Does My Friend Like My Romantic Partner? Links Among Relationship Work With Partners and Friends and Physiological Implications

ABSTRACT

Informed by the theoretical foundation of social exchange theory and polyvagal theory, we examined romantic, social, and psychophysiological functioning among a sample of 284 heterosexual, emerging adult romantic partners and their friends. Actor–partner interdependence model results revealed that both women and men reported greater romantic adjustment when also speaking frequently with their partner about romantic challenges. Additionally, men were found to have more favorable psychophysiological functioning, as captured by heart rate variability (HRV) during conversations with their partner, when men also spoke frequently to friends about romantic challenges. Moreover, men's HRV scores when talking to partners and to friends were more favorable if men perceived that friends' approval of their relationship was high. Clinical implications are presented, and social, romantic, and psychophysiological implications for emerging adult partners are discussed.

Romantic relationships evolve across life stages, and the nuances of communication between partners vary as relationships develop and progress (Arnett 2015; Courtain and Glowacz 2019). Emerging adulthood (ages 18–25) ushers in transitions not only for romantic relationships but also for how partners interact with their social networks (Erikson 1968). This stage is a time of tumultuous change, filled with choices that impact the remainder of the lifespan (Arnett 2015). Many emerging adults in the United States are simultaneously leaving home, exploring career and educational opportunities, and tentatively embracing newfound independence that accompanies unprecedented autonomy (Courtain and Glowacz 2019). Along with these personal developmental milestones, romantic relationships become increasingly important in young adulthood. Because these relationships do not exist in a vacuum, researchers have examined the personal and social contextual factors that are navigated while dating a romantic partner (Huston 2000). Notably, perceptions of approval from a partner's friends and family increase relationship stability, while social network disapproval can prove detrimental (Felmlee 2001).

Friends often become intertwined in romantic relationships through a phenomenon known as relationship work (RW; J. F. Jensen and Rauer 2016). RW occurs when partners discuss romantic challenges with one another, and then turn outside the relationship to discuss or vent about romantic issues to their friends (J. Jensen et al. 2019). Previous research suggests that it is most beneficial to the romantic union when partners discuss their relationship problems with one another before involving their friends (J. F. Jensen and Rauer 2016). The tone and frequency of such conversations with friends, as well as numerous other relationship components, are inevitably linked with measures of romantic functioning, such as relationship satisfaction (Proulx et al. 2004). Capturing antecedents and correlates of relationship adjustment is critical, given that successful romantic relationships have been linked to lower rates of emotional and physical distress, along with increased positive physiological functioning (Schneiderman et al. 2012).

1. Literature Review

1.1. Theoretical Foundation

To comprehensively assess the intersection of personal, social, and romantic functioning in emerging adulthood, two complementary theories guide the current study—Porges' (2009, 2015) polyvagal theory and social exchange theory (Cropanzano and Mitchell 2005). Both are rooted in the effects of relationships on individual functioning. These two theories serve as a foundation for examining how individuals, their partners, and their friends share information about romantic relationships and the resulting physiological changes that occur during such interactions. Porges' (2015) polyvagal theory captures how external stimuli affect internal physiological states. The body seeks to maintain homeostasis and perceive and respond to threats in the environment (Porges 2009). When the stress response is initiated, it works as a counterbalance by directing the reactions of various body systems. When security is established, humans' nervous systems facilitate growth and restoration, and they may emotionally connect with one another without the fear of physical or emotional threats (Porges 2022). Positive social processes, such as turning one's head toward another, or making eye contact, are referred to as “social engagement” behaviors or responses (Shaffer and Venner 2013). These responses increase connection and are activated when a person feels secure with those around them. The body initiates a physiological response, including a slowed heart and respiration rate and increased feelings of calmness and security (Shaffer and Venner 2013). Accordingly, via physiological functioning, individuals may coregulate and decrease defensiveness through the formation of safe and connected relationships resulting from supportive conversations (Porges 2022). Thus, capturing physiological responses as partners discuss romantic challenges may not only reveal internal biological patterns but may also elucidate interpersonal connections.

Although the examination of physiological functioning driven by polyvagal theory may indirectly inform social and romantic dynamics, social exchange theory more directly explains the driving factors behind the behaviors that formulate and sustain social relationships (Ahmad et al. 2023). Social exchange theory supplements the consideration of individual biological factors that inform social relationships by direct consideration of the assessment of the value of social interactions. Social exchange theory posits that behaviors are a good to be exchanged with rules and norms established between people that create guidelines by which actions are judged (Cropanzano and Mitchell 2005). The expectation of reciprocity (Ahmad et al. 2023) is a fundamental principle that drives social exchange (Cropanzano and Mitchell 2005). Partners generally seek to maintain approximate symmetry of perceived effort exerted to maintain a relationship. Consequently, conflict often ensues when the balance is unequal (Clark 2023). Remaining in a relationship is the result of the perceived net profit of this equation and when alternatives are deemed more rewarding, dependence on the relationship decreases (Nakonezny and Denton 2008). Critically important to the current study, romantic conflict associated with an imbalance in the relationship often results in triangulation of friends or others to seek validation or support where it is lacking (J. F. Jensen and Rauer 2016). Over time, this may result in the problematic process of turning more frequently to the social network, as opposed to one's partner, for relationship support (Helms et al. 2003; J. F. Jensen and Rauer 2016). Considering the discussion of romantic problems with friends as well as friends' approval, taps into the weighing of costs and benefits of said romantic relationship. Thus, the combination of Porges' polyvagal theory and social exchange theory lays the groundwork for examining emerging adult romantic relationships, accounting for the social network's impact on individual and relational well‐being.

