Abstract
Adolescence is a period of rapid social changes that may have important implications for the ways adolescents think, feel, and behave in their close relationships. According to family systems theory, adolescents’ attachment-related changes have the potential to spread throughout their family system, leading to coordinated changes in parents’ and adolescents’ attachment styles over time. The present study analyzed data from 205 adolescents (mean ageT1 = 14.0, SD = 0.9; 44% female, 56% male; 51% White, 33% African American, 3% Hispanic/Latino, 1% Asian American, 12% another race/ethnicity) and their parents (196 mothers, 105 fathers; median household incomeT1 = $100,000), who completed self-report measures of their general attachment styles annually for five years. Using a latent growth curve framework, we examined the extent of coordination among adolescents’ and parents’ long-term trajectories and shorter-term fluctuations in attachment security. The results revealed a push and pull dynamic between mothers’ and adolescents’ long-term trajectories of attachment security. Mothers who reported higher initial levels of attachment anxiety tended to have adolescent children who reported higher initial levels of avoidance. Additionally, adolescents who increased in attachment avoidance over time tended to have mothers who increased in attachment anxiety. Mothers and fathers mirrored each other’s patterns of attachment security as their children navigated their teenage years, reporting similar initial levels of attachment avoidance and synchronized shorter-term fluctuations in attachment anxiety and avoidance across time.
Keywords: attachment, adolescence, parent, longitudinal, change, coordination
Attachment theory was originally proposed to explain the bonds between young children and their caregivers (Bowlby, 1969/1982, 1973, 1980). According to the theory, children develop experience-based mental representations, known as internal working models (IWMs), that encompass their expectations and beliefs about the availability and responsiveness of attachment figures (Bowlby, 1969/1982; Bretherton & Munholland, 1999). Children who experience sensitive and responsive caregiving tend to develop secure representations that reflect a sense of confidence in their attachment figures’ (e.g., their parents’) ability to serve as a safe haven in times of distress and a secure base from which to explore the world. In contrast, children who experience unresponsive or inconsistently responsive caregiving tend to develop insecure representations, characterized by doubt or distrust that their attachment figures will be available and supportive when needed (see Thompson, 2021).
As children move into adolescence, these representations are thought to contribute to the general orientation with which they approach close relationships, including not only their continued relationships with parents (or caregivers), but also their relationships with friends and romantic partners (Feeney et al., 2008; Overall et al., 2003). Adolescents’ general attachment styles can be assessed with respect to two dimensions: attachment anxiety and attachment avoidance (Brennan et al., 1998). Attachment anxiety captures the extent to which a person feels worried about attachment-related concerns—for instance, worries that close others (e.g., parents, friends, romantic partners) will reject or abandon them. Higher attachment anxiety is characterized by hypervigilance for real or potential threats, intrusive caregiving toward others, and creation of a “self-amplifying cycle of distress” through rumination on feelings of discomfort or fear (Mikulincer & Shaver, 2019, p. 7). Attachment avoidance reflects the extent to which a person feels uncomfortable opening up to others and depending on them for support, and is characterized by the defensive exclusion (e.g., suppression, denial) of potential threats, negative feelings, and emotion-related thoughts (Mikulincer & Shaver, 2019). Within this dimensional framework (Brennan et al., 1998), people who are low on both attachment anxiety and avoidance are described as secure with respect to their attachment relationships (see Gillath et al., 2016, for a review of the dimensional approach to assessing attachment styles).1
Research suggests that these individual differences in the ways people think, feel, and behave in close relationships have a moderate degree of stability over time, but also remain open to change as people accumulate new relationship experiences (Sroufe & Waters, 1977; see Fraley & Roisman, 2019, for a review). Adolescence may be an especially important period for mean-level changes on the dimensions underlying general attachment styles, due to the whirlwind of biological and social changes that adolescents experience, including shifting relationships with childhood attachment figures (e.g., parents; Allen & Tan, 2016) and rapidly changing peer networks (Wrzus et al., 2013).
These issues raise a number of interesting questions, including questions about how much change adolescents experience and the extent to which changes in adolescents’ attachment styles may relate to other kinds of changes that occur within their social systems. In addition to attachment theory, family systems theory (Bowen, 1978; Cox & Paley, 2003; Minuchin, 1985) informs our consideration of these questions. Although family systems theory and attachment theory are distinct, these theories share a conceptual grounding in the notion that families are systems comprised of individuals, dyads, and perhaps triads or more, combined into subsystems (e.g., child-parent, parent-parent, child-child) that influence and are influenced by each other; anything that influences one part of the system may influence other parts (see Berlin & Cassidy, 1999; Byng-Hall, 1999; Byng-Hall & Stevenson-Hinde, 1991; Cowan, 1997; and Cowan & Cowan, 2009, for discussion of the interconnections between these two theories). From the perspective of these theories, the attachment-related changes experienced by adolescents may lead to changes in their parents’ attachment styles, and vice versa.
Cross-sectional research has provided some initial support for the idea that parents’ and adolescents’ attachment styles are interconnected. Studies have found that parents who report higher levels of attachment anxiety and avoidance tend to have children who also report higher levels of insecurity (e.g., Besser & Priel, 2005; Obegi et al., 2004; Roelofs et al., 2008). Research on the intergenerational transmission of attachment suggests that this association can be at least partially explained by the quality of caregiving experiences (van IJzendoorn, 1995). Parents who report higher levels of attachment insecurity provide less sensitive, responsive, and supportive caregiving; feel less close to their children; and express greater hostility toward their children (see Jones et al., 2015a, for a review). In turn, adolescents who perceive their parents as less sensitive, supportive, or accepting report higher levels of attachment insecurity, are less likely to use their parents as a secure base, and demonstrate less adaptive changes in attachment security over time (Gallarin & Alonso-Arbiol, 2012; Jones & Cassidy, 2014; Li et al., 2021; Ruhl et al., 2015).
Coordination Between Adolescents’ and Parents’ Attachment Styles
To the best of our knowledge, no research has examined the extent of coordination among changes in adolescents’ attachment styles and changes in their parents’ attachment styles. Based on prior research and theory, we can consider several ways in which these changes might play out. First, adolescents and parents might demonstrate mirrored changes in their attachment styles, moving in the same direction along the same dimension(s) of attachment insecurity over time. A mirroring dynamic can be good or bad depending on the direction of change. On the one hand, some parents and adolescents might experience co-occurring decreases in attachment anxiety and avoidance, reflecting a successful renegotiation of their attachment-caregiving relationship (Allen et al., 1994; Allen & Land, 1999; Moretti & Peled, 2004), in which parents allow their adolescents to have greater autonomy and explore new relationships with peers while remaining available as a secure base. At the same time, as adolescents gain more autonomy and begin to spend more time with peers, they can rely less on their parents to fulfill (at least some of) their attachment-related needs; this may allow parents greater space to balance their own and others’ attachment-related needs (e.g., partners, younger children) in a positive way. On the other hand, some adolescents and parents might demonstrate mirrored changes toward greater insecurity—for instance, showing co-occurring increases in attachment avoidance. If parents construe their teenagers’ desire for greater autonomy as abandonment, they might begin to emotionally withdraw from both their adolescent and their own attachment figures (e.g., their partner or co-parent), increasing their attachment avoidance. In turn, if adolescents feel that their parents are no longer available to support them when needed, their search for autonomy might devolve into more attachment-avoidant thoughts and behaviors (Gallarin & Alonso-Arbiol, 2012; Jones & Cassidy, 2014; Li et al., 2021; Ruhl et al., 2015).
Second, attachment scholars have proposed that changes in adolescents’ and parents’ attachment styles might demonstrate a push and pull dynamic (Allen & Tan, 2016; Martin et al., 2017; Moretti et al., 2017). Specifically, in families where parents increase in attachment anxiety as their children navigate the teenage years, their adolescents may increase in attachment avoidance. This push-and-pull is believed to reflect, in part, the emotional difficulties involved in renegotiating the caregiving relationship between parents and adolescents. If parents construe their adolescents’ desire for greater autonomy as abandonment, one sequela may be increased attachment anxiety, with concerns about others being less available to them. As such, parents might begin to act controlling, intrusive, or clingy (Feeney & Collins, 2001; Kunce & Shaver, 1994; Selcuk et al., 2010). When this occurs, adolescents might demonstrate more avoidant tendencies, such as seeking emotional distance from their parents and other close people. Alternatively, a push and pull dynamic could also emerge involving adolescents’ attachment anxiety and parents’ attachment avoidance. For instance, if the turmoil that adolescents often face leads them to experience periods of increased attachment anxiety (e.g., showing heightened sensitivity to cues of rejection, seeking excessive reassurance), parents might find themselves moving toward a more avoidant style, desiring greater emotional space from both their adolescents and close others.
Lastly, it could be the case that changes in adolescents’ and parents’ attachment styles show little-to-no coordination over time. Although both attachment theory and family systems theory suggest the possibility that changes in one person’s attachment are reflected in changes in another person’s attachment (Byng-Hall, 1999), these theories do not require that family members’ experiences always be linked. As adolescents begin to spend more time outside the home and online with peers (Barnes et al., 2007; Odgers & Jensen, 2020), there may be fewer opportunities for parents and adolescents to directly affect one another. Instead, changes in adolescents’ attachment styles may become increasingly coordinated with those of their peers (see Markiewicz et al., 2006, for a discussion on adolescents’ increasing use of peers for attachment needs). On a more pragmatic level, changes in parents’ and adolescents’ attachment styles might show little-to-no coordination if either parents or adolescents are not changing much in general. Previous research has shown that, although adolescents show considerable rank-order changes in their attachment styles over time (average test-retest correlation = .42), parents assessed over similar time periods exhibit more stability (average test-retest correlation = .72; Jones et al., 2018). This raises the possibility that, even if parents and adolescents are similar in their levels of insecurity overall, the changes that adolescents experience might not be coupled with the ways in which their parents’ attachment styles are changing if their parents are not changing much.
When considering links between adolescents’ and parents’ attachment styles, it is important to note the hierarchical nature of their attachment-caregiving relationship. According to theoretical definitions, an “attachment figure” is someone that a person uses as a safe haven in times of distress and a secure base from which to explore the world (Bowlby, 1969/1982). Therefore, whereas parents are attachment figures to their adolescents, parents do not typically view their teens as attachment figures (Ainsworth, 1989; Hazan & Zeifman, 1999). Nonetheless, adolescents and parents are bound to one another, interact on a daily basis, and often live in the same household. Thus, the residue of their activities may manifest in each other’s attachment styles. For example, conflicts between parents and adolescents may lead parents to experience heightened insecurity in their own attachment relationships (e.g., with romantic partners, close friends), if they feel unsupported by their romantic partner or close others in coping with and resolving the parent-teen conflict.
In the present research, we will examine the extent of coordinated changes in adolescents’ and parents’ general attachment styles—how they think, feel, and behave in close relationships in general. This approach offers a couple key advantages for our research purposes. First, research suggests that both parental and peer relationships (e.g., friends, romantic partners) play a critical role in adolescents’ socioemotional development (Steinberg & Morris, 2001). By assessing adolescents’ general attachment styles, we can gain a more holistic understanding of the changes that adolescents experience in their attachment styles (i.e., that accounts for all their close relationships), and the extent to which those changes are connected to changes in their parents’ attachment styles; notably, this approach avoids underestimating the roles that others (e.g., peers) may play in shaping adolescents’ overall, attachment-related functioning. Second, examining parents’ and adolescents’ general attachment styles allows us to be consistent in how we conceptualize and assess attachment styles across family members and dyads. As we explain in the following section, the present study will also examine the extent of coordination between mothers’ and fathers’ attachment styles. By focusing on general attachment styles, we can evaluate how each family member’s general sense of attachment security (i.e., the same attachment style) may be differentially linked with every other member’s general attachment security.
