The mediating role of psychological resilience in the relationship between health anxiety and burnout among healthcare workers working in closed environments

Abstract

Background

Burnout is a pervasive problem among healthcare professionals, particularly in high-stress environments where psychological demands and uncertainty are constant. Health anxiety, intensified by occupational stressors and perceived health risks, can increase vulnerability to burnout. However, psychological resilience may serve as a protective mechanism that buffers the impact of anxiety on occupational well-being. This study aimed to examine the mediating role of psychological resilience in the relationship between health anxiety and burnout among healthcare workers.

Methods

A cross-sectional, questionnaire-based design was employed. Data were collected from 348 healthcare professionals working in a university hospital in Türkiye. Standardized instruments were used to measure health anxiety, psychological resilience, and burnout. The hypothesized structural model was tested using structural equation modeling (SEM) with maximum likelihood estimation and bias-corrected bootstrapping (5,000 resamples). Model fit was evaluated using multiple indices, including χ²/df, RMSEA, CFI, TLI, and SRMR.

Results

The final model demonstrated an acceptable fit to the data (χ²/df = 3.83, RMSEA = 0.090, SRMR = 0.066, CFI = 0.89, TLI = 0.86). Health anxiety significantly and positively predicted burnout (β = 0.522, p < .001) and negatively predicted psychological resilience (β = –0.308, p < .001). Psychological resilience negatively predicted burnout (β = –0.210, p < .001). The indirect effect of health anxiety on burnout via resilience was significant (βindirect = 0.065, 95% BC bootstrap CI [0.032, 0.118]), confirming partial mediation. The model explained 38% of the variance in burnout and 9% in psychological resilience.

Conclusions

Health anxiety increases burnout among healthcare professionals, while psychological resilience serves as a partial buffer that mitigates this effect. These findings underscore the importance of enhancing resilience-focused interventions—such as stress management, coping skills, and mindfulness-based training—to protect healthcare workers from the psychological consequences of anxiety and emotional exhaustion. Future research should employ longitudinal designs and cross-cultural samples to validate these relationships and inform evidence-based mental health policies in healthcare institutions. This study highlights the importance of resilience-based interventions for sustaining the psychological well-being of healthcare professionals.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-026-14240-8.

Introduction

Working in closed clinical environments has not only pushed healthcare systems to their limits but has also severely tested the mental and physical resilience of healthcare workers. Burnout has emerged as one of the most common and damaging consequences of this process, as it reduces job performance and negatively affects both patient safety and quality of care [1–3].

Recent empirical evidence highlights that burnout has risen to unprecedented levels among healthcare professionals globally, particularly under conditions of high workload and systemic strain such as those seen during the COVID-19 pandemic [4]. Longitudinal data show that emotional exhaustion, depersonalization, and reduced personal accomplishment, core dimensions of burnout, have significantly increased over time in hospital workers evaluated throughout the pandemic period [5]. Complementing these trends, cross-sectional studies report high prevalence rates of burnout and psychological distress among healthcare workers, with emotional exhaustion and anxiety commonly observed across diverse settings [6]. Alongside burnout, health anxiety has also emerged as a salient psychological burden for healthcare personnel facing prolonged occupational risk, compounding distress in contexts marked by infection risk, extended work hours, and constrained resources [7]. Taken together, these findings underscore that health anxiety and burnout constitute pervasive, empirically documented challenges affecting healthcare workers’ psychological well-being in contemporary clinical environments.

However, little is known about how these mechanisms operate among healthcare professionals working in confined and daylight-limited clinical environments, where environmental stressors may further intensify psychological strain. Studies conducted in hospital settings show that insufficient exposure to natural daylight is associated with higher levels of emotional exhaustion, depersonalization, and depressive symptoms among nurses and operating room staff [8, 9]. Similarly, research examining hospital work environments across European clinical contexts has reported that poor spatial comfort, restricted physical mobility, and inadequate environmental design are robust predictors of burnout and occupational stress among healthcare workers [10]. Evidence-based healthcare design research further indicates that windowless and enclosed clinical environments are associated with impaired mood regulation, heightened physiological stress responses, and reduced psychological well-being among staff, underscoring the critical role of environmental conditions in shaping occupational mental health outcomes [11]. In parallel, studies on circadian disruption resulting from limited daylight exposure and irregular indoor work schedules demonstrate associations with sleep disturbance, affective dysregulation, and cardiovascular risk, which are well-established precursors of burnout and psychological strain in healthcare professionals [12]. Together, these findings substantiate the premise that confined and daylight-deprived healthcare environments represent a significant occupational risk context, thereby justifying focused investigation into their psychological consequences.

Burnout, defined as emotional exhaustion, depersonalization, and reduced personal accomplishment, is prevalent among stressful occupations such as healthcare [13, 14]. Prior research has shown that psychological resilience acts as a critical protective factor, mitigating the adverse impact of occupational stress and preventing burnout [15, 16]. For instance, in a study using structural equation modeling (SEM) among nurses, psychological resilience was found to directly and indirectly reduce burnout, mediating up to 48–87% of its effect [15]. Similarly, psychological flexibility was found to reduce burnout and the intention to leave among critical care nurses [17].

