Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Acute stress impairs cognitive flexibility in men, not women

Acute stress impairs cognitive flexibility in men, not women STRESS, 2016 VOL. 19, NO. 5, 542–546 http://dx.doi.org/10.1080/10253890.2016.1192603 SHORT COMMUNICATION Grant S. Shields , Brian C. Trainor, Jovian C. W. Lam and Andrew P. Yonelinas Department of Psychology, University of California, Davis, CA, USA ABSTRACT ARTICLE HISTORY Psychosocial stress influences cognitive abilities, such as long-term memory retrieval. However, less is Received 29 March 2016 Revised 4 May 2016 known about the effects of stress on cognitive flexibility, which is mediated by different neurobiological Accepted 18 May 2016 circuits and could thus be regulated by different neuroendocrine pathways. In this study, we randomly Published online 14 June assigned healthy adults to an acute stress induction or control condition and subsequently assessed participants’ cognitive flexibility using an open-source version of the Wisconsin Card Sort task. Drawing on work in rodents, we hypothesized that stress would have stronger impairing effects on cognitive KEYWORDS flexibility in men than women. As predicted, we found that stress impaired cognitive flexibility in men Cognitive flexibility; men; but did not significantly affect women. Our results thus indicate that stress exerts sex-specific effects on orbitofrontal cortex; sex cognitive flexibility in humans and add to the growing body of research highlighting the need to con- differences; stress; Wisconsin card sorting test; women sider sex differences in effects of stress. Introduction subsequently assessed participants’ cognitive flexibility using an open-source version of the Wisconsin Card Sorting Task. To Psychosocial stress has important effects on many cognitive confirm a stress induction, we assessed both negative affect processes (Allen et al., 2014; Shields et al., 2016a). Indeed, stress and salivary cortisol at both baseline and 15 min post-manipu- typically impairs in long-term memory retrieval and working lation. Consistent with prior research (Alexander et al., 2007), memory (Schoofs et al., 2013; Shields et al., 2016b) but enhan- we hypothesized that stress would impair cognitive flexibility ces other cognitive abilities, such as decision-making compe- and – based upon work with rodents (Laredo et al., 2015)– tence (Shields et al., 2016a; though see Starcke & Brand, 2012) that this effect would be stronger in men than women. and response inhibition (Schwabe et al., 2013). Understanding how stress influences higher cognitive abilities has important implications for solving problems that arise in daily life Methods (Diamond, 2013). Despite the well-known effects of stress on Participants many cognitive processes and the importance of these effects, it is still unclear how stress might impact many important cog- Participants were 113 healthy young adults attending the nitive processes, such as cognitive flexibility. University of California, Davis. We did not invite participants Prior research with both humans (Alexander et al., 2007; who had a current illness, diabetes, history of stroke, neuro- Plessow et al., 2011) and rodents (Laredo et al., 2015) has sug- logical disorders, current or former diagnosis of posttraumatic gested that stress may impair cognitive flexibility – that is, the stress disorder, hospitalization for a psychiatric disorder within ability to flexibly switch between thoughts or rules in a goal- the past year, current injury or illness within the past week, directed manner. However, when studying effects of stress, it is major sleep disturbances within the past six weeks, or con- important to examine potential sex differences, given the dra- sumption of more than eight caffeinated beverages a day. matically different effects stress can have on males and females Similarly, individuals who were pregnant, nursing, on any (Trainor et al., 2013); stress should not be presumed to influence form of medication (including hormonal birth control or males and females similarly (Cahill, 2012; Felmingham et al., asthma medication) or illegal drugs, had taken any mood- 2012; Schoofs et al., 2013; Zoladz et al., 2013). Indeed, work with altering medications within the past two months, or had rodents has suggested that the impairing effects of stress on taken corticosteroids within the past three months were not cognitive flexibility may be stronger in males than females invited to participate. Participants were instructed not to eat, (Laredo et al., 2015), but sex differences have not been exam- drink, use tobacco, brush their teeth or floss, or engage in ined in human studies of stress and cognitive flexibility. Based exercise for 2 h prior to the start of the study. Compliance on these findings, we tested whether sex moderates the effects with these instructions and inclusion criteria was assessed of stress on cognitive flexibility in human subjects. using a questionnaire at the beginning of the study; women To this end, we randomly assigned healthy adult men and also reported the date of the first day of their last menstrual women to either an acute stress induction or control task and period using that questionnaire. Menstrual cycle phase was CONTACT Grant S. Shields gsshields@ucdavis.edu Department of Psychology, University of California, Davis, CA 95616, USA 2016 Informa UK Limited, trading as Taylor & Francis Group STRESS 543 approximated by days since preceding cycle had begun (i.e. 5 min. Participants then completed the post-stressor affect five or less days: menstrual period; 6–13 d: follicular phase; measure, which took approximately 40 s. Participants then 14þ d: luteal phase). completed the cognitive flexibility task. We measured cogni- Fifty-six individuals (36 women, 20 men) were randomly tive flexibility using the Berg Card Sorting Test, which is an assigned to the stress induction condition and 57 individuals open-source version of the Wisconsin Card Sorting Test (39 women, 18 men) were randomized to the non-stressful (Mueller & Piper, 2014). This task is a well-validated task control condition. Participants ranged in age from 18 to 54 requiring cognitive flexibility (Miyake et al., 2000). The primary years old (M¼ 20.20, SD¼ 3.8). 