On decision-making theories’ crib, choices were described as the result of a conscious, deliberative, process in which known alternatives and their consequences would be analyzed and compared. According to classical theories of decision-making, people chose the most advantageous available option and considered the availability of their own resources (e.g., budget) to handle the consequences of their choices (1, 2). Later, theories started to acknowledge that human behavior, including choice behavior, was not purely rational and highlighted the influence of intuitive and emotional processes (2–7), a proposition now widely accepted.

A growing body of research has proposed people present personal tendencies to rely on either rational or intuitive processes when making decisions. This tendency may be described as “the typical manner by which individuals make decisions” and it is argued that it is built upon personality traits and habit-based inclinations (8). Intuitive and rational decision styles are not mutually exclusive (9) and may work together to provide better outcomes (8). Each one of them presents pros and cons that might be better suited for different contexts. To Payne (10), the efficacy and the expression of each decision style depends on contextual characteristics, in a way that personal tendencies be overrode by situational specificities, leading an individual to be more intuitive than usual, for example.

It is well established that decision-making is influenced by both environmental/contextual factors and individual differences, but, so far, researchers have focused on the study of direct/main effects rather than looking into interactions among them. On this paper, we seek to investigate whether Post-Traumatic Stress (PTS) symptoms reported during COVID-19 pandemic in Brazil impacts on measures of rational and intuitive decision-making styles. Post-Traumatic Stress Disorder (PTSD) is the most frequent and the most investigated psychiatric disorder on population exposed to disasters (11). Several studies associates PTSD to deficits on cognitive performance (12–19), including decision-making (20, 21). Neuroimaging studies suggest that brain regions associated with decision-making are sensitive to changes induced by stress response and behavioral research supports the hypothesis that stress influences decision processes [for a review, see (22)]. Uncertainty related to an unknown disease coursing with socioeconomic implications and frequent deaths seem to amplify the risk perception. So, COVID-19 pandemic can be characterized as a period higher than typical rates of perceived risk (23, 24) and psychological distress (23, 25–28). The rates of post traumatic symptoms and disorders in mixed populations reached 15% of prevalence during SARS and the COVID-19 pandemic including the health care professionals in Brazil (29, 30).

Trait Anxiety (TA) belongs to the neuroticism x emotional stability personality trait and “refers to the stable tendency to attend to, experience, and report negative emotions such as fears, worries and anxiety across many situations” (31: p. 1,989). Hartley and Phelps (32) point out that Pre-Frontal Cortex (PFC)-dependent cognitive and affective functions may be impaired in anxiety disorders. One of the impaired functions would be decision-making since anxiety and decision-making ability share neural substrates. Additionally, PFC activity may reduce susceptibility to biases and promote more rational decision-making (5, 33), which supports the hypothesis that more anxious individuals could make less rational choices. Studies that investigate the relationship between decision-making and anxiety point out that anxious individuals tend to have greater risk aversion, showing preference for safer choices in contexts of uncertainty. They would also exhibit a more pessimistic assessment of the situation (34). Other studies also suggest that anxiety is associated with higher levels of loss aversion (35, 36). As pointed by (37), the relationship between TA and decision-making only recently received attention and little is known about the impact of anxiety on specific decision-making variables (32).

Here we analyze whether TA affects the relationship between PTS symptoms and decision-making during COVID-19 pandemic. Building upon the literature about the impact of stress response on cognitive functioning, and considering TA’s cognitive characteristics, we hypothesize that more anxious individuals might display different patterns of response to PTS symptoms on decision-making strategies when compared to less anxious people. Specifically, we predict that TA might be associated with the use of more intuitive and less rational decision-making strategies.

Demographic data is presented in Table 1, consisted of mostly by women in middle age reporting stress not related to COVID-19.

To test the hypothesis that decision-making style varies as a function of multiple factors and, more specifically, whether trait anxiety moderates the relationship between PTS and decision-making, simple linear regression and simple moderator analysis were conducted. In the first step, a simple linear regression was calculated to predict rational decision-making style based on PTS (Table 2, model 1). A significant regression equation was found [F(1,1,356) = 10.752, p = 0.001; R² = 0.008]. The results of the regression indicated that PTS significantly predicted rational decision-making style (β = −0.089, p = 0.001). Then, a simple linear regression was run to verify whether rational decision-making style varied as a function of TA (Table 2, model 2). Results indicated that TA significantly predicted rational decision-making style (β = −0.176, p < 0.001), accounting for approximately 3% of the variance on rational decision-making style reports [F(1,1,356) = 43.531, p < 0.001].

A similar process was conducted for the intuitive decision-making variable. Simple linear regression results (Table 2, model 3) suggest PTS is a valid predictor of intuitive decision-making style (β = 0.119, p < 0.001) on a model where [F(1,1,356) = 19.481, p < 0.001, R² = 0.014]. Another simple linear regression equation was tested to verify whether intuitive decision-making style varied as a function of TA (Table 2, model 4). The equation was not significant [F(1,1,356) = 0.001, p = 0.973] and TA was not pointed as a significant predictor of intuitive decision-making style (β = 0.001, p = 0.973).

A simple moderator analysis performed to investigate conditional effects of PTS on rational decision-making style (Table 3) showed that the interaction between PTS and TA was statistically significant (b = −0.1654, 95% C.I. = −0.2964, −0.0344, p < 0.05). According to the results, the conditional effect of PTS on rational decision-making style was only significant when TA was high (TA scale score ≥ 15), with effect = −0.0891, (95% C.I. = −0.1689, −0.0093, t = −2.1907, p < 0.05). When TA was low (TA scale score < 15), (conditional effect was 0.0763, 95% C.I. = −0.0276, 0.1802, t = 1.4412, p = 0.1479). These results suggest TA acts as a negative moderator of the relationship between PTS and rational decision-making style. Conditional effects for rational decision-making are represented on Figure 1A. Results also suggest PTS’ impact on rational decision-making is only significant when TA is high.

