Intimate Partner Violence in Adolescent Girls: The Role of Impulsivity and Emotional Dysregulation

Abstract

Adolescent dating violence is a critical issue with far-reaching consequences. This study explores the role of maladaptive emotional regulation strategies in the relationship between impulsivity and the perpetuation of violence, differentiating between violence received and exerted. A cross-sectional correlational design was used. For sample selection, a probabilistic sampling method was used, specifically a one-stage stratified random sampling with proportional allocation. The sample consisted of 700 female adolescents aged between 14 and 18 years (mean = 15.64, SD = 1.20) from Ecuador. Participants were selected if they had been in at least one romantic relationship lasting more than a month. The data analysis included descriptive statistics, Chi-square tests for categorical variables, t-tests and ANOVA for group comparisons, and Pearson’s correlation for continuous variables and mediation analysis. The results show that both physical violence exerted and received have significant positive correlations with impulsivity and maladaptive strategies, except for DERS Awareness, DERS Goals, and DERS Clarity. Impulsivity also maintains significant positive correlations with all emotional dysregulation strategies except for DERS Awareness. DERS Impulsivity and DERS Non-acceptance act as partial mediators in the relationship between impulsivity and violence (both exerted and received). Impulsivity itself is established as a clear predictor of physical violence in adolescent intimate relationships, highlighting the need for interventions and the development of educational programs aimed at improving emotional regulation strategies, such as controlling impulsivity and accepting emotions.

1. Introduction

Intimate partner violence is a prevalent problem in adolescence. This life stage is especially important, as individuals assess what they want in a long-term romantic relationship as well as acceptable and unacceptable traits and behaviors, such as violence [1].

Partner violence is defined by the manifestation of different actions, including physical (e.g., hitting or pushing), psychological (e.g., intimidation, manipulation, isolation, or humiliation), sexual (e.g., forcing the partner to perform unwanted sexual acts), and verbal (e.g., insults, constant criticism or disqualifications) actions [1].

Ecuador is a country with a serious problem with gender-based violence. According to data collected by the National Institute of Statistics and Census (INEC) macro-survey [2], 6.5 out of every 10 women aged 15 to 65 have experienced some form of violence in any setting throughout their lives, while 3.2 out of every 10 have experienced it in the last 12 months. Regarding the type of violence, psychological violence is the most prevalent, followed by physical, sexual, and economic or patrimonial violence. Specifically, in terms of gender-based violence, 43 out of every 100 women have experienced some form of violence from their partner. Around 80% of these women did not report the perpetrator. On the other hand, data collected in the National Survey on Family Relationships and Gender-Based Violence Against Women [3] indicate that the onset of cohabitation and the age of women are determining factors in the establishment of violence within a relationship. The younger a woman enters into cohabitation or marriage, the higher the likelihood of experiencing violence. However, as women age, violence appears to decrease. Among adolescent girls who got married or started living with their partners between the ages of 10 and 15, physical violence reaches 39.3%, a figure that doubles that reported by those who married between 26 and 30 years old (20%). These figures become even more alarming when it comes to sexual violence: 16.1% of the youngest group reported experiencing it, whereas the percentage drops significantly to 7.7% among those who married or cohabited between the ages of 26 and 30. In this regard, the age of consent for sexual relations or marriage plays a crucial role, as delaying this age until at least adulthood is fundamental for protecting girls and adolescents. According to the INEC Yearbook [4], in Ecuador, 27% of women get married between the ages of 12 and 20, and 30% do so between the ages of 21 and 25. Although the legal age for marriage in Ecuador is 18, it is possible to marry from the age of 16. However, data suggest that marriages of convenience or forced relationships often begin even earlier than these legal limits.

This problem is not exclusive to Ecuador; similar situations occur in other Latin American countries. In fact, 14 Latin American countries are among the 25 nations with the highest incidence of femicides worldwide [5,6]. Notably, El Salvador, Guatemala, Honduras, Colombia, and Bolivia stand out, with femicide rates ranging between 12 and 6 women murdered per 100,000 women [7]. The same is true for Spanish-speaking European countries, such as Spain, where a study detected that for more than 13,000 adolescents between 14 and 20 years of age, psychological violence is the type of violence most frequently exercised and received within the couple, followed by physical and sexual violence, in both boys and girls [8]. Other macro-surveys carried out in Spain on gender violence report that 46.6% of women aged between 16 and 24 years have suffered some type of violence from their partner or ex-partner. This prevalence decreases to 32.4% in women aged 25 years [9]. Based on these statistics, it can be confirmed that dating violence is a priority issue in prevention and health promotion strategies, as it leads to serious repercussions for the well-being of adolescents [10].

In line with the increasing statistics on teenage dating violence and the consequences that this experience can have on the individual well-being and future relationships of adolescents, recent scientific studies have shown interest in identifying the predictors and consequences of intimate partner violence. In general terms, the literature related to risk of or predispositional factors for intimate partner violence focuses on individual and family factors [8,9,10,11,12].

Within the individual factors, personality traits such as impulsivity stand out, since impulsivity is considered one of the promoters for developing stable behavioral patterns such as aggressiveness when intimidation or provocation is perceived. Impulsivity is an emotionally charged trait that, when studied together with violent behavior, can be considered to lead to a tendency to execute unplanned aggression without control. Although no differences have been found between the type of violence perpetrated or suffered and impulsive behaviors, it does appear that impulsive violence is more likely to be perpetrated by men towards women than vice versa [10,11,12].

