Intimate Partner Violence

Several reports have shown that having suffered some kind of partner violence is quite common for various ethnic or social groups. In fact, ^{Chan, Shen, and Takeuchi (2009)} found that 10.2% women and 12% men out of a representative Asian-American population from USA reported to have suffered violence from their current intimate partner. A recent meta-analysis on a combined population of 329.212 military soldiers from Canada and USA showed prevalence of physical violence committed during the last year of 26% for men and 20% for women (^{Kwan et al., 2020}). Other studies reported prevalence of intimate partner violence suffered along their lives of 24.3% for women and 13.8% for men (^{Black et al., 2010})

Concerning couple violence in adolescents, early reports informed a prevalence rate between 9% and 57% (^{Cascardi, Avery-Leaf, & O’Leary, 1994; Roscoe & Callahan, 1985}). Later, the Center for Disease Control and Prevention reported that 12% secondary students had suffered physical violence during intimate partner relationship (^{Center for Disease Control and Prevention, 2000}

The high prevalence of violence among young students dating was also reported by an international study using samples from 22 countries (^{Chan, Straus, Brownridge, Tiwari, & Leung, 2008}). Though this study did not include Ecuador, two Latin American countries were part of it (Mexico, Brazil). Results for the country in the median showed that 30% of students in the sample had committed physical violence against their partner during the previous 12 months (^{Chan et al., 2008})

Moreover, evidence on adolescent relationships shows concurrence between aggression and victimization (^{Chan et al., 2008}; ^{Malik, Sorenson, & Aneshensel, 1997}; ^{Straus & Ramirez, 2007}; ^{Taylor & Mumford, 2016})

Intimate partner violence may be expressed in different ways. ^{Burk and Seiffge-Krenke (2015)} grouped violent behavior into two categories: physical aggression and relational aggression. The first consists of the intentional use of force that can hurt the partner and includes mildly aggressive behavior -such as a shove or a scratch- and severely aggressive behavior -such as asphyxiation attempt, slap or attack with weapon. This second category includes insult, verbal attack, defamatory gossip, exclusion, flirting with somebody else with the purpose of provoking jealousy and breaking-up threat. This type of relational aggression is associated to poor psychological adjustment, depression, and dysfunctional relationship (^{Prinstein, Boegers, & Vernberg, 2001}). When comparing between sexes, women show higher disapproval of these relational aggressions, feel higher impact of this behavior on their relationship, and spend more time thinking and talking about it (^{Salmivalli & Kaukiainen, 2004}). When partner violence is mutual, it is frequently associated with less adaptive functioning, conflict, lack of relevant qualities such as confidence and support (^{Seiffge-Krenke & Burk, 2015}).

^{Malik et al. (1997)} reported that Los Angeles female adolescents showed higher probability of intimate partner violent behavior than men and that the risk factors favoring aggression were: previous exposure to violence, naturalization of the use of violence, and cannabis use. ^{Capaldi et al. (2012)} carried out a systematic review and among demographic risk factors found that dating violence risk decreases with age, reaching its highest level in late adolescence or early adulthood. ^{Johnson, Giordano, Manning, and Longmore (2014)} provided further support to this finding reporting that in men, dating violence increases from 13 to 20 years old and then decreases, and in women, though with a similar pattern, its highest level is observed between 21 and 24 years old.

Latin American research studies also showed high prevalence and symmetry of violence in dating relationships. The work performed with a sample of 963 students from the National University of Cordoba, Argentina, revealed 34% prevalence of physical aggression for women during the past 12 months and 22% for men; acceptance of having committed aggression increased significantly the probability of suffering similar or another kind of aggression (^{Arbach-Lucioni, Nguyen-Vo, & Bobbio, 2015}). A recent Argentine study on 528 students from the Medicine School of the National University of La Plata reported an IPVUS mean of 0.35 (SD= 0.27; possible range 0-3); verbal emotional violence as prevailing type (M= 0.85; SD= 0.5); punctuations concentrated in lowest values but 3.2% with values higher than 1 in total violence committed; 50.2% of total, 19.1 % women and 20.6 % men informed to have committed at least one physical aggression during the index relationship (^{Delucchi, Leon-Mayer, & Folino, 2020}).