Coordination Between Mothers’ and Fathers’ Attachment Styles
An examination of the family system would not be complete without considering the parent-parent subsystem2 (Erel & Burman, 1995): How, and to what extent, might parents’ attachment styles change together over time? According to family systems theory, “each subsystem [e.g., parent-parent, parent-child] has its own integrity, defined metaphorically by the boundaries that separate it from other subsystems” (Minuchin, 1985, p. 291). Therefore, parents may demonstrate distinct pattern(s) of coordinated changes in their attachment styles, that differ in degree or nature from the patterns of coordination observed between each parent and their adolescent child. Indeed, there is reason to expect that parents will exhibit a greater degree of coordination with each other than with their adolescents and may be more likely to exhibit a mirroring dynamic. Unlike adolescents and their parents, both parents are adults and thus tend to have more similar developmental goals and attachment-related needs. Parents also frequently experience shared emotional challenges as a co-parenting team (Paley & Hajal, 2022). Importantly, whereas parents and adolescents share an asymmetrical attachment relationship in which parents provide for their children’s needs (but not vice versa), co-parents can provide for each other’s needs as part of a symmetrical attachment relationship (Hazan & Zeifman, 1999). Research suggests that adults who are coupled tend to rely on each other as the primary source for their attachment-related needs (e.g., Heffernan et al., 2012; Trinke & Bartholomew, 1997). Thus, parents who are together likely turn to each other in moments of distress, which—for better or worse—can contribute to linking their attachment styles over time.
Previous research investigating coordinated changes in young adult couples’ attachment security found mirrored patterns of change in partner-specific security across one year (Hudson et al., 2014). It is possible that similar patterns will be observed in general attachment security reported by middle-aged parents of adolescents. Alternatively, parents of adolescents might demonstrate greater coordination due to their longer relationship length, or diminished coordination when considering their general attachment styles (i.e., in close relationships more generally, rather than partner-specific attachment; Collins & Read, 1994; Overall et al., 2003) and when other family dynamics are added to the picture.
Modeling Coordination Across Time
Research on stability and change in attachment suggests that people’s attachment styles demonstrate both trait-like properties and more dynamic, state-like features over time (Fraley et al., 2011; Gillath et al., 2016). Studies have found that people report substantial fluctuations in attachment anxiety and avoidance on a day-to-day and month-to-month basis, as they encounter everyday relationship experiences (e.g., fights with a loved one; Davila & Sargent, 2003; Dugan et al., 2023; Girme et al., 2018). In most cases, these changes in people’s attachment styles are short-lived, and people return to their “baselines” of attachment security after a brief period. Nonetheless, research also suggests that people’s “baselines” of attachment security are not stationary or immutable; rather, people also experience more enduring, mean-level changes in attachment anxiety and avoidance as they age, and in response to certain life events and transitions (Chopik et al., 2019; Chopik et al., 2013; Chopik et al., 2014; Fraley et al., 2021; Hudson et al., 2015). This trait-state model of attachment suggests that, when studying coordinated changes in parents’ and adolescents’ attachment styles, it is important to distinguish between longer-term changes in their attachment styles and shorter-term fluctuations. As described earlier, family systems theory (Bowen, 1978; Cox & Paley, 2003; Minuchin, 1985) posits that parents and adolescents are interdependent parts of a dynamic and continually responsive system, meaning that both longer-term changes and shorter-term fluctuations in adolescents’ or parents’ attachment styles may reverberate through the family system, influencing other members (see also Fogel et al., 2000; Lougheed, 2020). However, the ways in which adolescents and their parents change, as well as the patterns of coordination among their changes, may differ across timescales. For example, we may find that adolescents and parents demonstrate mirrored patterns of shorter-term fluctuations, reflecting the current, or salient, “highs” and “lows” in their relationships, whereas they may demonstrate a push and pull dynamic among their longer-term changes, reflecting the more gradual process of renegotiating their attachment-caregiving relationship (Allen & Tan, 2016).
Multivariate latent growth curve models, like the one shown in Figure 1, provide a way to examine coordinated changes in parents’ and adolescents’ attachment styles across multiple timescales (McArdle, 2009; see Khan et al., 2020, for a discussion of this approach). Within this framework, people’s trajectories of attachment security are modeled using latent intercepts and latent slopes. Latent intercepts capture people’s levels of attachment security at the start of an assessment period, and latent slopes reflect their rates of change over time.
Figure 1.
Multivariate Latent Growth Curve Model
Note. Observed scores for two individuals are shown as x1…x5 and y1…y5. Ix and Iy denote the latent intercepts, and Sx and Sy denote the latent slopes. The smaller circles connected by orange, double-sided arrows represent the residual variances at each time point.
First, we can examine the association between parents’ and adolescents’ initial levels of attachment security by estimating the covariance among their intercepts, represented by the blue, long-dashed line in Figure 1. Prior research on the intergenerational transmission of attachment (e.g., Besser & Priel, 2005; Obegi et al., 2004; Roelofs et al., 2008) suggests that there may be a positive association between parents’ and adolescents’ intercepts of attachment security. For example, a positive association among their intercepts for avoidance would indicate that parents who report higher initial levels of attachment avoidance tend to have children who report higher levels of avoidance at the beginning of their teenage years. Of note, intercept-intercept associations reflect factors that have contributed to linking parents’ and adolescents’ attachment styles, prior to any changes that occur within the study period (e.g., pre-study developmental processes, genetic effects). Still, examining this kind of coordination is important for capturing the full extent to which parents’ and adolescents’ attachment styles are interconnected across time.
Second, we can examine whether parents and adolescents exhibit coordinated long-term changes in attachment security across the teenage years. This is captured by the slope-slope association, represented by the green, short-dashed line in Figure 1. For example, if we find a positive association between parents’ and adolescents’ slopes for attachment avoidance, that would suggest that parents who become more avoidant over time tend to have adolescent children who also become more avoidant over time, and vice versa (i.e., they exhibit mirrored long-term growth).
Third, we can examine whether parents and adolescents exhibit co-occurring, shorter-term fluctuations by looking at the covariance among their residuals, represented by the smaller orange lines in Figure 1. Residual variances capture the “leftover” variance at each time point, or the deviations between parents’ and adolescents’ time-specific scores in attachment security from their longer-term, linear trajectories. To illustrate this idea more clearly, we have plotted hypothetical attachment scores for a mother-adolescent dyad in Figure 2. Imagine that we assessed this adolescent and their mother once per year for five years. The blue data points represent the mother’s scores for attachment anxiety at each measurement occasion, and the orange data points represent the adolescent’s scores for attachment avoidance at each measurement occasion. We have plotted each person’s long-term, linear trajectory over time, as shown by the two solid lines in Figure 2. Each person’s residuals, or their deviations from their own trajectory at each of the five time points, are represented by the red dashed lines.
Figure 2.
Correlation among Residuals: Co-occurring Shorter-term Fluctuations
In this case, at times when the adolescent reported feeling particularly avoidant in their close relationships—that is, their avoidance scores were higher than what we would expect based on their long-term growth trend—their mother reported feeling particularly high levels of attachment anxiety (e.g., Year 2, Year 4). Similarly, at times when the adolescent reported particularly low levels of attachment avoidance, their mother reported lower levels of attachment anxiety (e.g., Year 1, Year 3, Year 5). This pattern of residuals reflects a push and pull dynamic between this adolescent’s and mother’s shorter-term fluctuations in attachment security. We describe these residuals as short-term because they are transient (i.e., specific to one time point) and do not contribute to individuals’ long-term growth trends of attachment insecurity.
Of note, whereas the adolescent in this example demonstrated an increase in attachment avoidance across the study period, their mother demonstrated minimal long-term growth in attachment anxiety (i.e., the mother’s linear trajectory is flat). However, by separating individuals’ growth trajectories from their time-specific deviations around those trajectories (i.e., residuals), we can capture patterns of coordination in attachment processes that unfold across both long-term and short-term timescales.
Overview of the Present Research
The present research examined the dynamic relationships between parents’ and adolescents’ attachment styles over time. We analyzed data from adolescents and their parents, who completed self-reports of their general attachment styles annually for five years. Using multivariate latent growth curve models, we examined the extent of coordination among adolescents’ and parents’ attachment styles over time, including the associations between their initial levels of attachment security, long-term growth in security, and shorter-term fluctuations in security. As a second aim, we also examined the extent of coordination among mothers’ and fathers’ attachment styles across time. In addressing our research questions, we drew from a family systems perspective, viewing adolescents and their parents as interdependent parts of a complex and continuously changing system (Bowen, 1966, 1978). We pre-registered our planned analyses, but intentionally approached our research questions regarding coordinated changes in adolescents’ and parents’ attachment styles with an open mind. We expected to find positive associations between parents’ and adolescents’ initial levels of attachment security based on prior research (e.g., Besser & Priel, 2005; Obegi et al., 2004; Roelofs et al., 2008). Further, analyses conducted to address these core aims will also provide novel information about mean-level growth in general attachment security experienced by adolescents and their parents.
Method
Participants and Procedure
Data were drawn from a broader study on family and peer relationships in adolescence. The study was approved by the Institutional Review Board at the University of Maryland, College Park (IRB# 05–0191). Families were recruited from the Washington, D.C., metropolitan area. Families who were proficient in English and had a child between the ages of 9 and 13 were eligible to participate in the initial lab session. The original study sample included 277 adolescents and their parent(s). Following the initial session, families were asked to return to the lab each year to complete 1.5-hour assessments involving various surveys and tasks. At waves 4 through 8 of the study, parents and adolescents were asked to complete self-report measures of their attachment styles. These data were collected from 2008 to 2012. In the present investigation, we analyzed the data collected at these five time points, which we will henceforth refer to as T1–T5.
At T1, 176 adolescents, 177 mothers, and 99 fathers provided data on their attachment styles, for a total of 163 adolescent-mother dyads and 94 adolescent-father dyads. At this first time point (T1), only a subsample of participants completed the attachment measures. Therefore, the relevant sample sizes increased slightly at T2. At T2, a total of 205 adolescents, 196 mothers, and 105 fathers provided data on their attachment styles (195 adolescent-mother dyads and 104 adolescent-father dyads). At T1, adolescents were 14 years old on average (SD = 0.9) and assessed annually through approximately age 18 years. Forty-four percent of adolescent participants identified as female, and 56% of adolescents identified as male. Participants were racially diverse: 51% identified as White, 33% African American, 3% Hispanic/Latino, 1% Asian American, and 14% reported another race/ethnicity. At T1, mothers reported an average age of 45.5 years (SD = 6.3), and fathers reported an average age of 48.9 years (SD = 8.3). Most parents were married (74%). The median household income reported at T1 was $100,000 (M = $104,755; SD = $58,218).
Measures
Parents’ Attachment Styles
Parents’ attachment styles were assessed using the Experiences in Close Relationships scale (ECR; Brennan et al., 1998), a 36-item measure designed to capture individual differences along two dimensions: attachment anxiety (e.g., “I worry about being abandoned”) and attachment avoidance (e.g., “I try to avoid getting too close to others”). Parents rated each item on a 7-point Likert scale (α ≥ .89 across assessments). Although the original version of the ECR asked about people’s experiences with romantic partners, many researchers now use a modified version of the ECR that assesses people’s experiences “in close relationships” more generally (see Mikulincer & Shaver, 2016). The present study used the latter, modified version of the ECR to assess parents’ attachment in close relationships, broadly defined.
Adolescents’ Attachment Styles
Adolescents’ attachment styles were measured using the 12-item short form of the ECR (ECR-S; Wei et al., 2007). Adolescents rated each item on a 7-point Likert scale (α ranged from .63–76). Like their parents, they completed the version of the ECR-S that asks about people’s experiences “in close relationships,” broadly defined. Research has shown that the ECR-S has strong psychometric properties and is highly correlated with the 36-item ECR (r’s ≥ .94 between ECR-S and ECR subscales for attachment anxiety and avoidance; Wei et al., 2007).
Transparency and Openness
We pre-registered the analysis plan on Open Science Framework (OSF; https://osf.io/2uqsw; Dugan, 2023). The study materials, analysis code, and de-identified data are available on OSF (https://osf.io/85kqg/; Dugan, 2024).