In response to the stress generated by the experiences of violence, both the aggressor and the aggressed may deploy various strategies to adapt to and manage the impacts of the experiences lived in the couple’s relationship. Thus, the ways in which a person copes with a stressor, such as violence, can establish patterns of adaptive or maladaptive coping that, with a high probability, can crystallize in adulthood. Previous literature has divided these strategies into two main categories depending on whether the efforts made by the person are focused on changing or modifying the stressor (approximation coping) or, on the contrary, whether the efforts are devoted to escaping regulating the negative emotional state that triggered the stressor (avoidance coping) [13]. While it is true that approximation coping strategies are related to resilience processes and help to improve the person’s response to an aggression, avoidance strategies contribute to the appearance of behavioral and emotional problems (e.g., social withdrawal, anxiety, depression, etc.), and are also associated with an increased risk of suffering violence in other subsequent intimate partner relationships [14,15,16,17].

It is evident that reliance on the use of adaptive coping strategies is favorable for optimizing the recovery of victims experiencing of intimate partner violence. However, the use of some strategies or others differs depending on the stage of the relationship or breakup in which the partners are. For example, Calvete et al. [18] found that maladaptive strategies partially mediated the effect of the violence suffered on anxiety and depression symptoms in women. Another study confirms that avoidance strategies, such as alcohol consumption or lack of support in the close environment, are associated with posttraumatic symptoms one year after suffering violence [18,19,20]. A more current study conducted by Goldberg–Looney et al. [21] explored the ability to predict the use of coping strategies as a function of severity of violence. Their results reveal that as intimate partner violence increases, the ability to plan behavioral coping to exit the stressor such as creating a safety plan to escape the batterer or seeking help outside the home, is more likely to decrease. However, in relation to impulsivity, the literature does not show consistent results for its implications for the use of adaptive or maladaptive coping strategies and the violence experienced and perpetrated by adolescents.

On the other hand, in relation to family factors, there is a large proportion of existing research that focuses on the interaction of parents with their children, especially on parenting style. Specifically, parental hostility to children was positively associated with aggressive behaviors in adolescents and was found to be predictive of children’s perpetration of violence toward their partners [22,23]. Finally, exposure to family stress has also been explored as an important predictor of generational transmission of intimate partner violence. Parental work, financial worries, or marital conflict have been associated with higher rates of parental intimate partner violence and perpetuation of violence in children with their partners [24]. From an integrative perspective, and looking at family factors, the presence of childhood trauma deserves mention. The impact that childhood trauma can have on socio-emotional competence in adulthood has been highlighted [25]. In turn, this deficit in socio-emotional skills, indispensable in couple relationships, can trigger impulsive behavior and maladaptive emotional regulation in couple relationships. The importance of family factors in turn provides a contextual variable that helps us to understand the complexity of intimate partner violence in adolescence, beyond the individual risk factors that are the focus of this research.

Based on the literature, the present study attempts to address several hypotheses:

Hypothesis 1.

Individual risk factors as a predictors.

The first hypothesis questions whether individual risk factors such as impulsivity could be predictive of violence perpetrated and received in intimate partner relationships in adolescents.

Hypothesis 2.

Mediators between impulsivity and violence perpetrated and received.

We also propose that maladaptive coping strategies (emotional regulation) may mediate the relationship between impulsivity and violence perpetrated and received.

Hypothesis 3.

The role of family factors.

Finally, we speculate that family factors such as adolescents’ cohabitation with parents, or even with grandparents, may have a significant effect on the proposed mediation model.

2. Materials and Methods

2.1. Design

This study used a cross-sectional correlational study design. A probabilistic approach with stratified random sampling and proportional allocation was employed. The distribution was determined by population size, considering the number of educational units, adolescents per class, urban and rural parishes, and class distribution by gender. The study focused on adolescents from public schools in both urban and rural parishes of Portoviejo Canton, Manabí Province, Ecuador. Participants included students in the tenth grade and the first three years of high school.

2.2. Participants

The study included 700 female adolescents, aged 14 to 18 years, from public schools in various urban and rural areas of Portoviejo Canton, Manabí Province, Ecuador. All of them had been in a romantic relationship lasting more than a month. The initial sample consisted of a total of 1519 participants; the inclusion criteria for this study were being female and having had at least one romantic relationship lasting more than a month. The final sample size of this research, after application of the above criteria, was n = 700.

To determine the sample, guidelines from Ecuador’s official bodies were followed, specifically those from the National Council for the Control of Narcotic and Psychotropic Substances, now known as the Technical Secretariat on Drugs (SETED). These parameters are standard for studies conducted across the continent. The target population comprised public educational units with students in the specified grades. The educational units were selected from the Ministry of Education Zonal Coordination 4 district 13D01 database. These units are located in the canton, within designated parishes, and are categorized by urban and rural zones. They represent various socio-demographic areas of Portoviejo Canton.

To calculate the sample size, a confidence level of 0.95 and a margin of error of 0.01 were used. Due to the sampling characteristics, a correction factor was applied to account for the design effect, which was estimated at 2, to increase the sample size and reduce observation variability. Additionally, the sample size was further increased by 10% to account for potential non-responses.

2.3. Procedure

Data collection was conducted for paper, with researchers present in the classroom to explain aspects related to the purpose of the assessment, the importance of honesty in responses, and the complete confidentiality of the data and to address any questions before collecting the completed questionnaires. Efforts were made to create an atmosphere of trust, security, and privacy to encourage adolescents to respond sincerely. Participants were asked to remain silent, refrain from discussing the questions or their answers with peers, and maintain some physical distance from one another to ensure privacy while responding. As a token of appreciation, participating students received a pencil and a certificate of participation.

2.4. Ethical Approval and Informed Consent

This study adhered strictly to the ethical standards for psychological research. It was conducted in line with the Declaration of Helsinki and received approval from both the Ministry of Education of Portoviejo (Ecuador) and the academic committee of the doctoral program in Social Work at Universidad Complutense de Madrid (Spain).