1.2. Emerging Adult Romantic and Social Communication—Relationship Work

Understanding the strategies that emerging adults use to communicate well is crucial, as healthy relationships have positive impacts on social and emotional development (Crouter et al. 2005). Highlighting social exchange theory, emerging adults become increasingly adept at engaging in relationships that they perceive as beneficial and withdrawing from those they deem unworthy of their energy (Ahmad et al. 2023; J. F. Jensen and Rauer 2016). Relational problem‐solving skills during emerging adulthood are influenced by romantic and platonic relationships, with gender socialization playing a crucial role. Women are often taught to share feelings and rely on others, whereas men are encouraged to project self‐sufficiency (Brown 2012).

As emerging adults form romantic relationships, their social networks mesh, overlap, and critically impact relationship development (Milardo 1982). When individuals are unable to successfully incorporate their partners into their own social networks, increased relationship dissatisfaction and instability are likely (Jenson et al. 2019). Moreover, if a member of one's social network criticizes one's partner, the friendship may suffer, as this has been linked to a decline in emotional closeness with that person (Gillian et al. 2022). Conversely, if friends approve of one's partner, this expands the positive opinion of the relationship (Felmlee 2001). Social network approval has been linked with increased relationship satisfaction in both same‐sex and different‐sex relationships (Holmberg and Blair 2016). Even the perception of social network approval can influence relationship choices. For example, emerging adults are more likely to be in interracial relationships if they believe their family members will be supportive (Miller et al. 2022). On the other hand, disapproval from one's family and friends has been linked to an increased risk of infidelity and relationship dissolution (Zak et al. 2002). Yet, because partners may not be accurate in the assessments of friends' approval, often assuming friends are more supportive than in reality (Etcheverry et al. 2008), more attention should be given to a comparison of actual versus perceived approval.

The influence of the social network on a romantic relationship can be seen through a type of divulgence known as RW. RW occurs when an individual discusses relationship problems with a partner or someone else (J. Jensen, Rauer, et al. 2018). RW conversations are not inherently positive or negative but rather reflect an attempt to engage in a discussion about a challenge in the relationship. Researchers initially dubbed this process of disclosure to others about marital issues as “marriage work” (Helms et al. 2003, p. 963), stating they examined “husbands' and wives routine disclosures with their closest friends about their marriage—a process we label as ‘marriage work’ based on earlier research by Oliker (1989).” Oliker investigated middle‐class wives' discussions of marital problems with their closest female friends and the impact of such talks on wives' marital attitudes. She found that friends not only validated wives' marital feelings, but also often shared supportive views of the marriage.

Helms et al.'s (2003) results showed that husbands engaged in greater amounts of marriage work with their wives than with close friends, while wives participated in similar amounts of marriage work with their husbands and friends. When sharing with friends, women may have more lengthy and detailed discussions, while men's conversations tend to be briefer and more solution‐oriented (J. Jensen et al. 2019). Connections with marital quality were also examined. If wives took part in low levels of marriage work with husbands and higher levels with their friends, the less marital love and more ineffective arguing they reported. If women engaged in high levels of marriage work with their husbands, the amount of marriage work with friends was unrelated to their perception of the quality of their marriage. Though the direction of effects is uncertain, this suggests that discussing relationship issues with friends to the exclusion of such talks with partners may be detrimental.

J. F. Jensen and Rauer (2016) revamped the concept of marriage work and applied it to emerging adults in dating relationships. Their research specifically considered emerging adult women and their engagement in RW with partners and friends across two time periods. Results showed that women engaged in more RW with partners than with friends and that RW with partners increased over the duration of the relationship. Increased amounts of RW with partners predicted greater romantic stability, and increased RW with friends was connected to a greater likelihood of breaking up. J. F. Jensen and Rauer posited that if individuals are less inclined to turn toward their partner to address romantic conflict, issues may fester, leading to a breakup. Similar to conclusions drawn by Helms et al. (2003), J. F. Jensen and Rauer surmised that engaging in high RW with friends instead of turning to a partner, was likely harmful to the relationship.

In a subsequent study examining RW patterns among emerging adults, researchers found partners most frequently turned to one another to discuss relationship issues, but friends and others, such as parents, were also confided in during times of romantic distress (J. Jensen, Rauer, et al. 2018). The findings also highlighted mothers' residual role as an important fixture in many children's social networks, even into emerging adulthood. Friends and mothers were found to provide nearly equally high levels of emotional support. Clinical implications may suggest that while practitioners often include family members in a systemic therapy approach, friends may be an undervalued resource. Later, J. Jensen et al. (2019) reported that relationship approval from women's friends was significantly positively associated with women engaging in more RW with their partners. However, the same connection was not found when examining women's friends' approval and men's likelihood to engage in RW with their partners. Researchers found that closer friendships were associated with more self‐reported RW with those friends, especially for men. J. Jensen et al. (2018) further explored the connection between skin conductance—a type of physiological arousal—and RW. Skin conductance activity (SCA) was examined due to its connection to physiological stress through skin perspiration. SCA was captured via finger sensors while participants discussed their relationship trials with partners and friends. Men experienced more SCA than women during these conversations. Surprisingly, participants experienced greater SCA during RW with friends as compared to RW with partners. In other words, participants experienced greater physiological stress when discussing their romantic problems with friends than with their partners, which may suggest that venting to friends is not as beneficial as many believe it to be. When friends validate a partner's frustration or anger, they also often inadvertently heighten physiological stress reactions in the partner through mutual venting or passionate support. As a result, turning to friends to the exclusion of partners to discuss romantic trials may actually exacerbate individuals' physiological stress.

