income level,
contributed to their physical health symptoms.
Sexual Assault and Physical Health
Campbell and Soeken (1999b) investigated health consequences of IPV
and differences in types o f violent attack. They compared 159 women (79%
African American) who had forced sex with their partner with women who were
otherwise physically assaulted. It was found that those who were sexually
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
assaulted had higher scores on negative health symptoms, gynecological
symptoms, and risk factors for homicide even when controlling for physical
abuse and demographic variables. While these are important findings, again only
one ethnicity was primarily assessed. Not much is known about how these
findings differ among ethnic communities.
While this research strongly suggests that IPV can have severe longer-
term (chronic) physical and mental health consequences, ethnic differences have
not been carefully assessed. While some like Campbell, Jones, Dienemann, and
Kub (2002) control for ethnic differences rather than highlighting them, others
such as Sullivan and Bybee (2001) predominantly sample one ethnic population.
In order to advance the culturally competent research in this area, ethnicity
should be more carefully assessed.
In summary, research has substantiated that IPV has a variety o f physical
health consequences for women. In addition to direct injuries such as those to
head, back, neck and abdomen, other health conditions such as gastrointestinal
problems, gynecological problems, cardiac problems, and colds and flu, are
common among women experiencing IPV. It is hypothesized that other health
conditions may be an indirect result o f the stress due to IPV. Studies have
typically not investigated how the experience of IPV impacts women’s health
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
over time and whether or not there are ethnic differences with regard to
potentially longer-term health consequences (Campbell & Soeken, 1999a).
Mental Health Consequences
IPV has also been linked to a variety of mental health consequences.
Campbell (2002) reviewed mental health effects of being in violent relationships
and found that depression and Post-traumatic Stress Disorder (PTSD) are among
the most prevalent consequences for victims of IPV. In addition to depression
and PTSD, Sutherland, Bybee and Sullivan (2002) found in a community
sample of 397 women, half of whom had been assaulted by an intimate partner
within the prior six months, other mental health consequences including stress,
anxiety, and substance addiction. In addition, Roberts, Lawrence, Williams and
Raphael (1998) found in their study of 335 women that those who reported
intimate abuse received significantly more diagnoses of depression, dysthymia,
generalized anxiety and phobias.
Differences Across Ethnicity and Socioeconomic Status
Increases in depression in women experiencing IPV has been observed
across ethnicity. Nedd (2001) found in a sample of 128 battered black women
that emotionally abused black women had significantly higher levels of
depression than black women who were not emotionally abused. Among the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
battered black women, those who were more depressed also reported more
gastrointestinal upsets than those who were less depressed.
Similarities have also been found in poorer communities. Tolman and
Rosen (2001) examined IPV, health and economic well being of women on a
welfare caseload. They studied 753 women on welfare (43% white, 57% black)
between the ages of 18 and 54. They found that women who experienced
intimate partner violence in the past 12 months had nearly three times as many
health disorders as their nonabused counterparts. Compared to national norms of
women, the sample had two to three times the prevalence o f depression,
generalized anxiety disorder and lifetime PTSD. However, differences between
mental health issues due to poverty and those due to IPV are unclear.
Although depression, PTSD and anxiety have been found to be common
mental health consequences of IPV, it can also be argued that some battered
women may have preexisting mental health issues that are exacerbated by the
stress of a violent relationship. For example, depression in battered women is
also associated with other life stressors that are often highly correlated with
intimate partner violence including daily stressors such as multiple children,
marital discord, change in residence and child behavior problems (Street, King,
King & Riggs, 2003). A history of child sexual abuse (CSA) is also highly
correlated with IPV and related to devastating mental health effects including
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
PTSD, depression and anxiety (Dilillo, Giuffre, Tremblay & Peterson, 2001;
Noll, Horowitz, Bonanno, Trickett, & Putnam, 2003; Wyatt, 1990; Wyatt,
Axelrod, Chin, Carmona & Loeb, 2000; Wyatt, et al., 2002). It could be that
other life stressors linked to IPV, and not IPV in and of itself, result in declined
health.
Golding (1999) conducted a meta-analysis to address the onset of mental
health conditions in relationship to IPV. Specifically, she reviewed published
literature on battering as a risk factor for depression, suicidality, PTSD, alcohol
abuse/dependence, and drug use or dependence. The meta-analysis evaluated
consistency with regard to time, place and circumstance across multiple studies,
and the extent to which the IPV/mental health disorders are consistent with the
actual passage of time. She determined that IPV caused a mental disorder only if
the disorder occurred after the violence. She concluded that although most
research has not addressed many criteria for causal inferences, the existing
research is consistent with the hypothesis that intimate partner violence
increases risk for mental health problems. Results across 18 studies for
depression found that the prevalence of depression among battered women was
47.6%, which is higher than general population rates, (from 10.2% to 21.3 %).
Thirteen studies of suicidality found prevalence rates of suicide 4.6%-7.7%
among battered women, with a weighted mean o f 17.9%. 63.8% of 11 studies of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
PTSD, 18.5% in 10 studies of alcohol use and 8.9% in four studies of drug use.
Weighted mean odds ratios representing associations of these problems with
violence ranged from 3.55 to 5.62 and were typically consistent across studies
(Golding, 1999). While this is an important meta-analytic study, ethnic
differences among the studies were not discussed and leave important questions
unanswered.
Street, King, King and Riggs (2003) attempted to understand the nature
of the psychological distress and the cognitive perceptions that increase risk for
victim depression. They investigated the association among partner violence, a
married woman’s psychological distress and children’s behavior problems. They
found that while a direct relationship between married woman’s experiences and
her psychological distress was found, this direct relationship was small (not
significant). The association between violence and wife's psychological distress
flowed indirectly through an intermediary variable - wife’s perspective on
family functioning. In the model, wife’s perspective on family functioning was
influenced both by the presence of male violence in the relationship and by her
husband’s perspective on the functioning o f the family.
Overall, it has been substantiated that there is a strong relationship
between IPV and mental conditions such as depression, anxiety, PTSD and
suicidality. However, much work still needs to be done to determine a casual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
association between these mental conditions and IPV. Ethnic differences with
regard to these mental conditions should be more carefully investigated, as
women from different ethnic backgrounds may have different mental health
experiences of IPV as well.
To conclude, women with experiences of IPV are likely to suffer from an
array o f mental health problems. The associated physical and mental health
consequences of IPV are important findings. However, most studies have not
fully addressed the ethnic differences that may be confounding data in this area
nor have they fully addressed the length or extent of these physical and mental
health illnesses (Sutherland, Bybee & Sullivan, 2002). There is also much
debate about whether health conditions can be directly attributed to IPV. Some
argue that certain women may be biologically pre-disposed to some of these
health conditions, while other women may have life stressors that account for
these health consequences. In addition, Campbell and Soeken (1999a) report that
many of the physical health studies on women in violent relationships are cross-
sectional and maybe the result of interactions of other lifestyle and high risk
behaviors. These researchers argued that lifestyle choices may account for some
o f these conditions; therefore, more culturally competent longitudinal research
should be conducted to further investigate the extent of these physical health
problems (Campbell, 2002).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
More research is needed to address complex issues involved in women’s
health. While some of these physical conditions may be the direct result of IPV,
some women may be predisposed to physical health problems or they may be
the result of other life stressors that exacerbate poor health. In order to better
understand the physical and mental health consequences due to violence, it is
important to understand women’s health and the factors that predispose women
to poorer health in general.
General H ealth Concerns of Women
Striepe and Coons (2002) reviewed the research on women’s health in
primary care settings and found that some medical conditions such as
fibromyalgia, chronic fatigue syndrome, headaches and chronic pelvic pain are
more prevalent among women than men. It is also known that women with some
health issues are often labeled with psychological diagnoses as well (Striepe &
Coons, 2002). In addition to these physical conditions, major depression,
dysthymia, generalized anxiety, and panic disorder are also common conditions
for women in primary care settings.
Women’ s Health and Poverty
Studies have found that poverty affects the life expectancy and the
overall length of life and health o f both men and women (Karlsen & Nazroo,
2002; Nazroo, 2003). Astone, Ensminger, Juon (2002) investigated the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
predictors of life longevity in women and found that after adjustment for age,
the death rates were higher for women who lived in households with incomes
below the poverty line. The effect of low education also persisted after
adjustment for all other predictors.
Physical Health and Women o f Color
Great health disparities exist among ethnically diverse women as well
(Rritek, Hargraves, Cuella, et al., 2002). Across all racial/ethnic categories, the
mortality rates for women are considerably lower than for men. However,
minority women continue to lag about five years behind Caucasian women in
life expectancy (Agency for Healthcare Research and Quality, 2003). African
Americans, Hispanics, American Indians and Asian Americans/Pacific Islanders
face disproportionate rates of illness and disease (Littelfield, Robinson,
Engelbrecht, Gonzalez, et al., 2002). There are several conditions that are among
the 10 leading causes of death for only one ethnic population. Suicide is a
leading cause o f death (8th) for only Asian/Pacific Islander women. HIV/AIDS is
a leading cause of death (10th ) for only African American women, and
congenital anomalies (10th) only for Hispanics and Alzheimer disease (ranked
8th) only for whites. Therefore, it is important to understand health disparities
across ethnicity more specifically.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
However, it is difficult to determine reasons for many of the health
disparities among women. Data for Hispanics and from Asians/Pacific Islanders
have only recently become available (Ro, 2002). In addition, there may be
intraethnic differences that may confound overall interethnic differences.
According to Ro (2002), little is known about the health status o f Asian
American/ Pacific Islander (AAPI) women, their access to, and their use of,
health services and this may be the result of a great heterogeneity in the AAPI
community.
Another limitation in understanding causes of these health disparities is
the biological versus environmental debate. While biology is implicated in
accounting for differences between ethnic groups, other issues such as SES,
ethnicity or experiences associated with ethnic minority status such as racial
discrimination may also play a major role in these health disparities (Gee, 2002).
The U.S. Department of Health and Human Services (2003) has recently
presented a charge to the mental and physical health communities to better
understand and eliminate health disparities among women o f color.
Despite the difficulties in assessing health disparities between of
different ethnicities, what is known is that Caucasian women generally have
better health. Hypertension has been found to affect African American,
Mexican American, Puerto Rican, Native Hawaiian significantly more than
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
Caucasian women. Also, diabetes is the fourth leading cause of death for
African American, Indian and Hispanic women and these rates have increased in
recent years (Agency for Healthcare Research and Quality, 2003).
African American women
Caucasian women have a life expectancy at birth that exceeds that of
.African American women peers by 5.2 years (Kumanyika, Morssink & Nestle,
2001; Williams, 2002). African American women have higher mortality rates
from a number of diseases than do Caucasian women, including heart disease,
stroke, and most cancers (all among the leading causes of death for African
American women) (Office on Women’s Health, 2003). African American
women have the highest death rate from stroke of all women, at 78.1 deaths per
100,000 (in contrast to 57.8 for white women). Among younger and
reproductive-aged women, maternal mortality and homicide rates are 4 times
higher for black women than white women, and the rate of HIV related deaths is
12 times higher for black women (Office on Women’s Health, 2003; Williams,
2002). Moreover, Kumanyika, Morssink and Nestle (2001) reports that breast
cancer incidence is similar for black and white women, but black women have
higher breast cancer mortality.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
Latina Women
Among Hispanic women, mortality rates from several diseases are lower
than those of white women, including stroke, chronic obstructive pulmonary
disease, and cancers of the respiratory system and breasts. However, mortality
rates from AIDS and homicide are significantly higher for Hispanic women than
white women (The Office on Women’s Health, 2003). Interestingly, despite
decreased mortality rates among Latinas, Amaro and De la Torre (2002) found
that more Latinas than Caucasians perceive themselves as being in fair to poor
health. Self-perceptions of health differ among Latina subgroups, with larger
percentages of Puerto Ricans (17.5%) and Mexican Americans (16.2%) than
members of other Latinas reporting fair or poor health.
Mental Health and Women o f Color
In addition to physical health disparities, women of color also experience
mental health disparities as well. The Office on Women’s Health (1993) reports
that Hispanic/Latina women have the highest lifetime prevalence of depression
(24%) among all women. Nearly twice as many Hispanic women reported being
depressed (11%) as African American women (6%) and Caucasian women
(5%). Although, African American women are less likely to have this disorder
(16%) than Caucasian women (22%), Nedd (2001) reports significant depression
among African American women that often goes unreported. Leigh & Jimenez
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
(2002) reports that of those suffering from depression, almost half (47%) of
African American women are afflicted with more severe depression.
In general, poor and/or ethnic minority women experience worse health
than Caucasian women. Women of color (particularly African American
women) are at greater risk for mortality, for hypertension, diabetes, suicide, and
breast cancer and Latina women are more likely to perceive their health as
poorer than that of Caucasians. Women of color also experience mental health
disparities as well, with Latinas reporting the highest rates o f depression. Little
is known about how these findings intersect with the IPV literature.
Conclusion
First, it was found that many one dimensional theories of IPV including
Social Learning theory and resource/power theories give only a partial look at
domestic violence and that theories that address the intersection o f IPV and
other issues such as race, income and gender intersect should be considered.
Empirical work has investigated the merit of these theories; however, findings
have been inconsistent.
Moreover, with regard to sociodemographics, women are at more risk for
IPV victimization than men. In addition, IPV is found across income, race and
ethnicity. Poorer women, unemployed women, and women o f color appear to be
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
at increased risk. However, much of the sociodemographic data does not clearly
differentiate between results due to ethnicity versus income or education.
Next, physical and mental health consequences were found for women of
all income levels and all ethnicities. Many physical health consequences such as
direct injuries to head, back, neck and abdomen and other health conditions such
as gastrointestinal problems, gynecological problems, cardiac problems, and
colds and flu, are likely in women experiencing IPV. They are also likely to
