Exploring the effects of pillow talk on relationship satisfaction and physiological stress responses to couples’ difficult conversations

ABSTRACT

This study examined the effects of pillow talk (i.e., communication after sexual activity) on relationship satisfaction and physiological stress responses to difficult conversations. Fifty heterosexual couples were randomly assigned to either a pillow talk condition in which they doubled their pillow talk for three weeks or a control condition. After three weeks, participants came to the lab to engage in a conflict conversation. Saliva samples taken before and after the conversation were assayed for cortisol, a physiological marker of stress. The results indicate that increased pillow talk produces greater relationship satisfaction for men (but not women), but does not benefit either group’s physiological stress responses. The findings suggest that pillow talk has distinct benefits for romantic relationships.

KEYWORDS:

• Post sex communication
• pillow talk
• stress
• resilience
• affectionate communication

Acknowledgements

The authors would like to thank Dr. Paul Schrodt and the reviewers for their insightful suggestions and guidance in improving the manuscript, as well as Dr. Tamara Afifi for her feedback and advice. They would also like to thank the undergraduate research assistants in the Department of Communication at the University of Connecticut for their assistance with data collection.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes on contributors

Amanda Denes is an associate professor in the Department of Communication at the University of Connecticut (PhD, University of California, Santa Barbara).

John P. Crowley is an assistant professor in the Department of Communication at the University of Delaware (PhD, University of Washington).

Kara L. Winkler is a lecturer in the Department of Arts and Communication at the University of Houston-Downtown (MA, California State University, Sacramento).

Anuraj Dhillon is an assistant professor in the Communication Studies Department at the California Polytechnic State University, San Luis Obispo (PhD, University of Connecticut).

Ambyre Ponivas is a doctoral candidate in the Department of Communication at the University of Connecticut (MA, Chicago School of Professional Psychology).

Margaret L. P. Bennett is a visiting assistant professor in the Department of Communication at the University of the Pacific (PhD, University of Connecticut).

Notes

1 AUC_G captures changes in cortisol over time, accounting for “the difference between the single measurements from each other (i.e., the change over time) and the distance of these measures from the ground, or zero (i.e., the level at which the changes over time occur)” (Pruessner et al., 2003, pp. 918–919). AUC_I uses the time 1 measurement as a baseline of comparison for subsequent time points (Pruessner et al., 2003). Thus, higher AUC_G scores indicate that individuals consistently released more cortisol and were more stressed throughout the conversation (including before beginning the conversation), whereas higher AUC_I scores indicate a greater increase in cortisol output from baseline to post-conversation (i.e., individuals became more stressed after engaging in the conversation).

2 The larger study included four same-sex couples, as no restrictions were placed on sexual orientation or relationship type when recruiting participants. However, the present analyses only focused on the heterosexual couples in the sample. This allowed for the use of an actor-partner interdependence model with distinguishable dyads (by biological sex) to test the effects of the intervention on women’s and men’s relationship satisfaction and physiological stress responses, as well as to test both actor and partner effects of relationship satisfaction on physiological stress responses.

3 More specifically, participants had to be in good (self-reported) health and could not have an acute or chronic illness, an endocrine disorder or be taking any anti-anxiety, antipsychotic, antidepressant, or mood stabilizing medications, anti-inflammatory steroid based medications, or be on any hormone replacement therapies (not including hormonal birth control). If participants met these criteria, they were directed to an online study sign-up survey. Participants were also asked to list any medication use, health issues, endocrine disorders, hormone replacement therapies, or experiences with chemotherapy in a short health survey during their lab visit. No participants reported any medication use or health issues that would disqualify them from the study, as indicated by their responses in the health survey.

4 In total, 209 individuals completed the screening survey (including repeats and instances where both members of the couple completed the screening survey). Fifty-eight percent of those individuals indicated that they engaged in more than 30 minutes of pillow talk and were disqualified from the study. Individuals who completed the screening survey twice (if they failed to meet the requirements the first time; i.e., they reported engaging in more than 30 minutes of pillow talk, then entered the system a second time and reported engaging in less than 30 minutes of pillow talk) were also disqualified from participating.

5 Because couples entered the study on a rolling basis and to reduce selection bias (see Kahan, Rehal, & Cro, 2015), simple (unrestricted) randomization was used (i.e., each couple had a 50/50 chance of being assigned to the experimental or control condition).

6 Participants in the experimental condition received the following instructions via email: “This study will take place over the next 3 weeks. During these 3 weeks, we are asking that you DOUBLE THE AMOUNT OF PILLOW TALK you engage in with your partner after sexual activity. ‘Pillow talk’ refers to the time you talk with your partner immediately after sexual activity (which can include after sexual intercourse, oral sex, hand stimulation, etc. – any activity ‘below the belt’). For example, you said that you usually spend X minutes talking to your partner after sexual activity, so we ask that you now spend X*2 minutes talking to them after each time you engage in sexual activity. Your partner is receiving the same email, so it’s fine to talk to them about this assignment.” The “X”s were replaced with the specific amounts participants indicated in the sign-up survey, with X*2 representing the doubling of pillow talk for individuals in the experimental condition.

7 It is also worth noting that both women and men reported being highly satisfied with the post sex intimate talk occurring in their relationship, as indicated by their response to one item asking how satisfied they were with the post sex intimate talk that occurred in their relationship (on a scale from 1 to 5 with higher scores reflecting greater satisfaction; women’s M = 4.24, SD = 1.00; men’s M = 4.44, SD = .91).

8 We recognize that the findings may be a result of demand characteristics (Orne, 1962) or pressures the participants in the experimental condition may have felt to report that they had doubled their pillow talk. However, McCambridge, de Bruin, and Witton (2012) note that demand characteristics in non-laboratory settings lack empirical validation. Given that the experimental manipulation occurred outside of the laboratory setting, demand characteristics may not be of concern. Nonetheless, to address possible demand characteristics, intervention experts were consulted prior to the study and every effort was made to ensure that participants did not feel pressure to report inaccurate information. More specifically, participants were reminded that their responses could not be linked with any identifying information. Additionally, research credit was given at each phase of the study and, therefore, participants should not have felt pressure to report inaccurate information at T2 out of fear of losing credit. Participants were also reminded that they would not be penalized for any responses or for deciding not to complete the study, as per IRB procedures, and that research credit and monetary compensation would be provided before any participants’ data were reviewed. Additionally, research credit was recorded by instructors who were not involved in the study. Taken together, these procedures increase confidence in the veracity of the data, as participants had no reason to report inaccurate information.

9 Post hoc exploratory analyses indicated that men who participated in the pillow talk intervention reported perceiving more risks to disclosing to their partners than those in the control group, which adds support to the point that this group of men may have felt additional pressure to communicate in a positive and effective manner. However, perceiving greater risks to disclosing did not account for the significant difference between men’s T1 cortisol scores in the experimental and control conditions. Thus, future work would benefit from continuing to investigate potential explanatory mechanisms that may shed further light on the unintended consequences of the pillow talk intervention on men’s physiological reactivity.