1.3. Psychophysiology in Couples

Psychophysiological arousal during couple conflict has been examined by researchers for decades. Monitoring physiological changes during couple interactions was first pioneered by Levenson and Gottman (1983). They examined thirty married couples and how physiological changes and affective patterns were associated with marital satisfaction. They hypothesized that physiological arousal would be associated with more negative affective patterns by considering physiological markers, including heart rate, pulse transmission time, skin conductance level, and general somatic activity. After speaking together for 20 min while being observed by researchers, the couples returned to watch the video of their interactions and self‐reported the changing affect they observed. Results revealed negative affect (e.g., stereotyping, rigidity, and sequential restraint) was seen more frequently in dissatisfied pairings, and 60% of the variation in marital satisfaction was accounted for due to the physiological linkage. This study was expanded upon later by re‐examining 19 of the original couples (Levenson and Gottman 1985). Notably, the more physiologically aroused the couple during the original interactions, the more their satisfaction declined over 3 years.

Heart rate variability (HRV) is a measure of physiological arousal that captures the discrepancy in time between heartbeats (Shaffer and Ginsberg 2017). Tracking changes in HRV can be insightful as the body continuously adapts to stimuli during interactions. Quantifying HRV can be accomplished through a variety of time and frequency domain metrics via nonlinear dynamics (Ciccone et al. 2017). Root mean square of successive difference (RMSSD) is commonly accepted as an appropriate time domain measure for HRV (Ciccone et al. 2017). This metric is captured using an electrocardiogram and determining the time interval between adjacent R‐waves (Ciccone et al. 2017). Cardiac parasympathetic activity is reflected in RMSSD values, which can vary depending on age, fitness, and heart health (Penttilä et al. 2001). In general, RMSSD values below 20 ms (ms) are considered low and indicate higher stress levels. Values between 30 and 60 ms are considered normal or baseline, and values above 60 ms indicate a more relaxed state (Shaffer and Ginsberg 2017). Tracking RMSSD has proven effective in measuring stress and emotional responses in adolescents and adults. Specifically, slowed, paced breathing was found to provide physiological relief (i.e., greater RMSSD) for adolescents under high‐stress conditions (Laborde et al. 2017).

Although only a singular measure of cardiac function and response, higher RMSSD values generally indicate greater flexibility to cope with stressors due to increased HRV (Laborde et al. 2022; Shaffer and Ginsberg 2017). HRV is importantly associated with heart rate (HR) based upon the physiological and mathematical relationships between the two components (Sacha 2014). A healthy HR is generally considered to be between 60 and 100 bpm, with higher scores indicating increased exertion, stress, or anxiety. J. Jensen, Rauer, et al. (2018) included a physiological component in the previously discussed RW study. Participants were monitored for HR at baseline when engaged in RW with a friend, and when discussing the same romantic challenge with their partner. HR was measured with a photoplethysmogram sensor attached to the participants' index finger on the nondominant hand. Physiological arousal was associated with RW, but researchers noted that there was not a significant difference in HR when performing RW with a friend or partner. Perhaps due to the pressure women often feel to maintain relationships, compared to men in the study, women experienced significantly higher HR scores at baseline, when with a friend, and with their partner. Despite these findings, researchers remain uncertain regarding how HRV is associated with romantic relationship functioning. As such, it is important to examine whether relationship adjustment might be positively or negatively associated with HRV.

2. Purpose of Study

Limited research (e.g., Levenson and Gottman 1983) has been conducted on partners' physiological data when conversing about romantic problems. Yet, these findings have been crucial in adding to the overall understanding of how couples experience distress during conflict. To our knowledge, no previous research has examined how HRV is affected when discussing romantic challenges with friends. Such disclosures to the social network are typical (Helms et al. 2003), and a greater understanding of these conversations could contribute to a more holistic comprehension of romantic relationship functioning (Huston 2000). By assessing how romantic trials may impact physiological functioning, clinicians may be better able to conceptualize the significance that these problems carry. Further, by including friends' relationship perceptions, clinicians may gain a deeper understanding of social factors affecting relationships. Additionally, investigating whether or not relationship adjustment is impacted when discussing romantic problems with friends (who either approve or disapprove of the romantic union) may help determine the effects of friendship on relationships. Therefore, the current study sought to explore how emerging adults interact with their partner and with a friend when discussing relationship challenges, as well as the resulting physiological changes that occur during these conversations.

The following research questions were addressed: (1) In emerging adult romantic couples, is social network approval linked with relationship adjustment and RW with partners and friends? (2) How do emerging adults' perceptions of their friends' approval of their romantic relationship differ from their actual approval? (3) For emerging adults in romantic relationships, how are relationship adjustment and HRV associated with RW with a partner and friend as they engage in relationship discussions, accounting for perceptions of friends' relationship approval? Hypotheses corresponding to each of the research questions are as follows: (1) based on the research by J. Jensen et al. (2019), we predict that for both women and men, social network approval will be significantly positively associated with relationship adjustment and greater RW with partners and friends. (2) By taking into account the findings of Felmlee (2001), which demonstrated that perceived social network approval had a more significant positive effect on relationship stability than actual social network approval, we anticipate that emerging adults will be more likely to perceive their friends as more approving than they actually are. (3) We anticipate that relationship adjustment and HRV will be positively associated with RW with partners, based on the previous findings by Gottman and Levenson (1985).