Informed consent forms were signed by the parents or guardians of the adolescents who participated in the study. They were informed about the completion guidelines, the duration and nature of the questions, the voluntary participation, and the confidentiality and anonymity of the data, and they were provided with the contact details of the researchers.

2.5. Variables and Instruments

2.5.1. Intimate Partner Physical Violence

The relevant dimension of the Scale of Violence Received, Exercised, and Perceived in Dating Relationships of Young People and Adolescents (VREP) was used [26]. The VREP evaluates five types of violence (physical, sexual, psychological social, psychological humiliation–coercion, and psychological control–jealousy) and three aspects of violence (exerted, received, and perceived) through 28 items. The physical violence subscale used in this study includes 5 items with a Likert-type response format offering 6 options (0: “Never”, 1: “Once”, 2: “2 to 5 times”, 3: “6 to 10 times”, 4: “11 to 15 times”, and 5: “More than 15 times”). Examples of items include: “I pinched him with the intention of hurting him” and “I have deliberately bitten or pulled his hair”. A higher score on these items indicates a greater presence of violence, whether received, perpetrated, or perceived. Only violence received and perpetrated was taken into account in this study.

Participants needed to have had at least one relationship lasting more than a month to complete the scale. They were asked to indicate whether and how often the described situations had occurred or were occurring in their relationships. In the original study, the scale had a Cronbach’s alpha coefficient of 0.99 for violence exerted and 0.91 for violence received. In our study, the Cronbach’s alpha coefficient for the physical violence subscale was 0.94 (violence perpetrated) and 0.89 (violence received).

2.5.2. Impulsivity

A Spanish translation [27,28] of the BIS-11-A [29] was used. The Barratt Impulsivity Scale (BIS-11) measures impulsivity in adolescents and adults, allowing the measurement of the three dimensions of impulsivity—attention, motor, and non-planning—together with a global measure of impulsivity.

The BIS-11-A consists of 30 items with a Likert-type scale in which participants report the frequency of different behaviors (1: rarely or never, 2: occasionally, 3: often or 4: almost always or always). The total score ranged from 30 to 120, and the internal consistency with adolescents was good (α = 0.78) [28]. In our study, Cronbach’s alpha was 0.81.

2.5.3. Emotional Dysregulation

The Spanish version [30] of the Difficulties in Emotional Regulation Scale (DERS) [24] was used. The DERS is an instrument designed to measure emotional dysregulation developed by Gratz and Roemer (2004) [31], and it consists of 36 items that assess multiple aspects of emotional regulation. Participants respond to each item using a 5-point Likert scale ranging from “Almost never” to “Almost always”. The DERS measures six specific dimensions of emotional dysregulation. These include non-acceptance of emotional responses: difficulty accepting negative emotions; difficulty achieving goals: difficulty concentrating and accomplishing goals when emotionally upset; impulse: difficulty controlling impulsive behaviors when experiencing negative emotions; lack of emotional awareness: lack of attention to emotions; limited access to emotional regulation strategies: belief that there are no effective strategies for regulating emotions; emotional clarity: difficulty in identifying one’s emotions. Cronbach’s alphas in our study ranged from 0.79 (for emotional clarity and 0.91 for non-acceptance of emotional responses).

2.6. Statistical Analysis

In the descriptive analysis of the sample, suitable summary statistics were applied. Categorical variables were displayed using absolute (n) and relative (%) frequencies, while continuous variables were summarized with the mean and standard deviation. To examine the association between categorical variables, a Chi-square test of independence was conducted. Considering the sample size and following the recommendations of previous examples from the literature [32,33], we decided to conduct parametric tests. It has been pointed out that normality tests, such as the Kolmogorov–Smirnov test, are very sensitive to sample size. In large samples, these tests can detect minimal deviations from normality that are not relevant to the analysis [34]. Therefore, it seems common to find non-normal distributions in large samples and, according to previous literature, this does not necessarily invalidate the use of parametric tests, especially if the robustness of these tests and the sample size are considered [34,35].

Student’s t-test and one-way ANOVA were used to determine if there was a statistically significant difference between the means of two groups or of three or more independent groups, respectively. In the case of statistically significant differences in one-way ANOVA, the post hoc Scheffé test was applied in order to find out between which groups the significant differences occurred. Pearson’s correlation was employed to investigate the relationship between two continuous variables.

For the mediation analysis, the SPSS macro PROCESS (model 4) was utilized. Following Hayes’ recommendations [36], the regression coefficients were presented in non-standardized form, as standardized coefficients typically lack substantive interpretive value. The models tested included impulsivity as the predictor (X), lack of emotional regulation strategies as mediators (M), and intimate partner physical violence as the dependent variable. All analyses were carried out using SPSS version 28 (IBM Corp, 2021).

3. Results

3.1. Socio-Demographic Characteristics of Participants

Table 1. Socio-demographic characteristics of participants and their families.