experience an array of mental health problems including depression, anxiety,
PTSD, suicidality and increased substance use. Sutherland, Sullivan and Bybee
(2001) argue that the long-term aftermath of direct IPV injuries and the fear and
stress associated with IPV may lead to severe psychological consequences,
which then in turn may lead to more long-term less obvious physical health
problems. While these findings appear to be consistent ethic differences in
physical and mental health due to IPV are unclear.
There is much debate about whether or not these health conditions can be
directly attributed to IPV. Some women may be biologically pre disposed to
these health conditions, while others may have life stressors that account for
some of the mental and physical health consequences. Women’s health was
evaluated to better understand the physical and mental health consequences of
violence. Women of color are disproportionately affected by a myriad of health
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
problems including hypertension, diabetes, cancer, depression, anxiety, and
PTSD all which contribute to an increased risk for earlier death.
More research is needed to understand the experience o f IPV in women
of all ethnicities. Variables such as income, employment and ethnicity should be
more carefully extracted to better understand the prevalence and incidence of
IPV in all women. Furthermore, the physical and mental health consequences of
IPV in women should be carefully examined using more longitudinal data that
may better detail the length and extent of the health consequences. Finally, other
pre-existing health conditions, particularly in vulnerable populations such as
women living in poverty or women of color, should be considered so that
differentiations between health conditions can be determined.
The Current Study
Major Research Questions
A longitudinal investigation of the mental and physical health
consequences o f IPV in a multiethnic sample o f women was conducted. Two
major research questions were considered.
The first question posed was: what are the differential sociodemographic
risk factors for IPV among African Americans, Caucasians and Latinas? While
there are a number of sociodemographic risks factors that appear to put women
at risk for violent relationships, it is not known whether or not these risk factors
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
are different among women of different ethnic groups. It may be that different
sociodemographic factors, such as income, put one ethnic group o f women at
risk but not another.
The second major question posed was: to what extent does IPV impact
s he physical and mental health of Caucasian, African American and Latina
women over time above and beyond what can be explained by other life-
siressors, such as poverty or histories of sexual trauma? The current literature
suggests that there are great mental and physical health disparities among
women of color (U.S. Department of Health and Human Services, 2003). Some
argue that these health disparities may be due to biology or socioeconomic
stress, while others argue that racism contributes to these disparities. This study
attempts to identify the role IPV plays in these health disparities.
Hypotheses
Four hypotheses were explored:
1) African American and Latina would experience higher IPV than
Caucasian women. The current literature suggests that African American and
Latina women are likely to report higher IPV than Caucasian women. This study
attempted to confirm and support previous findings in this area.
2) Lower income, lower education, higher chronic burden, and a history
of sexual trauma would be related to higher IPV among all ethnic groups. Data
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
also suggest that, across ethnic group, these socioeconomic factors appear to put
women at risk for higher IPV.
In addition, as detailed in research question 2, analyses were conducted
to explore the relationship between ethnicity, IPV and sociodemographic
variables. Socioeconomic factors may put one ethnic group at risk for IPV more
than another. While these relationships were explored, hypotheses were not
made about which variables would be related to IPV for each ethnic group. It
was only hypothesized that in general, lower income and education, and higher
chronic burden were related to more IPV and that the strengths of these
associations differed across ethnic groups.
3) IPV is related to decreased health status among all women. While this
trend has been well documented in the current literature primarily among
Caucasian women, this study attempted to confirm this relationship among
African American and Latina women.
4) Finally, ethnicity and IPV would significantly interact, with IPV
moderating ethnic differences in health over time. Specifically, it was predicted
that for women of color (African Americans and Latinas), health would decline
over time, with ethnicity and IPV significantly contributing to this decline.
The project explored the relationship between ethnicity, SES, IPV, and
changes in health over time. It was hoped that information would be gained with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
regard to differential sociodemographic risk factors for IPV across ethnic
groups, how intimate partner violence impacts health over time, and the
contribution of IPV to health disparities among women of color.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
4 7
CHAPTER 3. METHODOLOGY
The Data Sample
The hypotheses were investigated using data from The UCLA-Drew
Women & Family Project (WFP). This NIMH-funded 7-year (1993-2001)
longitudinal study examined the impact of HIV on women’s lives, including life
stresses, sexual functioning, coping mechanisms and disease progression. For
the purposes o f this study, only the HIV negative sample was used in order to
eliminate confounds of decreased health due to HIV. The study used baseline
and 12-month follow-up data on a sample of 157 participants consisting of the
three largest ethnic groups in the data set, including 47 African Americans, 58
Caucasians, and 52 Latinas. Asian Americans and Native Americans were
excluded due to small sample size.
Recruitment
290 HIV positive women were recruited from Los Angeles sites
providing HIV services, including county hospitals, community-based clinics,
ethnic- and AIDS -specific organizations and drug rehabilitation centers.
Prospective participants responded to flyers, and other media sources (such as
print ads, and radio). The Institute for Social Science Research at UCLA used
random-digit dialing and 1990 census data to identify a stratified random sample
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
48
of 200 HIV negative women who matched the positive women in ethnicity, age,
education, marital status and geographic residence (Wyatt & Chin, 1999).
Prospective seropositive and seronegative participants were screened and
enrolled if they were female, 18 years of age or older, self-identified as members
of the five major ethnic groups: African American, Latina, Caucasian,
Asian/Pacific Islander, and American Indian; and showed no evidence of severe
psychiatric, neurocognitive or other physical limitations.
Procedure
Eligible women were invited to participate in a 3 to 4 hour face-to-face
interview conducted by a trained, ethnically and linguistically matched female
interviewer. This procedure was used in order to reduce possible culturally
mediated obstacles to effective communication. All participants were paid $50
per session.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
4 9
Measures
Demographic Information
Table 1 Demographic Description of the Sample
AA (N = 47) CA (N= 58) LA (N= 52) Total (N= 157)
M = M = M = M =
Age (In Years) 31.42 34.29 30.37 32.56
Income/Month $2 0 1 2 . 1 0 a $2590.21 b $898.66a£> $1868.32
Education (In
Years)
13.52 ab 15.31 ac 1 0 . \4bc 13.09
Chronic Burden 33.85 31.76 32.14 32.52
Relationship
Status
14(31 %)ab 39 (6 6 %)a 29 (56%)b 82 (53%)
CSA 46% 37% 39% 41%
ASA 35% 32% 1 2 % 26%
RIPV 1 2 . 0 2 11.24 12.67 11.94
PIPV 1 2 . 6 8 141.57 12.63 12.25
* = g < .05, **p=<.005
Ethnic differences are discussed in the Results Section (p. 47)- a = significant ethnic differences
between African Americans and Caucasians, b = significant differences between African
Americans and Latinas, c = significant differences between Caucasians and Latinas
Table 1 reports demographic information for the sample.
Ethnic/racial Heritage was based on self-identification. Participants
were asked “what is your ethnic or racial heritage?” They were given five
options: Black/African American, White/Caucasian, Latina/Hispanic, mixed or
other. Given the nature o f the question asked, racial versus ethnic distinctions
could not be made. Results are reported in terms o f ethnic identity. Data were
analyzed on 47 African American, 58 Caucasian, and 52 Latina women. Data on
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
50
Asian American and Native American women were excluded due to small
sample size.
Age was assessed via birth date. The sample was 32.5 years old.
Total Household Income was calculated by summing total income
earned from self, family members, general assistance and other sources. The
average income for the sample was $22,500/year.
Education included the number of years of formal schooling. Most
participants had at least a high school education.
Relationship Status A participant was considered in a relationship if they
reported either currently being married (living or not living with spouse) or
living with a partner (unmarried). Fifty-three percent of the sample reported
being involved in a relationship.
Immigration Status/Citizenship. Immigration status was assessed among
the Latina population via the question, “Are you a U.S. Citizen?” 44% of the
population reported being a U.S. citizen.
Chronic Burden was used to assess a global understanding of
socioeconomic stress including housing, financial and environmental stressors.
This 21-item, 4-point likert-scale measure asked the extent to which stressors,
including “not having enough money to cover the basic needs o f life (food,
clothing, housing),” “no reliable source o f transportation such as a car that
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
51
works or reliable bus service,” “living in a high crime area”, “not having any
savings to meet problems that come up,” “work hours or problems change for
the worse” and “problems arranging childcare” were problems in the last six
months, where 1 = not a problem in the last six months to 4 = A major problem
for me in the last six months. The measure was summed to yield an assessment
of the chronic life stressors or burden that may impact life functioning. Possible
scores ranged between 21 and 84. Item analyses were conducted on the 21 items
hypothesized to assess SES. Coefficient alphas were calculated for the entire
sample (.81) and for the African Americans (.77), Caucasians (.84) and Latinas
(.80) samples separately. The mean chronic burden for the sample was 32.52,
falling in the “a little problem in the last six months” range.
Trauma History
Child Sexual Abuse (CSA) Child Sexual Abuse (CSA) was determined
by an affirmative answer to any of the nine questions posed to respondents
regarding sexual experiences that occurred before the age of 18 years old, both
coerced and noncoerced. Examples o f questions include, “Did a relative, family
friend or stranger ever feel up or fondle your body, including your breasts/penis
or genitals, or attempt to arouse you sexually?”; “Did anyone rub their genitals
against your body in a sexual way?”; or “During childhood and adolescence, did
anyone attempt to have intercourse with you against your will?” (Wyatt Sex
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
52
History Questionnaire, 1985). Forty-one percent of the sample reported at least
one incident of CSA.
Adult Sexual Abuse (ASA). Adult Sexual Abuse was determined by an
affirmative answer to either o f two items “Since the age of 18, have you been
raped, or has someone forced their penis or object in your vagina or bottom?” or
“Since the age of 18, has anyone ever tried to rape you?”. ASA is assessed
separately from IPV in order to assess distinct differences between rape histories
and current IPV (Wyatt Sex History Questionnaire, 1985). Twenty-six percent
of the population reported at least one incident of Adult Sexual Abuse.
Intimate Partner Violence
Intimate Partner Violence (IPV) was assessed using seven items from the
revised Conflict Tactics Scale (Straus, Hamby, Boney-McCoy, & Sugarman,
1990). Each participant was asked whether in he last six months if she or her
partner quarreled, shouted, insulted, threw, smashed, hit, kicked something;
slapped/physically attacked, or threatened or used a knife or gun. Items were
added to create sum scores of Respondent and Partner Intimate Partner
Violence. Responses ranged from 1 = Never to 5 = Always. Possible scores
ranged from 5 to 25. Item analyses were conducted on the seven items
hypothesized to assess IPV. Respondent behaviors and partner behaviors
(through respondent report) were assessed at baseline for the entire sample and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
53
for each ethnic group. Initially each of the items was correlated with the total
score for IPV. All correlations were greater than .30 except for one item:
‘ Threatened with knife or gun” (r = .18). This item had little to no variance
among the sample and was eliminated from the scale. Coefficient alphas for the
IPV scales for the entire group and for each ethnicity are reported in Table
2. The coefficient alpha for Respondent IPV (RIPV) = .76, and for Partner IPV
(PIPV) = .86.
Once scaled, correlation coefficients were computed among two IPV
scales: respondent IPV (M = 11.94) and partner IPV (M = 12.25). The
correlation between the two measures was statistically significant (.47). Partner
IPV (PIPV) was used as the measure of IPV in the remaining data analysis given
its higher scale reliability and the project’s focus on violence against women.
Table 2 Scale Alphas for IPV Measures
IPV African
Americans
(AA) N = 47
Caucasians
(CA)
N = 58
Latinas
(LA)
N = 52
Total Sample
N = 157
Respondent
IPVT1
.76 .78 .76 .76
Partner IPV
T2
.90 .82 .86 .86
M ental Health Outcomes
Depressive Symptoms (Depressive Sx) The Center for Epidemiological
Studies-Depression (CES-D) Scale (Radloff, 1977). is a widely used, self-report
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
54
measure of depressive symptoms that includes 20 items, each rated on a 4-point
scale that ranges from rarely or none of the time (0) to most or all o f the time
(3). Total scores for the CES-D can range from 0-60, with scores greater than or
equal to 16 indicating clinically significant levels of depressive symptoms.
Reliability coefficients for the CES-D range between .80 and .90. Reliability for
this sample was found to be .90.
Anxiety Symptoms (Anxiety Sx) The Symptoms Checklist for Anxiety is
a widely used, 15-item, 4-point likert scale, self report measure of anxiety
symptoms. Test-retest reliability coefficients range between .80 and .90
(Degratis, 1976). Reliability for this sample was found to be .92.
Physical H ealth Outcomes
Physical health was assessed via questions measuring health quality of
life (HQL) measured by the Medical Outcomes Survey Short Form 20 item
scale, health diseases, and negative health symptoms.
Health Quality of Life (HQL). To evaluate general health functioning,
activities o f daily living and general health quality of life, the widely used
Medical Outcomes Survey-Short Form, a measure of quality o f life impacted by
general health, was used (Stewart, Hays, & Ware, 1988). Questions such as
“How much bodily pain have you generally had during the past month?” and
“How much of the time during the past month did you feel worn out?” were
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
55
summed to yield a score measuring health quality of life. Reliability coefficients
range from .81 to .88 for the multi-item scale (.80 for this sample).
Health Diseases ( Health Dx). Seven dichotomous health questions
(“Have you experienced any of the following in the past six months - yes or
no”) assessing diagnosable diseases and/or chronic illnesses: high blood
pressure, cancer, arthritis, asthma, heart disease, anemia, and Candida, were used
to compile the Health Diseases Scale. Means score represent mean number of
diseases reported at T1 and T2. The final coefficient alphas for the sample (.68
at T1 and .64 at T2) and individual ethnic groups are listed in
Table 3.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
56
Negative Health Symptoms (Health Sx) Seven dichotomous health
questions (“Have you experienced any of the following in the past 6 months -
yes or no”) assessing negative symptoms, such as diarrhea, night sweats, fever,
colds, skin rash, weight gain, and numbness, were used to compile the Negative
Health Symptoms Scale. Mean scores represent the average number of total
negative health symptoms experienced at T1 and T2. Item analyses were
conducted on the scales for the sample and individual ethnic groups at baseline
and post (12 month) time points (.68). The final coefficient alpha for each
ethnic group is listed in
Table 3.
Table 3 Scale Alphas for Outcome Measures
AA CA LA Total Sample
Health Diseases T1 .74 .70 .62 .68
Health Symptoms T1 .63 .72 .67 .68
Health Diseases T2 .67 .50 .61 .64
Health Symptoms T2 .70 .67 .35 .68
Table 4 reports correlation coefficients among the 10 outcome variables
(five baseline and five post) among all outcome measures.
Table 4 Correlations among all Mental and Physical Health Outcome Measures
DS AS HQLT1 HD HS DS AS HQL HD HS
T1 T1 T1 T1 T2 T2 T2 T2 T2
Depressive
S x T l
1.0
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
57
Anxiety
Sx T1
4 9 **
1 .0
H Q LT1
Health Dx
T1
.37**
.0 1
.45**
.04
1 .0
-.26** 1 .0
Health SX
T1
-.04 .18* -.24** .28** 1 .0