16.7% of women were tested outcome in this task is the number of perseverative errors a during their menstrual period, 28.3% during the follicular person makes, which indicates a continued application of a card-sorting rule that is no longer appropriate instead of shift- phase and 55% during the luteal phase of their menstrual ing to the use of a new rule (i.e. cognitive inflexibility); higher cycle. scores thus indicate worse performance. The card-sorting test took on average approximately 6 min Materials and procedure to complete. Participants then waited quietly until 15 min total Participants came to the laboratory at either 12 p.m. or 3 p.m. had elapsed since the offset of the stress or control manipula- for three- or four-participant group sessions. Upon arrival, an tion and then provided a second saliva sample. Finally, partici- experimenter immediately greeted each participant and pants completed the demographics questionnaire before brought the participant into a cubicle in order to prevent the being debriefed, thanked and dismissed. participants from interacting with each other. Once in the cubicle, each participant provided informed consent and com- Cortisol pleted miscellaneous measures, including a measure of base- Participants provided two saliva samples (baseline and post- line affect, for approximately 10 min to allow acclimation to manipulation) using a passive drool method. Immediately the testing environment. Prior to learning of the stressor task after collection, the saliva vials were placed in a 20 C and immediately after the task, participants used an freezer until assayed in duplicate using high-sensitivity unmarked scale, ranging from 1(Not at All)to 7(Very Much), to Salivary Cortisol ELISA Kits (Salimetrics LLC, State College, PA) indicate the extent they currently felt a variety of negative according to manufacturer instructions. The inter-assay CV affective states, which were then averaged to create a nega- was 7.45% and the average intra-assay CV was 2.82%. tive affect composite. Negative affect was assessed at baseline Sensitivity for these assays was 0.012 lg/dL. All controls were (a¼ 0.90) and after the stress manipulation (a¼ 0.92). in the expected ranges. Cortisol concentrations were con- Participants’ computers then reached a password-protected verted from lg/dL to nmol/L for consistency with most screen that instructed them to wait for instructions from the human stress literature. experimenters. Participants waited until all other participants for the session completed the initial measures, upon which Data reduction and analysis time the first (baseline) saliva sample was taken. Next, participants completed the laboratory-based stressor All variables were inspected for conformity to a normal distri- or control task, depending upon their time slot’s assigned bution and the natural logarithm transformation was applied condition. An experience of acute stress was induced using when variables evidenced significant skew (i.e. cortisol, both the Trier Social Stress Test for Groups (TSST-G; von Dawans baseline and post-manipulation and perseverative errors). et al., 2011). This task includes two conditions: a stress induc- Because the acute stress manipulation necessitated random- tion condition and a non-stressful control condition. In brief, ization of participant sessions to conditions (i.e. rather than participants in the stress induction condition were conspicu- participants), analyses required a multilevel model to account ously recorded while they spoke on their real qualifications for shared variability within sessions. Thus, all analyses were for the job they would like to have in front of a live panel of linear mixed models with participants nested within Session. two trained, stern evaluators and afterwards were evaluated We used a mixed model ANOVA nesting measurement occa- as they completed a difficult math task wherein each partici- sions within participants and further nesting participants pant counted backwards in steps of 16 starting from a large within Session to assess changes in cortisol from baseline to four-digit number (e.g. 3329) that differed for each participant post-manipulation. All analyses were conducted in R, version in the session. During the math task, the participant was peri- 3.2.1. Mixed models were fit using the lmerTest package. odically instructed to count faster, and if the participant made Least-squares means and their corresponding standard errors a mistake, the participant was told his/her answer was incor- were derived using the lsmeans package. rect and s/he had to restart. In contrast, participants in the control condition quietly read aloud a scientific article and Results subsequently completed a simple forward counting task with- out any social evaluation. Preliminary analyses The TSST-G lasted approximately 30 min (including antici- pation), after which time participants returned to their com- Participants in the stress induction versus control conditions puters and waited for the experimenters to enter a password did not differ with respect to menstrual cycle phase, age, race to allow the participant to continue. This transition from the or sex (p> 0.15, uncorrected). Thus, our random assignment stressor to the participants’ computers took approximately was successful. 