Shortened sleep fuels inflammatory responses to marital conflict: Emotion regulation matters

Abstract

Sleep problems can boost inflammation and may jeopardize interpersonal functioning, risks that may be magnified in couples. This observational study examined the effects of self-reported recent sleep duration on couples' inflammation, inflammatory responses to a problem discussion, interpersonal behavior, and use of emotion regulation strategies (emotion expression, cognitive reappraisal) during conflict. People who slept less had higher stimulated interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNF-α) production after the marital problem discussion than those who slept more. However, using emotion expression and cognitive reappraisal strategies during conflict protected couples who slept less from inflammatory reactivity. Specifically, people's short sleep did not relate to inflammatory increases when they expressed their own feelings more or when their partner reappraised or expressed their emotions more. When both partners slept less, couples interacted in a more hostile way than when at least one partner slept more. These data point to the combination of short sleep and marital conflict as a novel path to heightened inflammation, a risk that partners' emotion regulation strategies may counteract. The study also highlights the role of short sleep in more negative or punishing marital behavior.

Keywords: Couples; Emotion regulation; Inflammation; Inflammatory response; Marital conflict; Sleep duration.

Figure 1

One’s own shorter sleep duration in hours predicting percent of positive behavior displayed during the marital problem discussion, conditional on the partner’s sleep duration. The simple slope of sleep and positive behavior for those whose partner slept 7.5 hours (solid line) was not significant (p = .12). The sleep-behavior simple slope for those whose partner slept 5.5 hours (dashed line) was significant (p = .02).

Figure 2

One’s own shorter sleep duration in hours predicting log-transformed LPS-stimulated IL-6 (Panel A) and TNF-α (Panel B) after the marital problem discussion, conditional on one’s own emotion expression during discussion. Panel A shows a significant interaction for IL-6 (p = .03). The simple slope of shorter sleep and IL-6 for those who expressed 1 SD more than average (solid line) was not significant (p = .52). The sleep-inflammation simple slope for those who expressed 1 SD less than average (dashed line) was significant (p = .001). Like IL-6 patterns in Panel A, Panel B shows a nonsignificant simple slope between sleep and TNF-α at high emotion expression (solid line, p = .41). The sleep-inflammation tie was significant at low levels of expression (p = .005), but the difference in slopes was not statistically significant (p = .09).

Figure 3

One’s own shorter sleep duration in hours predicting log-transformed LPS-stimulated IL-6 after the marital problem discussion, conditional on one’s partner’s reappraisal during discussion. The simple slope of shorter sleep and IL-6 for those whose partners reappraised 1 SD more than average (solid line) was not significant (p = .98). The sleep-inflammation simple slope for those whose partners reappraised 1 SD less than average (dashed line) was significant (p = .002).