Another important construct that became prominent during the pandemic, especially in closed healthcare environments, is health anxiety—an intense and persistent worry about existing or potential health threats [18]. Elevated health anxiety not only impairs quality of life and work performance but also increases the likelihood of burnout [19, 20]. Studies have reported that healthcare professionals with higher psychological resilience experience less burnout and lower levels of health anxiety and depression [21]. Likewise, research in China has shown that low job satisfaction and psychological well-being (flourishing) play key roles in burnout formation, indicating that both individual and contextual factors interact in shaping occupational well-being [22].

Empirical evidence suggests that health anxiety contributes to burnout both directly, by increasing persistent cognitive preoccupation with bodily symptoms and perceived vulnerability to illness, and indirectly, by eroding psychological resilience and adaptive coping capacity, which in turn amplifies emotional exhaustion and depersonalization [16]. Individuals with lower tolerance for ambiguity and reduced psychological resilience are more prone to stress-induced exhaustion and depersonalization [23]. Parallel findings among university students revealed that resilience negatively predicted pandemic-related burnout, suggesting that resilience serves as a psychological buffer against anxiety and stress [15].

In addition to individual factors, organizational and social support mechanisms also play a vital role [24]. Longitudinal evidence from Canada indicates that despite high burnout rates, strong institutional support and positive social relationships significantly reduced psychological distress [25]. Similarly, job stress and work commitment have been identified as mediating variables linking pandemic severity to burnout [22]. Supportive work environments, emotional safety, and social support systems enhance resilience and reduce burnout levels [26]. Collectively, these findings underscore the importance of protecting the mental health of healthcare workers to sustain both workforce well-being and the quality of healthcare delivery.

Recent studies have consistently demonstrated the protective role of psychological resilience against burnout and health anxiety [27, 28]. However, emerging evidence indicates that healthcare professionals working in enclosed and daylight-limited clinical environments experience disproportionately higher levels of emotional exhaustion, psychological distress, and burnout due to circadian disruption, spatial confinement, and reduced environmental recovery opportunities [8–10]. Despite these documented risks, existing research has largely focused on general hospital settings or pandemic-related stress without specifically examining the psychological mechanisms linking health anxiety, resilience, and burnout in such environmentally constrained contexts. Therefore, the present study addresses this critical gap by empirically investigating these pathways among healthcare workers operating in confined clinical environments.

Addressing this gap, the present study investigates the direct and indirect relationships among health anxiety, psychological resilience, and burnout among healthcare professionals working in enclosed clinical environments in Türkiye, using a structural equation modeling approach. It is hypothesized that psychological resilience partially mediates the relationship between health anxiety and burnout, functioning as a psychological buffer against stress-induced exhaustion. By integrating these constructs into a unified analytical framework, this study aims to advance theoretical understanding of resilience as a mediator and offer practical insights for developing evidence-based interventions to reduce burnout in healthcare institutions.

Method

This section includes information on the model to be followed in the research, the study universe and sample group, the tools and methods to be used in the data collection process, the principles for conducting the research within the framework of ethical rules, and the method by which the obtained data will be analyzed.

Study design and setting

This research was designed as a cross-sectional study conducted among healthcare workers employed in confined hospital environments. Data collection took place between 02 February and 14 February 2024 at Tokat Gaziosmanpaşa University Hospital, Turkey. The study followed the STROBE reporting guidelines for cross-sectional observational studies.

Participants and sampling

The study population included physicians, nurses, and allied healthcare staff working in inpatient units without natural ventilation or limited space. A non-probability convenience sampling strategy was utilized to recruit participants from a tertiary public hospital in Türkiye. All healthcare professionals working in enclosed clinical units during the data collection period were approached through institutional announcements and in-person invitations during work shifts, ensuring broad coverage of eligible staff within the constraints of clinical workload and access limitations.

Sample size was calculated using Cochran’s [29] formula for an unknown population, yielding a minimum of 348 participants, which was achieved in the final dataset. Inclusion criteria were: being a full-time healthcare worker at the hospital and volunteering to participate; exclusion criteria included administrative staff or incomplete responses. The response rate was 31.6% (348 completed questionnaires out of 1100 invited healthcare workers).

The sample consisted of 348 healthcare professionals from a total population of 1,100, yielding a response rate of 31.6%. According to the 10:1 parameter-to-sample ratio rule [30] the sample size was adequate for SEM analyses.

Sample size adequacy was evaluated using G*Power 3.1. For a linear multiple regression (test: R² deviation from zero) with two predictors, the observed R² = 0.38 corresponds to an effect size f² = 0.613. Using α = 0.05, N = 348, and 2 predictors, the post-hoc calculated power (1 − β) is ≈ 1.00. Thus, the study had more than adequate power to detect the observed effects. In addition to the achieved post-hoc power (1 − β ≈ 1.00) based on the observed effect size (f² = 0.613, N = 348, α = 0.05), a sensitivity analysis was conducted using G*Power 3.1 to estimate required sample sizes for different hypothetical effect sizes. For detecting a small effect (f² = 0.02) with 80% power, 396 participants would be required; for a medium effect (f² = 0.15), 68 participants; and for a large effect (f² = 0.35), 38 participants. According to Cohen’s [31] benchmarks, this corresponds to a large effect size, confirming that the observed relationships are both statistically and practically significant. The current sample size (N = 348) therefore provides well above the required power for detecting even medium-sized effects, confirming that the study is statistically well-powered.

