Multiple Victimizations and Overdose Among Women With a History of Illicit Drug Use

Abstract

The experiences of violence and overdose are highly prevalent among women who use illicit drugs. This study sought to ascertain whether multiple victimizations during adulthood increase the frequency of women’s overdose. The sample comprised 218 women recruited at Philadelphia harm reduction sites during 2016–2017. Victimization was assessed as exposure to 16 types of adulthood violence. Three measures were constructed for multiple victimizations: continuous and categorical polyvictimization, and predominant violence domain. Negative binomial regression estimated the incidence rate ratio (IRR) of lifetime overdoses from multiple victimizations. Lifetime history of opioid use (88.6%) and drug injection (79.5%) were common. Among overdose survivors (68.5%), the median of lifetime overdoses was 3. The majority of participants (58.7%) were victims of predominantly sexual violence, 26.1% experienced predominantly physical abuse/assault, and 3.7% were victims of predominantly verbal aggression/coercive control. Participants reported a mean of seven violence types; the higher-score category of polyvictimization (9–16 violence types) comprised 41.7% of the total sample. In multivariable models, one-unit increase in continuous polyvictimization was associated with 4% higher overdose rates (IRR: 1.04, 95% confidence interval [CI]: [1.00, 1.08]). Compared to women who were not victimized (11.5%), those in the higher-score category of polyvictimization (IRR: 2.01; 95% CI: [1.06, 3.80]) and exposed to predominantly sexual violence (IRR: 2.10, 95% CI: [1.13, 3.91]) were expected to have higher overdose rates. Polyvictimization and sexual violence amplified the risk of repeated overdose among drug-involved women. Female overdose survivors need to be screened for exposure to multiple forms of violence, especially sexual violence. Findings underscore the need to scale-up victimization support and overdose prevention services for disenfranchised women.

Keywords

violence exposure alcohol and drugs adult victims sexual assault

Introduction

The everyday lives of women who use drugs are threatened by pervasive violence and the risk of death from overdose. In various samples of socially marginalized drug-using women, the lifetime prevalence of sexual victimization reaches 68% (Braitstein et al., 2003), markedly exceeding the estimated prevalence of sexual violence (43.6%) among women in the general population (Smith et al., 2018). Nonfatal overdose is also common, with lifetime rates ranging from 40% among polysubstance-using female injectors (Kerr et al., 2007) to 72% among women living with HIV and injecting drugs (Lunze et al., 2016). Moreover, in various studies, fatal overdose was found to be the most frequent (Hser et al., 2012) or the second frequent (Hayashi et al., 2016) cause of death among drug-involved women.

Given the high rates of violence and overdose among women who use drugs, it could be argued that frequently or severely victimized drug-using women may also be prone to overdose. Yet, the link between violence and overdose has been underexplored. Previous research pointed to several factors that may confound a relationship between violence and overdose. In particular, exposure to violence was found to be associated with unsafe drug use practices among women, particularly illicit opioid use (Hall et al., 2016; Jessell et al., 2017) and injection drug use (El-Bassel et al., 2007)—behaviors that have been also associated with increased risk of overdose (Bohnert et al., 2011; Colledge et al., 2019; Kerr et al., 2007; Mitra et al., 2015; Rudd et al., 2016). Studies have also found a higher risk for overdose among socially marginalized people, such as homeless (Baggett et al., 2013), ex-prisoners (Binswanger et al., 2013; Bukten et al., 2017), and people involved in sex trade (Seal et al., 2001)—populations with documented high rates of victimization among women (Radatz & Wright, 2017; Shannon et al., 2009; Wong et al., 2016).

To date, only four published studies have examined the effect of violence on overdose and the results are inconsistent. Two studies reported significant associations between violence and overdose in mixed-gender samples of drug injectors. Milloy et al. (2008) documented higher odds of past 6-month nonfatal overdose (odds ratio [OR]: 1.45) among victims of past 6-month abuse in a prospective cohort of drug injectors recruited from a safe injection facility in Vancouver, Canada. Bohnert et al. (2011) found that exposure to sexual (OR: 2.79) and physical (OR: 1.64) victimization before the age of 18 was significantly associated with both a nonfatal overdose and suicide attempt in a large U.S. national sample of adults entering drug addiction treatment. As the results were not disaggregated by gender, one cannot determine the specific effects of violence on women even though the data suggest an association. While the prevalence of nonfatal overdose may be comparable for women and men (Darke et al., 2007; Kinner et al., 2012), levels of violence are often higher among women. Studies have shown that compared to drug-using men, women report more sexual violence as children (Braitstein et al., 2003; Maloney et al., 2009) and adults (Braitstein et al., 2003), are more likely to be victimized by their intimate partners (Marshall et al., 2008; Schneider et al., 2009), and die from homicide (Hayashi et al., 2016; Miller et al., 2007).

Two other studies, however, reported nonsignificant associations between violence and overdose among women who use drugs. Braitstein and coauthors (2003) examined a prospective mixed-gender cohort of injection drug users in Vancouver, Canada, and found that among females, exposure to sexual violence at childhood, adolescence, or adulthood was not correlated with accidental overdose, adjusting for sociodemographic variables. Lunze and colleagues (2016) recruited a women-only sample of HIV-positive drug injectors in St. Petersburg, Russia, and reported that lifetime sexual violence perpetrated by the police was not significantly associated with lifetime overdose at p = .07 (OR: 2.0, 95% confidence interval [CI]: [0.94, 4.40]), controlling for childhood sexual violence, depression, stigma score, incarceration, and sex work.

These results indicate that the relation between violence and overdose among drug-using women needs to be further explored. First, no previous study examined multiple victimizations or the amount of violence (polyvictimization) as an independent variable. This is problematic because research indicates that women who use drugs are subjected to multiple types of violence, including sexual violence (Braitstein et al., 2003; Lunze et al., 2016), physical abuse (Marshall et al., 2008), and verbal aggression and coercive control (Bourgois et al., 2004; Epele, 2002). In addition, little is known about the association of adulthood violence with women’s risk of overdose, except for adult sexual violence as examined by Braitstein et al. (2003). Finally, all previous studies reported the effect of violence on the odds of overdose as a binary measure, which does not distinguish between women who may have just one overdose and those with multiple overdose history. Hence, the existing evidence does not estimate the effect of violence on frequency of nonfatal overdoses, which have been associated with increased risk of overdose death through a dose–response relationship (Caudarella et al., 2016).

To address these gaps, this study examined whether frequency and types of adulthood victimizations increase the frequency of women’s personal overdose. In particular, we sought to answer the following questions: (a) What is the prevalence of individual and multiple adulthood victimizations among women with a history of illicit drug use? (b) What is the frequency of lifetime overdoses in this group? and (c) Are frequency and combinations of adulthood victimizations associated with overdose frequency controlling for potential confounders, such as risky drug use and social marginalization?