Results
Descriptive statistics were calculated in R (Version 4.1.3; R Core Team, 2022). Correlations among adolescents’, mothers’, and fathers’ attachment styles at the first two assessment waves are shown in Table 2. Descriptive statistics for all study variables, which includes mothers’, fathers’, and adolescents’ attachment styles measured at all five assessment waves, can be found in Table S1 in the supplemental materials. The main analyses were conducted in Mplus Version 8.8 (Muthén & Muthén, 2017).
Table 2.
Descriptive Statistics and Bivariate Correlations Among Adolescent and Parental Attachment Styles
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1. | Adolescent Attachment Anxiety T1 | ||||||||||||
| 2. | Adolescent Attachment Anxiety T2 | .48* | |||||||||||
| 3. | Adolescent Attachment Avoidance T1 | .31* | .18* | ||||||||||
| 4. | Adolescent Attachment Avoidance T2 | .23* | .13 | .60* | |||||||||
| 5. | Mother Attachment Anxiety T1 | .18* | .08 | .28* | .15 | ||||||||
| 6. | Mother Attachment Anxiety T2 | .17* | .06 | .20* | .08 | .79* | |||||||
| 7. | Mother Attachment Avoidance T1 | .07 | .02 | .13 | .06 | .41* | .30* | ||||||
| 8. | Mother Attachment Avoidance T2 | .02 | .01 | .13 | .08 | .41* | .40* | .72* | |||||
| 9. | Father Attachment Anxiety T1 | .18 | .12 | −.08 | −.04 | .15 | .33* | .05 | .24* | ||||
| 10. | Father Attachment Anxiety T2 | .09 | .02 | −.06 | .04 | .15 | .26* | .02 | .20 | .82* | |||
| 11. | Father Attachment Avoidance T1 | .15 | .02 | .07 | .14 | .25* | .23* | .13 | .18 | .39* | .38* | ||
| 12. | Father Attachment Avoidance T2 | −.03 | −.00 | .23* | .16 | .21 | .18 | .04 | .16 | .26* | .42* | .65* | |
|
| |||||||||||||
| Mean | 2.93 | 2.76 | 2.94 | 3.07 | 2.46 | 2.63 | 2.73 | 2.88 | 2.51 | 2.99 | 2.88 | 3.06 | |
| SD | 1.16 | 1.07 | 1.00 | 1.12 | 1.04 | 1.15 | 0.99 | 0.99 | 1.09 | 1.07 | 0.88 | 1.01 | |
Note.
p < .05
Mean-Level Changes in Attachment Security
First, we estimated univariate growth curve models to examine mean-level changes in parents’ and adolescents’ attachment security over time (see Table 3). On average, adolescents increased in attachment anxiety and avoidance across the study period (mean ages 14–18 years). There was significant variability among the slopes for both variables, indicating substantial variability in the rates at which adolescents became more or less secure over time. Additionally, we found a negative association between adolescents’ intercepts and slopes for avoidance, suggesting that adolescents who reported higher initial levels of attachment avoidance demonstrated less growth in avoidance over time.
Table 3.
Univariate Latent Growth Curve Models (LGCM) Parameter Estimates
| Est | SE | p | 95% CI | |
|---|---|---|---|---|
| Adolescent Attachment Anxiety | ||||
| Intercept Mean | 2.777 | 0.072 | <0.001 | [2.636, 2.918] |
| Slope Mean | 0.078 | 0.025 | 0.002 | [0.030, 0.127] |
| Intercept Variance | 0.603 | 0.111 | <0.001 | [0.384, 0.821] |
| Slope Variance | 0.030 | 0.015 | 0.046 | [0.001, 0.059] |
| Intercept-Slope Covariance | −0.003 | 0.033 | 0.928 | [−0.067, 0.061] |
| Adolescent Attachment Avoidance | ||||
| Intercept Mean | 2.956 | 0.068 | <0.001 | [2.823, 3.088] |
| Slope Mean | 0.057 | 0.026 | 0.026 | [0.007, 0.107] |
| Intercept Variance | 0.689 | 0.094 | <0.001 | [0.504, 0.873] |
| Slope Variance | 0.059 | 0.013 | <0.001 | [0.034, 0.085] |
| Intercept-Slope Covariance | −0.069 | 0.032 | 0.029 | [−0.131, −0.007] |
| Mother Attachment Anxiety | ||||
| Intercept Mean | 2.508 | 0.073 | <0.001 | [2.365, 2.651] |
| Slope Mean | 0.050 | 0.019 | 0.008 | [0.013, 0.087] |
| Intercept Variance | 1.004 | 0.116 | <0.001 | [0.777, 1.230] |
| Slope Variance | 0.031 | 0.009 | <0.001 | [0.014, 0.048] |
| Intercept-Slope Covariance | −0.032 | 0.022 | 0.140 | [−0.076, 0.011] |
| Mother Attachment Avoidance | ||||
| Intercept Mean | 2.791 | 0.064 | <0.001 | [2.666, 2.917] |
| Slope Mean | 0.015 | 0.014 | 0.265 | [−0.012, 0.042] |
| Intercept Variance | 0.724 | 0.092 | <0.001 | [0.543, 0.904] |
| Slope Variance | 0.005 | 0.005 | 0.285 | [−0.004, 0.015] |
| Intercept-Slope Covariance | 0.006 | 0.017 | 0.742 | [−0.028, 0.040] |
| Father Attachment Anxiety | ||||
| Intercept Mean | 2.838 | 0.109 | <0.001 | [2.625, 3.051] |
| Slope Mean | 0.052 | 0.027 | 0.060 | [−0.002, 0.105] |
| Intercept Variance | 0.939 | 0.159 | <0.001 | [0.628, 1.251] |
| Slope Variance | 0.016 | 0.011 | 0.134 | [−0.005, 0.036] |
| Intercept-Slope Covariance | −0.013 | 0.035 | 0.702 | [−0.081, 0.054] |
| Father Attachment Avoidance | ||||
| Intercept Mean | 2.983 | 0.074 | <0.001 | [2.837, 3.129] |
| Slope Mean | 0.042 | 0.019 | 0.025 | [0.005, 0.079] |
| Intercept Variance | 0.626 | 0.076 | <0.001 | [0.478, 0.775] |
| Slope Variance | 0.000 | 0.000 | 0.999 | [0.000, 0.000] |
| Intercept-Slope Covariance | 0.000 | 0.000 | 0.999 | [0.000, 0.000] |
Note. In the univariate LGCM examining father attachment avoidance, we constrained the slope variance to be equal to 0 (this sets the intercept-slope covariance equal to 0 as well) to facilitate convergence.
Mothers increased in attachment anxiety, on average, across the study period, with significant variability among their slopes for anxiety. Mothers did not, however, demonstrate growth in attachment avoidance, with little variation among their slopes for avoidance. Fathers did not show significant growth in attachment anxiety across the study period, with little variation among their slopes for anxiety. On average, fathers increased in attachment avoidance across the study period. However, for the univariate model examining fathers’ avoidance to converge, we had to constrain the slope variance to equal 0, indicating little-to-no variation in the extent to which fathers became more or less avoidant over time.
The minimal slope variance for mothers’ attachment avoidance and fathers’ attachment anxiety and avoidance suggested that we would not find slope-slope associations involving these variables in our planned multivariate models. Nevertheless, bivariate latent growth curve models can still allow us to examine other kinds of coordination among attachment styles (e.g., intercept-intercept associations, associations among residuals).
Coordinated Changes in Attachment Styles
Mothers’ and Adolescents’ Attachment Styles
We fit a series of bivariate latent growth curve models to examine the extent of coordination among mothers’ and adolescents’ attachment styles over time (see Figure 1). Specifically, we fit four models to examine each possible combination of mothers’ attachment anxiety and avoidance paired with adolescents’ attachment anxiety and avoidance (Models 1–4). In each model, we constrained the covariance among the residuals to be equal across waves to obtain a single estimate of the extent of coordination among shorter-term fluctuations. Standardized estimates were calculated separately using model constraints in Mplus Version 8.8 (Muthén & Muthén, 2017). In summarizing the results of each model, we will focus on the covariances/correlations between the growth factors and residuals. The means and variances of the growth factors were consistent with the univariate model results and are reported in each corresponding table.
Table 4 shows the results of Model 1 examining mothers’ and adolescents’ attachment anxiety. There were no associations among mothers’ and adolescents’ intercepts or slopes for attachment anxiety. There was a negative association between the residuals for mothers’ and adolescents’ attachment anxiety (r = −.110, SE = .056, p = .048), suggesting that, at times when adolescents reported particularly high levels of attachment anxiety, their mothers tended to report particularly low levels of anxiety. This result is not consistent with any of the three possibilities we have articulated (i.e., a mirroring dynamic, a push and pull dynamic, little-to-no coordination; see Table 1) but will be explored further in the discussion.
Table 4.
Model 1: Bivariate LGCM Mother Attachment Anxiety and Adolescent Attachment Anxiety
| Est | SE | p | 95% CI | |
|---|---|---|---|---|
| Unstandardized Parameter Estimates | ||||
| Covariances | ||||
| Mother Intercept-Adolescent Intercept (Covarying Initial Levels) |
0.152 | 0.081 | 0.060 | [−0.006, 0.309] |
| Mother Slope-Adolescent Slope (Coordinated Long-term Changes) |
0.010 | 0.007 | 0.181 | [−0.005, 0.024] |
| Mother Intercept-Adolescent Slope | 0.005 | 0.029 | 0.849 | [−0.051, 0.061] |
| Mother Slope-Adolescent Intercept | 0.009 | 0.019 | 0.624 | [−0.028, 0.047] |
| Mother Intercept-Mother Slope | −0.034 | 0.022 | 0.127 | [−0.078, 0.010] |
| Adolescent Intercept-Adolescent Slope | −0.003 | 0.033 | 0.938 | [−0.068, 0.062] |
| Covariance Among Residuals (Co-occurring Shorter-term Fluctuations) |
−0.046 | 0.022 | 0.033 | [−0.088, −0.004] |
| Means | ||||
| Mother Intercept | 2.508 | 0.073 | <0.001 | [2.365, 2.651] |
| Mother Slope | 0.049 | 0.019 | 0.009 | [0.012, 0.086] |
| Adolescent Intercept | 2.780 | 0.071 | <0.001 | [2.640, 2.919] |
| Adolescent Slope | 0.081 | 0.025 | 0.001 | [0.033, 0.130] |
| Variances | ||||
| Mother Intercept | 1.009 | 0.117 | <0.001 | [0.780, 1.238] |
| Mother Slope | 0.031 | 0.009 | <0.001 | [0.014, 0.048] |
| Adolescent Intercept | 0.602 | 0.112 | <0.001 | [0.382, 0.821] |
| Adolescent Slope | 0.029 | 0.015 | 0.054 | [0.000, 0.058] |
| Standardized Parameter Estimates | ||||
| Correlations | ||||
| Mother Intercept-Adolescent Intercept (Covarying Initial Levels) |
0.193 | 0.102 | 0.060 | [−0.008, 0.393] |
| Mother Slope-Adolescent Slope (Coordinated Long-term Changes) |
0.314 | 0.233 | 0.178 | [−0.143, 0.770] |
| Mother Intercept-Adolescent Slope | 0.033 | 0.166 | 0.841 | [−0.293, 0.360] |
| Mother Slope-Adolescent Intercept | 0.073 | 0.138 | 0.595 | [−0.197, 0.343] |
| Correlation Among Residuals (Co-occurring Shorter-term Fluctuations) |
−0.110 | 0.056 | 0.048 | [−0.219, −0.001] |
Note. The covariance/correlation among residuals was constrained to be equal across waves.
Table 1.
