Although the prevalence of women with substance use disorders (SUDs) is increasing worldwide, and the differences between the rates of SUDs among men and women are narrowing (Keyes et al., 2008; SAMHSA, 2018; Seedat et al., 2009; Steingrimsson et al., 2012), the research on SUDs has immoderately focused on men. In a call for action, the following statement was heard in June 2017 at the tenth annual meeting of the InWomen’s organization (International Women’s and Children’s Health and Gender Group): “although significant advances have been made in the field to date, gender-based issues for women remain a neglected area in much of substance use research” (Meyer et al., 2019, p. 158). In the current research, we aim to address one gap in the research on women with SUDs — the possible intermediate processes that facilitate the high co-occurrence between SUDs and compulsive sexual behavior disorder (CSBD; e.g., Ballester-Arnal et al., 2020). For instance, in a study comprising 588 women with a compulsive sexual behavior disorder, the comorbidity with substance use disorder and/or alcohol use was significantly high (45.9% and 40.8%, respectively; Carnes et al., 2005). Such high comorbidity complicates SUDs’ treatment outcomes, especially regarding treatment post-treatment and entry clinical outcomes (Greenfield & Grella, 2009). By exploring the intermediate processes that give rise to this co-occurrence, novel and effective treatments may be devised. Accordingly, in the current research, we will examine several candidates for the intermediate processes — emotion regulation strategies, mental health, and perceived social support.

Substance Use Disorders Among Women

To date, research on SUDs has immoderately focused on men because initial studies such as the 1992 Longitudinal Alcohol Epidemiologic Survey (Grant et al., 2004) and the National Household Survey on Drug Abuse (NHSDA; Branson et al., 1997) have revealed that men were more than three times as likely to have an alcohol-use disorder than women, and up to five times more likely to have SUDs. Although men have a higher tendency for illicit drug use, binge drinking, and alcohol use disorder than women (CBHSQ, 2016), recent research indicates that this gap is not detected among adolescent girls and boys. For example, research has revealed no gender differences among adolescents aged 12–17 in past-month illicit drug use (8.8% of boys and girls), current binge drinking (5.8% of boys and girls), and alcohol use (9.6% of boys vs. 9.9% of girls; CBHSQ, 2016). In addition, data from the National Interview Surveys from 1997 to 2014 demonstrates that this gap appears to be narrowing for older adults in the USA; for instance, whereas the rates of binge drinking remained unchanged among men over the age of 60, it increased by 3.7% annually among women (Breslow et al., 2017). On top of the increasing rates of SUDs among women, women with SUDs regularly report more significant functional impairment in domains such as medical and psychiatric functioning, employment, family and social relationships, and poorer overall quality of life than men (Foster et al., 2016; Griffin et al., 2015; Hernandez-Avila et al., 2004; McHugh et al., 2013; Sherman et al., 2017; Wu et al., 2010). Women also have specific struggles such as handling SUDs during pregnancy (Meyer et al., 2019).

Aside from the more significant negative consequences for women with SUD relative to men, there are also differences between men and women in the rates of psychiatric disorders among those with SUDs, with higher rates of depression, anxiety, and specific personality disorders such as avoidant, dependent, and paranoid disorders in women as compared with men (Conway et al., 2006; Khan et al., 2013). For example, using the 2001 to 2002 National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) in the USA, it was found that 33.09% and 13.54% of women with alcohol dependence had a major depressive disorder and/or generalized anxiety disorder, respectively, as compared with 16.79% and 6.49% among men with alcohol dependence (Khan et al., 2013). In addition, 7.44%, 1.74%, and 13.30% of women with alcohol dependence had avoidant, dependent, and paranoid personality disorders, respectively, compared with 5.23%, 1.46%, and 9.80% among men with alcohol dependence (Khan et al., 2013). One common comorbidity associated with SUDs is CSBD (e.g., Ballester-Arnal et al., 2020).

Substance Use Disorders and the Co-occurrence of Compulsive Sexual Behavior Disorder

CSBD is a disorder that characterizes between 3 and 10% of adults (Carnes et al., 2010; Coleman, 1992; Dickenson et al., 2018) and 12–18% of adolescents (Efrati & Dannon, 2018). Recently, Efrati and Mikulincer (2018) identified four facets of CSBD that are in keeping with the definition of ICD-11 and that manifest both among adults (e.g., Efrati & Gola, 2018, 2019a; Efrati et al., 2019) and adolescents (e.g., Efrati & Gola, 2019b; Efrati et al., 2020, 2021a, 2021b): (a) lack of behavioral control — relentless frenzied engagement with sexual fantasies, urges, and behaviors with numerous unsuccessful attempts to reduce repetitive sexual behavior; (b) unwanted consequences because of sexual fantasies — how sexual fantasies, urges, and behaviors carry harm to oneself (Reid et al., 2012) and/or to one’s close others such as family members (Reid et al., 2010), colleagues, and peers (Reid et al., 2011a, 2011b); (c) negative affect — adverse feelings and distress accompanied by guilt and shame because of sexual fantasies, urges, and behaviors; and (d) affect regulation — escape to sexual fantasies, pornography, and sexual behaviors because of pain, stress, and distress.