Method

Local Context, Recruitment, and Procedures

This analysis arose from the quantitative part of a larger mixed-methods study that examined personal and witnessed overdose among women in Philadelphia. In 2018, Philadelphia’s opioid overdose death rate of 49.2 per 100,000 residents (Department of Public Health, City of Philadelphia, 2020) was more than 3 times the national rate of 14.6 per 100,000 (Wilson et al., 2020). The larger study utilized venue-based recruitment at Prevention Point Philadelphia (PPP), the only harm reduction agency in the city, which serves marginalized populations, including people who use drugs, sex workers, people living with HIV, and homeless people. The organization’s main office is located in Kensington, an impoverished neighborhood in North Philadelphia known for a large concentration of open-air street-based drug sales (Fairbanks, 2011) and high density of opioid-involved overdose deaths (Philadelphia Department of Public Health, 2018). Services are also offered at mobile harm reduction sites in other parts of the city.

Inclusion criteria included women’s gender, adult age (18 or older), being nonpregnant (as self-report of violence could provoke traumatic memories and emotional reactions), and the ability to read in English. Between January 2016 and January 2017, 229 women who approached PPP for any service were contacted and screened. Screening involved approaching a female participant at the recruitment location and asking brief questions about age, literacy, and pregnancy; decisions to include/exclude were based on participants’ self-reports. Of 229 women screened, nine women did not qualify, including five who could not read, three who were pregnant, and one who was under 18. As a result, 220 women were enrolled in the quantitative arm of the larger study. This includes 199 women enrolled at the PPP main office in Kensington (through the drop-off center and weekly Women’s Night) and 21 enrolled at five mobile harm reduction points in Center City, North, and West Philadelphia.

Participants provided verbal informed consent and completed a 30- to 40-min interviewer-administered questionnaire in PPP main office’s private spaces, PPP mobile van (without the presence of staff), or other mutually agreed location, such as a nearby fast-food restaurant. The interviewer asked questions out loud and then recorded answers electronically in Qualtrics (Provo, UT, 2018). Participants provided answers verbally. Multiple-choice questions were accompanied by paper cards with available response options; participants were asked to look at the cards and read back applicable responses. All participants received US$10 in cash upon interview completion. Screening, enrollment, and interviewing were done on the same day, during weekdays. The Drexel University Institutional Review Board approved all study procedures.

For this study, the analytical sample comprised 218 of 220 women. One woman was excluded because she never used illicit drugs. A second woman did not disclose her age and her case was “missing” in all bivariate and multivariable models involving overdose rates based on age.

Measures

Sociodemographics

Participants were asked questions about age, race, Hispanic ethnicity, sexual orientation, education, family situation, employment status, and housing. Unstable housing was created as a composite binary variable (yes/no) that took a value of “yes” if participants endorsed at least one of the following items: living in a car, abandoned building/house, shelter, motel, transitional housing, on the streets, temporarily living with friends or family, or having any other unstable housing in the past 30 days or 12 months. Participants were also asked about their history of arrest and incarceration. Two items assessed the participants’ involvement in street economy: recent and lifetime survival sex (defined as exchanging sex for money, drugs, housing, or gifts) and selling drugs in public places in the past 30 days.

Drug use history and practices

The instrument assessed lifetime opioid use and lifetime drug injection. A series of questions also examined 30-day drug use practices, such as injection, opioid use (heroin or prescription opioids), street drug use (cocaine, crack, marijuana, and illicit drugs other than heroin), alcohol, as well as polydrug use, including speedballs (the concurrent use of heroin and cocaine or crack) and other drug combinations.

Overdose frequency (the dependent variable)

As polysubstance use is highly prevalent among opioid-using women (Jarlenski et al., 2017) and because of the study’s interest in overdoses from various substances, similar to previous research (Kerr et al., 2007), the question about the lifetime frequency of nonfatal overdose was not focused on specific substances and broadly asked: “In your lifetime, how many times have you overdosed on drugs? By ‘drug overdose’ I mean a situation when you’re taking more of a drug than your body can process. It can lead to some serious problems like blackout, inability to breathe or even heart failure.” If a participant never had an overdose, her response was coded as “0” (zero).

Individual victimizations (independent variables)

The exposure to various forms of adulthood violence was measured by 16 items, including 14 adapted and two original items (Table 1). Use of items from different sources was necessitated by the lack of a brief instrument that would measure multiple forms of violence by multiple perpetrators in women with a history of illicit drug use. Exposure to adulthood-only violence was assessed to improve recall bias and also because it has been rarely examined among drug-using women, to date.

Table 1.

Measurement of Individual Victimizations.

Violence Domain	Item	Source
1. Sexual violence	1.1. Attempted unwanted sex (vaginal/oral/anal)	Developed by the authors based on Tolin & Foa’s (2006) classification of sexual assault as “all unwanted contact (or attempted contact) of a sexual nature” (p. 961)
1. Sexual violence	1.2. Forced/coerced sex (vaginal/oral/anal)	Adapted from the Universal Violence Prevention Screening Protocol (Dutton et al., 1996)
2. Physical abuse/assault	2.1. Slapped, pushed, grabbed, or shoved
	2.2. Choked, kicked, bit, or punched
	2.3. Beaten up	Adapted from the Traumatic Life Events Questionnaire (Johnson & Lubin, 2015)
	2.4. Threatened with a knife or gun to become scared or hurt	Adapted from Universal Violence Prevention Screening Protocol (Dutton et al., 1996)
	2.5. Attacked with a knife or gun
	2.6. Abducted/kidnapped	Adapted from the Traumatic Life Events Questionnaire (Johnson & Lubin, 2015)
3. Verbal aggression/threats	3.1. Perpetrator’s repeated use of words, yelling or screaming in a way that frightened, threatened, or put down	Adapted from Heron and co-authors (2003) who added the item to the Universal Violence Prevention Protocol
	3.2. Threatened to be told on and was made to do something based on those threats	Developed by the authors upon receiving feedback on a survey instrument from Prevention Point Philadelphia staff
	3.3. Made to be afraid to be physically hurt	Adapted from the Universal Violence Prevention Screening Protocol (Dutton et al., 1996)
4. Coercive control	4.1. Was forced to take drugs or alcohol	Adapted from the Traumatic Life Events Questionnaire (Johnson & Lubin, 2015)
	4.2. Contacts with others were controlled
	4.3. Possessions were controlled
	4.4. Use of phone was controlled
	4.5. Was locked in a room/closet

The violence items had high internal consistency (Cronbach’s alpha = .912) and were organized into four broad domains: sexual violence, physical abuse/assault, verbal aggression/threats, and coercive control (Table 1). The identification of the domains was based on a typology used in prior research on violence against drug-using women (Bourgois et al., 2004; Braitstein et al., 2003; Epele, 2002; Lunze et al., 2016; Marshall et al., 2008). A participant was considered to be exposed to a particular violence domain if she endorsed at least one victimization type from that domain.

