Keywords

1 Introduction

During the last 60 years, attachment theory has acquired an important role in the conceptualization of emotional and social development, and it has become one of the most influential and investigated models in Developmental Psychology.

At the core of attachment theory are the dynamics that occur in the parent–child relationship and how these affect the future development of the child. More precisely, attachment can be defined as an innate disposition to seek proximity to and contact with a specific figure (usually parents or more generally caregivers) and to do so in situations where they experience distress: a child could smile or vocalize to express his interest in interacting, or he could cry and cling if he perceives some kind of risk. It is extremely important to respond to these signals, especially in times when the child is distressed, as the availability of the parental figure is the factor that creates a difference in the development of a secure attachment (Bowlby 1969/1982; Cassidy 2016). For example, if the child feels that the parent is present both physically and emotionally in case of need, he can slowly move away to explore the surrounding environment, aware that if something should happen (a loud noise, or a small accident in the game) he can always return to the caregiver to be comforted. For this reason the parent acts as “a secure base from which to explore” (Ainsworth 1970, p. 61). The need to be comforted and the need to explore can be expressed in various ways, thus leading to the formation of different attachment patterns. Attachment security represents a good balance between the attachment and exploration systems, and is also the natural theoretical outcome, as well as the most desirable one (Solomon and George 2016). The most important factor for the development of a secure attachment is to make sure that the child acquires the feeling that the parent is present and available, because in this way he will internalize this representation, adapt this model, and generalize it to other relationships (Bretherton 1985).

2 Theoretical Framework

2.1 Attachment in Adolescence

Adolescence is often seen as a sensitive developmental period, which is characterized by “storm and stress”, as many changes occur at a biological, cognitive, affective and social level (Arnett 1999; Moretti and Peled 2004). Emotional development in adolescence should lead to a realistic and coherent sense of identity that allows a person to learn how to cope with stress and manage emotions as well as to relate to others (Santrock 2001). As evidenced by several studies (e.g., Lee et al. 2014), adolescence is also a vulnerable period for the onset of serious mental disorders, which then tend to persist into adulthood. While there is ample evidence concerning risk factors for mental disorders in adolescence, less is known about protective factors, although one important protective factor to have emerged from recent research concerns attachment security (Moretti 2016).

According to Rice (1990), the quality of the attachment bond is strictly implicated in the adolescents’ development of identity and in their emotional and social adjustment. Bowlby (1969/1982) sustained that during adolescence most children still have a strong attachment toward their parents, even if the child-parent relationship is changing and the adolescent also develops significant bonds with other people. This change has been described as “a realignment and redefinition of family ties” (Steinberg 2001, p. 255) and consists in a process of individuation, characterized in terms of autonomy, independence and detachment from family members (Ryan and Lynch 1989). According to Kobak and Duemmler (1994), the relationship between adolescents and parents becomes increasingly goal-corrected: while in infancy this goal-corrected partnership was more “coordinated”, in adolescence it should be considered more as “negotiated” (Allen and Tan 2016). During the previous operations of the attachment system, the child became used to “habitual patterns of responding” (Allen and Tan 2016, p. 400) while in this new stage of life, this pattern does not fit anymore with the adolescent’s need for autonomy. Therefore, to accomplish social development, which is a major task in adolescence, the adolescent has to develop a new balance between his attachment behaviors and his exploration urges; to do so, he or she has to decrease their need for dependence on parental attachment figures to satisfy the need for exploration. Despite this disengagement and transformation of family interactions, it is important to emphasize that adolescents still feel that it is important to be connected with their parental figures, even if they share less time and activities (Larson et al. 1996). Furthermore, Liebermann Doyle and Markiewicz (1999) found that, during adolescence, the perception of parental availability is still as important as in childhood, but the need for parental help and support decreases as children grow up. Nevertheless, according to Fraley and Davis (1997) and Trinke und Bartholomew (1997), adolescents still seek parents, and in particular mothers, more than friends when they need a secure base (Doyle and Moretti 2000; Kerns et al. 2015).

2.2 Parental Reflective Function

Several studies acknowledge the influence of parents’ representations about their children on their development. There is evidence that parental representations of their children are linked to children’s emotional, social and cognitive development, and attachment security (Fonagy et al. 2002; Rosenblum et al. 2009; Slade 2005). The characteristics of the parents’ representations also seem related to later affect regulation, and social relationships (Slade 2005), and mental health in adolescence (Fonagy and Gergely 2000). Studies have highlighted the link between parental reflective functioning (PRF), i.e. the ability to think about one’s own thoughts and feelings and those of the child, and quality of child attachment, and the role of PRF in protecting adolescents from psychopathology (Fonagy et al. 2002).

