Pathological Affective Dependence (PAD) as an Antecedent of Intimate Partner Violence (IPV): A Pilot Study of PAD’s Cognitive Model on a Sample of IPV Victims ()
1. Introduction
In the last decades, the theme of PAD has evolved considerably as more and more patients have consulted psychotherapists and/or psychiatrists for their dysfunctional intimate relationships (Pugliese, Saliani, & Mancini, 2019; Pugliese et al., 2023) . However, PAD is not yet embedded in psychiatric nosology although there is informal recognition of its negative mental health consequences, its key role in IPV (Kane, Staiger, & Ricciardelli, 2000) , and in the intention to return to the violent partner (Crapolicchio et al., 2021) . Moreover, there is a lack of a psychometric instrument to assess symptoms.
According to Pugliese et al. (2019, 2023) (see also Iannucci et al., 2021; Perdighe et al., 2022 ) PAD can be considered as a relational phenomenon in which a person has an apparently indispensable relationship with an abusive, violent or manipulative partner. The relationship is seemingly “indispensable” to at least one of the two partners. This bond is a source of suffering for at least one of the partners. Despite this, the suffering partners feel that they are unable to end the relationship or tolerate the fact that their partner may choose to separate. When the breakup is in sight, they feel anxious and are willing to do anything to prevent it. After the breakup, they feel desperate and/or angry and tend to think about and hold on to the relationship for a long time to cope with the emotional distress caused by the separation.
Dependence in itself is not dysfunctional but can be considered as an etiologically adaptive and appropriate expression of the prosocial need for attachment, as an engine that creates bonds and forms groups for survival (Borgioni, 2015; Bowlby, 1979; Pugliese et al., 2023) .
However, when dependence is not balanced by adequate levels of self-confidence and autonomy, dependent partners become trapped. The ongoing frustration, caused by an abusive partner, of important relationship goals such as safety, self-esteem, and love can affect physical and mental health and is an indicator of PAD.
Pugliese et al. (2019, 2023) defined PAD as a dysfunction of the relationship potentially detectable in various personalities and not necessarily pathological, according to the criteria established in the DSM 5 (APA, 2013) . In the PAD condition, basic relational needs and goals become secondary to the terminal goal of maintaining the maladaptive bond with the partner. This occurs even when it has negative and significant consequences for the victim’s well-being and safety. As noted by the authors and consistent with clinical observations (Iannucci et al., 2021; Perdighe et al., 2022) , the sine qua non condition of these relationships is the presence of a conflict between the goal of maintaining the toxic relationship to preserve the bond and the goal of separating from the partner to protect oneself. Dependency leads to an intrapsychic conflict, sometimes on an unconscious level. The decisive factor that prevents them from ending the unsatisfactory relationship is the deep ambivalence of the painful oscillation between “stay or go”.
Most IPV victims are well aware that they have an abusive partner, but for many reasons continue to repeat the maladaptive relationship patterns described. For these reasons, victims of IPV often do not seek therapeutic help until they have repeatedly failed to leave the abusive partner.
In this state, individuals swing between the opposite poles of this conflict, choosing one option and the other, but never really breaking up with their partner. They evaluate this last option as a frightening and unacceptable state and, in the event of a real or imagined break-up, they chaotically try to reduce the resulting negative emotional dysregulation in various ways, e.g. by looking for the same or a similar relationship or by waiting for a positive change in the partner’s behavior and attitude that never occurs (Pugliese et al., 2023) . Over time, the illusion of this change becomes a powerful narcotic that can freeze the relationship for a long time, with serious consequences in the form of physical and/or psychological disorders; or the relationship can degenerate into intimate partner violence (physical or psychological).
In this article, we have looked more closely at PAD, which is often considered simply a form of addiction (Sussman et al., 2011) or a symptom or feature of a dependent personality disorder (Guerreschi, 2011; Iannucci et al., 2021) . Moreover, we consider PAD as a risk factor or antecedent of Intimate partner violence (IPV). IPV is becoming a widely recognized social and public health problem (Burelomova, Gulina, & Tikhomandritskaya, 2018; Heyman, Slep, & Foran, 2015; Pagliaro, Pacilli, & Baldry, 2020) . IPV refers to harmful acts (physical, verbal/ symbolic, or sexual) directed against a person and is rooted in gender inequality, abuse of power, and harmful norms. It is a serious violation of human rights and a life-threatening health and protection issue (Breiding et al., 2015) . Different definitions such as wife battering/abuse, and domestic/family violence have been used to indicate this condition (Van Parys et al., 2014) . These definitions assume that the victims of IPV were only women and the perpetrators were only men. Instead, several studies show that both genders can be victims or perpetrators (Tjaden & Thoennes, 2000) . Moreover, terms such as domestic violence or family violence have been used to refer to specific violence that occurs only among heterosexual couples and legally married, cohabiting, or divorced couples. According to empirical research, this phenomenon is not limited only to these conditions: IPV occurs among couples regardless of their sexual orientation and the condition of living under the same roof (Johnson & Ferraro, 2000) . IPV refers to all types of violence that occur between intimate partners.
In 2014, Hoeffler and Fearon (2014) established that the social costs of IPV were vastly higher than those of wars. Indeed, violence between “intimate partners” cost the community 4.423 billion, equivalent to 5.18% of the world’s Gross Domestic Product. As summarized in a World Health Organization (WHO) report by Heise and Garcia-Moreno (2002) , IPV has significant public health implications, including deterioration in numerous markers of morbidity (e.g., physical, sexual/reproductive, psychological/behavioral) and mortality. Physical violence also has an impact on children in the household (e.g., O’Brien et al., 1994 ). Kitzmann, Gaylord, Holt, and Kenny (2003) found that children exposed to IPV scored significantly higher on internalizing and externalizing problems, social issues, academic problems, negative affect/stress, and negative cognitions compared to children not exposed to IPV. The literature has highlighted that exposure to interparental violence in childhood is associated with later perpetration of violence, victimization, and violence towards one’s children, peers, or animals (Baldry, 2003a, 2003b; Malik & Heyman, 2014; Hong, Fisher, & Espelage, 2021) . To summarize, IPV negatively affects individuals and damages both physical and mental health (Giridhar, 2012; Lagdon et al., 2014; Trabold et al., 2020) . In terms of physical health, abused women report higher levels of problems such as hypertension, chronic pain, sexually transmitted diseases and diabetes (Breiding et al., 2014; Dolezal, 2009; Mittal et al., 2013) , poor pregnancy outcomes, and higher rates of HIV infection (Devries et al., 2011) . In terms of mental health victims of IPV (compared to women who have never experienced IPV) are three to five times more likely to have problems such as post-traumatic stress disorder (Golding, 1999; Wuest et al., 2009) , depression (Golding, 1999) , anxiety (Carlson, McNutt, & Choi, 2002; Hathaway et al., 2000) , eating disorders, substance abuse disorders (Danielson et al., 1998) , sleep disorders (Breiding et al., 2014; Hathaway et al., 2000) , and suicide attempts or thoughts (Devries et al., 2011) . Finally, victims who experience more than one form of violence and are re-victimized have an increased risk of mental disorders and comorbidity of disorders (Garcia-Moreno et al., 2005) .