3. Methods

3.1. Sample and Procedures

The data used for the current study was originally collected as part of the larger ongoing RW in Young Adults Study. This study considered emerging adults at a large, public university in the United States and was originally approved by the Institutional Review Board in May 2015. Only procedures and measures relevant to the current study are discussed. Although 75 emerging adult couples and a friend of each partner (i.e., 300 individuals) were recruited for participation, we retained 71 of those couples for analyses. Four identified as same‐sex couples, and we did not have sufficient power to analyze their results. Participants were Caucasian (73%), Black (9%), Hispanic (8%), Asian (7%), or “other” (3%). Women were on average 20.32 years old (SD = 1.45), and men 21.23 years old (SD = 2.42). Average romantic relationship length was 22.40 months. No participants had children.

Participants were recruited via classroom visits, research flyers, and university emails from an assortment of undergraduate and graduate‐level courses (e.g., Human Development, Biology, English, and Nursing). All participants fulfilled the following criteria: (1) In each group of four (two partners and two friends), at least one had to be a university student. (2) All participants were 18 years or older. (3) All partners indicated that they were in a romantic relationship. (4) Each partner identified a close friend and brought that friend to the research lab. Each group of four spent approximately 90 min participating in various tasks that captured personal, romantic, and social functioning. Additionally, all participants completed online questionnaires through Qualtrics about their personal characteristics, such as age and ethnicity, their romantic relationship (e.g., length and quality), and their friendship (e.g., length and quality). Along with questionnaires, participants were asked to engage in a video‐recorded conversation in which they discussed a romantic problem they were currently experiencing or had experienced. To match the typical social process (Helms et al. 2003), partners first spoke to one another and then to a friend and discussed the same romantic challenge with both. Before the video conversation, each partner received a list of common romantic relationship problems (e.g., needing to be more organized and wanting to make love more often) and were asked to independently rate what they felt were the most significant romantic problems on a six‐point Likert scale from 0 (“Not an issue”) to 5 (“Major problem”). Participants were encouraged to choose a topic from this list that received a higher score as the subject to discuss. Each partner identified a separate and unique issue that was a concern. The average score for the issue chosen was 3.6, suggesting a moderately high level of problem severity. After one partner discussed the issue with the partner and friend, the other partner followed the same pattern with their issue.

To account for potential gender dynamics, the order of participation was counterbalanced such that for odd‐numbered participants, the woman first discussed her problem with her partner and then her friend, and for even‐numbered participants, men went first. When discussing the issue with a partner, the friend left the room, and vice versa. While conversing about the romantic challenge, HRV of the partner leading the conversation was captured. Romantic partners, not friends, had psychophysiological data assessed and monitored. To compensate participants for their time, partners each received a $50 gift card, and each friend received a $35 gift card.

3.2. Measures

3.2.1. Dyadic Adjustment

Dyadic adjustment was captured using the 14‐item revised dyadic adjustment scale (Crane et al. 2000). The first six items were scored on a six‐point Likert scale, with 0 representing “Always disagree” and 5 representing “Always agree.” The next four items were also scored on a six‐point scale, with 0 representing “All the time” and 5 representing “Never.” Next, one item was scored on a five‐point scale ranging from 0, representing “Never,” to 4, representing “Every day.” The final three items were scored on a six‐point Likert scale ranging from 0, representing “Never,” to 5, representing “More than once a day.” Sample items included “How often do you and your partner quarrel?” and “How often do you and your partner work together on a project?” Items were summed and higher scores indicated greater dyadic adjustment. Reliability was acceptable (women, α = 0.64; men, α = 0.68).

3.2.2. Relationship Work

The frequency of RW with partners and friends was measured using a modified five‐item Marriage Work Scale (Helms et al. 2003). Some items from the original scale were dropped due to not being applicable for this population (i.e., items related to child rearing). Retained items were relevant to the sample, such as “How often do you bring up how well you and your partner get along with one another's families and how much and how often you see them?” Respondents selected a number on a nine‐point Likert scale, with 1 representing “Never” and 9 representing “Always.” Each question separately addressed RW with the respondent's partner and friend. Mean scores were created for each variable, and higher scores indicated more frequent RW with either partners or friends. Appropriate reliability was found for both the partner (women, α = 0.72; men, α = 0.78) and friend scale (women, α = 0.75; men, α = 0.85).

3.2.3. Social Network Approval

Partners' perceptions of their friends' approval of their romantic relationship were captured using a single‐item measure (“To what degree do you think that your close friend who accompanied you to the research visit today approves of your current romantic relationship?”), ranging from 1 (“Very much disapproves”) to 7 (“Very much approves”). Friends' approval of romantic relationships was captured using the six‐item Social Reactions Scale (Sprecher 2011). Responses were scored on a seven‐point Likert scale, with higher scores indicating greater approval of the romantic relationship. Mean scores were created, and reliability was found to be acceptable (women's friends, α = 0.95; men's friends, α = 0.84).

3.2.4. Heart Rate Variability

HRV was captured using Nexus‐10 psychophysiological equipment and Biotrace+ software. HRV can be measured in a variety of ways; we elected to examine the RMSSDs. RMSSD is a time domain measure for HRV that determines the amount of time between adjacent R‐waves (Ciccone et al. 2017). Higher RMSSD suggests greater flexibility to handle stressors and less physiological arousal (Laborde et al. 2022). Participants were monitored for RMSSD at baseline, during an RW conversation with the partner, and during RW with the friend. RMSSD was measured with a photoplethysmogram sensor attached to the participants' index finger on the nondominant hand.

