Chronic stress, defined as stress resulting from an “eliciting stimulus that remains in the environment for an extended period of time” (p. 28) (Miller et al., 2007), is considered an epidemic in modern society (Newbegin, 2014; Siervo et al., 2009). This epidemic results (Miller et al., 2007) in numerous long-term negative implications for mental and physical health, quality of life, and all-cause mortality (Da Estrela et al., 2021; Rodgers et al., 2021; Rohleder, 2019; Yaribeygi et al., 2017). Specifically, the presence of stress has been linked to the progression, onset, and severity of several long-term health outcomes including obesity, cardiovascular disease, cancer, insomnia, pain, anxiety, depression, executive and cognitive dysfunction, and death, including death by suicide (Chrousos, 2009; Cohen et al., 2007; Dai et al., 2020; Gallagher et al., 2018; Lagraauw et al., 2015; Østerås et al., 2015; Kanani and Sheikh, 2024).

Physical activity (PA), which is considered any bodily movement that expends energy above a basal level (U.S. Department of Health and Human Services, 2018), has emerged as an effective prevention and treatment strategy for perceived stress (PS). Current guidelines recommend at least 150 min of moderate PA or 75 min of vigorous PA and muscle strengthening exercises at least 2 days per week for optimum benefit, including reduced PS (U.S. Department of Health and Human Services, 2018). Engagement in these recommended levels of PA, regardless of the domain (e.g., biking, walking, swimming, running) is generally regarded as a useful lifestyle approach to stress management (Franklin et al., 2021), with benefits for both physical (Kaminsky et al., 2022; Viña et al., 2012) and mental (DeBoer et al., 2012; De la Rosa et al., 2020; Salmon, 2001) health, and a decreased risk for mortality (Zhao et al., 2020). Thus, manualized treatments that aim to increase engagement in PA, thereby reducing PS and improving mental health disorders, have become widely available. However, the relationship between PA and PS is not always linear (Salmon, 2001) and few people engage in the recommended levels of PA long-term (Evenson et al., 2015; Yang et al., 2019). Furthermore, emerging evidence suggests that many individuals do not in fact experience improvements in affect or PS during an activity bout, and factors such as activity intensity can contribute to worsening feelings of PS or aversion (Shimura et al., 2023; Petruzzello et al., 1991; Box and Petruzzello, 2020). The complex relationship between PA and affect involves several factors such as autonomy, weight, anxiety sensitivity, fitness level, fitness phobia, and exercise intensity which may increase negative affect, including PS (Ekkekakis and Lind, 2006; Ekkekakis et al., 2010; Ekkekakis et al., 2005). Given PS’ prevalence and far-reaching deleterious impacts, as well as the heterogeneous impact of PA on affect, understanding individual differences that would improve the impact of PA on PS are vital. One of these individual differences may be personality. After all, the same stressor may cause a variety of reactions that vary in duration, expression, and intensity in different people based on personality (Cohen et al., 2007).

Personality, (e.g., individual variations in persons Fowers et al., 2023), is one of the key contributors to individual differences in stress responses (Bowling et al., 2005), including both perception of stressors and stress reactivity (Cohen et al., 2007). Among the personality theories tested in the last several decades, the Five Factor Model of personality (i.e., The Big Five; FFM) produces personality profiles that are significantly correlated with PS (Kotov et al., 2010; Lou and Li, 2023; Ringwald et al., 2024; Williams and Carlson, 2024). FFM posits that personality is best characterized across five domains, including neuroticism, extraversion, conscientiousness, agreeableness, and openness to experience (Costa and McCrae, 1992). FFM has been validated globally (McCrae and Costa, 1987), and neuroticism appears to have the strongest relationship with PS such that those high in neuroticism experience more stressors and perceive stress as more intense (Ringwald et al., 2024). Studies have also validated FFM’s ability to predict stress-related negative outcomes, for example low openness and the experience of PS appear to interact to worsen sleep (Williams and Carlson, 2024) and increase poor cardiac outcomes (Gallagher et al., 2018).

One may posit that the complex, heterogeneous nature of the PA-PS relationship may be in part related to differences in personality. For instance, it may be that those with higher neuroticism are more likely to experience PA as a stressor, while those higher in conscientiousness tend to experience the same stimulus as relaxing or enjoyable. Recent studies have suggested personality factors relate to both short-term and long-term PA engagement, as well as PA enjoyment (Caille et al., 2024; Dominski et al., 2020; Dominski et al., 2021; Engels et al., 2022; Huang et al., 2024). Yet, to date, no study to our knowledge has explored the interaction between PA and personality on PS. This knowledge of individual differences impacting PA will be key in moving the field of stress management toward person-centered and precision medicine (Yardley and Campbell, 2020), an approach that encourages tailored treatments to the individual patient based on that patient’s individual differences. Thus, the purpose of this study is to determine if any FFM personality factor moderates the PA-PS relationship and if so in what direction. Based on the literature, we predict that neuroticism, extraversion, conscientiousness, and PA will predict PS. Given the paucity of the research, we did not have a priori hypotheses regarding interaction effects.