In 2018 one in three people died every day in Italy at the hands of a partner or former partner and approximately three million people suffered some form of violence in their lifetime. It has been estimated that the number of those who did not mention the mistreatment (abuse) was still very high (Istat data, 2018; Pagliaro et al., 2020) . During several global lockdowns for Covid-19, the threat of IPV has increased significantly (Gosangi et al., 2021; Istat, 2020) .
The urgency for scientific analysis and clarification of the antecedents of IPV is therefore evident. Among several social and psychological aspects that explain the phenomenon of IPV, the psychological characteristics of the victim play a central role: knowledge of how to quickly recognize a victim, their goals, their psychological functioning, and their intrapsychic, as well as interpersonal processes, can be considered an essential protective factor for the development of an effective social prevention strategy.
In this article, we aim to demonstrate that there is common psychological suffering and common maladaptive schemas (consisting of goals, anti-goals, and self-other beliefs) followed by people suffering from PAD.
We share the idea, confirmed by years of clinical observation and research (Kane, Staiger, & Ricciardelli, 2000; Pugliese et al., 2023) , that there is a common psychological condition suffered by IPV victims called PAD. In line with this, a study by Patsi Humérez and Requena Gonzales (2020) showed a strong correlation between maladaptive schemas and emotional dependence among women in situations of IPV. Furthermore, we emphasized the position that PAD is a democratic phenomenon that can affect both men and women of all ages, races, homosexuals, and heterosexuals, and regardless of socio-economic class and educational level (i.e., Caldwell, Swan, & Woodbrown, 2012; Messinger, 2017; National Coalition of Anti-Violence Programs, 2017; Walker et al., 2020 ).
Specifically, the aim of this article was to present the preliminary results and psychometric characteristics of a first PAD scale (PADS) tested in a pilot study on a clinical sample of IPV victims. The affective dependent profiles have been tracked, using a framework of anti-goals, self/other beliefs, and intrapsychic conflicts (absent, alternate, or akrasic conflict) as obstacles to healthy and safe separation.
The ultimate goal of the above project is to develop a definitive tool to identify the phenomenon in both clinical and non-clinical populations, considering PAD as the cognitive and emotional basis of gender-based violence. The identification of PAD profiles is useful to intervene before a cycle of violence occurs and could also help in working with psycho-educational programs (e.g., in different contexts such as schools, universities, anti-violence centers, social services, etc.). Indeed, the overall aim of all these works is to fully grasp the complexity of PAD and prevent its cost to our society.
2. Measure of PAD
The implementation of the PAD Scale (PADS) is based on the theoretical model developed from a cognitive approach to the PAD ( Pugliese et al., 2019, 2023 ; see also Iannucci et al., 2021; Perdighe et al., 2022 ) and based on the theory of goal-directed behaviors (see Miceli & Castefranchi, 1995; Miller, Galanter, & Pribram, 1960; Castelfranchi & Parisi, 1980; Weiner, 2010 ). In the following sections, we briefly describe the model and then the preliminary development of the PAD scale (PADS).
2.1. A Cognitive Model of Pathological Affective Dependence
With the reference to the cognitive model of PAD published by Pugliese et al. (2019, 2023) the following description will provide 1) the first description of a cognitive behavioral model of PAD and 2) a cognitive-behavioral profile of a typical affective dependent (TAD).
Pugliese et al. (2019, 2023) have shown that the TADs generally address the clinician with the stated aim of understanding what they want from their partner or ex-partner; why they are being mistreated by their partner psychologically, emotionally, economically, physically, and/or sexually; why they return to (or remain in) a pathological relationship despite being well aware of the distress this relationship is causing them; why they keep choosing the wrong partners.
To answer these questions, it is possible to refer to the theory of goal-directed behavior (see Miceli & Castelfranchi, 1995; Miller, Gallanter, & Pribram, 1960; Castelfranchi & Parisi, 1980; Weiner, 2010 ) and to the central role of goals in explaining the development of psychopathological symptoms.
The instrumental goal of TAD is not to break the bond with the partner. There are three-terminal goals that serve the main purpose. TADs terminal purposes are to guarantee their value, lovability, safety, and attachment by others who are unavailable, fragile, and/or mistreating (psychologically, emotionally, physically, sexually, and/or economically). They develop a real obsession with these partners. Caring partners oppress him/her while they prefer elusive or distant partners so that a partner who loves him/her or who positively responds to his/her needs is often described as “boring” (Pugliese et al., 2019, 2023) . This last aspect leads to an interpretation of the phenomenon of PAD that is far from the explanations in the literature that explain this problem as the effect of a deep need for caring, a key factor in the diagnosis of Dependent Personality Disorder (DSM-5, APA, 2013; Faith, 2009 ). If the main purpose of the TAD was to be cared for, he/she would not be bored with a partner who took care of his/her needs.
