Background and Aim: During adolescence, many physical and psychological changes occur which can unexpectedly and significantly reduce self-esteem. The current study investigated the ability to predict self-esteem in adolescents as it relates to their attachments to parents and peers. 
Materials and methods: A total of 314 high school students were enrolled as subjects. Each completed the Inventory of Parent and Peer Attachment and the Rosenberg Self-Esteem Scale. The data were analyzed using SPSS. 
Findings: Self-esteem had a significantly positive correlation with trust and communication in parents and peers and a significantly negative correlation with alienation from parents and peers. Regression analysis also indicated that trust in the mother, trust in peers, and alienation from peers could significantly predict the level of adolescent self-esteem. Trust in the mother played the most significant role in this prediction.
Conclusion: Self-esteem is significantly related to parental and peer attachment. Trust in the mother, trust in peers, and alienation from peers could be used to significantly predict adolescent self-esteem levels.

Adolescence is a critical period of life in which many surprising physical and psychological changes emerge (Kennedy et al., 2018; Eshghizadeh et al., 2014; Basiri-Moghaddam et al., 2017). One of these changes is a significant decrease in self-esteem levels (Eshghizadeh et al., 2014; Basiri-Moghaddam et al., 2017). Self-esteem is the understanding and perceptions of individuals about themselves and reflects their feelings of self-worth and self-acceptance (Mackie and Smith, 2007; Harter, 2005). This psychological entity gradually increases throughout childhood and then suddenly decreases during adolescence and can cause an individual to behave anti-socially and aggressively (Hutteman, 2015; Barry et al., 2018; Harris et al., 2018; Forat-Yazdi et al., 2015; Thibaut, 2017). 
Positive self-esteem protects people against stress and anxiety and affects their emotional, affective, and cognitive processes (Oshri, 2017; Ciarma & Mathew, 2017). Children and adolescents with high self-esteem levels can communicate better with others, express their feelings in different situations, and are able to take responsibility for possible failures (Orth et 
al., 2010). They also have positive feelings and are able to exert internal control over themselves (Orth et al., 2010). On the other hand, people with low self-esteem tend to be dependent to others and cannot tolerate failure or making a mistake (Orth et al., 2010).
Self-esteem usually decreases during adolescence (Eshghizadeh, 2014). One reason for this is that, during this period, individuals are seeking to earn an identity while physical and psychological changes of puberty and maturity may cause them to have a negative perception of themselves and feelings of inadequacy (Marshall et al., 2015). Several psychological factors and changes occur during adolescence that can significantly affect self-esteem levels (Jones et al., 2018). Attachment style is one of these factors and is closely related to and affects adolescent self-esteem (Jones et al., 2018). Attachment style is an individual’s approach to deep emotional bonding to specific people in life (Sanford, 2017). It is categorized into three main types: secure, avoidant, anxious, and ambivalent (Ainswoth & Waters, 2015).
In Bowlby’s attachment theory, an individual’s relationship with the primary sources of attachment during early childhood creates psychological patterns which are the basis of their adult relationships (Music, 2016). Positive attachment styles can help individuals to deal with life challenges and define parts of their behavior, whereas a problematic attachment style can cause behavioral problems for adolescents (Moffitt, 2017). Attachment theory has provided a solid theoretical basis for understanding adolescent behavioral and emotional problems (Music, 2016). 
Greenberg et al. (1983) studied 213 adolescents aged 12 and 19 years and found that the quality of their relationships with their parents and peers significantly helped to predict their levels of life satisfaction and self-esteem (Greenberg et al., 1983). They concluded that parental attachment is the most important predictor of adolescent psychological well-being compared to peer attachment (Greenberg et al., 1983). Other studies have agreed with these findings (Dhal et al., 2007; Arbona & Power, 2004; Lee & Hankin, 2009; Parker & Benson, 2005; Raja et al., 1992; Luke et al., 2004). 