544 G. S. SHIELDS ET AL. We examined changes in negative affect from baseline to errors than participants in the control group (M¼ 1.85, post-manipulation to confirm the success of both our stress SE¼ 0.10), the effect of condition was not significant, F (1, manipulation and our control condition. As expected, the time 113.0)¼ 1.82, p¼ 0.180. However, as hypothesized, the condition interaction was significant, F (1, 111.0)¼ sex condition interaction was significant, F (1, 113.0)¼ 4.32, 11.22, p¼ 0.001. Participants in the stress induction group sig- p¼ 0.040, g ¼ 0.037, indicating stress differentially influ- partial nificantly increased in negative affect from pre- to post- enced cognitive flexibility by sex. As shown in Figure 2, men manipulation, t(111.0)¼2.66, p¼ 0.009, whereas participants in the stress condition (M¼ 2.19, SE¼ 0.15) committed signifi- in the control group significantly decreased in negative affect cantly more perseverative errors than men in the control from pre- to post-manipulation, t(111.0)¼2.08, p¼ 0.040. We did not find evidence for a sex time condition interaction, F (1, 109.0)¼ 2.15, p¼ 0.146. We next examined cortisol reactivity over time for the stress induction group and a randomly selected 10 participants in the control group. As expected, the time condition interaction was significant, F (1, 58.6)¼ 12.66, p< 0.001 (Figure 1). Participants in the stress induction group signifi- cantly increased from pre- to post-manipulation, t(61.0)¼ 4.89, p< 0.001, whereas participants in the control group tended to decrease from pre- to post-manipulation, t(60.1)¼1.84, p¼ 0.071. In addition, post-manipulation cortisol was signifi- cantly greater in the stress-induction group than the control group, t(47.1)¼ 2.86, p¼ 0.006. We did not find evidence for a sex time condition interaction, p¼ 0.411, but because we only had 10 participants with cortisol assayed in the control group, we had very little power to determine a sex by condi- tion effect. Primary analyses We next examined what effect, if any, acute stress had on cognitive flexibility using a mixed-model ANOVA with sex and condition as between-subjects factors. There was no main effect of sex, indicating that men and women Figure 2. Sex-specific effects of acute stress on cognitive flexibility. Men in the committed the same amount of perseverative errors overall stress induction group committed significantly more perseverative errors (log transformed) than men in the control group, p¼ 0.044, indicating that stress F (1, 113.0)¼ 0.69, p¼ 0.794. Similarly, although there was a impaired cognitive flexibility in men. In contrast, women in the stress induction descriptive tendency for participants in the stress induction condition did not commit a different number of perseverative errors than women group (M¼ 2.03, SE¼ 0.10) to commit more perseverative in the control group, p¼ 0.543. Figure 1. Effect of the stress manipulation on cortisol. Values are presented in nmol/L for comparison with other studies, but analyses were conducted using log-trans- formed cortisol values to correct for skew. As expected, cortisol increased from pre- to post-manipulation in the stress induction group, p< 0.001, whereas cortisol tended to decrease in the randomly-selected 10 participants in control group we assayed, p¼ 0.071. There was no evidence of sex differences in reactivity to the stressor. STRESS 545 condition (M¼ 1.73, SE¼ 0.16), t(95.5)¼ 2.04, p¼ 0.044, participants in the control group weakened our neuroendo- d¼ 0.68; however, women in the stress condition (M¼ 1.89, crine conclusions, as it limited our ability to detect a stress by SE¼ 0.11) did not commit more perseverative errors than sex by time effect on cortisol. Fourth, our panels of evaluators women in the control condition (M¼ 1.98, SE¼ 0.11), were not always mixed-sex panels, and the sex of the evalua- t(51.9)¼0.61, p¼ 0.543, d¼0.15. An additional analysis tors may have influenced stress responses; however, our use of determined that there was no significant moderating effect of a multi-level model that nested participants within sessions menstrual cycle phase on cognitive flexibility,F (2, 60)¼ 1.85, partially controls for this limitation, as this model accounts for similarities between participants in sessions (e.g. being eval- p¼ 0.167. Across all participants, changes in cortisol (post- uated by the same panel). Fifth, we did not obtain cardiovascu- manipulation minus baseline) did not predict perseverative lar markers or other biological indices of a stress response. errors, p¼ 0.781; changes in cortisol did not predict persevera- Sixth, sampling cortisol 15 min after our stressor finished tive errors when restricting analyses to only males, p¼ 0.205 (which lasted 30 min) may have missed the cortisol peak, which or only females, p¼ 0.844. Thus, acute stress impaired cogni- typically occurs 20–30 min after stressor onset. Finally, our tive flexibility in men, but not women. stress manipulation was an acute stressor, and it is unknown whether chronic stress would produce the same sex-specific Discussion effects on cognitive flexibility as we observed in response to Although stress is known to influence cognitive processes acute stress. such as memory, the influence of stress on cognitive flexibility has received comparatively little attention. In addition, prior Conclusion research in humans has not examined potential sex differen- ces in stress effects on cognitive flexibility. Consistent with In conclusion, consistent with prior work with rodents, we work in rodents, we found that acute stress impaired cogni- found that stress impaired cognitive flexibility in men but not tive flexibility in men but not women. Thus, sex appears to be women. These results add to a growing literature that cognitive flexibility is more sensitive to stress or other insults in men an important factor for understanding stress effects on cogni- compared to women. Our research thus adds to the growing tive flexibility in humans. body of work highlighting the need to investigate sex differen- Prior studies examining stress effects on cognitive flexibility ces in effects of stress. In addition, our results suggest that have found that stress impaired cognitive flexibility in a num- women may be more suited to high-stress environments ber of ways – such as solving fewer anagrams, engaging in requiring cognitive flexibility than men. Future research should less creative thinking and reducing trial-by-trial modulation of be careful to consider sex as an important factor in stress behavior in a Simon task – in samples composed of 50% men effects on cognitive flexibility or any potential effect of stress. and 50% women (Alexander et al., 2007; Plessow et al., 2011). Our results suggest that these previous studies of stress and cognitive flexibility may have found larger effects if their anal- Acknowledgements yses were restricted to males. The