Jealousy and Intimate Partner Violence

Scientific literature provides broad evidence on the influence of jealousy on partner violence. Jealousy and violence of man on his woman partner have been early described as a psychological tendency that evolved in the Homo Sapiens to strengthen his confidence in paternity and ensuring sexual exclusiveness and control (^{Daly, Wilson, & Weghorst, 1982}). More recently, studies have focused on the relation of jealousy and other individual variables with aggression or other menacing consequences, and left aside that restricted view of male sex and reproductive functionality to also deal with transgender and nonconformist young people (^{Goldenberg, Jadwin-cakmak, & Harper, 2018})

Jealousy has been related to psychological and physical aggression in North American young people (^{Davis, Ace, & Andra, 2000}; ^{Giordano, Soto, Manning, & Longmore, 2010}), and Spanish students (^{Muñoz-Rivas, Graña, O’Leary, & González, 2007}). A study carried out in Boston (^{DeSteno, Valdesolo, & Bartlett, 2006}) associated the intensity of jealousy, mediated by self-esteem, with aggression incitement towards the partner.

^{Murphy y Russell (2016)} found that jealousy resulted to mediate between the sensitivity to rejection and aggression, and ^{DiBello, Rodriguez, Hadden, and Neighbors (2015)} reported the mediation of cognitive jealousy in the association of self-esteem in the couple relationship, alcohol use, and problems related with this addiction.

A Latin American contribution from a feminist qualitative research study showed the relation between jealousy, infidelity, and substance use with conflict and partner violence (^{do Nascimento Paixao et al., 2014}). Another qualitative study from the same country found significant association between non delirious pathological jealousy and anxiety, which could increase violent reactivity (^{Costa, Sophia, Sanches, Tavares, & Zilberman, 2015}). The Argentine research study aforementioned provides support to the existence of a significant relation between jealousy and partner violence (^{Delucchi, Leon-Mayer, & Folino, 2018}). Particularly, this study revealed that behavioral and cognitive jealousy associated positively and significantly with IPVUS, even controlling other factors such as stress and substance use during partner relationship and the age at the beginning of the relationship (^{Delucchi et al., 2020}).

Infidelity, substance use, stress and intimate partner violence

Jealousy, partner infidelity, and alcohol or substance use are closely related with partner violence (^{Paixão et al., 2014}). Early evidence led to consider that previous infidelity, either committed or suffered, increases the individual predisposition to react jealously (^{Pines & Aronson, 1983}). More recently, partner infidelity was reported to be a significant predictor of jealousy, particularly in men (^{Burchell & Ward, 2011}), and to trigger rage and propension to violence in response (^{Miller & Maner, 2008}).

According to evidence from community and assistance institutions, substance use is associated with violence risk (^{Swanson, Holzer, Ganju, & Jono, 1990}; ^{Swanson, Monahan, & Steadman, 1994}), even in partner relationships (^{Capaldi et al., 2012}). ^{Brem, Shorey, Rothman, Temple, and Stuart (2018)} found alcohol use in jealous men significantly related with partner violence. Moreover, ^{Ihongbe and Masho (2018)} reported that synthetic cannabis use is associated with a higher risk of violence victimization in young couples. Though substance use interacts with other variables affecting young partners violence -such as family support and student expectations (^{Gomez, Orchowski, Pearlman, & Zlotnick, 2019}), the study of this variable results quite relevant since it may be the target for further preventive actions.

Stress is known to be part of the diathesis for violence and is one of the factors analyzed to evaluate violence risk (^{Douglas, Hart, Webster, & Belfrage, 2013}; ^{Greer, Taylor, Cella, Stott, & Wykes, 2020}). In young couples, it is considered a direct consequence of violence, or indirect, when suffering is caused by the social rejection for young violent partners (^{Temple et al., 2016}). Recently, a significant independent contribution of stress to partner violence has been reported (^{Buunk & Massar, 2019}).