The concept of reflective functioning (RF) or mentalization was first introduced in the context of the “London Parent–Child Project”, which investigated the intergenerational transmission of attachment security (Steele and Steele 2008). RF has been defined as the capacity to understand one’s own and others’ mental states in terms of thoughts, feelings and desires. It is a very specific human feature that allows for emotion regulation and providing meaning to social interactions (Bateman and Fonagy 2016). RF has been proposed to be the missing link in the transmission of attachment security (Fonagy and Target 2005). Fonagy and colleagues (1991a, b), in their very first study on RF using the Adult Attachment Interview (AAI; George and Main 1984), could not only show that there was a high concordance between parent and child attachment, but also that the ability to reflect upon one’s own history of attachment was significantly correlated with the child’s attachment security. They concluded that the development of mentalizing ability originates within the parent–child relationship, where the child experiences the caregiver as being able to recognize and understand her or his mental states. Several years later, this idea was confirmed, by expanding it into a new line of research, i.e., by applying it to the Parent Development Interview – Revised (PDI-R; Slade et al. 2012), designed to evaluate parents’ mental representations of their child, of the relationship with him or her and of themselves as a parent. Instead of assessing a person’s ability to understand mental states with regard to his or her parents, this interview focuses on the capacity of parents to keep their child in mind, defined as parental reflective function (PRF). The caregivers’ capacity to hold in mind that their children have feelings, desires and intentions, allows children to discover their internal experience through the parents’ representation (Slade 2005). When children experience a change in their own mental state, parents need to be able to understand this moment by representing it and acknowledging this change through gestures, actions, words and play. The sensitization and the understanding of self-states occur through parental affect mirroring (Gergely and Watson 1996), where the caregiver produces an exaggerated version of realistic emotion expression, mirroring the infant’s state. The parent’s ability to observe changes in the child’s mental states lays the foundations for the development of secure attachment, which, in turn, provides the psychological basis in the acquisition of mentalizing abilities (Fonagy and Target 1997). The inter-subjective process of reflectiveness allows the child to understand the caregivers’ behavior and to perceive an image of him- or her-self as mentalizing, desiring and believing. Lastly, sensitive parenting also implies that parents are aware of their own mental states, so that they are able to differentiate between their own internal experience, and the one of their child (Slade 2005). Studies over the last decades have shown that PRF does not only affect children’s attachment security (Fonagy et al. 2002; Slade 2005) but parents’ emotion regulation as well (Schultheis et al. 2019).

Despite the growing evidence of the importance of PRF for the emotional development of children, most studies have focused on parents of infants, neglecting the role of PRF during adolescence. To our knowledge, only five studies have explored PRF in parents of older children. In the first study of this kind, Benbassat and Priel (2012) assessed PRF with the PDI in a group of mothers and fathers of adolescents and found that there was a positive association between PRF and the adolescents’ socio-emotional adjustment. A second study investigated a sample of clinically anxious children, including both mothers and fathers who completed the AAI (Esbjørn et al. 2013). The results showed that lower levels of maternal but not paternal reflective functioning (RF) were associated with higher child anxiety. A third study on a group of adopted adolescents, showed a strong relation between maternal RF, assessed with the PDI and child’s attachment (Molina et al. 2015). In a fourth study, researchers administered the PDI to a sample of parents, which mainly included mothers (Borelli et al. 2016). The results show that child-focused RF was significantly associated with child attachment security. Finally, the last study highlighted that mothers with higher levels of PRF would exhibit less overcontrolling behaviors with their children (Borelli et al. 2017). In summary, the results of these studies suggest that PRF is significantly associated with outcomes concerning the socioemotional adjustment and the well-being of children during middle childhood and adolescence.

2.3 The Project “ATTACH”: Studies and Methodology

More recently, we conducted a project (“Mental health and well-being during adolescence: The role of child attachment and parents’ representations of their children – ATTACH”) at the University of Luxembourg, which included families of adolescents from Luxembourg and Germany and which took place between October 2015 and November 2019. The project aimed to shed light on some of the potential factors that can affect attachment, mental health and well-being in adolescence. We assessed adolescents as part of a family system and, therefore, investigated the role of parental influence on adolescents’ psychological adjustment, and explored psychological and psychophysiological processes that are associated with attachment and parenting. The project involved three studies:

  • In the first study, we investigated the relationship of PRF with parental gender, parental cortisol responses and parenting behaviors (Decarli, Schulz, Pierrehumbert & Vögele, in press);

  • In the second study, we examined the factors that are associated with different attachment classifications and the association between attachment and internalizing and externalizing behavioral problems (Decarli, Pierrehumbert, Schulz & Vögele, in press);

  • In the third study, we tested whether adolescents with different attachment classifications show different patterns of emotion regulation strategies, both from a behavioral and physiological perspective, during two attachment relevant situations: an attachment interview and a parent–child conflict interaction task (Decarli et al. 2022).