3.3. Data Analysis

To answer our research questions, we first examined univariate descriptive statistics followed by bivariate correlational analyses among all study variables. Actor–partner interdependence modeling (APIM) was conducted to investigate relationships among multiple independent and dependent variables simultaneously. For our first research question regarding potential links between social network approval, relationship adjustment, and RW, bivariate correlations were conducted. Regarding our second question, in examining comparisons between partners' perception of friends' approval versus friends' actual approval of the relationship, we conducted Kendall's tau correlation comparisons in addition to examining the mean scores and other descriptive statistics. Path modeling (i.e., APIM) was conducted to answer our third research question, which considered the potential associations among all study variables simultaneously and examined actor and partner effects.

4. Results

Descriptive statistics and correlations can be found in Table 1. The sample showed a mean score of 49.31 (SD = 5.75) for dyadic adjustment for women and 49.24 (SD = 6.91) for men, suggesting a satisfied sample. The clinical cutoff score is 48, with scores ≤ 47 indicating relationship distress. Our first research question considered whether social network approval was linked with relationship adjustment and RW with partners and friends. Correlations suggested for women, friends' relationship approval was significantly correlated with women's self‐reported relationship adjustment (r = 0.23, p < 0.05). Similarly, for men, friends' relationship approval was significantly associated with men's dyadic adjustment (r = 0.24, p < 0.05). For women, we did not find a significant correlation between friends' social network approval and the amount of RW with partners or friends. We found marginal associations for men (p < 0.10; see Table 1).

Table 1.

Descriptive statistics and intercorrelations for study variables (N = 284).

Variables	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.	13.	14.	15.	16.	17.
1. Rom Rel length	—
2. W age	0.32**	—
3. M age	0.22†	0.44**	—
4. W RDAS	0.05	0.26*	0.06	—
5. M RDAS	−0.09	0.01	0.10	0.38**	—
6. W RW w/partner	0.29*	0.01	0.09	0.41**	0.25*	—
7. M RW w/partner	−0.10	0.07	−0.05	0.22*	0.21*	0.15	—
8. W RW w/friend	−0.13	−0.07	−0.13	−0.04	0.03	0.27*	−0.16	—
9. M RW w/friend	−0.08	−0.02	−0.01	−0.07	0.18	0.06	0.45**	0.25*	—
10. W Perc of Approv	0.07	−0.01	0.12	0.11	0.18	0.01	0.07	0.06	0.13	—
11. M Perc of Approv	0.12	−0.07	0.17	0.20†	0.20†	0.30*	0.11	0.10	0.22†	0.46**	—
12. W friend Approv	0.20†	0.09	0.07	0.23*	0.23*	0.15	0.08	−0.16	−0.11	0.68**	0.41**	—
13. M friend Approv	0.12	−0.08	0.02	0.25*	0.24*	0.23†	0.21†	0.04	0.16	0.36**	0.79**	0.42**	—
14. W RMSSD w/partner	0.29*	−0.10	−0.05	−0.05	−0.18	−0.16	−0.12	−0.11	−0.02	−0.01	−0.15	−0.12	−0.14	—
15. M RMSSD w/partner	−0.10	−0.08	−0.11	−0.24*	−0.04	−0.02	0.04	0.14	0.30*	−0.05	0.13	−0.04	−0.05	0.09	—
16. W RMSSD w/friend	−0.13	−0.17	−0.13	0.06	−0.21†	0.07	−0.08	−0.16	−0.09	−0.18	−0.15	−0.17	−0.12	0.84**	0.04	—
17. M RMSSD w/friend	−0.08	0.01	−0.12	−0.19	−0.06	−0.02	−0.04	0.12	0.18	−0.11	0.07	−0.06	−0.15	0.06	0.89**	0.04	—
M	22.40	20.52	20.59	49.31	49.24	7.11	6.57	5.43	4.49	6.29	6.00	5.65	5.53	80.63	74.60	84.73	82.32
SD	19.42	1.66	1.36	5.75	6.91	1.23	1.45	1.69	1.87	1.06	1.11	0.80	0.90	43.75	39.39	41.18	45.87

Note: Romantic Relationship Length is in months.

Abbreviations: Approv = approval, M = men, Perc = perception, RDAS = revised dyadic adjustment scale (sum score provided), RMSSD = root mean square of successive difference, Rom Rel = romantic relationship, RW = Relationship Work, W = women.

^_† p ≤ 0.10.

^* p < 0.05

^** p < 0.01.

To answer our second research question regarding comparisons between partners' perception of friends' approval versus friends' actual reported approval, we conducted Kendall's tau correlation comparisons and also examined mean scores and bivariate correlations. To note, both women's and men's perceptions of approval and actual friends' approval were significantly, positively correlated. Kendall's tau comparisons revealed that for both women (r = 0.53, p < 0.001) and men (r = 0.64, p < 0.001), partners' perceptions of friends' approval and friends' actual self‐reported approval were significantly correlated, demonstrating moderate concordance between the two scores. The tau correlations were then transformed to z scores to determine if the correlations significantly varied; the tau correlations were not significantly different (z = − 0.98, p = 0.16), indicating the concordance relationship did not vary between women and men.