Context

The present research study was carried out in the University of Cuenca, City of Cuenca, Ecuador. This University is a pubic institution with around 18,000 pre- and postgrade students. This study contributes to stregthen university policies since it promotes the development of care attention and orientation programs, drug and violence prevention campaigns, and assistance in case of rights vulnerability (^{Aula de Derechos Humanos, 2018}; ^{Consejo Universitario, 2018}). Institutional policies for the prevention of violence demonstrate the relevance of the promotion of empiric investigation on partner violence issues in university students and the need of developing evidence-based preventive actions. However, systematic studies on these issues are still scarce. A recent national report allows to perceive the scope of this problem in Ecuador, showing that, during their lives, 40.8% women have suffered psychological violence and one out of four of them has suffered physical violence (^{Instituto Nacional de Estadísticas y Censos, 2019}).

Present Study

The aim of the present study is to contribute to the gathering of information that can help develop IPVUS prevention programs. Particularly, the objectives of this study are: to describe violent behaviors and explore their relation with jealousy and other variables theoretically relevant in a sample of students from the University of Cuenca. We hypothesize here about the existence of a positive significant association of jealousy, stress, and substance use, with IPVUS.

Materials and Method

Results

Participants provided information about their index partner relationship, which had started at 18 years old in average (SD 2.74; min. 13; max. 26) and had a mean duration of 33 months (SD 30; min 1; max 144). Out of the whole sample, 49% participants informed that the relationship still continued at the moment of data collection.

IPVUS values distribution according to the CADRI differed significantly from normal for all sub-scales and for total. Total values distribution according to the CADRI for Abuse committed is shown in Graph 1, in which the abscissa is the possible score range. Scores concentration was low; however, value for Total abuse committed was 1 or more in a 5.4% of the participants, which suggests a concerning level of violence. Out of all participants, 60.2% informed to have committed at least one act of physical violence during the index relationship.

Graph 1: Distribution of Total Abuse Committed values according to the CADRI and possible range in abscissa. Mean= 0.41 (SD 0.331). N= 186. 

Table 1 shows descriptive statistics for CADRI sub-scales and totals, stratified by gender. Verbal-emotional violence was the subtype most committed and suffered. Parity in results stands out when they are compared between sexes (Table 1). Correlation between total violence committed and suffered was .77 (p< .001).

Table 1: Violence committed and suffered according to CADRI sub-scales and totals 

Measured by MJS, jealousy showed normal distribution only for total values (Kolmogorov-Smirnov= 0.063; p= .07). Values for emotional sub-scale showed left asymmetry, and cognitive and behavioral sub-scales, right.

For both sexes, the highest mean value was observed in emotional sub-scale (Table 2). Stratification by sex revealed that there were no statistically significant differences between means.

Table 2: Descriptive statistics of MJS values and sex comparison 

Co-variables showed an average medium stress during the index relationship (M= 4.4; SD 2.6; min. 0; max. 10) and higher level of fidelity than that of their previous partners with them (M= 2.6; SD 3.1; min. 0; max. 10 vs. M= 3.8; SD 3.2; min. 0; max. 10).

Stratification by sex revealed that their stress estimation by the time of index partner had no significant differences between men and women (M= 4.5; M= 4.4 respectively; p= .67). In average, women informed having committed less infidelity (M= 2.47; M= 3.14 respectively; p= .23) but having suffered more infidelity (M= 3.89; M= 3.18 respectively; p= 0.28), though differences were not statistically significant.

All participants informed that during the index relationship they had not used methamphetamines, heroin, or inhalants but 1.6% of them had used hallucinogens, 1% cocaine, and 1% ecstasy. Moreover, 6.5% had used cannabis some days in a month and 3.3% had used it at least half a month. Out of total participants, 32.8% informed no alcohol use, 53.8% had used it some days in a month, and 12.9% had used it at least half a month or more. Analgesics use without medical prescription was informed by 17.2% participants and 21% had smoked tobacco during the index relationship. A summary variable “Substances used” was developed adding up all substances used except for tobacco, and the mean was 1.48 (SD 1.9). Mean for men was 2 (SD 2.5) and for women, 1.4 (SD 1.7).