To date, only limited research has examined compulsive sexual behavior disorder among women (Kraus et al., 2021), with only a few studies establishing the link between CSBD and SUDs among women. For instance, in a survey of 588 women and 894 men who sought treatment for “sexual addiction,” the comorbidity of substance use and alcohol use disorder was high, with approximately 1 out of 2 men and/or women showing a co-occurrence (the exact values are 45.8% and 45.9% for men, and 40.1% and 40.8% for women; Carnes et al., 2005). In addition, research showed that between 12.6% and 19.6% of women with SUDs were “at-risk” of CSBD — i.e., at risk of clinically significant compulsive sexual behavior disorder (Brem et al., 2018; Deneke et al., 2015); above half (53.2%) of these women endorsed at least one concern for CSBD on the SAST-R (Sexual Addiction Screening Test–Revised) Core subscale.

Aside from the comorbidity of substance use disorder and CSBD, there are indications that women with SUDs also engage in risky sexual behaviors (Mezzich et al., 1997). Unlike CSBD, risky sexual behavior could characterize individuals with high levels of sexual interest and behavior (e.g., because of a high sex drive) who do not exhibit impaired control over their sexual behavior and significant distress or impairment in functioning (Kraus et al., 2018). A study on 125 female adolescents with SUDs and 78 controls between the ages of 14–18 years found a weak-to-moderate (r = 0.25) association between substance use and risky sexual behavior and showed that SUD is not only linked with CSBD but also with more dangerous sexual behaviors. Substance use comprised the quantity times the frequency of alcohol, cannabis, and other drug involvement, and risky sexual behavior consisted of sex without a condom, sex with a person who self-injects drugs, pregnancy, sexually transmitted disease, prostitution, and multiple partners.

Although the comorbidity of substance use disorders, compulsive sexual behavior disorder, and/or risky sexual behavior is high, it is not always clear which condition developed first. In some cases, CSBD has led to the development of a SUD. An example would be someone with long-standing CSBD who starts using alcohol excessively to cope and subsequently develops a SUD. In other instances, substance use precipitated CSBD or risky sexual behavior — such as when someone with SUD begins engaging in risky sexual behavior because of poor judgment under the effect of substance and/or to manage mounting distress temporarily. With that being said, in more traditional and/or religious countries, such as Israel, the average age of onset for SUD is commonly earlier than that of CSBD, especially for women (Efrati & Spada, 2022). Furthermore, co-occurring disorders can be bidirectional such that regardless of the temporal-causal relationship between a person’s SUD and CSBD, the two are likely to affect, and possibly intensify, one another — that is, the disorders are entangled in a vicious cycle of co-exacerbation. In addition, research has noted that the co-occurrence of substance use disorder with other disorders has dire consequences on people’s mental and physical state (Becker et al., 2017; Stone et al., 2019) and success of treatment (Krawczyk et al., 2017; McCabe et al., 2017).

The Effects of Co-occurring Disorders for Women With SUDs

Compared with people with SUDs alone, people with co-occurring disorders are more likely to be re-hospitalized. For example, in a study on more than 1.5 million recipients of Florida Medicaid, 28 percent of people with SUDs and a co-occurring disorder were re-hospitalized within 30 days of the discharge as compared with 22 percent, on average, of people with a single disorder (Becker et al., 2017). Regarding treatment, research revealed that half (48.6 percent) of adults ages 18 and older with SUD and a co-occurring disorder received no treatment at all in 2018, and only 7 percent received treatment for both conditions (McCance-Katz, 2019). Of people who did not complete treatment, 42 percent had a SUD and a co-occurring disorder, versus 36 percent without an additional disorder (Krawczyk et al., 2017). In addition, individuals with comorbidity between SUDs, CSBD, and/or risky sexual behavior are also at an elevated risk of relapse following SUD treatment if CSBD is not addressed (Schneider & Irons, 2001). Given the vicious cycle believed to entangle SUD with CSBD and/or risky sexual behavior and the consequences of this co-occurrence, it is essential to explore the intermediate processes that tie SUDs, on the one hand, with CSBD and risky sexual behavior, on the other hand. Identifying these processes may direct future research and treatment and allow the breakdown of the vicious cycle between the co-occurring disorders and the facilitation of better prognosis for women with SUD and CSBD co-occurrence. The current study will examine several candidates for these intermediate processes — emotion regulation strategies, mental health, and perceived social support.

