Occupational Stressors and Mental Health–Related Quality of Life Among Hospital Nurses in Northern Border Region Saudi Arabia: A Multi‐Site Cross‐Sectional Study

ABSTRACT

Aim

To examine the association between occupational stress and mental health–related quality of life (HRQoL) among nurses in the Northern Border Region of Saudi Arabia and to determine whether this relationship varies across professional and clinical subgroups.

Design

A cross‐sectional, correlational study.

Methods

A stratified sample of 257 registered nurses and nurse managers was recruited from three hospitals in the Northern Border of Saudi Arabia (53.2% female, 86.0% staff nurses and 44.0% with > 10 years of experience). Occupational stress was assessed using a culturally adapted and forward‐backward translated version of the Nursing Stress Scale (NSS), which measures seven stress domains. Mental HRQoL was evaluated using the Mental Component Summary (MCS) of the 12‐item Short Form Health Survey (SF‐12). Data were analysed using descriptive statistics, independent samples t‐tests, one‐way analysis of variance with post hoc comparisons, Pearson correlations and four‐step hierarchical linear regression.

Results

The mean MCS score was 52.01 (SD = 8.12), slightly above the SF‐12 normative mean of 50. All seven occupational stress domains were significantly and negatively associated with MCS scores (r = −0.22 to −0.33, all p < 0.001). Workload was the highest‐rated stressor (M = 2.91, SD = 0.65), whereas nurse–physician conflict demonstrated the strongest inverse correlation with mental HRQoL (r = −0.33). The final hierarchical regression model explained 34% of the variance in MCS scores (R ² = 0.34, adjusted R ² = 0.30). Occupational stress domains contributed an additional 15% of explained variance beyond demographic and professional characteristics (ΔR ² = 0.15, F‐change [7, 244] = 7.85, p < 0.001). No significant interaction effects were identified.

Patient or Public Contribution

No Patient or Public Contribution.

Abbreviations

CFA

confirmatory factor analysis

CI

confidence interval

COR

Conservation of Resources

EF

exploratory factor analysis

HRQoL

health‐related quality of life

I‐CVI

Item‐level Content Validity Index

JD‐R

Job Demands–Resources

MCS

Mental Component Summary

NSS

Nursing Stress Scale

PCS

Physical Component Summary

S‐CVI

Scale‐level Content Validity Index

SD

standard deviation

SF‐12

12‐item Short Form Health Survey

SPSS

Statistical Package for the Social Sciences

STROBE

Strengthening the Reporting of Observational Studies in Epidemiology

VIF

variance inflation Factor

1. Background

Occupational stress is no longer conceptualized as a private psychological burden carried by individual clinicians, but as a systems‐level phenomenon that propagates across care teams, institutional performance and population health outcomes. It arises when job demands persistently exceed an individual's available resources, coping capacity or perceived control, and its consequences in healthcare settings extend well beyond individual workers to encompass team functioning, patient safety, service quality and staff retention (Leger et al. 2021; Werke and Weret 2023). Work‐related stress is widely recognized as a major occupational health concern that adversely affects employee well‐being, job performance and organizational outcomes. The World Health Organization (WHO) defines work‐related stress as the response that may occur when workers are confronted with work demands and pressures that are not aligned with their knowledge and abilities and that challenge their capacity to cope (WHO, n.d.). The International Labor Organization (ILO) identifies psychosocial hazards, including excessive workload, low job control, inadequate social support, role ambiguity and interpersonal conflict, as critical workplace factors that contribute to work‐related stress and require systematic organizational assessment and management (ILO 2016). Similarly, the European Agency for Safety and Health at Work (EU‐OSHA) recognizes psychosocial risks and work‐related stress as among the most significant occupational health challenges, contributing to absenteeism, presenteeism, reduced productivity and impaired worker well‐being (EU‐OSHA 2022). In the United States, the National Institute for Occupational Safety and Health (NIOSH) identifies job stress as a harmful physical and emotional response resulting from a mismatch between job demands and workers' capabilities, resources or needs, with substantial implications for worker health, safety and organizational sustainability (NIOSH 2020). Within Saudi Arabia, healthcare workforce development and nursing sustainability have been identified as strategic priorities under Vision 2030 and the National Transformation Program, reflecting the need to strengthen the nursing workforce, improve retention and enhance the quality of healthcare delivery (Alluhidan et al. 2020). International nursing organizations converge on the same position: the International Council of Nurses (2021) identifies nurses' mental health, safe staffing and workforce well‐being as inseparable from patient safety, and the WHO (2020), in its State of the World's Nursing report, prioritizes investment in decent nursing jobs, manageable workloads, education and leadership as a global workforce imperative.