To prevent this destructive phenomenon, it is important to understand the reasons why TAD fails to get out of the relationship even though he/she is aware of the negative consequences and is stuck in the conflict. In fact, all TADs reported in the clinical sessions that they were more afraid of separating from their toxic partner than they were attracted to a stable relationship; on the contrary, they felt more attracted to problematic partners than to secure ones. They explained this pathological pattern as the result of their terrible fear of making their partner suffer, not deserving anything better, and feeling lonely/lost (Pugliese et al., 2023) . Accordingly, in this first study, we assume that TAD cannot interrupt the dysfunctional relationship for these three specific reasons. The worst-case scenario for the TAD is a relationship breakdown with the partner (we called it anti-goal 0 or AG0). The TAD will therefore defend the occurrence of the AG0 as much as possible. His/her purpose is to maintain the toxic relationship at all costs or to be loved as he/she dreams (we called it Goal 0 or G0) even if the relationship is causing great suffering to at least one of the partners and the life of one is in danger. This happens because of these three main concerns listed above, which were included in a recently published cognitive model of PAD and were called the three terminal anti-goals (Pugliese et al., 2023) . Each anti-goal (the worst-case scenario that refers to one of the main concerns) relates to a specific goal (i.e. the purpose they would achieve if they thought about the relationship and/or the partner) and to self/other beliefs (crystallized images of themselves and the other that they have built up over their life or through their childhood experiences). Anti-goal 1 (AG1): avoid the other’s suffering; this goal is to change/save someone who is emotionally fragile, violent, and unattainable (Goal 1). Anti-goal 2 (AG2): avoid losing one’s dignity; this goal is to feel worthy/seen or regain one’s worth that was lost during the dysfunctional relationship with a denigrating, humiliating/neglecting, and abusive partner (Goal 2). Anti-goal 3 (AG3): avoid feeling unsafe and alone; this goal is security and attachment. You need to feel safe and protected or regain your sense of safety that was lost during the relationship with a neglecting or physically and sexually abusive partner (Goal 3).
These anti-goals have developed from the frustration of basic needs (attachment and morality) and/or from modes of functioning learned in early relationships, and over time have given rise to separate parts of the self (and the relative image of the other), each with a specific purpose (see also Dweck (2017) ).
Holding on to a relationship despite the discomfort experienced has its origins in the unconscious desire to make up for the traumas of emotional deprivation or abuse experienced in childhood. The TAD adult may have been a child who, for the sake of his/her survival, was not allowed to lose the love of his/her parents, even if he/she was abused. It may have been a child who identified with the parent, with the mission of love and sacrifices towards the problematic partner, or a child dedicated to loving and saving the problematic parent who was despised and neglected by the partner. In other words, the role played by both parents seems to be crucial to the dynamics of the psychological development of the future TAD. There are three-terminal goals related to the following self-other beliefs: Altruistic problematic self/fragile other (Goal 1); Deontological self-humiliating other (Goal 2); Vulnerable/Traumatic self-emotionally unavailable/abusing other (Goal 3).
In summary, then the distinctive feature of this dysfunctional relationship is that the TAD finds it difficult to end the maladaptive relationship. They choose to stay in the relationship even though their psychological and physical well-being deteriorates on a personal, social, and professional level. At this moment, the TAD experiences a strong conflict between the purpose of maintaining the maladaptive relationship and the purpose of separating from the abusive partner.
As soon as the disadvantages of the pathological relationship become apparent, the TAD begins to feel the conflict between the will to stay in the relationship and the will to end it. If they choose the healthy direction (to get out of the maladaptive relationship) they risk realizing the anti-goal and therefore must quit. The TAD remains stuck in an impasse, oscillating between the two poles of the conflict. According to Pugliese et al. (2019, 2023) , there are three main intrapsychic conflicts: absent, alternate, and akrasic.
The first conflict is absent: the person does not feel this conflict, it is only seen by others (e.g. a family member, friend or therapist, etc.). PAD is therefore ego-syntonic and the person chooses not to separate because, according to his/her values, he/she considers the benefits of staying in the relationship as greater than the costs.
The second type of conflict can be defined as alternate. In this type of conflict, the person oscillates between different mental states: the same event or the possibility of a break-up, for example, is valued positively when it is about returning to feeling good by ending the relationship, and negatively when it is about something else, such as the loss of an exclusive guide or the partner’s (abusive) love. Because mental states are activated alternatively, they move from one to the other without being aware of it. For example, they may advocate separation on the one hand and do everything to avoid it on the other, without integrating these two states into a coherent pattern. The result is separate mental states and behaviors. To end the suffering, the person may threaten separation or decide to separate from the partner, but as soon as the pain of loss and emptiness is perceived, they fill the feeling of emptiness holding on to the pathological relationship.
The third conflict is characterized by a state of Akrasia. In this conflict, the TAD is ego-dystonic and there is an oscillation within the same mental state due to the overlap of two very different goals, such as the desire to separate due to the perceived discomfort and the will to stay in the relationship because they do not want to be alone.
In line with the described cognitive model of PAD (Pugliese et al., 2019, 2023) , we can hypothesize that there are four main factors, which can describe the mental functioning of TAD. These specific TAD factors are Altruistic, Deontological, Vulnerability, and Conflict.
As you can see in Table 1, the first three factors refer to the above-mentioned three main anti-goals, goals and self, and other beliefs. The fourth factor describes three conflicts (absent, alternate, and akrasic). Moreover, Pugliese et al. (2023) hypothesized that there are more types of TADs. All of them shareda are common fear of ending the relationship and the three conflicts. The authors described four prototypical profiles, consistently with the hyper-invested anti-goal: Saver (Anti-goal 1), Unworthy (Anti-goal 2), Traumatic (Anti-goal 3), and Mixed (all anti-goals or two of them)
Although several scales have been developed to assess constructs related to PAD constructs such as love addiction or emotional dependence, they are now outdated and/or lack a theoretical basis. For example, the Love Addiction Scale (LAS, Hunter et al. (1981) ) which was later superseded by the Passionate Love Scale (PLS) of Hatfieldand Sprecher (1986) and the more recent scale of Feeney and Noller (1990) . However, none of these scales is widely used and they do not assess the core characteristics of TAD as defined by the cognitive model of PAD. Other popular scales are available online that assess PAD similar symptoms such as the love addiction quiz (Gaba, 2018) , the love addiction self-assessment (Falango, 2012) , and the 40 questions of Sex and Love addicts Anonymous (Augustine Fellowship, 1985) .
However, despite their ease of use, these scales have not been published in peer-reviewed journals, do not include information on psychometric characteristics, and no theoretical underpinnings or constructs are presented. Given these many limitations of the existing instruments, a psychometrically robust scale is needed that assesses PAD using a strong theoretical framework that is consistent with recent developments in clinical psychology. Consequently, the present study aimed to fill this gap, and develop a preliminary version of a psychometrically robust scale for assessing PAD using a strong theoretical framework (theory goal-directed behavior, see Miceli & Castefranchi, 1995; Miller, Gallanter, & Pribram, 1960; Castelfranchi & Parisi, 1980; Weiner, 2010 ) and is consistent with years of clinical observations. Using these four factors to understand the psychological functioning of TAD would likely lead to more reliable decision-making in this area and more consistency across studies.