Depressive
Sx T2
.56** .58** -.51** .17
2 9 **
1 . 0
Anxiety
Sx T2
.47** .52** -.63** .17 .18* .72** 1 .0
H Q T 2
Health Dx
T2
.45**
.0 1 -.08
.60**
-.13
.25**
.54**
-.15
. 0 2
7]**
.06
.63**
.19*
1 . 0
-.15 1 .0
Health Sx
T2
.09 .09 _ 2 2 ** .47** .46** .23** .28**
.28**
.41** 1.0
* =
= g < .05
* * E =
<.005
Data Analyses
Demographics
First, a series of analyses o f variance, Chi-Square tests, and Pearson’s
product moment correlations were conducted to test hypotheses 1, 2 and 3.
Ethnic group experiences in all demographic characteristics including age,
economic status, and education, chronic burden, relationship status, U.S.
citizenship (among Latinas) experiences of sexual abuse (CSA and ASA), the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
58
experience of IPV and mental and physical health status at T1 and T2 were
explored.
Assessing Health Stability (Changes in Health status) Over Time
Next, repeated-measures ANOVAs were conducted by ethnicity and IPV
to detect changes in mental and physical health over time.
Multiple Regression
To answer the question o f whether or not ethnicity and IPV have a
unique influence on health status at T2, after controlling for health at T l,
chronic burden and ASA (hypothesis 4), 15 blockwise multiple regression
models were run (3 ethnic groups X 5 outcome variables), with four blocks of
variables.
Step 1 included each respective health outcome variable at Time 1 and
other significant background variables, chronic burden and ASA, in order to
control for them.
At Step 2 ethnicity was entered as a dichotomous dummy-coded variable
(either AA or not, CA or not or LA or not).
At Step 3 IPV was entered.
A Step 4 an ethnicity X IPV interaction term was entered in order to
assess the combined effect of ethnicity and IPV on mental and physical health
(Aiken & West, 1991). The regression equation was: Y = b (health at Tl + SES
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
and sexual trauma) + b (ethnicity) + b IPV + b (ethnicity X IPV), where Y
each health outcome variable at T2.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
60
CHAPTER 4. RESULTS
Demographic Description
Ethnic Comparisons of Demographic Variables
Table 1 reports the demographic description of the sample including
means for age (M^] g ro u p s = 32.5 years), income, (M^i g ro u p s = $ 1,868/month)
education (Man g ro u p s = 13.8 years), Chronic Burden (Man g r0 u p s = 32.5) and
number of participants in relationship (N % - 53% in a relationship). 44% o f the
Latina population reported being a U.S. Citizen.
Significant ethnic differences were found for income [F (2, 158) = 13.32,
p = .00], and education [F(2, 158) = 13.82, p = .00], African Americans reported
significantly higher income than Latinas. All groups were significantly different
in education status, with Caucasians reporting the most education, followed by
African Americans, then Latinas.
A two-way contingency table analysis was conducted to evaluate ethnic
differences in relationship status. Significant ethnic differences were found
[Pearson % 2 (2, N = 158) = 12.86, p_ = .00, Cramer’s V = .27]. Follow-up
pairwise comparisons were conducted to evaluate the difference among these
proportions. African American were significantly less likely to be in a
relationship than Caucasians or Latinas.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
61
Correlations Among Demographic Variables Within Groups
Correlation coefficients for each ethnic group were conducted among the
demographic variables: age, education, total household income, chronic burden
and relationship status and citizenship status among Latinas.
For the entire group, chronic burden was found to be significantly
negatively correlated with education -.17. It was found that the lower the
education, the higher the chronic burden. Also, those who were married were
likely to be older (f (156) = 13.88, p = .00), and have higher incomes [F (156) =
4.7p_ = .03] (see Table 5). Among Latinas, citizenship was related to age (-.32, p
= .02), education (.57, p = .00). Latinas either bom here or immigrated here were
more likely to be younger and more educated.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 5 Ethnic Differences in Correlation Coefficients for Dem ographic Variables
N = 172 Age Education Income Chronic
Burden
Age -
AA -
CA -
LA -
Education .1 1 -
AA -.14(b) -
CA -.0 1 (c) -
LA -.56** (be) -
Income -.07 - -
AA -.26 .69**(ab) -
CA -.36** ,28*(a) -
LA - . 0 2 . 2 0 (b) -
Chronic Burden .0 1 -.17* -.05 -
AA - . 0 2 -.17 -.47**(a) -
CA .0 1 -.27* . 1 1 (a) -
LA -.05 -.19 - . 2 2 -
Relationship Status .2 0 * . 0 2 .1 2 -.04
AA .04 (b) - .0 1 .06 . 0 0
CA .09 .05 .14 . 0 2
LA .37** (be) -.07 .07 -.04
* = £ < . 0 5 , **£ = < 0 0 5
a = Significant correlation between African American and Caucasian
b = Significant correlation Between African American and Latina
c = Significant correlation between Latina and Caucasian
Fisher’s R-Z correlation conversions (Cohen, 1988) were conducted
order to assess ethnic differences in the correlations among the variables.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
63
Table 5 reports significant differences between African Americans and
Caucasians (a), African Americans and Latinas (b) and Caucasians and Latinas
(c).
Correlations between income and education; income and chronic burden;
and age and whether one was involved in a relationship or not, were
significantly different among the three ethnic groups. The relationship between
income and education was much stronger among African Americans than
Caucasians or Latinas.
In addition, the relationship between income and chronic burden was
significantly different. For African Americans, lowered income significantly
increased chronic burden; however, this relationship was significantly different
among Caucasians, for whom this was not the case (See
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
64
Table 5). Finally, age significantly influenced relationship status
among Latinas. This was significantly different among African Americans. Age
did not significantly influence relationship status.
In sum, the sample was approximately 32.5 years old, had at least a high
school education, earned approximately $22,400/year, suffered, on average, “a
little” chronic burden (M = 1.5; scale item range 1-4) and most (53%) were in a
relationship. 44% OF Latinas reported being U.S. citizens. Ethnic differences
were also found. Latinas reported significantly lower income and education,
while Caucasians reported the highest education. African Americans were
significantly less likely to be in a relationship than Latinas or Caucasians.
Among Latinas, citizenship was related to age and education. No differences in
age or chronic burden were found.
Trauma Characteristics Among The Population
In order to gain a better understanding o f the experience of background
trauma that may be correlated with IPV and health outcomes among the women,
sexual violence was assessed.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
65
Table 6 reports percentages of sexual trauma in the participant
population. Forty-one percent reported at least one incident of child sexual abuse
Ti'SA) and 26% reported at least one incident of adult sexual abuse (ASA). A
Chi-Square analysis was conducted to evaluate the relationship between those
who had experienced CSA and those who had experienced ASA. For the
sample, CSA and ASA were found to be significantly related [Pearson
157) = 7.63, pl= -01, Cramer’s V = .26]. Those with histories of CSA were
likely to also have histories of ASA. This relationship existed among Caucasians
[Pearson % 2 (2, N = 58) = 11.62, p_ = .00, Cramer’s V = .44], however, not
among African Americans or Latinas.
Table 6 Rates of Sexual Trauma in the Sample
AA Caucasian Latina Total Pearson
X 2 (d f = 2)
Cramer’s
v
CSA 22 (46%) 22 (37%) 20 39%) 64(41% ) 1.47 . 1 0
ASA 17 (35%)b 19 (32%)c 6 ( 1 2 %)bc 42 (26%) 8.61** .23
* = g < .05, **^ = <.005
Ethnicity Differences in Traumatic Sexual Experiences
A two-way contingency table analysis was conducted to evaluate ethnic
differences in Child Sexual Abuse (CSA) and Adult Sexual Abuse (ASA) .
Ethnicity and ASA were found to be significantly related [Pearson % ( 2, N =
157) = 8.61, p_ = .01, Cramer’s V = .23]. Follow-up pairwise comparisons were
conducted to evaluate the difference among these proportions. Three times as
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
66
many African Americans and Caucasians reported ASA as Latinas (g <.05).
Ethnicity and CSA were not found to be significantly related (see Table 6).
CSA and ASA were also found to be significantly related to chronic
burden among the sample, t (154) =-3 .27, g_ = .00 and t (154) =-2.15, g_ = .03
respectively (See
Table 7 and Table 8). Fisher’s R-Z calculations (Cohen, 1988) were
conducted to assess ethnic differences in the correlation between sexual trauma
and the SES variables. No ethnic differences were found. Among the sample of
women, CSA and ASA were related to SES in a similar fashion.
Table 7 Relationship between CSA and Demographics in the Sample
AA (N =45)
T =
CA (N = 58)
T =
LA (N = 52)
T =
Total (N = 158)
T =
Age .07 -.23 .16 . - . 0 2
Income -.24 - . 1 2 .0 1 -.11
Education -.26 -.07 - . 0 2 -.08
Chronic .19 .33* .19 .26*
Burden
Relationship
Status
-.08 .26 . 1 0 .07
* = £ < . 0 5 , * * £ = <.005
Table 8 Relationship between ASA and Demographics in the Sample
AA (N =45)
T =
Caucasian
(N = 58) T
Latina
(N = 52) T =
Total (N = 158)
T =
Age .08 .1 1 .03 . 1 0
Income - . 2 0 - . 1 2 -.14 -.03
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
67
Education -.26 ..28* .18 .05
Chronic . 0 0 .35** .03 .17*
Burden
Relationship - . 1 0 .1 1 -.04 - . 0 2
CSA .07 .47 .1 1 .23
* = E < .05, ** g = c.0 0 5
Overall, approximately one fourth of the women reported an incident of
adult sexual abuse and two out of five women reported an incident of CSA.
Ethnic differences were found among those with histories of ASA. Three times
more African Americans and Caucasians reported ASA than Latinas. In
addition, those with experiences o f CSA or ASA were likely to have higher
chronic burden. Among Latinas, citizenship was not found to be related to CSA
or ASA.
Intimate Partner Violence
Hypothesis 1. Intimate Partner Violence and Ethnicity
Hypothesis 1 proposed ethnic differences in IPV among the sample.
One-way analyses o f variance were conducted to evaluate this. Mean ethnic
differences for both Respondent IPV (RIPV) and Partner IPV (PIPV) scores are
presented in Table 9. Hypothesis 1 was not supported. No ethnic differences
were found in experiences of intimate partner violence [F (2, 157) = 1.05, p =
.35], Although RIPV rates are reported here, this project’s focus is on the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
68
relationship between violence against women and subsequent health status,
therefore, only partner IPV is examined in follow-up analyses.
Table 9 Mean Ethnic Differences in IPV
AA CA LA Total F
£
RIPV M = 12.02 M = 11.24 M = 12.67 M = 11.94 2.36 . 1 0
SD = 3.45 SD = 2.89 SD = 3.83 SD = 3.41
PIPY M = 12.68 M = 11.57 M = 12.63 M = 12.25 1.05 .35
SD = 5.13 SD = 3.44 SD = 4.81 SD = 4.46
* = j> < .05, **£ = <.005
Hypothesis 2. IPV Differences Among Demographic Variables
Hypothesis 2 proposed a relationship between IPV and background
characteristics: age, income, education and chronic burden. A series of one-way
ANOVAs and Pearson’s R Correlations were conducted to test these
relationships. Significant correlations are presented in Table 10. Hypothesis 2
was partially supported. Chronic burden was found to be related to IPV (.313,
p <.01) among the sample, specifically African Americans (.469, p <. 01) and
Caucasians (.380, p <.01).
In order to explore the differential sociodemographic risk factors for IPV
within ethnic groups (research questions 2), Fisher’s R-Z calculations (Cohen
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
69
1988) were conducted based on the correlation coefficients presented in Table
10. Although IPV was found to be significantly related to chronic burden
among African Americans and Caucasians, this was not found among Latinas.
Additionally, income was related to IPV among African Americans but not
Caucasians or Latinas; however, this correlation difference was not found to be
significant. No other ethnic differences among IPV X demographic variable
correlations were found. Overall, the sociodemographic risk factor for IPV
appears to be chronic burden (or SES stress) and this appears to be stable across
ethnic groups.
Table 10 Relationship between IPV and Demographic Variables
IP V X AA Caucasian Latina Total
Age -.02 .15 .18 .08
Income -.33* .03 -.11 -.12
Education .01 -.14 -.20 -.16
Chronic Burden .47** (b) .38** .11(b) .31**
Relationship Status .21 -.04 -.08 -.00
CSA -.03 .13 -.14 -.01
ASA .14 .43* .43* .16
citizenship (Latinas Only) -.10
* = g < .05, ** p = <.005
Additional bivariate correlations among individual chronic burden scale
' items and partner IPV were conducted among the ethnic groups to gain better
perspective on the type o f chronic burden linked to IPV (see Table 11). IPV was
related to chronic burden items: lack of savings, transportation problems,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
70
housing problems, employment problems, long-term unresolved conflicts, crime
problems and being discriminated against (significant for African Americans).
Caucasians appeared to have the most chronic burden items correlated with IPV,
while Latinas had only one significant chronic burden item, correlated with IPV,
housing problems. When Fisher’s R-Z calculations were conducted, only the
correlation between IPV and being fired/laid off was significant. This appeared
to be a significantly related to IPV among African Americans, however, not
among Caucasians or Latinas.
Overall, among African Americans, Caucasians and Latinas, different
chronic burden items were found to be related to IPV. Employment and
discrimination problems were found to be significant risk factors for IPV among
African Americans. Among Caucasians IPV was found to be particularly related
to financial and transportation problems. Among Latinas, IPV was related to
housing problems.
Table 11 Correlation Coefficients b/w IPV & Chronic Burden Scale Items
IPV X AA Caucasian Latina Total
Not enough $ for basic needs .11 .31* -.08 .12
No savings to meet problems . 1 2 .18* -.07 .18*
No source o f Transportation . 1 2 .31* .08 .16*
Housing Problems .13 .09 .29* .18*
Problems arranging Child Care .04 .09 . 0 0 .05
Being a caregiver for someone .06 . 0 0 .06 .05
Divorce/separation from partner .17 - .0 1 •17 .15
Long-term, unresolved conflict .30* .40** . 1 0 .25*
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
71
Being fired/laid off .43* (b) .23 .06 (b) .23*
Trouble with employer .44* .31* -.07 .25*
W ork Hours change for worse . 2 2 .14 -.03 .1 1
Partner Work Hrs Change for worse .05 . 2 0 -.05 .06
Serious Accident/Injury/Illness .04 .0 1 .26 .08
Victim Crime .31* .04 - .0 1 .16*
Chronic pain .13 .07 -.07 .03
Long-Term Medical problems .04 .2 1 . 1 2 . 1 0
Immigration/Citizens Problems .16 - .1 1 - . 0 2 ..0 1
Arrest/Sent to Jail .0 1 -.03 -.11 - . 0 2
Living in High Crime Area .17 .15 -.03 . 1 0
Losing help you depend on .04 .04 .26 . 1 0
Being Discriminated Against .44** .25 .15 ..30**
* = p < .05, ** g = <.005
Sexual Trauma and IPV.
One-way ANOYAs were conducted to assess the relationship between
sexual trauma and IPV. ASA was found to be correlated with IPV for
Caucasians and Latinas, [F (150) = .4.078, p = .05] while CSA was not [F (150)
= .029, j> = .87], Fisher’s R-Z calculations found no ethnic differences for this
relationship.
In Sum, hypothesis 1 was not supported, and partial support for
hypothesis 2 was found. While no ethnic differences in IPV were detected, IPV
was found to be differentially related to other sociodemographic variables,
specifically, chronic burden. This appears to be a stable risk factor across ethnic
groups. Those with higher IPV reported higher levels o f chronic burden,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
72
specifically, lack of savings, transportation problems, housing problems,
employment problems, long-term unresolved conflicts, crime problems and
being discriminated against (significant for African Americans).. Additionally,
among African Americans, IPV was found to be negatively related to. income.
Those with lower income had higher IPV. Finally, those with histories of ASA
were more likely to report higher IPV.
Hypothesis 3. IPV Influences on Health
Hypothesis 3 proposed that IPV would negatively impact health.
Pearson’s product correlations were conducted to explore the relationship
between IPV and health among the ethnic groups. Table 12 reports Pearson’s
Product-moment correlations. Partner IPV was found to be significantly related
to all mental and physical health outcomes except depressive symptoms at T l,
and health diseases. Among the sample, those with more IPV reported more
depressive symptoms, more anxiety, more negative health symptoms and less
health functioning (HQL). Various IPV X health relationships existed among the
different ethnic groups as well (see Chart 1 and Table 12).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
73
Among African Americans, IPV was related to depression and anxiety.
Among Caucasians, IPV was related to depression, anxiety and health quality of
life. Among Latinas, IPV was related to depression and anxiety; however,
interestingly only at T2. These ethnic differences in the correlations among IPV
and the health variables were assessed for significance using Fisher’s R-Z
calculation (Cohen, 1988). The relationship between IPV and anxiety was
significantly different among African Americans and Latinas. Latinas
experienced significantly more anxiety with IPV than African Americans. IPV
and HQL were also stronger among Caucasians than Latinas, suggesting that
IPV affects health differently based on ethnicity. Overall, IPV negatively
impacts health.