authors wish to thank numerous research assistants for assistance The biological mechanism(s) behind our observed effects with data collection. are currently unclear. Although differentially influenced by stress in men and women (Shields et al., 2016b), cortisol does Disclosure statement not appear to be playing a role, as it was not related to cogni- The authors declare no conflict of interest with respect to this work. tive flexibility in our data or in a recent meta-analysis (Shields et al., 2015). Work in mice has identified the l-opioid receptor Funding information as an important mechanism mediating stress’s sex-specific effects on cognitive flexibility (Laredo et al., 2015), indicating This research was supported by a University of California, Davis Psychology that the l-opioid system may contribute to our observed Department Summer Grant in Aid of Research to Grant S. Shields, NIH MH103322 to Brian C. Trainor, a University of California, Davis Provost’s effects. Alternatively, sex differences have been observed in Undergraduate Fellowship to Jovian C. W. Lam, and NIH MH059352 and noradrenergic activity following stress (Bangasser & Valentino, EY025999 to Andrew Yonelinas. These organizations had no role in design- 2014), and noradrenergic activity is necessary for producing ing the study; in collecting, analyzing, or interpreting the data; in writing stress effects on cognitive flexibility (Alexander et al., 2007). this report; or in deciding to submit this report for publication. Similarly, sex differences in dopaminergic activity exist, and dopamine is important for executive functioning (Shansky & ORCID Lipps, 2013). Thus, future work should attempt to elucidate Grant S. Shields http://orcid.org/0000-0002-0827-4669 the biological mechanism(s) underpinning sex-specific stress effects on cognitive flexibility in humans. This study has several limitations worth noting. First, References because we used a sample of undergraduate students, it is Alexander JK, Hillier A, Smith RM, Tivarus ME, Beversdorf DQ. (2007). Beta- unknown if our results would generalize to a nonstudent popu- adrenergic modulation of cognitive flexibility during stress. J Cogn lation. Second, the cognitive flexibility task we used is accur- Neurosci 19:468–78. acy-based, and it is unknown whether our results generalize to Allen AP, Kennedy PJ, Cryan JF, Dinan TG, Clarke G. (2014). Biological and other tasks requiring cognitive flexibility, such as tasks with psychological markers of stress in humans: focus on the Trier Social shift costs as an outcome. Third, assaying cortisol for only ten Stress Test. Neurosci Biobehav Rev 38:94–124. 546 G. S. SHIELDS ET AL. Bangasser DA, Valentino RJ. (2014). Sex differences in stress-related psy- Shansky RM, Lipps J. (2013). Stress-induced cognitive dysfunction: hor- chiatric disorders: neurobiological perspectives. Front Neuroendocrinol mone-neurotransmitter interactions in the prefrontal cortex. Front Hum 35:303–19. Neurosci 7:123. Cahill L. (2012). A half-truth is a whole lie: on the necessity of investigat- Shields GS, Bonner JC, Moons WG. (2015). Does cortisol influence core ing sex influences on the brain. Endocrinology 153:2541–3. executive functions? A meta-analysis of acute cortisol administration Diamond A. (2013). Executive functions. Annu Rev Psychol 64:135–68. effects on working memory, inhibition, and set-shifting. Felmingham KL, Tran TP, Fong WC, Bryant RA. (2012). Sex differences in Psychoneuroendocrinology 58:91–103. emotional memory consolidation: the effect of stress-induced salivary Shields GS, Lam JCW, Trainor BC, Yonelinas AP. (2016a). Exposure to acute alpha-amylase and cortisol. Biol Psychol 89:539–44. stress enhances decision-making competence: evidence for the role of Laredo SA, Steinman MQ, Robles CF, Ferrer E, Ragen BJ, Trainor BC. (2015). DHEA. Psychoneuroendocrinology 67:51–60. Effects of defeat stress on behavioral flexibility in males and females: Shields GS, Sazma MA, McCullough AM, Yonelinas AP. (2016b). The modulation by the mu-opioid receptor. Eur J Neurosci 41:434–41. effects of acute stress on memory: a meta-analysis and integrative Miyake A, Friedman NP, Emerson MJ, Witzki AH, Howerter A, Wager TD. review. (2000). The unity and diversity of executive functions and their contri- Starcke K, Brand M. (2012). Decision making under stress: a selective butions to complex ‘‘Frontal Lobe’’ tasks: a latent variable analysis. review. Neurosci Biobehav Rev 36:1228–48. Cogn Psychol 41:49–100. Trainor BC, Takahashi EY, Campi KL, Florez SA, Greenberg GD, Laman- Mueller ST, Piper BJ. (2014). The Psychology Experiment Building Maharg A, Laredo SA, et al. (2013). Sex differences in stress-induced Language (PEBL) and PEBL test battery. J Neurosci Methods 222:250–9. social withdrawal: independence from adult gonadal hormones and Plessow F, Fischer R, Kirschbaum C, Goschke T. (2011). Inflexibly focused inhibition of female phenotype by corncob bedding. Horm Behav under stress: acute psychosocial stress increases shielding of action 63:543–50. goals at the expense of reduced cognitive flexibility with increasing von Dawans B, Kirschbaum C, Heinrichs M. (2011). The Trier Social time lag to the stressor. J Cogn Neurosci 23:3218–27. Stress Test for Groups (TSST-G): a new research tool for controlled Schoofs D, Pabst S, Brand M, Wolf OT. (2013). Working memory is differen- simultaneous social stress exposure in a group format. tially affected by stress in men and women. Behav Brain Res Psychoneuroendocrinology 36:514–22. 241:144–53. Zoladz PR, Warnecke AJ, Woelke SA, Burke HM, Frigo RM, Pisansky JM, Schwabe L, Hoffken € O, Tegenthoff M, Wolf OT. (2013). Stress-induced Lyle SM, Talbot JN. (2013). Pre-learning stress that is temporally enhancement of response inhibition depends on mineralocorticoid removed from acquisition exerts sex-specific effects on long-term receptor activation. Psychoneuroendocrinology 38:2319–26. memory. Neurobiol Learn Mem 100:77–87. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Stress: The International Journal on the Biology of Stress Taylor & Francis

Acute stress impairs cognitive flexibility in men, not women

Loading next page...