Among all healthcare professions, nursing occupies a particularly exposed position: it integrates acute physical workload, sustained emotional labour, high clinical responsibility, time pressure, repeated exposure to patient suffering and constant interpersonal demands (Okuhara et al. 2021; Sarafis et al. 2016). When such pressures accumulate over time without adequate organizational buffers, nurses may experience deterioration in health‐related quality of life (HRQoL), psychological distress, reduced performance capacity and diminished ability to provide compassionate and safe care (Babapour et al. 2022; Bakker and Demerouti 2017). In an era in which global nursing shortages are projected to intensify through 2030, understanding the specific mechanisms by which occupational stressors erode mental HRQoL has shifted from a workforce concern to a patient‐safety imperative.

Occupational stress remains a significant challenge in the nursing profession, affecting both workforce well‐being and quality of care. Nurses frequently experience high levels of work‐related stress due to demanding workloads and complex clinical responsibilities (Aserri et al. 2021; Yan et al. 2022). Evidence indicates that occupational stress is associated with poorer psychological well‐being, reduced quality of life, emotional exhaustion and impaired caring behaviours among nurses (Babapour et al. 2022; Yan et al. 2022). Furthermore, elevated stress and burnout have been linked to lower job satisfaction, increased intention to leave, higher turnover, and adverse organizational and patient outcomes (Lasater 2024; Okuhara et al. 2021; Sarafis et al. 2016). Therefore, implementing evidence‐based strategies to mitigate occupational stress is essential for improving nurse retention, workforce sustainability, and the delivery of safe, high‐quality patient care (Lasater 2024; Okuhara et al. 2021).

A substantial international literature documents that occupational stress among nurses is associated with poorer mental health outcomes, including emotional exhaustion, reduced subjective well‐being and clinically meaningful declines in HRQoL (Sarafis et al. 2016; Aserri et al. 2021; Yan et al. 2022). Common stressor domains include excessive workload, inadequate staffing, role uncertainty, insufficient collegial and supervisory support, and conflict within clinical hierarchies, particularly with physicians (Gray‐Toft and Anderson 1981; Okuhara et al. 2021). However, the empirical literature has increasingly moved beyond mere prevalence estimation towards a more discriminating question: which stressor domains, under which organizational conditions, generate the most corrosive effects on nurses' mental health? Answering this question requires a shift from descriptive epidemiology to theoretically grounded mechanism testing.

1.1. Theoretical Framework

This study is grounded in two complementary frameworks. The Job Demands–Resources (JD‐R) model posits that high job demands, when insufficiently buffered by organizational or interpersonal resources, lead to psychological strain and diminished well‐being (Bakker and Demerouti 2017). Consistent with this architecture, the WHO (2022) guidelines on mental health at work organize their recommendations around precisely the demand–resource logic of the JD‐R model, prioritizing organizational‐level interventions, workload modulation, role clarity, supportive management practices and structured manager training, over individual‐level measures alone, thereby endowing the theoretical apparatus of this study with direct policy correspondence.

Conservation of Resources (COR) theory further explains that sustained depletion of valued resources, such as autonomy, social support or professional efficacy, elicits cumulative stress responses and disproportionate adverse outcomes (Hobfoll 2011; Hobfoll et al. 2018). Accordingly, the WHO (2022) guidelines designate safeguards for workers facing hardship as a core recommendation, the ILO (2016) framework conceptualizes psychosocial hazards as collectively produced and collectively remediable, and the NIOSH Total Worker Health program (NIOSH, n.d.) advocates integrated interventions that address the organization of work rather than individual coping alone, positions that jointly frame resource depletion as a preventable occupational exposure rather than an unavoidable personal vulnerability.