Table 3: Correlations for CADRI total abuse committed and independent variables 

In order to consider CADRI variable Total Abuse Committed as the dependent variable in a multivariate model, and taking into account distribution asymmetry, a logistic regression model was formulated. Thus, we dichotomized its total according to the median (Md= 0.3) and assigned code 1 to cases with values higher than median, and 0 to those with equal or lower value; they were labeled as “High violence group” (47.8%) and “Low violence group” (52.2 %). A logistic regression model was formulated considering as independent variables all those for which correlation was informed in Table 3, adding sex. Estimation was performed using the “forward conditional” method with entry probability of 0.05 and removal probability of 0.10.

Variables remaining in the model were MJS Behavioral (OR= 1.09; p= .003; 95% IC 1.03, 1.2) and Stress (OR= 1.41; p< .001; 95% IC 1.21, 1.64). The model classified correctly 71.4% cases and adjusted data properly (Hosmer and Lemeshow= 8.11 (gl 8) p= .42). Variables in the model explain no less than 22% variance (Cox & Snell R-square= 0.22; Nagelkerke R-square= 0.30).

Discussion

The aims of the present study were to describe violent behavior in a sample of students from the University of Cuenca, Ecuador, and examine their relation with jealousy and other factors theoretically relevant. Data were obtained through a survey which asked participants to inform about the variables within the context of an intense romantic relationship, one that had moved their feelings, romanticism, or love, according to their own interpretation. Students provided information about partners of rather long relationships which in almost half the cases still continued. Since couples had started their relationship at 18 years old in average, they are far from the time in which first adolescent contacts start. These characteristics provide reliability to references since memories are less affected by the course of time, and in coincidence with our request, they are memories of significant experiences. Thus, since information required was vulnerable to memory bias, this risk was minimized.

Any expression of violence is undesirable but it is even more distressing when it is more frequently or intensely externalized. The general IPVUS profile obtained with the CADRI showed low values; prevalence of verbal emotional violence was observed with the highest values, in agreement with findings from a recent Latin American study (^{Delucchi et al., 2020}). However, more than 5% students in the sample showed concerning values and 60.2% informed to have committed at least one physically violent act during the index relationship.

Another finding to be discussed is the parity observed between values for violence committed by the group of women and those by the group of men. Previous studies reported aggression bilaterality in young couples (^{Chan et al., 2008}; ^{Malik et al., 1997}; ^{Straus & Ramirez, 2007}; ^{Taylor & Mumford, 2016}), and even higher values in the group of women (^{Arbach-Lucioni et al., 2015}; ^{Elmquist et al., 2016}). Though the present study does not assess those subtypes determined by aggressive unidirectionality -only from man to woman or only from woman to man- or bidirectionality (^{Straus & Gozjolko, 2014}), our findings show a prevailing reciprocity of violence in young couples. Besides parity, the present work confirmed that violence committed varied concomitantly with violence suffered. Moreover, this result suggests the existence of bilateral violent interactions in couples since the context for both types of violence was the index couple itself. Interestingly, ^{Fernández-Fuertes, Fuertes and Pulido (2006)} also found high correlations between violence committed and suffered, and made a similar interpretation suggesting the possible existence of mutual violence patterns. This IPVUS aspect could be associated with shared behavioral and expression habits, and even with the predisposition to action, either defensive or offensive, that young people have. In agreement with ^{Taylor and Munford (2016)}, our findings allow to suggest that prevention/treatment services should not contemplate groups as completely separated into victims and aggressors.

No matter whether the level of violence that women declared is similar or higher than that informed by men, the interpretation will never attenuate the historical regularity of the higher impact of masculine violence but reveals a relevant aspect of this issue: violent behavior in partners belonging to this age group goes beyond biological sex. And this is relevant not only for prevention policies but also for forensic assessment since, at the moment of assessing a specific case, this evidence forces the expert to explore about the existence of an eventual violent interaction behavior.