As adolescents strive to become independent of their attachment sources (Rosenberg, 2015), the quality of their relationships with their parents and peers can significantly define their self-esteem levels and behaviors (Raja et al., 1992; Luke et al., 2004). A feeling of alienation from parents during childhood is associated with increased depression, lower self-esteem, and a more insecure attachment style during adolescence (Ben-Ami and Baker, 2012). On the other hand, an insecure attachment style is associated with lower self-esteem and hypersensitivity to rejection; thus, self-esteem mediates the relationship between an insecure attachment style and hypersensitivity to rejection (Dhal et al., 2007; Arbona & Power, 2004; Lee & Hankin, 2009; Parker & Benson, 2005).
A literature review indicates that positive parental attachment provides a feeling of safety for adolescents and significantly correlates with their level of self-esteem (Sim & Yow, 2010). If parents care adequately for their child, that child will be more likely to consider others as being reliable and predictable people, experience feelings of self-worth, safety, self-control, self-efficacy, self-acceptance, and happiness and have high self-esteem (Luke et al., 2004). Self-esteem also mediates the relationship between parents and peer attachment with mental health (Wilkinson, 2004). Those who have been neglected by their parents during childhood can feel worthless during adolescence and consider people to be unreliable (Luke et al., 2004).
The experiences of individuals with their primary attachment sources affect their future perceptions and judgments about others as well as their emotional and social development (Luke et al., 2004; Bowlby, 1982). For example, someone with a secure attachment style will likely have a better self-concept and fewer psychological problems than someone with an insecure attachment style (Nishikawa et al., 2009). Children who have been hospitalized with psychiatric disorders have reported more ambivalent attachment styles than the healthy controls (Goodman, 2012).
Ultimately, an attachment style reflects the developmental context of the feelings, emotions, cognitions, interpersonal relationships, and self-concept of the individual and is significantly related to self-esteem. The present study investigated the ability to predict self-esteem as it relates to parental and peer attachment among adolescents.
Method
2.1 Sample and study design
In this cross-sectional study, the target population included all high school students in a large city (Tehran). A total of 400 high school students were enrolled in the study using convenience sampling. Written and verbal informed consent was properly obtained from all participants and they were assured of the lack of any personal data interpretation. They then were asked to complete the Inventory of Parent and Peer Attachment (Armsden and Greenberg, 1987) and the Rosenberg Self-Esteem Scale (Rosenberg, 1965). Any questionnaires that had incomplete responses were excluded from the study and a total of 314 subjects with fully completed questionnaires were entered into the study and data analysis. The sample size was determined to be sufficient according to the table developed by Krejcie and Morgan (1970). Data were analyzed using SPSS software.
2.2 Data collection
The Inventory of Parental and Peer Attachment (Armsden & Greenberg, 1987) was used to assess the amount and quality of attachment to parents and peers in the areas of trust, communication, and alienation for responses to 75 statements using a 5-point Likert scale. The test-retest reliability was 0.93 for parent attachment and 0.86 for peer attachment (Armsden & Greenberg, 1987). The internal reliability using Cronbach’s alpha was 0.87 for mother attachment, 0.89 for father attachment, and 0.92 for peer attachment (Armsden & Greenberg, 1987).
The Rosenberg Self-Esteem Scale was used to measure feelings of self-worth and self-acceptance using responses to ten self-report statements (Rosenberg, 1965). Agreement with the first five statements will add one point to the total score and disagreement will subtract one point. Agreement with statements 6 through 10 subtracts one point to the total score and disagreement adds one point (Rosenberg, 1965). A total score of over 20 indicates a very high level of self-esteem and a score of less than 10 indicates a very low level of self-esteem (Rosenberg, 1965).
Pullmann and Allik (2000) have reported that the test has acceptable validity and an internal consistency of 0.84. Rusticus et al. (2004) calculated the internal consistency of the test using Cronbach’s alpha for American, Canadian, and New Zealander adolescents as being 0.87, 0.86, 0.83, respectively. Mohammadi (2006) calculated the test reliability using Cronbach’s alpha as 0.69, the test re-test reliability as 0.78 and the split-half reliability as 0.68.

