Occupational Stress and Anxiety Among Korean Employees: Moderated Mediation Model of Burnout and Grit

Abstract

Objective

Although occupational stress is a major risk factor for high anxiety in employees, the specific mechanisms underlying this relationship are not sufficiently established. This study investigated an interpersonal model of anxiety development in employees, wherein occupational stress is associated with burnout and burnout affects risk for anxiety, and examined whether this mediation is moderated by grit.

Methods

The 11,421 participants, aged 19–65 years, were employees of 18 private companies and local government organizations in South Korea. They completed the Korean versions of the Occupational Stress Scale, Oldenburg Burnout Inventory, Clinically Useful Anxiety Outcome Scale, and Grit Scale. Mediation and moderation analyses were performed using the SPSS PROCESS macro.

Results

The association between occupational stress and anxiety is mediated only by exhaustion (b=0.348, 95% confidence interval [0.330, 0.367]), not by disengagement. Moreover, the effect of exhaustion on anxiety is moderated by grit, with the effect being stronger for employees with low grit (passion: b=1.245, p<0.001; perseverance: b=1.274, p<0.001) than for those with high grit (passion: b=0.797, p<0.001; perseverance: b=1.004, p<0.001).

Conclusion

The study findings contribute to the understanding of how occupational stress is associated with anxiety in workplace, and have practical implications for preventing burnout and nurturing grit to protect employees’ mental health.

INTRODUCTION

Occupational stress refers to an employee’s response to work demands that exceed their resources and coping abilities [1]. While moderate stress can be stimulating and pleasantly challenging [2], excessive stress is linked to psychiatric conditions like anxiety [3-5]. Anxiety, characterized by tension and worry, is a normal stress response that can improve coping but, when excessive, harms workplace productivity through absenteeism and presenteeism [6]. A recent study [7] reported higher absenteeism in employees with anxiety, underscoring the need for more research on workplace anxiety, which is less explored compared to workplace depression.

Occupational stress is explained by the demand–control model [8], where high demands and low control lead to psychological disorders, and the effort–reward imbalance model [9], which emphasizes the stress caused by disproportionate effort and insufficient rewards. Studies show a positive correlation between occupational stress and anxiety [10-12], including in Korean employees [13,14]. Korea’s collectivist culture, stress is often driven by interpersonal relationships rather than individual factors, highlighting the need for context-specific research [15-17].

Burnout, defined as exhaustion and disengagement due to excessive demands [18], is linked to mental health issues such as depression and anxiety [19,20]. Studies link high job demands, extrinsic effort, and over-commitment to increased anxiety [21], with emotional exhaustion and cynicism positively associated with anxiety [22]. Burnout, particularly emotional exhaustion, correlates with anxiety [23]. While the learned helplessness theory explains burnout-depression links, its application to anxiety requires further exploration.

Grit, the combination of passion and perseverance for long-term goals [24], reflects the ability to stay committed to meaningful objectives despite challenges [25]. Passion drives initial engagement, while perseverance sustains effort [26]. Grit is linked to better mental health and healthcare management [27], particularly in reducing depression [27,28]. However, research on its impact on anxiety is limited, and studies are needed to explore whether grit can mitigate anxiety symptoms, especially given the role of over-commitment.

Our team previously studied depressive symptoms in the same sample [29]. Given the different mechanisms behind depressive and anxiety symptoms, this study focuses on the impact of occupational stress on anxiety among Korean employees. It also explores burnout’s mediating role between stress and anxiety, and grit’s moderating effect between burnout and depression, within a conceptual framework (Figure 1). Our hypotheses are as follows:

Figure 1.

Conceptual framework.

Hypothesis 1: Occupational stress has a positive correlation with anxiety.

Hypothesis 2: High level of occupational stress is associated with a high level of burnout, which in turn, causes a high level of anxiety.

Hypothesis 3: The effect of burnout on anxiety is stronger for employees with low grit than for those with high grit.