A key refinement within the JD‐R framework is the challenge–hindrance distinction proposed by Crawford et al. (2010). Challenge demands (e.g., workload) may stimulate engagement, whereas hindrance demands (e.g., interpersonal conflict, inadequate support) obstruct goal attainment and are more consistently linked to poor psychological outcomes. Applied to nursing, this synthesis generates the central proposition of the present study: hindrance stressors, such as nurse–physician conflict and lack of staff support, are expected to show stronger associations with impaired mental health–related quality of life (HRQoL) than more frequently reported challenge demands like workload. This anticipated dissociation between stressor frequency and psychological impact provides the study's primary conceptual contribution. This distinction converges with the NIOSH (2020) position that modifiable organizational stressors, as opposed to clinically inherent demands, constitute the appropriate target of workplace intervention and with EU‐OSHA (2022) guidance that psychosocial risk assessment must differentiate inherent occupational exposures from modifiable organizational hazards.

1.2. The Saudi Arabian Context

This framework is particularly salient in the Saudi Arabian healthcare context, where rapid system expansion under Vision 2030 and the National Transformation Program has intensified service demands amid persistent nursing shortages and reliance on expatriate staff (Alluhidan et al. 2020; Alsadaan et al. 2021). The resulting multinational, linguistically diverse workforce may heighten hindrance‐type stressors, including role ambiguity and interprofessional conflict, beyond levels typical of more homogeneous Western settings. Seen through the lens of the WHO Global Strategy on Human Resources for Health (WHO 2016), which projects a global shortfall of approximately 18 million health workers by 2030 and identifies retention in high‐demand systems as a central policy lever, the Saudi experience exemplifies the accelerating demand–capacity tension confronting health systems worldwide.

These structural pressures coexist with sociocultural norms related to family roles, gender expectations and religious practices, which may shape stress perception and coping in ways distinct from Western contexts (Aserri et al. 2021; Alharbi and Alshehry 2019; Alsadaan et al. 2021). In this respect, the composition and conditions of the Saudi nursing workforce also engage the WHO Global Code of Practice on the International Recruitment of Health Personnel (WHO 2010), which commits destination countries to safeguarding the well‐being and equitable treatment of internationally recruited health workers, a commitment with direct implications for the psychosocial working conditions of the predominantly expatriate nursing workforce studied here.

Notably, it remains unclear whether internationally observed gender‐related stress patterns are replicated within Saudi hospital nursing, given the country's unique institutional and cultural environment (Al Mutair et al. 2022).

1.3. Research Gap and Rationale for a Northern Border Region Focus

Although prior Saudi studies have established that occupational stress is prevalent among hospital nurses, several gaps remain. Most research has focused on stress prevalence or burnout without linking specific stressor domains to validated mental HRQoL outcomes. Subgroup analyses by gender, role or clinical setting have often been limited, constraining inference regarding differential risk profiles. The present study responds to this gap and aligns with the WHO/ILO priority of strengthening national‐level evidence on psychosocial occupational hazards (WHO and ILO 2021), as well as with ILO (2016) calls for systematic, context‐specific surveillance of workplace stress in under‐represented settings; regionally disaggregated evidence of this kind is a prerequisite for targeted, risk‐based intervention.

Moreover, Saudi nursing research is geographically concentrated in the central and western regions, leaving northern settings underrepresented despite differences in workforce composition, expatriate‐to‐national staffing ratios, patient catchment characteristics and healthcare infrastructure (Aserri et al. 2021; Alharbi and Alshehry 2019; Alsadaan et al. 2021). The Northern Border Region, characterized by a dispersed population, heavier reliance on expatriate nurses and fewer tertiary facilities, offers a strategically informative context for examining how domain‐specific occupational stressors translate into mental HRQoL outcomes. Prior studies have also not systematically examined whether gender, professional role or clinical setting modifies these associations or functions as an additive source of risk.

1.4. Operational Definitions and Study Aim

To ensure conceptual precision, two constructs were operationally defined. Occupational stress was defined, consistent with Gray‐Toft and Anderson (1981) and Okuhara et al. (2021), as nurses' perceived frequency and burden of work‐related demands across seven domains: death and dying, conflict with physicians, inadequate preparation, lack of support, conflict with other nurses, workload and uncertainty concerning treatment. Mental health–related quality of life (mental HRQoL) was defined in line with Ware Jr. et al. (1996) as self‐reported psychological and social functioning, operationalized via the Mental Component Summary (MCS) of the SF‐12. This domain‐specific operationalization addresses limitations of earlier Saudi research that relied on global or single‐item indicators.