Emotion Regulation as a Possible Intermediate Factor

Substance use may alter the naturally occurring emotion regulation processes and produce dysfunctional strategies for emotion regulation because they externally regulate emotions by pharmacologically altering one’s emotional state. Specifically, various drugs are described as euphoric — increasing positive emotions (Jaffe & Jaffe, 1989). Studies on self-administration drugs such as alcohol, methamphetamine, cocaine, and marijuana have revealed a significant increase in feeling “high” following the administration (e.g., Hart et al., 2001). In addition to increasing positive emotions, various licit and illicit drugs are known to alleviate negative emotional states such as anxiety (e.g., alcohol), sadness and depression (e.g., cocaine and amphetamines), and pain (e.g., heroin). Consistently, it has been proposed that the drug’s positive and palliative effects lead to reinforcement and increase the likelihood of future drug use (Kober et al., 2009). Ultimately, however, SUDs would cause deficient regulation of emotions as implicated in several theories on ongoing drug use and relapses, such as the relapse prevention model (Marlatt & Witkiewitz, 2005), the affective processing model (Baker et al., 2004), and the self-medication hypothesis (Khantzian, 1987). For example, in a study on 50 abstinent treatment-seeking alcohol users (with only nine women, however) and 62 social drinkers, it was found that alcohol-dependent participants had significantly more difficulties in emotion regulation (e.g., “When I’m upset, I become angry with myself for feeling that way” and “When I’m upset, I lose control over my behaviors”; Fox et al., 2008). In addition, in a study on 121 cigarette smokers, it was found that they utilized less effective emotion regulation strategies (expressive suppression over cognitive reappraisal), which implicated in more extended smoking history, more significant attentional bias to smoking cues, negative mood, and a greater urge to smoke (Fucito et al., 2010). Accordingly, SUDs often lead to unmanageable emotions and defective strategies to alleviate negative emotions and distress (Kober, 2014) and, therefore, to hunting for other means to regulate emotions. Because of dysfunctional strategies for emotion regulation, SUDs could also lead to risky decision-making (Chen et al., 2020). Given that research has noted that for some people, sexual behaviors often are the primary means of regulating distressing or undesirable emotions (Cashwell et al., 2017), and because risky decision-making may result in poor judgment and engagement in risky sexual behavior, it may be that the vicious cycle behind the comorbidity between substance use disorders, compulsive sexual behavior disorder, and/or risky sexual behavior reflects an effort to alleviate negative emotions and distress and/or poor judgment because of deficient emotion regulation strategies.

Mental Health as a Possible Intermediate Factor

A construct related to deficient emotion regulation strategies may also serve as an intermediate factor underlying the vicious cycle between substance use disorders, compulsive sexual behavior disorder, and/or risky sexual behavior in participants’ mental health. Epidemiological, cross-sectional studies have documented an elevated prevalence of substance use disorders in people with emotional disorders (i.e., mood and anxiety disorders) and found a high prevalence of emotional disorders in people with SUDs (e.g., Conway et al., 2006; Grant et al., 2004; Kessler et al., 2005). This comorbidity is present across various emotional disorders and with different primary substances (e.g., Chilcoat & Breslau, 1998; Compton et al., 2007; Conway et al., 2006; Schneier et al., 2010). Although longitudinal studies, in keeping with the “self-medication” theory, indicate that emotional disorders such as anxiety and unipolar mood disorders are associated with a later onset of SUDs but not vice versa (Wolitzky-Taylor et al., 2012), SUDs are reliably linked with increased severity of distress (e.g., Booth et al., 2010). For example, a 3-year longitudinal study on 710 drug users revealed that the greater the number of drugs used (a mean of 5.2 substances), the greater the reported distress. This latter effect was especially pronounced among women (Booth et al., 2010). One core and a repeatedly documented component of compulsive sexual behavior is the escape to sexual fantasies, pornography, and sexual behaviors because of pain, stress, and distress in the hope of alleviating these sensations (Efrati & Mikulincer, 2018). Given that this exacerbated distress needs to be regulated, people with SUDs might turn to compulsive sexual behavior and/or risky sexual behavior not only because of risky decision-making (Chen et al., 2020) but also because they seek other means to alleviate difficult emotions momentarily (e.g., sadness, shame, loneliness, boredom, or anger), distress, or painful experiences (Efrati & Mikulincer, 2018; Gola et al., 2020; Lew-Starowicz et al., 2020; Reid et al., 2008, 2011a, 2011b).

Perceived Social Support as a Possible Intermediate Factor

In addition to deficient emotion regulation strategies and elevated distress, a third candidate for a possible intermediate factor in the vicious cycle entangling SUD, compulsive sexual behavior, and risky sexual behavior, is a lack of an essential protective element because of SUDs — perceived social support from close others. Lack of social support may serve as both a precursor for the development of SUDs and the results of SUDs (Gliksberg et al., 2022). Specifically, research has indicated that compared with people of the general population, individuals with substance use disorder report less perceived social support. Research has also noted a possible explanation for this association by indicating that SUDs result in decreased social network size and diversity, which in turn were associated with lower perceived social support (Gliksberg et al., 2022). Given that high perceived social support among people with SUDs is essential for securing significant cognitive resources for recovery, such as self-efficacy (Stevens et al., 2015), and protects against the development of health problems, psychosomatic symptoms, and emotional distress over time (Newcomb & Bentler, 1988), the low levels of perceived social support of people with SUDs may leave them with fewer coping resources; scares coping resources may drive them to seek solace elsewhere, and plausibly in compulsive sexual behavior and/or risky sexual behavior.