Chart 1. Map of Relationship between IPV and Health among AA, CA and LA groups
X IP V Depressive
Sx
Anxiety Sx HQL Health Dx Health Sx
AA X X X
CA X X X
LA X X
Table 12 Correlation Coefficients between IPV and Health Outcomes
Outcomes P IP V AA PIPV CA PIPV LA PIPV
Depressive Sx T1 .13 . 1 2 .38** .04
Anxiety Sx T1 .31** .51** (b) .42** .13 (b)
HQLT1 -.2 1 * - .0 1 (c) -,40**(c) -.25
Health Dx T1 .1 1 - .1 1 .25 .18
Health Sx T1 .18* .28* .25 .17
Depressive Sx T2 .43** .52**
3g**
.40*
Anxiety Sx T2 .30** .17 .38* .41*
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
74
S . 1Q L T 2 -.25** -.15 -.38* -.23
Health 1 )\ 12 .0 1 -.03 -.06 .1 1
Health Sx T2 . 1 2 . 2 0 .05 . 0 0
* = 2 < .05, ** g = <.005
. ALTH STATUS
In order to assess general ethnic differences in health, without taking into
account experiences of IPV, one-way ANOVAs were conducted among the
health variables across the three ethnic groups.
Table 13 reports significant ethnic differences in depressive symptoms,
health diseases and health symptoms.
Table 13 Ethnic Differences in M ental and Physical Health Outcomes
AA M CA M = Latina M Total M = F
E =
Depressive Sx T1 11.096 11.93c 20.316c 14.48 5.96** . 0 0
Anxiety T1 4.96 5.58 9.45 6.67 2.94 .06
HQL T1 1 0 1 . 6 99.68 98.51 99.86 .64 .53
Health Dx T1 2.18 ab 1.38a 1.216 1.55 7.45 . 0 0
Health Sx T1 2.52 3.04 2.40 2.67 2 . 0 0 .14
Depressive Sx T2 10.44 8.87c 15.37c 1 1 . 2 2 3.51* .03
Anxiety Sx T2 3.4 3.55 6.51 4.32 2.83 .06
H QLT2 104.0 105.11 103.51 104.27 . 2 0 .82
Health Dx T2 2.006 1.59 1.086 1.59 6.83** . 0 0
Health Sx T2 2.766 2.33 2.296 2.29 3.72* .03
* = 2 < -05, ** 2 = <-005
a = Significant correlation between African American and Caucasian
b = Significant correlation Between African American and Latina
c = Significant correlation between Latina and Caucasian
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
75
There were significant ethnic differences in mental and physical health.
Specifically, Latinas had significantly more depressive symptoms. Disturbingly,
they reported clinical depression (M < 16.00) on average at T l. A trend toward
significant ethnic differences in anxiety symptoms was found, with Latinas
reporting substantially more anxiety symptoms than African Americans and
Caucasians. African Americans experienced a greater number o f health diseases
and negative health symptoms than Caucasians, but not Latinas. As suggested in
previous literature, health disparities appear to exist among women of color.
Health Changes over time.
To answer the question o f whether or not health significantly changed
over time among the sample, mixed ANOVAs were conducted (See Table
14). Anxiety symptoms and the number o f negative health symptoms
significantly declined over time and health quality of life improved over the 12-
month period (Anxiety A = .95, F (2, 156) = 6.75, p = .01, health symptoms A
= .95, F (2,156) = 7.33 p = .01; and HQL A = .13, F (2, 156) = 18.03, p = .00).
However, one interaction was found (see Table 12). Ethnicity seemed to
determine whether or not health diseases changed over time A = .94, F (2, 156)
= 3.99, p = .02. African Americans appeared to increase in number of negative
health symptoms while the other groups significantly decreased in negative
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
76
health symptoms between Time 1 and Time 2 (See Figure 1). In general, health
improved from T1 and T2 among the sample, except for African Americans,
whose negative health symptoms increased.
Table 14 Mixed ANOVAs of health outcomes
Source Wilke’s Lambda F
Depressive Symptoms .98 3.00
Ethnicity X Depressive .99 .36
Symptoms
Anxiety Symptoms .95 6.75*
Ethnicity X Anxiety 1.0 .176
Symptoms
Health Quality o f Life .13 18.03*
Ethnicity X Health Quality of .98 1.59
Life Sx
Health Diseases 1.0 .52
Ethnicity X Health Diseases .98 1.38
Health Symptoms .95 7.33*
Ethnicity X Health Symptoms .94 3.99*
* = p < .05, **g = <.005
Figure 1 Mixed ANOVA for Negative Health Symptoms
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
77
Estimated Marginal Means of MEASURE_1
2.6
2,4
2.2
2.0
Ethnicity
1.8
Q .
A frican A m erican
>v
1.6
C au casian
1.4 Latina
1 2
HEALTHS
Background Influences on Health
The relationship between health and other sociodemographic variables
was explored as well. Health was strongly related to a number o f background
experiences including age, income, education, chronic burden and adult sexual
abuse (See Table 15).
Table 15 Correlation Coefficients for Demographics and Health Outcomes (Total Sample)
Age Income Education Chronic
Burden
Relationship
Status
CSA ASA
Depressive Sx
T1
-.05 . 27** -.26**
23**
-.04 . 1 2 .17*
Anxiety Sx T1 .03 -.25* -.2 0 *
3 7 **
- .0 1
2 3 **
.09
HQLT1 -.13 .14 . 1 2 -.39** -.09 - .1 1 - . 1 2
Health D x T l .2 2 ** .05 .06 .08 -.13 .09 . 1 2
Health Sx T1 .05 -.07 .15 .2 0 * . 0 0 .08 .17*
Depress Sx T2 .14 -.30** -.33** .42** .04 .19* .06
Anxiety Sx T2 .26** -.2 2 * -.38** .32** .13 . 8 - .0 1 ■
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
78
HQL T2
. 27**
.18* .19* -.40** -.09 - .1 1 -.09
Health Dx T2
28**
.19 .18 .1 1 -.03 .16 .19*
Health Sx T2
24**
-.04 .04 .27** - .1 1 .06 .19*
*=£<.05, **£=<.005
Overall, mental health (depressive and anxiety symptoms) was related to
income, education and chronic burden. Mental health was also found to be
related to CSA and ASA. HQL was related to these variables as well. In terms
of physical health, number of health diseases was related to age and ASA (T2),
while number of negative physical symptoms was related to age, adult sexual
abuse and chronic burden as well.
These relationships were also explored specifically within each ethnicity.
Similar SES X health relationships existed among African Americans (see
Table 16) mental health among African Americans was related to income,
education, chronic burden and CSA. Physical health was related to age, chronic
burden, and CSA.
Table 16 Correlation Coefficients for Demographics and Health Outcomes for AA
Age Income Education Chronic
Burden
Relationship
Status
CSA ASA
Depressive Sx
T1
-.09 -.46** -.32* .26 -.15 .33* .14
Anxiety Sx
T1
-.13 -.41** -.15 .52** -.09 .25 .25
HQLT1 -.13 .089 . 0 0 -.23 - .1 1 -.27 -.04
Health Dx T1
.40**
. 0 2 -.05 . 1 0 . - . 0 1 .47** .07
Health Sx T1 . 0 0 - . 0 0 .04 .32* .04 .16 .05
Depress Sx
T2
.0 1 -.45** -.25 .46* . 0 0 .36* .25
Anxiety Sx .28 -.28 -.15 .32* .04 .25 .18
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
79
T2
H Q LT2 -.25 .17 .17 -.25 -.04 -.15 -.07
Health Dx T2
- .1 1 .03 . 1 0 -.04 .23 . 1 0
Health Sx T2
.41**
-.38 .0 1 .03 .15 -.06 -.04 .09
= j > < .05, ** g = <.005
Fewer health X SES relationships existed among Caucasians.
Among Caucasians, chronic burden and adult sexual abuse seem to be most
related to the health variables (see
Table 17).
Table 17 Correlation Coefficients for Demographics and Health Outcomes for CA
Age Income Education Chronic
Burden
Relationship
Status
CSA ASA
Depressive Sx
T1
- . 1 2 -.23 -.29
4 9 **
. 0 2 .25 .38**
Anxiety Sx
T1
-.03 - . 2 2 -.11 .35** - . 1 0 .37** .42**
HQLT1 -.04 .18 .32* .58** -.06 -.24 -.40
Health D x T l .27* . 1 0 .05 - . 0 0 - .0 1 -.25 .25
Health Sx T1 .16 - . 2 2 - . 2 0 .30* -.03 .14 .25
Depress Sx
T2
.17 - .2 1 -.03 .51** - . 0 0 . 2 0 .38**
Anxiety Sx
T2
.17 -.17 -.06 .47** -.04 .17 .38*
HQL T2 - .2 1 .17 .08 -.55** . 0 0 -.26 -.38*
Health Dx T2 .23 .15 . 1 2 - .0 1 .19 -.06 -.06
Health Sx 12 .17 -.16 - .2 1 .35* - .1 1 -.05 .05
* = E<.05,**e = <.005
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
8 0
Interestingly, among Latinas, mental health and health quality of life
were significantly related to relationship status (see Table 18). This correlation
•M ss not found among the other ethnic groups. Being a U.S. citizen was related
lc lower depression and anxiety.
Table 18 Correlation Coefficients for Demographics and Health Outcomes for LA
Age Income Education Chronic
Burden
Relationship
Status
CSA ASA Immigr
ation
Status
Depressive
<-'v T1
.07 -.07 -.06 .16 .09 .0 2 .33* -.31*
;O A 1 1
Anxiety
S x T l
. 2 2 -.17 -.13 .38** . 1 0 .13 .13 -.16
HQL Ti -.23 .15 .06 -.35* -.08 -.03 -.25 .03
Health Dx
Tl
.15 - . 1 0 - . 0 0 . 0 0 -.14 .13 .18 .03
Health Sx
Tl
- . 1 0 -.09 .28* .05 -.17 - .0 1 .27 .26
Depress
SxT 2
.37* -.23
_ 4 4 **
.41* .63** .09 . 0 0 -.23
Anxiety
Sx T2
.41* -.26 -.54** .39* .47** -.03 -.15 -.36*
H QLT2 -.39* .27 .31 -.40* -.48** .1 1 .05 .29
Health Dx
T2
. 1 0 .33* .18 -.06 .07 .25 .16 .2 1
Health Sx
T2
.17 -.23 .06 .26 .19 .43** .35* . 0 2
* = £ < .05, **p = <.005
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
81
In sum, African Americans and Latinas experienced decreased mental
health and more health diseases and negative health symptoms than Caucasians,
lending more support to the literature on ethnic minority health disparities
among women. In addition, SES factors: income, education and chronic burden
and sexual trauma (ASA) were also found to influence health (particularly
depression, anxiety, and health symptoms) among the three ethnic groups.
Among Latinas, being a U.S. citizen was related to decreased depression and
anxiety. Once differences in health were established, confirming SES factors
contribute to health differences, the relative contribution of IPV to health
disparities was investigated.
Regression Outcomes
Hypothesis 4. The Contribution o f Ethnicity and IPV to Health Status
Finally, in order to assess whether or not IPV impacted health at T2
(after controlling for health at T l), and the relative contribution o f ethnicity and
SES variables to mental and physical health, a series of multiple regressions
were conducted. Analyses were conducted in four steps.
Step 1 included the respective health variable at time 1 (i.e., depressive
symptoms time 1) and the background variables, chronic burden and ASA.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
8 2
Chronic burden was controlled for because it was found to be
significantly related to IPV across the ethnicities and was found to be
significantly related to many of the health outcomes.
ASA was selected because it was found to be significantly related to IPV
among the sample and found to be significantly related to the outcome variables
as well.
Step 2 included whether being African American or not, Caucasian or
not, or Latina or not, contributed to health status. Regressions were run
separately by ethnicity.
Step 3 included IPV.
Step 4 included an ethnicity X IPV interaction term. This term combined
separate ethnicity and IPV terms into one term expressing the magnitude of IPV
as a function of ethnicity. This step is conducted to test whether or not IPV
moderates the impact on health among the three different ethnicities (Aiken &
West, 1991).
Hierarchical Regression Table Entries (See Tables 19.20 and 211
2 2
Y M o d e l i = Adjusted R , Change in R and beta for: Health at Tl, Chronic
Burden, and ASA.
Y M o d e l 2 = Adjusted R2, Change in R2 and beta for: Ethnicity
Y M o d e l 3 = Adjusted R2, Change in R2 and beta for: IPV
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
83
• 9 9
Y M o d ei 4 - Adjusted R , Change in R and beta for: Ethnicity X IP V
Flie final regression equation was Y = bfhealth at Tl + Chronic Burden and
ASA) + b (ethnicity) + b(IPV) + b(ethnicity x IPV).
African Americans
Table 19 and Table 20 reports significant predictors for the five health
outcomes among African Americans. For all health outcomes, health at Tl
significantly predicted health at T2. Chronic burden was a significant
contributor, as well, for depression, HQL, health diseases and health symptoms.
ASA was not found to be significant in predicting any health outcomes. One
ethnicity main effect was found for negative health symptoms (see Figure 2).
Being African American predicted more negative health symptoms at T2.
After controlling for health at T l, chronic burden, ASA and ethnicity,
IPV main effects were found for depressive symptoms (See Figure 2) and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
84
anxiety symptoms at T2. Specifically, among African Americans, the more IPV,
the more depression and anxiety experienced at T2. Also, a significant
interaction was found for the combined effect of I IPV and being African
American (See Figure 3). Data suggest that African American women appear to
experience the most anxiety with moderate levels of IPV. However, as IPV
increases, anxiety tends to decrease. Hypotheses about this finding are suggested
in the Discussion section. Health quality of life and health diseases were
predicted neither by ethnicity nor IPV.
Table 19 Regression Results for Mental Health Outcomes among African Americans
Depression T2 A nxiety T2
A djusted
R2
A R 2 B A djusted R2 A R 2 B
M odel 1
,39**
.32** .28** .29**
Health Tl .48* .44*
Chronic Burden .23* .15
ASA .1 1 -.09
M odel 2 .39 . 0 0 .29 . 0 0
Ethnicity
Model 3 .48* .09*
-.03
.29 . 0 0
.07
IPV .24* -.47
Model 4 .47 . 0 0 .33 .05
Ethnicity X IPV -.08 -.65
* = j) < .05, ** j> = <.005 T l = R espective H ealth at T l , C B = C hronic B u rd en , A S A = A d u lt
Sexual A buse. Y = b(heaith at T l + Chronic Burden and ASA) + b (ethnicity) + b(IPV) + b(ethnicity x
IPV).
Table 20 Regression Results for Physical Health among African Americans
H e a lth Q u a lity o f L ife H e a lth D isea ses N e g a tiv e H ea lth
S y m p to m s
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
85
Adjust A R 2 B
ed R 2
Adj. R2 A R 2 B Adj.R2 A R 2 B
Model 1 .38** .39** .29** .31*
*
.16 .18**
Health Tl .51* .50* .32*
Chronic -.17 .04 .18*
Burden
ASA .03 .1 2 .09
Model 2 .38 .00 .30 .01 .18* .03*
Ethnicity - . 0 2 - . 1 2 -
.17*
Model 3 .38 .01 .30 .00 .18 . 0 0
IPV -.06 .08 . 2 0
Model 4 .37 .00 .16 .29 .00 .18 .0 1
Ethnicity X .16 .25
IPV
= 2 < -05, ** g = <.005 Tl = Respective Health at T l, CB = Chronic Burden, ASA = Adult
Sexual Abuse. Y = b(health at T l + Chronic Burden and ASA) + b (ethnicity) + b(IPV) + b(ethnicity x
IPV).
Caucasians
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
86
Table 21 and Table 22 report predictors of health outcomes among
Caucasians. For all health outcomes, health at Tl significantly predicted health
at T2. Chronic burden was also found to significantly add to the depressive
symptoms, HQL, and health symptoms models. No ethnicity main effects or
interactions were found.
After controlling for health at T l, chronic burden, ASA and ethnicity,
IPV main effects were found for depressive symptoms at T2 (see Figure 2) and
health diseases at T2 (see
Table 21) among this group. The more IPV a Caucasian woman
experienced, the higher the depressive symptoms and the greater the number of
reported health diseases at T2.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
87
Table 21 Regression Results for Mental Health Outcomes among Caucasians
D ep ressio n T2 A n x ie ty T2
Adjusted R2 A R2 B Adjusted R2 A R 1 B
Model 1 .28** .29**
Health Tl .49* .44*
Chronic .23* . 1 2
Burden
ASA - . 1 0 -.09
Model 2 .39 .00 .27 .00
Ethnicity .03 - . 0 2
M odel 3 .48** .09** .28 . 0 2
IPV .2 2 ** .14
Model 4 .47 .00 .28 . 0 0
Ethnicity X -.09 . 0 0
IPV
* = g < .05, ** g = <.005 T l = R esp ectiv e H ealth at T l , C B = C hronic B u rd en , A S A =
A d u lt Sexual A buse. Y = b(health at T l + Chronic Burden and ASA) + b (ethnicity) + b(IPV)
+ b(ethnicity x IPV).
Table 22 Regression Results for Physical Health Outcomes among Caucasians
H e a lth Q u a lity o f
m _______________
H e a lth D isea ses N e g a tiv e H e a lth
S y m p to m s
Adjusted
R 2
A R 2 B Adj R2 A R 2 B Adjusted A R 2 B
R 2
Model 1
Health
Tl
.38** .39*
.51
*
.30** .31**
.54*
.16** .18**
.32*
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
88
Chronic -
.07 .2 1 *
Burden .16
*
ASA .03 .16* .13
Model 2 .38 .01 .29 .00 .16 . 0 0
Ethnicity .03 .0 2 -.05
Model 3 .38 .01 .28* .0 1 * .15 .00
(E = .07)
IPV -.27 - -.38
.56*
Model 4 .37 .00 .30 .02 .15 .10
Ethnicity -.19 -.50 -.36
X IPV
= E < -05, ** £ = <.005 T l = R espective H ealth a t T l , C B = C hronic B u rd en , A S A = A dult
Sexual A buse. Y = b(health at T l + Chronic Burden and ASA) + b (ethnicity) + b(lPV) + b(ethnicity
xIPV ).
Latinas
Table 23 and Table 24present predictors for the five health outcomes
among Latinas. For all health outcomes, health at Tl significantly predicted
health at T2. Chronic Burden was found to be a significant contributor to anxiety
symptoms, HQL, and health symptoms. ASA was not found to contribute to the
models. Ethnicity main effects were found for health diseases and health
symptoms (see
Figure 4 and Figure 5). Being Latina predicted more health diseases and health
symptoms at T2 than other groups.
After controlling for health at T l, chronic burden, ASA and ethnicity, an
IPV main effect was found for depressive symptoms at T2 (See
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
89
Table 24). IPV predicted depressive symptoms at T2 among Latinas.
Additionally, the combined effect of ethnicity and IPV was found to
significantly contribute to anxiety. Latinas experiencing high IPV appear to
have more anxiety symptoms than the other groups at low or high IPV (See
Figure 3).Error! Reference source not found.).
Table 23 Regression Results for Mental Health Outcomes among Latinas
D ep ressio n T2 A n x ie ty T2
Adjusted R2 A R 2 B Adjusted
R 2
A R 2 B
Model 1 .39** .41 ** .28**
2 9 **
Health Tl .46** .43*
Chronic Burden .24** .17*
ASA -.08 -.03
Model 2 .40 .0 1 .29 .0 1
Ethnicity .09 - . 1 2
Model 3 .48** .08** .29* .0 2 *
IPV .41* .69**
Model 4 .48 . 0 0 .33* .04*
Ethnicity X IPV . 1 2 .62**
* = E < 05, ** 2 = < 005 T l = Respective Health at T l, CB = Chronic Burden, ASA = Adult Sexual Abuse
Y = b(health at T l + Chronic Burden and ASA) + b (ethnicity) + b(IPV) + b(ethnicity x IPV).
Table 24 Regression Results for Physical Health Outcomes among Latinas
H ealth Quality o f L ife H ealth Diseases Negative Health
Sym ptom s
Adj.
R2
A R 2 B Adj R2 A R 2 B Adj. R2 A R2 B
Model 1 .38**
3 9 **
.29** .31** .16* .18*
Health Tl
Chronic
Burden
.52*
.17*
.50*
.05
.29*
.2 0 *
ASA
Model 2 .37 . 0 0
. 0 2
.32** .004*
*
. 1 0
.18* .0 2 *
.06
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