 
/lp/taylor-francis/acute-stress-impairs-cognitive-flexibility-in-men-not-women-M5n0kg86rE

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Taylor & Francis
Copyright
© 2016 Informa UK Limited, trading as Taylor & Francis Group
ISSN
1607-8888
eISSN
1025-3890
DOI
10.1080/10253890.2016.1192603
Publisher site
See Article on Publisher Site

Abstract

STRESS, 2016 VOL. 19, NO. 5, 542–546 http://dx.doi.org/10.1080/10253890.2016.1192603 SHORT COMMUNICATION Grant S. Shields , Brian C. Trainor, Jovian C. W. Lam and Andrew P. Yonelinas Department of Psychology, University of California, Davis, CA, USA ABSTRACT ARTICLE HISTORY Psychosocial stress influences cognitive abilities, such as long-term memory retrieval. However, less is Received 29 March 2016 Revised 4 May 2016 known about the effects of stress on cognitive flexibility, which is mediated by different neurobiological Accepted 18 May 2016 circuits and could thus be regulated by different neuroendocrine pathways. In this study, we randomly Published online 14 June assigned healthy adults to an acute stress induction or control condition and subsequently assessed participants’ cognitive flexibility using an open-source version of the Wisconsin Card Sort task. Drawing on work in rodents, we hypothesized that stress would have stronger impairing effects on cognitive KEYWORDS flexibility in men than women. As predicted, we found that stress impaired cognitive flexibility in men Cognitive flexibility; men; but did not significantly affect women. Our results thus indicate that stress exerts sex-specific effects on orbitofrontal cortex; sex cognitive flexibility in humans and add to the growing body of research highlighting the need to con- differences; stress; Wisconsin card sorting test; women sider sex differences in effects of stress. Introduction subsequently assessed participants’ cognitive flexibility using an open-source version of the Wisconsin Card Sorting Task. To Psychosocial stress has important effects on many cognitive confirm a stress induction, we assessed both negative affect processes (Allen et al., 2014; Shields et al., 2016a). Indeed, stress and salivary cortisol at both baseline and 15 min post-manipu- typically impairs in long-term memory retrieval and working lation. Consistent with prior research (Alexander et al., 2007), memory (Schoofs et al., 2013; Shields et al., 2016b) but enhan- we hypothesized that stress would impair cognitive flexibility ces other cognitive abilities, such as decision-making compe- and – based upon work with rodents (Laredo et al., 2015)– tence (Shields et al., 2016a; though see Starcke & Brand, 2012) that this effect would be stronger in men than women. and response inhibition (Schwabe et al., 2013). Understanding how stress influences higher cognitive abilities has important implications for solving problems that arise in daily life Methods (Diamond, 2013). Despite the well-known effects of stress on Participants many cognitive processes and the importance of these effects, it is still unclear how stress might impact many important cog- Participants were 113 healthy young adults attending the nitive processes, such as cognitive flexibility. University of California, Davis. We did not invite participants Prior research with both humans (Alexander et al., 2007; who had a current illness, diabetes, history of stroke, neuro- Plessow et al., 2011) and rodents (Laredo et al., 2015) has sug- logical disorders, current or former diagnosis of posttraumatic gested that stress may impair cognitive flexibility – that is, the stress disorder, hospitalization for a psychiatric disorder within ability to flexibly switch between thoughts or rules in a goal- the past year, current injury or illness within the past week, directed manner. However, when studying effects of stress, it is major sleep disturbances within the past six weeks, or con- important to examine potential sex differences, given the dra- sumption of more than eight caffeinated beverages a day. matically different effects stress can have on males and females Similarly, individuals who were pregnant, nursing, on any (Trainor et al., 2013); stress should not be presumed to influence form of medication (including hormonal birth control or males and females similarly (Cahill, 2012; Felmingham et al., asthma medication) or illegal drugs, had taken any mood- 2012; Schoofs et al., 2013; Zoladz et al., 2013). Indeed, work with altering medications within the past two months, or had rodents has suggested that the impairing effects of stress on taken corticosteroids within the past three months were not cognitive flexibility may be stronger in males than females invited to participate. Participants were instructed not to eat, (Laredo et al., 2015), but sex differences have not been exam- drink, use tobacco, brush their teeth or floss, or engage in ined in human studies of stress and cognitive flexibility. Based exercise for 2 h prior to the start of the study. Compliance on these findings, we tested whether sex moderates the effects with these instructions and inclusion criteria was assessed of stress on cognitive flexibility in human subjects. using a questionnaire at the beginning of the study; women To this end, we randomly assigned healthy adult men and also reported the date of the first day of their last menstrual women to either an acute stress induction or control task and period using that questionnaire. Menstrual cycle phase was CONTACT Grant S. Shields gsshields@ucdavis.edu Department of Psychology, University of California, Davis, CA 95616, USA 2016 Informa UK Limited, trading as Taylor & Francis Group STRESS 543 approximated by days since preceding cycle had begun (i.e. 5 min. Participants then completed the post-stressor affect five or less days: menstrual period; 6–13 d: follicular phase; measure, which took approximately 40 s. Participants then 14þ d: luteal phase). completed the cognitive flexibility task. We measured cogni- Fifty-six individuals (36 women, 20 men) were randomly tive flexibility using the Berg Card Sorting Test, which is an assigned to the stress induction condition and 57 individuals open-source version of the Wisconsin Card Sorting Test (39 women, 18 men) were randomized to the non-stressful (Mueller & Piper, 2014). This task is a well-validated task control condition. Participants ranged in age from 18 to 54 requiring cognitive flexibility (Miyake et al., 2000). The primary years old (M¼ 20.20, SD¼ 3.8). 