Batteries of almost identical questions were successively asked in relation to six types of potential perpetrators: intimate partner, family member, sex work client, police officer, drug dealer, and stranger. Each series of questions started with identifying a particular perpetrator type, for example: “Has your partner, boyfriend or husband (current or former) . . .,” following with the list of victimization items.

The exposure to a particular victimization was assessed as “yes” if the participant endorsed any of the two time frames for that victimization: since the age of 18 excluding the past year and during the past year.

As the five coercive control items were mainly relevant to intimate partners or family members (i.e., the woman’s immediate social circle), all 16 violence questions were asked in relation to these two perpetrator types only. Violence from the other perpetrators—sex work clients, police officers, drug dealers, and strangers—was measured by 11 items representing sexual violence, physical abuse/assault, and verbal aggression/threats domains. Respondents received questions about perpetration from a sex work client only if they endorsed lifetime survival sex.

Multiple victimizations (independent variables)

To assess the prevalence of multiple victimizations, we constructed three aggregate measures. The first two measures examined the amount of adulthood violence, based on Finkelhor et al. (2007) research on polyvictimization. To construct a continuous measure of polyvictimization, we summed the number of victimization types that resulted in a composite score of 0 to 16. A categorical measure of polyvictimization converted the continuous measure into three groups: (a) “Higher score”—participants reporting nine to 16 victimizations, that is, above the mean number of victimizations among victimized participants, that replicated the procedure of defining “poly-victims” in prior polyvictimization research (Finkelhor et al., 2007); (b) “Lower score”—participants who reported one to eight victimizations, that is, equal to or below the mean number of victimizations that corresponds to the “less victimized” category in Finkelhor et al. (2007) study; and (c) “Not victimized”—participants not reporting any adulthood violence. Prior research on adult women’s polyvictimization (Radatz & Wright, 2017) adapted similar procedures from Finkelhor et al. (2007) by assigning the “polyvictim” status to those who scored above the mean number of victimizations in the total sample.

A third measure for multiple victimizations evaluated the relative effects of different combinations of violence types. We developed this measure based on Lau et al. (2005) hierarchical classification of predominant maltreatment types. This approach treats the active forms of violence, such as sexual and physical, as having higher importance in terms of severity, compared to the passive forms of violence, such as psychological or emotional violence. As a result, this variable divided the sample into four mutually exclusive categories: (a) “Predominantly sexual violence,” to classify victims of sexual violence and potentially other forms of violence; (b) “Predominantly physical abuse/assault,” to identify victims of physical abuse/assault and potentially other forms of violence except sexual violence; (c) “Verbal aggression or coercive control,” to identify women who reported any verbal aggression/threats and/or coercive control, but did not experience sexual or physical violence; and (d) “Not victimized” women.

Analysis

The lifetime number of overdoses, the dependent variable, was a count variable with overdispersed variance (greater than the mean). We examined several distributions suitable for modeling count dependent variable, including Poisson, negative binomial, and zero-inflated models. Negative binomial model provided better estimates of the incidence rate ratio (IRR) for the dependent variable based on such criteria as deviance statistic, Akaike information criterion (AIC) and Bayesian information criterion (BIC). To account for various lengths of time, during which lifetime overdoses have happened, we modeled overdose rates (counts per 1 year of age) instead of simple overdose counts and used the natural log of age as an offset in both bivariate and multivariable negative binomial regression analyses.

Bivariate analyses examined associations of individual or multiple victimizations with lifetime overdose rates. Multivariable models tested for associations between multiple victimizations and lifetime overdose rates after controlling for five covariates selected a priori, including opioid use, drug injection, unstable housing, survival sex, and incarceration. As was described above, those factors were found to be associated with increased risk of both overdose and violence.

As the dependent variable was a lifetime measure, only lifetime covariates were entered in the model with the exception of unstable housing for which only past 12-month measure was available. Variance inflation factors (VIFs) did not exceed 2.0 for each independent variable and a set of control variables, indicating the lack of multicollinearity in multivariable models. We also tested associations between the covariates and used the Phi coefficient since all covariates were categorical variables.

The comparative fit of Poisson, negative binomial, and zero-inflated models was assessed in SAS 9.4 (SAS Institute Inc, 2013) and all other analyses were conducted in IBM SPSS Statistics for Windows, version 25 (IBM Corp., 2017).

Results

Sociodemographic Characteristics and Drug Use Practices

Of 218 participants, the majority were White (63.8%) and in their 30s–40s (mean age = 38.2, SD = 10.1). Almost half (49.5%) were married or lived with a partner. More than three quarters (78.1%) had children and almost one third of mothers (54 out of 171) reported their parental rights had ever been terminated or suspended. The rate of formal employment was low (under 4%). However, in the past 30 days, 42.2% women reported survival sex and 26.3% sold drugs in public places. More than half (55.3%) were unstably housed in the past 30 days, and 70.6% in the past 12 months. A majority had lifetime involvement with criminal justice system, including arrest (80.8%) or incarceration (69.7%) (see Table 2).

Table 2.

Sociodemographic Characteristics, Drug Use, and Overdose History (N = 218).

Variable	Total, % (n)
Age
M (SD)	38.2 (10.1)
Race and ethnicity
White	63.8 (140)
African American	19.6 (43)
Hispanic	7.3 (16)
Native American	1.4 (3)
Mixed or other race	7.8 (17)
Hispanic ethnicity, any race (n = 217)	13.8 (30)
Sexual orientation
Sexual minority (homosexual, bisexual, or transsexual) (n = 201)	28.4 (57)
Family situation
Married or living with a partner	49.5 (109)
Having children	78.1 (171)
Parental rights ever terminated or suspended	24.7 (54)
Education
High school graduate or GED	69.3 (152)
Employment, full- or part time (n = 214)	3.7 (8)
Involvement in street economy
Survival sex, lifetime	75.3 (165)
Survival sex, past 30 days	42.2 (93)
Drug sales in public places, past 30 days	26.3 (56)
Criminal justice history
Arrest, lifetime	80.8 (177)
Incarceration (prison or jail), lifetime	69.7 (153)
Unstable housing
Past 12 months	70.6 (155)
Past 30 days	55.3 (121)
Drug use practices, lifetime
Injection	79.5 (174)
Any opioid use	88.6 (194)
Drug use practices, past 30 days
Any illicit drug use	90.4 (198)
Injection	68.0 (149)
Any opioid use	76.3 (167)
Heroin	72.6 (159)
Cocaine/crack	68.5 (150)
Any polysubstance use	63.9 (140)
Speedball (heroin and cocaine/crack)	44.3 (97)
Drug use in public places	71.1 (155)
History of drug overdose
Ever overdosed	68.5 (150)
Lifetime number of overdoses among overdose survivors, median	3
Range	1–30

Note. GED = General Education Development.