Patterns of Coordination and Bivariate LCGM Estimates
| Mirroring Dynamic: | Parents and adolescents move in the same direction on the same dimension(s) of attachment insecurity over time | |
|
| ||
| Bivariate Latent Growth Curve Model Estimates | Interpretation | |
|
|
||
| Among initial levels | Positive correlation between parents’ intercepts for attachment anxiety and adolescents’ intercepts for attachment anxiety and/or positive correlation between parents’ intercepts for attachment avoidance and adolescents’ intercepts for attachment avoidance | Example for attachment anxiety: Parents who report higher initial levels of attachment anxiety tend to have adolescent children with higher initial levels of attachment anxiety |
| Among long-term changes | Positive correlation between parents’ slopes for attachment anxiety and adolescents’ slopes for attachment anxiety and/or positive correlation between parents’ slopes for attachment avoidance and adolescents’ slopes for attachment avoidance | Example for attachment avoidance: Adolescents who become more avoidant over time (from ages 14 to 18) tend to have parents who also become more avoidant over time |
| Among shorter-term fluctuations | Positive correlation between parents’ residuals for attachment anxiety and adolescents’ residuals for attachment anxiety and/or positive correlation between parents’ residuals for attachment avoidance and adolescents’ residuals for attachment avoidance | Example for attachment anxiety: At times when adolescents report particularly high levels of attachment anxiety compared to their expected trajectories, their parents also tend to report higher levels of attachment anxiety compared to their expected trajectories |
|
| ||
| Push and Pull Dynamic: | As one member of a parent-adolescent dyad increases in attachment anxiety, the other member increases in attachment avoidance | |
|
| ||
| Bivariate Latent Growth Curve Model Estimates | Interpretation | |
|
|
||
| Among initial levels | Positive correlation between parents’ intercepts for attachment anxiety and adolescents’ intercepts for attachment avoidance and/or positive correlation between parents’ intercepts for attachment avoidance and adolescents’ intercepts for attachment anxiety | Example of adolescents “pulling away”: Parents who report higher initial levels of attachment anxiety tend to have adolescent children with higher initial levels of attachment avoidance |
| Among long-term changes | Positive correlation between parents’ slopes for attachment anxiety and adolescents’ slopes for attachment avoidance and/or positive correlation between parents’ slopes for attachment avoidance and adolescents’ slopes for attachment anxiety | Example of adolescents “pulling away”: In families where parents increase in attachment anxiety over time, their adolescents tend to increase in attachment avoidance (from ages 14 to 18) |
| Among shorter-term fluctuations | Positive correlation between parents’ residuals for attachment anxiety and adolescents’ residuals for attachment avoidance and/or positive correlation between parents’ residuals for attachment avoidance and adolescents’ residuals for attachment anxiety | Example of adolescents “pulling away”: At times when parents report particularly high levels of attachment anxiety compared to their expected trajectories, their adolescents tend to report higher levels of attachment avoidance compared to their expected trajectories |
|
| ||
| Little-to-No Coordination: | Parents and adolescents show little-to-no coordination in their attachment styles over time; this may occur if other relationships (e.g., with peers) hold greater weight in shaping each members’ general attachment styles, or if either member is not changing much overall | |
|
| ||
| Bivariate Latent Growth Curve Model Estimates | Interpretation | |
|
|
||
| Near-zero correlations between parents’ and adolescents’ intercepts, slopes, and residuals for attachment avoidance and anxiety | Parents and adolescents show little-to-no coordination in their initial levels, long-term changes, or shorter-term fluctuations in their attachment styles | |
Table 5 shows the results of Model 2 examining mothers’ attachment anxiety and adolescents’ attachment avoidance. We found a positive association between the intercepts for mothers’ attachment anxiety and adolescents’ attachment avoidance (r = .234, SE = .090, p = .010). This suggests that mothers who reported higher initial levels of attachment anxiety tended to have adolescents who reported higher initial levels of avoidance. Additionally, we found a positive slope-slope association (r = .327, SE = .152, p = .032), suggesting that mothers who increased in attachment anxiety across the study period tended to have adolescents who increased in avoidance. These findings are indicative of a push and pull dynamic between mothers and adolescents. No association among residuals was found.
Table 5.
Model 2: Bivariate LGCM Mother Attachment Anxiety and Adolescent Attachment Avoidance
| Est | SE | p | 95% CI | |
|---|---|---|---|---|
| Unstandardized Parameter Estimates | ||||
| Covariances | ||||
| Mother Intercept-Adolescent Intercept (Covarying Initial Levels) |
0.198 | 0.077 | 0.010 | [0.047, 0.349] |
| Mother Slope-Adolescent Slope (Coordinated Long-term Changes) |
0.015 | 0.006 | 0.019 | [0.002, 0.027] |
| Mother Intercept-Adolescent Slope | −0.030 | 0.028 | 0.291 | [−0.086, 0.026] |
| Mother Slope-Adolescent Intercept | −0.022 | 0.019 | 0.248 | [−0.058, 0.015] |
| Mother Intercept-Mother Slope | −0.033 | 0.022 | 0.140 | [−0.076, 0.011] |
| Adolescent Intercept-Adolescent Slope | −0.067 | 0.032 | 0.035 | [−0.129, −0.005] |
| Covariance Among Residuals (Co-occurring Shorter-term Fluctuations) |
−0.017 | 0.018 | 0.337 | [−0.053, 0.018] |
| Means | ||||
| Mother Intercept | 2.505 | 0.073 | <0.001 | [2.363, 2.648] |
| Mother Slope | 0.052 | 0.019 | 0.005 | [0.015, 0.089] |
| Adolescent Intercept | 2.955 | 0.068 | <0.001 | [2.822, 3.087] |
| Adolescent Slope | 0.060 | 0.026 | 0.019 | [0.010, 0.110] |
| Variances | ||||
| Mother Intercept | 1.003 | 0.116 | <0.001 | [0.776, 1.229] |
| Mother Slope | 0.031 | 0.009 | <0.001 | [0.014, 0.049] |
| Adolescent Intercept | 0.681 | 0.094 | <0.001 | [0.497, 0.866] |
| Adolescent Slope | 0.059 | 0.013 | <0.001 | [0.033, 0.085] |
| Standardized Parameter Estimates | ||||
| Correlations | ||||
| Mother Intercept-Adolescent Intercept (Covarying Initial Levels) |
0.234 | 0.090 | 0.010 | [0.057, 0.412] |
| Mother Slope-Adolescent Slope (Coordinated Long-term Changes) |
0.327 | 0.152 | 0.032 | [0.028, 0.625] |
| Mother Intercept-Adolescent Slope | −0.117 | 0.116 | 0.313 | [−0.343, 0.110] |
| Mother Slope-Adolescent Intercept | −0.136 | 0.129 | 0.292 | [−0.388, 0.117] |
| Correlation Among Residuals (Co-occurring Shorter-term Fluctuations) |
−0.037 | 0.047 | 0.431 | [−0.129, 0.055] |
Note. The covariance/correlation among residuals was constrained to be equal across waves.
The results of Model 3 examining mothers’ attachment avoidance and adolescents’ attachment anxiety are summarized in Table 6. There were no associations among mothers’ and adolescents’ intercepts and slopes. We did find a negative association between the residuals for mothers’ attachment avoidance and adolescents’ attachment anxiety (r = −.123, SE = .052, p = .019), suggesting that, at times when adolescents reported particularly high levels of anxiety, their mothers tended to report particularly low levels of avoidance. This pattern of coordination between mothers’ and adolescents’ shorter-term fluctuations in general attachment security does not align with the alternatives outlined in Table 1 (see the Discussion).
Table 6.
Model 3: Bivariate LGCM Mother Attachment Avoidance and Adolescent Attachment Anxiety
| Est | SE | p | 95% CI | |
|---|---|---|---|---|
| Unstandardized Parameter Estimates | ||||
| Covariances | ||||
| Mother Intercept-Adolescent Intercept (Covarying Initial Levels) |
0.036 | 0.078 | 0.648 | [−0.118, 0.189] |
| Mother Slope-Adolescent Slope (Coordinated Long-term Changes) |
0.000 | 0.005 | 0.999 | [−0.010, 0.010] |
| Mother Intercept-Adolescent Slope | 0.035 | 0.025 | 0.174 | [−0.015, 0.084] |
| Mother Slope-Adolescent Intercept | 0.006 | 0.014 | 0.640 | [−0.021, 0.034] |
| Mother Intercept-Mother Slope | 0.005 | 0.017 | 0.764 | [−0.029, 0.040] |
| Adolescent Intercept-Adolescent Slope | −0.003 | 0.033 | 0.929 | [−0.068, 0.062] |
| Covariance Among Residuals (Co-occurring Shorter-term Fluctuations) |
−0.049 | 0.020 | 0.015 | [−0.089, −0.010] |
| Means | ||||
| Mother Intercept | 2.793 | 0.064 | <0.001 | [2.667, 2.919] |
| Mother Slope | 0.015 | 0.014 | 0.285 | [−0.012, 0.042] |
| Adolescent Intercept | 2.781 | 0.071 | <0.001 | [2.641, 2.921] |
| Adolescent Slope | 0.078 | 0.024 | 0.001 | [0.030, 0.126] |
| Variances | ||||
| Mother Intercept | 0.726 | 0.093 | <0.001 | [0.544, 0.909] |
| Mother Slope | 0.005 | 0.005 | 0.280 | [−0.004, 0.015] |
| Adolescent Intercept | 0.605 | 0.112 | <0.001 | [0.385, 0.824] |
| Adolescent Slope | 0.030 | 0.015 | 0.043 | [0.001, 0.059] |
| Standardized Parameter Estimates | ||||
| Correlations | ||||
| Mother Intercept-Adolescent Intercept (Covarying Initial Levels) |
0.056 | 0.119 | 0.641 | [−0.178, 0.289] |
| Mother Slope-Adolescent Slope (Coordinated Long-term Changes) |
−0.010 | 0.392 | 0.979 | [−0.778, 0.757] |
| Mother Intercept-Adolescent Slope | 0.227 | 0.168 | 0.177 | [−0.102, 0.556] |
| Mother Slope-Adolescent Intercept | 0.113 | 0.236 | 0.633 | [−0.350, 0.575] |
| Correlation Among Residuals (Co-occurring Shorter-term Fluctuations) |
−0.123 | 0.052 | 0.019 | [−0.226, −0.021] |
Note. The covariance/correlation among residuals was constrained to be equal across waves.
Table 7 shows the results of Model 4 examining mothers’ and adolescents’ attachment avoidance. No associations were found.
Table 7.
Model 4: Bivariate LGCM Mother Attachment Avoidance and Adolescent Attachment Avoidance
| Est | SE | p | 95% CI | |
|---|---|---|---|---|
| Unstandardized Parameter Estimates | ||||
| Covariances | ||||
| Mother Intercept-Adolescent Intercept (Covarying Initial Levels) |
0.111 | 0.069 | 0.106 | [−0.024, 0.246] |
| Mother Slope-Adolescent Slope (Coordinated Long-term Changes) |
0.004 | 0.005 | 0.430 | [−0.006, 0.014] |
| Mother Intercept-Adolescent Slope | 0.005 | 0.025 | 0.829 | [−0.044, 0.054] |
| Mother Slope-Adolescent Intercept | −0.017 | 0.016 | 0.294 | [−0.049, 0.015] |
| Mother Intercept-Mother Slope | 0.007 | 0.017 | 0.694 | [−0.027, 0.041] |
| Adolescent Intercept-Adolescent Slope | −0.067 | 0.031 | 0.034 | [−0.129, −0.005] |
| Covariance Among Residuals (Co-occurring Shorter-term Fluctuations) |
0.005 | 0.017 | 0.777 | [−0.028, 0.037] |
| Means | ||||
| Mother Intercept | 2.789 | 0.064 | <0.001 | [2.663, 2.915] |
| Mother Slope | 0.016 | 0.014 | 0.233 | [−0.011, 0.043] |
| Adolescent Intercept | 2.954 | 0.068 | <0.001 | [2.821, 3.087] |
| Adolescent Slope | 0.058 | 0.026 | 0.022 | [0.008, 0.108] |
| Variances | ||||
| Mother Intercept | 0.722 | 0.092 | <0.001 | [0.541, 0.903] |
| Mother Slope | 0.005 | 0.005 | 0.307 | [−0.005, 0.015] |
| Adolescent Intercept | 0.686 | 0.094 | <0.001 | [0.502, 0.870] |
| Adolescent Slope | 0.058 | 0.013 | <0.001 | [0.033, 0.084] |
| Standardized Parameter Estimates | ||||
| Correlations | ||||
| Mother Intercept-Adolescent Intercept (Covarying Initial Levels) |
0.157 | 0.097 | 0.104 | [−0.032, 0.347] |
| Mother Slope-Adolescent Slope (Coordinated Long-term Changes) |
0.231 | 0.288 | 0.422 | [−0.333, 0.795] |
| Mother Intercept-Adolescent Slope | 0.027 | 0.121 | 0.824 | [−0.211, 0.264] |
| Mother Slope-Adolescent Intercept | −0.289 | 0.258 | 0.263 | [−0.795, 0.217] |
| Correlation Among Residuals (Co-occurring Shorter-term Fluctuations) |
0.017 | 0.047 | 0.719 | [−0.075, 0.108] |
Note. The covariance/correlation among residuals was constrained to be equal across waves.