METHODS

Participants

This study included male and female employees aged 19–65 who underwent mental health screenings at the Workplace Mental Health Institute of Kangbuk Samsung Hospital, Seoul, South Korea, between April 2020 and March 2022. These screenings were part of regular occupational health examinations offered to employees under contracts with private companies and public institutions. Participation was voluntary, and the sample included a broad range of occupational groups, such as office workers, manufacturing workers, and employees from public organizations, thereby providing heterogeneity across industries. Inclusion required completion of both questionnaires and socio-demographic information. Out of 12,344 respondents, 923 were excluded for incomplete clinical assessments, resulting in 11,421 participants (7,749 male and 3,672 female) for analysis.

All study procedures were approved by the Institutional Review Board of Kangbuk Samsung Hospital and complied with the Declaration of Helsinki and Good Clinical Practice (approval number: KBSMC 2022-03-046). Informed consent was waived as only de-identified data collected during routine health screenings were utilized.

Clinical assessments

Socio-demographic information encompassed details such as age, sex, educational attainment (high school or below, college graduate, university graduate, master’s degree, doctorate degrees), and marital status (married, unmarried, other [divorced, widowed, separated]). Additionally, job-related demographic data included the duration of work at the current workplace (years), weekly hours of work, and monthly earned income.

Occupational stress was assessed using the Korean Occupational Stress Scale-Short Form (KOSS-SF), a 24-item self-report questionnaire scored on a 4-point Likert scale (1=strongly disagree to 4=strongly agree), with higher scores indicating greater stress. The composite score comprises seven subscales: high job demands, insufficient job control, inadequate social support, job insecurity, organizational injustice, lack of reward, and discomfort in the organizational climate. The Cronbach’s alpha for KOSS in this study was 0.895. To address the differing number of items across subscales, the following formula [17] was applied instead of simply summing subscale scores:

$$\begin{array}{l}\text{Corrected KOSS subscale score}\\ \text{=(Actual score-Number of questions)/}\\ \text{(Highest possible score-Number of quetstions)}\end{array}$$

$$\text{KOSS total score=(Sum of corrected KOSS subscale scores)/7}$$

Anxiety was measured with the Clinically Useful Anxiety Outcome Scale (CUXOS), a 20-item self-report questionnaire evaluating anxiety severity over the past week. It has a 6-item psychic anxiety subscale and a 14-item somatic anxiety subscale, scored on a 5-point Likert scale (0=not at all true to 4=almost always true). The Cronbach’s alpha for CUXOS was 0.952.

Burnout was assessed using the Oldenburg Burnout Inventory (OLBI) [30], comprising 16 items measuring exhaustion (8 items) and disengagement (8 items), scored on a 5-point Likert scale (1=strongly disagree to 5=strongly agree). High scores indicate higher burnout levels. The Cronbach’s alpha for exhaustion was 0.882 and for disengagement, 0.833.

Grit was evaluated with the Short Grit Scale [31], an 8-item self-report questionnaire scored on a 5-point Likert scale (1=not at all like me to 5=very much like me), assessing passion (3 items, Cronbach’s alpha: 0.585) and perseverance (5 items, Cronbach’s alpha: 0.779). The overall Cronbach’s alpha for grit in this study was 0.794.

Statistical analysis

The descriptive analyses were conducted on demographic characteristics and correlations among six variables: occupational stress, exhaustion, disengagement, severity of anxiety symptoms, passion, and perseverance. Differences in anxiety levels based on categorical demographic characteristics were examined using independent t-tests and analysis of variance.

The mediation and moderation models were analyzed using Models 4 and 14 in the PROCESS Macro [32] in SPSS. To reduce multicollinearity, mean centering was applied to independent and moderating variables. Mediation analyses involved occupational stress as the independent variable, severity of anxiety symptoms as the dependent variable, and exhaustion and disengagement as mediators. Bootstrapping (5,000 samples) was used to estimate indirect effects and 95% confidence intervals (CIs). Moderated mediation effects with two moderators (passion and perseverance) were tested. A significant interaction between burnout and grit indicated a moderated mediation effect, assessed by examining conditional effects at one standard deviation (SD) above and below the mean of grit. Control variables included age, sex, education, marital status, years of service, weekly hours worked, and monthly income. A CI including zero indicated a lack of significant mediating effect at the 5% significance level. Sensitivity analyses were performed using Model 1 in the PROCESS Macro. All analyses were conducted with SPSS 28.0 for Windows (IBM, Corp.).