2.2. Pathological Affective Dependence Scale (PADS)
The aim of the paper was to present the preliminary results and psychometric characteristics of a measure of PAD. The PAD scale (PADS) was developed by creating 27 items that were applied to each dimension of the PAD theoretical model. Specifically, some items of PADS were adaptations of items that related to already-existing scales (all scales were validated in the Italian context) measuring similar factors1. Other items were created ad hoc based on clinical observations conducted by experts in the field. As highlighted in the above-described cognitive model of PAD, for each factor, PADS explores three sub-dimensions that better explain the phenomenon of PAD: beliefs about the self, beliefs about the other (partner), and main anti-goals. Below is a description of the subdimensions (and an example of an item) for each of the four PADS factors hypothesized:
Factor 1: Altruistic
• Beliefs about the altruistic self: e.g. “I feel responsible for my partner’s happiness/well-being”;
• Anti Goal about the altruistic self: e.g. “If I leave my partner, he/she would suffer too much”;
• Beliefs about the fragile other: e.g. “I think my partner needs me”.
Factor 2: Deontological
• Beliefs about the unworthy self: e.g. “There is something wrong with me”; item from Inadequacy Trap Questionnaire (Young, Klosko, & Weishaar, 2003) ;
• Anti-Goal about the unworthy self: e.g. “I would not leave my partner because I do not believe I deserve a better man or woman”. It is based on item 15 of the Pathogenic Beliefs Scale (PBS, Curtis & Silberschatz (2005) ).
Factor 3: Vulnerability
• Beliefs about the vulnerable self: e.g. “I believe I need my partner to make me feel safe”;
• Anti-Goal about the vulnerable self: e.g. “If I left my partner, I would feel lonely and lost”;
• Beliefs about the performing other: e.g. “I believe that if I disagree with my partner(s), it will lead to dismissive, angry, and rejecting reactions toward me”;
• Beliefs about the emotionally unavailable other: e.g. “I believe my partner is cold and incapable of meeting my needs”.
Factor 4: Conflict
• Absent Conflict: e.g. “Others point out to me that I often change my goals, plans, beliefs, etc… since I have been with my partner(s) but I do not realize it”. It is based on item 95 of Scid-II (First et al., 1995) ;
• Alternate Conflict: e.g. “I feel satisfied with my relationship at certain times while at other times I consider ending the relationship permanently”;
• Akrasic conflict: e.g. “I am aware that I need more love than I receive, but at the same time I cannot give up on my partner”. It is based on item 1 of the Emotional deprivation traps questionnaire (Young & Klosko, Weishaar, 2003) .
For each item, the current state (whether it occurred in the present moment) and the habitual relationship trait (whether it was typical for the person when in a relationship) were assessed.
For both trait and state, participants rated their responses on a 5-point Likert scale: 1 = not at all, 2 = slightly, 3 = sometimes, 4 = often, and 5 = always. Higher scores indicated that the cognitive components of the theoretical model of PAD were more prominent. The original items (in Italian) are listed in Appendices 1 and 2, along with the respective factors and the associated descriptive and reliability statistics.
To advance research on PAD and to enable further theoretical and scientific refinement, it is essential to further investigate the construct validity of the PADS. Construct validity is the extent to which a measurement assesses the theoretical construct and constitutes the starting point for evaluating any measurement (Chmielewski et al., 2016; Clark & Watson, 1995) . In addition to measurement-issues, construct validity provides substantial information about the construct of interest itself. Without initial evidence supporting the construct validity of this new measurement and a clear understanding of the construct and how it is assessed, advances in theories of PAD and prevention and implementation efforts may be hampered. For example, inadequate construct validity makes it difficult to properly test interventions and treatments, identify new treatment targets, or plan educational programs. To address this aim we assessed the internal consistency, construct validity, and convergent and discriminant validity of the PADS. We evaluated Pearson r correlations both between the subscales of the PADS and between each of them and the subscales of some selected scales (measures of depression, self-esteem, perceived relationship qualities, maladaptive attachment styles, moral orientation, dysfunctional personality traits and conflict with a partner) to test convergent and discriminant validity.
In this regard, we hypothesized that the subscales of PADS would be positively correlated with dysfunctional personality traits relative to the dominance/submission domains, such as Negative affect, Detachment, Antagonism, Disinhibition, and Psychoticism; Depression; Conflict with partners components (Negotiation, Psychological Aggression, Physical Assault, Sexual Coercion and Injury); Different types of Sense of guilt; Anxious Attachment Style. Moreover, we hypothesized that the subscales of PADS would be negatively correlated with self-esteem; Perceived relationship quality component subscales (Relationship Satisfaction, Commitment, Intimacy, Trust, Passion, and Love). Finally, we hypothesized that the subscales of PADS would not be (or weakly, i.e. Pearson’s r < .30) correlated with the Avoidant attachment style.
3. The Pilot Study for the Preliminary PADS Development
A pilot study was conducted to present the psychometric properties of the preliminary PADS. In this section, we present the method and results of the study.
3.1. Participants
A total of 25 participants (F = 21.84%; M = 4.16%; 56% not married, 20% separated/divorced, 4% waiting for separation, 16% married, 4% widowed; 76% heterosexual, 12% homosexual and 12% bisexual), recruited specifically for the pilot study, completed an online questionnaire that included the newly developed 27 items questionnaire, as well as measures useful for testing the convergent and discriminant validity of the PADS. Participants’ ages ranged from 29 to 61 years (M = 41.32; sd = 9.88). Participants presented symptoms characteristic of the cognitive model of PAD. Data were collected with the involvement of anti-violence centers, the Mutual Help Group of Millemé (an online group on Facebook that includes victims and survivors of IPV, both psychological and physical; https://www.milleme.it/), supervised and supported by psychotherapists and lawyers (each specializing in relational trauma and penal law), and patients currently undergoing private psychotherapy studies (https://apc.it/). According to the specific characteristics of PAD (Pugliese et al. 2019, 2023) , only patients with experience of maladaptive relationships were selected. Specifically, all participants were in the three conflict conditions (assent, alternate and akrasic). Finally, they reported being (or having been in the past) in a relationship with an abusive partner.