The Current Study

In keeping with the suggested intermediate factors for the vicious cycle between SUD, compulsive sexual behavior, and risky sexual behavior, in the current study, we hypothesize that women with SUDs would show higher levels of compulsive sexual behavior disorder and riskier sexual behaviors as compared with women of the general community, and that these association would be mediated by deficient emotion regulation strategies (more expressive suppression and less cognitive reappraisal), greater distress, lower well-being, and lower levels of perceived social support. Ecologically speaking and given the vicious cycle, we believe that each construct may serve as both a precursor and the outcome of the other construct: SUD may lead to dysfunctional emotion regulation strategies, greater distress, lower well-being, and lower levels of perceived social support. In turn, these states may result in higher likelihood for CSBD and engagement in risky sexual behavior, which may further intensify the dysfunctional emotion regulation strategies and distress and lower well-being and perceived social support. Coming full circle, these latter processes may exacerbate the severity and diversity of SUDs. Because in the traditional and religious society of Israel, especially for women, SUDs tend to develop before CSBD (Efrati & Spada, 2022), we modeled the vicious cycle accordingly: SUD ⇒ dysfunctional emotion regulation strategies, distress, lower well-being, and lower perceived social support ⇒ CSBD and risky sexual behavior.

Method

Participants

The study consisted of 132 Israeli women — 62 with SUDs (drug and/or alcohol) and 70 healthy, matched controls. To sample the substance abuse group from “Dolphin” and “Elah” — two Israeli detox program for girls and women (97% response rate; n = 52 from Elah and n = 11 from Dolphin). The healthy, matched control group (n = 70) was sampled out of 105 women residing in the same regions as the women with SUDs (i.e., 66.7% response rate). Demographic details are presented in Table 1.

Table 1 Means, standard deviations, test statistics, and effect sizes for differences in study measures as a function of study group
Full size table

Measures

Individual-Based Compulsive Sexual Behavior (I-CSB; Efrati & Mikulincer, 2018)

To examine women’s CSBD, we used the I-CSB scale. The scale is a 24-item self-report questionnaire measuring four factors of CSBD: lack of control (e.g., “I find that I often prefer to engage in sexual fantasies and not do other things”), unwanted consequences (e.g., “Watching sexual content on the internet damages my relations with my family or friends”), affect regulation (e.g., “I use sexual fantasies when I feel unpleasant feelings”), and negative affect (e.g., “I feel bad after being exposed to sexual content on the internet”). Participants rated the degree that each item describes their feelings on a 7-point Likert scale ranging from 1 “not at all” to 7 “very much.” The questionnaire was successfully administered in research on non-clinical and clinical populations of Sexaholics Anonymous patients (Efrati & Gola, 2019a; Efrati & Mikulincer, 2018; Efrati et al., 2021a, 2021b). The reliability of I-CSB factors was high: α = 0.87 for lack of control, α = 0.91 for affect regulation, α = 0.86 for unwanted consequences, and α = 0.79 for negative affect. We also computed a total compulsive sexual behavior disorder score for each woman by averaging all I-CSB answers (α = 0.93).

Risky Sexual Action Tendencies (Ariely & Loewenstein, 2006)

Women’s risky sexual action tendencies were estimated by the Risky Sexual Action Tendencies (RSAT) scale comprising 19 items (translated to Hebrew by Efrati & Amichai-Hamburger, 2021). The scale consists of risky sexual action tendencies in three domains: 9-item abusive sex (α = 0.79; e.g., “I would consider adding a drug to woman’s drink to increase the change of intercourse”), 8-item domination sex (α = 0.86; e.g., “It could be fun to smack my sexual partner?”), and 2-item unsafe sex (α = 0.76; e.g., “Condoms hinder the spontaneous of sex”). Each woman reported the degree that she agreed with each action tendency on a 5-point Likert scale ranging from 1 not at all to 5 very much. We calculated the three scores of risky sexual action tendencies for each woman by averaging the relevant answers.

Mental Health Index (MHI-5; Ware et al., 1993)

To examine participants’ mental health (i.e., well-being and distress), we used the MHI-5 — a subscale of the RAND SF-36 Quality of Life Scale (Ware et al., 1993). The MHI-5 is a non-specific measure of mental health, which is used to assess well-being and the occurrence and degree of psychological distress (usually of anxiety and depression-related distress) during the past month (Lavikainen et al., 2006). Specifically, three items assess psychological distress (α = 0.83; “Have how much of the time you been a very nervous person?” [anxiety], “How much of the time have you felt so down in the dumps that nothing could cheer you up?”, and “How much of the time have you felt downhearted and blue?” [depression]) and two items assess well-being (α = 0.83; “How much of the time have you felt calm and peaceful?” and “How much of the time were you a happy person?”). Participants were asked to describe the frequency that they felt the content of each item on a 6-point scale ranging from 1 “none of the time” to 6 “all of the time.” The validity and reliability of the MHI-5 have been extensively studied in the past (e.g., Rumpf et al., 2001). In the current study, we calculated well-being and psychological distress scores by averaging the relevant items.

