An inverted U-shaped relationship between cortisol awakening response and same-day error monitoring function in healthy males

Highlights

• CAR is thought to provide an energetic “boost” for the coming day.

• No association was observed between CAR and post-error behavior.

• A significant quadratic relationship was observed between CAR and same-day ERN amplitude.

• Moderate CAR was associated with best error monitoring function in healthy males.

• Cortisol awakening response could be a biological predictor of cognitive function in the same day.

Abstract

The cortisol awakening response (CAR) is thought to provide an energetic “boost” for the coming day and has been shown to be associated with prefrontal dependent function. The aim of the current study was to examine the relationship between the CAR and same-day neural activity following an error response task. Forty-six healthy males (22.25 years ± 1.98) collected saliva samples at 0, 30 and 45 min post-awakening to measure the CAR, and on the same afternoon event-related potential activity during error processing was measured. Although no association was observed between CAR and post-error behavior, a significant quadratic relationship was observed between CAR and Error-related Negativity (ERN) amplitude, and this association remained while controlling for confounding factors. This finding suggested the existence of an inverted U-shaped relationship between CAR and the same-day error-monitoring function in healthy males.

Introduction

Given the established link between cortisol and cognition (de Kloet, Oitzl, & Joëls, 1999; Domes, Rothfischer, Reichwald, & Hautzinger, 2005; Joels, 2006; Lupien et al., 2005; Lupien, Juster, Raymond, & Marin, 2018; Lupien, McEwen, Gunnar, & Heim, 2009), efforts have focused on the cortisol awakening response (CAR), a rapid rise in cortisol levels that peaks between 30- and 45-min following morning awakening (Pruessner et al., 1997). Historically, studies treated the CAR as a trait measure for HPA-axis status (Wust et al., 2000). Its associations with both cognition and affective disorder such as depression have been reported (Almela, van der Meij, Hidalgo, Villada, & Salvador, 2012; Kofman et al., 2019; Kuehner, Holzhauer, & Huffziger, 2007; Quevedo, Doty, Roos, & Anker, 2017), though there is some inconsistency in the direction of relationship. For example, some research show greater CAR is associated with poorer cognition (Butler, Klaus, Edwards, & Pennington, 2017; Franz et al., 2011), but more recent research with accurate monitoring of sampling in the CAR period show greater CAR is associated with better cognitive functioning (Baumler et al., 2014; Evans, Hucklebridge, Loveday, & Clow, 2012).

At the same time, further studies argued that the CAR was prone to significant state variation which is greater than trait variation (Stalder, Evans, Hucklebridge, & Clow, 2010; Stalder, Hucklebridge, Evans, & Clow, 2009). For example, Hellhammer et al. (2007) reported that the CAR of a single day was determined to a great extent by state factors (61 %–82 %). It is proposed that state-related increases in the magnitude of the CAR may play a homeostatic role in ‘boosting’ cognitive performance upon awakening (Adam, Hawkley, Kudielka, & Cacioppo, 2006; Clow, Hucklebridge, Stalder, Evans, & Thorn, 2010). Recent studies found that CAR could predict same-day brain function, especially the prefrontal cortex (PFC) mediated executive function (Law, Hucklebridge, Thorn, Evans, & Clow, 2013). Law, Evans, Thorn, Hucklebridge, and Clow (2015) measured CAR and cognitive function over 50 days in one healthy male and found that on days in which CAR was higher executive function (measured by attention-switching task) in the morning of the same day was better. In an intensive 4-day protocol of 9 males, CAR and brain plasticity later on in the day was measured one week apart, on days in which CAR was greater this was associated with better same-day neuroplasticity measured in motor cortex (Clow et al., 2014). In addition, research from our laboratory also supported a similar positive relationship between the CAR and same-day brain function in prefrontal cortex; higher CAR was predictive of better function of response inhibition (Shi et al., 2018) and stronger positive medial prefrontal cortex connectivity (Wu et al., 2015). However, in another study, Moriarty et al. (2014) showed a more complex U-shaped relationship between CAR magnitude and same-day spatial working memory performance.

Error processing, another critical prefrontal cortex-related high-level ability, plays an important role in adapting to the environment (Botvinick, Cohen, & Carter, 2004; Notebaert et al., 2009). Electrophysiological studies indicate that error processing includes different stages and two electrophysiological indices are of interest: the Error-related Negativity (ERN) and the Error Positivity (Pe) (Bediou, Koban, Rosset, Pourtois, & Sander, 2012; Luijten et al., 2014). The ERN is a negativity elicited approximately 50 ms after erroneous responses and it is thought to reflect the automatic error monitoring mechanism of the individual’s responses to his/her own errors (Gehring, Goss, Coles, Meyer, & Donchin, 2018; Gehring, Goss, Coles, Meyer, & Donchin, 1993). The Pe, a positivity following ERN, possibly reflects the conscious awareness of errors or the motivational significance attributed to the error (Helenius, Laasonen, Hokkanen, Paetau, & Niemivirta, 2010; Lievaart et al., 2016).

Prior studies have addressed the relationship between cortisol and error processing. For example, less cortisol change was correlated with the larger ERN amplitude, indicating that participants with larger ERNs were associated with successful stress regulation (Compton, Hofheimer, & Kazinka, 2013). On the contrary, Tops, Boksem, Wester, Lorist, and Meijman (2006) reported that ERN amplitudes were positively related with cortisol mobilization during a task session. With regards to the relationship between CAR and error processing, to our knowledge, there is only one study that showed a higher CAR was associated with slowed latency of the ERN (Zhang et al., 2015). However, one limitation of this work is that CAR was collected on two consecutive days after the ERP measurement. Since the CAR is characterized by substantial state variation, it is unclear whether CAR is associated with error processing of the upcoming day (i.e. measurement of CAR prior to the measurement of error processing on the same day).

In the present study, we set out to further elucidate whether the CAR was associated with same-day error processing using event-related potentials. To avoid possible confound eff ;ects of sex we examined an all-male, healthy young population (Almeida, Piazza, & Stawski, 2009). We hypothesized that the greater CAR was associated with better error processing (indicated by ERN/Pe). However, we did not exclude quadratic relationship because results from early studies investigating the relationship of the CAR and PFC function have been mixed.