16.7% of women were tested outcome in this task is the number of perseverative errors a during their menstrual period, 28.3% during the follicular person makes, which indicates a continued application of a card-sorting rule that is no longer appropriate instead of shift- phase and 55% during the luteal phase of their menstrual ing to the use of a new rule (i.e. cognitive inflexibility); higher cycle. scores thus indicate worse performance. The card-sorting test took on average approximately 6 min Materials and procedure to complete. Participants then waited quietly until 15 min total Participants came to the laboratory at either 12 p.m. or 3 p.m. had elapsed since the offset of the stress or control manipula- for three- or four-participant group sessions. Upon arrival, an tion and then provided a second saliva sample. Finally, partici- experimenter immediately greeted each participant and pants completed the demographics questionnaire before brought the participant into a cubicle in order to prevent the being debriefed, thanked and dismissed. participants from interacting with each other. Once in the cubicle, each participant provided informed consent and com- Cortisol pleted miscellaneous measures, including a measure of base- Participants provided two saliva samples (baseline and post- line affect, for approximately 10 min to allow acclimation to manipulation) using a passive drool method. Immediately the testing environment. Prior to learning of the stressor task after collection, the saliva vials were placed in a 20 C and immediately after the task, participants used an freezer until assayed in duplicate using high-sensitivity unmarked scale, ranging from 1(Not at All)to 7(Very Much), to Salivary Cortisol ELISA Kits (Salimetrics LLC, State College, PA) indicate the extent they currently felt a variety of negative according to manufacturer instructions. The inter-assay CV affective states, which were then averaged to create a nega- was 7.45% and the average intra-assay CV was 2.82%. tive affect composite. Negative affect was assessed at baseline Sensitivity for these assays was 0.012 lg/dL. All controls were (a¼ 0.90) and after the stress manipulation (a¼ 0.92). in the expected ranges. Cortisol concentrations were con- Participants’ computers then reached a password-protected verted from lg/dL to nmol/L for consistency with most screen that instructed them to wait for instructions from the human stress literature. experimenters. Participants waited until all other participants for the session completed the initial measures, upon which Data reduction and analysis time the first (baseline) saliva sample was taken. Next, participants completed the laboratory-based stressor All variables were inspected for conformity to a normal distri- or control task, depending upon their time slot’s assigned bution and the natural logarithm transformation was applied condition. An experience of acute stress was induced using when variables evidenced significant skew (i.e. cortisol, both the Trier Social Stress Test for Groups (TSST-G; von Dawans baseline and post-manipulation and perseverative errors). et al., 2011). This task includes two conditions: a stress induc- Because the acute stress manipulation necessitated random- tion condition and a non-stressful control condition. In brief, ization of participant sessions to conditions (i.e. rather than participants in the stress induction condition were conspicu- participants), analyses required a multilevel model to account ously recorded while they spoke on their real qualifications for shared variability within sessions. Thus, all analyses were for the job they would like to have in front of a live panel of linear mixed models with participants nested within Session. two trained, stern evaluators and afterwards were evaluated We used a mixed model ANOVA nesting measurement occa- as they completed a difficult math task wherein each partici- sions within participants and further nesting participants pant counted backwards in steps of 16 starting from a large within Session to assess changes in cortisol from baseline to four-digit number (e.g. 3329) that differed for each participant post-manipulation. All analyses were conducted in R, version in the session. During the math task, the participant was peri- 3.2.1. Mixed models were fit using the lmerTest package. odically instructed to count faster, and if the participant made Least-squares means and their corresponding standard errors a mistake, the participant was told his/her answer was incor- were derived using the lsmeans package. rect and s/he had to restart. In contrast, participants in the control condition quietly read aloud a scientific article and Results subsequently completed a simple forward counting task with- out any social evaluation. Preliminary analyses The TSST-G lasted approximately 30 min (including antici- pation), after which time participants returned to their com- Participants in the stress induction versus control conditions puters and waited for the experimenters to enter a password did not differ with respect to menstrual cycle phase, age, race to allow the participant to continue. This transition from the or sex (p> 0.15, uncorrected). Thus, our random assignment stressor to the participants’ computers took approximately was successful. 