Six hundred thirty-six participants were screened for eligibility in the current study. Of those, 595 consented to participate and self-reported being eligible for participation. Another 56 were removed after validation checks deemed them ineligible due to age and self-reported history of an anxiety disorder. Two hundred sixty-three participants were missing data. Those who completed surveys showed very high conscientiousness scores which may have contributed to their decision to complete the surveys. Statistical comparisons between groups were not possible since those who did not complete the surveys did not have personality data. Taken together, these data did not appear to be missing at random and the use of multiple imputation was precluded. These participants were excluded from this study, resulting in a final sample of 276.

In total, 47% of participants were female, 38% were White, 26% were Asian American, 6% were Black, and 6% were Hispanic/Latino, with the majority of participants reporting regular engagement in, enjoyment of, and motivation to engage in exercise (see Tables 1, 2). Ninety three percent of the sample reported at least some college education and 75.4% reported participating in high school sports. Fourteen percent reported being diagnosed with a mental illness other than anxiety and nearly half (48.2%) reported experiencing an event that caused more PS than normal in the last month.

Prior to conducting regression modeling, visual inspection confirmed the assumption of normal residuals, and correlations and variance inflation factor did not suggest issues with collinearity. In single predictor models, neuroticism (B = 0.259, p < 0.001) explained the most variance in PS, followed by conscientiousness (B = −0.191, p < 0.001), extraversion (B = −0.137, p < 0.001), and agreeableness (B = −0.099, p < 0.001). PA and openness to experience did not significantly predict PS. In simple interaction models, only openness to experience X PA was significant (B = 3.35, p = 0.014) suggesting that among those with high openness scores, more PA was associated with higher PS. The opposite was true among those with low openness scores (see Figure 1). The openness to experience X PA interaction was somewhat mitigated in a final model (see Tables 3, 4), such that the pattern remained consistent but was no longer statistically significant (p = 0.066). In this final model, conscientiousness was negatively associated with PS (B = 0.087, p < 0.001) and neuroticism was positively associated with PS (B = 0.221, p < 0.001). Agreeableness was not retained in the final model.

The current study expanded previous work on the relationship between the FFM model of personality, PS, and PA by exploring the relationship amongst these three constructs (i.e., personality, PA, and PS) in a diverse adult sample. The results revealed that neuroticism and conscientiousness predicted PS but not PA, which offered partial support for our hypothesis that personality and PA would predict PS. Thus, this study provides additional support for the relationship between aspects of the FFM model and PS. Notably, an interaction between PA and openness to experience was also observed, whereby those with relatively low levels of openness were more likely to demonstrate lower levels of PS when they were active and vice versa. This pattern was not significant in the final model. Still, further investigation—particularly leveraging more accurate and objective measures of PA—are warranted to determine whether individuals who are highly active and highly open may be in need of additional stress management support.

The lack of relationship between PA and PS in this sample was unusual in view of the literature to date connecting personality and PA (Wilson and Dishman, 2015), as well as PA and PS (Chauntry et al., 2022; Perchtold-Stefan et al., 2020). However, participants in this study reported the equivalence of nearly 1,000 min of moderate-intensity PA per week, or nearly seven times the current PA guideline recommendations. These data may suggest that the relationship amongst these variables is less marked in those who self-identify as frequently engaging in exercise (i.e., structured PA) as part of their daily active energy expenditure or those who are high in conscientiousness similar to those in this sample. This surprising result could also be due to the use of self-report measures of PA versus objective measures or the use of total METs rather than specific domains of PA. These findings might suggest the importance of assessing current levels PA and personality characteristics during treatment as this could inform the decision of including a PA component to the intervention. Finally, our study revealed a surprising interactive relationship between openness to experience and PA on PS that may be related to novelty seeking. Given the preliminary nature of these exploratory findings, these possible explanations should be investigated further in future studies that include a less educated and less active sample.

Expanding the understanding of how to match patients to stress management techniques that they are more likely to enjoy and therefore adhere to may guide the field toward precision medicine and decrease the far-reaching negative impacts of chronic stress. Beyond stress reduction, physical therapy and structured exercise have been proposed as novel frameworks for suicide prevention and depression management, underscoring their broader mental health benefits (Kanani, 2025).