Fathers’ and Adolescents’ Attachment Styles
We fit a series of bivariate growth curve models to examine the extent of coordination among fathers’ and adolescents’ attachment styles over time. For the models involving fathers’ attachment avoidance (paired with adolescents’ attachment anxiety and avoidance, Models 7–8), we constrained the slope variance for fathers’ avoidance to equal 0 to achieve convergence. The results are shown in the supplement (see Tables S2–S5, Models 5–8). There were no associations between fathers’ and adolescents’ attachment styles (i.e., their intercepts, slopes, or residuals). Overall, this suggests that fathers and adolescents showed little-to-no coordination in their attachment styles over time.
Mothers’ and Fathers’ Attachment Styles
Lastly, we fit bivariate growth curve models to examine the extent of coordination among mothers’ and fathers’ attachment styles. For Model 12 examining mothers’ and fathers’ attachment avoidance, we fixed the slope variance for fathers’ avoidance to 0 to achieve convergence. The results are provided in the supplement (see Tables S6–S9, Models 9–12). We found a positive association between mothers’ and fathers’ intercepts for attachment avoidance (r = .249, SE = .118, p = .035), indicating that mothers who reported higher initial levels of avoidance tended to be paired with fathers who also reported higher initial levels of avoidance. No other associations were found among mothers’ and fathers’ intercepts and slopes. Interestingly, we found a positive association between the residuals in all four models examining mothers’ and fathers’ attachment styles. At times when mothers reported feeling particularly anxious in their relationships, fathers reported feeling particularly anxious (r = .242, SE = .076, p = .001) and avoidant (r = .167, SE = .061, p = .006) in their relationships (and vice versa). Similarly, at times when mothers reported feeling particularly avoidant in their relationships, fathers reported particularly high levels of both anxiety (r = .286, SE = .081, p < .001) and avoidance (r = .183, SE = .062, p = .003). Taken together, these findings suggest that parents experienced mirrored patterns of shorter-term fluctuations on both dimensions of attachment insecurity.
Discussion
The present research investigated the dynamic relationships among parents’ and adolescents’ attachment styles over time. We analyzed self-report attachment data collected from adolescents and their parents annually for five years (ages 14 to 18) to examine the extent of coordination between their long-term trajectories and shorter-term fluctuations in general attachment security. As a second aim, we also examined the extent of coordination among mothers’ and fathers’ attachment styles across time.
The results revealed a push and pull dynamic between mothers’ and adolescents’ attachment styles over time. Adolescents who increased in attachment avoidance from mid- to late-adolescence—who became less comfortable confiding in others and depending on them for support—tended to have mothers who increased in attachment anxiety, expressing greater concerns about rejection and abandonment. Additionally, adolescents who were higher in avoidance initially tended to have mothers who reported higher initial levels of attachment anxiety. Adolescents and mothers exhibited coordinated patterns of shorter-term fluctuations in attachment security as well—at times when adolescents reported particularly high levels of attachment anxiety, their mothers tended to report lower levels of both attachment anxiety and avoidance. We did not find any coordination among fathers’ and adolescents’ attachment styles across time.
Mothers and fathers demonstrated mirrored shorter-term fluctuations in attachment security as their children navigated adolescence. At times when one parent reported feeling particularly anxious or avoidant with respect to their close relationships, the other parent tended to report higher levels of both anxiety and avoidance. We also found a positive association between mothers’ and fathers’ initial levels of attachment avoidance, indicative of non-random pairing of attachment styles within co-parent dyads. Each of these findings has important implications for research on the development of attachment styles and coordination within family systems, which we explore in the following sections.
The Push and Pull Among Mothers’ and Adolescents’ Attachment Styles
Our findings suggest that, in families where adolescents tended to pull away from their attachment figures as they navigated the teenage years, their mothers felt an increased need to cling tighter to close others in their lives. This push-and-pull trend could reflect varying degrees of tension involved in adolescents’ and mothers’ renegotiation of the caregiving relationship (Allen & Tan, 2016). Mothers who are high in attachment anxiety might pose a threat to adolescents’ autonomy-striving, leading adolescents to feel uncomfortable opening up to their mothers and turning to them in times of need. Research suggests that people who are high in attachment anxiety tend to be hypervigilant of real or potential threats, prefer to closely monitor loved ones’ behaviors, and exacerbate their own distress and fear through rumination (Mikulincer & Shaver, 2019). It is not difficult to imagine, therefore, how adolescents’ bids for greater autonomy might not be received well by mothers who are high in attachment anxiety. Highly anxious mothers may resist the idea of their adolescents spending more time outside the home, where many real and potential dangers exist. Mothers who are high in attachment anxiety may also resist the negotiation process itself, as figuring out a new set of guidelines for their teenagers often requires having tense or conflict-ridden conversations with their adolescents and other family members (e.g., co-parents, non-adolescent children; Kobak et al., 1994). From adolescents’ perspective, however, receiving these kinds of reactions from their mothers, who are often their primary attachment figures (Rosenthal & Kobak, 2010), can have harmful consequences. Adolescents may learn that disclosing their feelings to others tends to be unproductive, and they cannot count on others to be supportive of their goals (i.e., hallmarks of attachment avoidance).
It is important to note that the associations between adolescents’ intercepts (i.e., initial levels) and slopes (i.e., growth over time) for attachment avoidance and mothers’ intercepts and slopes for attachment anxiety are non-directional. Thus, the underlying dynamics connecting mothers’ and adolescents’ growth in attachment insecurity could also be triggered by adolescents’ behavior or flow in both directions. For instance, teenagers may approach the challenges of adolescence with different strategies that vary in the extent to which they reflect healthy autonomy-striving versus compulsive self-reliance (i.e., attachment avoidance). Research suggests that children who are higher in avoidance tend to become emotionally “closed off” during adolescence, showing high levels of disengagement and unresponsiveness during interactions with their mothers (Allen & Land, 1999; Becker-Stoll et al., 2001; Becker-Stoll & Fremmer-Bombik, 1997; Kobak et al., 1993). Research has also found that adolescents who are higher in avoidance tend to report greater internalizing and externalizing problems (Pace & Zappulla, 2011) and suicidal behaviors (Falgares et al., 2017), develop less satisfying relationships, and engage in fewer pro-relationship behaviors (Collins et al., 2002). If mothers notice their adolescents becoming closed off and experiencing these real threats, they may become hypervigilant of their adolescents, heightening their own attachment anxiety regarding the availability of close others and exacerbating their teenagers’ avoidance. Ein-Dor and colleagues (2011) have proposed that increased attachment anxiety can represent an adaptive response to threats in one’s environment. Specifically, the hypervigilance that characterizes attachment anxiety (e.g., monitoring close others’ behaviors, exaggerated distress reactions) can allow people to detect dangers more quickly and alert close others (e.g., co-parents, friends) who can provide support during challenging times.
Interestingly, in his earliest writings, Bowlby (1969/1982) also described a set of adaptive, push and pull tendencies between mothers and their (young) children, that served to promote maximal exploration for the child, while keeping the child safe. Drawing from classic theories on system dynamics (e.g., control theory, General System Theory; von Bertalanffy, 1968)—perspectives that also influenced, and are reflected in, family systems theory—Bowlby (1969/1982) wrote that attachment behavior was, “…organized so that movement toward a specified goal-object would become increasingly probable the greater the distance from the goal object” (p. 235). He observed that, as one member of a mother-child dyad moved away, the other compensated by seeking proximity, so that the total distance between mother and child did not exceed what was safe. The push and pull trend we observed between mothers’ and adolescents’ attachment styles could reflect a similar kind of realignment, at the psychological level, to compensate for the new, independence-striving goals of adolescents.
The observed coordination between adolescents’ growth in attachment avoidance and mothers’ growth in attachment anxiety could also reflect different, yet simultaneous, reactions to the same external pressures. For example, grandparents becoming ill could contribute to both mothers’ and adolescents’ growth in general attachment insecurity. Adolescents, who are faced with the developmental challenge of becoming more self-reliant, may respond to this stressor by becoming increasingly avoidant towards close others (Erikson, 1994). Alternatively, mothers, who are faced with the developmental task of investing in social roles and relationships (see Lodi-Smith & Roberts, 2007), may respond to this same stressor with increased attachment anxiety, trying to cling too tightly to close others.
Besides these coordinated patterns of long-term growth, mothers and adolescents also demonstrated coordinated patterns of shorter-term fluctuations in attachment security. At times when adolescents reported particularly high levels of general attachment anxiety (i.e., expressed greater concerns about abandonment and the availability of close others), their mothers tended to report lower levels of both attachment anxiety and avoidance. Of note, these results do not align with the possibilities we outlined earlier (see Table 1). It could be the case that, on occasions when adolescents were feeling particularly worried about their close relationships, they tended to turn to their secure bases, their mothers, to ease their distress. These moments of mother-adolescent bonding may have evoked feelings of warmth and security among mothers.
Coordination among Fathers’ and Adolescents’ Attachment Styles
We found no coordination between fathers’ and adolescents’ attachment styles over time. This lack of coordination could reflect, in part, a greater disconnect between adolescents and their fathers, compared to adolescents and their mothers. According to the American Time Use Survey, men who are caregivers report spending about 25–33% less time per week caring for children, on average, than women (U.S. Bureau of Labor Statistics, 2023, based on the past 15 years). Fathers also tend to be less aware than mothers of their children’s whereabouts, associates, and activities (e.g., Cottrell et al., 2003; Jones et al., 2015b). Research suggests that, during the teenage years, most adolescents continue to view their mothers as their primary attachment figures and an important source for their attachment-related needs (especially secure base needs), but their fathers tend to get “demoted” to a status equal to or less than their best friends. Markiewicz and colleagues (2006) found that adolescents were more likely to turn to their best friends than their fathers for all of their attachment-related needs. Only about a third of adolescents turned to their fathers for any of their attachment-related needs. Other research by Rosenthal and Kobak (2010) found that best friends and fathers were equally likely to hold the primary positions in adolescents’ attachment hierarchies.
When interpreting the present findings in light of prior work, one caveat to keep in mind is that fathers have been underrepresented in developmental research for decades (Cabrera et al., 2018; Fabiano & Caserta, 2018; Parent et al., 2017; Yaremych & Persky, 2023), and studies that have included fathers have often not been well-positioned to study their experiences (e.g., small sample sizes of fathers). This has likely introduced substantial bias into our understanding of the role of father-child relationships in adolescent development and family processes. We would caution against placing too much interpretive weight in the present findings for fathers until they have been replicated among larger samples of father-child dyads. Furthermore, it is important to continue collecting new data on fathers’ and adolescents’ attachment styles as gender and parenting roles continue to evolve. Future research can investigate whether family-specific expectations about gender roles and caregiving responsibilities relate to the extent of coordination between adolescents’ and fathers’ attachment styles versus adolescents’ and mothers’ attachment styles.
Another avenue for future research is to examine the extent of coordination among adolescents’ attachment security and that of their peers (e.g., friends, partners). The extent to which adolescents exhibit coordinated patterns of security with their family members versus their potentially more turbulent peers could have implications for their mental health, academic success, and later relationship functioning.