RESULTS

Participants’ demographic, job, and clinical characteristics

Table 1 displays participants’ demographic, job, and clinical characteristics. Participants’ mean age was 36.7±9.4 years, and 7,749 (67.9%) participants were male. CUXOS scores showed significant positive correlation with age (r=0.022, p=0.021), years of service (r=0.021, p=0.027), and weekly hours of work (r=0.096, p<0.001), and significant negative correlation with monthly earned income (r=-0.067, p<0.001). Female sex (t=-23.789, p<0.001), lower degree of education (F=33.328, p<0.001), and unstable marital status (i.e., divorced, widowed, separated; F=5.667, p=0.003) were associated with higher CUXOS scores.

Table 1.

Participants’ demographic, job, and clinical characteristics (N=11,421)

	Value
Sex
Male	7,749 (67.9)
Female	3,672 (32.1)
Age (yr)	36.7±9.4
Education (graduate)
High school or below	1,404 (12.3)
College graduate	1,402 (12.3)
University graduate	6,840 (59.9)
Master’s degree	1,452 (12.7)
Doctorate degrees	323 (2.8)
Marital status
Married	6,370 (55.8)
Unmarried	4,806 (42.1)
Other	245 (2.1)
Years of service (yr)	10.5±9.1
Hours of work per week (hours)	46.8±7.5
Monthly earned income (million won)	4.3±2.7
Clinical characteristics
KOSS total score	40.82±14.01
OLBI exhaustion score	23.63±6.24
OLBI disengagement score	23.54±5.93
CUXOS score	17.09±14.59
Grit passion subscale score	9.79±2.07
Grit perseverance subscale score	17.28±3.45

Data are presented as mean±standard deviation or number (%).

KOSS, Korean Occupational Stress Scale; OLBI, Oldenburg Burnout Inventory; CUXOS, Clinically Useful Anxiety Outcome Scale.

Occupational stress–burnout–anxiety pathways

Bivariate correlations among study variables are presented in Supplementary Table 1. As age, sex, education level, marital status, years of service, weekly hours worked, and monthly income significantly correlated with workplace anxiety symptoms, these variables were controlled in subsequent mediation effect evaluations. The PROCESS macro assessed the mediating roles of exhaustion and disengagement in the relationship between occupational stress and anxiety symptoms among workers.

The significance of the mediating effects was tested using a bootstrap method. Table 2 shows the results of the bootstrap analysis. Occupational stress was directly and positively predictive of anxiety symptoms (b=0.172, p<0.001) and positively related to exhaustion (b=0.307, p<0.001) and disengagement (b=0.315, p<0.001). Exhaustion, in turn, was positively related to anxiety (b=1.133, p<0.001); however, disengagement was not significantly related to anxiety (b=0.022, p=0.512). The total, direct, and total mediating effects of exhaustion and disengagement between occupational stress and anxiety symptoms were 0.527, 0.172, and 0.355, respectively. The effect of the path occupational stress → exhaustion → anxiety symptoms was 0.348 (95% CI [0.330, 0.367]); the effect of the path occupational stress → disengagement → anxiety symptoms was not significant. These results indicate that exhaustion partially mediate the relationship between occupational stress and anxiety symptoms.

Table 2.