3.2. Procedures
Questionnaires were completed online and anonymously. The study was conducted in accordance with the ethical standards of the Declaration of Helsinki, and the completion of the survey took approximately 20 minutes. Informed consent was obtained from all subjects participating in the study (Pr. 2/21).
3.3. Measures
Demographic information: Demographic information was requested including specific questions concerning gender, age, education level, relationship status, and the length of time in a relationship.
PAD scale: the PADS is a 27-items self-report questionnaire that measures both state and trait conditions of PAD for the four theoretical factors (see the theoretical model of PAD mentioned above): altruistic factor, trait and state vulnerability factor, trait and state deontological factor, trait and state conflict factor (α indexes in Appendix 1 and Appendix 2).
The Revised Conflict Tactic Scale (CTS-2, Straus et al. (1996) ): the CTS-2 is a 39-items scale that measures both how much each partner engages in psychological and physical attacks on each other and also their use of reasoning or negotiation to deal with conflicts. Items are divided into five categories: “Negotiation” (α = .69), “Psychological Aggression” (α = .80), “Physical Assault” (α = .92), “Sexual Coercion” (α = .81), and “Injury” (α = .86). Each of the five categories is then further subdivided into two subscales for a total of 78 questions.
The Perceived Relationship Quality Component Inventory (PRQC; Fletcher et al. (2000) ): the PRQC was developed to assess multiple aspects of relationship quality (Fletcher et al., 2000) , and it is a widely used 18-item self-report inventory consisting of six subscales: Relationship Satisfaction (α = .59), Commitment (α = .41), Intimacy (α = .60), Trust (α = .65), Passion (α = .64) and Love (α = .55). Participants rated their responses according to a 7-point Likert scale ranging from 1 (not at all) to 7 (extremely).
The Experience in Close Relationship (ECR; Picardi et al. (2002) ; Italian versione, Brugnera et al. (2019) ): the ECR is a 36-item self-report adult attachment style questionnaire focused on close relationships. This scale measures maladaptive attachment in adults who are in a romantic relationship. The ECR measures individuals on two subscales of attachment: Avoidance (α = .71) and Anxiety (α = .65). The 7-point Likert scale ranges from 1 = strongly disagree to 7 = strongly agree.
Rosenberg Self-Esteem Scale (RSES; Rosenberg (1965) ; Italian version Prezza et al. (1997) ): the RSES is a self-assessment test to measure self-esteem (α = .93). The scale consists of ten statements you should apply to yourself and estimates to which extent you agree with each of them. Participants respond to questions on a 4-points Likert scale ranging from 0 = strongly disagree to 3 = strongly agree.
Moral Orientation Guilt Scale (MOGS; Mancini et al. (2022) ): the MOGS is a 17-items scale to assess the different types of guilt referring to 4 domains: Dirtiness (α = .84), Moral Norm Violation (α = .89), Empathy (α = .69) and Harm (α = .85). Dirtiness and Moral Norm Violation scores can be added together to obtain the Deontological guilt total score (α = .90), while Empathy and Harm scores can be summed to obtain Altruistic guilt total score (α = .90). Participants responded to the questions on a 5-points Likert scale ranging from 1 = (not at all) to 5 = (extremely).
The Beck Depression Inventory-Second Edition (BDI-II; Beck et al. (1996) ; Italian version, Ghisi et al. (2006) ): the BDI-II is an improvement and renewal of the first edition of the Beck Depression Inventory (BDI, Beck et al., 1961 ) and was developed to measure the severity of depression in adolescents and adults (α = .94). The BDI-II consists of 21 multiple-choice self-report questions. Raw scores from 0 to 13 indicate minimal depression, 14 to 19 indicate mild depression, 20 to 28 indicate moderate depression and 29 to 63 indicate severe depression. The BDI-II was developed to assess symptoms consistent with the criteria for the diagnosis of depressive disorders listed in the DSM-IV. In this version, four items were deleted (weight loss, change in body image, somatic rumination, and work difficulties) and four new items were added (restlessness, worthlessness, difficulty concentrating, and loss of energy) to capture symptoms typical of major depression.
The Personality Inventory of the DSM-5, PID-5-Adults (Fossati et al., 2013) : The DSM-5 Personality Inventory, PID-5-Adults, is a self-report questionnaire consisting of 220 items answered on a 4-point Likert scale ranging from 0 (strongly disagree) to 3 (strongly agree). The scale is intended for adults 18 years of age and older and assesses 25 facets of personality traits: Anhedonia, Anxiousness, Attention seeking, Callousness, Deceitfulness, Depressiveness, Distractibility, Eccentricity, Emotional Lability, Grandiosity, Hostility, Impulsivity, Intimacy Avoidance, Irresponsibility, Manipulability, Perceptual Dysregulation, Perseveration, Restricted Affectivity, Rigid Perfectionism, Risk Taking, Separation Insecurity, Submissiveness, Distrust, Unusual Beliefs and Experiences, and Withdrawal. Certain triplets of facets (groups of 3), can be combined to assess the five domains of traits: Negative affect (α = .72), Detachment (α = .67), Antagonism (α = .78), Disinhibition (α = .81), and Psychoticism (α = .81).
3.4. Data Analysis
Data were analyzed with SPSS 25.0. Descriptive statistics, reliability analysis (based on analysis of Cronbach’s alpha), inter-correlations (based on Pearson’s r) between items of the scale, and correlation analysis (Pearson’s r) were computed for the study variables.
3.5. Results
Table 2(a) and Table 2(b) report means, standard deviations, skewness, kurtosis, and intercorrelation indexes of all the study variables.
For a clearer reading of the statistical indexes, we have reported only significant correlations. Concerning the relationships between State and Trait Factors of PADS, positive correlations are significant only between the State and Trait components of the Deontological Factor (r = .47, p < .05) and negative between State and Trait components of the Alternate Conflict Factor (r = −.54, p < .05).