Guided by the challenge–hindrance refinement of the Job Demands–Resources (JD‐R) model, this multi‐site cross‐sectional study advances three exploratory, directional propositions rather than confirmatory hypotheses:

• Workload will be the most frequently reported stressor; however, hindrance stressors, particularly nurse–physician conflict and lack of staff support, will show the strongest associations with poorer mental HRQoL.

• Staff nurses will report poorer mental HRQoL than nurse managers, reflecting higher demands and more limited resources.

• Nurses in high‐acuity settings will report greater stress exposure and poorer mental HRQoL than those in lower‐acuity units.

These propositions serve to structure interpretation and guide future hypothesis‐driven research.

1.5. Contribution and Significance

This study contributes conceptually, methodologically and practically by examining associations between seven theoretically grounded stressor domains and a validated measure of mental HRQoL in a multi‐site sample from an understudied Saudi region. Conceptually, it extends the application of the challenge–hindrance JD‐R framework (Crawford et al. 2010; Bakker and Demerouti 2017) to a non‐Western nursing workforce. Methodologically, it combines psychometric evaluation of the Nursing Stress Scale (Gray‐Toft and Anderson 1981) with hierarchical regression to assess the incremental contribution of distinct stressors beyond demographic and role‐related factors. This study contributes to the evidence base on occupational stress among nurses and aligns with international priorities for promoting workplace mental health and managing psychosocial risks. The WHO (2022) emphasizes evidence‐based approaches to protecting workers' mental health and well‐being. Likewise, the ILO (2016), the EU‐OSHA (2022) and the NIOSH (2020) advocate for systematic assessment, surveillance and organizational interventions to address occupational stress and improve workforce health. Accordingly, this study provides context‐specific evidence that may inform strategies to enhance nurses' well‐being, retention, and workforce sustainability.

Practically, findings inform Vision 2030–aligned workforce strategies by identifying stressors most strongly linked to mental HRQoL impairment, supporting targeted interventions in interprofessional relations, staff support, and acuity‐responsive staffing (Alluhidan et al. 2020; Alsadaan et al. 2021; Al Mutair et al. 2022). Moreover, by mapping region‐specific stressor profiles onto a validated mental HRQoL outcome, the study operationalizes the surveillance function called for by the WHO (2022) guidelines and EU‐OSHA (2022) risk‐assessment frameworks, offering Saudi policymakers an empirical baseline against which the effectiveness of Vision 2030–aligned workforce interventions can be monitored.

2. Methods

2.1. Study Design and Setting

This study used a descriptive, multi‐site cross‐sectional design to examine associations between specific occupational stressor domains and mental health–related quality of life (HRQoL) among hospital‐based nurses in the Northern Border Region of Saudi Arabia. Study design, conduct, and reporting adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE; Appendix S1) guidelines for cross‐sectional studies (Von Elm et al. 2008). In doing so, the study responds to the WHO (2022) recommendation that surveillance of psychosocial working conditions form an integral component of national occupational health systems, extending such surveillance to a region not previously characterized.

Data collection was conducted from April to October 2024 in three Ministry of Health public‐sector hospitals in the Northern Border Region, including two secondary referral hospitals (Hospitals A and B; 220 and 180 beds, respectively) and one tertiary referral hospital (Hospital C; 310 beds). Collectively, these hospitals constitute the primary inpatient‐care infrastructure serving the region's resident and expatriate populations.

The data‐collection period excluded the peak Hajj season but overlapped with Ramadan; participants recruited during Ramadan were flagged to allow sensitivity analyses. No hospital reported a declared COVID‐19 or other public health surge during the study period, and findings therefore reflect routine clinical operating conditions.

2.2. Participants and Sampling

The study population comprised registered nurses and nurse managers actively employed in clinical, administrative or quality‐assurance roles across the three participating hospitals.

Inclusion criteria were as follows: (i) current active nursing practice at one of the three study hospitals; (ii) a minimum of 1 year of professional nursing experience; (iii) sufficient fluency in Arabic or English to complete the study questionnaires; and (iv) cognitive and physical capacity to provide informed consent. Exclusion criteria were as follows: (i) long‐term sick leave, maternity leave or other documented absence exceeding four consecutive weeks within the data‐collection window; (ii) nursing students, interns or agency‐contract nurses without continuous institutional affiliation; and (iii) senior nursing administrators holding non‐clinical executive appointments, to preserve comparability of the staff‐nurse and nurse‐manager subgroups.