Finally, to answer our third research question, we fit an APIM testing all links among partners' self‐reported RW with their partner and friends and perceptions of friends' social network approval with RMSSD with partner, RMSSD with friend, and romantic adjustment, controlling for age and romantic relationship length (see Figure 1). Noting that perceived support for the relationship may have a greater effect than actual approval (Felmlee 2001), acknowledging power constraints due to sample size and number of variables, and considering the large associations between perceived and actual friend approval in this sample, we only included partner's reports of friends' approval in the path model. Data were analyzed using Mplus Version 8, and missing data were handled using full information maximum likelihood. Residuals of the dependent variables were allowed to covary. Goodness of fit was evaluated using the Chi‐square statistic, comparative fit index (CFI), Tucker–Lewis index (TLI; Bentler 1990), and root mean square error of approximation (RMSEA; Hu and Bentler 1995). The model indicated a good fit with χ ² = 18.83, df = 22, and p = 0.66. Additionally, CFI = 1.00, TLI = 1.00, and RMSEA = 0.00. Results especially pointed to significant actor effects (see Figure 1).

Figure 1.

Actor–partner interdependence model of links among women's and men's romantic, social, and psychophysiological measures, controlling for age and relationship length (n = 284). Note: Standardized path coefficients are shown. Model fit statistics: χ² = 18.83, p = 0.66, df = 22, CFI = 1.00, TLI = 1.00, RMSEA = 0.00. *p < 0.05; **p < 0.01; ^† p ≤ 0.10. Although all potential pathways were tested, only significant pathways are depicted. Although not depicted, we controlled for partners' age. Women's age was significantly positively associated with women's romantic adjustment. CFI, comparative fit index; RMSEA, root mean square error of approximation; RMSSD, root mean square of successive difference; RW, relationship work; TLI, Tucker–Lewis index.

Aligning with the correlational results, women's (β = 0.43, p < 0.01) and men's (β = 0.27, p < 0.05) RW with their partners were significantly, positively associated with their respective dyadic adjustment. Consistent with our hypothesis and previous RW literature (Helms et al. 2003; J. F. Jensen and Rauer 2016), this suggested that the more RW women and men reported doing with partners, the greater the relationship adjustment. Additionally, men's RMSSD during conversations with partners (β = 0.51, p < 0.01) and friends (β = 0.60, p < 0.01) was significantly, positively associated with their perception of their friend's approval. This indicates that when men perceive their friends as approving of the relationship, they demonstrate better HRV during stressful conversations with both friends and partners. Moreover, men's RMSSD was significantly, positively linked with men's RW with friends (β = 0.29, p < 0.05), suggesting that men enjoy better physiological functioning when talking with their partners, if they report more frequently engaging in RW with friends. Meanwhile, women's RMSSD during conversations with friends was found to be marginally positively associated with their RW with their partner, and marginally, negatively linked to their RW with friends. This suggests a possible trend that women's physiological functioning when discussing relationship issues may be enhanced by talking about romantic trials more frequently with partners and less frequently with friends. Additionally, we noted the following significant covariances among the residuals of the dependent variables: women's RMSSD with a partner and women's RMSSD with a friend (r = 0.88, p < 0.01); men's RMSSD with a partner and men's RMSSD with a friend (r = 0.87, p < 0.01); women's dyadic adjustment with men's RMSSD with a partner (r = − 0.26, p < 0.05); and, demonstrating interdependence, women's dyadic adjustment with men's dyadic adjustment (r = 0.43, p < 0.01). In other words, the “leftover” variance in each dependent variable noted above not accounted for by the model (could be attributed to measurement error, individual‐level factors, etc.) are related. When covariances among the same variable for each member of the dyad are significant, this indicates the relationship between the leftover variance among the two variables not accounted for by the model is, at least in part, due to the interdependence of the dyad. This model accounted for 8.8% of the variance in women's RMSSD with a partner, 13.1% of the variance in women's RMSSD with a friend, 34.4% of the variance in women's romantic adjustment, 22.1% of the variance in men's RMSSD with a partner, 21.7% of the variance of men's RMSSD with a friend, and 24.0% of the variance of men's dyadic adjustment.

5. Discussion

Relationships do not exist in isolation (Felmlee 2001). When partners enter a romantic union, the opinions and messages of those around them can have lasting effects on how the relationship progresses, the values partners hold, and the problems they may experience (Jenson et al. 2021; Zak et al. 2002). By capturing the influence of the social network, clinicians can expand their knowledge regarding how to effectively help couples navigate relationship challenges (Blair and Pukall 2015; Milardo 1982). Researchers will also benefit from considering how partners share romantic trials with their friends, as well as the level of approval friends provide. Our results add importance to what is known regarding individual physiological, romantic, and social functioning in young adulthood. Consistent with previous literature (J. Jensen, Rauer, et al. 2018; Proulx et al. 2004), both women and men enjoyed greater romantic adjustment when they frequently turned to their partner to process romantic challenges. Additionally, men's healthy physiological functioning was importantly linked to both their ability to turn frequently to friends to talk about their relationship, as well as to men's perceptions of friends' approval.

5.1. Turning to One's Partner Matters

Aligning with numerous RW studies (Helms et al. 2003; J. F. Jensen and Rauer 2016; Proulx et al. 2004; J. Jensen, Rauer, et al. 2018), we found that both women and men benefit when frequently turning to their partner to talk about the challenges they face in their relationship. Controlling for age and romantic relationship length, and accounting for the perception of friend's approval, our path analysis suggested that both women and men who reported turning more often toward their partners to discuss romantic challenges, also reported greater romantic adjustment. It is notable that actor effects were present for people of both genders, linking RW with partner—but not RW with friend—to romantic adjustment. Punctuating the work of J. F. Jensen and Rauer (2016), this finding suggests that although many may opine that triangulating a friend into romantic conflict to “blow off steam” or act as a sounding board may help a partner, turning to the partner appears to remain the better choice. Results of this study support the notion that engaging in frequent RW with one's partner is beneficial and serves as a protective factor against relationship dissolution as the romantic adjustment is higher (J. F. Jensen and Rauer 2016).