Concerning the intercorrelations among State Factors of PADS, positive significant correlations are between:
Deontological and Altruistic (r = .45, p < .01); Deontological and Vulnerability (r = .81, p < .01); Deontological and Absent Conflict (r = .50, p < .01); Deontological and Akrasic Conflict State (r = .41, p < .05); Altruistic and Vulnerability (r = .51, p < .01), Altruistic and Absent Conflict State (r = .44, p < .05); Altruistic and Alternate Conflict State (r = .60, p < .01); Altruistic and Alternate Conflict State (r = .60, p < .01); Absent Conflict and Akrasic Conflict (r = .65, p < .01); Absent Conflict and Alternate Conflict (r = .65, p < .01); Akrasic Conflict and Alternate Conflict (r = .50, p < .05).
Concerning the intercorrelations among Trait Factors of PADS, positive significant correlations are between: Deontological and Altruistic (r = .80, p < .01);
![]()
![]()
(a)Notes. Sk = skewness; C = Kurtosis; S = state; T = Trait.
(b)Notes. **p < .01; *p < .05, S = state, T = trait.
Table 2. (a) reports means, standard deviations, skewness, kurtosis, of all the study variables; (b) Zero-order correlations (Pearson’s r) between variables (N = 25).
Deontological and Vulnerability (r = .45, p < .05); Deontological and Absent Conflict (r = .46, p < .05); Deontological and Alternate Conflict (r = .62, p < .01); Altruistic and Vulnerability (r = .51, p < .05); Altruistic and Absent Conflict (r = .67, p < .01); Altruistic and Akrasic Conflict (r = .54, p < .01); Altruistic and Alternate Conflict (r = .75, p < .01); Vulnerability and Absent Conflict (r = .67, p < .01); Vulnerability and Akrasic Conflict (r = .49, p < .05); Vulnerability and Alternate Conflict (r = .64, p < .01); Absent Conflict and Akrasic Conflict (r = .64, p < .01); Absent Conflict and Alternate Conflict (r = .68, p < .01); Akrasic Conflict and Alternate Conflict (r = .76, p < .01).
Finally, there is only one significant positive State-Trait inter-component correlation between the Vulnerability State and Deontological Trait (r = .41, p < .05).
We present here the relationships between PADS factors and the other measured constructs (see Table 2(b)) for the verification of the hypotheses.
As we expected, all dimensions of State PADS—Deontological, Altruistic, Vulnerability, Absent Conflict, Akrasic Conflict, and Alternate Conflict—are positively correlated with all Dysfunctional personality traits, such as Negative affect (.45 < r < .73; p < .05), Detachment (.43 < r < .74; p < .05), Antagonism (.41 < r < .68; p < .05), Disinhibition (.45 < r < .52; p < .05) and Psychoticism (.49 < r < .75; p < .05), except for Detachment and Alternate Conflict, Disinhibition and Altruistic, Akrasic Conflict and Alternate Conflict (non-significant correlations). Concerning Trait dimensions, only Deontological and Altruistic Traits have a positive and significant correlation with Dysfunctional personality traits, in particular with Antagonism (r = .41; r = .44; p < .05, respectively), Disinhibition (r = .59; r = .46; p < .05, respectively), and Psychoticism (r = .41; p < .05, for each).
As hypothesized, almost all PADS factors positively correlate with Depression (.42 < r < .73; p < .05), except for Vulnerability Trait, Absent Conflict Trait, Akrasic Conflict Trait, and Alternate Conflict State and Trait.
Concerning the expected positive correlations between Conflict with partners components, the only positive significant correlations are between Psychological Aggression and Absent Conflict State (r = .53, p < .05), and Akrasic Conflict State (r = .44, p < .05); Sexual Coercion and Akrasic Conflict Trait (r = .53, p < .05); Physical aggression and Akrasic Conflict State (r = .51, p < .05).
Concerning the hypotheses about the positive correlations between PADS dimensions and Different types of Sense of guilt, we found that all PAD States are positively correlated with Dirtiness (.47 < r < .75; p < .05), Empathy (.46 < r < .85; p < .05), Harm (.42 < r < .79; p < .05), and Moral Norm Violation (.43 < r < .68; p < .05), except for the correlations between Harm and Alternate Conflict State, and Moral Norm Violation and Vulnerability State. PADS Traits are not significantly correlated with the Sense of guilt, except for the positive correlation between Deontological Trait and Dirtiness (r = .55, p < .01), and negative correlations between Moral Norm Violation and Vulnerability (r = −.43, p < .05) and Akrasic Conflict Trait (r = −.48, p < .05).
Anxious Attachment Style, as expected, positively correlates with Deontological State (r = .42, p < .05) and Trait (r = .47, p < .05), Vulnerability State (r = .51, p < .05), and Akrasic Conflict Trait (r = .61, p < .01), but do not correlate with other PADS factors.
The hypothesis about the negative correlations between Self-esteem and PADS state Factors is not confirmed. Self-esteem positively correlates with all the PADS Traits: Deontological (r = .43, p < .05); Altruistic (r = .44, p < .05); Vulnerable (r = .74, p < .01); Absent Conflict (r = .67, p < .01); Akrasic Conflict (r = .44, p < .05); Alternate (r = .57, p < .01). But it does not significantly correlate with PADS State dimensions.
As hypothesized, there are no correlations between PADS Dimensions and Avoidant Attachment.
The other hypothesis about the negative correlations between PADS factors and Perceived relationship quality component subscales is only partially confirmed: only the Altruistic State negatively correlates with Relationship Satisfaction (r = −.47, p < .05), Commitment (r = −.48, p < .05), Intimacy (r = −.54, p < .01), and Trust (r = −.43, p < .05), but not with Passion and Love; also Alternate Conflict State negatively correlates with Commitment (r = −.42, p < .05) and Intimacy (r = −.49, p < .05). Deontological Trait positively correlates with Trust (r = .40, p < .05), Passion (r = .52, p < .01), and Love (r = .49, p < .05). Other positive correlations are between Relationship Satisfaction and Vulnerability Trait (r = .53, p < .01) and Absent Conflict State (r = .59, p < .01).
4. Discussion
The objective of this paper was to present the preliminary version of the PAD Scale (PADS) in a pilot study conducted on a clinical sample (victims of IPV). Moreover we presented PADS psychometric properties to capture the four dimensions of PAD in terms of state and trait: deontological, altruistic, vulnerable, and conflict (absent, alternate, akrasic). The preliminary PADS consists of 27 items, with the ultimate aim of having a tool to timely detect the early manifestation of the PAD phenomena and to intervene before the cycle of IPV (Walker, 2006) is activated. As a clinical case by Perdighe et al. (2022) demonstrates, the lack of clinical tools that can identify a TAD prevents therapists from recognizing severe and insidious psychological violence due to PAD. The tragic end was the victim’s suicide after years of IPV. Clinicians need to be aware of the cognitive factors of PAD to reduce the delay in diagnosing mental illness and improve the management of psychological outcomes in this patient population.