90
Ethnicity - .0 2 - .16*
..19*
(E
.07)
Model 3 .38 .01 .32 .00 .17 .00
i'PV -.08 .08 - .0 1
Model 4 .37 .00 .31 .00 .16 . 0 0
n c ty X .0 1 .15 - .0 1
05, ** j> = <.005 T1 = R espective H ealth at T l, C B = C hronic B urden, A SA = A dult
Sexual A buse. Y = b(health at T l + Chronic Burden and ASA) + b (ethnicity) + b(lPV ) + b(ethnicity x
IPV).
F ig u re 2 IP V M a in E ffects fo r D e p re ssiv e S y m p to m s a t T 2
30
tn
£
o
< D
>
'tn
V)
£
CL
0)
O
c
<0
0 )
Ethnicity
A fric a n A m e ric a n
C a u c a sia n
Latina
IFV H ig h IFV A v e ra g e
IPV
Figure 3 Ethnicity Interaction for Anxiety Symptoms at T2
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
91
1 4
12
1 0 -
C M
h-
c o
e
o
Q -
E
> .
CO
> *
E thnicity
0 ) A frican A m erican
x
c
<
C a u c a sia n
c
C O
( D
Latina
Low IFV High IP'/ A v e ra g e IFV
IPV
Figure 4 Main Effects for Health Diseases T2
2.4
22
2.0
C M
I-
tn
C O
C D
1.4
Ethnicity M
b
. c
A frican A m erican
to
0 3
X 1.0
C
C O
2 8
C aucasian
Low IFV A v erag e IFV High IFV
IPV
F ig u re 5 Main Effects for Health Symptoms T2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
3.5
O t
H . 3-0
£
> .
sz
To 2.5
a >
X
0)
>
03
05 2.0
0 5
X
c
C D
C D
S 1-5 .
Ethnicity
A frican A m erican
C au casian
Latina
Low IPV A v e ra g e FV High IFV
iPV
Figure 6 Significant Main Effects for Health Quality Of Life
o
E thn icity
to
a
0)
X
c:
to
< D
Low IPV A verage IFV
IPV
In general, while chronic burden significantly added to the regression
models for the health outcomes variables, ASA did not.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
Increased IPV significantly predicted more depressive symptoms at T2
among all three ethnic groups and more anxiety symptoms among African
Americans at T2. The combined effect o f ethnicity X IPV produced an
interaction result for African Americans and Latinas. Latinas experiencing high
IPV appear to have more anxiety symptoms than the other groups at low or high
IPV. African Americans appear to experience the most anxiety with moderate
levels of IPV. However, as IPV increases, anxiety tends to decrease.
Interestingly, IPV predicted more health diseases among Caucasians at T2. This
was not for African Americans or Caucasians.
Additionally, ethnicity also predicted an increase in health diseases and
negative health symptoms. Being Latina increased the risk of more health
diseases at T2 than at Tl and being African American or Latina increased the
risk of more negative health symptoms at T2 than at T l.
CONCLUSIONS
Ethnic differences were found among demographic variables including
education, income and relationship status, with Latinas reporting the lowest
monthly income and least amount o f education and African American women
least likely to be married. Among Latinas, citizenship was related to age and
education.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
94
Ethnic differences were also found with respect to the experiences of
ASA. African Americans and Caucasians three times more likely to have
reported an incident of ASA. Those with CSA and ASA were also likely to
experience more chronic burden and those with histories of ASA also
experienced more IPV. Finally, ASA was found to increase a number of
negative health consequences including depression, anxiety, health diseases, and
health symptoms. CSA was also related to anxiety. Among Latinas, CSA and
ASA were not related to immigration status/citizenship.
Overall, no ethnic differences were found in IPV. However, IPV was
related to chronic burden among African Americans and Caucasians,
specifically, IPV was related to lack of money, transportation, housing
problems, employment problems, long-term unresolved conflicts, crime
problems and being discriminated against (significant for African Americans).
When follow-up analyses were conducted, being fired/laid off seemed to be
significantly related to IPV among African Americans. IPV was also found to be
related to adult sexual abuse among Caucasians and Latinas and income among
African Americans.
IPV was found to be related to health. Those with more IPV reported
more depressive symptoms, more anxiety symptoms, more negative health
symptoms and less health functioning/quality o f life.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
95
Mixed ANOVAs assessed the change in overall health status within each
of the three ethnic groups. The sample became less anxious over time, health
quality of life increased and an ethnicity X time interaction was found for
negative health symptoms. African Americans appeared to increase in number
of negative health symptoms over time, while the other groups significantly
decreased in negative health symptoms between Time 1 and Time 2.
Health disparities among women of color were found. When mental
health status was assessed between the groups, Latinas had more depressive
symptoms and on average, Latinas were experiencing clinical levels of
depression. Interestingly, mental health among Latinas was related to citizenship
status. When physical health was assessed, African Americans were found to
have significantly more health symptoms and health diseases than the other
groups. Chronic burden and ASA also influenced health. However, when ASA
was entered into regression models to assess the contribution of ASA to health
changes over time, it was not found to be significant.
Finally, while health generally improved over time among the groups,
after controlling for SES, ASA and ethnicity, IPV predicted a change in health
over time. Increased IPV significantly predicted more depressive symptoms at
T2 among all three ethnic groups and more anxiety symptoms among African
Americans at T2. The combined effect o f ethnicity X IPV produced an
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
96
interaction result for African Americans and Latinas. Latinas experiencing high
IPV, appeared to have more anxiety symptoms than the other groups at low or
high IPV. African Americans appear to experience the most anxiety with
moderate levels of IPV. However, as IPV increased, anxiety tended to decrease.
Finally, IPV seemed to increase health diseases at T2 among Caucasians. This
was not found among African Americans or Latinas.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
97
CHAPTER 5. DISCUSSION
The current study investigated sociodemo graphic risk factors for IPV
across ethnic groups, how intimate partner violence impacts health over time,
and the contribution of IPV to health disparities among women of color. Two
major questions were probed: 1) What are the differential sociodemographic risk
factors for IPV across ethnicity?; and 2) To what extent does IPV impact the
mental and physical health of Caucasian, African American and Latina Women
above and beyond what can be explained by other life-stressors such as poverty
or histories of sexual trauma?
Four hypotheses were investigated: 1) African Americans and Latinas
would experience higher IPV than Caucasian women; 2) Lower income, lower
education, higher chronic burden and a history of sexual trauma would be
related to higher IPV among all ethnic groups; 3) IPV would be related to
decreased health status among all women; and 4) Ethnicity and IPV would
interact, with IPV moderating ethnic differences in health over time.
Specifically, African American and Latina women’s health would decline over
time, with ethnicity and IPV significantly contributing to this decline.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
Research Question 1, Hypothesis 1 and 2
While ethnic differences were not found for IPV, socioeconomic factors
appear to be related to IPV. Data further suggest that these sociodemographic
factors may place women at differential risk based on ethnicity.
Hypothesis 1
First, although hypothesis 1 predicted ethic differences in IPV, data did
not substantiate the claim. This supports previous findings in the literature by
the Office on Women’s Health (2003) that violence is not linked specifically to
racial or ethnic factors, but is linked rather to socioeconomic status. This seems
to suggest that a woman’s ethnicity in and of itself is not related to IPV, but is
linked to socioeconomic experiences that are then related to IPV. In other words,
socioeconomic status appears to be a mediating factor between ethnicity and
IPV, as further discussed in hypothesis 2. As suggested by Anderson (1997),
research should look beyond ethnic factors and more toward women’s life
experiences for a more complete understanding of IPV.
Hypothesis 2
Hypothesis 2 predicted IPV would be related to lower income and
education, higher chronic burden and a history of sexual trauma. Data partially
supported claims. IPV was found to be significantly related to chronic burden
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
among African Americans and Caucasians. In addition, income was related to
IPV for African Americans and a history o f adult sexual abuse among Latina
and Caucasians.
Chronic Burden
Chronic burden was found to be significantly related to IPV among
African Americans and Caucasians, but not Latinas. When individual chronic
burden items were analyzed, it was found that lack of savings, transportation
problems, housing problems, employment problems, long-term unresolved
conflicts, crime problems and being discriminated against (significant for
African Americans) were the items significantly related to IPV. Although a
causal relationship cannot be inferred - it cannot be said that having employment
problems or being discriminated against places a woman at more risk for IPV - it
can be said that IPV and these experiences are related. As employment problems
and discrimination increase, so does the experience of IPV. This issue warrants
further attention as there could be several plausible explanations.
It could be that problems at work or discrimination in social settings
exacerbate stress, which then increases conflict in relationships, which then
leads to emotional and physical abuse in the relationship. This partially supports
resource theories that suggest that lower income contributes to stress in a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
relationship, thereby increasing IPV, particularly among communities where
lower SES is more chronic.
It could also be that IPV actually causes SES problems, as suggested by
feminist and power theories. When one is being emotionally or physically
assaulted, problems could arise in other settings such as at work. For instance, a
perpetrating partner could deny transportation or harass the victim at work as
part of the IPV in the relationship. This could make work very difficult,
resulting in numerous employment problems or even being fired/laid-off,
thereby causing an increase in chronic burden.
Chronic burden and IPV could also both be linked to a third variable,
such as income or education, which might make leaving an IPV relationship
very difficult. In this way, lower income affects both chronic burden and IPV.
All of these are plausible pieces of a multidimensional problem. More
research is needed to address the complex relationship between
sociodemographic background, IPV and the impact of IPV on chronic life
burden. Clearly, many theories receive partial support given the complexity of
the IPV phenomenon.
Income
The relationship between income and IPV is well established (Rennison
& Welchans, 2000; Riger & Krieglstein, 2000; Schecter, 1999). In this sample,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
101
actual reported income was found to be related to IPV only among African
Americans. However, interestingly, when individual chronic burden items were
analyzed, it was found that among Caucasians, “not having enough money for
basic needs” and “not having a savings to meet problems” were significantly
related to IPV as well. For Caucasians, although actual income was not as
relevant to IPV, whether or not this income was perceived to meet basic needs
or address emergencies was significantly related to IPV. This suggests an
interesting difference in the meaning of income that warrants much further
investigation. It could be that IPV occurs across income levels, however, IPV
may impact whether or not income can actually meet the needs o f the victim.
Among Caucasians, there may be enough income, but due to financial control by
their partners, this income may not be sufficient. Future studies should focus on
the role of income (not just amount of income, but how it is used) in IPV
relationships.
Age. Education. Relationship Status and U.S. Citizenship
Age, and education were not found to be related to IPV among any o f the
ethnic groups, suggesting that IPV exists among women across the life span and
additionally across education levels. Interestingly, it was also found that those
in a serious relationship (either currently married or living together) were no
more likely to have experiences o f IPV than those who were not in a serious
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
102
relationship. Also, contrary to other findings, IPV was not found to be related to
citizenship status. However, this could be due to the dichotomous nature of the
question. More specific questions, such as length of time in the U.S may better
assess the experience of immigration.
Sexual Trauma
While child sexual abuse did not appear to be related to IPV among the
sample, an experience of adult sexual abuse was related among Caucasians and
Latinas. While this is an interesting finding, confounds could exist. It is not
known whether or not these women were sexually assaulted in the current
relationship or had a history of sexual assault. What is known is that among
Caucasians and Latinas, being sexually assaulted is significantly related to being
in an IPV relationship. This was not found among African Americans. This
could mean that there may be ethnic differences in types of violence that occur
within the context of IPV relationships as suggested by Natera, Tiburcio, and
Villatoro (1997) and Tjaden and Thoennes (1998). It could also mean that being
sexually assaulted may increase risk for IPV in subsequent relationships among
specific ethnicities. These questions warrant further examination.
Research Question 1
In addition to hypothesizing that socioeconomic factors would be related
to IPV, it was also suggested that ethnic differences in the relationship between
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
103
specific socioeconomic factors and IPV exist, although no predictions about
exact relationships were hypothesized. This was confirmed.
While some socioeconomic factors were related to IPV among African
Americans, such as income and chronic burden (specifically employment
problems and discrimination), other factors such as a history of adult sexual
abuse and chronic burden factors, including problems with transportation, and
not enough money for basic needs, were related to IPV among Caucasians.
Latinas reported more housing problems related to IPV.
This could mean that different ethnic communities may be more
vulnerable to specific SES factors. African Americans may experience more
employment issues, while Latinas may have more housing problems, which then
influence IPV relationships.
However, this could also mean that violence in a relationship may be
perpetrated in different ways, affecting different aspects o f life. Caucasians and
Latinas appear to suffer more sexual assaults related to IPV. Caucasians may
also experience more financial control. Among African American women,
perpetrators may impact employment via verbal abuse, harassment at work, or
more physical injuries resulting in more days off from work resulting in being
fired. IPV may impact housing for Latinas.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
104
The interaction between ethnicity, SES and IPV appears very complex
and h as many implications. It is not known whether or not SES increases risk for
IPV among the three ethnic groups or whether or not IPV decreases SES, which
then leads to more IPV. Again, much more work needs to be done to make these
important distinctions in an ethnically specific way.
Hypothesis 3
Hypothesis 3 proposed that IPV impacts health among all women. This
was supported. Those with more IPV reported more depressive symptoms, more
anxiety, more negative health symptoms and less health quality of life. This is
consistent with findings by Sutherland, Bybee and Sullivan (2002), Campbell
(2002), and others. IPV is associated with a host of negative mental and physical
health symptoms including depression, anxiety, the number of health symptoms
and decreased quality of life.
Ethnic differences in health, regardless of IPV, were found as well,
supporting research suggesting health disparities among women of color.
Latinas were likely to have more depressive symptoms, while African
Americans were likely to have more health diseases than Caucasians or Latinas.
These health disparities appear to be related to income, education and chronic
burden among African Americans and interestingly, appear to be linked to
relationship status and U.S. citizenship among Latinas.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
105
Research Question 2 and Hypothesis 4
Research question 2 examined to what extent IPV impacts the mental
and physical health o f Caucasian, African American and Latina women above
and beyond what can be explained by other Iife-stressors, such as poverty or
histories of sexual trauma. Hypothesis 4 proposed that ethnicity and IPV would
interact, with IPV moderating ethnic differences in health over time.
When analyzed over time, most participants’ health did not further
decline over time. Although African Americans increased in number of negative
health symptoms while the other groups significantly decreased in negative
health symptoms between Time 1 and Time 2, changes in other health variables
among the three groups were not found. However, in order to better determine
the impact of IPV on health at T2, after controlling for health at T l, additional