Measures

The questionnaire used in this study was not developed by the authors. It consisted of previously validated and published scales:

• Health Anxiety Inventory (HAI): Developed by Salkovskis et al. [32], the scale is a self-report measure consisting of 18 items, rated on a four-point Likert-type scale. The scale was translated and adapted into Turkish by Aydemir et al. [33]. The body dimension, comprising the first 14 items, consists of statements with four-point responses. The other four items constitute the additional dimension, which assumes the mental state of individuals in the event of a serious illness. The total score of the scale is calculated as the arithmetic mean of each item. The first 14 items represent the body dimension and a sample item is: “I generally think that my risk of developing a serious illness is high” are called the “Dimension of Hypersensitivity to Somatic Symptoms.” The other four items will be titled “Dimensions Related to the Negative Consequences of Illness,” and an example item is: “If I were to contract a serious illness, I would feel that my dignity would be completely lost.” The internal consistency coefficient was 0.91, and the test-retest reliability was 0.57. The scale was also found to be valid in assessing anxiety about somatic symptoms.

• Psychological Resilience Scale: The scale was developed by Friborg et al. [34]. The purpose of the scale is to measure individuals’ psychological resilience. Its validity and reliability study in Turkey was conducted by Basım and Çetin [35]. It is a 5-item Likert-type scale consisting of 33 items. Four items constitute the “Structural Style” (example item: “I am at my best when I have a clear goal that I want to achieve.“) dimension, four items constitute the “Perception of the Future” (example item: “The plans I make for the future are difficult to achieve.“) dimension, six items constitute the “Family Cohesion” (example item: “My family’s understanding of what is important in life is different from mine.“) dimension, six items constitute the “Self-Perception” (example item: “When the unexpected happens, I always find a solution.“) dimension, six items constitute the “Social Competence” (example item: “I enjoy being with other people.“) dimension, and seven items constitute the “Social Resources” (example item: “I do not discuss personal matters with anyone.“) dimension. The internal consistency values ​​of the structural equation model for the reliability of the scale were found to be 0.75 for “Perception of the Future”, 0.82 for “Social Competence”, 0.84 for “Social Resources”, 0.80 for “Self-Perception”, 0.86 for “Family Cohesion”, and 0.76 for “Structural Style”.

• Maslach Burnout Inventory (MBI): Originally developed by Maslach and Jackson [36] and adapted into Turkish by Ergin [37], the MBI assesses burnout across emotional exhaustion, depersonalization, and personal accomplishment. Sample items include: “I feel emotionally drained from my work” (emotional exhaustion), “I feel very energetic” (Personal Achievement Dimension) and “Since I started this job, I’ve become more insensitive to people.” (Desensitization Dimension). Cronbach’s α values were 0.88, 0.86, and 0.84, respectively.

All instruments were rated on Likert-type scales. Higher scores indicated greater health anxiety, resilience, and burnout, respectively.

Data collection procedure

Data were collected from healthcare professionals working in closed environments using both online and paper-based questionnaires administered during working hours. Participation was voluntary and anonymous, and written informed consent was obtained from all participants prior to data collection. Ethical approval for the study was obtained from the relevant institutional review board. To reduce potential response bias, participants were assured of confidentiality, and no identifying information was collected.

Measurement reliability and validity

Reliability and validity analyses confirmed the adequacy of all constructs. Cronbach’s α coefficients ranged from 0.80 to 0.92, composite reliability (CR) values exceeded 0.70, and average variance extracted (AVE) values exceeded 0.50, indicating satisfactory internal consistency and convergent validity.

Statistical analysis

All analyses were conducted using AMOS v.24 (IBM Corp., Armonk, NY, USA). Prior to structural modeling, descriptive statistics, reliability coefficients (Cronbach’s α, composite reliability), and confirmatory factor analyses (CFA) were performed. Convergent and discriminant validity were assessed using average variance extracted (AVE) and the Fornell–Larcker criterion.

Data were screened for univariate and multivariate normality. Skewness and kurtosis values for all observed indicators ranged between − 1.25 and + 1.42, indicating mild deviations from normality. Mardia’s multivariate kurtosis coefficient (5.73) suggested slight non-normality in the multivariate distribution. Multivariate outliers were identified using the Mahalanobis D² statistic (p < .001), and 11 extreme cases were removed prior to analysis. Missing values (< 2%) were handled using full information maximum likelihood (FIML). Given the approximate normality of the data and adequate sample size (n = 348), maximum likelihood (ML) estimation was used as the primary estimator [38]. A sensitivity analysis using robust maximum likelihood (MLR) produced nearly identical parameter estimates and fit indices; therefore, ML results are reported for consistency and interpretability. Although alternative estimators such as WLSMV were considered due to the Likert-type nature of the items (1–5), ML estimation was retained because of the continuous treatment of Likert scores and adequate sample size [39].