The scale was tested on a clinical sample of 25 participants. We conducted correlational analyses on the factors of PADS and other measures to preliminarily test PADS.
4.1. Correlations between Factors (Construct Validity)
The first relevant result is the positive significant correlation between the factors of PAD in both the state and trait conditions, which confirms the construct validity of the scale. Moreover, these results are consistent with previous literature on intimate relationships and attachment systems; in particular, as Spencer et al., (2021) pointed out, contact itself is reminiscent of patients’ maltreatment in early relationships. Specifically, we found a positive significant correlation between the three main factors of TAD (Deontological, Altruistic, and Vulnerability) in both state and trait conditions. According to the theoretical model of PAD (Pugliese et al., 2019, 2023) , all these factors contribute (with different weights depending on the specific personality profile and early problematic relationship experiences) to the description of the general functioning of the TAD. At the same time the positive correlations among the three conflict factors (absent, alternate, Akrasic), under both state and trait conditions, confirm that they constitute a unique conflict factor (Mancini & Giacomantonio, 2018) . All types of conflict can be activated by a specific moment of the pathological relationship. As for the correlations between the PADS factors and other measures, we hypothesized that they correlate positively with dysfunctional personality traits (Pico-Alfonso et al., 2008; Collison & Lynam, 2021) , depression (Beydoun et al., 2012) , couple conflict components, different types of sense of guilt (Beck et al., 2011) and anxious attachment style (Spencer et al., 2021) . Our results confirm the positive associations between dysfunctional personality traits and the factors of PADS under state conditions, which support the idea that IPV affects the victim’s mental health and becomes a trigger of specific dysfunctional beliefs and behaviors that degenerate into a kind of “transient” personality disorder. Among the trait factors, the only positive correlations between Deontological and Altruistic were Antagonism, Disinhibition, and Psychoticism. This can be explained by a different personality profile (PAD traits vs PAD state—when they are currently in the relationship vs when they have ended it) of TAD when involved in the abusive relationship, in which the dependency/submission domains (i.e. negative affect), typical of PAD, and the pathological opposite domain (detachment) were not present.
4.2. PAD and Depression
As for the positive correlation of PADS with Depression, the hypothesis was confirmed: depression is associated with the main factors of the PADS, both for the state and trait conditions, but not for the conflict factors, except for the Akrasic Conflict State. As we expected, when the conflict is absent or alternate, the victim is completely (for absent conflict) or partially (for alternate conflict) unaware that he/she is in a violent relationship. This dissociation could be a buffer for depression and would explain the non-significant correlation between depression and conflict factors. Furthermore, this confirms increased negative mood, loss of energy and vitality in TADs, and overall reduced mental health and quality of life due to PAD—both in state and trait conditions. These associations were also noted in a recent review based on 13 retrospective cohort studies, which highlighted the bidirectional association between depression and physical and mental health in IPV survivors (Bacchus et al., 2018) .
These results show how difficult it is to discern the temporality of this relationship. As expected, another retrospective cohort study from the UK, comparing 18,547 women who were victims of violence with 74,188 unexposed women (Chandan et al., 2020) , confirmed that there is a significant association between surviving violence and poor mental health, particularly concerning depression and anxiety. Unfortunately, none of these studies were conducted in Italy.
4.3. PAD and Couple Conflict
The Conflict Tactic Scale (Straus et al., 1996) , which measures how much partners attack each other psychologically and physically, was positively and significantly correlated with absent and akrasic conflict: In the absent condition (when the conflict is absent in the victim’s mind), victims could be hypothesized to dissociate the partner’s abusive behaviors after the trauma (Basu, Levendosky, & Lonstein, 2013) . They may be in a positive moment of the dysfunctional relationship (so-called honeymoon; e.g., Baldry & Roia, 2011; Pugliese et al., 2019; Walker, 2006 ) and remain in the toxic condition with “only” psychological violence. In the akrasic conflict (when victims know they are in a violent relationship but do not want to sever the bond with the toxic partner as a result of a long period of abuse), victims might eventually be aware of physical aggression and sexual coercion. Despite the abuse, they might not be able to disengage from the relationship, as described in the definition of PAD.
For the alternate conflict, the correlation with the couple conflict is not significant because it could depend on one of the two states in which the victim was separately present at that time. If he/she does not integrate, he/she may not be aware of the abuse, to invest in the future of the relationship. The psychological, emotional, and verbal abuse, on the other hand, may not even be recognized by the victim as abuse and therefore does not bring him/her into conflict (Perdighe et al., 2022) .
4.4. PAD and Guilt
As for the positive association between PAD and a sense of guilt, we found it only for the State condition, which confirms the hypothesis based on the theoretical model: during the PAD state, TAD tends to have a higher sensitivity to the partner’s suffering, since it is a mission, they may have undertaken in early childhood, and also a strong sense of morality and integrity, consistent with the deontological factor, due to a lack of personal value and a sense of duty to maintain the relationship.
Regarding the anxious attachment style, as expected, there was a positive correlation with Deontological State and Trait, Vulnerability State, and Akrasic Conflict Trait. Individuals with an anxious attachment style tend to fear rejection and abandonment. They often worry that their partner does not want to be with them, feel more vulnerable and lonely, and believe that they do not deserve genuine and sincere love. This can be considered typical of people who suffer from PAD.
4.5. PAD and Self-Esteem
The hypothesis about the negative correlations between Self-esteem and PADS state Factors is not confirmed. Self-esteem positively correlates with all the PADS traits, but not with PADS states. It seems that when TADs recognize their PAD as a temporary state associated with a particular situation, it does not affect (either negatively or positively) their self-esteem, but is independent of it, whereas when they perceive it as their characteristic trait, this strengthens their self-esteem.