544 G. S. SHIELDS ET AL. We examined changes in negative affect from baseline to errors than participants in the control group (M¼ 1.85, post-manipulation to confirm the success of both our stress SE¼ 0.10), the effect of condition was not significant, F (1, manipulation and our control condition. As expected, the time 113.0)¼ 1.82, p¼ 0.180. However, as hypothesized, the condition interaction was significant, F (1, 111.0)¼ sex condition interaction was significant, F (1, 113.0)¼ 4.32, 11.22, p¼ 0.001. Participants in the stress induction group sig- p¼ 0.040, g ¼ 0.037, indicating stress differentially influ- partial nificantly increased in negative affect from pre- to post- enced cognitive flexibility by sex. As shown in Figure 2, men manipulation, t(111.0)¼2.66, p¼ 0.009, whereas participants in the stress condition (M¼ 2.19, SE¼ 0.15) committed signifi- in the control group significantly decreased in negative affect cantly more perseverative errors than men in the control from pre- to post-manipulation, t(111.0)¼2.08, p¼ 0.040. We did not find evidence for a sex time condition interaction, F (1, 109.0)¼ 2.15, p¼ 0.146. We next examined cortisol reactivity over time for the stress induction group and a randomly selected 10 participants in the control group. As expected, the time condition interaction was significant, F (1, 58.6)¼ 12.66, p< 0.001 (Figure 1). Participants in the stress induction group signifi- cantly increased from pre- to post-manipulation, t(61.0)¼ 4.89, p< 0.001, whereas participants in the control group tended to decrease from pre- to post-manipulation, t(60.1)¼1.84, p¼ 0.071. In addition, post-manipulation cortisol was signifi- cantly greater in the stress-induction group than the control group, t(47.1)¼ 2.86, p¼ 0.006. We did not find evidence for a sex time condition interaction, p¼ 0.411, but because we only had 10 participants with cortisol assayed in the control group, we had very little power to determine a sex by condi- tion effect. Primary analyses We next examined what effect, if any, acute stress had on cognitive flexibility using a mixed-model ANOVA with sex and condition as between-subjects factors. There was no main effect of sex, indicating that men and women Figure 2. Sex-specific effects of acute stress on cognitive flexibility. Men in the committed the same amount of perseverative errors overall stress induction group committed significantly more perseverative errors (log transformed) than men in the control group, p¼ 0.044, indicating that stress F (1, 113.0)¼ 0.69, p¼ 0.794. Similarly, although there was a impaired cognitive flexibility in men. In contrast, women in the stress induction descriptive tendency for participants in the stress induction condition did not commit a different number of perseverative errors than women group (M¼ 2.03, SE¼ 0.10) to commit more perseverative in the control group, p¼ 0.543. Figure 1. Effect of the stress manipulation on cortisol. Values are presented in nmol/L for comparison with other studies, but analyses were conducted using log-trans- formed cortisol values to correct for skew. As expected, cortisol increased from pre- to post-manipulation in the stress induction group, p< 0.001, whereas cortisol tended to decrease in the randomly-selected 10 participants in control group we assayed, p¼ 0.071. There was no evidence of sex differences in reactivity to the stressor. STRESS 545 condition (M¼ 1.73, SE¼ 0.16), t(95.5)¼ 2.04, p¼ 0.044, participants in the control group weakened our neuroendo- d¼ 0.68; however, women in the stress condition (M¼ 1.89, crine conclusions, as it limited our ability to detect a stress by SE¼ 0.11) did not commit more perseverative errors than sex by time effect on cortisol. Fourth, our panels of evaluators women in the control condition (M¼ 1.98, SE¼ 0.11), were not always mixed-sex panels, and the sex of the evalua- t(51.9)¼0.61, p¼ 0.543, d¼0.15. An additional analysis tors may have influenced stress responses; however, our use of determined that there was no significant moderating effect of a multi-level model that nested participants within sessions menstrual cycle phase on cognitive flexibility,F (2, 60)¼ 1.85, partially controls for this limitation, as this model accounts for similarities between participants in sessions (e.g. being eval- p¼ 0.167. Across all participants, changes in cortisol (post- uated by the same panel). Fifth, we did not obtain cardiovascu- manipulation minus baseline) did not predict perseverative lar markers or other biological indices of a stress response. errors, p¼ 0.781; changes in cortisol did not predict persevera- Sixth, sampling cortisol 15 min after our stressor finished tive errors when restricting analyses to only males, p¼ 0.205 (which lasted 30 min) may have missed the cortisol peak, which or only females, p¼ 0.844. Thus, acute stress impaired cogni- typically occurs 20–30 min after stressor onset. Finally, our tive flexibility in men, but not women. stress manipulation was an acute stressor, and it is unknown whether chronic stress would produce the same sex-specific Discussion effects on cognitive flexibility as we observed in response to Although stress is known to influence cognitive processes acute stress. such as memory, the influence of stress on cognitive flexibility has received comparatively little attention. In addition, prior Conclusion research in humans has not examined potential sex differen- ces in stress effects on cognitive flexibility. Consistent with In conclusion, consistent with prior work with rodents, we work in rodents, we found that acute stress impaired cogni- found that stress impaired cognitive flexibility in men but not tive flexibility in men but not women. Thus, sex appears to be women. These results add to a growing literature that cognitive flexibility is more sensitive to stress or other insults in men an important factor for understanding stress effects on cogni- compared to women. Our research thus adds to the growing tive flexibility in humans. body of work highlighting the need to investigate sex differen- Prior studies examining stress effects on cognitive flexibility ces in effects of stress. In addition, our results suggest that have found that stress impaired cognitive flexibility in a num- women may be more suited to high-stress environments ber of ways – such as solving fewer anagrams, engaging in requiring cognitive flexibility than men. Future research should less creative thinking and reducing trial-by-trial modulation of be careful to consider sex as an important factor in stress behavior in a Simon task – in samples composed of 50% men effects on cognitive flexibility or any potential effect of stress. and 50% women (Alexander et al., 2007; Plessow et al., 2011). Our results suggest that these previous studies of stress and cognitive flexibility may have found larger effects if their anal- Acknowledgements yses were