Structural equation modeling (SEM) was applied to test the hypothesized mediating effect of psychological resilience between health anxiety and burnout. Model fit was evaluated using χ²/df, RMSEA with 90% confidence intervals, CFI, TLI, SRMR, and GFI, following Hu and Bentler’s [40] thresholds. Indirect effects were tested using bias-corrected bootstrap resampling with 5,000 iterations, and mediation was considered significant when the 95% confidence interval did not include zero. Standardized path coefficients (β), effect sizes, and the proportion of the total effect explained by the indirect pathway were also reported.

To assess potential common method variance (CMV), both procedural and statistical remedies were applied. Procedurally, data were collected anonymously, respondents were assured of confidentiality, and items measuring predictor and criterion variables were interspersed to minimize response patterning. Statistically, Harman’s single-factor test indicated that the first factor accounted for 31.2% of the total variance, below the recommended 40% threshold, suggesting that no single factor dominated the data. In addition, a CFA common latent factor test showed no meaningful improvement in model fit (ΔCFI = 0.001, ΔRMSEA = 0.001), further indicating that common method bias was not a major concern.

Ethical considerations

Ethical approval was obtained from Tokat Gaziosmanpaşa University Clinical Research Ethics Committee (approval numbers: 02.02.2024/24-KAEK-023 and 14.02.2024/24-KAEK-029). Institutional permission was granted by the hospital administration (date: 05.02.2024). All procedures complied with the Declaration of Helsinki.

Conceptual framework and research model

This study investigates the relationships among health anxiety, burnout, and psychological resilience among healthcare professionals working in enclosed clinical environments. Specifically, the research examines how health anxiety directly influences burnout and indirectly affects it through the mediating role of psychological resilience. Furthermore, the study explores whether these psychological constructs differ significantly according to key demographic (gender, marital status, educational level, professional specialization) and environmental (exposure to daylight, type of clinical setting) variables.

The research was designed within the framework of a descriptive and correlational model, aiming to identify existing relationships among variables rather than manipulate them experimentally. In accordance with the methodological approach proposed by Karasar [41], this model describes the current state of the phenomena as they naturally occur within the healthcare context.

Within the conceptual model, health anxiety is hypothesized to have a direct positive effect on burnout and a negative effect on psychological resilience, whereas psychological resilience is expected to have a negative effect on burnout. Moreover, psychological resilience is proposed to partially mediate the relationship between health anxiety and burnout, serving as a psychological buffer that mitigates the detrimental influence of health-related concerns on emotional exhaustion.

Accordingly, the model provides an integrative framework that links anxiety-driven psychological processes to occupational outcomes through resilience mechanisms. The hypothesized structural relationships were empirically tested using structural equation modeling (SEM) to determine the strength, direction, and significance of the proposed paths.

Figure 1 Conceptual model illustrating the hypothesized direct and indirect relationships among health anxiety, psychological resilience, and burnout among healthcare professionals. Health anxiety is hypothesized to have a direct positive effect on burnout and a negative effect on psychological resilience, whereas psychological resilience is expected to negatively predict burnout, indicating a partial mediating effect between health anxiety and burnout.

Fig. 1.

Conceptual model. Health Anxiety (HA), Psychological Resilience (PR), Burnout (BO), Mediated pathway: HA → PR → BO

Conceptual model illustrating the hypothesized direct and indirect relationships among health anxiety, psychological resilience, and burnout. Health anxiety is proposed to have a direct positive effect on burnout and a negative effect on psychological resilience, whereas psychological resilience is expected to negatively predict burnout.

Research hypotheses

H1

Health anxiety significantly and positively predicts burnout levels of healthcare professionals.

H2

Health anxiety significantly and negatively predicts the psychological resilience levels of healthcare workers.

H3

Psychological resilience significantly and negatively predicts burnout levels of healthcare professionals.

H4

Psychological resilience plays a partial mediator role in the relationship between health anxiety and burnout.

Results

This section presents the results of confirmatory factor analysis (CFA) and structural equation modeling (SEM), including measurement model tests, descriptive statistics, structural relationships, and mediating effects.

The measurement model demonstrated acceptable fit: χ²(238) = 546.21, χ²/df = 2.30, RMSEA = 0.061 (90% CI [0.055, 0.068]), CFI = 0.911, TLI = 0.903, SRMR = 0.041. All standardized factor loadings were significant (p < .001) and ranged from 0.41 to 0.92, indicating convergent validity. The structural model also exhibited an adequate fit: χ²(238) = 912.87, χ²/df = 3.83, RMSEA = 0.090 (90% CI [0.082, 0.097]), CFI = 0.89, TLI = 0.86, SRMR = 0.066. These indices collectively support the acceptability of the final model for hypothesis testing.

Descriptive statistics

A total of 348 healthcare workers participated in the study, all of whom were employed at a university hospital in Türkiye. The mean age of the participants was 33.10 years (SD = 7.85), and their average professional experience was 10.61 years. Of the total sample, 67.5% were women and 32.5% were men. In terms of marital status, 63.5% were married and 36.5% were single.

Regarding educational status, 48.3% of the participants held a bachelor’s degree, 28.2% had an associate degree, 13.8% were high school graduates, and 9.8% held a postgraduate degree. The majority of participants were nurses or midwives (45.4%), followed by health technicians/technologists (23.6%), healthcare support staff (17.0%), physicians (6.6%), healthcare license holders (5.7%), and healthcare administrators (1.7%).

