The role of emotional dependence in the relationship between attachment and impulsive behavior

Título: El papel de la dependencia emocional en la relación entre el apego y la conducta impulsiva. Resumen: La dependencia emocional se define como una necesidad afectiva extrema que una persona siente hacia su pareja. Se ha relacionado con el déficit de control de impulsos y con el estilo de apego preocupado. El presente estudio tiene como objetivos analizar la relación entre la dependencia emocional, el apego y la conducta impulsiva, así como, analizar si estas dos últimas predicen la dependencia emocional. Además, se analiza si la dependencia emocional media en la relación entre el apego y la impulsividad y, finalmente, se estudian las diferencias de género en las tres variables mencionadas. La muestra está formada por 1533 adolescentes (707 mujeres y 826 hombres), con edades que oscilan entre los 14-18 años (M = 15.76; DT = 1.25). Los resultados muestran relaciones positivas entre la dependencia emocional y la permisividad parental, traumatismo infantil e impulsividad. Asimismo, la impulsividad se relaciona de manera positiva con evitar estar solo, relación asimétrica, necesidad de agradar, permisividad parental y traumatismo infantil. Igualmente, se comprueba como la permisividad parental, autosuficiencia y rencor contra los padres y el traumatismo infantil predicen la dependencia emocional. Finalmente, se observa como la relación entre el apego y la conducta impulsiva es mediada por la dependencia emocional y se aprecian diferencias en las tres variables en función del género. Palabras clave: dependencia emocional; apego; conducta impulsiva; adolescentes. Abstract: The emotional dependence has been related with the problems in the impulse control and with the preoccupied attachment style. The objectives of this study are to analyze the relation between the emotional dependence, attachment and the impulsive behavior, as well as, to analyze if these final two can predict the emotional dependence. Furthermore, it analyze if the emotional dependence mediate in the relation between attachment and impulsivity and, finally, it studies the differences of gender in the three mentioned variables. The sample is formed by 1533 teenagers (707 women and 826 men), with ages between 14-18 years old (M = 15.76; SD = 1.25). The results showed of positives relations between the emotional dependence and the parental permission, childhood trauma and impulsive behavior. Moreover, the impulsivity is positively related with the fact to avoid to be alone, asymmetric relation, the need to please others, parental permission and childhood trauma. In the same way, it proves how the parental permission, self-sufficiency and resentment against parents and childhood trauma can predict the emotional dependence. Finally, it observes how the relation between the attachment and the impulsive behavior is mediated with the emotional dependency and it appreciates differences between the three variables in gender basis.