Results from the simple moderator analysis performed to investigate conditional effects of PTS on intuitive decision-making style (Table 3) showed that the interaction between PTS distress and TA was not statistically significant (b = 0.0111, 95% C.I. = −0.1310, 0.1533, p = 0.8778). Regression analysis with mean-centered variables showed that only PTS was a significant main predictor of intuitive decision-making (b = 0.1598, 95% C.I. = 0.0471, 0.2726, t = 2.7808, p = 0.0055). There was a trend for the main effect of TA on intuitive decision-making (b = −0.3639, 95% C.I. = −0.7658, 0.0381, t = −1.7759, p = 0.076). These results suggest no moderation effect of TA on the relationship between PTS and intuitive decision-making style. Effects for intuitive decision-making are visually displayed at Figure 1B.

Taken together, results suggest the direction of the impact of PTS on rational decision-making style depends on TA level.

This study investigated the relationship between Post-Traumatic Stress and decision-making styles, and whether Trait Anxiety moderates it. Our findings suggest that the impact of exposure to stressful contexts, measured by PTS, may change how people make decisions. Furthermore, the direction of this change depends on how vulnerable to anxiety people are. People who reported higher TA exhibited lower levels of both rational and intuitive decision-making styles when compared to subjects who reported lower tendency to anxiety. More importantly, on contexts of trauma, the more anxious people adopted a less rational decision-making style while subjects with lower TA displayed more reason-based strategies.

Some mechanisms might be behind the observed results. The first one is cognitive overload hypothesis. Literature on decision-making suggests that, in uncertain scenarios, our mental resources, necessary for self-control, are drained, which leads to an increase of the preference for “wants” over “shoulds” (44). The cognitive overload hypothesis is also present in the reasoning used to explain why people exposed to contexts of poverty, debt and other stressful scenarios have worse cognitive functioning when compared to other populations and themselves in less challenging contexts (45, 46). TA might compromise the ability to use effort-based strategies under stress via cognitive depletion due to heightened mental activity toward anxiety-influenced processes, such as worry, catastrophizing and planning for hypothetical scenarios.

Uncertainty is also a factor that might influence the observed results. With the beginning of the COVID-19 pandemic, the world experienced the biggest periods of uncertainty of the recent history. In Brazil, the scenario is aggravated by factors such as government misinformation through contradictory directions by authorities responsible for coordinating actions to deal with the current health crisis (47). Brazilians were therefore immersed in uncertainties that encompass the economic, political, social, and health spheres. We suggest it is plausible to consider that the general context of uncertainty might play a part on how people experienced stress response. In a study of two different stressful scenarios, such as in medical activities after earthquake and in COVID-19, a population also composed mostly by women perceive stress as anxiety, somatization, and depression acutely. The organization and participation on work related to the response to the event were important to perception of distress to improve health and wellbeing of professionals (48).

According to our analysis, PTS was positively associated to a more intuitive decision-making style. This observation is consistent with the findings that suggest uncertainty elicits intuitive/automatic processes of decision-making (22). It is possible to argue that TA’s heightened vulnerability to uncertainty might be part of the mechanism through which PTS decreases the use of analytical/rational decision-making strategies on high TA individuals.

Another point that might be helpful on the interpretation of our results concerns the heightened sensitivity to internal sensation associated to the experience of emotions exhibited by high TA individuals (49). Baradell and Klein (50) conducted a study in which subjects with higher inner body consciousness showed increased susceptibility to experience impact of critical life events and daily struggles on decision-making. It is possible to argue that TA increases subjective perception of stress through sensitivity to emotion-related inner body sensations, therefore modulating the impact of distress on decision-making processes.

One of the greatest limitations of this study is the conceptualization on decision-making styles and its influence on results interpretation. While some authors understand decision-styles as crystalized constructs (primarily defined by personality traits and, therefore, constant throughout life), others highlight the influence of habit-based learning on personal tendencies on how to approach decision tasks. However, as pointed earlier, authors believe decision styles might be influenced by acute contextual factors, such as time pressure (7, 10). Our work tries to attend earlier recommendation to try identifying such factors. Finally, this study presents limitations regarding its sample. Subjects who participated on this research were self-selected and had to be able to assess the internet, which might create a sampling bias toward subjects with higher socioeconomic condition.

In conclusion, this work results suggest that people with higher tendency to display more frequent, intense, and dysfunctional levels of anxiety may suffer greater cognitive impact on stressful scenarios. Our findings align with and may contribute to the theory of differential sensitivity to context, which proposes that people with different levels of environmental vulnerability experience different outcomes when exposed to stress (51). Our work might contribute to the understanding of how stress impacts cognitive functioning and help to identify the most vulnerable individuals.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Universidade Federal de Minas Gerais - CONEP. The patients/participants provided their written informed consent to participate in this study.

MR: designed the study, wrote the draft, revised literature, and analyzed data. LM-D, AS, and DM: supervision, work on project design and search for funding. RJ, AS, APa, JP, DC, and MR-S: help on data collect and study design and discussions. APi: analyzed data and revise the manuscript. All authors contributed to the article and approved the submitted version.

This study was funded by governmental funding by the agencies: CAPES, CNPq, OPAS, MS-Brasil, and FAPEMIG.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.