Consistent with the findings of others, we observed widespread individual differences in the magnitude of change in circulating levels of IL-6 in response to acute psychological stress. The primary goal of the current study was to examine whether task-induced affective states contributed to this variability in inflammatory response. In our sample of relatively healthy, mid-life adults, angry and anxious responses to a simulated public speaking task were associated positively with changes in circulating levels of IL-6, as measured 30 minutes following the task. These associations were independent of age, race, BMI, menopausal status for women, and baseline levels of IL-6. In regard to gender, anxious reactions predicted increases in IL-6 for males and females; however, associations of task-induced anger with IL-6 were observed primarily among males. In sum, our findings provide initial evidence that affective responses to acute psychological stress contribute to heightened inflammatory responses. These findings are consistent with recent evidence that negative affective states result in activation of innate inflammatory pathways. For example, negative affective responses to an anger recall interview (ARI) are associated with increased B2-integrin expression (
Greeson et al., 2009) and monocyte-stimulated proinflammatory cytokine production among insulin resistant men (
Suarez et al., 2006). Similarly, negative affective responses to the Trier Social Stress Test (TSST) have been positively associated with magnitude of inflammatory response, with perceived stress predicting increases in circulating levels of IL-1β (
Yamakawa and colleagues,2009) and fear predicting increases in IL-6 in oral fluids (
Moons et al., 2010). In contrast to the current findings,
Moons et al (2010) did not observe an association of anger in response to the TSST and inflammatory markers in oral fluids. Reasons for these discrepant findings are unclear, but may relate to the examination of oral versus circulating inflammatory markers or be a function of the timing of the assessment of affective state, with measures being taken immediately following the task in the current study and 30 minutes after the task in the
Moons et al. (2010) study.
Interestingly, although we observed that angry and anxious responses to the speech task covaried, the positive associations of these activated negative emotions with inflammatory responses following acute challenge were largely independent. Stress-induced increases in depression and fatigue, which represent less physiologically activating negative moods than anxiety and anger, were not significantly associated with IL-6 reactivity after adjusting for covariates. Changes in positive affect (i.e., calm, vigor, and well-being) were also unrelated to the magnitude of IL-6 responses. These findings raise the possibility that activated negative emotional states contribute to the increase in circulating levels of IL-6 that follows acute psychological challenge. However, it is also possible that our failure to find more significant associations of depression, calm, vigor, fatigue, and well-being with IL-6 reflects the relatively small task-related changes in these affective states, when compared with anxiety and anger, limiting our ability to detect relationships.
Although there is growing evidence that circulating levels of IL-6 typically increase following acute psychological stress (see
Steptoe et al 2007 for review), mean effect sizes are modest (
Steptoe et al., 2007), with a number of studies finding no significant IL-6 response to laboratory stress (e.g.,
Miller et al., 2005;
Lutgendorf et al., 2004). The current findings are generally consistent with this literature and show a trend toward a mean increase in IL-6 from pre- to 30 minutes post-task. Our findings extend this literature by providing the first evidence to show that individual differences in affective response to the task are associated with the magnitude of this inflammatory response in healthy mid-life adults. We also provide initial evidence to suggest that there may be gender differences in the impact of acute affective states on systemic inflammation, with both sexes showing similar inflammatory responses to anxiety, but males showing greater responses to anger than females. Thus, inconsistencies across studies may reflect gender differences and/or heterogeneity of affective responses, with results varying as a function of the type or intensity of emotions induced by the laboratory task. It is also likely that neurological (
Slavich et al., 2010) or genetic (
Cole et al., 2010) vulnerabilities contribute to individual differences in affective and/or inflammatory responses. The clinical significance of the increases in systemic inflammation that follow acute laboratory challenge remains to be determined. In this regard,
Brydon & Steptoe (2005) showed a positive association of magnitude of IL-6 response to acute laboratory stress with increases in blood pressure over a 3 year follow-up period; an effect that was independent of initial blood pressure, baseline IL-6, age, BMI, gender and smoking status. This raises the possibility that individuals prone to stress-induced increases in inflammation, possibly by virtue of their affective responses to environmental challenges, may be at greater risk for cardiovascular disease.