Perceived Social Support (MPSS; Zimet et al., 1988)

We used the Multidimensional Scale of Perceived Social Support (MPSS; Zimet et al., 1988) to examine participants’ perceived social support. The MPSS, a 12-item self-report scale, has three subscales measuring perceived social support from family (e.g., “My family really tries to help me”), friends (e.g., “I have friends with whom I can share my joys and sorrows”), and significant others (e.g., “There is a special person in life who cares about my feelings”). Participants were asked to rate their answers on a 6-point Likert scale ranging from 1 “strongly disagree” to 6 “very strongly agree.” Cronbach’s αs indicated that reliability was high: 0.94 for support from friends, 0.91 for support from significant others, and 0.92 for support from family. Accordingly, we calculated three scores of perceived social support for each participant by averaging the relevant items in each domain.

Emotion Regulation (ERQ; Gross & John, 2003)

To examine participants’ emotion regulation strategies, we used the Emotion Regulation Questionnaire (ERQ), which is a 10-item self-report questionnaire comprising two scales corresponding to two different emotion regulation strategies: cognitive reappraisal (6 items; e.g., “When I want to feel more positive emotion (such as joy or amusement), I change what I’m thinking about”) and expressive suppression (4 items; e.g., “I keep my emotions to myself”). Instructions state, “We would like to ask you some questions about your emotional life, in particular, how you control (that is, regulate and manage) your emotion.” The ten items are rated on a 7-point Likert scale ranging from 1 strongly disagree to 7 “strongly agree.” In the current study, Cronbach’s αs indicated adequate reliability: 0.69 for cognitive reappraisal and 0.71 for expressive suppression. Accordingly, we calculated two scores of emotion regulation strategies for each participant by averaging the relevant items in each domain.

Alcohol Use (The Michigan Alcoholism Screening Test (MAST); original in Selzer, (1971), and adapted to healthy participants in Philippe-Labbé, (2006)).

To confirm that women with SUDs are different in alcohol use than healthy controls, each woman completed the MAST scale — a 24-item screening test that is used to detect alcohol problems over the last year (e.g., “Can you stop drinking without a struggle after one or two drinks?”, “Do you feel guilty about your drinking?”). Each woman indicated whether each item was relevant to her situation. The current study’s reliability was high (α = 0.92). Accordingly, we calculated for each woman a total score by summing all items with a “yes” response. The cut-off of 8 was used to indicate alcohol use disorder (Horn et al., 1992; Philippe-Labbé, 2006). The specificity and sensitivity of this cutoff point are 0.92 and 0.88, respectively (Martin et al., 1990).

Drug Use (The Drug Abuse Screening Test-20 (DAST-20); Skinner, 1982; Skinner & Goldberg, 1986)

To confirm that women with SUDs are different in drug use than healthy controls, each woman completed the DAST scale — a 28-item screening test that is used to detect drug use over the last year (e.g., “Can you get through the week without using drugs?”, “Are you always able to stop using drugs when you want to?”). Each woman indicated whether each item was relevant to her situation. In the current study, the reliability was high (α = 0.91). Accordingly, we calculated for each woman a total score by summing all with a yes response. A cut-off of 6 was used to indicate drug use (Yudko et al., 2007). The specificity and sensitivity of this cutoff point are 0.83 and 0.74, respectively (Cocco & Carey, 1998).

Procedure

Women With SUDs

The inclusion criterion was known substance use; the exclusion criteria were women with a dual mental illness or an acute mental illness and women who do not speak Hebrew at a level that allows them to complete the research questionnaires. The SUD group was sampled from women in the treatment units of Dolphin and Elah. After the research assistants arrived in the treatment units, a meeting was held to explain the aim and rationale of the research. In that meeting, the research assistants explained how participants’ anonymity would be protected and addressed the right to cease participation at any point in the study. Upon the completion of the meeting, in-patients who volunteered to participate in the study signed an informed consent form and completed a battery of questionnaires, which were administered in the following order: MAST, DAST, CTQ-SF, LES, risky sexual action tendencies, I-CSB, and a socio-demographic questionnaire. The study procedure and materials (questionnaires and informed consent form) were submitted and approved by the Helsinki Commission of the Israel Ministry of Health (decision number MOH-122–2017) and Beit-Berl’s Institution Review Board (IRB).

Healthy Controls

The healthy control group consisted of a convenience sample of women who lived in the same regions as the SUD group. Women were approached by a research assistant, who explained the aim and rationale of the study and its importance. Women who volunteered to participate were asked to sign an informed consent form and complete the same questionnaires as the SUD group.