Participants had a high prevalence of drug use practices that increase overdose risks. The majority had a lifetime history of opioid use (88.6%) and drug injection (79.5%). In the past 30 days, more than three quarters (76.3%) used some opioids, including heroin or prescription painkillers, 63.9% reported polysubstance use (the concomitant use of two or more drugs), more than two thirds (68.0%) injected drugs, and 71.1% used drugs in public places.

We examined associations between five drug use and social marginalization characteristics used as covariates in multivariable models: lifetime opioid use, lifetime drug injection, lifetime incarceration in multivariable, lifetime survival sex, and 12-month unstable housing. We found several significant correlations, including those between lifetime opioid use and lifetime injection (φ = .67, p < .001), lifetime survival sex and lifetime injection (φ = .31, p < .001), lifetime opioid use and lifetime survival sex (φ = .26, p < .001), lifetime survival sex and lifetime incarceration (φ = .22, p < .01), and lifetime opioid use and past 12-month unstable housing (φ = .15, p < .05). Those correlations could increase the uncertainty in the estimates of association between the covariates and the outcome, but they did not affect our ability to control for them as confounders in the relationship between our primary exposure (adulthood victimizations) and outcome (overdose frequency).

Overdose History

More than the two thirds of participants (68.8%) experienced at least one lifetime overdose. Among those that overdosed, half reported more than three lifetime overdoses, and the total number of overdoses ranged from 1 to 30.

Individual Victimizations and Violence Domains

A significant proportion of participants (88.5%) experienced some form of interpersonal violence during their adulthood (data not shown). Prevalence of individual adulthood victimizations ranged from 8% for being “forced to take drugs or alcohol” to 78% for being “slapped, pushed, grabbed, or shoved.” Exposure to each violence domain was endorsed by the majority of the sample, including physical abuse or assault (83.5%), verbal aggression or threats (78.9%), sexual violence (58.7%), and coercive control (53.2%) (see Table 3).

Table 3.

Prevalence of Adulthood Victimizations and Their Unadjusted Associations With Overdose Rates (N = 218).

Violence Domains and Items	% (n)	Unadjusted IRR [95% CI]^a
1. Any sexual violence	58.7 (128)	2.47 [1.73, 3.51]***
1.1. Attempted unwanted sex (vaginal/oral/anal)	49.1 (107)	2.08 [1.47, 2.94]***
1.2. Forced/coerced sex (vaginal/oral/anal)	50.9 (111)	2.03 [1.43, 2.87]***
2. Any physical abuse/assault	83.5 (182)	2.73 [1.64, 4.56]***
2.1. Slapped, pushed, grabbed, or shoved	78.0 (170)	2.43 [1.56, 3.80]***
2.2. Choked, kicked, bit, or punched	64.7 (141)	2.02 [1.39, 2.94]***
2.3. Beaten up	57.3 (125)	1.63 [1.14, 2.34]**
2.4. Threatened with a knife or gun to become scared or hurt	47.7 (104)	1.23 [0.86, 1.76]
2.5. Attacked with a knife or gun	34.4 (75)	1.05 [0.72, 1.52]
2.6. Abducted/kidnapped	19.3 (42)	1.42 [0.91, 2.22]
3. Any verbal aggression/threats	78.9 (172)	3.01 [1.90, 4.78]***
3.1. Perpetrator’s repeated use of words, yelling or screaming in a way that frightened, threatened, or put down	72.9 (159)	2.48 [1.65, 3.75]***
3.2. Threatened to be told on and was made to do something based on those threats	40.4 (88)	1.59 [1.11, 2.27]*
3.3. Made to be afraid to be physically hurt	61.0 (133)	2.19 [1.53, 3.15]***
4. Any coercive control	53.2 (116)	1.57 [1.10, 2.24]*
4.1. Was forced to take drugs or alcohol	8.3 (18)	2.41 [1.32, 4.42]**
4.2. Contacts with others controlled	43.1 (94)	1.60 [1.12, 2.29]**
4.3. Possessions controlled	34.9 (76)	1.71 [1.19, 2.46]**
4.4. Use of phone controlled	35.8 (78)	1.68 [1.17, 2.41]**
4.5. Was locked in a room/closet	19.3 (42)	1.48 [0.95, 2.30]

Note. IRR = incidence rate ratio; CI = confidence interval.

All models include the natural log of age as offset.

*p < .05.

**p < .01.

***p < .001.

Multiple Victimizations

The mean number of total victimizations was 7.17 in the total sample (Table 4) and 8.1 among all victimized women (data not shown). Two fifths of participants (41.7%) were in the higher-score category of polyvictimization (i.e., those who endorsed 9–16 violence types) and a slightly larger proportion (46.8%) was lower-scored (1–8 violence types). When victimization was classified by a predominant violence domain, predominantly sexual violence, which included all victims of sexual violence and all other types of violence, was the most prevalent (58.7%). Predominantly physical abuse/assault—a category that excluded sexual violence victimizations, but was potentially mixed with verbal aggression or coercive control victimizations—comprised 26.1% participants. Verbal aggression or coercive control category, which excluded victims of sexual or physical violence, was the smallest (3.7%) (see Table 4).

Table 4.

Prevalence of Multiple Victimizations and Their Associations With Overdose Rates (N = 218).

Variable	Prevalence% (n)	Unadjusted IRR[95% CI]^a	Adjusted IRR[95% CI]^a,b
Multiple victimizations
Polyvictimization—continuous
M (SD)	7.17 (4.84)	1.09 [1.05, 1.13]***	1.04 [1.00, 1.08]*
Polyvictimization—categorical
Higher score (9–16 types)	41.7 (91)	3.43 [1.81, 6.52]***	2.01 [1.06, 3.80]*
Lower score (1–8 types)	46.8 (102)	2.28 [1.20, 4.33]*	1.71 [0.91, 3.21]
Not victimized	11.5 (25)	1.0 [Ref]	1.0 [Ref]
Predominant violence domain
Predominantly sexual violencec	58.7 (128)	3.48 [1.89, 6.41]***	2.10 [1.13, 3.91]*
Predominantly physical abuse/assaultd	26.1 (57)	1.64 [0.84, 3.19]	1.47 [0.76, 2.83]
Verbal aggression or coercive control	3.7 (8)	1.06 [0.36, 3.19]	1.38 [0.47, 4.07]
Not victimized	11.5 (25)	1.0 [Ref]	1.0 [Ref]

Note. IRR = incidence rate ratio; CI = confidence interval.

All models include the natural log of age as offset.

Adjusted by 12-month unstable housing, lifetime opioid use, lifetime drug injection, lifetime survival sex, and lifetime incarceration.

Sexual violence alone or in any combination with physical abuse/assault, verbal aggression, and coercive control.