regression analyses were conducted. It was found that although in general health
improved over time, ethnicity and IPV impacted the rate and degree of
improvement.
After controlling for health at T l, SES, ASA and ethnicity, IPV
predicted decreased health at T2. IPV significantly predicted more depressive
symptoms at T2 among all three ethnic groups and more anxiety symptoms
among African Americans at T2. While high IPV predicted increased depression
among the three ethnic groups, Latinas appear to be particularly impacted (See
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
106
Figure 2). At high levels of IPV, Latinas appear to have almost clinically
significant levels of depression M = 15.37 (16.00 = clinically depressed). IPV
also predicted more health diseases among Caucasians.
Interestingly, there was a differential anxiety change among the ethnic
groups based on the amount of IPV. The combined effect of ethnicity X IPV
produced an interaction result for African Americans and Latinas. Latinas
experiencing high IPV appeared to have more anxiety symptoms than the other
groups at low or high IPV. African Americans appeared to experience the most
anxiety with moderate levels of IPV. However, as IPV increased, anxiety tended
to decrease. Being African American also significantly predicted a change in
health diseases and negative health symptoms between Tl and T2 (See Figures 3
and 4).
These findings have substantial implications. IPV seems to impact all
women, yet, not al 1 women are impacted in the same way or to the same degree.
African Americans
At moderate rates o f IPV, African American women appear to
experience the highest rates o f anxiety, health symptoms and health diseases
(See Figures 1-4). While depression continues to increase at higher rates of
abuse, other health conditions seem to improve as the abuse increases. These
are very interesting patterns. It appears that these women could possibly be in
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
107
some ways adapting to abuse. However, these women may also be seeking
services, getting help from friends and/or family, or beginning to leave the
relationships at higher rates of abuse. African American women were also less
likely to be in relationships in the sample. It could be that there were more
women in relationships when there were moderate rates of abuse rather than
high rates of abuse.
Although reasons for these patterns are unclear, it is clear that African
Americans experiencing moderate rates of IPV suffer high rates of depression as
well as other health problems. Given that African American women are already
disproportionately impacted by a myriad of health problems (see Figures 1-4),
IPV can exacerbate these conditions, further declining health. Intervention
appears to be needed, particularly as low amounts of violence begin to increase
to moderate amounts of violence; this appears to be when African Americans are
most vulnerable to a decline in health.
Caucasians
Caucasian women also suffer diminished mental and physical health as a
result o f IPV. This is consistent with findings that women suffer from a variety
of health factors as result of IPV. Caucasians had higher rates of depression at
high IPV and significantly higher rates of health diseases at moderate rates of
IPV. It appears that among this sample of women, Caucasian women were more
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
108
likely to be impacted physically by IPV than Latinas (See Figures 1-4). It
appears that among Caucasians, IPV plays a bigger role in physical health.
Reasons for this are unclear at this time. It could be that these women are more
prone to somaticizing their feelings or they could be suffering from more
physical abuse, resulting in more physical consequences.
It could also be that for African Americans, ethnic background decreases
health in such a way that additional experiences of IPV do not
disproportionately affect health in the same way. There may be a floor effect
among African Americans- these women may already be suffering so much that
IPV does not significantly add to health decline. However, among Caucasians,
who may generally experience better health, IPV may cause a significant
decline.
Interestingly, Caucasians seem to have worse physical health than
Latinas overall. This is an interesting finding supported by existing data from
the Office on Women’s Health (2003). Latinas were reported to have better
health yet perceived more suffering. This data should be further explored. This
could be due to the nature of the self-report measures o f symptoms and diseases
o f this study.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
109
Latinas
Among Latinas, mental health appears to be particularly impacted by
IPV. Latinas have substantially more depressive symptoms (almost clinically
significant) and appear to experience extremely high anxiety at high rates of
abuse. However, Latinas appear to have the lowest number of health symptoms
and diseases among the groups.
Health status among this group should be further investigated. It appears
that on average Latinas have significantly high rates of depression, and that not
only IPV, but simply being involved in a relationship, increases depression and
anxiety symptoms as well as decreases health quality o f life (see
Table 18). Additionally, immigration status was related to mental health
as well. Causal factors o f these correlations are unknown. It could be that
Latinas experience such high rates o f depression and anxiety because o f lack of
resources to get help when in an IPV relationship. In addition, among Latinas,
age and relationship status were highly correlated, it could be that many of these
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
110
women are younger and thus even more vulnerable to socioeconomic stressors
as well as relationship stressors.
Other traumatic experiences may also play a role. Although Latinas
reported the lowest rate of ASA, experiences of ASA predicted both IPV and
negative health .symptoms. Sexual trauma may be disproportionately adding to
depression among this group. This supports other research by Wyatt (1990)
suggesting that sexual trauma has extremely detrimental mental and behavioral
health effects for women.
Socioeconomic status could also play a role. Latinas had the lowest
income and education among the sample. These could significantly impact
access to resources and increase isolation, leading to more depression.
Finally, acculturation and gender roles could be also be factors. This
population of women is known to have high religious values that stress the
importance of relationships, even if personal needs are sacrificed, as well as,
stringent gender roles that condone submissiveness and subservience to
husbands (Amaro, 1988; Carmona, Romero, & Loeb, 1999) This may prevent
women from leaving their husbands or speaking of their abuse to outsiders.
These women could be experiencing painful verbal and physical abuse, yet do
not have persons to turn to for help. As the Latino community expands in this
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
I l l
country, it is critical that community services expand and become better
equipped to meet these mental health needs.
Summary o f Findings
SES and IPV
Overall, the data indicate a relationship between ethnicity, IPV and
health exists and that this relationship is complex at many levels. There are no
clear answers to whether or not ethnic groups are particularly at risk for IPV; it
appears to depend on socioeconomic status and history of sexual trauma. It is
also not known whether or not SES influences IPV or IPV influences SES. It
appears that both of these relationships exist and may work in a cyclical fashion
to maintain high rates of abuse in disenfranchised communities. This supports
Anderson’s (1997) integrated theory emphasizing the intersectionality o f race,
gender, class and resources in understanding the roots and risk factors o f IPV
and the Office on Women’s Health (2003) health findings that race and ethnicity
may not be the only important factors in influence violence in intimate
relationships. All in all, current research must move past only ethnic
comparisons in research on IPV. Persons must be more completely understood,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
112
not just ethnically but by their income, their environment and their history of
rape and other trauma. Each person must be fully evaluated in order to receive
competent care.
Health and IPV
In addition, it was also found that IPV may differentially impact mental
health among women of color. Latinas appear to be particularly vulnerable to
depression as a result of IPV. African Americans appear to experience
significantly increased anxiety at moderate rates of abuse, but this anxiety
appears to lessen as the abuse increases. In addition, Caucasians appear to
experience a significantly higher report o f health diseases at moderate rates of
abuse. Reasons for these health patterns are not yet known. It appears that
cultural and socioeconomic factors, as well as IPV, play a significant role in
mental and physical health status among ethnically diverse women. Finally,
while health in general improved among the sample, results indicate that IPV
impacts the rate and degree of improvement over time. This is significant
information, IPV appears to determine the extent to which women can improve
their health over time. While health in this sample generally improved, IPV
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
113
predicted health improvement among the population. IPV appears to be a
hinderence to women’s improvement in health.
Other Findings
Ethnic differences in health status were also found. Latinas reported
poorest mental health, while African Americans reported poorest physical
health. This supports well documented literature that women o f color experience
disparities in health. It was also found that disparities in mental health
(particularly anxiety) are exacerbated by the experience o f IPV. Research is only
beginning to take note of these health differences and a call to action by the U.S.
Department of Health And Human Services (2003) suggests the need for
rigorous research to better understand these ethnic health vulnerabilities. It
appears that some of the health vulnerabilities may be due to life stressors such
as income, chronic burden, immigration status, the experience of adult sexual
abuse and the experience of IPV. However, much more work is needed to fully
address this issue.
All in all, IPV appears to remain a complex phenomenon with many
multidimensional aspects. This may be why IPV remains such a pervasive and
deadly problem in our country. Hopefully, this study has provided some insight
into the diverse ways in which IPV impacts the lives of women between and
within ethnic group.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
114
Implications
Clinicians working with clients experience IPV in relationship should be
made aware of the possible other life-stressors exacerbating and contributing to
the experience of IPV. When treating clients with current intimate partner
violence, it seems important to assess demographic vulnerabilities and history of
sexual trauma that have not been explored. It is also crucial for those working
with IPV clients to ensure the physical health of these clients. It appears that it is
not enough to assess immediate threat or injury, psychotherapy must move
toward investigating long-term indirect health effects such as disease
vulnerability, and negative physical symptoms that could also put a client in
danger. The link between declined physical, as well as mental health and IPV,
should be made clear to clients. They may be unaware of the “hidden” dangers
of IPV. This appears especially true for ethnic women, and women of lowered
SES, who may be disproportionately vulnerable to a diminished mental and
physical health status.
Limitations
The first limitation of the study has to do with the inability to clearly
describe the sample, racially and ethically. It is not known if all o f the women
labeled as African American were in fact African American. These women
could have also been African, Affo-Caribbean, or o f mixed heritage. Likewise,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
115
the Caucasian and Latina groups could have been heterogeneous, as well. As the
research in this area becomes more culturally competent, more must be done to
better detail the racial and ethnic experiences o f women.
Next, with regard to socioeconomic status, immigration status was not
fully explored. Only U.S. citizenship was assessed. Given the large Latina
population, more was needed on the extent to which immigration status,
including length of time in the U.S., may have added to the health decline
among the Latinas.
With regard to sexual trauma, ASA was assessed separately from IPV. It
is unknown whether or not participants were raped by their intimate partner
and/or whether or not these women were raped in the context of an IPV
relationship. More research should be done to explore the difference between a
history of ASA and it’s contribution to risk for subsequent IPV relationship, and
ASA in the context of an IPV relationship.
Also, IPV was described in a very global way. Distinctions between
verbal, sexual and physical abuse were not made. In addition, respondent versus
partner or couples mutual violence was not fully explored. There could be
differences in SES or a differential impact on health based on type of IPV.
Finally, some hypotheses about changes in physical health may not have
been supported due to the self-report nature o f the health variables. It may also
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
116
be due to the chronic nature of many of the illnesses examined in the study
including cancer and diabetes. This may have been why some physical health
symptoms changed, yet diseases did not. In addition, it may not just be the
r«i mber of illnesses but the experience o f the illness that declines as a result of
IPV. This study assessed whether or not ilhess or diseases increased in number.
; i could be that while the number o f illnesses or symptoms do not necessarily
increase, present conditions could be exacerbated over time as a result o f IPV.
Future Studies
Much more study is needed to address the complex relationship between
sociodemographic background, IPV and the impact of IPV on chronic life
burden. In addition, studies should further explore the relationship between
health and IPV among women o f other ethnicities including Asians, Native
Americans, and Middle Eastern women. The impact of IPV is rarely explored
among these populations.
The findings o f this study suggest the need for more research in terms of
how violence is perpetrated in the context of intimate relationships. While this
study only reported rates of respondent IPV versus partner IPV and found no
ethnic differences, it was found that these two types of violence are significantly
related and may be differentially impacted by background and environment and
may differentially impact health. Sexual trauma history and ASA in the context
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
117
o f IPV should also be explored. As suggested by Johnson and Ferarro (2000),
more work is warranted in this area.
Finally, in addition to the exploration of risk factors, other health issues
should be explored including, investigation of health behaviors and substance
use issues. Flealth behaviors may be moderating the influence of ethnicity or
other background variables on health status. It may not be background life-
stressors, but behaviors such as smoking or sexual risk taking that may actually
more accurately influence health in ethnic and Caucasian women. Finally,
distinctions between the impact of verbal abuse and control versus physical and
sexual assault should be better assessed to best serve those most vulnerable to
diminished health.
Conclusion
The project explored the relationship between ethnicity, SES, IPV, and
changes in health over time. At the end o f this study, information was to be
gained with regard to differential sociodemographic risk factors for IPV across
ethnic groups, how intimate partner violence impacts health over time, and the
contribution of IPV to health disparities among women o f color.
Overall, the current data supports the idea that ethnicity in and o f itself
does not increase risk for IPV, however, there may be ethnic differences in other
sociodemographic variables that increase risk for IPV. Previous studies suggest
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
118
a need to move past ethnic differences in IPV toward other differences in life
experience that may give insight into why and how IPV occurs in relationship;
particularly who is at risk for IPV and the devastating long-term mental and
physical health consequences of IPV. This study support these conclusions, as it
appears that many factors influence IPV and IPV appears to influence many life
factors, including SES and health. While health among the sample generally
improved over time, ethnicity and IPV affected the extent to which this
occurred.
More research and intervention are desperately needed. IPV appears to
have a detrimental health impact, particularly among the Latina community.
Latinas appear to be disproportionately suffering from IPV and SES stress.
Education, prevention and clinical work are warranted to assist this community.
Also, more research and intervention are needed to address the apparent
health disparities among women of color. While IPV appears to contribute to
some of these disparities, it is not the only answer. It appears that other factors
increase health risk as well, such as income, education and socioeconomic
burden. More exploration of these relationships is imperative to improve the
lives of women of color.
In conclusion, IPV has been found to have a number o f devastating
health consequences across ethnic groups that impedes women’s health over