Data Analysis

To validate the differences between groups in SUDs using the MAST and DAST cut-off points, chi-square analyses were performed for the independence of measures with Fisher’s exact tests to estimate significance. Women in the control group who had suspected substance use (based on the MAST and DAST) were omitted from subsequent analyses. Next, we conducted a second set of chi-square analyses to examine differences between groups in the relevant background measures of acute physical health disorder (yes, no) and family status (married, unmarried). In addition, we conducted a series of independent sample t-tests to examine the differences between study groups in the background measures of residence quality, socioeconomic status, and years of education. Because differences in the latter three measures were observed, we controlled for their contribution in primary analyses. Our model regarding the vicious cycle between SUD, CSBD, and risky sexual behavior via the intermediate factors of dysfunctional emotion regulation strategies, distress, and lower well-being and perceived social support was appraised in the following steps: (1) to validate the association between SUD, CSBD, and risky sexual behavior, we first assessed differences between women with documented SUDs and controls in CSBD and risky sexual behavior factors by two multivariate analyses of covariance (MANCOVAs); (2) to examine which intermediate processes are associated with SUD, we assessed the differences between women with documented SUDs and controls in emotion regulation strategies (reappraisal and suppression), distress, well-being, and perceived social support (from family, friends, and significant others). Specifically, we conducted two MANCOVAs for mental health and perceived social support and ANCOVA for each emotion regulation strategy. We did not employ MANCOVA for emotion regulation strategies because of the independence of the two strategies. (3) We selected the intermediate measures that were significantly different between women with documented SUDs and controls (i.e., those examined in Step 2), and examined whether they mediate the links between the study group (women with documented SUDs and controls) and CSBD and risky sexual behavior. To do so, we conducted a multi-path mediation model using MPlus 8.8 Structural Equation Modeling (SEM) package. In this model, study group (women with documented SUDs and controls) served as the predictor, the selected intermediate factors as mediators, CSBD and risky sexual behavior as the outcome measures, and economic status, residence status, and years of education as covariates. Specifically, well-being, distress, and perceived social support from friends and significant others were observed variables. In contrast, compulsive sexual behavior and risky sexual behavior were latent variables on which their clusters were loaded (sexual-related unwanted consequences; negative affect, lack of control, and affect dysregulation for CSB; and abusive, domination, and unprotected sex for risky sexual behavior). Significance was appraised using a bias-corrected bootstrap analysis with 1000 resampling cycles; the Model fit was appraised by Root Mean Square Error of Approximation (RMSEA), Tucker Lewis Index (TLI), and Comparative Fit Index (CFI). The model is presented in Fig. 1.

Fig. 1
figure 1

Multi-path mediation model linking study group with CSB and risky sexual behavior (RSB) via distress, well-being, and perceived social support from friends and significant others

Full size image

Results

Drug and Alcohol Use

Chi-square tests revealed whereas the control group comprised 14.3% of women above the DAST cut-off point, 95.2% of the SUD group were above the DAST cut-off point (χ2(1) = 86.20, p < 0.001, φ = 0.81, relative risk [rr] = 6.66). Likewise, whereas the control group comprised 12% of women above the MAST cut-off point, 77.3% of the SUD group were above the MAST cut-off point (χ2(1) = 27.31, p < 0.001, φ = 0.63, rr = 6.44).

Do People Who Use Drugs and People Who Use Alcohol (i.e., the SUD Group) Differ in Background Measures From Healthy Controls?

Table 1 presents means, standard deviations, statistics, and effect sizes. Chi-square tests revealed that the percentage of married participants was not significantly different between the study groups (4.9% of the SUD group and 12.9% of the control group; χ2(1) = 2.46, p = 0.139, φ = 0.14) or was the percentage of physical health disorder (20.3% of the SUD group and 19.4% of the control group; χ2(1) = 0.02, p = 1.00, φ = 0.01). Independent sample t-tests indicated that women with SUDs had significantly lower economic status, lived in less expensive areas, and were less educated than the healthy controls. Age differences were not significant. Accordingly, in the primary analyses, we controlled for the contribution of economic and residence quality and years of education.

Do People Who Use Drugs and People Who Use Alcohol (i.e., the Study Group) Differ in CSBD and Risky Sexual Behavior as Compared With Healthy Controls?

Chi-square tests revealed that the I-CSB classified 14.0% of the SUD group as having CSBD as compared to 2.9% of the healthy controls (χ2(1) = 5.19, p = 0.042, φ = 0.20, rr = 4.78). The MANCOVA has indicated that women with SUDs differed in the multivariate factor of CSBD severity from healthy controls (based on four factors of lack of control, unwanted consequences, affect dysregulation, and negative affect), Pillai’s t = 0.10, F(4, 103) = 2.89, p = 0.026, η2p = 0.10. Subsequent univariate ANCOVAs revealed that the SUD group had significantly greater severity of sex-related lack of control, unwanted consequences, and affect dysregulation; no differences were observed in sexual-related negative affect (see Table 1). Finally, a second MANCOVA has indicated that women with SUDs and healthy controls also differed in the multivariate factor of risky sexual behavior (abusive, domination, and unprotected sex), Pillai’s t = 0.18, F(3, 102) = 7.54, p < 0.001, η2p = 0.18. Subsequent univariate ANCOVAs revealed that the SUD group takes part in more abusive (marginally significant, p = 0.08), domination, and unprotected sex as compared to healthy controls (see Table 1).

Do Women With SUDs and Controls Differ in Emotion Regulation Strategies, Mental Health, and Perceived Social Support?

A series of one-way analyses of covariance (ANCOVA) indicated that the groups did not differ in the prevalence of utilizing reappraisal and/or suppression as emotion regulation strategies (see Table 1). Conversely, a MANCOVA has indicated that women with SUDs and healthy controls differed in the multivariate factor of well-being and distress, Pillai’s t = 0.23, F(2, 110) = 16.49, p < 0.001, η2p = 0.23. Subsequent univariate ANCOVAs revealed that the SUD group had significantly lower well-being and higher distress as compared to healthy controls (see Table 1). A second MANCOVA also shows significant differences between the groups in the multivariate factor of perceived social support (family, friends, significant others), Pillai’s t = 0.20, F(3, 107) = 8.85, p < 0.001, η2p = 0.20. Subsequent univariate ANCOVAs revealed that the SUD group had significantly lower perceived social support from friends and significant other but not from family as compared to healthy controls (see Table 1).