Physical abuse/assault alone or in any combination with verbal aggression and coercive control.

*p < .05.

**p < .01.

***p < .001.

Association of Victimizations With Overdose Rates

In bivariate analyses involving individual victimizations and violence domains, 12 out of 16 violence types and all violence domains were positively associated with overdose rates (see Table 3).

In bivariate analyses, the following multiple victimizations were expected to increase overdose rates: one-unit increase in polyvictimization as a continuous measure (IRR: 1.09, 95% CI: [1.05, 1.13]), the higher-score (9–16 violence items) category of polyvictimization (IRR: 3.43, 95% CI: [1.81, 6.52]), the lower-score (1–8 violence items) category of polyvictimization (IRR: 2.28, 95% CI: [1.20, 4.33]), and predominantly sexual violence (IRR: 3.48, 95% CI: [1.89, 6.41]) (see Table 4).

In multivariable models, each additional victimization was expected to increase the overdose rate, on average, by 4%, while holding the other variables in the model constant (IRR: 1.04, 95% CI: [1.00, 1.08]). Compared to not-victimized women, the expected overdose rate was, on average, 2.01 times higher for women in the higher-score category of polyvictimization (IRR: 2.01; 95% CI: [1.06, 3.80]) and 2.10 times higher for women exposed to predominantly sexual violence (IRR: 2.10, 95% CI: [1.13, 3.91]), while holding all other variables constant in the model (Table 4).

Discussion

To our knowledge, this is the first study that quantified both the number of overdoses and victimizations, and examined how multiple victimizations were associated with the frequency of overdose in a sample of women recruited from harm reduction sites. Overall, results suggest that exposure to many violence types during adulthood may amplify the number of lifetime overdoses even after accounting for risky drug use behaviors, such as opioid use and injection, and features of social marginalization, such as unstable housing, incarceration, and survival sex. Our findings highlight the potential utility of addressing violence as a means of reducing overdose frequency among vulnerable women.

The results elucidated the multiplicity of various violence types in this sample. On average, women reported more than seven types of violence. Moreover, two out of five women were exposed to the high amount of violence during adulthood, that is, more than eight violence types. These results extend previous research by demonstrating that not only sexual violence (Braitstein et al., 2003) or physical abuse (Marshall et al., 2008) are very common among drug-involved women, but also that various types of violence are interconnected in this vulnerable population. This fits with polyvictimization research showing that experiences of violence seldom happen in isolation and that one victimization often indicates the presence of another (Finkelhor et al., 2007; Price-Robertson et al., 2013).

We documented significant association of higher risk of overdose with both continuous and categorical measures of polyvictimization. Each additional type of violence slightly increased the risk of higher overdose rates; exposure to the higher amount of violence (more than eight types of violence) doubled this risk. One possible explanation for this finding is that after a certain threshold, multiple victimizations may interfere with normal coping (Finkelhor et al., 2007); unsafe drug use, a form of unhealthy coping, implies a risk of overdose. The link between polyvictimization and adverse health outcomes has also been found in prior research, which showed that the higher amount of co-occurring victimizations was associated with psychological distress (Richmond et al., 2009), psychiatric impairment and delinquency (Ford et al., 2010), and suicidal behavior (Charak et al., 2016).

Out of many forms of violence, sexual assault may have a particularly detrimental effect on the number of overdose events among drug-using women. The experience of attempted or forced unwanted sex was tightly intertwined with other victimizations and strongly associated with the risk of higher overdose rates. Furthermore, the experience of any violence other than sexual was not associated with higher overdose rates. Our results are somewhat in contrast with those of Lunze et al. (2016) and Braitstein et al. (2003) that did not find a link between sexual violence and lifetime overdose. While not all victims of sexual violence are predisposed to overdose, those who did experience a personal overdose may have greater susceptibility to subsequent overdoses for at least two reasons. First, sexual violence can cause long-term deleterious physical and mental health outcomes, which are also independently associated with overdose. In particular, sexual violence has been associated with long-lasting physical pain (Ulirsch et al., 2014), which in its turn has been linked to overdose, particularly through the nonmedical use of prescription opioids (Bonar et al., 2014). The other possible pathway involves chronic mental health conditions resulting from sexual violence, including posttraumatic stress disorder (PTSD) and depression (Campbell et al., 2009), which have been associated with overdose (Bartoli et al., 2014; Havens et al., 2011). Additionally, as prior research indicated, women’s exposure to childhood violence may result in long-term sequelae, such as adulthood substance abuse (Salisbury & Van Voorhis, 2009; Stein et al., 2017) or adulthood sexual violence (Tripodi & Pettus-Davis, 2013). Therefore, it is also likely that women’s early victimizations can create pathways leading to the observed association between adulthood sexual violence and overdose.

It is noteworthy that after multivariable adjustment, the associations between overdose rates and multiple victimizations became attenuated. This suggests a multiplicity of factors contributing to women’s overdose. Still, the fact that the relationship between multiple victimizations and overdose rates did not disappear in the final models point to violence as an overdose risk factor, which can work through many mechanisms, but is also modifiable through direct public health interventions (Beattie et al., 2010; Gilbert et al., 2001).

Significantly, our study demonstrated an association between multiple victimizations and higher overdose frequency among women with a history of illicit drug use. Our findings have implications for future research, public health, and clinical practice. Given that childhood physical and sexual abuse is strongly correlated with violence experienced during adulthood (Parks et al., 2011), future research should assess the effects of both childhood and adulthood violence on repeated overdose. In addition, health conditions, especially pain, depression, and PTSD, should be tested as potential mediators between violence and overdose, as can be implied from previous literature. While this study did not distinguish between accidental and intentional overdose, the exploration of the effect of multiple victimizations on overdoses with and without the intent of self-ham can be a fruitful area for future research. To untangle gender-specific effects of violence on overdose, future studies should also examine the exposure to multiple forms of violence and personal overdose in mixed-gender samples of people who use drugs. Finally, future research should utilize longitudinal designs to establish a causal or bidirectional relationship between victimization and overdose and also investigate whether the association between polyvictimization and overdose frequency is due to an additive or synergistic effect of multiple violence types.

These findings have public health implications and call for the scale-up of overdose prevention services and victimization support services for drug-involved women. Preferably, such services should be integrated in the same place, such as harm reduction programs, mental health clinics, substance abuse treatment programs, and other services frequented by drug-using women. In addition, emergency departments treating female overdose survivors, especially those with multiple overdose histories, should screen women for possible exposure to multiple types of violence, particularly sexual assault, and offer referrals to appropriate services, including psychosocial support, accommodation/shelter, and legal counseling for violence victims.