Mediation analysis results

Predictor	Model 1 (Exhaustion)				Model 2 (Disengagement)				Model 3 (Anxiety)
	B	SE	t	p	B	SE	t	p	B	SE	t	p
Age	-0.080	0.007	-10.813	<0.001	-0.078	0.007	-11.923	<0.001	0.142	0.020	7.117	<0.001
Sex	1.103	0.089	12.406	<0.001	0.774	0.078	9.884	<0.001	2.317	0.239	9.703	<0.001
Education	0.146	0.049	2.979	0.003	-0.013	0.043	-0.292	0.770	-0.045	0.130	-0.348	0.728
Marital status	-0.045	0.096	-0.475	0.635	-0.408	0.074	-4.845	<0.001	0.288	0.255	1.127	0.260
Years of service	0.007	0.007	0.946	0.344	0.019	0.006	3.097	0.002	0.023	0.019	1.229	0.219
Working hours	0.072	0.005	13.636	<0.001	0.011	0.005	2.374	0.018	0.013	0.014	0.884	0.377
Income	0.020	0.016	1.219	0.223	-0.063	0.014	-4.442	<0.001	-0.122	0.043	-2.856	0.004
Occupational stress	0.307	0.003	105.055	<0.001	0.315	0.003	122.298	<0.001	0.172	0.012	13.850	<0.001
Exhaustion									1.133	0.029	38.815	<0.001
Disengagement									0.022	0.033	0.656	0.512
R2	0.549				0.613				0.414
F	1,729.180***				2,245.854***				802.132***
Bootstrap			Occupational stress → Burnout (Exhaustion and Disengagement) → Anxiety
			B		SE		LCI		UCI		Conclusion
Total effect			0.527		0.009		0.510		0.543
Direct effect			0.172		0.012		0.147		0.196
Total indirect effect			0.355		0.010		0.335		0.376		Partial mediation
Indirect effect of exhaustion			0.348		0.009		0.330		0.367		Partial mediation
Indirect effect of disengagement			0.007		0.010		-0.308		0.374		No mediation

^***p<0.001.

B, estimate of the regression coefficient; SE, standard error of estimate; t, t-test value; p, significance probability; R², explanatory power; F, F-test value; LCI, lower bound of 95% confidence interval; UCI, upper bound of 95% confidence interval.

Grit as a moderator in the occupational stress–burnout–anxiety pathways

Table 3 presents the moderated mediation analysis results, highlighting grit (passion and perseverance) as a moderator in the relationship between burnout (exhaustion and disengagement) and anxiety symptoms among employees using mean-centered variables. Significant interactions were observed between exhaustion and passion (b=-0.108, p<0.001) and exhaustion and perseverance (b=-0.039, p<0.001), while disengagement and grit did not show significant interaction. For sensitivity analysis, we reassessed grit’s moderating effect on the exhaustion-anxiety relationship using Model 1 in the SPSS PROCESS Macro. Results confirmed that both passion and perseverance significantly moderate this association (Table 3).

Table 3.