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
119
time. Intervention is urgently needed, not only from a legal standpoint but from
an economic, cultural and health standpoint as well. IPV cannot be considered in
a monolithic way. Many variables intersect, which may account for how and
why IPV is such as devastating phenomenon that is difficult to change.
Overall, IPV should be considered a clear and present danger to
women’s mental and physical health over time; and is related to a consistent and
chronic burden that impacts their mind, body and spirit. This phenomenon, gone
unaddressed, will continue to plague future generations of men, women and
children worldwide.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
120
References
Abbott, P., & Williamson, E. (1999). Women, health and domestic violence.
Journal o f Gender Studies, 8(1), 83-102.
Aciemo, R., Resnick, & Kilpatrick, D.G. (1997). Health impact of interpersonal
violence 1: prevalence rates, case identification, and risk factors for
sexual assault, physical assault and intimate partner violence in men and
women. Behavioral Medicine, 23(2), 53-64.
Agency for Healthcare Research and Quality (2003). Health care fo r minority
women: Program B rief U.S. Department o f Health and Human Services
Public Health Service. Washington, DC. Retrieved from
http://www.ahrq.gov/ research/minority.htm
Aiken, L. & West, S. (1991). Multiple Regression: Testing and Interpreting
Interactions._ Newbury Park: Sage Publications.
Amaro, H. (1988). Considerations for prevention of HIV infection among
Hispanic women. Psychology o f Women Quarterly, 12, 429-443.
Amaro, H. & de la Torre, A. (2002). Public health needs and scientific
opportunities in research on Latinas. American Journal o f Public Health,
92 (4), 525-529.
American Psychological Association (2003). Guidelines on multicultural
education, training, research, practice and organizational change for
psychologists. American Psychologist, (5), 377-402.
Anderson, K. L. (1997). Gender status, and domestic violence: An integration of
feminist and family violence approaches. Journal o f Marriage and
Family, 59(3), 655-669.
Anderson, M.J. (1993) A License to Abuse: the Impact of Conditional Status on
Female Immigrants. Yale Law Journal, 102(6), 1401-1431.
Astone, N. M., Ensiminger, M., & Juon, H. S. (2002). Early adult characteristics
and mortality among inner-city African American women. American
Journal o f Public Health, 92, (4), 640- 645.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
121
Bandura, A. (1977) Social Learning Theory, Prentice-Hall, Englewood Cliffs,
N.J.
Bennice, J. A., Resick, P. A., Mechanic, M., & Astin, M. (2003). The relative
effects of intimate partner physical and sexual violence on
posttraumatic stress disorder symptomatology. Violence & Victims,
18(1), 87-94
Bograd, M. (1999). Strengthening intimate partner violence theories:
Intersections of race, class, sexual orientation and gender. Journal o f
Marital andFamily Therapy, 25(3), 275-289.
Browne A., &Bassuk, E. (1997). Intimate violence in the lives of homeless and
poor housed women: Prevalence and patterns in an ethnically diverse
sample. American Journal o f Orthopsychiatry, 67, 261-278.
Campbell, J. C. (2002). Health consequences of intimate partner violence. The
Lancet, 359, 1331-1336.
Campbell, J. C., Jones, A. S., Dienemann, J. & Kub, J.(2002). Intimate partner
violence and physical health consequences. Archives O f Internal
Medicine, 162(10), 1157-1163.
Campbell, J. C., & Soeken, K. L . (1999a). Women's responses to battering over
time: An analysis of change. Journal O f Interpersonal Violence, 14 (1),
21-40.
Campbell, J. C., & Soeken, K. L. (1999b). Forced sex and intimate partner
violence: Effects on women’s risk and women’s health. Violence Against
Women,5 (9), 1016-1035.
Carmona, J. V., Romero, G. J., & Loeb, T. B. (1999). The impact of HIV status
and acculturation on Latinas’ sexual risk taking. Cultural Diversity &
Ethnic minority Psychology. Special Issue: HLV/A1DS and Ethnic
Minority Women, Families and Communities, 5(3), 209-221.
CDC National Center for Injury Prevention and Control, (2003). Costs of
Intimate Partner Violence Against Women in the United States. Atlanta
(GA): Centers for Disease Control and Prevention. Retrieved from
http://www.cdc.gov/ncipc/pub-res/ipv cost/ipv.htm.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
122
Cohen, J. W. (1988). Statistical Power Analysis fo r the Behavioral Sciences (2nd
Edition). Hillsdale, NJ: Lawrence Erlbaum Associates.
Crenshaw, K. (1991). Mapping the margins: Intersectionality, identity politics,
and violence against women of color. Stanford Law Review, 43a 1241-
1299.
Degratis LR, Rickels K, Rock AF (1976). . The SCL-90 and the MMPI: a step in
the validation of a new self-report scale. Behavioral Journal o f
Psychiatry, 128, 280-289.
DiLillo, D., Giuffre, D., Tremblay, G. C., & Peterson, L. (2001). A closer look
at the nature of intimate partner violence reported by women with a
history of child sexual abuse. Journal o f Interpersonal Violence, 16(2),
116-132.
Erickson, R. A., & Hart. S. J. (1998). Domestic violence: Legal, practice and
educational issues. Medsurg Nursing, 7(3), 142-164.
Fee, E., Brown, T. M., Lazarus, J., & Theerman, P. (2002). Domestic violence-
Medieval and modem American Journal o f Public Health, 92(2), 5.
Findings from the National Violence against Women Survey. (2000, July).
Retrieved from Family Education Network September. 25, 2003 at
Infoplease.com, Http://www.infoplease.com/ipa/A0875303 .html
Fox , G. L., Benson, M. L., DeMaris, A. A, & Van Wyk, J. (2002). Economic
distress and intimate violence: Testing family stress and resources
theories. Journal O f Marriage A nd Family, 64(3), 793-807.
Gee, G. (2002). A multilevel analysis of the relationship between institutional
and individual racial discrimination and health status. American Journal
o f Public Health, 92(4), 615-623.
Gelles, R. J., & Cornell, C. P. (1990). Intimate violence in fam ilies (2nd ed.).
Newbury Park, CA: Sage.
Golding, J. M. (1999). Intimate partner violence as a risk for mental disorders: A
meta-analysis. Journal o f Family Violence, 14(2), 99-132.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
123
Greenwood, G.L., Relf, M.V., Huang, B. Pollack, L.M, Canchola, J. A., &
Caania, J. (2002). Battering victimization among a probability-based
sample of men who have sex with men. American Journal o f Public
Health, 92(12), 1964-170.
Heise L, Pitanguy J, Germain A. (1994). Violence against women: the hidden
health burden. World Bank Discussion Paper 255. Washington, DC: The
World Bank.
Health Behavior News Service (2003, March). Heart disease costs the most:
it as most expensive medical condition. News Release. Health on the Net
Foundation. Retrieved from http://www.hon.ch/News/HSN/512273.html.
Hotaling G. T., & Sugarman, D. B. (1986). An analysis of risk markers in
husband to wife violence: The current state of knowledge. Violence and
Victims, 1, 101-124.
Johnson, M. P. (1999, November). Two types o f violence against women in the
American family: Identifying patriarchal terrorism and common couple
violence. Paper presented at the National Council on Family Relations
Annual Meetings, Irvine, CA.
Johnson, M. P. (2000). Conflict and control: Images of symmetry and
asymmetry in domestic violence. In A. Booth, A. C. Crouther, & M.
Clements (Eds.) Couples in Conflict, Hillsdale, NJ: Erlbaum.
Johnson, M. P., Ferraro, K. J. (2000). Research on intimate partner violence in
the 1990s: Making distinctions. Journal o f Marriage and the Family, 62,
948-963.
Kalmuss, D. S. & Straus, M. A. (1990). Wife’s marital dependency and wife
abuse. In M. A. Straus & R. J. Gelles (Eds.), Physical violence in
American families: risk factors and adaptations to violence in 8, 145
families, (pp. 277-286). New Brunswick, NJ. Transaction.
Karlsen, S., & Nazroo, J. Y. (2002). Relationship between racial discrimination,
social class and health among ethnic minority groups. American Journal
o f Public Health, 92(4), 624-631.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
124
Kemic, J., Wolf, M. E., & Holt, V. L. (2000). Rates and relative risk o f hospital
mission among women in violent intimate partner relationships.
American Journal o f Public Health, 90(9), 1416-1420.
Kritek, P. B., Hargraves, M., Cuellar, E. H., Dallo, P., Gauthier, D. M., Holland,
C. A., Iikiw, C., Swanson, J. W., & Swanson, R. (2002). Eliminating
health disparities among minority women: A report on conference
workshop process and outcomes. American Journal o f Public Health,
92(4), 580-587.
Kumanyika, S. K., Morssink, C. B., Nestle, M. (2001). Minority women and
advocacy for women’s health. American Journal o f Public Health, 91
(9), 1383-1388.
Lane, T. (2003). Women have different risk factors for verbal, physical partner
abuse. Perspectives on Sexual and Reproductive Health, 25(2), 106-107.
Lawrence, S. (2003). Intimate partner violence and men. Nursing Standard,
17(40), 41-43.
Lee, R. K., Thompson, V. L. S., & Mechanic, M. B. (2002). Intimate partner
violence and women of color: A call for innovations. American Journal
o f Public Health, 92(4), 530-534.
Leigh, W. A., & Jimenez, M. A. (2002). Women o f Color Health Data Book.
Office o f the Director: National Institute of Mental Health, Washington
DC. Retrieved from http://www.4woman.gov/owh/pub/woc.
Lenton, R. L. (1995a). Power versus feminist theories o f wife abuse. Canadian
Journal o f Criminology, 37(3), 305-318.
Lenton, R. L. (1995b). Feminist versus interpersonal power theories of wife
abuse revisited. Canadian Journal o f Criminology, 37(4), 567-571.
Lloyd, S., &Taluc, N. (1999). The effects o f male violence on female
employment. Violence Against Women, 5, 370-392.
Lown, E. A., Vega, W. A. (2001). Prevalence and predictors of physical partner
abuse among Mexican American women. American Journal o f Public
Health, 91(3), 441-445.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
125
Meloy, J. R. (1998). The psychology of stalking. In J. R. Meloy (Ed.), The
psychology o f stalking: Clinical and Forensic Perspectives (pp. 1 -23).
New York: Academic Press.
Natera, G. R., Tiburcio, M. S., & Villatoro, J. V. (1997). Marital violence and its
relationship to excessive drinking in Mexico. Contemporary Drug
Problems, 24, 787-804.
National Research Council (1996). Understanding Violence Against Women,
Washington, B.C.: National Academy Press, 9-10 & 74-80. Retrieved
from http://www.cdc.gov/ncipc/factsheets/ipvfacts.htm
Nazroo, J. Y. (2003). The structuring of ethnic inequalities in health: Economic
position, racial discrimination and racism. American Journal o f Public
Health, 93(2), 277-284.
Nedd, D. M. (2001). Self-supported health status and depression of battered
black women,_ABNF Journal, 12(2), 32-35.
Noll, J. G., Horowitz, L. A., Bonanno, G. A., Trickett, P. K., & Putnam, F. W.
(2003). Revictimization and self-harm in females who experienced child
sexual abuse. Results from a prospective study. Journal o f Intimate
Partner Violence, 18(12), 1452-1471.
Office on Women’s Health (2003, July). The health o f minority women. U.S.
Department of Health and Human Services, Washington, DC. Retrieved
from http://www.4woman.gov/owh/pub/minority/index.htm
Radloff, L. S. (1977). The CES-D Scale. A Self-report depression scale for
research in the general population. Applied Psychological Measurement,
f 385-401.
Rennison, C. M., & Welchans, S. (2000). Bureau o f justice statistics special
report.: Intimate Partner violence. US Department of Justice. Office of
Justice Programs. NCJ 17824. Press Release Retrieved from
http://www.ojp.usdoj .gov/bjs/ abstract/ipv.htm.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
126
Resnick, H. S., Aciemo, R., & Kilpatrick, D. G. (1997). Health impact of
interpersonal violence 2: Medical and mental health outcomes.
Behavioral Medicine, 23(2) 65-78.
Riger, S. & Krieglstein, M. (2000). The impact of welfare reform on men's
violence against women. American Journal o f Community Psychology,
28 (5), 631-648.
Ro, M. (2002). Moving forward: Addressing health of Asian American and
Pacific Islander women. American Journal o f Public Health, 92 (4), 516-
519.
Roberts, G., L., Lawrence, J. M., Williams, G. M., & Raphael, B. (1998). The
impact of intimate partner violence on women’s mental health.
Australian and New Zealand Journal o f Public Health, 22(7), 796-801.
Roberts, G. L., Williams, G. M., Lawrence, J. M., & Raphael, B. (1998). How
does intimate partner violence affect women’s mental health. Women &
Health, 28(1), 117-129.
Rodriguez, E., Lasche, K. E., Pinky, C.., & Lee, J. (2001). The rehtion of family
violence, employment status, welfare benefits and alcohol drinking in the
United States. Western Journal o f Medicine, 174(5), 317-323.
Rodgers, K. (1994). Wife assault: The findings of a national survey. Juristat, 14,
(9).Retrieved from http://www.hc-sc.gc.ca/hppb/ familyviolence/html/
femnational_e.html.
Saltzman, L.E., Fanslow, J. L., McMahon, P.M. & Shelley, G. A. (1999).
Intimate Partner Violence Surveillance: Uniform Definitions and
Recommended Data Elements, Atlanta: National Center for Injury
Prevention and Control, Centers for Disease Control and Prevention.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
127
Schechter, Susan (Dec 2002) Expanding Solutions fo r Intimate partner violence
and Poverty: What Battered Women with Abused Children Need from
Their Advocates Building Comprehensive Solutions to Domestic Violence:
Publication # 13, A Vision Paper. National Resource Center on Domestic
Violence, a project of the Pennsylvania Coalition Against Domestic
Violence. Retrieved from http://www.vaw.umn.edu/documents
Sharma, S. (1997). Intimate partner violence against minority women:
Interventions, preventions and health implications. Equal Opportunities
International, 16(2), 1-14.
Smith, M. (1990) Patriarchal ideology and wife beating. A test of a feminist
hypothesis. Violence and Victims, 5, 257-273.
Straus, M., Hamby, Sherry L., Boney-McCoy, S., & Sugarman, D. B.(1990).
The revised conflict tactics scales (CTS2). Journal o f Family Issues,
17(3), 283-317.
Straus, M.A. & Gelles, R. J. (1990). Physical violence in American Families:
Risk factorsand adaptations to violence in 8,145 families. New
Brunswick, NJ.
Stewart, A. L., Hays, R. D.,&Ware, J. E. (1988). The MOS short-form general
health survey: Reliability and validity in a patient population. Medical
Care, 26, 724-735.
Street, A. E., King, L. A., King, D. W., & Riggs, D. S. (2003). The associations
among male perpetrated partner violence, wives’ psychological distress
and children’s behavior problems: A structural equation modeling
analysis. Journal o f Comparative Family Studies, 34(1), 23-39.
Striepe, M. I., & Coons, H. L. (2002). W omen’s health in primary care:
Interdisciplinary interventions. Families, Systems & Health, 20(3), 237-
251.
Sutherland, C. L, Bybee, D. I., & Sullivan, C. M (2002). Beyond bruises and
broken bones: The joint effects o f stress and injuries on battered
women’s health. American Journal o f Community Psychology, 30 (5),
609-636.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
128
Sutherland, C. A., Sullivan, R. M., & Bybee, D. I. (2001). Effects of intimate
partner violence versus poverty on women’s health. Violence Against
Women, 7(2), 1122-1143.
Tjaden, P., & Thoennes, N. (1998). Prevalence, Incidence, and Consequences
of Violence Against Women: Findings From the National Violence
Against Women Survey, Research in Brief; Washington, D.C.: U.S.
Department of Justice, National Institute of Justice, 1998, NCJ 172837.
Tjaden, P., & Thoennes, N. (2000). Extent, Nature, and Consequences o f
Intimate Partner Violence: Criminal victimization reports; domestic
violence; rape and sexual assault. Retrieved from National Institute of
Justice and the Centers for Disease Control and Prevention,
http://www.ncjrs.Org/txtfilesl/nij/l 81867.txt
Tolman, R. M., & Rosen, D. (2001). Intimate partner violence in the lives of
women receiving welfare: Mental health, substance dependence, and
economic well-being. Violence Against Women, 7(2), 141-158.
U.S. Department of Health and Human Services Office of Minority Health
(2003, September. Initiative to Eliminate Racial and Ethnic Disparities
in Health. Retrieved from http://raceandhealth.hhs.gov/
Williams, D. R. (2002). Racial/ethnic variations in women’s health: The social
embeddedness of health. American Journal o f Public Health, 92(4), 588-
597.
World Health Organization (2002) World report on violence and health. Geneva
Switzerland Retrieved from http://www.who.int/ violence injury
prevention/ violence/world_report/en/ abstract_en.pdf.
Wyatt, G. E. (1985). The Wyatt Sex History Questionnaire. Los Angeles: The
University o f California Los Angeles. Department o f Psychiatry and
Biobehavioral Sciences.
Wyatt, G. E. (1990). The aftermath o f child sexual abuse of African American
and White American women: The victim's experience. Journal o f Family
Violence, 5, 61-81.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
129
Wyatt, G. E., Axelrod, J., Chin, D., Carmona, J. V., & Loeb, T. (2000).
Examining patterns of vulnerability to intimate partner violence among
African American women. Violence Against Women, 6 (5), 495-514.
Wyatt, G. E., & Chin, D. (1999). HIV and ethnic minority women, families and
communities: An overview, Cultural Diversity & Ethnic Minority
Psychology, 5(3), 179-182.
Wyatt, G. E., Myers, H. F., Williams, J. K., Kitchen, C. R., Loeb, T., Carmona,
J. V., Wyatt, L., Chin, D., & Presley, N. (2002). Does a history of
trauma contribute to HIV risk for women of color? Implications for
prevention and policy. American Journal o f Public Health, 92(4), 660-
665.
Yoshihama, M. (2002). Battered women’s coping strategies and psychological
distress: Differences by immigration status. American Journal o f
Community Psychology, 30(3), 429-452.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Predictors of dysphoria and substance abuse among Latino and African American youth