2.2.1. Sampling Strategy

A stratified quota‐sampling strategy was employed, with strata defined by hospital (three levels) and clinical work area (critical care, inpatient ward and administrative/quality‐assurance). Within each stratum, participants were recruited consecutively as availability permitted, without randomization. This approach is therefore more accurately characterized as stratified quota sampling rather than stratified probability sampling; the earlier label of ‘stratified non‐probability sampling’ is retained in the manuscript's audit trail but should be read in this more precise methodological sense. The implications of this sampling procedure for generalizability are explicitly addressed in the Limitations (Section 4.6).

Participant flow is reported in Figure 1 and in Section 3.1 of the Results. Of approximately 428 eligible nurses invited, 290 consented and returned a questionnaire (consent rate = 67.8%); 33 questionnaires were excluded because they exceeded the pre‐specified 20% item‐level missingness threshold or failed embedded attention checks, yielding the final analytic sample of N = 257 (final response rate = 60.0%).

FIGURE 1.

Participants flow diagram.

2.3. Sample‐Size Calculation

The target sample size was determined a priori using G*Power 3.1.9.7 (Faul et al. 2009), specifying a two‐tailed α = 0.05, minimum statistical power of 1 − β = 0.80, a small‐to‐medium anticipated effect size for hierarchical multiple regression (f ² = 0.15, consistent with Cohen's (1988) convention and with meta‐analytic effect estimates for occupational‐stress‐to‐mental‐health associations in nursing samples; Yan et al. 2022; Podsakoff et al. 2023), and a maximum of 14 predictors in the full model. Under these specifications, the minimum required sample was N = 118 to detect the overall R ² of the fully specified model. The achieved sample of N = 257 therefore exceeded the minimum requirement by a factor of 2.18 for main‐effects detection.

The a priori calculation was explicitly calibrated for main‐effects detection and not for the detection of two‐way interaction terms. Standard heuristics (Aiken and West 1991) indicate that interaction terms in moderated regression typically require 2–4 times the sample size needed for comparable main effects, so the present sample is underpowered for definitive moderation testing. This limitation was anticipated a priori, is reported transparently here (rather than only in the Limitations section), and directly conditions the interpretation of the three interaction terms reported in Section 3.7 of the Results.

2.4. Measures

2.4.1. Demographic and Professional Characteristics

A structured questionnaire collected data on age, gender, marital status, highest educational qualification, years of nursing experience, professional role (staff nurse or nurse manager), primary work area (critical care, inpatient ward or administrative/quality assurance) and prior occupational‐safety training.

2.4.2. Occupational Stress: Nursing Stress Scale

Occupational stress was measured using the Nursing Stress Scale (NSS; Gray‐Toft and Anderson 1981), a 34‐item self‐report instrument comprising seven subscales: death and dying, conflict with physicians, inadequate preparation, lack of support, conflict with other nurses, workload and uncertainty concerning treatment. Items are rated on a 4‐point frequency scale (1 = never/rarely stressful to 4 = very frequently stressful), with higher scores indicating greater stress exposure. The original 34‐item structure was retained to preserve the scale's established seven‐domain factor architecture (Gray‐Toft and Anderson 1981; Okuhara et al. 2021). The seven NSS domains map closely onto the psychosocial hazard taxonomy elaborated by the ILO (2016) and EU‐OSHA (2022), including workload, interpersonal conflict, role ambiguity and deficient social support, supporting the construct coverage of the instrument against contemporary international risk‐assessment frameworks.

2.4.2.1. Translation and Cultural Adaptation

The Arabic NSS was produced using forward–backward translation consistent with WHO guidelines (WHO 2023) and ISPOR recommendations (Wild et al. 2005). Procedures included independent translations, expert‐panel reconciliation, blinded back‐translation and cognitive debriefing with 15 nurses, resulting in minor linguistic refinements without altering construct content.

2.4.2.2. Content Validity

Expert review by seven nursing academics and senior clinicians yielded item‐level CVIs ≥ 0.80, modified kappa values of 0.67–0.97, and an S‐CVI/Ave of 0.90, exceeding recommended thresholds (Polit and Beck 2021).