Introduction
Early childhood experiences form the schemes about oneself and others that will serve for the later development of personality (Young, 1999).Early affective deficiencies may be responsible for the etiology of emotional dependence (Castelló, 2012).Many of the first affective experiences in people with emotional dependence are negative (Scantamburlo, Pitchot & Ansseau, 2013).As a result, relationships could be used to try to meet unmet emotional needs in childhood (Lemos & Londoño, 2006;Michelena, 2007;Skvortsova & Shumskiy, 2014).Emotional dependence is defined as an extreme affective need that a person feels towards their partner in the course of their different relationships (Castelló, 2005).Their intense fear of rupture contaminates the emotional bond, making them vulnerable and pathological (Del Castillo, Hernández, Romero & Iglesias, 2015).They usually establish relationships as a couple since adolescence and experience ruptures as catastrophic events (Castelló, 2012).They are also accompanied by loss of iden-tity (Schaeffer, 1998), role asymmetry, possessive behaviors, intense energetic attrition, emotional instability and continuous inner emptiness (Moral & Sirvent, 2008).The fear of rupture is such that they remain in relationships that cause them discomfort in order to retain the couple alongside them, sacrificing their own desires and needs and leading to deterioration in their quality of life (Cubas, Espinoza, Galli & Terrones, 2004).However, despite experiencing negative feelings such as sadness, despair, abrupt emotional changes and irrepressible temptations, they also perceive very intense positive feelings such as passion, immense happiness and reduction of tension (Skvortsova & Shumskiy, 2014).
An important aspect is that the first love relationships during adolescence are positive because they could influence the way relationships behave in adulthood (Pradas & Perles, 2012).In this line, the acquisition of emotional dependence could be explained by inadequate attachment during childhood which would remain stable throughout life (Izquierdo & Gómez-Acosta, 2013).Attachment is considered a basic necessity throughout the entire life cycle (Gómez-Zapiain, Ortiz & Gómez-Lope, 2012), influencing the choice of the couple and the way they act within it (Barroso, 2014;López, 2003).Previous studies have linked emotional dependence to attachment styles in adulthood by pointing to the predominance of preoccupied attachment (Momeñe & Estévez, anales de psicología / annals of psychology, 2018, vol. 34, nº 3 (october) accepted publication; Yárnoz-Yaben, 2010).This style of attachment implies a high anxiety for abandonment, a high desire for closeness, intimacy (Alonso-Arbiol, Shaver & Yárnoz, 2002), commitment (Feeney & Noller, 1990), jealousy and obsessive concern (Feeney & Noller, 2001).In this sense, dependent persons emit behaviours aimed at preventing the couple's abandonment by employing manipulation, which could endanger their lives or threaten to inflict harm (Villegas & Sánchez, 2013).In this line, it has been proved how the type of attachment influences the functional or dysfunctional expression of anger (Mikulincer & Shaver, 2011), being the preoccupied attachment, predominant in people with emotional dependence, prone to experience greater anger and impulsivity (Loinaz, 2011) and inability to regulate it (Loinaz & Echeburúa, 2012).Consequently, both real or imaginary rupture attempts and emotional estrangement (Mayseless, 1991) can disproportionately activate dysfunctional attachment by precipitating aggressive behaviors (Loinaz & Echeburua, 2012).In the same way, attachment figures in childhood play a crucial role in the acquisition of a good capacity for emotional regulation, being the absence of contact and lack of care those who predispose to react more impulsively and aggressively in adulthood (Barroso, 2014).This is what happens in emotional dependents where their attachment figures show negligent behavior, seriously neglecting their obligations, relating in an affectionately poor way and even providing them with ill-treatment (Castelló, 2005).This is how they would obtain, therefore, dysfunctional schemes of insufficient self-control of impulses and emotions and low tolerance to frustration (Jaller & Lemos, 2009).As can be seen, impulsivity plays an important role in substance addictions (Belin, Mar, Dalley, Robins & Everitt, 2008), however, emotionally dependent people due to their similarity with addictive pathologies also show deficits in impulse control (Villegas & Sánchez, 2013), difficulties in emotional regulation (Momeñe, Jáuregui & Estévez, 2017), and a great comorbidity with anxious-depressive symptomatology (Deave & McGraw, 2003).Impulsivity may refer to rapid, unplanned actions where immediate reinforcement is experienced as a matter of urgency (Moeller, Barrat, Dougerty, Schmitz & Swann, 2001), showing an inability to delay or inhibit a response (Papachristou, Nederkoorn, Havermans, Van der Horst & Jansen, 2012), actions expressed prematurely, risky or inappropriate behaviours that are often undesirable for the situation (Evenden, 1999), leading to behavioural disinhibition (Verdejo-García, Lawrence & Clark, 2008).
Because of this, the objectives of this study have been to study the relationship between dependence, attachment and impulsive behavior, as well as to analyze whether the latter two predict emotional dependence.The differences in the variables mentioned above are also analyzed according to gender and, finally, it has been verified whether the average emotional dependence in the relationship between attachment and impulsivity.

Method
It is a non-experimental, cross-sectional, correlational design study.The sampling method used is probabilistic with a stratified-unietapic random sampling type with proportional affixing; the distribution was made according to weight or size within the population.In order to stratify, the following factors were taken into account: the number of educational units, the number of adolescents per class, urban and rural parishes, and the distribution of courses by sex.The reference population is made up of adolescents attending school in the Fiscal Educational Units of the Urban and Rural Parishes of the Portoviejo Canton in the Province of Manabí in the Republic of Ecuador, which encompass different sociodemographic sectors that share common characteristics and constitute the object of study.