Age, Mean (SD)	15.64 (1.20)
Areas, n (%) Rural Urban	245 (35.1) 455 (64.9)
Cohabitation Living with their father, n (%) Living with their mother, n (%) Living with both parents, n (%) Living with your father’s partner, n (%) Living with your mother’s partner, n (%) Living with siblings, n (%) Living with your parent’s partner’s children, n (%) Living with grandparents, n (%) Living with their own partners, n (%) Living alone, n (%)	499 (71.4) 608 (87) 476 (68.1) 24 (3.4) 70 (10) 523 (74.8) 55 (7.9) 180 (25.8) 68 (9.7) 4 (0.6)
Mother’s employment status Works alone on housework, n (%) Working outside the home, n (%) Unemployed, n (%) Pensioner or retired, n (%) Deceased, n (%) I don’t know, n (%)	483 (69.3) 175 (25.1) 16 (2.3) 3 (0.4) 6 (0.9) 17 (2)
Father’s employment status Works alone on housework, n (%) Working outside the home, n (%) Unemployed, n (%) Pensioner or retired, n (%) Deceased, n (%) I don’t know, n (%)	40 (5.8) 489 (70.6) 51 (7.4) 13 (1.9) 24 (3.5) 83 (10.9)
Mother’s level of education * No studies, n (%) Level 1 and 2. Primary education, n (%) Level 3. Secondary education, n (%) Level 4. Post-secondary non-tertiary, n (%) Level 7. University studies, n (%) I don’t know, n (%)	139 (20) 182 (26.2) 113 (16.3) 84 (12.1) 64 (9.2) 117 (16.3)
Father’s level of education * No studies, n (%) Level 1 and 2. Primary education, n (%) Level 3. Secondary education, n (%) Level 4. Post-secondary non-tertiary, n (%) Level 7. University studies, n (%) I don’t know, n (%)	110 (15.9) 176 (25.5) 112 (16.2) 79 (11.4) 52 (7.5) 170 (23.3)
Participants combining work and studies, n (%)	91 (13)

* Level of education was organized according to The International Standard Classification of Education (ISCED, 2011).

shows the socio-demographic characteristics of the participants. The sample consisted of a total of 700 adolescent girls. All of the adolescent girls were born in Ecuador. As can be seen in Table 1, the average age was 15.64 years, ranging from 14 to 18 years of age. Most of the participants (64.9%) were from schools in urban areas in Portoviejo Canton, Manabí Province, Ecuador.

Almost three quarters lived with their father, and 87% lived with their mother. Nearly 70% of the participants lived with both parents. Three quarters of the participants lived with siblings at home. About 25% of the adolescents lived at home with their grandparents, and approximately 10% lived with their partners.

Most mothers were housewives and did not work outside the home (69.3%). Most fathers (70.6%) worked outside the home.

Regarding the mothers’ levels of education, most of them had primary (26.2%) or secondary school education (16.3%), and 20% of mothers had no education. Regarding the fathers’ levels of education, most of them had primary (25.5%) or secondary school education (16.2%), and 15.9% of fathers had no education.

In the last month, 13% (n = 91) of the sample had combined work and studies.

3.2. Descriptive Data and Correlations Between Variables of Interest

Table 2. Descriptive data and correlations between variables of interest.

	Mean (SD)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
(1) Violence Exercised	10.85 (16.58)	0.76 **	0.20 **	−0.03	0.17 **	0.11 *	0.07	0.08	0.12 **
(2) Violence Received	11.96 (19.16)		0.14 **	−0.02	0.16 **	0.10 *	0.07	0.07	0.11 *
(3) Impulsivity	65.60 (8.69)			0.04	0.43 **	0.24 **	0.35 **	0.31 **	0.36 **
(4) DERS Awareness	18.47 (5.20)				0.24 **	0.39 **	0.31 **	0.09	0.29 **
(5) DERS Impulse	14.72 (5.34)					0.54 **	0.60 **	0.40 **	0.70 **
(6) DERS Non Accept.	16.50 (7.25)						0.57 **	0.30 **	0.63 **
(7) DERS Goals	13.86 (4.06)							0.36 **	0.59 **
(8) DERS Clarity	13.08 (3.79)								0.37 **
(9) DERS Strategies	16.93 (5.74)

* p < 0.05, ** p < 0.01; DERS Awareness: lack of attention to emotions; DERS Impulse: difficulty controlling impulsive behaviors when experiencing negative emotions; DERS Non-acceptance: difficulty accepting negative emotions; DERS Goals: difficulty concentrating and accomplishing goals when emotionally upset; DERS Clarity: difficulty in identifying one’s emotions; DERS Strategies: limited access to emotional regulation strategies, belief that there are no effective strategies for regulating emotions.

shows the descriptive data of the variables of interest as well as the correlations between them. As can be seen, both physical violence exerted and received maintained significant positive correlations with impulsivity and with maladaptive strategies, with the exception of DERS Awareness, DERS Goals, and DERS Clarity. Physical violence exerted and physical violence received maintained high positive correlations with each other. Impulsivity maintained significant positive correlations with all emotional dysregulation strategies except for DERS Awareness.

As can be seen in Table 2, the highest correlation was observed between exercised violence and received violence (0.76). Impulsivity maintained the highest correlations with DERS Impulse, DERS Strategies, and DERS Goals (between 0.43 and 0.35). The values of the correlations of violence exercised and violence received with emotional dysregulation strategies were very low (all less than 0.20).

3.3. Relationships Between Socio-Demographic Variables and Variables of Interest

In relation to age, negative correlations were observed, although with a very low effect, between age and lack of regulation based on impulse (r = −0.074, p = 0.050), lack of regulation due to lack of emotional clarity (r = −0.082, p = 0.030), and lack of emotional regulation based on absence of strategies (r = −0.083, p = 0.029).

Regarding the status of living with the father, no statistically significant differences were observed in the variables of interest. Tendencies to significance were only observed in the case of lack of emotional clarity (p = 0.070) and lack of regulation strategies (p = 0.051). In both cases, adolescents who lived with their fathers had lower scores in these variables; that is, they had fewer difficulties in their emotional regulation. Regarding the status of living with the mother, no statistically significant differences were observed in the variables of interest.

Concerning living with siblings in the home, statistically significant differences were observed in violence exercised (p = 0.029). Specifically, those adolescents who lived with siblings had lower scores (mean = 10.05, SD = 15.41) than those who did not live with siblings (mean= 13.21, SD = 19.52) in violence exercised. Considering the presence of grandparents in the home, statistically significant differences were found in the lack of emotional regulation due to insufficient emotional awareness (p = 0.035) and insufficient emotional acceptance (p = 0.028). Specifically, adolescents who lived with their grandparents had lower scores (mean = 17.77, SD = 5.06 for lack of emotional awareness; mean = 16.15, SD = 7.21 for lack of emotional acceptance) than those who did not live with grandparents (mean = 18.71, SD = 5.23 for lack of emotional awareness; mean = 17.52, SD = 7.29 for lack of emotional acceptance).