2.4.2.3. Structural Validity

Exploratory factor analysis (EFA) conducted on half the sample (n = 128) using principal‐axis factoring with Promax rotation supported a seven‐factor solution (KMO = 0.91; Bartlett's test p < 0.001), accounting for 87.1% of cumulative variance, consistent with parallel analysis and scree‐plot inspection (Horn 1965). Confirmatory factor analysis (CFA) on the remaining sample (n = 129) produced marginal overall fit: χ ²(506) = 1088.4, p < 0.001; RMSEA = 0.077 (90% CI [0.071, 0.083]); SRMR = 0.060; TLI = 0.827; CFI = 0.816 (Hu and Bentler 1999). Despite marginal CFA indices, the model was retained based on extensive prior validation (Gray‐Toft and Anderson 1981; French et al. 2000; Milutinović et al. 2012), avoidance of data‐driven re‐specification and strong internal consistency.

This limitation is addressed in Section 4.6.

2.4.2.4. Internal Consistency

Reliability was high (total NSS α = 0.96; subscales α = 0.85–0.95; ω = 0.96), with estimates interpreted in light of the multidimensional scale structure.

2.4.3. Mental and Physical HRQoL: SF‐12

Mental and physical HRQoL were assessed using the SF‐12v1 (Ware Jr. et al. 1996), administered in its licensed Arabic version with established psychometric support (Haddad et al. 2021). The survey yields norm‐based Physical (PCS) and Mental (MCS) Component Summary scores (mean = 50, SD = 10), with higher scores indicating better functioning. Scores were calculated using US‐based norms due to the absence of Saudi reference values. The MCS was specified a priori as the primary outcome. Implications of norm use and lack of formal measurement‐invariance testing are discussed in Section 4.6.

2.5. Data‐Collection Procedure

Data collection was coordinated through nursing‐administration liaisons at each site. Participants completed questionnaires either in paper format or via a secure, password‐protected institutional online platform, requiring approximately 15–20 min.

To minimize response bias and perceived coercion, questionnaires were distributed and returned using sealed envelopes handled by an independent research assistant, confidentiality assurances were clearly communicated, and data were collected during breaks or off‐shift periods. No incentives were provided.

2.6. Ethical Considerations

Ethical approval for the study was obtained from the Local Research Ethics Committee at Northern Border University (reference no. HAP‐09‐2023‐43), following review during the committee meeting held on 30 November 2023, with formal approval issued on 5 December 2023. In addition, administrative permissions were secured from the Directors of Nursing at each of the participating hospitals prior to data collection.

The study was conducted in accordance with the principles of the Declaration of Helsinki (World Medical Association 2013). Written informed consent was obtained from all participants before questionnaire administration. Participation was entirely voluntary, and participants were informed of their right to withdraw at any stage without any repercussions for their employment or professional status.

Confidentiality and anonymity were strictly maintained by assigning unique study codes to all questionnaires, storing consent forms separately from response data and limiting access to identifiable information to the principal investigator only. No personally identifiable data were collected or included in the final analytical dataset.

2.7. Statistical Analysis

Analyses were conducted in SPSS v28.0 using a pre‐specified analytic plan. Two‐tailed tests (α = 0.05) were applied, with exact p values reported. Item‐level missingness was < 5% and handled via listwise deletion; sensitivity analyses using pairwise deletion produced negligible coefficient changes (Δβ ≤ 0.04). Distributional assumptions, variance homogeneity and multicollinearity were assessed using standard diagnostics (all VIFs < 5).

Before Pearson correlation analyses, the assumptions of normality and linearity were evaluated using Kolmogorov–Smirnov tests (p > 0.05 for all continuous variables), Q–Q plot inspection, and scatterplot examination between each stressor domain and the MCS. These checks confirmed acceptability of the parametric approach. For hierarchical linear regression, the additional assumptions of homoscedasticity and absence of influential outliers were verified using residual‐versus‐fitted plots and Cook's distance statistic (all values < 1), respectively. The independence of observations was supported by the intraclass correlation coefficient for the hospital‐level clustering effect (ICC = 0.021), which was negligible.

Hospital‐level clustering was minimal (ICC = 0.021); given only three sites, primary analyses used single‐level regression, with hospital fixed effects examined in sensitivity analyses, yielding comparable estimates (ΔB ≤ 0.03).