Results of the moderated mediation analysis of burnout and grit on anxiety

Predictor	Anxiety				Anxiety moderated by passion				Anxiety moderated by perseverance
	B	SE	t	p	B	SE	t	p	B	SE	t	p
Age	0.142	0.020	7.117	<0.001	0.151	0.019	7.873	<0.001	0.146	0.020	7.381	<0.001
Sex	2.317	0.239	9.703	<0.001	1.791	0.232	7.733	<0.001	2.197	0.238	9.252	<0.001
Education	-0.045	0.130	-0.348	0.728	-0.030	0.126	-0.234	0.815	0.021	0.130	0.161	0.872
Marital status	0.288	0.255	1.127	0.260	0.461	0.247	1.867	0.062	0.316	0.254	1.244	0.213
Years of service	0.023	0.019	1.229	0.219	0.014	0.018	0.760	0.447	0.021	0.018	1.147	0.251
Working hours	0.013	0.014	0.884	0.377	0.011	0.014	0.761	0.446	0.012	0.014	0.854	0.393
Income	-0.122	0.043	-2.856	0.004	-0.114	0.041	-2.747	0.006	-0.120	0.043	-2.818	0.005
Occupational stress	0.172	0.012	13.850	<0.001	0.183	0.012	15.205	<0.001	0.188	0.012	15.139	<0.001
Exhaustion	1.133	0.029	38.815	<0.001	1.022	0.029	35.134	<0.001	1.139	0.030	37.749	<0.001
Disengagement	0.022	0.033	0.656	0.512	-0.032	0.032	-0.990	0.322	0.002	0.033	0.044	0.965
Passion					-1.297	0.056	-23.208	<0.001
Perseverance									-0.085	0.036	-2.328	0.020
Exhaustion×Passion					-0.108	0.012	-8.912	<0.001
Exhaustion×Perseverance									-0.039	0.007	-5.262	<0.001
Disengagement×Passion					-0.020	0.013	-1.513	0.130
Disengagement×Perseverance									-0.015	0.008	-1.791	0.073
R2	0.414				0.453				0.422
F	802.132***				722.539***				638.217***
Mediator					Index of moderated mediation
					Effect		SE		LCI		UCI
Occupational stress → Exhaustion → Anxiety
Passion					-0.033		0.004		-0.040		-0.026
Perseverance					-0.012		0.002		-0.017		-0.008
Occupational stress → Disengagement → Depression
Passion					-0.006		0.004		-0.014		0.002
Perseverance					-0.005		0.003		-0.009		<0.001
	Occupational stress → Anxiety				Exhaustion → Anxiety				Disengagement → Anxiety
	B	LCI		UCI	B	LCI		UCI	B	LCI		UCI
Low passion (-SD)	1.246	1.170		1.321	0.383	0.357		0.408	0.003	-0.025		0.031
Medium passion (SD)	1.022	0.965		1.079	0.314	0.296		0.332	-0.010	-0.030		0.010
High passion (+SD)	0.797	0.722		0.873	0.245	0.225		0.266	-0.023	-0.047		<0.001
Low perseverance (-SD)	1.274	1.196		1.353	0.392	0.366		0.419	0.016	-0.013		0.043
Medium perseverance (SD)	1.139	1.080		1.199	0.350	0.331		0.370	0.001	-0.021		0.021
High perseverance (+SD)	1.004	0.927		1.081	0.309	0.287		0.331	-0.015	-0.042		0.010

^***p<0.001.

B, estimate of the regression coefficient; SE, standard error of estimate; t, t-test value; p, significance probability; R², explanatory power; F, F-test value; LCI, lower bound of 95% confidence interval; UCI, upper bound of 95% confidence interval; SD, standard deviation.

The simple main effects analysis evaluated the effect of exhaustion on anxiety symptoms at one SD above and below the mean grit levels. Figure 2 illustrates that predicted anxiety symptoms are influenced by exhaustion and grit levels. Simple slope tests indicated that exhaustion has a stronger effect on anxiety for employees with low grit passion (b=1.246, 95% CI [1.170, 1.321]) compared to high grit passion (b=0.797, 95% CI [0.722, 0.873]). A similar pattern was found for grit perseverance, with lower grit levels showing a stronger effect on anxiety (low perseverance: b=1.274, 95% CI [1.193, 1.353]; high perseverance: b=1.004, 95% CI [0.927, 1.081]). The conditional indirect effects of occupational stress on anxiety symptoms via exhaustion were also assessed. The indirect effect at one SD below the mean grit passion (b=0.383, 95% CI [0.357, 0.408]) was stronger than at one SD above (b=0.245, 95% CI [0.225, 0.266]). Similarly, the indirect effect for grit perseverance was stronger below the mean (b=0.392, 95% CI [0.366, 0.419]) than above (b=0.309, 95% CI [0.287, 0.331]), supporting the moderated mediation effects.

Figure 2.

Grit (A) passion and (B) perseverance as moderators in the association between the exhaustion dimension of burnout and anxiety in employees.

DISCUSSION

This study investigated the association between occupational stress and anxiety symptoms and the mediating effect of burnout, comprising exhaustion and disengagement, among Korean employees. Moreover, grit, comprising passion and perseverance, was validated as a protective factor buffering the adverse effects of burnout on anxiety symptoms. To our knowledge, this is the first study to investigate the protective effects of grit in the pathway of burnout to anxiety among a sample of Korean employees.