4.6. PAD and Perceived Relationship Quality
The other hypothesis about the negative correlation between the PADS factors and the component subscales of the Perceived Relationship Quality was confirmed: The more the victims suffer from PAD, the more positively they describe the relationship according to the main anti-goal “do not break the bridge with the partner at any cost”. Specifically, the deontological Trait is positively correlated with Trust, Passion, and Love. Other positive correlations exist between Relationship Satisfaction and Trait Vulnerability and the State of Absence of Conflict However, the Altruistic State correlates negatively with Relationship Satisfaction, Commitment, Intimacy, and Trust, but not with Passion and Love. This confirms the idea that TADs in the altruistic state because everything is focused on the toxic partner, are unhappy in their relationship and feel overwhelmed. However, they reported in the sessions that they still love their partner but at the same time do not know/want how to break the toxic bond. In addition, the Alternate Conflict State correlates negatively with Commitment and Intimacy. This correlation would relate to the specific, non-integrated state of the relationship.
4.7. PAD and Avoidant Attachment (Discriminant Validity)
As hypothesized, there were no correlations between PADS dimensions and Avoidant Attachment, according to which individuals had the goal of avoiding intimacy, a very different condition of the TAD (Goal 0). Such a result confirmed the discriminant validity of the PADS, while the previous significant positive and negative correlations confirmed a convergent validity.
5. Conclusion
PADS seems to have good validity. According to the cognitive model of Pugliese et al. (2019, 2023) , there are four main factors describing a profile of TAD (altruistic, deontological, vulnerability and conflict). Each factor is linked to a specific self/other image (altruistic self/fragile partner; deontological self/humiliating partner; vulnerable self/abusing partner). The pivotal feature of people with PAD is the fear of separation and loss of relationship with an unavailable, fragile, and/or abusive partner. Thus, the purpose of TAD is to maintain the relationship and ensure closeness to the other person. In case of a risk of break up, the TAD would experience a strong conflict between the desire to be with that person and the desire to separate from him or her. Three types of conflict have been hypothesized. In the first conflict (absent), PAD is ego-syntonic: the person considers the benefits of remaining in the relationship greater than the costs. PAD is ego-dystonic in the other two conflicts (alternate and akrasic). When the conflict is alternate, the person vacillates between different mental states that are activated separately and are therefore not coherently connected. In the third (akrasic) type of conflict, the person is fully aware that staying in this type of relationship is harming him/her; he/she would like to get rid of it, but does not.,
Concerning the preliminary scale of DAP, these first results showed that PAD is both a trait and a latent psychological condition triggered by the abusive partner. These results are also consistent with the observations made about the different associations between the components of state vs. trait PADS (for example, with the dysfunctional personality traits). Indeed, the trait condition lacks the correlation with the negative effect that was hypothesized to correlate with PAD, while the same positive correlation is significant in the state condition (concerning the current relationship or the last dysfunctional relationship experienced by respondents). Thus, it is as if one has a hidden illness that only occurs when exposed to an interpersonal trigger factor (such as the correlation between rheumatism and humidity as an environmental trigger factor). This has several implications for clinical and social interventions that focus on key factors, such as community education programs in schools (for adolescents) and universities (for young adults), in urban social centers, with a focus on disease prevention (e.g. how to quickly recognize an abusive partner) and on direct interventions with perpetrators, as well as on psychosocial education for healthy intimate relationships with a focus on health promotion (interventions to promote self-compassion, self-assertion, empowerment, self-determination, autonomy, economic independence, etc.) and on interpersonal capital, improving relationship assets, social networking as a supportive pattern for individual well-being, and interdependent happiness (Maricchiolo et al., 2021) .
5.1. Limitations and Future Developments
The present study has evident limitations. The small sample size could have prevented us from observing true relationships among variables in our pilot study; in any case, we want to underline that our sample was very difficult to recruit. It was a sample of survivors of IPV and the research team was interested to collect useful information for scale development but with the contemporary attention to not be focused only on reasonings about adequate statistical power but especially to not expose victims to secondary trauma. It is noteworthy to mention also the fact that the sample is composed of IPV survivors who can be classified as fully mature (i.e. mean age 41.32) and so our preliminary results could not be generalized to adolescents and young people. Moreover, to better understand the differences and peculiarities of the state and trait conditions, a larger sample should be tested, which would allow also for testing the structure of the PADS with exploratory and confirmatory factor analyses approaches, in so guaranteeing a solid basis for application and use of the scale in both prevention and treatment contexts. Finally, future investigations should aim to compare people of different sexes, gender identities, sexual orientations, social classes, and education levels as well as cultures and nationality (by cross-cultural or cross-national research), to demonstrate the transversal spread of the phenomenon or eventually describe its differences in terms of impact from these diverse social categories.
5.2. Practical Implication
PAD model and scale (PADS) has a practical contribution to the field of pathological affective dependence and its role in Intimate Partner Violence. For example, in-depth knowledge of the psychological profile of the TAD, its cognitive functioning, and the interpersonal cycles, would promote early recognition of the specific dynamics activated in violent relationships with abusive partners. The PAD scale would allow psychologists and psychotherapists (or all the professionals working in the field of IPV (i.e., lawyers, police officers, doctors, etc.) to protect the TAD from the severe consequences of IPV. Moreover, this cognitive model and scale can help professionals manage IPV dynamics, and implement the right interventions aimed at reducing the resulting psychosocial impact. On the patient’s side, recognizing one’s mental functioning specifically activated in IPV situations would promote the important awareness of the psychological dynamics, negative automatic thoughts and self-images, etc., that blind them to a healthy separation from the partner. Hence, this is necessary to break the typical cycle of violence. From a social-psychological point of view, the model and scale of PAD (PADS) can be used for psychoeducational interventions in schools, but also could help train police officers or all the professionals dealing with victims of IPV.
Acknowledgements
Erica Pugliese would like to acknowledge all the IPV victims and survivors of the self-mutual-help group Millemé (https://www.milleme.it/) and patients of APC (https://apc.it/) for their voluntary research collaboration.
NOTES
1The existing measures employed are: The Pathogenic Beliefs Questionnaire (Caron, 1992) ; PID-5 (Fossati et al., 2015) ; Coping Schemas Inventory-Revised (Wong et al., 1993) ; Young-Rygh Avoidance Inventory (Young & Rygh, 1994) ; Parental Bonding Instrument (Parker et al., 1979) . SCID-II (First et al., 1995) ; Traps Questionnaire (Young, Klosko, & Weishaar, 2003) .