No statistically significant differences were observed in the variables of interest in those adolescents who lived with their own partners compared to those who did not.

Regarding the employment situation, no statistically significant differences were observed in any of the variables of interest regarding the employment situation of the mother, nor regarding the employment situation of the father. Regarding the level of education, there were no statistically significant differences in any of the variables of interest regarding the mother’s studies or regarding the father’s studies.

Regarding the status of combining work with studies, statistically significant differences were observed in impulsivity (p = 0.017). Specifically, those adolescents who combined work with studies had higher scores in impulsivity (mean = 67.64, SD = 8.37) than those who were only studying (mean = 65.30, SD = 8.71).

3.4. Mediation Analysis: Influence of Impulsivity on Intimate Partner Physical Violence Through Lack of Emotional Regulation

Figure 1 shows the proposed simple mediation model. The model is based on impulsivity as the predictor variable and intimate partner violence (with violence perpetrated and violence received as different variables) as the outcome variable. In this relationship, the different emotional dysregulation strategies are proposed as mediating variables.

Figure 1 proposed conceptual model: effect of impulsivity on physical partner violence through emotional dysregulation.

The preconditions established for the mediation models (significant correlations between predictor and mediator, between mediator and consequent, and between predictor and consequent) show the following possible mediation models:

(a)

For physical violence received, the following mediation models (meeting the preconditions) are proposed: impulsivity–DERS Impulse–received violence; impulsivity–DERS Non-acceptance–received violence; impulsivity–DERS Strategies–received violence.

(b)

For physical violence perpetrated, the following mediation models (meeting the preconditions) are proposed: impulsivity–DERS Impulse–perpetrated violence; impulsivity–DERS Non-acceptance–perpetrated violence; impulsivity–DERS Goals–perpetrated violence; impulsivity–DERS Strategies–perpetrated violence.

After testing the 7 previous mediation models (3 for violence received and 4 for violence perpetrated), only 4 of them were statistically significant for mediation. Specifically, for both received and perpetrated physical violence, DERS Impulse and DERS Non-acceptance were shown to be significant mediators between impulsivity and violence. In all cases, mediation was partial. The following tables (Table 2,

Table 3. Mediation model: regression of impulsivity on violence received through lack of regulation of impulsivity (emotional dysregulation) as mediator (n = 700).

Regression of Impulsivity on Violence Received Through Lack of Regulation of Impulsivity (DERS Impulsivity)
Outcome variable: DERS Impulsivity
Impulsivity	B (SE) 0.26 (0.02)	t 12.67	p <0.001	[LLCI–ULCI] [0.225/0.308]
R = 0.44	R2 = 0.19	F = 33.52	p < 0.001
Outcome variable: Violence received
	B (SE)	t	p	[LLCI–ULCI]
Impulsivity	0.28 (0.09)	3.14	<0.001	[0.108/0.467]
M: DERS Impulsivity	0.36 (0.14)	2.43	0.015	[0.069/0.653]
Age	−0.39 (0.60)	−0.646	0.518	[−1.56/0.79]
Siblings coexistence	−2.03 (1.65)	−1.23	0.218	[−5.27/1.20]
Grandparents coexist.	0.33 (1.63)	0.205	0.837	[−2.87/3.54]
Working	2.28 (2.13)	1.069	0.285	[−1.91/6.47]
Model summary	R = 0.21	R2 = 0.04	F = 5.35 p < 0.001
Direct Effect of X on Y	Eff. (SE) 0.28 (0.09)	t 3.14	p 0.001	[LLCI–ULCI] [0.108/0.467]
Indirect effect of X on Y DERS Impulsivity	Eff. (SE) 0.09 (0.04)			[LLCI–ULCI] [0.016/0.177]

B: unstandardized beta coefficient; (SE): standard error; [LLCI–ULCI]: lower-level confidence interval–upper-level confidence interval; Eff.: effect size; M: mediator.

,

Table 4. Mediation model: regression of impulsivity on violence received through non-acceptance (emotional dysregulation) as mediator (n = 700).

Regression of Impulsivity on Violence Received Through Non-Acceptance (DERS Non-Acceptance)
Outcome variable: DERS Non-acceptance
Impulsivity	B (SE) 0.18 (0.03)	t 5.93	p <0.001	[LLCI–ULCI] [0.123/0.245]
R = 0.24	R2 = 0.06	F = 8.57	p < 0.001
Outcome variable: Violence received
	B (SE)	t	p	[LLCI–ULCI]
Impulsivity	0.34 (0.08)	4.01	<0.001	[0.172/0.505]
M: DERS Non-acceptance	0.26 (0.10)	2.53	0.011	[0.057/0.456]
Age	−0.48 (0.60)	−0.798	0.425	[−1.65/0.69]
Siblings coexistence	−2.03 (1.65)	−1.23	0.218	[−5.28/1.21]
Grandparents coexist.	0.12 (1.64)	0.076	0.939	[−3.09/3.34]
Working	2.44 (2.13)	1.142	0.253	[−1.75/6.63]
Model summary	R = 0.21	R2 = 0.04	F = 5.46	p < 0.001
Direct Effect of X on Y	Eff. (SE) 0.34 (0.08)	t 4.01	p <0.001	[LLCI–ULCI] [0.173/0.505]
Indirect effect of X on Y DERS Non-acceptance	Eff. (SE) 0.05 (0.02)			[LLCI–ULCI] [0.011/0.091]

B: unstandardized beta coefficient; (SE): standard error; [LLCI–ULCI]: lower-level confidence interval–upper-level confidence interval; Eff.: effect size; M: mediator.

and

Table 5. Mediation model: regression of impulsivity on violence exercised through lack of regulation of impulsivity (emotional dysregulation) as mediator (n = 700).