Do Mental Health and Perceived Social Support Mediate the Links Between Study Group, Compulsive Sexual Behavior, and Risky Sexual Behavior?

In this section, we examined whether mental health (well-being and distress) and perceived social support (friends, significant others) mediate the links between study group (women with substance use disorder, controls), compulsive sexual behavior, and risky sexual behavior. The model had adequate fit to the observed data, χ2(53) = 133, p < 0.05, CFI = 0.95, TLI = 0.94, RMSEA = 0.07.

The model indicated that whereas the study group was higher on distress and lower on well-being and perceived support from friends and significant others as compared with the control group, only higher distress was significantly related to higher CSB severity (indirect = 0.39, 95% bias-corrected confidence interval [CIBC] 0.16, 0.76) and riskier sexual behavior (indirect = 0.30, CIBC 0.04, 0.68). Thus, higher distress significantly mediated the link between study group, CSB, and risky sexual behavior, whereas well-being and perceived social support did not.

Discussion

In the current study, we hypothesized that a vicious cycle underlines the high co-occurrence between SUDs and compulsive sexual behavior and/or risky sexual behavior among women. In keeping with research that addressed the co-occurrence of SUDs and compulsive sexual behavior (Brem et al., 2018; Deneke et al., 2015; Kraus et al., 2021) and/or risky sexual behavior (Mezzich et al., 1997), we found that women with SUDs had higher severity of the CSBD factors of sexual-related lack of control, affect dysregulation, and unwanted consequences as compared to healthy controls. Put differently, women with SUDs are engulfed in more sexual urges, fantasies, and behaviors that harm their mental state and the well-being of their social networks, such as peers, colleagues, and family members. They also have a constant uncontrolled preoccupation with sexual urges and fantasies in the hope of alleviating pain, stress, and distress.

Results also indicated that women with SUDs engaged more heavily in risky sexual behaviors — abusive, dominating, and unsafe sex — than healthy controls. One factor that could explain this finding is experiential avoidance — the unwillingness to remain in contact with distressing internal experiences and attempts to control and/or avoid distressing inner experiences (Chawla & Ostafin, 2007). People high on experiential avoidance tend to compulsively engage in problematic behaviors that induce short‐term feelings of comfort (e.g., sexual activity) to silence distressing and painful internal experiences (Hayes et al., 2006). Because of the short‐term benefit of this coping response, people are negatively reinforced to continue engaging in maladaptive behaviors at the expense of processing difficult experiences (Hayes et al., 2006). Research indicated that people with SUDs (Brem et al., 2018; Chawla & Ostafin, 2007; Stewart et al., 2002) and/or CSBD (Brem et al., 2018) have high levels of experiential avoidance.

We also predicted that in between SUD and CSBD reside intermediate factors that fuel the vicious cycle in the form of deficient emotion regulation strategies (more expressive suppression and less cognitive reappraisal), more significant distress, lower well-being, and lower levels of perceived social support. We found, in keeping with our predictions, that women with SUDs had a greater sensation of distress (and lower well-being) and lower perceived social support from their non-family social network as compared with their healthy counterparts. These results corroborate both cross-sectional and longitudinal research that SUDs produces high amount of distress (e.g., Booth et al., 2010) and is linked with lower availability and satisfaction of perceived social support (e.g., Dorard et al., 2013; Gliksberg et al., 2022). For example, Booth and colleagues (2010) found in a longitudinal study over a course of 3 years that the greater the number of substances used, the higher was people’s distress, and that this association is constant over time. Whereas Booth and colleagues (2010) studied men and women with SUDs, we concentrated solely on women and compared those with SUDs to healthy controls, revealing much higher rates of distress among women with SUDs. Regarding perceived social support, research has established a link between SUDs (e.g., the level of cannabis use in Dorard et al. (2013) and Gliksberg et al., (2022)) and smaller size and lower diversity of social networks (Gliksberg et al., 2022), overall lower perceived social support (Gliksberg et al., 2022), and lower satisfaction regarding social support (Dorard et al., 2013). To date, most studies examined a mixture of men and women and did not relate to the source of support. Here, we examined solely women and studied the potential source of support and found that less support is given, particularly from non-family members.

In contrast, we did not find that women with SUDs used less effective emotion regulation strategies than healthy controls — i.e., more expressive suppression and less cognitive reappraisal. It may be that women with SUDs had more erratic emotion regulation strategies and showed disorganized difficulties in emotion regulation that are not captured by expressive suppression and/or cognitive reappraisal. Using the measures of the current study, we are unable to appraise the likelihood of this latter proposal, and future research would be needed to examine the emotional regulation state of women with SUDs.