Analysis of working conditions showed that 94% of participants reported working in closed environments, and 76% stated that they had limited or no exposure to natural daylight during working hours. Descriptive findings further revealed that the mean score of health anxiety was 38.83 (SD = 10.03), the mean burnout score was 71.47 (SD = 10.77), and the mean psychological resilience score was 103.02 (SD = 8.72), indicating moderate levels of burnout and psychological resilience among the participants, along with a relatively elevated level of health anxiety.

Findings regarding the measurement tool

Before testing the structural relationships, the measurement properties of the study variables were examined through reliability and validity analyses. Internal consistency reliability was assessed using Cronbach’s alpha coefficients, and the results indicated that all measurement instruments demonstrated acceptable to high levels of reliability. Specifically, Cronbach’s alpha was 0.91 for the Health Anxiety Inventory, 0.80 for the Psychological Resilience Scale, and ranged between 0.84 and 0.88 for the subdimensions of the Maslach Burnout Inventory, indicating strong internal consistency.

Convergent validity was evaluated using standardized factor loadings, Composite Reliability (CR), and Average Variance Extracted (AVE). All standardized loadings were statistically significant (p < .001) and ranged from 0.57 to 0.86, exceeding the recommended threshold of 0.50. CR values ranged between 0.84 and 0.92, surpassing the acceptable level of 0.70, while AVE values were between 0.52 and 0.58, meeting the minimum criterion of 0.50. These findings demonstrated that the observed variables adequately represented their corresponding latent constructs.

Discriminant validity was also confirmed based on the Fornell–Larcker criterion. The square root of the AVE values for each construct was greater than the inter-construct correlation coefficients, indicating that each latent variable was empirically distinct from the others. Taken together, the results confirmed that the measurement model demonstrated satisfactory reliability, convergent validity, and discriminant validity, supporting its adequacy for further structural equation modeling.

The measurement model consists of three latent factors (psychological resilience, burnout, and health anxiety) and eleven observed variables (structural style, perception of the future, family cohesion, self-perception, social competence, social resources, emotional exhaustion, depersonalization, personal accomplishment, hypersensitivity anxiety, and negative consequences). The path diagram of the measurement model is shown in Fig. 2.

Fig. 2.

Measurement model (CFA Results). PRS: Psychological Resilience Scale, BLS: Burnout Level Scale, HAS: Health Anxiety Scale

Confirmatory factor analysis (CFA) was conducted to assess the factorial validity of the measurement model. All standardized factor loadings were statistically significant (p < .001), indicating that the observed indicators adequately reflected their respective latent constructs. The standardized factor loadings for the Health Anxiety construct ranged from 0.78 to 0.90 across its dimensions (e.g., Negative Consequences, Hypersensitivity-Anxiety). For Psychological Resilience, factor loadings varied between 0.41 and 0.78 across dimensions such as Personal Competence, Family Cohesion, Self-Perception, and Social Resources. The factor loadings for the three subdimensions of Burnout ranged between 0.72 and 0.92 for Emotional Exhaustion, 0.59 and 0.81 for Depersonalization, and –0.28 to 0.62 for Personal Accomplishment (reverse scored). All corresponding t-values exceeded the critical threshold of 1.96 and were statistically significant, supporting the measurement quality of the model. These findings confirm acceptable convergent validity.

The analyses revealed that the standardized path coefficients ranged from − 0.277 to 0.925. Except for the personal accomplishment sub-dimension, the T value of all dimensions was greater than 1.96, and the standardized path coefficients in the measurement model were statistically significant (p < .001).

Confirmatory factor analysis (CFA) was performed to evaluate the factorial validity of the measurement model, which consisted of three latent constructs: health anxiety, psychological resilience, and burnout (with three subdimensions: emotional exhaustion, depersonalization, and personal accomplishment). All observed variables were loaded on their respective latent factors, and all standardized loadings were 708 statistically significant (p < .001).

The standardized factor loadings for the Health Anxiety Inventory ranged between 0.60 and 0.86, indicating strong item representation of the latent construct. The Psychological Resilience Scale showed item loadings between 0.57 and 0.79, while the Maslach Burnout Inventory subdimensions demonstrated loading ranges of 0.61–0.85 for emotional exhaustion, 0.59–0.81 for depersonalization, and 0.61–0.82 for personal accomplishment. All t-values exceeded 1.96, confirming statistical significance at the 0.05 level.

Convergent validity was further supported by high standardized factor loadings and AVE values greater than 0.50, indicating that the items sufficiently explained their respective constructs. Discriminant validity was established as the square roots of AVE values were greater than the inter-construct correlations.

Overall, the CFA results confirmed that the three-factor model provided a satisfactory fit, validating the measurement structure for subsequent structural equation modeling (SEM).

Structural model findings

Figure 3 direct and indirect relationships between variables included in the structural model were analyzed. Health anxiety was considered a predictor of burnout, and psychological resilience was considered a mediating variable in this relationship. Data on the path coefficients for structural equation modeling are shown below.

Fig. 3.

Structural model (SEM results)

Taken together, the findings suggested that healthcare professionals experiencing higher levels of health anxiety are at increased risk of burnout. However, psychological resilience serves as a protective factor that weakens this relationship. These results highlight the importance of resilience-enhancing interventions to mitigate burnout in high-stress healthcare environments.