The findings show that high occupational stress is associated with elevated anxiety symptoms, aligning with previous studies [33-35]. For instance, a study [13] found significant correlations between KOSS-SF subscales, except insufficient job control, and anxiety in Korean female manufacturing workers. Similarly, another study14 reported significant links between most KOSS subscales and anxiety in Korean male office workers. Consistent with prior research, in our multiple linear regression analysis, all subscales except insufficient job control and inadequate social support showed significant associations with anxiety symptoms (Supplementary Table 2). This study examined work-related strain’s psychological impact on anxiety, considering the job demand–control model and the effort–reward imbalance model [36]. The job demand–control model links stress to high demands and low control, with social support as a mitigating factor [37], while the effort–reward imbalance model emphasizes balance between effort and rewards like money, esteem, and job security [36]. The results of this study indicate that the effort–reward imbalance model better explains the relationship between occupational stress and anxiety symptoms, as insufficient job control and inadequate social support do not account for anxiety levels, contrasting with the predictions of the job demand–control model.

This study demonstrates that the relationship between occupational stress and anxiety is mediated by exhaustion rather than disengagement, consistent with previous research linking job demands, burnout, and anxiety [19,22,38,39]. Discrepancies arise from differences in how burnout components are categorized; this study used the OLBI, which separates burnout into exhaustion and disengagement, whereas other instruments may differ. Future research should explore various methodologies for a more comprehensive understanding of burnout-anxiety relationships. A previous study [12] highlighted the importance of considering both extrinsic and intrinsic stressors, particularly over-commitment, which moderates the link between occupational stress and anxiety. Research involving Chinese civil aviation pilots [40] and nurses [12] found that individuals with high over-commitment scores experience increased anxiety with rising occupational stress. Overcommitment, linked to the effort–reward imbalance model, contributes to mental health decline [41] and may result in repeated effort–reward imbalances leading to exhaustion and anxiety. A related study by the same research team [29] found mediating effects of exhaustion and disengagement on depressive symptoms from occupational stress, suggesting that while depressive symptoms may relate to learned helplessness, anxiety symptoms are more influenced by over-commitment.

Notably, disengagement did not mediate the stress–anxiety association once exhaustion was modeled concurrently, despite a moderate bivariate correlation with anxiety (r=0.523). Within the Job Demands–Resources framework, exhaustion reflects affective and physiological depletion that aligns closely with the arousal-based phenomenology of anxiety, whereas disengagement captures cognitive detachment or cynicism that has been more frequently linked to depressive outcomes [22,38,39]. Our prior work using the same cohort showed that both exhaustion and disengagement mediated the association between stress and depressive symptoms [29], while in the present study focusing on anxiety, only exhaustion emerged as a significant mediator. This symptom-specific pattern suggests that exhaustion may constitute a more proximal pathway to anxiety, with disengagement functioning as a more distal or context-dependent process. In addition, measurement characteristics may have influenced the results. The CUXOS focuses on psychological and somatic arousal that closely corresponds to exhaustion, whereas disengagement represents cognitive withdrawal from work tasks measured by the OLBI. Beyond these conceptual and measurement explanations, the cultural context of Korean workplaces may further clarify this pattern. In collectivistic and hierarchical organizational settings, employees are often expected to maintain high interpersonal engagement and compliance even under stress [15,16]. As a result, the immediate manifestation of occupational stress may be emotional and physiological depletion, captured as exhaustion, rather than cognitive withdrawal from work. Disengagement may develop more gradually and be more strongly linked to depressive than to anxiety pathways. Taken together, these findings warrant cautious interpretation and motivate longitudinal, multi-method studies to test temporal ordering and construct-specific pathways.