Our results also indicated that although women with SUDs had a greater sensation of distress (and lower well-being) and lowered perceived social support from their non-family social network relative to healthy controls, the sensation of distress was the sole mediator of the co-occurrence of SUDs, compulsive sexual behavior, and risky sexual behavior. It supports the argument that SUDs cause exacerbated distress that needs to be regulated. Because compulsive sexual behavior and risky sexual behavior are performed partly to alleviate distress momentarily, difficult emotions (e.g., sadness, shame, loneliness, boredom, or anger), and/or painful experiences (Efrati & Mikulincer, 2018; Gola et al., 2020; Lew-Starowicz et al., 2020; Reid et al., 2008, 2011a, 2011b), we may speculate that women with SUDs could turn to CSBD and risky sexual behavior in the hope to manage their distress. This process could also fuel a vicious cycle because CSBD and/or risky sexual behavior are also ineffective ways of coping with distress, and in keeping with research on experiential avoidance (e.g., Brem et al., 2018), women could turn back to SUDs. In addition, it may be that because of the low perceived social support, women with SUDs feel that they have fewer resources to co-regulate their distress (i.e., with the help and care from others) but to focus on self-regulation. In keeping with their tendencies to use external means of emotion regulation in the form of substance use, it may be reasonable to assume that in unmanageable distress, they will turn to additional external means of emotion regulation — CSBD and/or risky sexual behavior.

Discovering that unmanageable distress is at the core of the co-occurrence between substance use disorders among women, on the one hand, and CSBD and riskier sexual behavior, on the other hand, may promote new therapy methods. Novel therapeutic approaches are essential, especially for women, because only one in five people in substance use treatment is a woman, which is far less than the rates of women with SUDs (World Health Organization, 2015). This implies that gender-specific barriers to treatment access and engagement are present for women. Women might be afraid to seek help or get treatment for their SUDs and/or compulsive sexual behavior disorder because of fright of possible legal matters and social stigma if pregnant, the absence of childcare while in treatment, or because of other family obligations related to the role of women as mothers and caregivers (Stone, 2015). Focusing on the uncontrollable distress and combining that with the applicability of digital health to the assessment and treatment of substance use disorders (Marsch et al., 2020) may promote more efficient and reachable treatment for women who seek help for SUDs. One treatment schedule that could be promising, given the significant role of distress, is the somewhat novel distress tolerance treatment, Skills for Improving Distress Intolerance (SIDI; Bornovalova et al., 2012). Psychological distress tolerance is the ability to persist in goal-directed activity when experiencing psychological distress (Brown et al., 2002; Lejuez et al., 2003; Strong et al., 2003). SIDI was developed to teach people with SUD skills for tolerating distress, including increasing the ability to experience emotional distress and controlling behaviors in the context of emotional distress. SIDI was found as a promising therapy for improving psychological distress tolerance among people with SUD (Bornovalova et al., 2012). Hence, it may be considered an adjunctive treatment for women with co-occurrence of SUD, CSBD, and risky sexual behavior.

Although the current study revealed potential intermediate factors between SUDs, compulsive sexual behavior disorder, and/or risky sexual behavior among women, it has several limitations that ought to be recognized. First, the study consisted of women with diverse SUDs; it is unclear whether different SUDs share a similar vicious cycle that ties them to CSBD and risky sexual behavior. For instance, the third research aim in Meyer and colleagues’ (2019) roadmap was to “address the heterogeneity of substance use among women in both treatment and research, incorporating differences by substances used, race/ethnicity, or gender identity.” Further research is warranted to examine the possible diversity in substance use-related processes among women and the SUD-CSBD relationship. Second, although the current samples of women with SUDs and healthy controls are unique, the sample sizes are not ideal and hinder an in-depth examination of the SUD-CSBD relationship. Further studies are needed to replicate the current pattern of results in larger samples. Third, we hypothesize that a vicious cycle entangles SUD, CSBD, and risky sexual behavior via intermediate factors. Using the current cross-sectional study, we could not examine the full circle of processes but only a portion that fits well with Israel’s traditional and religious society (i.e., SUD → intermediate processes → CSBD and risky sexual behaviors). Specifically, using a cross-sectional design, it is mathematically identical to examine whether SUD is associated with CSBD and risky sexual behaviors via intermediate factors or vice versa — whether CSBD and risky sexual behavior is linked with SUD via the same intermediate factors. Longitudinal studies with at least 3 time points would be needed to examine the directionality of the suggested vicious cycle. Here, we only supported its existence but not its directionality. We speculate that longitudinal studies would indeed find support for a vicious cycle that entangles SUD with CSBD. Finally, in the current study, we only explored two specific emotion regulation strategies — expressive suppression and cognitive reappraisal; future research would benefit from exploring other venues of emotion regulation, such as specific difficulties in emotion regulation.

Despite these limitations, we view this study as a necessary step toward understanding women with SUDs — an overlooked area in SUD research (Meyer et al., 2019). This knowledge may allow us to better understand the inner world of women with substance use disorders, the comorbidity between SUDs, CSDB, and risky sexual behavior, and the factors at the core of this comorbidity. In addition, the present study may help tailor better interventions to reduce SUDs and their adverse outcomes by targeting the vicious cycle of unmanageable distress to disrupt the escapism to addictive behaviors.