The study has several limitations. The sample was not randomly selected, which may constrain the generalizability of findings to a broader group of female participants with a history of illicit drug use. However, the demographic characteristics of the sample were consistent with the characteristics of PPP female clients (S. Mazzella, email communication, August 18, 2016). The number of overdoses is subject to recall bias. Moreover, despite the fact that we defined an overdose for participants, it is one of many other overdose definitions that may not necessarily coincide with the layperson’s concept of drug overdose. In addition to overdose, all other measures were self-reported, but studies have found self-reported data by people who use drugs to be reliable (Darke, 1998; Kokkevi et al., 1997). In addition, as with other cross-sectional studies, results do not prove causality. It is possible that some overdoses (dependent variable) occurred before the age of 18, that is, before adulthood victimizations (independent variable). This is not vital however, because the regression models tested an association between independent and dependent variables rather established directionality of the association. Finally, the results may not be generalizable to women who use drugs in other urban contexts given Philadelphia’s high rates of overdose and violence, especially in the Kensington area. Still, our results may be indicative for the overall direction of the relationship between violence and overdose among disenfranchised drug-using women in the United States.

In conclusion, we found that polyvictimization and sexual violence may have contributed to a greater number of overdoses among drug-involved women in Philadelphia. The potential confluence of violence and overdose should raise the importance of more complex approaches to overdose prevention—approaches that are not limited to safer drug use per se, but also take into account complex social realities facing women with a history of illicit drug use.

Footnotes

Acknowledgments

We would like to thank Drs. Greg Falkin and George De Leon for their input into the development of this manuscript. We are indebted to Prevention Point Philadelphia (PPP) anonymous participants for sharing their violence and overdose experiences and PPP staff for their advisory role and feedback on the results.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This publication was made possible by a National Institute on Drug Abuse (NIDA)-funded predoctoral fellowship to Janna Ataiants (T32DA007233-33) in the Behavioral Sciences Training in Drug Abuse Research Program at New York University.

ORCID iD

Janna Ataiants

Author Biographies

Janna Ataiants, DrPH, is a postdoctoral research fellow in the Department of Community Health and Prevention at Drexel University’s Dornsife School of Public Health. Her research examines the experiences of people who use illicit drugs and the role of social realities in the production of adverse health outcomes in this disenfranchised group.

Silvana Mazzella, MA, is the associate executive director at Prevention Point Philadelphia, a private nonprofit organization providing harm reduction services to Philadelphia. She is responsible for program planning, initial implementation, and evaluation of all programs and services. She is also an adjunct faculty in the Department of Geography and Urban Studies at Temple University teaching research methods and drug policy.

Alexis M. Roth, PhD, is an associate professor in the Department of Community Health and Prevention at Drexel University’s Dornsife School of Public Health. Her research is focused on understanding how individual, social, and environmental factors influence sexual health and drug use–related health disparities. She utilizes mixed methods and community-engaged research approaches to co-create sustainable solutions to public health problems.

Lucy F. Robinson, PhD, is an assistant professor in the Department of Epidemiology and Biostatistics at Drexel University’s Dornsife School of Public Health. Her research has focused on modeling and inference techniques for functional MRI (fMRI) data and on change point detection in network data.

Randall L. Sell, ScD, is a professor in the Department of Community Health and Prevention at Drexel University’s Dornsife School of Public Health. He has more than 30 years of experience addressing health disparities experienced by lesbian, gay, bisexual and transgender people. He is the director of Drexel University’s Program for Lesbian, Gay, Bisexual and Transgender Health.

Stephen E. Lankenau, PhD, is a professor and associate dean for research at Drexel University’s Dornsife School of Public Health. A sociologist by training, his research focuses on the use and misuse of substances at the forefront of current public health concerns, including medical cannabis and opioids, and associated morbidities, such as overdose, HIV, HCV, and drug dependence.

References

Baggett

T. P.

Hwang

S. W.

O’Connell

J. J.

Porneala

B. C.

Stringfellow

E. J.

Orav

E. J.

Rigotti

N. A.

(2013). Mortality among homeless adults in Boston: Shifts in causes of death over a 15-year period. JAMA Internal Medicine, 173(3), 189–195. https://doi.org/10.1001/jamainternmed.2013.1604

Bartoli

Carrà

Brambilla

Carretta

Crocamo

Neufeind

, … Clerici

(2014). Association between depression and non-fatal overdoses among drug users: A systematic review and meta-analysis. Drug and Alcohol Dependence, 134, 12–21. https://doi.org/10.1016/j.drugalcdep.2013.10.007

Beattie

T. S.

Bhattacharjee

Ramesh

B. M.

Gurnani

Anthony

Isac

, … Blanchard

J. F.

(2010). Violence against female sex workers in Karnataka state, south India: Impact on health, and reductions in violence following an intervention program. BMC Public Health, 10(1), Article 476. https://doi.org/10.1186/1471-2458-10-476

Binswanger

I. A.

Blatchford

P. J.

Mueller

S. R.

Stern

M. F.

(2013). Mortality after prison release: Opioid overdose and other causes of death, risk factors, and time trends from 1999 to 2009. Annals of Internal Medicine, 159(9), 592–600. https://doi.org/10.7326/0003-4819-159-9-201311050-00005

Bohnert

A. S.

Roeder

K. M.

Ilgen

M. A.

(2011). Suicide attempts and overdoses among adults entering addictions treatment: Comparing correlates in a U.S. National Study. Drug and Alcohol Dependence, 119(1), 106–112. https://doi.org/10.1016/j.drugalcdep.2011.05.032

Bonar

E. E.

Ilgen

M. A.

Walton

Bohnert

A. S.

(2014). Associations among pain, non-medical prescription opioid use, and drug overdose history. The American Journal on Addictions, 23(1), 41–47. https://doi.org/10.1111/j.1521-0391.2013.12055.x

Bourgois

Prince

Moss

(2004). The everyday violence of hepatitis C among young women who inject drugs in San Francisco. Human Organization, 63(3), 253–264. https://doi.org/10.17730/humo.63.3.h1phxbhrb7m4mlv0

Braitstein

Tyndall

Spittal

O’Shaughnessy

M. V.

Schilder

, … Schechter

M. T.

(2003). Sexual violence among a cohort of injection drug users. Social Science & Medicine, 57(3), 561–569. https://doi.org/10.1016/S0277-9536(02)00403-3

Bukten

Stavseth

M. R.

Skurtveit

Tverdal

Strang

Clausen

(2017). High risk of overdose death following release from prison: Variations in mortality during a 15-year observation period. Addiction, 112, 1432–1439. https://doi.org/10.1111/add.13803

10.

Campbell

Dworkin

Cabral

(2009). An ecological model of the impact of sexual assault on women’s mental health. Trauma, Violence, & Abuse, 10(3), 225–246. https://doi.org/10.1177/1524838009334456

11.

Caudarella

Dong

Milloy

M. J.

Kerr

Wood

Hayashi

(2016). Non-fatal overdose as a risk factor for subsequent fatal overdose among people who inject drugs. Drug and Alcohol Dependence, 162, 51–55. https://doi.org/10.1016/j.drugalcdep.2016.02.024

12.