This study suggests that both facets of grit—passion and perseverance—can mitigate the negative effects of exhaustion on anxiety symptoms. Previous research [42-45] has consistently linked high levels of grit to lower stress, burnout, and anxiety, particularly in educational contexts. For instance, grit-passion was negatively correlated with fear of failure [46], indicating that individuals with high grit-passion demonstrate better stress management and a reduced fear of failure due to their commitment and persistence in their careers. Individuals with high levels of grit, characterized by a strong commitment to long-term goals and a continuous willingness to face challenges, exhibit quicker recovery from negative emotions to positive emotions compared with individuals with low levels of grit, even in the face of unfavorable feedback [47]. They also demonstrate lower levels of psychological exhaustion [48] and tend to perceive relatively low stress in adversity, viewing it as an opportunity for personal growth [49]. Overall, gritty individuals are more capable of coping with exhaustion, which reduces their risk of developing anxiety. This study underscores the protective role of grit in reducing anxiety stemming from exhaustion among Korean employees.

This study had several limitations. First, recruiting participants from the workplace mental health assessment at Kangbuk Samsung Hospital may limit the generalizability to all Korean workers. While including diverse occupational groups broadens the scope, it introduces variability due to unique professional demands. This study aimed to explore factors influencing overall employee mental health, not specific groups; future research could categorize participants for detailed analysis. Second, the cross-sectional design limits our ability to determine longitudinal anxiety symptom trajectories, raising the possibility that identified mediators may be covariates rather than true mediators. Longitudinal studies are necessary for deeper exploration of these associations. Additionally, as all major variables in this study were assessed using self-report questionnaires, there is a possibility of shared method variance, which may have inflated the observed associations among stress, burnout, anxiety, and grit. Although we attempted to minimize careless responses, this limitation remains and suggests that our findings should be interpreted with caution. Future research incorporating objective or multimethod assessments, such as physiological stress indicators or supervisor ratings, would provide a more comprehensive understanding of these relationships. Finally, although the large sample size strengthened statistical power, it also means that even very small effects could reach statistical significance. To address this, we placed greater emphasis on effect sizes in interpreting the results. The mediation pathway from exhaustion to anxiety demonstrated a relatively large effect (b>1.0), suggesting clinical and organizational importance. In contrast, the moderation effects of grit were small (b=-0.108 for passion; b=-0.039 for perseverance), yet these should not be dismissed. Even modest reductions in anxiety symptoms can accumulate across large employee populations, leading to meaningful improvements in absenteeism, presenteeism, and productivity at the organizational level. Therefore, while the moderation effects are statistically small, they retain practical significance within workplace mental health interventions.

This study’s comprehensive model highlights the direct and indirect impact of occupational stress on anxiety, mediated by burnout. It also underscores the mitigating role of grit in reducing burnout’s adverse effects on anxiety, contributing to a robust theoretical framework for workplace mental health (Supplementary Figure 1). From a practical standpoint, our findings suggest several actionable strategies for workplace mental health interventions. To alleviate exhaustion, organizations could implement workload adjustments, enforce reasonable working hours, and provide structured recovery programs such as stress management or mindfulness training. To address disengagement, leadership development programs that enhance organizational justice and role clarity, as well as opportunities for career development, may reduce detachment from work. In addition, resilience training, mentoring, and coaching initiatives can foster grit by promoting long-term commitment and perseverance among employees. In collectivistic workplace cultures such as Korea, team-based interventions including peer-support groups and mentoring networks may be particularly effective. Together, these interventions can help organizations move beyond individual-level coping to systemic approaches that prevent burnout and strengthen protective traits like grit. In conclusion, this study provides valuable insights into the dynamics of occupational stress, burnout, grit, and anxiety, offering a foundation for strategies to promote mental well-being in the workplace.

Supplementary Materials

The Supplement is available with this article at https://doi.org/10.30773/pi.2025.0305.

Supplementary Table 1.

Correlations among variables

Supplementary Table 2.

Multiple linear regression analysis of the association between occupational stress types and anxiety symptom severities

Supplementary Figure 1.

Visual representation of grit’s moderating effects on the occupational stress–burnout–anxiety pathways. (A) exhaustion and (B) disengagement. Solid arrows indicate statistically significant pathways, whereas dashed arrows indicate non-significant pathways. Standardized regression coefficients (B) are presented along the significant paths. B′ denotes indirect effects. **p<0.001.