After receiving ethics approval, we recruited participants in the German speaking area of the Greater Region of Luxembourg. Exclusion criteria were mental health problems of adolescents or parents, and participants needed to be fluent in either German, English, or French. Participants were 49 adolescents and their mothers and fathers. Not all fathers took part in the study. Adolescents had a mean age of 14.2 years and had different family backgrounds: 23 were living in their biological families, 19 had divorced or separated parents, and 7 had been adopted. The study design included three separate visits. During the first meeting we met with parent(s) on their own (consent form, interview and questionnaires), while the second meeting took place with the adolescents alone (consent form, attachment interview, questionnaires). These meetings took place at the family home, while the third meeting was scheduled with the whole family at the Clinical Psychophysiology Laboratory of the University of Luxembourg (for a detailed overview see Decarli et al. 2022, in press).

We used a multi-method approach with interviews, the observation of parent–child interactions, questionnaires and the assessment of psychophysiological stress responses during attachment-related situations. Concerning interviews and questionnaires, both parents were administered an interview to assess their RF capacities (PDI-R; Slade et al. 2012), then one of them took part in an interview to assess possible attachment difficulties of the child during childhood, and both parents (if possible) completed a questionnaire to evaluate their child’s behavioral problems. The adolescents were administered an interview to assess their attachment representations (Friends and Family Interview – FFI; Steele and Steele 2014) and then completed a self-reported questionnaire to evaluate behavioral problems. Concerning the parent–child interaction (Family Interaction Task – FIT; Allen et al. 2012), mothers and fathers took part separately in a conflict interaction task with their child and were asked to discuss a disagreement and to try to reach a solution. The interaction task allowed for the assessment of the adolescents’ ability to express their own opinions with confidence while maintaining an engaged and amicable connectedness with their parents. Parenting behaviors were operationalized as sensitive autonomy support and psychological control. We monitored heart rate and electrodermal activity as physiological indicators of stress and arousal from the adolescents both during the attachment interview, selecting the most attachment-relevant questions, and during the interaction task, focusing on the first 4 min, since some dyads after this time started discussing non-problematic issues. For analyses a reactivity score was calculated, by subtracting baseline means from the respective means of the selected questions during the FFI and means during the 1-min periods of the FIT. Concerning the parents, we collected 5 salivary cortisol samples during the FIT and calculated a reactivity score in terms of total hormone concentration over a period of reactivity.

3 Results

3.1 PRF, Parenting Behaviors and Attachment

We could show that PRF plays an important role in regulating the parents’ own stress responses in the context of parenting (operationalized in terms of cortisol reactivity). However, parental physiological reactivity alone did not predict parenting behaviors. The findings showed that parents who display higher mentalizing capacities show lower cortisol reactivity (i.e., stress in the context of parenting) and at the same time, thanks to their RF abilities, better caregiving behaviors, with higher levels of sensitive autonomy support and lower levels of psychological control. Current research has shown that emotional arousal and psychosocial stress are factors that can cause a reduction of the capacity to understand someone else’s mental states (Nolte et al. 2013). This is particularly true in the context of attachment relationships. There is evidence that when emotional arousal is high, the capacity to mentalize in a conscious, controlled and reflective way switches to an automatic, non-conscious and unreflective mode (e.g., Fonagy and Luyten 2018). During adolescence, the intensity and the frequency of conflicts between parents and children increase, and although these conflicts are also considered to be a means to negotiate relational changes, they can also enhance parenting stress (e.g., Branje 2018).