Although engaging in a conversation about a romantic problem with one's partner may be uncomfortable, and it may seem easier to turn to a friend, partners should consider the social context that informs such processes. Previous work indicates that friends generally report a more negative perception of one's romantic relationship than romantic partners themselves (Agnew et al. 2001). McDermott et al. (2013) reported that romantic instability occurs in clusters within a social network. Individuals experiencing romantic challenges who break up may influence not only their friends but also their friends' friends as the propensity to end the relationship spreads (J. F. Jensen and Rauer 2016). Consequently, mutual sharing of romantic problems within a social network appears to have the potential to negatively impact multiple relationships. Thus, if romantic relationship maintenance is valued, caution should be exercised as partners share romantic challenges with their friends. Conversely, numerous studies now reveal that despite its potential discomfort, addressing these issues with the partner is linked with positive individual and relationship outcomes (J. F. Jensen and Rauer 2014; J. Jensen et al. 2019).

For family therapists, the notion of turning outside the romantic relationship to share with a third party likely calls to mind the conceptual literature of Bowen's healthy versus unhealthy triangles (Kerr and Bowen 1988). Within healthy triangles, each member communicates directly with the others (i.e., open RW in all directions), emotional stability is maintained in lieu of amplifying anxiety, and problem‐solving strategies are used to reduce stress (Goldenberg and Goldenberg 2013). Conversely, unhealthy triangles occur when the third person plays the role of anxiety buffer, and avoidance of conflict or emotional cutoff is perpetuated (Nichols and Davis 2020). Given the results of the current study and of previous RW studies, it appears that frequently, when partners engage in RW with friends, an unhealthy triangle is formed. Therapists may need to provide guidance regarding romantic relationship boundaries when talking to friends, and encourage partners to turn toward one another more often.

5.2. Talking About Relationships With Others Is Critical for Men

In addition to uncovering the importance of RW with a partner for both women and men, when examining our third research question, we also discovered that men's physiological functioning during RW with a partner was significantly healthier when they were more open with their friends about romantic struggles. Rather than undoing the findings associated with the importance of RW with partners discussed above, we posit that this discovery aligns with previous literature highlighting the importance of men doing RW with anyone. Although frequently engaging in RW with friends—especially to the exclusion of partners—has proven problematic for women (i.e., greater likelihood of breakup; J. F. Jensen and Rauer 2016), the same pattern does not seem to hold true for men. J. F. Jensen and Rauer (2014) noted that men who engaged in frequent RW with both partners and friends reported the highest level of happiness among women and men. In fact, taken together across all RW studies, it appears that heterosexual partners in the United States may need to find the “Goldilocks” zone when it comes to RW with friends. Relationships appear to benefit most when women share slightly less about romantic problems with their friends than they are socialized to do, and men slightly more. Due to gendered socialization patterns, women tend to be more apt than men to disclose relationship details to friends and family (Busby et al. 2017). Nevertheless, when men display more sensitivity to relationship issues by opening up to partners, friends, and others, they often experience personal and relationship benefits (e.g., improved communication, deeper emotional intimacy, and greater relationship satisfaction; Brown 2012). Further highlighting the importance of men's connections to their social networks, we found that men's perception of friends' approval was linked with healthier physiological functioning during their RW conversations with both partners and friends. Aligning with the tenets of polyvagal theory (Porges 2022), men may feel more secure and at ease—reflected by better observed cardiovascular functioning—when discussing romantic challenges with friends, if they perceive friends approve of their partner. Personal health benefits may be associated with perceptions of friends' approval that are linked with men talking vulnerably with friends about their partner (Levant and Wong 2017).

5.3. Links Among Physiological, Social, and Romantic Functioning

In partial answer to our third research question, we found that HRV is linked with social factors only for men. HRV, as measured by RMSSD, was found to be higher, and therefore, indicative of less emotional stress (Shaffer and Ginsberg 2017) when men reported their friend as more approving of the relationship. Given that perception of greater social network approval has been linked to greater romantic stability (Blair et al. 2023; Felmlee 2001), men may feel calmer during RW conversations when they perceive their friends as approving. Due to the significant links between HRV and personal health (e.g., greater cardiovascular health, stress resilience, sleep quality, cognitive functioning, and emotional regulation; Shaffer and Ginsberg 2017; Thayer et al. 2012), men should be educated regarding the personal benefits of being vulnerable with friends about romantic trials and the importance of friends' approval of their relationships.

Although social networks are commonly examined, most researchers do not include third‐party members in study designs (Blair and Pukall 2015; Zak et al. 2002). Therefore, in answering our second research question and extending beyond existing literature, it was important to analyze how perceptions of approval may differ from actual approval ratings provided by those outside the relationship. Results revealed moderate concordance between the perceived approval score and the actual score reported by the friends. Men and women were found to have similar alignment, indicating both genders are moderately accurate in their perceptions of their friend's approval rating. Social network approval impacts romantic relationships by influencing factors, such as relationship longevity (Lehmiller and Agnew 2007), stability (Felmlee 2001), and affection‐sharing behavior (Blair et al. 2023). Consistent with our hypothesis for our first research question and with previous literature (Busby et al. 2017), our correlational results showed that both women and men reported greater relationship adjustment when their friends approved of the romantic relationship. Harkening back to social exchange theory (Cropanzano and Mitchell 2005), the meaningful benefits of friends' approval may contribute to romantic satisfaction and stability, as partners determine that the pros of remaining in a romantic union (e.g., “My friend really likes my partner!”) outweigh the costs.