Regression of Impulsivity on Violence Exercised Through Lack of Regulation of Impulsivity (DERS Impulsivity)
Outcome variable: DERS Impulsivity
Impulsivity	B (SE) 0.27 (0.02)	t 12.671	p <0.001	[LLCI–ULCI] [0.225/0.308]
R = 0.44	R2 = 0.20	F = 33.529	p < 0.001
Outcome variable: Violence exercised
	B (SE)	t	p	[LLCI–ULCI]
Impulsivity	0.36 (0.07)	4.68	<0.001	[0.211/0.516]
M: DERS Impulsivity	0.37 (0.12)	2.92	0.003	[0.121/0.617]
Age	−0.574 (0.51)	−1.127	0.260	[−1.57/0.43]
Siblings coexistence	−3.25 (1.40)	−2.32	0.020	[−6.00/−0.51]
Grandparents coexistence	−1.78 (1.38)	−1.28	0.198	[−4.51/0.939]
Working	0.389 (1.81)	0.215	0.829	[−3.17/3.94]
Model summary	R = 0.28	R2 = 0.08	F = 10.112	p < 0.001
Direct Effect of X on Y	Eff. (SE) 0.36 (0.08)	t 4.68	p <0.001	[LLCI–ULCI] [0.211/0.516]
Indirect effect of X on Y DERS Impulsivity	Eff. (SE) 0.10 (0.03)			[LLCI–ULCI] [0.031/0.175]

B: unstandardized beta coefficient; (SE): standard error; [LLCI–ULCI]: lower-level confidence interval–upper-level confidence interval; Eff.: effect size; M: mediator.

) show the results obtained from these mediation models. Taking into account the significant bivariate relationships found, age, living with siblings, living with grandparents, and combining work and studies were considered as covariates.

3.4.1. Physical Violence Received

Table 3 shows the results of the effect of impulsivity on violence received through DERS Impulsivity as a dysfunctional emotional regulation strategy. As the results show, DERS Impulsivity plays a mediating role between impulsivity and violence received. Similarly, there is a direct effect between impulsivity and violence received (p < 0.001). None of the covariates considered have significant effects on the model. Impulsivity is positively related to DERS Impulsivity (B = 0.26, p < 0.001), which in turn contributes to an increase in violence received (B = 0.36, p = 0.015).

Table 4 shows the results of the effect of impulsivity on violence received through DERS Non-acceptance as a dysfunctional emotional regulation strategy. As the results show, DERS Non-acceptance plays a mediating role between impulsivity and violence received. Similarly, there is a direct effect between impulsivity and violence received (p < 0.001). None of the covariates considered have significant effects on the model. Impulsivity is positively related to DERS Non-acceptance (B = 0.18, p < 0.001), which in turn contributes to an increase in violence received (B = 0.26, p = 0.011).

3.4.2. Physical Violence Perpetrated

Table 5 shows the results of the effect of impulsivity on violence exercised through DERS Impulsivity as a dysfunctional emotional regulation strategy. As the results show, DERS Impulsivity plays a mediating role between impulsivity and violence exercised. Similarly, there is a direct effect between impulsivity and violence exercised (p < 0.001). Living with siblings is shown to be a significant covariate on the model (p = 0.020), particularly as a buffer against violence. Impulsivity is positively related to DERS Impulsivity (B = 0.27, p < 0.001), which in turn contributes to an increase in violence exercised (B = 0.37, p = 0.003).

Table 6. Mediation model: regression of impulsivity on violence exercised through non-acceptance (emotional dysregulation) as mediator (n = 700).

Regression of Impulsivity on Violence Exercised Through Non-Acceptance (DERS Non-Acceptance)
Outcome variable: DERS Non-acceptance
Impulsivity	B (SE) 0.18 (0.03)	t 5.938	p <0.001	[LLCI–ULCI] [0.123/0.245]
R = 0.24	R2 = 0.06	F = 8.573	p < 0.001
Outcome variable: Violence exercised
	B (SE)	t	p	[LLCI–ULCI]
Impulsivity	0.42 (0.07)	5.87	<0.001	[0.281/0.563]
M: DERS Non-acceptance	0.23 (0.09)	2.65	0.008	[0.060/0.398]
Age	−0.675 (0.51)	−1.32	0.185	[−1.67/0.326]
Siblings coexistence	−3.24 (1.40)	−2.30	0.021	[−5.99/−0.48]
Grandparents coexist.	−1.96 (1.39)	−1.40	0.160	[−4.69/0.778]
Working	0.389 (1.81)	0.215	0.829	[−3.17/3.94]
Model summary	R = 0.28	R2 = 0.08	F = 9.85	p < 0.001
Direct Effect of X on Y	Eff. (SE) 0.42 (0.07)	t 5.87	p <0.001	[LLCI–ULCI] [0.281/0.563]
Indirect effect of X on Y DERS Non-acceptance	Eff. (SE) 0.04 (0.02)			[LLCI–ULCI] [0.008/0.082]

B: unstandardized beta coefficient; (SE): standard error; [LLCI–ULCI]: lower-level confidence interval–upper-level confidence interval; Eff.: effect size; M: mediator.

shows the results of the effect of impulsivity on violence exercised through DERS Non-acceptance as a dysfunctional emotional regulation strategy. As the results show, DERS Non-acceptance plays a mediating role between impulsivity and violence exercised. Similarly, there is a direct effect between impulsivity and violence exercised (p < 0.001). Living with siblings is shown to be a significant covariate on the model (p = 0.021), particularly as a buffer against violence. Impulsivity is positively related to DERS Impulsivity (B = 0.18, p < 0.001), which in turn contributes to an increase in violence exercised (B = 0.23, p = 0.008).