High levels of RF in both mother and fathers were the best predictors of secure attachment, confirming previous studies on infant samples (e.g., Steele and Steele 2008) and adding new important information about its role in adolescence. We would however have expected that parenting behaviors predicted attachment security as well, which occurred only in the case of attachment disorganization. Unexpectedly, its best predictor were indeed maternal behaviors undermining relatedness (but not the paternal ones), which also might explain the relation between disorganized attachment and maternal RF. Therefore, in the case of disorganization, maternal hostile and threatening behaviors seem to be the primary mechanism through which mothers’ RF is translated and communicated in the relationship with their adolescent children. These results might be explained as a consequence of the task we used to measure parental sensitivity and parenting behaviors, in that it might lacked ecological validity. The FIT is mainly designed around the notions of autonomy and relatedness, and more connected to the concept of “secure base” (Allen et al. 2003). Therefore, on the one hand, we might not have been able to capture sensitive behaviors that a highly reflective parent may use in other contexts, as for instance when the child is upset. On the other hand, an interaction task that challenges autonomy negotiations, which is conflictual by nature, might have maximized the opportunity for mothers with low RF capacities to display more often and at a higher extent hostile and threatening behaviors, since they might have felt more helpless and frustrated (Borelli et al. 2017; Hennighausen et al. 2011). Future research on parents’ sensitivity towards their adolescent children should include also other problem-solving situations that might be encountered by these dyads, as for example helping with a difficult task (Borelli et al. 2017), planning a vacation together (Becker-Stoll et al. 2008) or offering emotional support after a stressful task as the Trier Social Stress Test for Children (Buske-Kirschbaum et al. 1997).

Of significant interest are also the results concerning differences between mothers and fathers in the RF capacities. This finding was somewhat surprising, since research has shown that women usually have higher or at least similar RF levels as men (e.g., Steele and Steele 2008; Cooke et al. 2017). However, there may be several reasons for this result: (1) the possible presence of an assortative mating effect among intact and adoptive families (Borelli et al. 2017), (2) the possible influence of a divorced family status on maternal RF (Amato 2014), (3) the missing information concerning most of the fathers in the divorced families, and (4) the sociocultural changes concerning the role of fathers in families of western societies over the last fifty years (Cowan and Cowan 2019). It is important to note that these results (i.e., gender differences in RF levels and maternal behaviors predicting disorganized attachment) may have been possibly influenced at most by the fact that the data of the fathers were not randomly missing. Their participation could have explained more variance in the regression models, especially for what concerns autonomy support, and could serve as one of the possible interpretations as to why only maternal RF and behaviors were associated with disorganization. Recent studies have indeed shown that paternal sensitive support of autonomy is linked to attachment security and better psychological adjustment both in infancy and adolescence (Benbassat and Priel 2015; Grossmann et al. 2002). It remains entirely speculative, of course, whether or not the inclusion of fathers from divorced couples would have changed our findings and future studies should try to involve both parents, to better understand how each of them can uniquely contribute to the psychological adjustment of their children.

3.2 Attachment, Emotion Regulation and Mental Health

The most important results of the project ATTACH concern the converging evidence that disorganized attachment is linked to (1) more behavioral withdrawal during parent–child interactions, (2) heightened stress reactivity during attachment relevant situations and (3) more internalizing behavioral problems as compared to organized categories. With regard to the FIT, the lack of affect and engagement suggests that especially the interactions with the mothers were dominated by fear or by dissociative processes, which resulted in the full withdrawal of the adolescent (Allen et al. 2012; Duschinsky 2018). Concerning psychophysiological responses, our findings are in line with previous results from studies on infant samples, which showed that disorganized children during the SSP had higher heart rate responses when alone in the room (Spangler and Grossmann 1993), and increased heart rate during separation from and a decrease during reunion with the caregiver (Willemsen-Swinkels et al. 2000). However, our findings also differed from those of Beijersbergen, Bakermans-Kranenburg, van IJzendoorn and Juffer (2008), who did not find any differences between adolescents with resolved and unresolved attachment representation, but the authors argued that this might be due to how they measured physiological reactivity concerning loss, abuse and trauma experiences. Studies on adult samples have also shown that disorganized individuals displayed enhanced psychophysiological reactions during attachment related tasks (e.g., De Rubeis et al. 2016; Reijman et al. 2016). Finally, our results concerning the relation of attachment disorganization and internalizing behavioral problems were consistent with previous research (Brumariu and Kerns 2010; Madigan et al. 2016), further indicating its negative effects also on mental health. The pervasive negative influence that disorganized attachment has on the psychological, behavioral and physiological systems and its repeatedly demonstrated associations with maladjustment and psychopathology, has led to a call for disorganization to be included in psychiatric diagnostic systems as a developmental disorder (see Lyons-Ruth and Jacobvitz 2016). More recent reviews have pointed out, however, that disorganized attachment only shows modest stability over time, giving room for interventions with children and their parents (Granqvist et al. 2017).