The Mirroring of Mothers’ and Fathers’ Attachment Styles
The present study revealed a mirroring dynamic between mothers’ and fathers’ shorter-term fluctuations in attachment security as their children navigated the teenage years. At times when either parent reported feeling particularly anxious or avoidant in their close relationships, their co-parent reported feeling particularly insecure as well (i.e., anxious and avoidant). This finding may suggest that mothers and fathers tended to be highly in tune with one another’s current (i.e., time-specific) attachment-related signals. Most parents in the present study reported being married to one another, and thus, were also romantic partners. Research suggests that adults in well-established romantic relationships tend to rely on their partners as the primary source for their attachment-related needs (e.g., Heffernan et al., 2012; Trinke & Bartholomew, 1997). Consequently, mothers and fathers in our sample likely spent a considerable amount of time together and turned to one another when they were feeling distressed. As such, some of their coordinated fluctuations in security may reflect a “spillover” of attachment-related feelings from one partner to the other (Erel & Burman, 1995). For instance, one partner’s feelings of insecurity regarding external relationships with friends or family members may have led the other partner to feel insecure about their own relationships. Mothers and fathers may have also reported simultaneous dips in security due to conflict within their own relationship or co-occurring boosts in security following joint participation in security-enhancing activities (Arriaga et al., 2018).
It is interesting to compare the patterns of coordination we observed between mothers’ and adolescents’ attachment styles across time to the patterns of coordination between mothers’ and fathers’ attachment styles. Attachment scholars maintain that one of the critical distinctions between mother-child versus romantic attachment bonds is the level of reciprocity involved in each relationship. Mother-child attachment bonds involve a largely, if not entirely, one-sided flow of resources, in which mothers provide for their children’s attachment-related needs but not vice versa (Hazan & Zeifman, 1999). In contrast, mothers and fathers often share a reciprocal romantic attachment bond. Aligning with this distinction, we found that mothers’ and adolescents’ attachment styles exhibited an asymmetric demand-withdraw dynamic (in which mothers who pushed for closeness had adolescents who pulled away), whereas mothers’ and fathers’ levels of attachment (in)security moved up and down in tandem (see also Hudson et al., 2014).
The present study also revealed that mothers who reported higher initial levels of avoidance tended to be coupled with fathers who were also higher in avoidance. The similarity-attraction hypothesis offers one possible explanation for this finding: People with similar levels of avoidance may be more likely to be attracted to one another, and therefore, more likely to have children together (see Holmes & Johnson, 2009, for a review). Interestingly, research by Klohnen and Luo (2003) found that attachment avoidance was a stronger predictor of attraction to romantic partners than attachment anxiety (that is, people seem to prioritize their preferences for intimacy and independence when selecting a partner; see also Holmes & Johnson, 2009). Nonetheless, it is also possible that parents who are coupled may have become more similar in their levels of avoidance over the course of their relationship. Further research is needed to examine how co-parents’ attachment styles might change together throughout different stages of parenting and the unique demands they bring (e.g., adolescents leaving home for work or college).
Mean-Level Growth in General Attachment Security among Adolescents and their Parents
In addition to our primary findings, the present research sheds new light on mean-level changes in general attachment security experienced by adolescents. Few studies to date have assessed adolescents’ attachment security across at least three time points, which is required to estimate growth trajectories across adolescence. As such, the present study offers a more fine-tuned picture of growth in general attachment (in)security from mid- to late-adolescence (ages 14–18). On average, adolescents increased in both general attachment anxiety and attachment avoidance as they navigated the teenage years (i.e., they became less secure). Further, we found that mothers increased in general attachment anxiety, on average, as their children navigated adolescence, whereas fathers of adolescents tended to increase in avoidance. The average increase in mothers’ attachment anxiety stands in contrast to the slight decline in attachment anxiety across middle adulthood found in several large-scale studies (Chopik et al., 2013; Chopik et al., 2019; Chopik & Edelstein, 2014; Hudson et al., 2015). This suggests that mothers might undergo an emotionally salient transition when their children become adolescents that leads their trajectories to deviate from average trends, at least for a few years. Though beyond the scope of this paper, we discuss how these findings contribute to understanding age-related trends in attachment security in the supplement.
Limitations
The present study has several strengths, including a racially diverse sample, a multi-wave design, and data from multiple family members. Nonetheless, it also has several limitations. First, participating families were relatively financially comfortable, on average (median household income = $100,000), and the study sample only included different-sex parents (i.e., mother-father dyads), the majority of which were married (74%). Additional research is needed to test whether the present findings may generalize to more economically diverse samples and samples involving different family structures (e.g., families with same-sex parents, single parent families, families with non-married parents, parents living apart).
Second, the data are correlational in nature and thus cannot support causal conclusions. The observed associations could be explained by causal processes flowing in multiple directions or external pressures that shape both adolescents’ and parents’ attachment styles.
Third, in the present research, we define shorter-term fluctuations in attachment styles as participants’ time-specific residuals, or deviations, from their full-study growth trajectories of attachment security. Although these residuals are short-term in that they are transient and do not contribute to parents’ and adolescents’ long-term growth in attachment security, we only assessed parents’ and adolescents’ attachment styles once per year. Therefore, we do not know how long these fluctuations may have lasted within the one-year test-retest intervals. Previous work has found that people demonstrate substantial fluctuations in attachment anxiety and avoidance on a day-to-day and month-to-month basis (Davila & Sargent, 2003; Dugan et al., 2023; Girme et al., 2018). However, further research involving intensive longitudinal designs is needed to investigate the coordination of parents’ and adolescents’ fluctuations in attachment security across shorter intervals. Daily diary studies can offer exciting new insights into attachment processes that occur at the daily level. Such studies would also allow for the examination of cross-lagged associations using more complex models like Latent Curve Models with Structured Residuals (LCM-SR; see Curran et al., 2014), which could better capture day-to-day “spillover effects” from parents’ relationships to parent-child relationships (Enger, 1988; Erel & Burman, 1995). For example, an unresolved fight between parents could lead to more negative interactions between each parent and their adolescent in the days that follow, thereby increasing their child’s attachment insecurity. Studies with measurement burst designs, in which people are studied over longer periods with shorter “bursts” (e.g., 14-day, one month) of experience sampling in between, could offer additional insights into both long-term and short-term attachment processes during adolescence (Nestler, 2021; Sliwinksi, 2008).
Third, the number of fathers who participated was less than ideal for estimating fathers’ long-term growth in attachment security. This limitation could partially explain the lack of coordination between fathers’ and adolescents’ attachment styles. Because there was little-to-no variation among the slopes of fathers’ trajectories of attachment security, this excluded the possibility of finding coordinated long-term changes in fathers’ and adolescents’ attachment styles. However, we also found no associations between fathers’ and adolescents’ initial levels or shorter-term fluctuations in attachment security, which we did find for mothers’ and fathers’ attachment styles. The lack of variation among fathers’ slopes of attachment security also suggested that triadic models3, which could reveal more complex patterns of coordinated change among mothers, fathers, and adolescents, would not be advisable given the data. Examining attachment dynamics across triads and in relation to other family members (e.g., siblings) is an important direction for future research. Additionally, future studies with larger total samples can examine the extent to which coordinated changes in family members’ attachment security might be moderated by demographic characteristics such as gender identity, sex, race, ethnicity, culture, and socioeconomic status.
Lastly, the present study focused on parents’ and adolescents’ general attachment styles (i.e., generalized across close relationships). This allowed for consistency of measurement across all three family members and dyads assessed and accounted for the importance of both parental and peer relationships in guiding adolescents’ development (Steinberg & Morris, 2001). Attachment researchers suggest that people also develop relationship-specific attachment styles that reflect their thoughts, feelings, and behaviors towards specific close others, including their parents, partners, and friends (Collins & Read, 1994; Overall et al., 2003). Future studies can expand upon this work by assessing both general and relationship-specific attachment styles and examining how the strength of the observed associations may differ across levels of specificity. For example, research may find even stronger associations between changes in mothers’ attachment anxiety and changes in adolescents’ mother-specific attachment avoidance (Fraley et al., 2011). Assessing changes in adolescents’ mother-specific attachment avoidance might better capture the extent to which they are “pulling away” from their mothers specifically. Some adolescents who demonstrate increases in general attachment anxiety and avoidance may instead be encountering difficulties in navigating friendships and romantic relationships. If adolescents turn to their mothers for support at times when they encounter external relationship difficulties, there may actually be negative associations between certain forms of adolescents’ attachment avoidance (e.g., partner-specific avoidance, best friend-specific avoidance) and mothers’ attachment anxiety. Moreover, associations might emerge between fathers’ and adolescents’ attachment styles, if adolescents’ father-specific attachment styles are assessed. Future studies that measure relationship-specific attachment styles can also investigate more nuanced questions regarding how relationship-specific experiences may bleed into other family member dyads (Enger, 1988; Erel & Burman, 1995). For example, following a conflict with one parent, an adolescent might seek support from another parent (or family member). Depending on the other parent’s response, the adolescent may exhibit different or similar changes in their specific attachment style towards each parent; the conflict-involved parent may also experience changes in their partner-specific attachment style, if the two parents are also partners. As this example illustrates, assessing relationship-specific attachment styles can open the door to studying more complex family dynamics from an attachment-informed lens.
Conclusion
During the teenage years, adolescents and their parents must strike a new balance in their relationship that allows adolescents greater autonomy while also ensuring that they continue to return to their secure base (i.e., their parents) when needed. The present research suggests that mothers’ and adolescents’ patterns of growth in attachment security exhibit a push and pull dynamic across the teenage years. Mothers who report higher initial levels of attachment anxiety—who are concerned that close others will reject or abandon them—tend to have children who report higher levels of avoidance at the beginning of adolescence, who feel uncomfortable confiding in others and relying on them to support their needs and goals. Furthermore, mothers who increase in attachment anxiety as their children navigate the teenage years—who try to hold on too tightly to their loved ones—have adolescent children who become increasingly closed off and avoidant. During this challenging transition, mothers and fathers appear to remain highly in tune with one another’s current attachment-related signals. Echoing a key tenet of family systems theory–that different family subsystems (e.g., parent-adolescent, parent-parent) can operate in different ways (Minuchin, 1985)—our findings indicate that the parent-parent subsystem demonstrates a different dynamic: Mothers and fathers mirrored each other’s patterns of attachment security as their children navigated their teenage years, reporting similar initial levels of attachment avoidance and synchronized shorter-term fluctuations in attachment anxiety and avoidance across time.
Supplementary Material
Public Significance Statement:
This study suggests that parents and adolescents show coordinated changes in attachment security over time, highlighting the importance of family dynamics during the teenage years. Mothers who become increasingly worried that close others will abandon them tend to have adolescents who “pull away” and become more avoidant. Mothers and fathers report co-occurring ups and downs in attachment security, remaining in tune with one another as their children navigate adolescence.
Acknowledgements:
This research and the preparation of this article were supported by the National Institutes of Health (NIH; R01DA018647, R21DA025550, T32MH015755).
Footnotes
The study materials, analysis code, and de-identified data are available on Open Science Framework (OSF; https://osf.io/85kqg/; Dugan, 2024). We pre-registered the analysis plan on OSF (https://osf.io/2uqsw; Dugan, 2023).
Earlier, categorical approaches to assessing attachment styles often assigned people to one of four prototypical attachment styles: secure, preoccupied, dismissing-avoidant, and fearful-avoidant (Bartholomew & Horowitz, 1991). However, research has shown that individual differences in attachment are more appropriately conceptualized as continuously distributed, with empirical evidence supporting dimensional approaches to assessing attachment over categorical approaches (Fraley & Waller, 1998; Fraley et al., 2015). Within the dimensional framework of attachment (Brennan et al., 1998; see Gillath et al., 2016 for a review), “attachment security” is conceptualized as being low in attachment anxiety and low in attachment avoidance. Throughout this paper, we use the term “secure” as shorthand to describe people who are relatively low on both dimensions of attachment insecurity. In doing so, we are not referring to a separate category or dimension of attachment; rather, “attachment security” exists within the two-dimensional space defined by (low) attachment anxiety and (low) attachment avoidance.
The “parent-parent subsystem” refers to the relationship between the parents of a family.
We included triadic growth curve models in our pre-registration. After finding little-to-no variation among fathers’ slopes for attachment anxiety and avoidance, we decided not to move forward with these analyses.
References
- Ainsworth MS (1989). Attachments beyond infancy. American Psychologist, 44(4), 709–716. 10.1037/0003-066X.44.4.709 [DOI] [PubMed] [Google Scholar]
- Allen JP, Hauser ST, Bell KL, & O’Connor TG (1994). Longitudinal assessment of autonomy and relatedness in adolescent-family interactions as predictors of adolescent ego development and self-esteem. Child Development, 65(1), 179–194. 10.1111/j.1467-8624.1994.tb00743.x [DOI] [PubMed] [Google Scholar]
- Allen JP, & Land D (1999). Attachment in adolescence. In Cassidy J & Shaver PR (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 319–335). Guilford. [Google Scholar]
- Allen JP, & Tan J (2016). The multiple facets of attachment in adolescence. In Cassidy J & Shaver PR (Eds.), Handbook of attachment: Theory, research, and clinical applications (3rd ed., pp. 399–415). Guilford. [Google Scholar]
- Arriaga XB, Kumashiro M, Simpson JA, & Overall NC (2018). Revising working models across time: Relationship situations that enhance attachment security. Personality and Social Psychology Review, 22(1), 71–96. 10.1177/1088868317705257 [DOI] [PubMed] [Google Scholar]
- Bowen M (1966). The use of family theory in clinical practice. Comprehensive Psychiatry, 7(5), 345–374. [DOI] [PubMed] [Google Scholar]
- Bowen M (1978). Family therapy in clinical practice. Jason Aronson, Inc. [Google Scholar]
- Bowlby J (1969/1982). Attachment and loss: Volume I: Attachment. Basic Books. [Google Scholar]
- Bowlby J (1973). Attachment and loss: Volume II: Separation, anxiety and anger. Basic Books. [Google Scholar]
- Bowlby J (1980). Attachment and loss: Volume III: Loss, sadness and depression. Basic Books. [Google Scholar]
- Barnes GM, Hoffman JH, Welte JW, Farrell MP, & Dintcheff BA (2007). Adolescents’ time use: Effects on substance use, delinquency and sexual activity. Journal of Youth and Adolescence, 36(5), 697–710. 10.1007/s10964-006-9075-0 [DOI] [Google Scholar]
- Becker-Stoll F, Delius A, & Scheitenberger S (2001). Adolescents’ nonverbal emotional expressions during negotiation of a disagreement with their mothers: An attachment approach. International Journal of Behavioral Development, 25(4), 344–353. 10.1080/01650250143000102 [DOI] [Google Scholar]
- Becker-Stoll F, & Fremmer-Bombik E (1997, April). Adolescent-mother interaction and attachment: A longitudinal study [Conference session]. Biennial Meetings of the Society for Research in Child Development, Washington, DC, United States. [Google Scholar]
- Berlin LJ, & Cassidy J (1999). Relations among relationships: Contributions from attachment theory and research. In Cassidy J & Shaver PR (Eds.), Handbook of attachment: Theory, research and clinical applications (pp. 688–712). Guilford. [Google Scholar]
- Besser A, & Priel B (2005). The apple does not fall far from the tree: Attachment styles and personality vulnerabilities to depression in three generations of women. Personality and Social Psychology Bulletin, 31(8), 1052–1073. 10.1177/0146167204274082 [DOI] [PubMed] [Google Scholar]
- Brennan KA, Clark CL, & Shaver PR (1998). Self-report measurement of adult attachment: An integrative overview. In Simpson JA & Rholes WS (Eds.), Attachment theory and close relationships (pp. 46–76). Guilford. [Google Scholar]
- Bretherton I, & Munholland KA (1999). Internal working models in attachment relations: A construct revisited. In Cassidy J & Shaver PR (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 89–111). Guilford. [Google Scholar]
- Byng-Hall J (1999). Family and couple therapy: Toward greater security. In Cassidy J & Shaver PR (Eds.), Handbook of attachment: Theory, research and clinical applications (pp. 625–645). Guilford. [Google Scholar]
- Byng-Hall J, & Stevenson-Hinde J (1991). Attachment relationships within a family system. Infant Mental Health Journal, 12(3), 187–200. [Google Scholar]
- Cabrera NJ, Volling BL, & Barr R (2018). Fathers are parents, too! Widening the lens on parenting for children’s development. Child Development Perspectives, 12(3), 152–157. 10.1111/cdep.12275 [DOI] [Google Scholar]
- Chopik WJ, & Edelstein RS (2014). Age differences in romantic attachment around the world. Social Psychological and Personality Science, 5(8), 892–900. 10.1177/1948550614538460 [DOI] [Google Scholar]
- Chopik WJ, Edelstein RS, & Fraley RC (2013). From the cradle to the grave: Age differences in attachment from early adulthood to old age. Journal of Personality, 81(2), 171–183. 10.1111/j.1467-6494.2012.00793.x [DOI] [PubMed] [Google Scholar]
- Chopik WJ, Edelstein RS, & Grimm KJ (2019). Longitudinal changes in attachment orientation over a 59-year period. Journal of Personality and Social Psychology, 116(4), 598–611. 10.1037/pspp0000167 [DOI] [PubMed] [Google Scholar]
- Collins NL, Cooper ML, Albino A, & Allard L (2002). Psychosocial vulnerability from adolescence to adulthood: A prospective study of attachment style differences in relationship functioning and partner choice. Journal of Personality, 70(6), 965–1008. 10.1111/1467-6494.05029 [DOI] [PubMed] [Google Scholar]
- Collins NL, & Read SJ (1994). Cognitive representations of attachment: The structure and function of working models. In Bartholomew K & Perlman D (Eds.), Attachment processes in adulthood (pp. 53–90). Kingsley. [Google Scholar]
- Cottrell L, Li X, Harris C, D’Alessandri D, Atkins M, Richardson B, & Stanton B (2003). Parent and adolescent perceptions of parental monitoring and adolescent risk involvement. Parenting: Science and Practice, 3(3), 179–195. 10.1207/S15327922PAR0303_01 [DOI] [Google Scholar]
- Cowan PA (1997). Beyond meta-analysis: A plea for a family systems view of attachment. Child Development, 68(4), 601–603. 10.2307/1132111 [DOI] [PubMed] [Google Scholar]
- Cowan PA, & Cowan CP (2009). Couple relationships: A missing link between adult attachment and children’s outcomes. Attachment & Human Development, 11(1), 1–4. 10.1080/14616730802500149 [DOI] [PubMed] [Google Scholar]
- Cox MJ, & Paley B (2003). Understanding families as systems. Current Directions in Psychological Science, 12(5), 193–196. 10.1111/1467-8721.01259 [DOI] [Google Scholar]
- Curran PJ, Howard AL, Bainter SA, Lane ST, & McGinley JS (2014). The separation of between-person and within-person components of individual change over time: A latent curve model with structured residuals. Journal of Consulting and Clinical Psychology, 82(5), 879–894. 10.1037/a0035297 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Davila J, & Sargent E (2003). The meaning of life (events) predicts changes in attachment security. Personality and Social Psychology Bulletin, 29(11), 1383–1395. 10.1177/0146167203256374 [DOI] [PubMed] [Google Scholar]
- Dugan KA (2023, January 28). Coordination among parents’ and adolescents’ attachment patterns. Open Science Framework. https://osf.io/2uqsw [Google Scholar]
- Dugan KA (2024, April 22). Coordination among parents’ and adolescents’ attachment patterns. Open Science Framework. https://osf.io/85kqg/ [Google Scholar]
- Dugan KA, Fraley RC, Gillath O, & Deboeck PR (2024). Testing the canalization hypothesis of attachment theory: Examining within-subject variation in attachment security. Journal of Personality and Social Psychology, 126(3), 511–541. 10.1037/pspp0000488 [DOI] [PubMed] [Google Scholar]
- Ein-Dor T, Mikulincer M, & Shaver PR (2011). Attachment insecurities and the processing of threat-related information: Studying the schemas involved in insecure people’s coping strategies. Journal of Personality and Social Psychology, 101(1), 78–93. 10.1037/a0022503 [DOI] [PubMed] [Google Scholar]
- Enger A (1988). The interrelatedness of marriage and the motherchild relationship. In Hinde R & Stevenson-Hinde J (Eds.), Relationships within families: Mutual influences (pp. 83–103). Oxford University Press. [Google Scholar]
- Erel O, & Burman B (1995). Interrelatedness of marital relations and parent-child relations: A meta-analytic review. Psychological Bulletin, 118(1), 108–132. 10.1037/0033-2909.118.1.108 [DOI] [PubMed] [Google Scholar]
- Erikson EH (1994). Identity and the life cycle. WW Norton & company. [Google Scholar]
- Fabiano GA, & Caserta A (2018). Future directions in father inclusion, engagement, retention, and positive outcomes in child and adolescent research. Journal of Clinical Child and Adolescent Psychology, 47(5), 847–862. 10.1080/15374416.2018.1485106 [DOI] [PubMed] [Google Scholar]
- Falgares G, Marchetti D, De Santis S, Carrozzino D, Kopala-Sibley DC, Fulcheri M, & Verrocchio MC (2017). Attachment styles and suicide-related behaviors in adolescence: The mediating role of self-criticism and dependency. Frontiers in Psychiatry, 8, Article 36. 10.3389/fpsyt.2017.00036 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Feeney BC, Cassidy J, & Ramos-Marcuse F (2008). The generalization of attachment representations to new social situations: Predicting behavior during initial interactions with strangers. Journal of Personality and Social Psychology, 95(6), 1481–1498. 10.1037/a0012635 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Feeney BC, & Collins NL (2001). Predictors of caregiving in adult intimate relationships: An attachment theoretical perspective. Journal of Personality and Social Psychology, 80(6), 972–994. 10.1037/0022-3514.80.6.972 [DOI] [PubMed] [Google Scholar]
- Fogel A (2000). Systems, attachment, and relationships. Human Development, 43(6), 314–320. 10.1159/000022692 [DOI] [Google Scholar]
- Fraley RC, Heffernan ME, Vicary AM, & Brumbaugh CC (2011). The experiences in close relationships—Relationship Structures Questionnaire: A method for assessing attachment orientations across relationships. Psychological Assessment, 23(3), 615–625. 10.1037/a0022898 [DOI] [PubMed] [Google Scholar]
- Fraley RC, Hudson NW, Heffernan ME, & Segal N (2015). Are adult attachment styles categorical or dimensional? A taxometric analysis of general and relationship-specific attachment orientations. Journal of Personality and Social Psychology, 109(2), 354–368. 10.1037/pspp0000027 [DOI] [PubMed] [Google Scholar]
- Fraley RC, & Roisman GI (2019). The development of adult attachment styles: Four lessons. Current Opinion in Psychology, 25, 26–30. 10.1016/j.copsyc.2018.02.008 [DOI] [PubMed] [Google Scholar]
- Fraley RC, & Waller NG (1998). Adult attachment patterns: A test of the typological model. In Simpson JA & Rholes WS (Eds.), Attachment theory and close relationships (pp. 77–114). Guilford. [Google Scholar]
- Gallarin M, & Alonso-Arbiol I (2012). Parenting practices, parental attachment and aggressiveness in adolescence: A predictive model. Journal of Adolescence, 35(6), 1601–1610. 10.1016/j.adolescence.2012.07.002 [DOI] [PubMed] [Google Scholar]
- Girme YU, Agnew CR, VanderDrift LE, Harvey SM, Rholes WS, & Simpson JA (2018). The ebbs and flows of attachment: Within-person variation in attachment undermine secure individuals’ relationship wellbeing across time. Journal of Personality and Social Psychology, 114(3), 397–421. 10.1037/pspi0000115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hazan C, & Zeifman D (1999). Pair bonds as attachments: Evaluating the evidence. In Cassidy J & Shaver PR (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 336–354). Guilford. [Google Scholar]