4. Discussion

Adolescent dating violence is a widespread problem with deleterious effects that can be sustained in later life. However, the relationships between some personality traits associated with violence, such as impulsivity, and the type of strategies employed to buffer the effects of the perceived stress of staying in a dating relationship are still unclear. This is why this study addressed important gaps in the literature on the reasons why adolescents maintain intimate partner relationships in which they experience or exert violence towards theirs partners. To this end, an explanatory model was provided that sought to explore the predictive effect of impulsivity on the violence perpetrated and received, attending to the mediating role of emotional strategies in adolescent females. The findings emphasize the importance of paying attention to these variables with the intention of preventing violence, in any of its manifestations, in adolescent couple relationships.

Taking an interactionist approach to the construction of couple relationships, the results showed that exercised and received partner violence maintained similar mean scores in the participants. As for psychological variables, mean scores were observed on the impulsivity scale, and coping strategies focused on difficulty paying attention or invalidation of emotions (non-awareness), and strategies based on non-acceptance of unpleasant emotions were the most predominant in buffering the effects of exerted and perpetrated violence. These data coincide with statistics collected in previous studies conducted in Spain in which it has been confirmed that the perpetration of violence by adolescent females was carried out by means of controlling behaviors and/or verbal–emotional aggression, while boys tended to exercise relational and/or sexual violence [37,38,39]. These data are important, since the manifestations of violent behavior in adolescents allow us to take a step further in understanding the mechanisms of couple relationships and to explore the ways of coping with this stressor.

In addition to socioeconomic and cultural aspects, it is important to emphasize that prevalence rates of dating and intimate partner violence can be sensitive to a variety of factors that bias the interpretation of the results. For example, the selection of participants, the type of instrument used in the assessment, the method of data collection, the operationalization of variables, and social desirability bias make the disclosure of sensitive data on the private lives of minors available to the reader. Even so, these data are part of a question to be further investigated in future longitudinal studies to understand behavioral patterns, both within a relationship and during a breakup.

Few studies have focused on the role of impulsivity in the prediction of dating violence received and perpetrated by adolescent females. The present study makes a significant contribution, finding answers to this question in the significant correlations found in the variables of interest. First, as confirmed in Hypothesis 1, the tendency to impulsivity in the study participants predicts both perpetrated and received dating violence. This result follows the line proposed by previous studies in which it has been emphasized that the lack of self-regulation resources is associated with aggressive behaviors towards the partner. Specifically, behaviors associated with sensation seeking without thinking about the consequences of their actions increased the predisposition to behave aggressively towards an intimate partner [40]. According to the model proposed by Whiteside and Lyam [41], impulsivity is made up of four factors: negative urgency, understood as the tendency to act immediately when unpleasant emotions arise; lack of premeditation, which refers to the lack of reflection on the consequences of a given behavior; lack of perseverance, that is, the lack of concentration to stick to difficult tasks; and, finally, sensation seeking, understood as the tendency to enjoy and seek activities that are exciting. Based on this model and the results found in the present study, it is possible that the adolescent participants may have developed impulsive patterns tending to minimize the impact of unpleasant emotions and maximize pleasant emotions, despite being subjected to stressors that favor the perception of intimate partner violence.

Second, and in line with the tendency to minimize unpleasant emotions, Hypothesis 2 is confirmed given the significant associations found between impulsivity and the violence perpetrated and suffered, taking into account the role played by the coping emotional strategies used by the participants to buffer the perceived stress of the violence perpetrated and received. Specifically, it is the maladaptive or avoidance coping strategies that maintain significant relationships with the rest of the psychological variables that make up the proposed model. According to previous literature, adolescents who consider themselves victims of intimate partner violence more frequently use avoidance strategies compared to adolescents who do not report victimization experiences [42,43]. This may result in learned helplessness as well as the development of later psychological conditions such as anxiety or depression in adulthood [17].

Regarding family factors and addressing Hypothesis 3 on adolescent cohabitation, the environment can influence adolescents’ experiences in their dating relationships [23,24]. As indicated by various studies conducted in Ecuador, it is highlighted that the younger the age at which adolescents begin cohabiting with their partners, the higher the risk of experiencing violence in any form [1,2,3,4]. In the results of the present study, it was found that living with parents has no effect on violence or emotional regulation. No differences were observed between adolescents living with their own partners compared to those who do not. However, cohabitation with siblings has acted as a significant covariate in the tested model and is highlighted as a protective factor against partner violence. Furthermore, adolescents living with their grandparents have lower scores in emotional awareness and emotional acceptance deficits, meaning that they seem to have better emotional regulation strategies compared to those who do not live with their grandparents. This would lead us to partially accept Hypothesis 3. It is worth noting that the literature has not been able to provide an objective answer on the effect of living with parents or grandparents on the risk of perpetrating or experiencing violence in dating relationships. Specifically, it has been shown that family communication style, emotional support, or romantic relationships between parents can help adolescents establish healthier or less-healthy relationships [44,45].