The findings concerning dismissing attachment support the notion that during conflict interactions dismissing individuals tend to employ a fight or flight strategy, as evidenced by their display of behaviors undermining autonomy and relatedness at a higher extent and frequency than secure and preoccupied adolescents. It also appears that dismissing individuals experienced more stress than secure and preoccupied ones during the FFI suggesting that the attachment interview was able to activate their attachment-related defensive processes (Mikulincer et al. 2009) both from a narrative perspective (i.e., using idealization or derogation) as well as from a physiological one (i.e., using ineffectively a deactivating strategy). Our results are in line with previous findings from Dozier and Kobak (1992) and Roisman et al. (2004), who found that deactivation is associated with increased stress reactivity. In contrast, our findings differ from those of Beijersbergen and colleagues (2008) who found that dismissing individuals seemed to be less stressed than secure and preoccupied ones during the AAI. These contrasting results might be due to differences in the task that we used. We would argue that the questions of the FFI were better able to trigger attachment-related physiological responses in dismissing adolescents, because the interview is not focused on their early attachment experiences, but on current ones (Steele and Steele 2005). Our results concerning the psychophysiological responses during the FIT are also different from the ones of Beijersbergen and colleagues (2008) and this might be due to the different setting were the interaction took place (home vs. laboratory in our study). Moreover, the sample of Beijersbergen et al.’s study was only composed of adopted adolescents, and studies have shown that interactions between adopted children and their parents are more intense than those that occur in other families (Rueter et al. 2009). Future studies should therefore focus on different family types to find possible similarities and differences. Finally, our results are also consistent with the literature on problem behaviors (e.g. Madigan et al. 2016), since dismissing individuals reported higher externalizing symptoms and were judged by their mothers as having more externalizing difficulties.

In conclusion, it seems plausible that dismissing individuals might have developed some self-regulatory abilities that help them to deal physiologically with stressful circumstances (as for instance an interaction). Adolescents with disorganized attachment may not have this capacity, because their fear system is continuously activated, leading simultaneously to an excessive deactivation of the vagal system (freezing) and to the activation of the sympathetic nervous system (fight or flight), which does not allow for any response strategies. This process is reminiscent of what has been described in the literature on disorganized attachment as “fright without solution” (Hesse und Main 1999, p. 484).

4 Limitations and Future Research

Limitations of the current project concern, for example, the sample size, which was small in statistical terms and which might have led to insufficient statistical power to detect further differences between different attachment representations. Furthermore, the high rates of insecurity and, in particular of disorganization, present in our (non-clinical) sample might suggest that participant recruitment might have been biased. Families were mostly self-selected and the possible presence of an assortative mating effect might suggest that we have recruited in a way that maximized the possibility of finding thoughtful, reflective and sensitive parents among intact and adoptive families, whereas parents of the divorced group were probably more motivated by personal reasons (e.g., they felt that they struggled more in their parenting role or that their son or daughter had difficulties). Furthermore, the complexity of data collection, as for instance the inclusion of both parents, the fact of being interviewed and video-recorded, and last but not least the collection of physiological data might have limited participations. The selection of a bigger sample would probably bypass this “selection bias” and it would allow for better generalizability of the results. It is also important to point out that we did not assess other factors that might affect child attachment, PRF, parenting behaviors and parenting stress, and parents’ own attachment representations and parenting alliance. Therefore, some important information was missing from this project that might have added further explanatory power to our model. We can only hypothesize that parents with lower reflective capacities had an insecure (or unresolved) state of mind with respect to attachment, which is amply supported by the literature (Steele and Steele 2008). In this case, however, this remains speculative. The same applies to parenting alliance, since assessing how cooperative, communicative and mutually respectful parents are with regard to caring for their children has important effects on the parenting behaviors and this might have been especially important to explore in families with divorced parents (Becher et al. 2019). Future studies should assess both these variables when investigating PRF, parenting behaviors, parenting stress and the child’s psychological outcomes.