Descriptive statistics summarized sample characteristics. Pearson correlations assessed bivariate associations with Holm–Bonferroni correction. Group differences were examined using Welch‐corrected t‐tests and ANOVA with Games–Howell post hoc tests; effect sizes are reported as Cohen's d or partial η ².

Hierarchical linear regression evaluated associations between occupational stress and mental HRQoL. Demographics, role variables, total stress exposure and pre‐specified interaction terms were entered sequentially, with predictors mean‐centred before interaction testing. Models report coefficients, robust standard errors, exact p‐values, VIFs and ΔR ². Pre‐specified sensitivity analyses confirmed robustness across alternative specifications.

3. Results

3.1. Participant Flow

Figure 1 summarizes participant flow in accordance with STROBE reporting conventions. Of approximately 428 eligible registered nurses invited across the three study hospitals, 290 provided written informed consent and returned a questionnaire (consent rate = 67.8%). Thirty‐three questionnaires were subsequently excluded because they exceeded the pre‐specified threshold of 20% item‐level missingness or failed embedded attention checks, yielding a final analytic sample of 257 nurses (final response rate = 60.0%). Documented reasons for non‐participation included scheduling constraints during the data‐collection window (n = 62), decline without specified reason (n = 51) and long‐term or maternity leave (n = 25).

Item‐level missingness on retained questionnaires was minimal (< 5%) and was handled by listwise deletion in the primary analysis. A sensitivity analysis employing pairwise deletion produced substantively identical regression coefficients (all Δβ ≤ 0.04 units from the primary estimates), supporting the robustness of the analytic approach.

3.2. Sample Characteristics

Demographic and professional characteristics of the 257 participating nurses are summarized in Table 1. The sample was predominantly female (n = 118, 53.2%) and married (n = 201, 78.4%); the majority held a bachelor's degree as their highest professional qualification (n = 146, 56.8%). Inpatient wards (n = 118, 45.9%) and critical care units (n = 96, 37.4%) were the most frequently represented clinical settings; the remaining 43 nurses (16.7%) worked in administrative or quality‐assurance roles. Staff nurses constituted the majority of the sample (n = 221, 86.0%), and 186 nurses (72.5%) reported having previously received formal occupational‐safety training. Professional experience was distributed across early‐career (< 5 years, 14.4%), mid‐career (5–10 years, 41.6%) and senior‐career (> 10 years, 44.0%) groups. Age and gender were reported by 222 of the 257 participants (35 missing responses); percentages for these two variables are therefore calculated on valid responses (n = 222), whereas all remaining characteristics were reported by the full sample.

TABLE 1.

Demographic and professional characteristics of study participants (N = 257).

Variable	n	%
Age
20–30 years	37	16.7
31–40 years	102	45.9
41–50 years	72	32.4
> 50 years	11	5.0
Gender
Female	118	53.2
Male	104	46.8
Marital status
Single	43	16.8
Married	201	78.4
Divorced/Widowed	13	5.1
Education
Diploma	67	26.1
Bachelor's degree	146	56.8
Postgraduate	44	17.1
Professional position
Staff nurse	221	86.0
Nurse manager	36	14.0
Primary work area
Critical care	96	37.4
Inpatient ward	118	45.9
Admin/QA	43	16.7
Years of experience
< 5 years	37	14.4
5–10 years	107	41.6
> 10 years	113	44.0
Occupational safety training
Yes	186	72.5
No	71	27.5

3.3. Preliminary Psychometric and Assumption Checks

Before inferential analyses were undertaken, the psychometric adequacy of the adapted 34‐item Nursing Stress Scale (NSS) and the statistical assumptions underlying parametric testing were examined.

3.3.1. Internal Consistency

Cronbach's α for the total adapted NSS was 0.96, with subscale αs ranging from 0.85 to 0.95. McDonald's ω for the total scale was also 0.96, indicating that the high α estimate was not an artefact of an unduly homogeneous item set.

3.3.2. Distributional Assumptions

Kolmogorov–Smirnov and Shapiro–Wilk tests, in combination with visual inspection of Q–Q plots and residual‐versus‐fitted plots, indicated acceptable approximation of normality and homoscedasticity for all continuous variables entered into the regression model.

3.3.3. Multicollinearity

Variance inflation factors (VIFs) for all predictors in the final regression model ranged from 1.08 to 3.71, all below the conventional threshold of 5 (O'Brien 2007), supporting coefficient stability.