Concerning the discrepancies between the data from this study on cohabitation with one’s partner and previous data from national surveys in Ecuador, they can be explained from several perspectives. On one hand, the differences in study design and methodology may play a role, as the data from INEC [2,4] were collected on a large scale, which allows for the identification of more general population patterns. However, in this study, with a more specific sample of adolescents, the results may be more varied and reflect a less generalized reality. In fact, only 68 participants (9.7%) lived with their partners. On the other hand, the age and emotional maturity of adolescents may influence the results, as during adolescence, young people are still learning to manage emotions and handle interpersonal relationships [46], so the lack of emotional skills and conflict resolution can contribute to violence in romantic relationships. However, in the sample of this study, the adolescents may still be in the process of adaptation or maturation in which their cohabitation with their partner has not yet reached a critical level to generate violence. Family protection factors should also be highlighted, as, as found in this study, the proximity of grandparents and siblings can act as protective factors, which may have influenced the results. It is possible that adolescents living with their partners but who also maintain a close relationship with other family members have more emotional support and resources to manage conflicts in their relationship, thereby reducing the risk of violence [47]. Lastly, another factor that could have influenced the results is the socioeconomic and cultural status of the adolescents analyzed, as gender violence in romantic relationships is strongly influenced by the socioeconomic and cultural conditions of the region. In many cases, national surveys reflect a broader picture of economic inequality, social exclusion, and cultural factors that perpetuate gender violence [48].

A novelty contributed by this study is that not only are coping strategies based on non-acceptance and lack of impulse regulation are related to the violence suffered, but they are also related to the violence perpetrated by the participants, taking into account the predictive role of impulsivity. There is currently no literature available to compare these results, as the present study is the first to investigate the mediating role of maladaptive coping strategies during adolescence in relation to dating violence, taking into account impulsivity. The simple mediation analysis verifies that the disconnection with unpleasant emotions and the lack of acceptance of them leads to any type of violence towards the partner, since there is no reflection on the consequences of one’s actions. Despite the lack of literature, the results are consistent with the model of emotion evaluation proposed by Gross [49], which emphasizes that emotions involve a process of evaluation of perceptions about the world and perceptions of the desired world; to regulate the discrepancies found, action responses are given rise to. In line with studies on violence and emotional regulation, it is suggested that the inability to effectively manage unpleasant emotions, such as anger, predicts violence in young couples [50].

In summary, as to the results that confirm the second hypothesis of our study, in adolescents, we observed that attempts to regulate adverse emotional states by means of lack of contact with emotions, lack of identification, and even lack of acceptance of them, cause precisely the opposite effect by dysregulating behavior in the couple’s relationship. In other words, these strategies, rather than attempting to reduce the levels of violence, increase them. Although it is necessary to interpret these results with caution, it is possible that these early experiences in dating relationships may consolidate into stable relational patterns in adulthood, as supported by the literature that addresses these issues [15,16,17].

One of the strengths of this study is the detection of the problem of dating violence among adolescent women who not only suffer but also perpetrate violence against their partners. To further advance the understanding of this social phenomenon, psychological factors have been included to help predict the behavioral pattern of a person in a relationship. The contribution of these results can guide the development of educational programs aimed at reducing violence in this vital stage. This prevention should be carried out in early stages such as adolescence, since resistance to any intervention will be lower compared to adults [51]. This deconstruction of violence can be conducted through interventions focused on reflecting on new models of healthy couples by working on distortions about control in the couple or delegitimizing violence as a means to conquer the partner [52]. It should also be borne in mind that, although our research has not investigated them, social and contextual aspects have to be taken into account. In particular, family relationships, family stress, and the possible presence of childhood trauma should be the object of assessment, prevention, and intervention because of the role they can play in violent intimate partner relationships [25]. Another aspect of interest and to be addressed in future research is the permanence in the situation of intimate partner violence. Individuals often remain trapped in violent relationships due to a combination of fear, emotional attachment, financial dependency, and lack of support networks [53,54]. Factors that can contribute to reducing such situations include promoting healthy relationship skills, providing economic support, and ensuring access to victim-centered services [55]. Addressing these elements can empower individuals to leave abusive environments and build safer lives [55].

Despite the practical implications mentioned above, this study has certain methodological limitations. First, the cross-sectional design prevents us from drawing conclusions on cause and effect between the variables of the model proposed. Second, the sample has been collected in the same geographical context and shares the same nationality. It should be borne in mind that previous studies have found significant relationships between culture and dating violence [56], so it would be interesting for future studies to analyze possible cultural differences under a longitudinal design. Another limitation found is the lack of contemplation of different sexual orientations. Finally, in the digital era into which society is developing, the use of social networks is predominant in the formation of couples. In the study of teen dating violence, it would be interesting to include the effect of technology as one of the most frequent socialization environments among young people.

5. Conclusions

Given the lack of knowledge about coping through emotional regulation strategies with the stress underlying an adolescent dating relationship and impulsivity, the present study presents a series of empirical findings that support the role of these variables in the violence perpetrated and suffered among adolescent girls. In this regard, it is necessary to bear in mind that, although violence is exercised and perpetrated by both genders, it can manifest itself in different ways [57]. Research on intimate partner violence among adolescents has addressed both males and females, although most studies tend to focus more on the experiences of females as victims. This is partly due to the higher prevalence and severity of violence experienced by females compared to males [58]. Studies indicate that verbal and emotional violence is common in both genders, while physical and sexual violence tends to be reported more by women as victims [59]. On the other hand, this study highlights the role of the cohabitation variable, specifically with grandparents and siblings, acting as a protective factor against gender violence and facilitating better emotional regulation tools. Attention should be paid to socio-demographic variables such as cohabitation, as they can influence the initiation and maintenance of violent relationships [2,3].

Our findings, focusing specifically on a sample of adolescent females, support what has come to be called bidirectional violence, i.e., when both partners perpetrate and experience violence [60]. Specifically, our results confirm that suppression-based coping strategies (non-awareness of impulses and non-acceptance of unpleasant emotions) mediate the association between high impulsivity and experienced and perpetuated dating violence. Fostering egalitarian relationships between adolescents and associating them with coping strategies that enhance pleasant emotions may help to reduce this problem that has its point of origin in adolescence.