In our study on physiological responses during attachment relevant situations, we examined the physiological reactivity of the adolescents during an attachment interview and during a conflict interaction with their mothers and fathers. However, we used different methods to assess the physiological responses (heart rate and electrodermal activity for the adolescents and cortisol for the parents). It would be interesting in future research to examine the same indicators of physiological reactivity in all participants. Cortisol is widely used in research but also carries some limitations, as it can be imprecise and influenced by other factors (Nicolson 2007). Therefore, the assessment of heart rate and electrodermal activity could have provided greater time-specificity for their associations with PRF and parenting behaviors. Furthermore, future research may also focus on the concordance of physiological responses of parents and adolescents during a conflict interaction task in relation to attachment representation. In a previous study, Zelenko et al. (2005) reported that heart rate changes during the SSP were more consistent in secure mother-infant dyads than in insecure dyads. Other studies have also shown the powerful impact of maternal-infant social contact on the infant’s physiological systems and the results evidenced that the concordance between maternal and infant biological rhythms increased significantly during episodes of affect and vocal synchrony compared to non-synchronous moments (Feldman et al. 2011). This focus would give more insight into the importance of attachment representations for psychophysiological attunement also during adolescence. Moreover, another area that needs to be further explored concern peer relationships during adolescence. Although parents are still the main attachment figures in the life of adolescents, close friends and romantic acquire new significant importance (Allen and Tan 2016). Future studies should also investigate whether adolescents with different attachment representations would show the same physiological patterns of activation during conflict interactions with close friends as with their parents. This has been previously done with adult couples and the results showed that adults classified as insecure showed more physiological reactivity during interactions with romantic partners than secure adults (Roisman 2007). Future research should also focus on other factors that might influence attachment, as for instance neural mechanism behind attachment and parenting, and child temperament and genetics (Verhage et al. 2016).

A final important consideration concerns the impact that divorce has on the well-being of family members. In our study a divorced family status was significantly associated with lower maternal RF capacities, lower sensitivity and higher levels of adolescents’ disorganization. Previous research has shown that divorce increases the probability of developing insecure attachment, since there is less parental availability, more conflicts between parents and more parent–child negative interactions (Howes and Markman 1989; Waters et al. 1993). Divorce, however, can also cause discontinuity of attachment: individuals who experienced divorce during their childhood are more at risk of displaying insecure attachment in adolescence, although they had been assessed as secure in infancy (Lewis et al. 2000). However, the relationship between divorce and quality of attachment has been predominantly investigated in infancy/childhood (Solomon and George 1999; Page and Bretherton 2001) whereas only two studies have addressed this topic in an adolescent sample (Lewis et al. 2000; Hamilton 2000). We would argue that divorced mothers often do not have the possibility to rely on someone else in situations of (parenting) stress and this might activate feelings of helplessness, which in turn would activate their attachment system and might induce the use of automatic mentalization and display subsequently less sensitive behaviors and more psychological control (Fauber et al. 1990; Fonagy and Luyten 2009). Nevertheless, it is probably not divorce per se that causes parents to show more psychologically controlling behaviors but the sequelae of stressful factors that divorce brings with it. These findings advance our understanding of stress in the context of parenting as influencing the parental ability to cope with the daily challenges that child rearing brings with it, especially during the sensitive period of adolescence, and on the effects that these factors have on the quality of their children’s attachment. Therefore, more longitudinal studies are needed in order to infer causality and to understand what are the different factors that influence the developmental trajectories of an individual in one direction rather than in another, leading to significant differences in their psychological and social adjustment.

5 Clinical Implications and Concluding Remarks

The results of this project are particularly innovative and relevant for Luxembourg and Germany, since they highlight the importance of secure attachment representations during adolescence. The study can also provide significant guidance for the services that are involved in mental-health interventions, indicating the factors that enhance attachment-related security in families, to prompt the development of targeted interventions, both preventive and rehabilitative. Furthermore, the study of the attachment representations of adolescents with divorced parents could support clinical services in the establishment of interventions that can also help parents in better understanding the children’s needs and difficulties, thus improving their well-being.

Therefore, clinical interventions should help parents in developing their reflective skills, which can then be translated in more sensitive parenting behaviors, and adolescents in better integrating their attachment representations, to develop a more coherent and reflective state of mind. Attachment-based interventions (for an overview see Kobak and Kerig 2015; Slade 2007) are receiving increasing attention and might be a first choice, given their efficacy in helping parents to better understand their children’s thoughts, feelings and needs, and adolescents in better regulating their emotions (e.g., Attachment-Based Family Therapy; Connect Program; Mentalization-Based Treatment for Adolescents; Reflective Parenting Program). Other interventions could aim at involving more the paternal figures, mediating parental conflicts, and socially and financially supporting single mothers, who most often after a divorce might face social isolation and financial challenges. Moreover, it would be also important to train professional figures who are in close contact with families, as pediatricians or general practitioners, to recognize possible difficulties at an early stage, in order to promptly activate appropriate measures to support the child and his or her parents (Cohen and Weitzman 2016).