As shown in Table 1, health anxiety positively predicted burnout (β = 0.522, SE = 0.062, t = 8.401, p < .001; 95% BC CI [0.401, 0.643]), and negatively predicted psychological resilience (β = –0.308, SE = 0.027, t = − 4.039, p < .001; 95% BC CI [–0.362, –0.241]). Psychological resilience negatively predicted burnout (β = –0.210, SE = 0.058, t = − 3.208, p < .001; 95% BC CI [–0.310, –0.119]). The indirect effect of health anxiety on burnout via resilience was significant (βindirect = 0.065, 95% BC CI [0.032, 0.118]), confirming partial mediation. The final model explained 38% of the variance in burnout and 9% in resilience.

Table 1.

Path coefficients table in the structural model

Path	β (Standardized)	SE	t-value	p-value	95% Bootstrap CI	Effect Type
Health Anxiety → Burnout	0.522	0.062	8.401	< 0.001	[0.401, 0.643]	Direct
Health Anxiety → Psychological Resilience	–0.308	0.027	–4.039	< 0.001	[–0.362, –0.241]	Direct
Psychological Resilience → Burnout	–0.210	0.058	–3.208	< 0.001	[–0.310, –0.119]	Direct
Indirect Effect (HA → PR → BO)	0.065	0.021	—	< 0.01	[0.032, 0.118]	Indirect
Total Effect (HA → BO)	0.587	—	—	< 0.001	[0.473, 0.708]	Total

R² (Burnout) = 0.38 — R² (Psychological Resilience) = 0.09

Note. Standardized beta coefficients are reported. Bootstrap confidence intervals were calculated using 5,000 resamples (bias-corrected). HA = Health Anxiety; PR = Psychological Resilience; BO = Burnout

The structural model explained 38% of the variance in burnout and 9% of the variance in psychological resilience. Health anxiety had a significant direct positive effect on burnout and a negative effect on psychological resilience. Psychological resilience partially mediated the relationship between health anxiety and burnout.

Table 1 presents the direct effects estimated in the structural equation model. The findings revealed that health anxiety had a significant positive effect on burnout (β = 0.522, p < .001), indicating that healthcare professionals with higher levels of health-related anxiety were more likely to experience burnout. In addition, health anxiety had a significant negative effect on psychological resilience (β = –0.308, p < .001), demonstrating that increasing levels of anxiety reduced individuals’ psychological coping capacity. Psychological resilience, in turn, had a significant negative effect on burnout (β = –0.210, p < .001), suggesting that resilience served as a protective factor by mitigating burnout levels among healthcare workers. All regression paths were statistically significant, and the bootstrap confidence intervals did not include zero, supporting the robustness of these relationships. These results support the hypothesized model and provide empirical evidence for the mediating role of psychological resilience in the relationship between health anxiety and burnout.

The observed effect sizes were large (f² > 0.35 for the primary paths), and the structural model explained a substantial proportion of variance in burnout (R² = 0.38) and psychological resilience (R² = 0.09), indicating that health anxiety and resilience exert not only statistically significant but also practically meaningful effects on occupational well-being among healthcare professionals.

Table 2 the correlation analysis revealed significant relationships among the study variables. Health anxiety was positively correlated with emotional exhaustion (r = .481, p < .001) and depersonalization (r = .462, p < .001), and negatively correlated with personal accomplishment (r = –.405, p < .001), indicating that higher levels of health anxiety were associated with greater burnout among healthcare workers. Psychological resilience showed a negative correlation with emotional exhaustion (r = –.226, p < .001) and depersonalization (r = –.204, p < .001), and a positive correlation with personal accomplishment (r = .233, p < .001), suggesting that resilience served as a protective factor against burnout. The correlations among the subdimensions of burnout were also moderate to strong; emotional exhaustion and depersonalization were positively correlated (r = .612, p < .001), while both were negatively associated with personal accomplishment. These findings support the hypothesized relationships among variables and justify further structural equation modeling to test direct and indirect effects.

Table 2.

Means, standard deviations, and correlations among study variables

Variable	M	SD	1	2	3	4	5
1. Health Anxiety	38.83	10.03	—
2. Psychological Resilience	103.02	8.72	–0.308***	—
3. Emotional Exhaustion (Burnout-EE)	28.14	6.21	0.481***	–0.226***	—
4. Depersonalization (Burnout-DP)	16.34	4.82	0.462***	–0.204***	0.612***	—
5. Personal Accomplishment (Burnout-PA)*	26.98	5.43	–0.405***	0.233***	–0.388***	–0.374***	—

Note. ***p < .001. Higher scores in PA indicate lower burnout, as the subscale is reverse interpreted relative to EE and DP

According to the findings obtained as a result of the analysis, the acceptance and rejection status of the hypotheses are as follows:

H1, which proposed that health anxiety would positively predict burnout levels, was supported. Health anxiety significantly and positively predicted burnout among healthcare professionals (β = 0.522, SE = 0.062, t = 8.401, p < .001; 95% BC CI [0.401, 0.643]). This finding indicates that as health anxiety increases, burnout levels also rise, reflecting a direct risk association between psychological stress and emotional exhaustion.