PDF

Impact of language and culture on a neuropsychological screening battery for Hispanics

PDF

Stress, appraisal and coping among Japanese-American, Anglo-American, African-American and Mexican-American spousal caregivers of persons with dementia

PDF

The effects of race and gender on respondent attitudinal perceptions of acquaintance rape

PDF

Substance use and disordered eating as outcomes of childhood maltreatment in a community sample of adolescent Latina females

PDF

The role of affect in the marital satisfaction of monoethnic Latino, biethnic Latino, and European -American newlywed couples

PDF

Role of ethnicity, body dissatisfaction, and stressful life events in disordered eating among Asian-American, Chicana/Latina, and European American college females

PDF

The roles of acculturation and family relationships in the career decision self -efficacy of Asian American college students

PDF

Psychological well-being, religious affiliation, and cultural identification in adult children of Jewish-Gentile interfaith marriages

PDF

Protective factors influencing resiliency among African -American and Latino students

PDF

Early adolescent drug use among multiethnic males: A prospective examination of the influences of psychological distress, relationship with family and school, law abidance, guilt and peer drug use

PDF

The impact of undergraduate multicultural courses on White students' racial identity development

PDF

Narrative analysis of Korean American women's acculturation and changing self

PDF

An adolescent -parent conflict resolution skills training program for ethnically diverse families: A program evaluation study

PDF

The relationship between changes in the type and frequency of contact in open and closed adoptions and behavioral outcomes of adopted children

PDF

The impact of sociocultural and media influences, self -esteem and acculturation on the development of eating disordered behaviors in African -American women

PDF

An integrative analysis of racism and quality of life: A comparison of multidimensional moderators in an ethnically diverse sample

PDF

The therapy relationship: Developing a sound metaphysical description of personal relations

PDF

Depression, hopelessness, and problem-solving in three Latino groups

PDF

Psychologists' perceptions of older clients: The effect of age, gender, knowledge, and experience

Asset Metadata

Creator Green, Kristen Nicole Presley (author)

Core Title Mental and physical health consequences of intimate partner violence in a multi-ethnic sample of women

Degree Doctor of Philosophy

Degree Program Education (Counseling Psychology)

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag health sciences, public health,OAI-PMH Harvest,Psychology, clinical,sociology, ethnic and racial studies,women's studies

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-403625

Unique identifier UC11336004

Identifier 3145206.pdf (filename),usctheses-c16-403625 (legacy record id)

Legacy Identifier 3145206.pdf

Dmrecord 403625

Document Type Dissertation

Rights Green, Kristen Nicole Presley

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA