Intrapersonal Emotion Regulation Difficulties and Maladaptive Interpersonal Behavior in Adolescence

Abstract

Although most research conceptualizes emotion regulation as an internal process (i.e., intrapersonal), emotions are frequently regulated in an interpersonal context. Adolescents may be particularly prone to turn to peers for assistance, given that honing emotion regulation abilities is a key task and peers become increasingly important at this stage. Adolescents each participated with a self-nominated same-gender friend (total N = 186; M age = 15.68 years). All participants provided self-reports of their own emotion regulation difficulties and their engagement in interpersonal behaviors (excessive reassurance seeking, negative feedback seeking, conversational self-focus, self-disclosure) as well as their perceptions of their friends’ engagement in excessive reassurance seeking and conversational self-focus. Using a cross-lagged panel model design, the current study tested associations between adolescents’ intrapersonal emotion regulation difficulties and engagement in self- and friend-reported interpersonal emotion regulation behaviors at 3 time points over 6 months. Gender and age group differences were considered. Results evidenced concurrent and longitudinal associations between emotion regulation difficulties and self-reported maladaptive (excessive reassurance seeking, conversational self-focus, negative feedback seeking), but not adaptive (self-disclosure), interpersonal regulatory behaviors. Friends’ reports of adolescents’ excessive reassurance seeking and conversational self-focus were associated concurrently, but not longitudinally, with self-reported emotion regulation difficulties. Implications for clinical intervention with adolescents struggling to regulate emotions are discussed.

The ability to regulate one’s emotions is a central task of adolescence (Steinberg, 2005), and adolescents who struggle with emotion regulation face increased risk for social and emotional problems (Silk et al., 2003; Zeman et al., 2006). To date, much of the empirical literature on emotion regulation focuses on intrapersonal emotion regulation (i.e., internal regulation of affect; Hofmann et al., 2016; Zaki & Williams, 2013). Researchers have been increasingly interested in the ways in which adolescents may employ their interpersonal context to assist with regulating emotion (Dixon-Gordon et al., 2015), although studies of specific interpersonal behaviors are lacking. What is more, whether adolescents’ intrapersonal emotion regulation difficulties give rise to interpersonal attempts to regulate emotion is unknown. The current study addressed these gaps by examining longitudinal associations of adolescents’ intrapersonal emotion regulation difficulties and interpersonal behaviors.

Intrapersonal Emotion Regulation

Emotion regulation (awareness, assessment, and modification of emotions; e.g., Cicchetti et al., 1991; Compas et al., 2017; Eisenberg et al., 2010), has traditionally been conceptualized as intrapersonal. Intrapersonal emotion regulation reflects internal efforts to regulate one’s own emotions. These processes involve identification of the emotion(s) to be regulated, selecting and implementing a regulation strategy, and evaluating the effectiveness of the strategy and whether additional emotion regulation is necessary (Fernandez et al., 2016).

The ability to regulate emotion develops across the lifespan and is a central task of adolescence. Emotion regulation involves higher-order cognitive processes that continue to develop through childhood (Perlman & Pelphrey, 2010; Tottenham et al., 2011). By adolescence, youths’ cognitive abilities are increasingly complex, enabling the use of internal strategies to regulate emotion such as cognitive reappraisal and distraction (Skinner & Zimmer-Gembeck, 2007; Thompson & Goodman, 2010; Zimmer-Gembeck & Skinner, 2016).

Youths’ ability to employ adaptive intrapersonal emotion regulation skills promotes positive adjustment in multiple domains (Gross, 2013; Schaäfer et al., 2017). Difficulties in intrapersonal emotion regulation, however, are associated with myriad emotional adjustment problems. Given that risk for these problems (e.g., depression) increases dramatically during adolescence (Merikangas et al., 2010), emotion regulation difficulties during this period are an important focus of empirical research. Indeed, intrapersonal emotion regulation difficulties have transdiagnostic implications (Aldao, 2012; Kring & Sloan, 2010), with meta-analyses consistently finding significant associations between youths’ intrapersonal emotion regulation difficulties and increased psychological distress (Compas et al., 2017; Schäfer et al., 2017).

Interpersonal Emotion Regulation

Emotion regulation is not solely an intrapersonal process, however. Conceptualizations of emotion regulation do include the engagement of others to assist with internal regulation of emotion (e.g., Hofmann, 2014; Morris et al., 2007). Scholars have been increasingly interested in these processes, often referred to as interpersonal emotion regulation (e.g., Zaki & Williams, 2013). Interpersonal emotion regulation is the provision of regulatory assistance between individuals (e.g., seeking regulation of one’s own emotions through help from others). It is not surprising that individuals often enlist others to regulate emotions given that the ability to regulate emotions initially develops and is honed through social interactions (Hofmann, 2014).

Specifically, caregivers provide regulation in early attachment relationships (e.g., infants’ crying elicits soothing). As development progresses, caregivers continue to socialize and model emotion identification, expression, and regulation (Morris et al., 2007). As youth mature, they rely more heavily on peers to model appropriate emotion expression and to provide regulation of one's own emotions (Compas et al., 2017). Adolescents with intrapersonal emotion regulation difficulties may be especially prone to rely on peers to help regulate their emotions. As friendships are a central focus of adolescents’ social support system (Buhrmester, 1990; Furman & Buhrmester, 1992) and active coping strategies, such as seeking social support, are associated with positive outcomes (e.g., Compas et al., 2001; Herres & Ohannessian, 2015), it is likely that youth with problems navigating their own distress utilize friends to help them do so.

Proposed Regulation Behaviors.

Scholars have been increasingly interested in the ways in which individuals may utilize interpersonal relationships to regulate emotions and the specific interpersonal behaviors they may employ to do so. Adaptively, peers offer opportunities for adolescents’ self-disclosure, the reciprocal sharing of thoughts and feelings, which occurs frequently among friends for the purposes of navigating problems and other challenges. Self-disclosure with friends provides a supportive context for modeling ways to identify and express emotion, receiving social support, and/or gaining problem solving assistance (e.g., Helsen et al., 2000; Klimes-Dougan et al., 2014). These normative functions of friendship, afforded by this active approach to seeking support for coping purposes (Compas et al., 2001), have clear social and emotional benefits for youth (Hartup, 1996; Hartup & Stevens, 1999).

There appear to be, however, more maladaptive ways that adolescents attempt to seek assistance from others to regulate their emotions. Interpersonal theories of psychopathology (e.g., Coyne, 1976; Joiner, 2002) posit specific interpersonal behaviors—such as excessive reassurance seeking, negative feedback seeking, and conversational self-focus—that may serve as attempts to downregulate negative affect during periods of distress that in fact are less than effective. Such behaviors are thought to be reinforcing to distressed individuals in the short term, as relationship partners may indeed initially offer the requested attention, reassurance, support, validation, and assistance. However, over time, the distressed individuals doubt the support provided and repeatedly request additional attention to their concerns. These behaviors are perceived by others as increasingly aversive, and relationship partners may distance themselves from the distressed individual, which can result in increased distress. In this way, such attempts to garner support from friends may reflect avoidant, emotion-focused coping strategies that actually serve to prolong distress (e.g., Lynch et al., 2001). The present study tested associations of one adaptive (self-disclosure) and three maladaptive (excessive reassurance seeking, negative feedback seeking, conversational self-focus) interpersonal regulatory behaviors with intrapersonal emotion regulation difficulties.

Excessive reassurance seeking.

Excessive reassurance seeking is the tendency to seek confirmation from others that one is truly liked or cared for (Joiner et al., 1999). Individuals, especially those who are distressed, are thought to engage in excessive reassurance seeking as a way to assuage fears about self-worth and social connectedness (Joiner et al., 1999). However, they then doubt the truthfulness of reassurances provided, attributing them to others’ feelings of pity or obligation. Although individuals engage in excessive reassurance seeking in part to regulate negative affect, excessive reassurance seeking appears to be ineffective in doing so. In fact, Van Orden and Joiner (2006) termed excessive reassurance seeking "Band-Aid emotion regulation” such that it may feel helpful in the moment but can actually worsen distress, should underlying problems go unresolved. Excessive reassurance seeking has been shown to trigger increased depressive affect and continued requests for reassurance, thus continuing a problematic cycle (Cougle et al., 2012; Prinstein et al., 2005; Nesi & Prinstein, 2015).

Conversational self-focus.

Conversational self-focus is the redirection of conversations to focus on the self (Schwartz-Mette & Rose, 2009). Individuals may self-focus during interactions with others because they find it difficult to disengage from personal worries and concerns (Pyszczynski & Greenberg, 1987), and they may see problem talk with friends as a way to gain support for their troubles. Yet monopolizing conversations in this way may undermine their attempts to gain support and reduce distress as healthier, more reciprocal interactions are likely impeded. Studies have yet to evaluate whether conversational self-focus longitudinally predicts increased negative affect, but conversational self-focus does predict increased rejection by friends over time (Schwartz-Mette & Rose, 2016). Two studies of conversational self-focus indicate that adolescents’ redirection and monopolization of conversations to focus on their own problems is viewed as aversive by friends (Schwartz-Mette & Rose, 2009; 2016). Conversational self-focus has yet to be formally included in the study of interpersonal emotion regulation. However, given its theoretical links to excessive reassurance seeking as well as the one-sided and problematic nature of this behavior, it is unlikely to serve an adaptive interpersonal emotion regulation purpose and may increase negative affect as a function of interpersonal rejection.

Negative feedback seeking.

Negative feedback seeking is looking for confirmation from others that one’s own negative self-appraisals are indeed correct via solicitation of negative feedback (Joiner & Metalsky, 1995). Individuals who already are experiencing depressive affect are believed to engage in negative feedback seeking to maintain a sense of predictability and control (Swann et al., 1992). Rather than providing comfort that may alleviate negative emotion, efforts to seek negative feedback from others instead serve to maintain depressive affect (Pettit & Joiner, 2001). Negative feedback seeking leads to increased depressive symptoms (Borelli & Prinstein, 2006) and rejection by others (Joiner & Metalsky, 1995) over time. Thus, although individuals may engage in negative feedback seeking in an effort to gain relief through reassurance that their negative beliefs are correct, they instead experience greater distress. Like self-focus, negative feedback seeking has yet to be formally conceptualized as an interpersonal emotion regulation strategy, but given its theoretical and functional ties to excessive reassurance seeking, which has been studied in this context, it appears to be a logical candidate.

Self-disclosure.

Unlike the three previous interpersonal behaviors, self-disclosure is considered more adaptive. Self-disclosure is the normative, reciprocal process of sharing thoughts and feelings with others (Cozby, 1973), and it serves multiple functions within adolescents’ friendships (e.g., social validation, self-clarification, self-expression, relationship development; Buhrmester & Prager, 1995). Although excessive reassurance seeking, conversational self-focus, and negative feedback seeking are linked with increased negative affect and relational consequences, self-disclosure has an inverse association with psychopathology (Adams & Cantin, 2012; Schwartz-Mette & Rose, 2009; 2012). Self-disclosure is thought to serve a protective function in terms of socioemotional adjustment, given its centrality to positive friendship quality (e.g., Buhrmester, 1990; Markovic & Bowker, 2017; Parker & Asher, 1993). Moreover, sharing feelings with others via self-disclosure is conceptualized as a type of interpersonal emotion regulation (Rime, 2007; Zaki & Williams, 2013) and thus may be a more constructive interpersonal emotion regulation strategy.

The Current Study

Researchers have long hypothesized that individuals with intrapersonal emotion regulation difficulties may turn to their interpersonal context to assist with regulating difficult emotions. Although identification of specific interpersonal behaviors is lacking, the literature on the interpersonal context of distress presents likely candidates for examination--namely excessive reassurance seeking, conversational self-focus, and negative feedback seeking. These one-sided, aversive behaviors appear to have poor utility for emotion regulation and carry negative relationship consequences, which leads to increased distress. Self-disclosure may reflect a more adaptive interpersonal emotion regulation strategy that is more palatable for relationships.

Empirical examination of whether intrapersonal emotion regulation difficulties are associated with increased use of these interpersonal strategies has, to date, not yet occurred. Additionally, few studies have examined the longitudinal consequences of employing these interpersonal behaviors on intrapersonal emotion regulation. Moreover, an added benefit of studying interpersonal emotion regulation strategies is that they are observable to others; yet, others’ perspectives on interpersonal regulation behaviors have rarely been obtained. It is especially critical that these questions be tested in adolescence and within friendships, as peers are increasingly important and refining emotion regulation skills is a central developmental task. It was expected that intrapersonal emotion regulation difficulties would predict adolescents’ increased engagement in excessive reassurance seeking, conversational self-focus, and negative feedback seeking. Reciprocal associations also were predicted such that increased engagement in maladaptive interpersonal emotion regulation strategies would predict increased intrapersonal emotion regulation difficulties. We hypothesized that these effects would hold for both adolescents’ and friends’ reports of adolescents’ interpersonal behavior. Regarding self-disclosure, elevated emotion regulation difficulties may also predict engagement in this adaptive interpersonal strategy. On the other hand, given that self-disclosure is normative and generally unrelated to psychopathology symptoms, it may be that self-disclosure is unrelated to emotion regulation difficulties. As such, no firm hypotheses were put forth regarding self-disclosure.

Hypotheses were tested in a community sample of adolescents assessed at three time points spaced approximately 3 months apart. The use of cross-lagged structural equation models enabled testing of specific directions of effect for each interpersonal behavior. Self-reports of adolescents’ emotion regulation difficulties and interpersonal behaviors and friend-reports of excessive reassurance seeking and conversational self-focus were available to test for variability in associations based on reporter. Specifically, models tested whether intrapersonal emotion regulation difficulties longitudinally predicted self- and friend-reported interpersonal behavior and/or whether each interpersonal behavior predicted later intrapersonal emotional regulation difficulties, over and above the between-subjects stability of these constructs. Models tested relations with each behavior and reporter separately, given the potential for differences across behaviors and reporters. Gender and age group differences also were considered in each model.

Method

Participants and Procedures

Study procedures were approved by the University of Maine Institutional Review Board (IRB). Data for the current study were collected as part of a larger project examining social relationships and health in adolescence. Adolescents (n = 93) were recruited through local print and social media advertisements as well as in-person at community and/or school events. They then identified a close, same-gender friend (n = 93) with whom they would like to participate. From here on, “participants” refers to both adolescents and the friends with whom they participated. Prior to data collection, written parent/guardian consent was obtained for youth under age 18, and youth were scheduled to attend a lab session.

Participants were 186 middle and high school adolescents (M = 15.68 years; SD = 1.49; range: 12-19) recruited from rural communities surrounding a mid-sized public university in New England. 69.9% identified as female, and 30.1% identified as male. The self-reported racial and ethnic identities of the sample were: 87.6% Caucasian, 4.3% Black/African American, 3.2% Asian/Pacific Islander, 3.2% Hispanic/Latino(a), and 1.6% American Indian/Alaskan Native. Although these self-reported identities of participants reflect a relatively non-diverse sample, the sample is representative of the larger community from which it was drawn.

Participants 18 years or older provided written consent upon arrival to the lab, and minor participants provided assent to participate. Youth then completed a series of self-report questionnaires administered using Qualtrics© survey software and laboratory tasks not relevant to the current study. Approximately three and six months after the initial lab visit, participants were contacted to complete follow-up self-report questionnaires online or, if participants did not have a device and/or reliable internet service, they were sent paper questionnaires with a self-addressed, stamped envelope to return. All participants received $40.00 immediately following the lab visit and $10.00 for each completed follow-up questionnaire.

Measures

Demographics.

Participants reported age, gender identity, and racial and ethnic identities.

Intrapersonal Emotion Regulation Difficulties.

The 36-item Difficulties in Emotion Regulation Scale (DERS; Gratz & Roemer, 2004) assessed difficulties in intrapersonal emotion regulation. Items inquire about the frequency with which individuals respond to their emotions in adaptive or maladaptive ways. Responses are rated on a 5-point Likert scale ranging from 1 (Almost never/0-10% of the time) to 5 (Almost always/91-100% of the time). Items assessing adaptive responses to emotion were reverse-coded. Summed responses from all DERS items yield a total emotion regulation difficulties score with higher scores reflecting greater difficulties (Time 1 ɑ = .96, Time 2 ɑ = .96, Time 3 ɑ = .97).

Excessive Reassurance Seeking.

Participants responded to four items adapted from the Depressive Interpersonal Relationships Inventory (DIRI; Joiner & Metalsky, 2001) to assess excessive reassurance seeking at Time 2. Participants rated the frequency with which they ask family and friends if they are truly liked or cared for on a 3-point Likert scale ranging from 0 (Not very much) to 2 (Very Much). Scores were the mean of responses to the four items at each time point (Time 1 ɑ = .83, Time 2 ɑ = .73, Time 3 ɑ = .92). Participants then reported on their friends’ engagement in excessive reassurance seeking using the same four items edited to reflect perspectives on friends’ behavior (Time 1 ɑ = .80, Time 2 ɑ = .80, Time 3 ɑ = .83). Thus, all participants received a self-reported and friend-reported excessive reassurance seeking score.

Conversational Self-Focus.

To assess conversational self-focus, participants responded to six items assessing the degree to which they are prone to redirect and dominate problem-focused conversations with others (Schwartz-Mette & Rose, 2006). Responses were rated on a 5-point Likert scale ranging from 1 (Not at all true) to 5 (Very True), and participants’ scores were the mean of all ratings at each time point (Time 1 ɑ = .80, Time 2 ɑ = .77, Time 3 ɑ = .84). Participants then reported on their friends’ tendency to engage in conversational self-focus using the same six items edited to reflect perspectives on friends’ behavior (Time 1 ɑ = .88, Time 2 ɑ = .93, Time 3 ɑ = .96). As with excessive reassurance seeking, all participants received a self-reported and friend-reported conversational self-focus score.

Negative Feedback Seeking.

Negative feedback seeking was assessed using the Feedback Seeking Questionnaire (Swann et al., 1992), modified for use among adolescents (Joiner et al., 1997). Items assessed participants’ tendency to seek positive and negative feedback across four domains: social competency, academic competency, athletic skills, and physical attractiveness. For each domain, six questions were listed, and participants were instructed to choose up to three questions they would like to have a friend answer about them. Three questions within each domain were worded positively to assess inclinations to seek positive feedback (e.g., What is some evidence you have seen that your friend has good social skills?), and three questions were worded negatively to assess inclinations to seek negative feedback (e.g., What is some evidence you have seen that your friend does not have good social skills?). Each participant was assigned a score that was the total number of negatively-worded questions chosen.

Self-Disclosure.

Adolescents completed a five-item measure assessing normative self-disclosure with peers (Rose, 2002). Each item is rated on a 5-point Likert scale ranging from 1 (Not at all true) to 5 (Very True) with higher scores reflecting greater reciprocal sharing of thoughts and feelings with friends. Scores were the mean of all items at each time point (Time 1 ɑ = .90, Time 2 ɑ = .91, Time 3 ɑ = .92).

Missing Data

Some participants had missing data. Specifically, of the 186 youth who participated in the Time 1 assessment, seven youth did not participate at Time 2, 37 did not participate at Time 3, and 39 youth did not participate at either Time 2 or Time 3. Representative analyses indicated that these groups of participants did not differ in emotion regulation difficulties [Time 1: F(1, 184) = 3.67, p = .06; Time 2 F(1, 129) = 0.97, p = .33; Time 3 F(1, 101) = 0.74, p = .39], excessive reassurance seeking [Time 1 F(1, 185) = 1.00, p = .32; Time 2: F(1, 125) = 1.79, p = .18; Time 3: F(1, 102) = 0.43, p = .51], conversational self-focus [Time 1 F(1, 183) = 0.83, p = .77; Time 2 [F(1, 131) = 0.39, p = .54; Time 3: F(1, 102) = 1.08, p = .30], NFS [Time 1 F(1, 185) = 1.50, p = .22; Time 2: F(1, 137) = 0.69, p = .41; Time 3: F(1, 102) = 0.38, p = .54], or self-disclosure [Time 1 F(1, 184) = 1.24, p = .27; Time 2 [F(1, 131) = 1.26, p = .26; Time 3: F(1, 102) = .06, p = .81]. Little’s test further indicated that data were missing completely at random (MCAR), χ²(256) = 258.9122, p = .45. Thus, missing data were imputed using an expectation maximization procedure in SPSS, and the full sample was used for all analyses.

Results

Descriptive statistics and correlations are presented in Table 1. Generally, intrapersonal emotion regulation difficulties were positively correlated with self-reported excessive reassurance seeking, conversational self-focus, and negative feedback seeking, and self-reports of each of these three behaviors were generally positively correlated with one another. In contrast, correlations of self-disclosure with other variables were negative (e.g., associations with intrapersonal emotion regulation difficulties and negative feedback seeking) or not significant (e.g., with excessive reassurance seeking and conversational self-focus).

Table 1.

Descriptive Statistics and Correlations

	M (SD)	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.	13.	14.	15.
1. DERS 1	2.30 (.73)	.77***	.72***	.55***	.43***	.37***	.26***	.15*	.25***	.32***	.30***	.19**	−.02	−.15*	.11
2. DERS 2	2.20 (.69)	-	.88***	.39***	.44***	.40***	.21**	.12†	.28***	.36***	.44***	.34***	−.07	−.18*	−.21**
3. DERS 3	2.15 (.65)		-	.37***	.46***	.44***	.29***	.23***	.37	.40***	.56***	.50***	−.09	−.20**	−.24**
4. ERS 1	0.19 (.38)			-	.77***	.79***	.17*	.14*	.22**	.25***	.10	.04	−.01	.04	.09
5. ERS 2	0.17 (.28)				-	.79***	.19**	.20**	.46***	.15*	.19*	.14*	−.11	−.03	−.02
6. ERS 3	0.18 (.38)					-	.35***	.29***	.47***	.16*	.15*	.12†	.07	.10	.15*
7. CSF 1	1.33 (.45)						-	.61***	.64***	.18**	.14*	.14†	.08	.00	.03
8. CSF 2	1.33 (.37)							-	.72***	.16*	.12†	.26***	.00	−.13†	−.05
9. CSF 3	1.31 (.36)								-	.18**	.18**	.31***	.03	−.08	−.02
10. NFS 1	3.66 (2.24)									-	.47***	.59***	.06	−.03	.01
11. NFS 2	4.37 (2.13)										-	.59***	.01	−.13†	−.19**
12. NFS 3	4.33 (2.18)											-	−.12	−.27	−.28
13. SD 1	3.79 (1.00)												-	.76***	.74***
14. SD 2	3.75 (.98)													-	.84***
15. SD 3	3.69 (.94)														-

Notes. N = 186.

^†p = .09

^*p < .05.

^**p < .01.

^***p < .001. DERS = Difficulties in Emotion Regulation Scale scores. ERS = excessive reassurance seeking scale scores. CSF = conversational self-focus scale scores. NFS = negative feedback seeking scale scores. SD = self-disclosure scale scores.

Correlations between self- and friend- reports of excessive reassurance seeking (r = .36, .25, .60) and of conversational self-focus (r = .49, .27, .27) were significant at each time point (all p values < .01) and small-to-medium in size. Adolescents’ emotion regulation difficulties were associated with friend-reports of excessive reassurance seeking at Time 1 (r = .20, p = .005) and Time 3 (r = .22, p = .002) and with friend-reports of conversational self-focus at Time 1 only (r = .25, p =.001).

Primary hypotheses were tested using structural equation modeling in Amos 27.0. Separate cross-lagged panel models were fit to test the longitudinal associations of intrapersonal emotion regulation difficulties with each of the self-reported and friend-reported interpersonal behaviors of interest. In each model, the stability of intrapersonal emotion regulation difficulties over six months was modeled by including a path from Time 1 intrapersonal emotion regulation difficulties to Time 2 intrapersonal emotion regulation difficulties and a path from Time 2 difficulties to Time 3 difficulties. The stability of interpersonal behavior was modeled similarly. Cross paths were then added from the Time 1 variables to the Time 2 variables and from the Time 2 variables to the Time 3 variables. Covariances between intrapersonal emotion regulation difficulties and the interpersonal behaviors were added within time point. Fit of each cross-lagged panel model was assessed using standard fit indices including chi-square, comparative fit index (CFI), Tucker-Lewis fit index (TLI), and the root mean-square error of approximation (RMSEA). Smaller values of chi-square and RMSEA and larger values of CFI and TLI indicate better model fit (Cangur & Ercan, 2015).

Excessive Reassurance Seeking

Fit of the model for excessive reassurance seeking was adequate, χ² (1) = 2.61, p = .24; CFI = 0.99, TLI = .97, RMSEA = .09 (see Figure 1, Panel A). Intrapersonal emotion regulation difficulties and excessive reassurance seeking evidenced strong, positive associations from Time 1 to Time 2 and from Time 2 to Time 3 and were significantly associated within time point at Time 1. Greater Time 1 intrapersonal emotion regulation difficulties did not predict higher excessive reassurance seeking at Time 2, and higher Time 1 excessive reassurance seeking did not predict greater intrapersonal emotion regulation difficulties at Time 2. However, from 3 to 6 months, greater intrapersonal emotion regulation difficulties predicted higher excessive reassurance seeking, and higher excessive reassurance seeking predicted greater intrapersonal emotion regulation difficulties. Indirect effects were next tested. The indirect effect (IE) of Time 1 excessive reassurance seeking on Time 3 excessive reassurance seeking via Time 2 intrapersonal emotion regulation difficulties was significant, IE = .28 (95% CI: .11, .55), p < .05. The indirect effect of Time 1 intrapersonal emotion regulation difficulties on Time 3 difficulties via Time 2 excessive reassurance seeking also was significant, IE = .63 (.49, .77), p < .05.

Figure 1.

Panel A. Associations of Emotion Regulation Difficulties and Self-Reported Excessive Reassurance Seeking Over 3 and 6 Months

Panel B. Associations of Emotion Regulation Difficulties and Friend-Reported Excessive Reassurance Seeking Over 3 and 6 Months

Notes. N = 186. **p < .01. ***p < .001.

A model was next fit using friends’ reports of adolescents’ excessive reassurance seeking (see Figure 1, Panel B). This model was identical to the original model, except that friends’ reports were used in place of adolescents’ reports. This model fit less well than the self-report model, and modification indices suggested no alterations to improve fit, χ² (1) = 4.61, p = .03; CFI = 0.99, TLI = .92, RMSEA = .14. Greater emotion regulation difficulties were associated with higher levels of friend-reported excessive reassurance seeking at Time 1. Friends’ reports of adolescents’ excessive reassurance seeking showed moderate positive associations from Time 1 to Time 2 and from Time 2 to Time 3, but none of the cross paths were significant.

Conversational Self-Focus

Fit of the model including conversational self-focus was good, χ² (5) = 7.84, p = .17; CFI = 1.00, TLI = 0.99, RMSEA = .06 (see Figure 2, Panel A). Associations of conversational self-focus from Time 1 to Time 3 and from Time 2 to Time 3 were positive and strong. No significant cross paths from any Time 1 variable to any Time 2 variable were significant. However, greater Time 2 intrapersonal emotion regulation difficulties predicted higher conversational self-focus at Time 3, and higher Time 2 conversational self-focus predicted greater intrapersonal emotion regulation difficulties at Time 3. Regarding indirect effects, the indirect effect of Time 1 conversational self-focus on Time 3 conversational self-focus via Time 2 intrapersonal emotion regulation difficulties was significant, IE = .32 (.19, .48), p < .05. The indirect effect of Time 1 intrapersonal emotion regulation difficulties to Time 3 intrapersonal emotion regulation difficulties via Time 2 conversational self-focus also was significant, IE = .66 (.58, .74), p < .05.

Figure 2.

Panel A. Intrapersonal Emotion Regulation Difficulties and Self-Reported Conversational Self-Focus Over 3 and 6 Months

Panel B. Intrapersonal Emotion Regulation Difficulties and Friend-Reported Conversational Self-Focus Over 3 and 6 Months

Notes. N = 186. ***p < .001.

A parallel model was fit using friends’ reports of adolescents’ conversational self-focus (see Figure 2, Panel B). The initial fit of this model was poor, and modification indices suggested adding a covariance between friends’ reports of conversational self-focus at Time 1 and emotion regulation difficulties at Time 3. Fit of this model was good, χ² (4) = 7.10, p = .13; CFI = 0.99, TLI = 0.98, RMSEA = .07. Friends’ reports of adolescents’ conversational self-focus were moderately and positively associated across 3 and 6 months, but no cross-paths were significant.

Negative Feedback Seeking

Fit of the negative feedback seeking model was good, χ² (1) = 0.29, p = .59; CFI = 1.00, TLI = 1.01, RMSEA = .00 (see Figure 3). Positive associations of negative feedback seeking from Time 1 to Time 2 and from Time 2 to Time 3 were found. Greater Time 1 intrapersonal emotion regulation difficulties predicted higher negative feedback seeking at Time 2, and higher Time 2 negative feedback seeking predicted greater intrapersonal emotion regulation difficulties at Time 2. Additionally, the cross paths from Time 2 to Time 3 were also significant, such that greater Time 2 intrapersonal emotion regulation difficulties predicted higher negative feedback seeking at Time 3, and higher Time 2 negative feedback seeking predicted greater intrapersonal emotion regulation difficulties at Time 3. The indirect effect of Time 1 negative feedback seeking on Time 3 intrapersonal emotion regulation difficulties [IE = .18 (.07, .30), p < .05] was comparable in strength to the indirect effect of Time 1 intrapersonal emotion regulation difficulties on Time 3 negative feedback seeking [IE = .16 (.06, .26), p < .05].

Figure 3.

Associations of Intrapersonal Emotion Regulation Difficulties and Self-Reported Negative Feedback Seeking Over 3 and 6 Months

Notes. N = 186. *p < .05. **p < .01. ***p < .001.

Self-Disclosure

Finally, a model with adequate fit was tested for self-disclosure, χ² (5) = 10.71, p = .06; CFI = 0.99, TLI = 0.98, RMSEA = .08 (see Supplementary Materials). Strong positive associations were observed for self-disclosure from Time 1 to Time 2 and from Time 2 to Time 3. Self-disclosure did not significantly predict intrapersonal emotion regulation difficulties over 3 or 6 months. However, greater Time 1 intrapersonal emotion regulation difficulties predicted lower self-disclosure at Time 2, and greater Time 2 intrapersonal emotion regulation difficulties predicted lower self-disclosure at Time 3. The indirect effect of Time 1 intrapersonal emotion regulation difficulties on Time 3 self-disclosure was significant [IE = −.15 (−.24, −.08), p < .05].

Age and Gender Differences

Multiple group comparisons were tested for each model to examine potential age or gender differences. Given the range and distribution of the age variable, a group of younger (ages 12-15, n = 80) and a group of older (ages 16-18, n = 106) adolescents were identified. An unconstrained model in which all parameters were allowed to vary across age groups or gender was compared to a series of increasingly constrained models to test for group differences. In all cases, models in which the structural paths (stability and cross paths) were constrained to be equal across groups did not significantly differ from the unconstrained model for any of the self- or friend-report models. These results suggest that the structural paths of interest performed similarly for younger and older adolescents and for females and males in each model.

Discussion

The current study tested whether adolescents’ self-reported difficulties in intrapersonal emotion regulation were associated with self- and friend-reported engagement in interpersonal efforts to regulate emotion. Despite research finding that emotion regulation is commonly interpersonal (Zaki & Williams, 2013) and that the peer context is centrally important for adolescents (Hartup, 1993), no research to date had evaluated links between intrapersonal and interpersonal emotion regulation in this age group. Specifically, the current study contributed new data evidencing concurrent and longitudinal associations between adolescents’ intrapersonal emotion regulation difficulties and engagement in excessive reassurance seeking, conversational self-focus, negative feedback seeking, and self-disclosure.

First, results suggested that higher initial levels of intrapersonal emotion regulation difficulties, excessive reassurance seeking, conversational self-focus, negative feedback seeking, and normative self-disclosure predicted higher levels of these variables at 3 and 6 months. The possibility that adolescents who engage in high levels of maladaptive interpersonal behaviors continue to do so over time is underscored by theory (Coyne, 1976) suggesting that others reinforce such behaviors in the short-term, even if the behaviors are ineffective at reducing longer-term distress. Specifically, these maladaptive behaviors may initially be reinforced by peers through increased feelings of closeness or receipt of social support. Adolescents may even temporarily experience decreases in distress immediately following such interactions. Over time, adolescents may thus persist in these behaviors, even if they are ineffective in regulating one’s own emotions over the longer term, if they do not result in increased intrapersonal emotion regulation capacity, and/or if they may be perceived by friends as aversive.

Importantly, this study also provided initial evidence that difficulties in intrapersonal emotion regulation are associated with engagement in certain interpersonal behaviors aimed at regulating emotions. Greater difficulties with intrapersonal emotion regulation were associated with higher concurrent engagement in self- and friend-reported excessive reassurance seeking, self- and friend-reported conversational self-focus, and self-reported negative feedback seeking. Conversely, greater intrapersonal emotion regulation difficulties were associated with less self-reported self-disclosure, which is more normative and adaptive (Buhrmester & Prager, 1995).

It may be that adolescents with intrapersonal emotion regulation difficulties do not use just any available strategy to engage with peers, but rather they engage selectively in problematic interpersonal behaviors in an attempt to regulate emotions that they have difficulty regulating themselves. Internal difficulties regulating one’s own emotions could lead adolescents to experience greater distress, which in turn could make them less effective in their interpersonal efforts to gain support and regulation. Another possibility is that adolescents who have not yet acquired adaptive tools to regulate emotions also lack interpersonal skills necessary to effectively engage peers in assisting them with regulating emotion, rendering them ineffective at both.

Prior to the current study, longitudinal relations of initial intrapersonal emotion regulation difficulties with later interpersonal attempts to regulate emotion, and vice versa, had not been evaluated. The current study provided initial evidence that greater difficulties in intrapersonal emotion regulation and greater self-reported engagement in maladaptive interpersonal emotion regulation behaviors have transactional, downstream effects. Greater intrapersonal emotion regulation difficulties predicted higher self-reported, but not friend-reported, excessive reassurance seeking, conversational self-focus, and negative feedback seeking over time. Higher self-reported, but not friend-reported, engagement in these behaviors also longitudinally predicted greater intrapersonal emotion regulation difficulties. In contrast, greater intrapersonal emotion regulation difficulties predicted lower self-reported self-disclosure over time. These effects were significant after controlling for the longitudinal stability of the variables.

With regard to the first direction of effect, adolescents with greater underlying difficulties regulating emotion appear to engage more in socially maladaptive attempts to regulate emotion. Given that these interpersonal behaviors often are perceived as aversive to peers and do not appear to actually aid in intrapersonal emotion regulation, they may result in even greater negative affect, potential peer rejection, and lower confidence in one’s abilities to independently manage distress. Additionally, overreliance on friends for regulating strong emotions may preclude these adolescents from engaging in more adaptive social behavior (e.g., normative self-disclosure) and from learning effective intrapersonal emotion regulation skills.

It is perhaps not surprising that data also supported the other direction of effect. Greater self-reported excessive reassurance seeking, conversational self-focus, and negative feedback seeking were associated with greater intrapersonal emotion regulation difficulties over time. Adolescents’ reduced tendencies to utilize normative self-disclosure under conditions of higher levels of intrapersonal regulation problems may reflect yet another vulnerability, as they may be less willing or able to employ adaptive support-seeking behaviors within their friendships.

It is also noteworthy that findings from the current study did not support longitudinal associations between adolescents’ emotion regulation difficulties and friends’ reports of excessive reassurance seeking or conversational self-focus. There could be important variations in adolescents’ interpersonal behavior not captured by friends’ reports, as the friends participating in this study may not be privy to the range of interpersonal behavior exhibited by adolescents. Future research incorporating more objective measurements of adolescents’ behavior (e.g., observations) alongside self- and friend-reports could clarify what methods of assessment are most accurate. Studies also could extend our understanding of friends’ perceptions of adolescents' negative feedback seeking and self-disclosure behaviors, as friend-reports of these behaviors were unavailable in the current study. Finally, friends’ reports of adolescents’ maladaptive interpersonal behaviors may be particularly important to obtain in the context of clarifying the impact of these behaviors on friendship outcomes.

Future research also is needed to provide more textured information about transactional associations of adolescents’ intrapersonal emotion regulation difficulties and each interpersonal behavior. Use of sophisticated cross-lagged panel models (Hamaker et al., 2015) would extend this work to isolate within-person processes and between-person differences. Future studies could also determine whether and how these behaviors may differ in form, function, and impact on downstream socioemotional outcomes. For instance, excessive reassurance seeking and conversational self-focus may not be immediately aversive to friends and could, according to theory (Coyne, 1976), elicit initial peer support and even serve a mild regulatory function. Friends may first respond by offering the reassurance sought or by allowing the adolescent to dominate conversations about their problems. At first, the adolescent may feel validated, supported, and relieved. Over time, however, as these patterns of behavior become more entrenched, friends may respond with less and less assistance and may even pull away from the friendship (Prinstein et al., 2005; Schwartz-Mette & Rose, 2014), leading to more distress and fewer options for regulation, especially if friends withdraw from the relationship.

Negative feedback seeking, on the other hand, may be more likely to violate friends’ expectations for normative disclosures as it is both a one-sided (i.e., non-reciprocal) process and more negatively focused. As such, negative feedback seeking could be more immediately aversive to friends who are asked to either confirm negative statements about the adolescent, which likely feels unsupportive and in contrast to typical provisions of social support, or to disagree with the adolescent, which may lead to friendship conflict. Either response is likely to be interpreted as invalidating by the distressed adolescent, further increasing the adolescent’s negative affect and emotion dysregulation. Distressed adolescents also may experience less positive reinforcement from friends when engaging in negative feedback-seeking, as compared to when they engage in excessive reassurance seeking and/or conversational self-focus.

Results from the current study also did not suggest that the processes under investigation differed significantly for younger or older adolescents, or for those adolescents identifying as female or male. Given mean-level age and gender differences in emotion regulation that may have origins in or around adolescence (Zimmerman & Iwanski, 2014); emerging evidence for gender differences in excessive reassurance seeking (e.g., Prinstein et al., 2005; Starr & Davila, 2008), negative feedback seeking (Nesi & Prinstein, 2015), and conversational self-focus (Schwartz-Mette & Rose, 2016); and strong evidence for age and gender differences in self-disclosure (see Rose & Rudolph, 2006), future research should continue to test age and gender differences in and among these variables in order to enhance confidence in these results.

Limitations and Additional Directions for Future Research

The current study provides new evidence that intrapersonal emotion regulation difficulties are associated with maladaptive behaviors both concurrently and over time, but it is not without limitations. One limitation is that effects of these processes on psychopathology were not examined in the current study. An important next step is to examine implications of these strategies for mental health outcomes. Previous research has separately evidenced associations between intrapersonal emotion regulation deficits and internalizing symptoms (e.g., Berking et al., 2014; Garnefski et al., 2005) and between internalizing symptoms and engagement in the studied intrapersonal behaviors (e.g., Borelli & Prinstein, 2006; Joiner et al., 1999; Schwartz-Mette & Rose, 2016). It may be that individuals who are experiencing depressive and anxiety symptoms have more challenging emotions to regulate and fewer intrapersonal skills to do so, leading them to rely on these maladaptive interpersonal interactions for regulatory functions. Moreover, depressive symptoms are associated with lower friendship quality and fewer friendships (e.g., Biggs et al., 2011; Glass et al., 2012; Prinstein et al., 2005; Scharfstein et al., 2111; for review see Schwartz-Mette et al., 2020). Thus, when quality social partners are lacking and social interactions already tend to be poor, it may follow that individuals engage in these more maladaptive interpersonal behaviors when dysregulated. It will therefore be important to determine if and how internalizing symptoms impact and are impacted by transactional relations between intrapersonal and interpersonal emotion regulation.

Similarly, peer rejection was not addressed in the present study. Excessive reassurance seeking, conversational self-focus, and negative feedback seeking are hypothesized to be associated with psychopathology in part because they lead to reduced friendship quality and, at times, dissolution of peer relationships. Assessment of impacts on friendship, as perceived by both members of the friendship dyad, would be beneficial in clarifying how characteristics of the relationship impact, and are impacted by, intrapersonal and interpersonal deficits in emotion regulation. For example, it may be that individuals with difficulties intrapersonally regulating emotions engage in maladaptive efforts to regulate emotions interpersonally (as shown in the present study), which then leads to their friends to distance themselves or to dissolve the friendship. This peer rejection could then lead to greater distress and intrapersonal difficulties for the adolescent attempting to regulate these now stronger negative emotions, without the interpersonal opportunities to gain much-needed support.

Future research should also explore the overlap and interplay between strategies that are enacted completely internally and interpersonal regulatory behaviors, as even if others are recruited to assist in regulating emotions, the regulation itself may still occur as a largely internal process. What is not yet known is how engagement with others for the purposes of regulating distress impacts adolescents’ internal regulation processes in real time. The presence of and responses from a trusted and supportive friend may assist distressed adolescents in the moment with feeling calm, generating new problem-solving ideas, and enhancing self-efficacy, all of which may boost intrapersonal emotion regulation capacities. Moreover, there may be yet undiscovered individual differences in adolescents' responses to interpersonal assistance.

An additional limitation is that friend reports were available for only two of the four behaviors studied. Social interaction inherently involves multiple people, and inclusion of both dyad members’ reports of their engagement in each of these behaviors could provide a more comprehensive picture of dyadic interactions. Additional research evaluating dyadic (as opposed to one-sided) interpersonal efforts to regulate emotion (e.g., co-rumination, or the excessive dwelling on problems and negative affect with peers; Rose, 2002) also would add to this work. Larger and more diverse samples will certainly be required to test these more complex models and to evaluate whether results can be replicated and generalized to more diverse populations.

Despite its limitations and the need for further empirical research, the current study contributes new evidence that intrapersonal emotion regulation deficits are associated with engagement in maladaptive interpersonal efforts to regulate emotions in adolescence and that these behaviors further exacerbate intrapersonal emotion regulation difficulties. To be maximally beneficial, intervention efforts aimed at improving regulatory capabilities should therefore consider incorporating both intrapersonal and interpersonal strategies. By doing so, adolescents may become better equipped to regulate emotions while also preserving strong peer support that is so important during this developmental stage.