H2 predicted that health anxiety would negatively predict psychological resilience. The results confirmed this hypothesis (β = –0.308, SE = 0.027, t = − 4.039, p < .001; 95% BC CI [–0.362, –0.241]). Thus, higher levels of health anxiety were associated with lower levels of resilience, implying that persistent health-related worries reduce individuals’ capacity for adaptive coping.

H3, which posited that psychological resilience negatively predicts burnout, was also supported (β = –0.210, SE = 0.058, t = − 3.208, p < .001; 95% BC CI [–0.310, –0.119]). These results suggest that resilience functions as a protective psychological resource that reduces the likelihood of burnout symptoms among healthcare employees.

H4 examined the mediating role of psychological resilience in the relationship between health anxiety and burnout. The bootstrap analysis revealed a significant indirect effect of health anxiety on burnout via resilience (βindirect = 0.065, 95% BC CI [0.032, 0.118]). Because the direct effect of health anxiety on burnout remained significant (βdirect = 0.522, p < .001), partial mediation was confirmed. The indirect pathway accounted for approximately 11% of the total effect (βtotal = 0.587), indicating that resilience partially buffers the detrimental impact of health anxiety on burnout.

Overall, all four hypotheses (H1–H4) were statistically supported, confirming the proposed conceptual model. The final model explained 38% of the variance in burnout and 9% of the variance in psychological resilience, demonstrating satisfactory explanatory power for psychological and occupational outcomes among healthcare workers.

Discussion

This study tested a theoretically grounded structural model examining the relationships among health anxiety, psychological resilience, and burnout among healthcare professionals working in confined clinical environments. The overall structural model demonstrated acceptable fit and explained a substantial proportion of variance in burnout (38%), indicating that the combined influence of cognitive-emotional vulnerability (health anxiety) and adaptive psychological resources (resilience) provides a meaningful explanatory framework for occupational distress in enclosed healthcare contexts. Importantly, the model supports a dual-process mechanism whereby burnout is shaped both directly by anxiety-driven threat appraisals and indirectly through erosion of resilience-based coping resources.

Current findings indicate that health anxiety is significantly associated with burnout, and psychological resilience partially mediates this relationship among healthcare workers employed in enclosed environments. These results suggest that environmental stressors in healthcare settings (such as limited daylight exposure, spatial constraints, and suboptimal workspace design) exacerbate psychological stress and occupational burnout. Evidence in the literature supports these findings, showing that insufficient access to natural daylight and windowless clinical environments are associated with higher levels of emotional exhaustion and depersonalization among healthcare workers, particularly in high-density settings such as operating rooms [8, 9]. Consistent with these findings, the current study demonstrates that healthcare workers predominantly employed in enclosed environments report high levels of health anxiety and burnout, suggesting that environmental deprivation may intensify perceived vulnerability to illness and occupational stress.

Environmental and occupational health research has consistently shown that inadequate physical work environments, including insufficient natural light and poor spatial conditions, are associated with higher emotional exhaustion and reduced occupational well-being among healthcare staff [8, 11]. Moreover, disruption of circadian rhythms due to limited daylight exposure has been linked to impaired sleep quality, affective dysregulation, and heightened vulnerability to stress-related outcomes in healthcare workers, which are well-established predictors of burnout [12]. Similarly, assessments of the comfort of hospital workspaces indicate that deficiencies in environmental design, including lack of visual privacy, poor ergonomic conditions, and inadequate spatial comfort, are associated with higher burnout symptoms and decreased occupational well-being among healthcare workers [10].

Although RMSEA and incremental fit indices (CFI/TLI) indicate borderline acceptable model fit, these values ​​are considered acceptable in complex SEM models containing multiple latent constructs and indicators, particularly in the context of applied health research where strict cutoff criteria may be overly conservative [42, 43].

The study provides empirical support for the hypothesized relationships between health anxiety, psychological resilience, and burnout among healthcare professionals working in confined clinical settings. All four hypotheses were supported, indicating that resilience plays a protective yet partial mediating role in the link between health anxiety and burnout. The results provide robust empirical evidence supporting all proposed hypotheses. Consistent with H1, health anxiety was found to be a significant positive predictor of burnout. This result aligns with prior studies showing that sustained anxiety related to health threats or infection risk can heighten emotional exhaustion and depersonalization in healthcare settings [44–47]. In environments with high uncertainty and continuous exposure to patient suffering, anxiety may escalate job stress and resource depletion, ultimately leading to burnout. Importantly, the magnitude of these associations was substantial, with large effect sizes and high explained variance in burnout, underscoring that health anxiety and psychological resilience constitute clinically meaningful determinants of occupational functioning rather than merely statistically detectable correlates.

Supporting H2, health anxiety was negatively associated with psychological resilience. This finding is congruent with previous literature suggesting that anxiety reduces self-efficacy and adaptive coping capacities, thereby undermining resilience [48–50]. When healthcare workers perceive themselves as vulnerable to illness or incapable of controlling health-related risks, their psychological flexibility tends to decline.