Several potential mechanisms may underlie associations of activated negative emotions and inflammatory responses to acute stress. Based on existing literature, we expected that negative emotional states would activate the sympathetic division of the autonomic nervous system, resulting in increases in circulating levels of IL-6 as the consequence of (1) activation of peripheral blood mononuclear cells (
Bierhaus et al., 2003), (2) the release of IL-6 from adipocytes (
Burysek & Houstek, 1997), and (3) pro-coagulant/proinflammatory responses elicited from damaged vascular walls under sheer stress (
Moshage, 1997;
Thrall et al., 2007;
von Kanel et al., 2001). However, in contrast to the findings of others (
Owen & Steptoe, 2003), our analyses did not provide clear evidence for an association of cardiovascular responses to the stress task, an indirect measure of autonomic activity, with the magnitude of inflammatory response. Although anxious responses to the speech stressor were associated with concomitant increases in SBP and HR, this cardiovascular reactivity did not account for variability in IL-6 response associated with anxious mood. Moreover, there was no association of task-induced anger with cardiovascular reactivity in the current sample. Thus, our findings do not provide evidence that activation of the sympathetic nervous system drives the increase in peripheral inflammation associated with activated negative mood states. That said, measures of HR and BP provide an indirect measure of sympathetic activation, and it remains possible that a more direct measure of autonomic activation, such as circulating catecholamine levels, could yield a different pattern of results.
Given that stress-related increases in markers of peripheral inflammation are typically not detected until at least 30 minutes following the stressor, it is also possible that the kinetics of the autonomic response to stress are more influential than the peak stress response. For example, individuals who take longer to recover to baseline levels after a task may be at greater risk for inflammatory responses. In support of this possibility, delayed blood pressure recovery following acute stress has been positively associated with the magnitude of inflammatory response (
Steptoe & Marmot, 2006). This delay in blood pressure recovery may be the result of prolonged sympathetic activity, decreased parasympathetic activity, or greater peripheral damage as a consequence of prolonged blood pressure. In regard to the parasympathetic division of the autonomic nervous system, lower heart rate variability, an indirect measure of parasympathetic (vagal) control over variations in heart rate, accompanies both negative emotional states (
Brosschot & Thayer, 1998;
Friedman & Thayer, 1998) and inflammation (
Frasure-Smith et al., 2009;
Marsland et al., 2007). Thus, it is possible that stress-induced decreases in parasympathetic activation may contribute to associations of negative emotional states with markers of inflammation. Finally, activation of the hypothalamic-pituitary adrenal (HPA) axis may also play a role, as cortisol shows a delayed response to acute psychological stress that parallels the timing of the inflammatory response. In this regard, it has been proposed that individuals who are emotionally sensitive to psychological stressors may exhibit blunted cortisol responses to stress and thus increased activation of inflammatory processes (
Burke et al., 2005).
Several limitations of the current research should be noted. First, the sample was healthy, nonsmoking, and free of medications affecting immune, nervous, and endocrine systems; thus, results may not generalize to clinical or at-risk samples. Second, analyses were based on affect and IL-6 reactivity from only one laboratory session -- repeated assessments across time and situations would provide a more reliable measure of individual differences. In addition, our protocol did not include assessment of changes in IL-6 later than 30 minutes after the task. Given that increased inflammatory activity have been observed up to 2 hours after acute stress, future work would benefit from exploring whether affective response predicts more sustained elevations in IL-6 (
Steptoe et al., 2007). Finally, although relatively small increases in circulating markers of inflammation (i.e. CRP, IL-6) have been shown to predict increased disease risk (
Libby & Ridker, 2004;
Libby et al., 2002;
Ward et al., 2009), the clinical significance of the small, transient, stress-induced increase in IL-6 remains to be determined.
In conclusion, the present findings suggest that individual differences in affective responses to a mildly stressful speaking task are associated with the magnitude of change in circulating IL-6 levels, a marker of systemic inflammatory activity. Gender may play a role in this relationship, with males showing greater inflammatory responses to acute anger than females, but both genders showing similar inflammatory responses to anxiety. Otherwise, positive associations of negative affective states with IL-6 were independent of demographic characteristics and cardiovascular reactivity. These results suggest that activated negative affect in response to acute stress is associated with stress-related increases in inflammation. Long-term or repeat activation of inflammatory pathways as a consequence of heightened negative affective responses to stressful situations in daily life may contribute to risk for inflammatory diseases.
The publisher's final edited version of this article is available at 