5.4. Strengths and Limitations

The current study has several strengths, including its mixed‐method design, which included psychophysiological assessments. To our knowledge, the investigation of physiological stress via HRV (i.e., RMSSD; Penttilä et al. 2001) in the context of dyadic relationship discussions, had not been previously examined. Accounting for friends' approval of romantic functioning added a triadic data component to the literature as we assessed individual physiological, social, and romantic functioning—something no other RW or psychophysiological study has done (J. F. Jensen and Rauer 2014, Proulx et al. 2004). Yet, despite these strengths, there were several limitations that merit attention. Although the sample size was justifiable given our hypotheses, future studies examining similar constructs should aim to increase the number of couples. Specifically, it would be advantageous to extend the research beyond heterosexual relationships given that same‐sex and nonbinary couples rely significantly more on their social networks than heterosexual couples (Blair and Pukall 2015). Despite our attempts to address the observer effect using baseline physiology, some participants may have felt uncomfortable being watched during conversations, which may have influenced physiological readings. We also recognize the potential limitation of requiring one of the four participants to be a university student, a decision tied to university compensation that was outside the researchers' control. Additionally, partners' perceptions of social network approval were captured using a single‐item measure. The use of a multiquestion assessment may have more thoroughly captured perceptions of social network approval. Furthermore, the reliability scores for the 14‐item dyadic adjustment (Crane et al. 2000) were lower than expected and should be interpreted with caution. Nevertheless, reliability scores should be interpreted within the context of the well‐established limitations of Cronbach's alpha, most notably when utilizing scales of a limited item count (Hinton et al. 2004). Moreover, we failed to counterbalance the order of partners speaking to one another and then to friends. Having every other set of participants engage in RW with friends first may have improved the study design and potentially avoided any physiological attenuation to the discussion, including partners feeling more at ease discussing the issue for the second time. Finally, the cross‐sectional nature of the study prevents us from commenting on the direction of effects.

5.5. Clinical Implications and Conclusion

Romantic relationships are nestled within the confines of robust and complex social networks (Milardo 1982). As such, clinicians should be more diligent in accounting for the social network when treating couples. Our findings point to the importance of the effect that perceived social network approval has on partner discussions, especially for males. Therapists should feel empowered to educate males not only on the relational benefits of their openness and vulnerability (Rosenfield and Mouzon 2013) but also on the individual physiological benefits of more open relationship discussions. This is especially important as it further clarifies the meaningful influence of the social network on romantic dynamics described by Huston (2000) in his socioecological model of romantic functioning.

Clinicians should account for how supported each partner feels by their larger social network through assessment. Therapists may benefit from speaking with couples more openly about social network involvement, influence, and approval. Such discussions may inform conclusions about the impact of clients' external relationships. While most clinicians consider the systemic effects parents and immediate family members can have on relationship functioning (Becvar et al. 2023), assessing for friend support or lack thereof, especially in young adulthood, may be important in developing a more accurate perception of couple functioning beyond the therapy room. By thoroughly investigating the involvement of friends, individuals who have a significant influence on the relationship may be identified. Clinicians may provide strategies to involve friends in romantic conflict in productive ways without causing detriment to the relationship (i.e., being a friend of the relationship; J. Jensen et al. 2019). Given previous findings that engaging with friends in RW does not affect relationship satisfaction provided that RW with partners occurs with equal or greater frequency (Helms et al. 2003), partners should not be dissuaded from discussing problems with friends so long as discussions are co‐occurring between partners. Instead, they should be encouraged to be thoughtful about whom to discuss these issues with, and how such discussions can prioritize the health of the relationship (Blair and Pukall 2015). For example, J. Jensen et al. (2019) noted that engaging in mutual venting with friends about partners may actually increase stress instead of reducing it. Rather, if romantic relationship maintenance is a priority, partners should be encouraged to disclose to friends who may validate their feelings, but also speak supportively of the relationship and the other partner.

Clinicians should also consider expanding the therapeutic system by inviting friends into therapy with the couple. Given that partners often discuss romantic conflicts with friends (Proulx et al. 2009), this may prove valuable in addressing romantic challenges. Communication issues may also be more effectively mitigated by engaging in more RW with their partner (Helms et al. 2003), and men being more open in general regarding their relationships. Clinicians should encourage the sharing of problems and willingness to engage in conversations about romantic conflict. Healthy strategies for these conversations should be emphasized, such as sharing underlying emotions, utilizing “I‐statements,” and active listening. As researchers and clinicians continue to educate couples on how to engage in effective RW, we are hopeful relationship outcomes will improve and create healthy and long‐lasting partnerships.

In conclusion, both men and women appear to benefit when frequently turning to one another to discuss romantic challenges. In fact, people of both genders report that they themselves enjoy greater romantic adjustment when they turn to their partners often. Moreover, men seem to especially benefit from being more open when it comes to discussing relationship issues with others. When men turn to their friends and are open about relationship challenges, they experience healthier physiological functioning during these conversations. Men must continue to receive the message in therapy that their openness, vulnerability, and willingness to engage in relational conversations bring numerous benefits not only to the relationship but also to men's own health and well‐being. Finally, clinicians are encouraged to more thoroughly account for and be aware of social influences when working with partners in couples therapy. Discovering how often partners involve their friends in romantic disputes and the strategies they use when engaging in these discussions may not only be worthwhile for improving romantic functioning but may also shed important light on individual physiological health.