H3 further confirmed the protective role of resilience against burnout. Individuals with higher levels of psychological resilience experience less emotional exhaustion and depersonalization and maintain a greater sense of personal accomplishment [51–53]. This finding supports the stress-buffering perspective, suggesting that resilient healthcare professionals can manage occupational challenges more effectively.

Finally, H4 demonstrated that psychological resilience partially mediates the relationship between health anxiety and burnout. The mediation pattern indicates that while health anxiety exerts a direct detrimental impact on burnout, part of this relationship operates indirectly through reduced resilience. In other words, resilience serves as a partial buffer that mitigates—but does not fully eliminate—the adverse effects of health anxiety on burnout [54–56]. This partial mediation highlights the dual role of resilience: as both an individual protective factor and a mechanism through which anxiety influences occupational outcomes.

The findings suggest that healthcare organizations should prioritize resilience-oriented interventions—such as cognitive-behavioral training, mindfulness-based stress reduction, and structured peer support programs—to mitigate the psychological burden of health anxiety and prevent burnout. Future research should employ longitudinal designs and cross-cultural samples to validate the temporal and contextual stability of these findings.

Conclusion

The present study provides empirical evidence that health anxiety is a critical psychological factor influencing burnout among healthcare professionals, and that psychological resilience serves as a partial mediator in this relationship. Consistent with theoretical perspectives on stress and coping, the findings indicate that elevated levels of health anxiety heighten emotional exhaustion and depersonalization, whereas resilience mitigates these adverse effects by enabling adaptive coping and emotional regulation. In other words, resilience functions as a protective psychological resource that buffers, but does not completely eliminate, the detrimental influence of anxiety on burnout.

By integrating health anxiety, psychological resilience, and burnout into a single structural model, this research contributes to a more comprehensive understanding of psychological mechanisms underlying occupational well-being in healthcare contexts. The results underscore the importance of strengthening resilience-oriented interventions—such as stress management training, cognitive-behavioral coping programs, and mindfulness-based practices—to enhance healthcare workers’ ability to cope with persistent health-related concerns and prevent burnout. From an organizational standpoint, these findings emphasize the need for hospital administrators to monitor health anxiety symptoms, provide supportive supervision, and design workplace environments that promote emotional safety and psychological recovery.

Despite its valuable contributions, this study has certain limitations. Although all variables were assessed via self-report instruments, CFA-based tests indicated that a single-factor model demonstrated poor fit, suggesting that common method variance was unlikely to substantially bias the observed relationships. The cross-sectional design precludes causal inference; therefore, future research should adopt longitudinal or experimental approaches to validate these relationships over time. Additionally, the data were collected from healthcare professionals working in a single hospital context, which may limit generalizability. Future studies are encouraged to replicate this model in different cultural and organizational settings, explore potential moderating factors such as gender or profession type, and examine intervention-based strategies to enhance resilience and reduce burnout.

In conclusion, this study highlights psychological resilience as a vital mechanism linking health anxiety and burnout among healthcare workers. Fostering resilience not only supports individual mental well-being but also contributes to the sustainability of healthcare systems facing chronic stress and uncertainty. The findings provide a theoretical and practical foundation for developing preventive strategies and mental health policies aimed at protecting healthcare professionals’ psychological resources and improving their long-term occupational health.

This study employed a cross-sectional design, which precludes causal inferences among health anxiety, burnout, and resilience. The observed relationships should therefore be interpreted as associations rather than cause–effect links. Future research using longitudinal or experimental designs could better clarify the directionality and potential causal mechanisms underlying these relationships.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

HA

Health Anxiety

PR

Psychological Resilience

BO

Burnout

EE

Emotional Exhaustion

DP

Depersonalization

PA

Personal Accomplishment

SEM

Structural Equation Modeling

CFA

Confirmatory Factor Analysis

AVE

Average Variance Extracted

CR

Composite Reliability

CI

Confidence Interval

SD

Standard Deviation

SE

Standard Error

β

Standardized Regression Coefficient (Beta)

χ²/df

Chi-square to Degrees of Freedom Ratio

CFI

Comparative Fit Index

TLI

Tucker–Lewis Index

RMSEA

Root Mean Square Error of Approximation

SRMR

Standardized Root Mean Square Residual

NFI

Normed Fit Index

IFI

Incremental Fit Index

GFI

Goodness of Fit Index

AGFI

Adjusted Goodness of Fit Index

α

Cronbach’s Alpha (Reliability Coefficient)

f²

Effect Size (Cohen’s f-squared)

R²

Coefficient of Determination

Author contributions

Study concept and design: AÇ, and İÇ; acquisition of data: AÇ, and İÇ; analysis and interpretation of data: AÇ; drafting of the manuscript: AÇ; critical revision of the manuscript: AÇ, and İÇ; statistical analysis: AÇ; and study supervision: AÇ, and İÇ.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets generated and/or analysed during the current study are not publicly available in accordance with institutional data protection policies, but are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Ethics Approval and Consent to Participate: Ethics committee approval for the study was obtained in accordance with the Declaration of Helsinki, with the decision of the Tokat Gaziosmanpaşa University Graduate Education Institute, dated February 2, 2024, and numbered 394121. Approval was also obtained with the letter of the Tokat Gaziosmanpaşa University Health Research and Application Center, dated February 14, 2024, and numbered 399109.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.