A total of 314 participants (153 males and 161 females) were enrolled in the study. The mean age of female participants was 16.81 and of male participants was 16.82, which showed no significant difference. Table 1 shows that, for all subjects, the mean trust in mother score was higher than for the father and the peers scores and the mean communication with mother score was higher than for the father and peers scores. The mean alienation from mother score was lower than for the father and peers scores among all subjects.

Table 1: Mean plus standard deviation on attachment to mother, father, and peers and self-esteem scores

Interestingly, the mean trust in mother score, mean communication with mother score, mean trust in father score, and the mean communication with father score were higher among male subjects than female subjects. Moreover, the mean alienation from mother score and the mean alienation from father score were both higher among male participants than among female participants. The mean trust in peers, mean communication with peers, and alienation from peers scores were lower among male than female subjects. Another interesting finding was that the mean self-esteem scores were higher among female subjects than male subjects.
Pearson’s correlation coefficient was used to assess the relationship between self-esteem and attachment to parents and peers and the results are presented in Table 2. As seen, self-esteem was significantly positively correlated with trust in mother (p<0 R=292), R=0.241), R=0.195), R=0.119), R=0.155), R=0.120). R=-0.268), R=-0.198), R=-0.161). >

Table 2: Correlation between attachment to mother, father, and peers and self-esteem

Stepwise regression analysis was used to assess the role of attachment to parents and peers as the independent variables for prediction of the level of self-esteem as the dependent variable. The results are presented in Table 3. It can be seen that self-esteem was significantly related to the variables of trust in mother, trust in father, trust in peers, communication with mother, communication with father, communication with peers, alienation from mother, alienation from father, and alienation from peers. 
The results in Table 3 and of regression analysis for all variables were entered into analysis in the first step. As seen, the only variable that was able to significantly predict self-esteem levels was trust in mother (R2=0.9). This variable was able to define 9% of the variance of self-esteem. Moreover, trust in mother and alienation from peers combined were able to significantly predict self-esteem levels (R2=0.11) and defined 11% of the variance of self-esteem. Thereafter, the variables of trust in mother, alienation from peers, and trust in peers combined were able to significantly predict self-esteem (R2=0.13) and defined 13% of the variance. The remaining variables were not able to significantly define and predict self-esteem.

Note: “Trust in mother” is the most significant predictor of self-esteem levels in this analysis. “Trust in mother”, “alienation from peers”, and “trust in peers” can together predict self-esteem levels. Other variables were not able to significantly predict self-esteem levels. 

Table 3: Results of stepwise regression analysis for prediction of self-esteem as it relates to attachment to father, mother, and peers

Overall, the variables of trust in mother, alienation from peers, and trust in peers combined were able to significantly predict self-esteem levels. The variable of trust in mother was the most significant predictor of self-esteem level compared to the other variables. 