Charak

Byllesby

B. M.

Roley

M. E.

Claycomb

M. A.

Durham

T. A.

Ross

, … Elhai

J. D.

(2016). Latent classes of childhood poly-victimization and associations with suicidal behavior among adult trauma victims: Moderating role of anger. Child Abuse & Neglect, 62, 19–28. https://doi.org/10.1016/j.chiabu.2016.10.010

13.

Colledge

Peacock

Leung

Larney

Grebely

Hickman

, … Degenhardt

(2019). The prevalence of non-fatal overdose among people who inject drugs: A multi-stage systematic review and meta-analysis. International Journal of Drug Policy, 73, 172–184. https://doi.org/10.1016/j.drugpo.2019.07.030

14.

Darke

(1998). Self-report among injecting drug users: A review. Drug and Alcohol Dependence, 51, 253–263. https://doi.org/10.1016/S0376-8716(98)00028-3

15.

Darke

Williamson

Ross

Mills

K. L.

Havard

Teesson

(2007). Patterns of nonfatal heroin overdose over a 3-year period: Findings from the Australian treatment outcome study. Journal of Urban Health, 84(2), 283–291. https://doi.org/10.1007/s11524-006-9156-0

16.

Department of Public Health, City of Philadelphia. (2020). Opioid misuse and overdose report. https://www.phila.gov/documents/city-of-philadelphia-opioid-related-reports/

17.

Dutton

M. A.

Mitchell

Haywood

(1996). The emergency department as a violence prevention center. Journal of the American Medical Women’s Association, 51(3), 92–95.

18.

El-Bassel

Gilbert

Chang

Gomes

Vinocur

Spevack

(2007). Intimate partner violence prevalence and HIV risks among women receiving care in emergency departments: Implications for IPV and HIV screening. Emergency Medicine Journal, 24(4), 255–259. https://doi.org/10.1136/emj.2006.041541

19.

Epele

M. E.

(2002). Gender, violence and HIV: Women’s survival in the streets. Culture, Medicine and Psychiatry, 26(1), 33–54. https://doi.org/10.1023/A:1015237130328

20.

Fairbanks

R. P.

(2011). The politics of urban informality in Philadelphia’s recovery house movement. Urban Studies, 48(12), 2555–2570. https://doi.org/10.1177/0042098011411943

21.

Finkelhor

Ormrod

R. K.

Turner

H. A.

(2007). Poly-victimization: A neglected component in child victimization. Child Abuse & Neglect, 31(1), 7–26. https://doi.org/10.1016/j.chiabu.2006.06.008

22.

Ford

J. D.

Elhai

J. D.

Connor

D. F.

Frueh

B. C.

(2010). Poly-victimization and risk of posttraumatic, depressive, and substance use disorders and involvement in delinquency in a national sample of adolescents. Journal of Adolescent Health, 46(6), 545–552. https://doi.org/10.1016/j.jadohealth.2009.11.212

23.

Gilbert

El-Bassel

Rajah

Foleno

Frye

(2001). Linking drug-related activities with experiences of partner violence: A focus group study of women in methadone treatment. Violence and Victims, 16(5), 517–536.

24.

Hall

M. T.

Golder

Higgins

G. E.

Logan

T. K.

(2016). Nonmedical prescription opioid use among victimized women on probation and parole. Addictive Behaviors, 53, 113–119. https://doi.org/10.1016/j.addbeh.2015.10.008

25.

Havens

J. R.

Oser

C. B.

Knudsen

H. K.

Lofwall

Stoops

W. W.

Walsh

S. L.

, … Kral

A. H.

(2011). Individual and network factors associated with non-fatal overdose among rural Appalachian drug users. Drug and Alcohol Dependence, 115(1), 107–112. https://doi.org/10.1016/j.drugalcdep.2010.11.003

26.

Hayashi

Dong

Marshall

B. D.

Milloy

M. J.

Montaner

J. S.

Wood

Kerr

(2016). Sex-based differences in rates, causes, and predictors of death among injection drug users in Vancouver, Canada. American Journal of Epidemiology, 183(6), 544–552. https://doi.org/10.1093/aje/kwv207

27.

Heron

S. L.

Thompson

M. P.

Jackson

Kaslow

N. J.

(2003). Do responses to an intimate partner violence screen predict scores on a comprehensive measure of intimate partner violence in low-income black women? Annals of Emergency Medicine, 42(4), 483–491. https://doi.org/10.1067/S0196-0644(03)00718-2

28.

Hser

Y. I.

Kagihara

Huang

Evans

Messina

(2012). Mortality among substance-using mothers in California: A 10-year prospective study. Addiction, 107(1), 215–222. https://doi.org/10.1111/j.1360-0443.2011.03613.x

29.

Jarlenski

Barry

C. L.

Gollust

Graves

A. J.

Kennedy-Hendricks

Kozhimannil

(2017). Polysubstance use among US women of reproductive age who use opioids for nonmedical reasons. American Journal of Public Health, 107(8), 1308–1310. https://doi.org/10.2105/AJPH.2017.303825

30.

Jessell

Mateu-Gelabert

Guarino

Vakharia

S. P.

Syckes

Goodbody

, … Friedman

(2017). Sexual violence in the context of drug use among young adult opioid users in New York City. Journal of Interpersonal Violence, 32(19), 2929–2954. https://doi.org/10.1177/0886260515596334

31.

Johnson

D. R.

Lubin

(2015). Principles and techniques of trauma-centered psychotherapy. American Psychiatric Publishing.

32.

Kerr

Fairbairn

Tyndall

Marsh

Montaner

Wood

(2007). Predictors of non-fatal overdose among a cohort of polysubstance-using injection drug users. Drug and Alcohol Dependence, 87(1), 39–45. https://doi.org/10.1016/j.drugalcdep.2006.07.009

33.

Kinner

S. A.

Milloy

M. J.

Wood

Zhang

Kerr

(2012). Incidence and risk factors for non-fatal overdose among a cohort of recently incarcerated illicit drug users. Addictive Behaviors, 37(6), 691–696. https://doi.org/10.1016/j.addbeh.2012.01.019

34.

Kokkevi

Richardson

Palermou

Leventakou

(1997). Reliability of drug dependents’ self-reports. Drug and Alcohol Dependence, 45, 55–61. https://doi.org/10.1016/S0376-8716(97)01339-2

35.

Lau

A. S.

Leeb

R. T.

English

Graham

J. C.

Briggs

E. C.

Brody

K. E.

Marshall

J. M.

(2005). What’s in a name? A comparison of methods for classifying predominant type of maltreatment. Child Abuse & Neglect, 29(5), 533–551. https://doi.org/10.1016/j.chiabu.2003.05.005

36.