Participants
The sample is composed of 1.533 adolescents with schooling of both sexes, 53.9% male (826) and 46.1% female (707), aged between 14 and 18 years (M = 15.76,DT = 1.25) belonging to twelve Fiscal Educational Units of the different urban (60%) and rural (40%) socio-demographic sectors of the Canton Portoviejo of the Province of Manabí of the Republic of Ecuador.In order to obtain the sample, Ecuador's official bodies (National Council for Narcotic and Psychotropic Substances Control, 2005) were considered, and today it is known as the Technical Secretariat for Drugs (SETED).The design of the sample and the determination of its size were defined and calculated following the same criteria as those used in the Report of the Second National Survey of Secondary Education Students on Drug Use (2005) of the Republic of Ecuador conducted by the National Council for Narcotic and Psychotropic Substances Control.These parameters used for the calculation, as referred to by the Council in that report, are those applied to all the studies carried out on the continent.The target population was the fiscal educational units of adolescents enrolled in tenth grade, first, second and third year of high school.The selection of the educational units was obtained from the database of the Ministry of Education Zonal Coordination 4 district 13D01, the educational units belong to the canton, parish established, institution zone (urban and rural), support related to the Fiscal Educational Units and representative of different socio-demographic zones of the Canton Portoviejo.For the calculation of the sample size, the confidence level of the sample and the relationship with the margin of error or variation that always exists between the results obtained in a sample and its inference towards the population were considered.The confidence level used is.95 with a margin of error of .015.Due to the sampling characteristics it is considered a correction factor by design effect to increase sample size and decrease the variability of observations, this factor was estimated at 2. Finally, this sample size was increased to compensate for 10% of possible non-response.
Each educational unit had the probability of selection directly proportional to the number of classes in the Tenth General Basic Education (GBS) and the Unified General Baccalaureate (BGU) (first, second and third years of Bachillerato).In the educational units with a number of classrooms greater than the sampling interval, several classes could be selected.The criteria for the selection of the strata were the representativeness criteria that correspond to the capital of Manabí of the Republic of Ecuador; this study will represent the two parishes (urban and rural) of the different educational units.The sample consisted of: a) The capital of Manabí Portoviejo with two parishes: urban and rural, and b) 12 educational units belonging to the two parishes, assuming a total of 1.533 adolescents belonging to the 12 educational units and parishes.

Procedure
The study was preceded by the provision of informed consent to the parents and/or guardians of the adolescents who completed the questionnaires.They were also informed about the rules of completion, duration and aspects to be measured, voluntarily conducting the study, confidentiality and anonymity of the data obtained and telephone and email addresses of the reference researcher to contact.Since the data collection has been done in paper format, during the administration of the questionnaires the researcher stayed in the classroom with the students until they were all returned to her.It was optional for teachers to remain in the classroom.The students collaborating in the research received a pencil and certificate of participation as a token of their appreciation.
This study was carried out following the criteria of the Declaration of Helsinki (World Medical Association, 2013).
The internal structure of the questionnaire consists of 7 dimensions, five of which refer to the representations of attachment (security: availability and support of attachment figures; family concern; parental interference; selfsufficiency, resentment against parents; and childhood trauma) and two refer to the representations of family structure (parental authority and parental permission).The scale of "security: availability and support of attachment figures" refers to one's own perception of having felt and feeling loved by their attachment figures, as well as being able to trust them and perceive their availability if needed.The scale of "family concern" refers to intense separation anxiety and excessive preoccupation at the present time with attachment figures.The "parental interference" scale refers to memories of being overprotected during childhood, being a fearful child, showing concern about being abandoned.The "parental authority" scale is the person's positive assessment of family values of authority and hierarchy, while the "parental permission" scale shows memories of a lack of limits and parental guidance during childhood.The scale of "selfsufficiency and resentment against parents" refers to a sense of resentment towards loved ones and rejection of feelings of dependence and emotional reciprocity.Finally, the scale of "childhood trauma" refers to memories of having experienced lack of availability, violence and threats from attachment figures during childhood.It is answered through a Likert scale from 1 ("totally disagree") to 5 ("totally agree").The internal consistency values in the Spanish adaptation are satisfactory (the value of Alpha in the different subscales is between .60 and .85).In this study Cronbach's alpha is .90.
Emotional dependence in dating relationships in young people.The Emotional Dependence Scale in the Bridehood of Youth and Adolescents (DEN; Urbiola, Estévez & Iraurgi, 2014).This scale consists of 12 items structured in 4 subscales: (a) Avoiding being alone, refers to the actions carried out by the dependent person in order to avoid being alone; (b) Necessity of exclusivity, refers to the dependent person's need to know at all times that his or her partner is available only for him or her; (d) Asymmetric relationship refers to the subordinate and asymmetrical nature of the relationship.It is answered through a Likert scale of 6 points, from 0 ("never") to 5 ("always"), indicating with which they identify with each other in their relationships.To complete the questionnaire it is necessary to have had at least one dating relationship.In relation to the internal consistency of the DEN, it obtained a Cronbach alpha of .82 in the validation study, being a value of reliability considered as high.In this study the value of Cronbach's alpha is .87.
Impulsivity.Impulsivity scale of BARRAT BIS 11 (Salvo & Castro, 2013).This scale has been designed to evaluate impulsivity considering the construct from a multidimensional point of view.It consists of 30 items that are grouped into three subscales: Cognitive Impulsivity, related to the restlessness of thinking and quick decision making, Motor Impulsivity, implies acting in a hasty way without previous reflection and Unplanned Impulsivity, carrying out actions without planning towards the future.It is answered by a Likert scale of 4 response options (0 "rarely or never"; 1 "occasionally"; 3 "often"; 4 "always or almost always").The internal consistency values in the Chilean adaptation, used here, are acceptable (Cronbach's alpha value for the full scale is .77).In this study Cronbach's alpha is .81.