Discussion
The results show that the mean trust in mother score was higher than the mean trust in father and mean trust in peers scores. The mean communication with mother score was higher than the mean communication with father and the mean communication with peers scores, while the mean alienation from mother score was lower than the mean alienation from father and the mean alienation from peers scores. Taken together these results indicate that adolescents rely more on their mothers as the most trustworthy source of secure attachment than on their fathers and peers. 
Bowlby (1982) concluded that being responsive to a child’s needs is the most important factor in the development of that child’s attachment style and that mothers are generally more responsive to a child’s needs, especially during childhood and adolescence, than fathers and peers. Thus, overall, children have better relationships and put more trust in their mothers than in fathers and peers (Bowlby, 1982). 
Attachment theory is highly affected by socio-cultural settings (Keller, 2013). In fact, attachment behaviors can be better explained and understood within cultural and psychosocial contexts. For example, attachment dimensions differ among different ethnic and racial groups (Keller, 2013). Culture, religion, and beliefs significantly affect parenting styles and attitudes toward children and the rearing of children and contribute to the development of attachment styles among people (Kirkpatrick, 1990; Granqvist, 2010; Keller, 2013). 
In Iranian culture, the mother has been found to be the most reliable and trusted source of secure attachment (Omidvar et al., 2014). This was evident in our findings. The reason for this is because, in Iran, the mother is traditionally more connected to the children than the father (Omidvar et al., 2014). The role of the mother is highly valued in Iranian culture and the mother plays a significant role in the development of the family. Mothers are considered to the managers of the home and family and communicate more regularly with the children than do fathers (Alinejad-Naeini et al., 2021). In this culture, fathers are primarily responsible for financial and other family-related matters (Alinejad-Naeini et al., 2021). 
In Iranian culture, religion, history, and traditions have attributed an honorable and spiritual role to the mother (Alinejad-Naeini et al., 2021). Most women prioritize motherhood and diligently care for their children, which creates a strong maternal-child connection and helps to develop a secure attachment style among the children (Omidvar et al., 2014). Moreover, children are more likely to put trust in their mothers and have good relationships with them, than in their fathers or peers, and feel less alienated from their mothers compared to fathers and peers (Omidvar et al., 2014).
Our results indicate that self-esteem had a significantly positive relationship with trust in the father, trust in the mother, trust in peers, communication with father, communication with mother, and communication with peers. It also had a significantly negative relationship with alienation from mother, alienation from father, and alienation from peers. These results indicate that the quality of an adolescent’s relationship with parents and peers affect that individual’s level of self-esteem. These results are consistent with the findings of Arbona and Power (2003), Armsden and Greenberg (1987), Ben-Ami and Baker (2012), Dhal et al. (2007), Goodman et al. (2012), Greenberg et al. (1983), Hoffman et al. (1988), Luke et al. (2004), McCormick and Kennedy (1994), Nishikawa et al. (2010), Papini et al. (1991), Paterson et al. (1995), Raja et al. (1992) and Sim and Yow (2011).
According to Bowlby’s model, a child moves from attachment toward independence. Bowlby (1963; 1973; 1980) believed that attachment to a mother or any other primary source of attachment is a basic psychological need and the origin of an individual’s sense of independence. Bowlby (1980) felt that a child who has been adequately loved and valued will be self-reliant in the future. Bowlby (1973) also emphasized the role of internal working models, which result from interactions and communications between mothers (or other attachment sources) and children. 
Such models can help people to predict attachment-source behavior and respond accordingly (Bowlby, 1973). For example, a secure attachment style which primarily results from having received adequate maternal care during childhood, creates healthy internal working models for individuals during adolescence and adulthood and makes them feel worthy and powerful (Bowlby, 1973). On the other hand, a problematic attachment style creates problematic internal working models, such as early maladaptive schemas, which can negatively affect information processing and cause people to have incorrect perceptions of situations and respond inappropriately (Bowlby, 1973). Moreover, people with problematic attachment styles along with disordered internal working models typically have feelings of worthlessness and inefficacy (Bowlby, 1973).
The results indicated that subjects with higher self-esteem levels had more secure attachment to their parents and peers and felt less alienated from them. Subjects with low self-esteem levels were insecurely attached to their parents and peers and felt alienated from them. Children of responsive, supportive, and logically authoritative parents tend to have higher self-esteem levels during adolescence and adulthood, while the children of non-responsive, controlling, hostile, and rejecting parents tend to have low self-esteem during adolescence and adulthood (Shahedifar et al., 2011). 