Lunze

Raj

Cheng

D. M.

Quinn

E. K.

Lunze

F. I.

Liebschutz

J. M.

, … Samet

J. H.

(2016). Sexual violence from police and HIV risk behaviours among HIV-positive women who inject drugs in St. Petersburg, Russia—A mixed methods study. Journal of the International AIDS Society, 19(4, Suppl. 3), Article 20877. https://doi.org/10.7448/IAS.19.4.20877

37.

Maloney

Degenhardt

Darke

Nelson

E. C.

(2009). Are non-fatal opioid overdoses misclassified suicide attempts? Comparing the associated correlates. Addictive Behaviors, 34(9), 723–729. https://doi.org/10.1016/j.addbeh.2009.04.011

38.

Marshall

B. D.

Fairbairn

Wood

Kerr

(2008). Physical violence among a prospective cohort of injection drug users: A gender-focused approach. Drug and Alcohol Dependence, 97(3), 237–246. https://doi.org/10.1016/j.drugalcdep.2008.03.028

39.

Miller

C. L.

Kerr

Strathdee

S. A.

Wood

(2007). Factors associated with premature mortality among young injection drug users in Vancouver. Harm Reduction Journal, 4(1), Article 1. https://doi.org/10.1186/1477-7517-4-1

40.

Milloy

M. J.

Kerr

Mathias

Zhang

Montaner

J. S.

Tyndall

Wood

(2008). Non-fatal overdose among a cohort of active injection drug users recruited from a supervised injection facility. The American Journal of Drug and Alcohol Abuse, 34(4), 499–509. https://doi.org/10.1080/00952990802122457

41.

Mitra

Wood

Nguyen

Kerr

DeBeck

(2015). Drug use patterns predict risk of non-fatal overdose among street-involved youth in a Canadian setting. Drug and Alcohol Dependence, 153, 135–139. https://doi.org/10.1016/j.drugalcdep.2015.05.035

42.

Parks

S. E.

Kim

K. H.

Day

N. L.

Garza

M. A.

Larkby

C. A.

(2011). Lifetime self-reported victimization among low-income, urban women: The relationship between childhood maltreatment and adult violent victimization. Journal of Interpersonal Violence, 26(6), 1111–1128. https://doi.org/10.1177/0886260510368158

43.

Philadelphia Department of Public Health. (2018). Fatal drug overdoses in Philadelphia, 2017. CHART, 3(1). https://www.phila.gov/media/20181106124514/chart-v3e1.pdf

44.

Price-Robertson

Higgins

Vassallo

(2013). Multi-type maltreatment and polyvictimisation: A comparison of two research frameworks. Family Matters, 93, 84–98.

45.

Radatz

D. L.

Wright

E. M.

(2017). Does polyvictimization affect incarcerated and non-incarcerated adult women differently? An exploration into internalizing problems. Journal of Interpersonal Violence, 32(9), 1379–1400. https://doi.org/10.1177/0886260515588921

46.

Richmond

J. M.

Elliott

A. N.

Pierce

T. W.

Aspelmeier

J. E.

Alexander

A. A.

(2009). Polyvictimization, childhood victimization, and psychological distress in college women. Child Maltreatment, 14(2), 127–147. https://doi.org/10.1177/1077559508326357

47.

Rudd

R. A.

Aleshire

Zibbell

J. E.

Matthew Gladden

(2016). Increases in drug and opioid overdose deaths—United States, 2000–2014. American Journal of Transplantation, 16(4), 1323–1327. https://doi.org/10.1111/ajt.13776

48.

Salisbury

E. J.

Van Voorhis

(2009). Gendered pathways: A quantitative investigation of women probationers’ paths to incarceration. Criminal Justice and Behavior, 36(6), 541–566. https://doi.org/10.1177/0093854809334076

49.

Schneider

Burnette

M. L.

Ilgen

M. A.

Timko

(2009). Prevalence and correlates of intimate partner violence victimization among men and women entering substance use disorder treatment. Violence and Victims, 24(6), 744–756. https://doi.org/10.1891/0886-6708.24.6.744

50.

Seal

K. H.

Kral

A. H.

Gee

Moore

L. D.

Bluthenthal

R. N.

Lorvick

Edlin

B. R.

(2001). Predictors and prevention of nonfatal overdose among street-recruited injection heroin users in the San Francisco Bay Area, 1998–1999. American Journal of Public Health, 91(11), 1842–1846. https://doi.org/10.2105/AJPH.91.11.1842

51.

Shannon

Kerr

Strathdee

S. A.

Shoveller

Montaner

J. S.

Tyndall

M. W.

(2009). Prevalence and structural correlates of gender based violence among a prospective cohort of female sex workers. BMJ, 339, Article b2939. https://doi.org/10.1136/bmj.b2939

52.

Stein, M. D., Conti, M. T., Kenney, S., Anderson, B. J., Flori, J. N., Risi, M. M., & Bailey, G. L. (2017). Adverse childhood experience effects on opioid use initiation, injection drug use, and overdose among persons with opioid use disorder. Drug and alcohol dependence, 179, 325--329.

53.

Smith

S. G.

Zhang

Basile

K. C.

Merrick

M. T.

Wang

Kresnow

M. J.

Chen

(2018). National Intimate Partner and Sexual Violence Survey: 2015 data brief. National Center for Injury Prevention and Control, Centers for Disease Control and Prevention. https://www.cdc.gov/violenceprevention/datasources/nisvs/2015NISVSdatabrief.html

54.

Tolin

D. F.

Foa

E. B.

(2006). Sex differences in trauma and posttraumatic stress disorder: A quantitative review of 25 years of research. Psychological Bulletin, 132(6), 959–992.

55.

Tripodi

S. J.

Pettus-Davis

(2013). Histories of childhood victimization and subsequent mental health problems, substance use, and sexual victimization for a sample of incarcerated women in the US. International Journal of Law and Psychiatry, 36(1), 30–40. https://doi.org/10.1016/j.ijlp.2012.11.005

56.

Ulirsch

J. C.

Ballina

L. E.

Soward

A. C.

Rossi

Hauda

Holbrook

, … Goodman

(2014). Pain and somatic symptoms are sequelae of sexual assault: Results of a prospective longitudinal study. European Journal of Pain, 18(4), 559–566. https://doi.org/10.1002/j.1532-2149.2013.00395.x

57.

Wilson

Kariisa

Seth

Smith

, IV, & Davis

N. L.

(2020). Drug and opioid-involved overdose deaths—United States, 2017–2018. Morbidity and Mortality Weekly Report, 69, 290–297. https://doi.org/10.15585/mmwr.mm6911a4

58.

Wong

L. H.

Shumway

Flentje

Riley

E. D.

(2016). Multiple types of childhood and adult violence among homeless and unstably housed women in San Francisco. Violence and Victims, 31(6), 1171–1182. https://doi.org/10.1891/0886-6708.VV-D-15-00132