restricted to males. The authors wish to thank numerous research assistants for assistance The biological mechanism(s) behind our observed effects with data collection. are currently unclear. Although differentially influenced by stress in men and women (Shields et al., 2016b), cortisol does Disclosure statement not appear to be playing a role, as it was not related to cogni- The authors declare no conflict of interest with respect to this work. tive flexibility in our data or in a recent meta-analysis (Shields et al., 2015). Work in mice has identified the l-opioid receptor Funding information as an important mechanism mediating stress’s sex-specific effects on cognitive flexibility (Laredo et al., 2015), indicating This research was supported by a University of California, Davis Psychology that the l-opioid system may contribute to our observed Department Summer Grant in Aid of Research to Grant S. Shields, NIH MH103322 to Brian C. Trainor, a University of California, Davis Provost’s effects. Alternatively, sex differences have been observed in Undergraduate Fellowship to Jovian C. W. Lam, and NIH MH059352 and noradrenergic activity following stress (Bangasser & Valentino, EY025999 to Andrew Yonelinas. These organizations had no role in design- 2014), and noradrenergic activity is necessary for producing ing the study; in collecting, analyzing, or interpreting the data; in writing stress effects on cognitive flexibility (Alexander et al., 2007). this report; or in deciding to submit this report for publication. Similarly, sex differences in dopaminergic activity exist, and dopamine is important for executive functioning (Shansky & ORCID Lipps, 2013). Thus, future work should attempt to elucidate Grant S. Shields http://orcid.org/0000-0002-0827-4669 the biological mechanism(s) underpinning sex-specific stress effects on cognitive flexibility in humans. This study has several limitations worth noting. First, References because we used a sample of undergraduate students, it is Alexander JK, Hillier A, Smith RM, Tivarus ME, Beversdorf DQ. (2007). Beta- unknown if our results would generalize to a nonstudent popu- adrenergic modulation of cognitive flexibility during stress. J Cogn lation. Second, the cognitive flexibility task we used is accur- Neurosci 19:468–78. acy-based, and it is unknown whether our results generalize to Allen AP, Kennedy PJ, Cryan JF, Dinan TG, Clarke G. (2014). Biological and other tasks requiring cognitive flexibility, such as tasks with psychological markers of stress in humans: focus on the Trier Social shift costs as an outcome. Third, assaying cortisol for only ten Stress Test. Neurosci Biobehav Rev 38:94–124. 546 G. S. SHIELDS ET AL. Bangasser DA, Valentino RJ. (2014). Sex differences in stress-related psy- Shansky RM, Lipps J. (2013). Stress-induced cognitive dysfunction: hor- chiatric disorders: neurobiological perspectives. Front Neuroendocrinol mone-neurotransmitter interactions in the prefrontal cortex. Front Hum 35:303–19. Neurosci 7:123. Cahill L. (2012). A half-truth is a whole lie: on the necessity of investigat- Shields GS, Bonner JC, Moons WG. (2015). Does cortisol influence core ing sex influences on the brain. Endocrinology 153:2541–3. executive functions? A meta-analysis of acute cortisol administration Diamond A. (2013). Executive functions. Annu Rev Psychol 64:135–68. effects on working memory, inhibition, and set-shifting. Felmingham KL, Tran TP, Fong WC, Bryant RA. (2012). Sex differences in Psychoneuroendocrinology 58:91–103. emotional memory consolidation: the effect of stress-induced salivary Shields GS, Lam JCW, Trainor BC, Yonelinas AP. (2016a). Exposure to acute alpha-amylase and cortisol. Biol Psychol 89:539–44. stress enhances decision-making competence: evidence for the role of Laredo SA, Steinman MQ, Robles CF, Ferrer E, Ragen BJ, Trainor BC. (2015). DHEA. Psychoneuroendocrinology 67:51–60. Effects of defeat stress on behavioral flexibility in males and females: Shields GS, Sazma MA, McCullough AM, Yonelinas AP. (2016b). The modulation by the mu-opioid receptor. Eur J Neurosci 41:434–41. effects of acute stress on memory: a meta-analysis and integrative Miyake A, Friedman NP, Emerson MJ, Witzki AH, Howerter A, Wager TD. review. (2000). The unity and diversity of executive functions and their contri- Starcke K, Brand M. (2012). Decision making under stress: a selective butions to complex ‘‘Frontal Lobe’’ tasks: a latent variable analysis. review. Neurosci Biobehav Rev 36:1228–48. Cogn Psychol 41:49–100. Trainor BC, Takahashi EY, Campi KL, Florez SA, Greenberg GD, Laman- Mueller ST, Piper BJ. (2014). The Psychology Experiment Building Maharg A, Laredo SA, et al. (2013). Sex differences in stress-induced Language (PEBL) and PEBL test battery. J Neurosci Methods 222:250–9. social withdrawal: independence from adult gonadal hormones and Plessow F, Fischer R, Kirschbaum C, Goschke T. (2011). Inflexibly focused inhibition of female phenotype by corncob bedding. Horm Behav under stress: acute psychosocial stress increases shielding of action 63:543–50. goals at the expense of reduced cognitive flexibility with increasing von Dawans B, Kirschbaum C, Heinrichs M. (2011). The Trier Social time lag to the stressor. J Cogn Neurosci 23:3218–27. Stress Test for Groups (TSST-G): a new research tool for controlled Schoofs D, Pabst S, Brand M, Wolf OT. (2013). Working memory is differen- simultaneous social stress exposure in a group format. tially affected by stress in men and women. Behav Brain Res Psychoneuroendocrinology 36:514–22. 241:144–53. Zoladz PR, Warnecke AJ, Woelke SA, Burke HM, Frigo RM, Pisansky JM, Schwabe L, Hoffken € O, Tegenthoff M, Wolf OT. (2013). Stress-induced Lyle SM, Talbot JN. (2013). Pre-learning stress that is temporally enhancement of response inhibition depends on mineralocorticoid removed from acquisition exerts sex-specific effects on long-term receptor activation. Psychoneuroendocrinology 38:2319–26. memory. Neurobiol Learn Mem 100:77–87.

Journal

Stress: The International Journal on the Biology of StressTaylor & Francis

Published: Sep 2, 2016

Keywords: Cognitive flexibility; men; orbitofrontal cortex; sex differences; stress; Wisconsin card sorting test; women

References