Statistical analysis
First, bivariate relationships between emotional dependence, attachment and impulsive behavior were analyzed using Pearson's r.The effect size was interpreted as described by Cohen (1992) in the following way: values under .20 were considered small, those around .50 medium and those higher anales de psicología / annals of psychology, 2018, vol.34, nº 3 (october) than .80large.Second, a multi-step multiple regression analysis was performed to test the predictive role of attachment and impulsivity over emotional dependence.Third, a mediation was conducted to see whether the dependence on the relationship between attachment and impulsive behavior was mediated.Finally, the differences in mean differences according to gender in emotional dependence, attachment and impulsive behavior were analyzed.

Correlation between attachment, emotional dependence and impulsivity
First, the correlations between attachment, emotional dependence and impulsivity were measured (Table 1).The degree of relationship was determined according to Cohen (1992) as follows: values below .20 would be small; around .50 would be medium and above .80would be large.
The results show small, though significant, positive correlations between emotional dependence and parental permission, childhood trauma and impulsivity.In terms of impulsivity, the results reflect positive relationships with avoidance of loneliness, asymmetric relationship, need to please, parental permission and childhood trauma.Likewise, unplanned impulsivity is negatively related to security, family concern, parental interference and parental authority.Finally, motor impulsivity is positively related to the need for exclusivity and self-sufficiency and resentment against parents.

Influence of emotional dependence on the relationship between attachment and impulsivity
Thirdly, the relationship between attachment and impulsivity is analyzed, being this relationship mediated by emotional dependence.
Attachment is initially associated with impulsivity (Table 3).As can be seen in security, parental authority, parental permission, self-sufficiency and resentment against parents and child trauma are statistically significantly associated with impulsivity.Later, attachment is associated with emotional dependence (Table 4).Parental interference, parental permission, self-sufficiency and resentment against parents, and child-hood trauma are statistically significantly associated with emotional dependence.Finally, emotional dependence is associated with impulsivity by controlling the association between attachment and impulsivity (Table 5).The results show how safety, parental authority, parental permission, self-sufficiency and resentment against parents, child trauma and emotional dependence are statistically significantly associated with impulsivity.
As can be seen, the non-standardised coefficient of the safety variable changes from the first step -.28 to -.27 in the third step.This means that 3.57% of the variance in the relationship between security and impulsivity is explained by emotional dependence.Likewise, the non-standardized coefficient of the variable parental authority is changed from the first step -.29 to -.28 in the third step.This means that 3.44% of the variance in the relationship between parental authority and impulsivity is explained by emotional dependence.Similarly, the non-standardised coefficient of the pa-rental permission variable changes from .30 in the first step to .28 in the third step.This means that 6.66% of the variance in the relationship between parental permission and impulsivity is explained by emotional dependence.Similarly, the non-standardized coefficient of the variable selfsufficiency and resentment against parents goes from being .27 in the first step to .28 in the third step, assuming that -3.70% of the variance in the relationship between selfsufficiency and resentment against parents and impulsivity is explained by emotional dependence.Finally, the nonstandardized coefficient of the variable childhood trauma, changes from being .14 in the first step to .12 in the third step, which means that 14.28% of the variance in the relationship between childhood trauma and impulsivity is explained by emotional dependence.

Gender differences in attachment, emotional dependence and impulsivity
Finally, gender differences in attachment, emotional dependence and impulsivity were analyzed (Table 6).The results obtained in a statistically significant way are shown below.The men scored significantly higher in avoiding being alone, need to please, asymmetric relationship, emotional dependence, cognitive impulsivity and unplanned impulsivity with a small effect size in all of them (avoiding being alone: d = .25;need to please: d = .27;asymmetrical relationship: d = .16;emotional dependence: d = .24;cognitive impulsivity: d = .05;unplanned impulsivity: d = .15).The women scored significantly higher on safety, family concern and parental authority, with a small effect size (security: d = .08;family concern: d = .18;parental authority: d = .15).