Bowlby (1969) stated that a newborn who has available and responsive caregivers will feel worthy and have a higher self-esteem level in the future. Moreover, people with secure attachment styles positively perceive people, while people with insecure attachment styles will perceive people negatively (Luke et al., 2004). Bowlby (1969) further stated that an individual who was adequately cared for during childhood will have a feeling of self-worth and consider others as reliable and predictable. An individual who has received inadequate care and support during childhood may feel worthless and believe that other people are unreliable and unpredictable. This indicates that an individual’s development and relationship with parents during childhood can predict their self-esteem level as an adult (Luke et al., 2004).
Some parents exercise too much control over their children, which does not allow them to become independent and responsible. The excessive control and criticism by such parents can cause children to feel helpless and ashamed and significantly decrease their self-esteem levels (Shahedifar et al., 2011). Such children may become unable to fulfill everyday tasks without outside support (Shahedifar et al., 2011). They are unable to envision themselves as being independent of their parents’ identity, and usually feel empty, confused, and inadequate (Shahedifar et al., 2011). They usually have a sense of failure and tend to be taken advantage of by others (Shahedifar et al., 2011).
The results of this study indicated that the variables of trust in mother, alienation from peers, and trust in peers significantly predicted self-esteem levels. Trust in the mother was the most significant predictor of self-esteem. Our results were consistent with those of several other studies that found that attachment to mother can significantly predict self-esteem levels, while attachment to father could not be used to predict self-esteem levels among adolescents (Patterson et al., 1995; Luke et al., 2004; Ainsworth, 1989; Bowly, 1982). 
Studies have revealed that fathers and mothers play different roles in the development of internal working models and self-esteem in children (Luke et al., 2004). Ainsworth (1989) and Bowly (1982) reported that the quality of an individual’s relationship with their mother is a stronger predictor of self-esteem, as opposed to the quality of the relationship with their father (Patterson et al., 1995). Maternal support is important to the development of an adolescent’s self-esteem while, in the absence of a mother, an individual’s peers are the most important sources of attachment and can affect the development of self-esteem (Hoffman et al., 1988).
Hoffman et al. (1988) found that paternal support was not related to adolescent self-esteem levels. Peer relationships, however, increase in importance during adolescence and the quality of this relationship will significantly affect an adolescent’s level of self-esteem (Berndt, 1982). During the transitional adolescence period, adolescents tend to communicate with their peers more than before and their relationship with parents gradually decreases (Cooper & Cooper, 1992). During this period, attachment behavior is directed more toward close friends and peers and those peers play important roles in the development of an adolescent’s self-esteem (Cooper & Cooper, 1992). Armsden and Greenberg (1987) reported that relationships between adolescents and their peers is more important than their relationship with their parents in the development of their self-esteem.
It can be seen that an individual’s relationship with their parents and peers significantly affects their self-esteem levels (Raja, 1992). Moreover, the role of peer relationships during adolescence is very important to the development of self-esteem and identity (Armsden & Greenberg, 1987; Cooper & Cooper, 1992). 
Conclusion
This study confirmed that parents and peer attachment can significantly predict self-esteem levels among adolescents and that trust in the mother and in peers and alienation from peers are the most significant predictors of adolescent self-esteem levels. Moreover, the results of this study show that adolescents tend to rely on their mothers as the most trustworthy source of a secure attachment rather than their fathers and peers. 
These findings are the basis for understanding the development of adolescent self-esteem. The results also indicate that, when a child receives adequate care and support, that child will tend to develop a secure attachment style and higher self-esteem level in adulthood. An individual who has been neglected by parents during childhood will tend to have a disordered attachment style and low self-esteem level during adolescence and adulthood. Attachment styles indicate how people perceive and react to their surrounding environment and can help predict their levels of self-esteem.
Limitation
Access to this large and young target group and asking them to cooperate properly was among limitations of this study. It is suggested to researchers to study this topic on students of this age in other parts of the world to compare the results with this study.
Declaration of interests
All authors declare that they have no conflict of interests for this study.
Data AvailabilityAll data sets is available from corresponding author upon request.
Ethics Approval
All content of this research adheres with the ethical guidelines developed by the Committee on Publication Ethics (COPE) during the 2nd World Conference on Research Integrity in Singapore in 2010. All parts of this study meets the American Psychological Association's (APA) Ethical Principles of Psychologists and Code of Conduct (the Ethics Code) and adheres to the legal requirements of the study country, Iran.