Findings suggest that the provision of specific prevention services to the local university students population is not only morally fair but also empirically demonstrated. Comparison with studies from other countries show agreement in this sense. Mean values were higher for our sample than for a similar age group -19 years old- from a Spanish sample (^{Fernández-Fuertes et al., 2006})⁠ and an Argentine sample (^{Delucchi et al., 2020}). These considerations highlight the need for special preventive care of local students and make evident for future studies the significance of analyzing the eventual influence of macro factors such as gender inequality (^{Okeke, Mumford, & Rothman, 2019}) or cultural standards related with the early start of romantic relationships.

Some comments on descriptive findings in the present study will help in the discussion. Regarding total jealousy, we observed a normal distribution, but when analyzing emotional jealousy on the one hand, and cognitive and behavioral jealousy on the other hand, a clear contrast was found. In general, students informed that jealousy was mainly emotional. Differences between women and men were not significant, unlike Argentine results where women informed significantly higher jealousy quantum than men (^{Delucchi et al., 2020}).

Concerning data on the estimation about their own stress, infidelity committed, and that committed by previous partners, students informed medium stress level and no significant differences were found between sexes.

The assessment of substance use during index couple relationship showed that almost 10% had used cannabis and two thirds, alcohol. These data reveal that substance use for the sample in the present study was lower than that for the student population of La Plata, Argentina (^{Delucchi et al., 2020}). In fact, this information should be assessed in future research studies since Ecuadorian mountain area shows the highest level of alcohol use in young population (^{Instituto Nacional de Estadisticas y Censos, 2014}).

Except for emotional jealousy, the relation between each of the rest of co-variables with IPVUS was statistically significant. Interestingly, unlike results with the remaining measures, as lower the age at the start of the relationship, the higher IPVUS value. This finding agrees with those from foreign studies (^{Delucchi et al., 2020}; ^{Johnson et al., 2014}) though these last showed higher correlation. This association remained negative when the multivariate model was formulated but lost statistical significance; the same was observed for the association with other factors. Only associations of behavioral jealousy and stress with cases in the high violence group kept statistically significant. Even though the study design does not allow the assessment of causality, findings are compatible with a multifactorial explanatory model in which externalized jealousy and stress finally affect violent behavior between partners. The association between these factors and IPVUS cannot explain the complete variance of the phenomenon. Further investigation will allow identify other influencing factors, but for the time being, this finding unveiled an empirically delimited target for preventive policies. Not only behavioral externalization of jealousy but also concomitant stress are factors to be considered in order to reach an optimistic perspective on intimate violence decrease.

Besides its cross-sectional design, this study presents some further limitations. Results do not represent the whole population of university students. Though the sample includes most of the students from two university degree courses (186 out of 220), it does not encompass the complete universe of local students. The marked disproportion between women and men, and eventual characteristics of students from those particular courses might have biased our results.

Data obtained do not allow determine whether substance use started after violence and was consumed to relieve discomfort, or if it started before, so that eventual psychoactive effects could have been present at the moment of the violent interaction. We cannot either determine whether substance use caused just transitory effects or if it was a long lasting disorder. This factor should be further investigated and considered, at least theoretically, for possible preventive actions. Its harmful influence should not be underestimated since substance use is related with violence in complex and bidirectional ways (^{Levitt & Cooper, 2010}). Substance use may directly act as disinhibiting factor and indirectly when used as palliative, postposing the resolution of conflict, or affecting human relationship.

The characteristics of this study that allowed reduce the risk of memory and social desirability biases, such as the context of an affectively significant relationship and anonymity, have been already addressed. However, we admit the possibility of a certain level of influence on information obtained.

Finally, as far as we could confirm with indexed bibliographic data, this research study is the first empirical attempt to relate jealousy with violence in young students couples from Ecuador. Factors associated with IPVUS should be considered when evidence-based prevention programs are designed. Authors from other countries have already remarked the need for implementing those programs (^{Khubchandani et al., 2017}) and the present study can contribute to make this need known and locally accomplished.