Discussion
The first objective of this study has been to study the relationship between emotional dependence, attachment and impulsive behavior.The results show that as emotional dependency increases, parental permissiveness, childhood trauma and impulsivity will also increase.These results are in line with previous studies where emotional dependence has been positively related to parental permissiveness and child traumatism (Momeñe & Estévez, accepted in the publication), thus situating their origin on occasion in the emotional or physical abuse received by attachment figures during childhood (Moral & Sirvent, 2009).Regarding the relationship between emotional dependence and impulsivity, previous studies point to deficits in impulse control in people with emotional dependence (Villegas & Sánchez, 2013) and difficulties in emotional regulation (Momeñe et al., 2017).In this study, impulsivity reflects relationships in the positive sense of avoiding being alone, asymmetric relationship, need to please, parental permissiveness and childhood trauma.These results coincide with previous studies where it is proved how the lack of care and the absence of contact with attachment figures in childhood predisposes to the development of impulsive behaviors (Barroso, 2014).In this vein, in a study conducted by Patock-Peckham & Morgan-Lithuania (2006), they found that the permissive parenting style, on a same-sex parent, was linked to high levels of impulsivity.Likewise, one of the novel results was found in the relationship between unplanned impulsivity and security, family concern, parental interference and parental authority, being this relationship in a negative sense, as well as the relationship between motor impulsivity and the need for exclusivity and self-sufficiency and resentment against parents, being in this case, a relationship in a positive sense.
Secondly, the predictive role of attachment and impulsive behavior on emotional dependence is analyzed.The results of this study indicate that parental permission, selfsufficiency and resentment against parents, and childhood trauma predict emotional dependence.These results would go along the lines of previous studies where it is pointed out how emotionally dependent people experienced in their childhood by their attachment figures absence of boosters, aggression (Castelló, 2005) and unsatisfied emotional needs (Barroso, 2014).
Thirdly, the relationship between attachment and impulsive behavior is studied, being this relationship mediated by emotional dependence.The results obtained show how emotional dependence acts as a mediator in the relationship between security and impulsivity, parental authority and impulsivity, parental permission and impulsivity, selfsufficiency and resentment against parents and impulsivity, childhood trauma and impulsivity.These results could be explained by the fact that emotional dependence has been linked to affective deficits in childhood (Castelló, 2012), as well as impulsive behavior, with previous studies mentioning how impulsivity levels vary depending on the parenting style received by attachment figures (Chapple & Johnson, 2007) and how low self-control would be the result of attachment disruptions during childhood (Hayslett-Mccall & Bernard, 2002).Because of this, these are novel results in mentioning emotional dependence as a mediating variable in the relationship between attachment and impulsivity.
Finally, the differences in the mean differences of emotional dependence, attachment and impulsive behavior according to sex are analyzed.It should be noted that men scored significantly higher in avoiding being alone, need to please, asymmetric relationship, emotional dependence, cognitive impulsivity and unplanned impulsivity, while women scored significantly higher in safety, family concern and parental authority.These results are consistent with recent studies where they have found a greater number of emotionally dependent men to their partners (Urbiola & Estévez, 2015), which carry with them aspects related to rejection of loneliness, the need to please and the asymmetric relationship (Castelló, 2005).Similarly, previous studies have found a greater number of impulsive men (Sauceda-García, Lara-Muñoz & Fócil-Márquez, 2006), despite this, it is mentioned how impulsivity varies according to age, with the adolescent stage being the highest rate of impulsive men reporting, while at later ages such differences between men and women would decrease (Escorial & Navas, 2006).
This study is not without limitations.On the one hand, the design of the study is transversal, making it impossible to establish causal relationships.In this respect, longitudinal studies could help to obtain such information.On the other hand, people who suffer from emotional dependency show difficulties in recognizing the problem, which may be due to the social acceptance of the thoughts and manifestations that surround it (Cogswell, Alloy, Karpinski & Grant, 2010).In addition, it is a sample composed solely of adolescents.
In conclusion, due to the scarce research on the relationship between attachment and impulsive behavior, as well as the mediating role of emotional dependence in this relationship, this study provides novel and highly clinical results for the intervention and treatment of emotional dependence and impulsivity.Among many other aspects, it should be noted that parental permissiveness may be related to more impulsive relationships based on emotional dependence on adolescents.Attachment in childhood may be an important factor for the development of impulsive behaviors in adolescence, such as emotional dependence, which is closely linked to the establishment of problematic relationships.

Table 1 .
Correlation between attachment, emotional dependence and impulsivity.

Table 2 .
Multiple linear regression of impulsive behavior and attachment to emotional dependence.

Table 3 .
The predictive role of attachment on impulsivity

Table 5 .
Mediation analysis of the indirect effect of attachment on impulsive behavior, this relationship being mediated by emotional dependence.

Table 6 .
Mean differences in attachment, emotional dependence and impulsivity according to gender