- Heffernan ME, Fraley RC, Vicary AM, & Brumbaugh CC (2012). Attachment features and functions in adult romantic relationships. Journal of Social and Personal Relationships, 29(5), 671–693. 10.1177/0265407512443435 [DOI] [Google Scholar]
- Holmes BM, & Johnson KR (2009). Adult attachment and romantic partner preference: A review. Journal of Social and Personal Relationships, 26(6–7), 833–852. 10.1177/0265407509345653 [DOI] [Google Scholar]
- Hudson NW, Fraley RC, Brumbaugh CC, & Vicary AM (2014). Coregulation in romantic partners’ attachment styles: A longitudinal investigation. Personality and Social Psychology Bulletin, 40(7), 845–857. 10.1177/0146167214528989 [DOI] [PubMed] [Google Scholar]
- Hudson NW, Fraley RC, Chopik WJ, & Heffernan ME (2015). Not all attachment relationships develop alike: Normative cross-sectional age trajectories in attachment to romantic partners, best friends, and parents. Journal of Research in Personality, 59,44–55. 10.1016/j.jrp.2015.10.001 [DOI] [Google Scholar]
- Jones JD, & Cassidy J (2014). Parental attachment style: Examination of links with parent secure base provision and adolescent secure base use. Attachment & Human Development, 16(5), 437–461. 10.1080/14616734.2014.921718 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jones JD, Cassidy J, & Shaver PR (2015a). Parents’ self-reported attachment styles: A review of links with parenting behaviors, emotions, and cognitions. Personality and Social Psychology Review, 19(1), 44–76. 10.1177/1088868314541858 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jones JD, Ehrlich KB, Lejuez CW, & Cassidy J (2015b). Parental knowledge of adolescent activities: Links with parental attachment style and adolescent substance use. Journal of Family Psychology, 29(2), 191–200. 10.1037/fam0000070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jones JD, Fraley RC, Ehrlich KB, Stern JA, Lejuez CW, Shaver PR, & Cassidy J (2018). Stability of attachment style in adolescence: An empirical test of alternative developmental processes. Child Development, 89(3), 871–880. 10.1111/cdev.12775 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Khan F, Chong JY, Theisen JC, Fraley RC, Young JF, & Hankin BL (2020). Development and change in attachment: A multiwave assessment of attachment and its correlates across childhood and adolescence. Journal of Personality and Social Psychology, 118(6), 1188–1206. 10.1037/pspi0000211 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Klohnen EC, & Luo S (2003). Interpersonal attraction and personality: What is attractive—self similarity, ideal similarity, complementarity or attachment security? Journal of Personality and Social Psychology, 85(4), 709–722. 10.1037/0022-3514.85.4.709 [DOI] [PubMed] [Google Scholar]
- Kobak RR, Cole HE, Ferenz-Gillies R, Fleming WS, & Gamble W (1993). Attachment and emotion regulation during mother-teen problem solving: A control theory analysis. Child Development, 64(1), 231–245. 10.2307/1131448 [DOI] [PubMed] [Google Scholar]
- Kobak R, Ferenz-Gillies R, Everhart E, & Seabrook L (1994). Maternal attachment strategies and emotion regulation with adolescent offspring. Journal of Research on Adolescence, 4(4), 553–566. 10.1207/s15327795jra0404_7 [DOI] [Google Scholar]
- Kobak R, Rosenthal N, & Serwik A (2005). The attachment hierarchy in middle childhood: conceptual and methodological issues. In Kerns KA & Richardson RA (Eds.), Attachment in middle childhood (pp. 71–88). Guilford. [Google Scholar]
- Kunce LJ, & Shaver PR (1994). An attachment-theoretical approach to caregiving in romantic relationships. In Bartholomew K & Perlman D (Eds.), Attachment processes in adulthood (pp. 205–237). Jessica Kingsley Publishers. [Google Scholar]
- Li M, Chen X, Gong H, Ji W, Wang W, Liang S, & Kong A (2021). The predictive effect of parental adult attachment on parent-adolescent attachment: The mediating role of harsh parenting. Frontiers in Psychology, 12, 710167–710167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lodi-Smith J, & Roberts BW (2007). Social investment and personality: A meta-analysis of the relationship of personality traits to investment in work, family, religion, and volunteerism. Personality and Social Psychology Review, 11(1), 1–19. 10.1177/1088868306294590 [DOI] [PubMed] [Google Scholar]
- Lougheed JP (2020). Parent–adolescent dyads as temporal interpersonal emotion systems. Journal of Research on Adolescence, 30(1), 26–40. 10.1111/jora.12526 [DOI] [PubMed] [Google Scholar]
- Markiewicz D, Lawford H, Doyle AB, & Haggart N (2006). Developmental differences in adolescents’ and young adults’ use of mothers, fathers, best friends, and romantic partners to fulfill attachment needs. Journal of Youth and Adolescence, 35(1), 127–140. 10.1007/s10964-005-9014-5 [DOI] [Google Scholar]
- Martin MJ, Sturge-Apple ML, Davies PT, Romero CV, & Buckholz A (2017). A process model of the implications of spillover from coparenting conflicts into the parent–child attachment relationship in adolescence. Development and Psychopathology, 29(2), 417–431. 10.1017/S0954579417000086 [DOI] [PMC free article] [PubMed] [Google Scholar]
- McArdle JJ (2009). Latent variable modeling of differences and changes with longitudinal data. Annual Review of Psychology, 60, 577–605. 10.1146/annurev.psych.60.110707.163612 [DOI] [PubMed] [Google Scholar]
- Mikulincer M, & Shaver PR (2016). Attachment in adulthood: Structure, dynamics, and change (2nd ed.). Guilford. [Google Scholar]
- Mikulincer M, & Shaver PR (2019). Attachment orientations and emotion regulation. Current Opinion in Psychology, 25, 6–10. 10.1016/j.copsyc.2018.02.006 [DOI] [PubMed] [Google Scholar]
- Minuchin P (1985). Families and individual development: Provocations from the field of family therapy. Child Development, 56(2), 289–302. 10.2307/1129720 [DOI] [PubMed] [Google Scholar]
- Moretti MM, Pasalich DS, & O’Donnell KA (2017). An attachment-based program for parents and adolescents. In Steele H & Steele M (Eds.), Handbook of attachment-based interventions (pp. 375–400). Guilford. [Google Scholar]
- Moretti MM, & Peled M (2004). Adolescent-parent attachment: Bonds that support healthy development. Paediatrics & Child Health, 9(8), 551–555. 10.1093/pch/9.8.551 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Muthén LK and Muthén BO (2017). Mplus User’s Guide. Eighth Edition. Muthén & Muthén. [Google Scholar]
- Nestler S (2021). Modeling intraindividual variability in growth with measurement burst designs. Structural Equation Modeling, 28(1), 28–39. 10.1080/10705511.2020.1757455 [DOI] [Google Scholar]
- Obegi JH, Morrison TL, & Shaver PR (2004). Exploring intergenerational transmission of attachment style in young female adults and their mothers. Journal of Social and Personal Relationships, 21(5), 625–638. 10.1177/0265407504045891 [DOI] [Google Scholar]
- Odgers CL, & Jensen MR (2020). Annual research review: Adolescent mental health in the digital age: Facts, fears, and future directions. Journal of Child Psychology and Psychiatry, 61(3), 336–348. 10.1111/jcpp.13190 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Overall NC, Fletcher GJO, & Friesen MD (2003). Mapping the intimate relationship mind: Comparisons between three models of attachment representations. Personality and Social Psychology Bulletin, 29(12), 1479–1493. 10.1177/0146167203251519 [DOI] [PubMed] [Google Scholar]
- Pace U, & Zappulla C (2011). Problem behaviors in adolescence: The opposite role played by insecure attachment and commitment strength. Journal of Child and Family Studies, 20, 854–862. 10.1007/s10826-011-9453-4 [DOI] [Google Scholar]
- Paley B, & Hajal NJ (2022). Conceptualizing emotion regulation and coregulation as family-level phenomena. Clinical Child and Family Psychology Review, 25(1), 19–43. doi: 10.1007/s10567-022-00378-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Parent J, Forehand R, Pomerantz H, Peisch V, & Seehuus M (2017). Father participation in child psychopathology research. Journal of Abnormal Child Psychology, 45(7), 1259–1270. 10.1007/s10802-016-0254-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- R Core Team. (2022). R: A language and environment for statistical computing. https://www.R-project.org/ [Google Scholar]
- Roelofs J, Meesters C, & Muris P (2008). Correlates of self-reported attachment (in) security in children: The role of parental romantic attachment status and rearing behaviors. Journal of Child and Family Studies, 17, 555–566. 10.1007/s10826-007-9174-x [DOI] [Google Scholar]
- Rosenthal NL, & Kobak R (2010). Assessing adolescents’ attachment hierarchies: Differences across developmental periods and associations with individual adaptation. Journal of Research on Adolescence, 20(3), 678–706. 10.1111/j.1532-7795.2010.00655.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ruhl H, Dolan EA, & Buhrmester D (2015). Adolescent attachment trajectories with mothers and fathers: The importance of parent–child relationship experiences and gender. Journal of Research on Adolescence, 25(3), 427–442. 10.1111/jora.12144 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Selcuk E, Günaydin G, Sumer N, Harma M, Salman S, Hazan C, Dogruyol B, & Ozturk A (2010). Self-reported romantic attachment style predicts everyday maternal caregiving behavior at home. Journal of Research in Personality, 44(4), 544–549. 10.1016/j.jrp.2010.05.007 [DOI] [Google Scholar]
- Sliwinski MJ (2008). Measurement-burst designs for social health research. Social and Personality Psychology Compass, 2(1), 245–261. 10.1111/j.1751-9004.2007.00043.x [DOI] [Google Scholar]
- Sroufe LA, & Waters E (1977). Attachment as an organizational construct. Child Development, 48(4), 1184–1199. 10.2307/1128475 [DOI] [Google Scholar]
- Steinberg L, & Morris AS (2001). Adolescent development. Annual Review of Psychology, 52(1), 83–110. 10.1146/annurev.psych.52.1.83 [DOI] [PubMed] [Google Scholar]
- Thompson RA (2021). Attachment networks and the future of attachment theory. New Directions for Child and Adolescent Development, 2021(180), 149–156. 10.1002/cad.20455 [DOI] [PubMed] [Google Scholar]
- Trinke SJ, & Bartholomew K (1997). Hierarchies of attachment relationships in young adulthood. Journal of Social and Personal Relationships, 14(5), 603–625. 10.1177/0265407597145002 [DOI] [Google Scholar]
- U.S. Bureau of Labor Statistics. (2023). American Time Use Survey. https://www.bls.gov/tus/ [Google Scholar]
- van IJzendoorn MH (1995). Adult attachment representations, parental responsiveness, and infant attachment: A meta-analysis on the predictive validity of the Adult Attachment Interview. Psychological Bulletin, 117(3), 387–403. 10.1037/0033-2909.117.3.387 [DOI] [PubMed] [Google Scholar]
- von Bertalanffy L (1968). General system theory: Foundations, development, applications. G. Braziller. [Google Scholar]
- Wei M, Russell DW, Mallinckrodt B, & Vogel DL (2007). The Experiences in Close Relationship Scale (ECR)-short form: Reliability, validity, and factor structure. Journal of Personality Assessment, 88(2), 187–204. 10.1080/00223890701268041 [DOI] [PubMed] [Google Scholar]
- Wrzus C, Hänel M, Wagner J, & Neyer FJ (2013). Social network changes and life events across the life span: A meta-analysis. Psychological Bulletin, 139(1), 53–80. 10.1037/a0028601 [DOI] [PubMed] [Google Scholar]
- Yaremych HE, & Persky S (2023). Recruiting fathers for parenting research: An evaluation of eight recruitment methods and an exploration of fathers’ motivations for participation. Parenting, 23(1), 1–32. 10.1080/15295192.2022.2036940 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.