3.3.4. Hospital‐Level Clustering

The intraclass correlation coefficient (ICC) for the Mental Component Summary (MCS) by hospital was 0.021 (95% CI [0.000, 0.067]), indicating that approximately 2% of MCS variance was attributable to between‐hospital differences. A sensitivity analysis including hospital as a set of fixed‐effect dummy variables produced stressor‐domain coefficients within 0.03 units of the primary model, supporting the decision to present the single‐level regression as the primary analysis while formally acknowledging the clustering issue as a design limitation (see Section 4.6).

3.4. Occupational Stress and Mental HRQoL: Descriptive Findings

Descriptive statistics for the seven stressor domains, the composite stress score, and the SF‐12 summary indices are presented in Table 2. All stressor items were rated on a 1–4 frequency scale, with higher scores indicating more frequent exposure. Among the seven domains, workload received the highest mean rating (M = 2.91, SD = 0.65, 95% CI [2.83, 2.99]), followed by inadequate preparation (M = 2.50, SD = 0.66, 95% CI [2.42, 2.58]) and lack of staff support (M = 2.47, SD = 0.73, 95% CI [2.38, 2.56]). The composite stress score was 2.46 (SD = 0.52, 95% CI [2.40, 2.52]), reflecting a moderate level of perceived occupational stress on the 1–4 metric.

TABLE 2.

Descriptive statistics for occupational stressors and SF‐12 summary scores (N = 257).

Domain	M	SD	95% CI
Stressor domains (1–4 scale)
Workload	2.91	0.65	[2.83, 2.99]
Inadequate preparation	2.50	0.66	[2.42, 2.58]
Lack of staff support	2.47	0.73	[2.38, 2.56]
Uncertainty concerning treatment	2.44	0.71	[2.35, 2.53]
Death and dying	2.39	0.69	[2.30, 2.48]
Conflict with other nurses	2.32	0.72	[2.23, 2.41]
Nurse–physician conflict	2.17	0.75	[2.08, 2.26]
Composite stress score	2.46	0.52	[2.40, 2.52]
SF‐12 summary scores (norm mean = 50, SD = 10)
Physical Component Summary (PCS)	53.91	7.20	[53.02, 54.80]
Mental Component Summary (MCS)	52.01	8.12	[51.02, 53.00]

Abbreviations: CI, confidence interval; M, mean; MCS, Mental Component Summary; PCS, Physical Component Summary; SD, standard deviation.

The mean Physical Component Summary (PCS) score was 53.91 (SD = 7.20, 95% CI [53.02, 54.80]), and the mean Mental Component Summary (MCS) score was 52.01 (SD = 8.12, 95% CI [51.02, 53.00]). Because the SF‐12 is norm‐scored to a US general‐population mean of 50 (SD = 10) (Ware Jr. et al. 1996), the present MCS mean lies marginally above the international normative benchmark. This counterintuitive observation warrants three interpretive caveats: (a) well‐coping or higher‐functioning nurses may have been disproportionately likely to complete an in‐workplace survey; (b) culturally influenced response tendencies towards negatively framed items may have contributed to underreporting of psychological distress in the Saudi context; and (c) measurement invariance of the SF‐12 between the US normative sample and Arabic‐speaking Saudi nurses was not formally tested. In line with the ILO (2016) and EU‐OSHA (2022) emphasis on interpreting exposure and outcome data jointly, the pattern observed here, an apparently favourable population mean set against uniformly significant inverse stressor–MCS associations, illustrates why single‐summary indicators are insufficient for psychosocial risk surveillance.

3.5. Bivariate Associations Between Stressor Domains and Mental HRQoL (Evaluating Proposition 1)

Pearson correlations between the seven stressor domains, the composite stress score, and the SF‐12 summary indices are presented in Table 3. All seven stressor domains were inversely and reliably associated with the MCS (r = −0.22 to −0.33, all p < 0.001), and all seven associations remained reliable after Holm–Bonferroni correction across the pre‐specified family of seven stressor–MCS tests. Consistent with Proposition 1, nurse–physician conflict exhibited the numerically largest inverse association with MCS (r = −0.33, 95% CI [−0.43, −0.22], p < 0.001), followed by lack of staff support (r = −0.30, 95% CI [−0.41, −0.19], p < 0.001) and nurse–peer conflict (r = −0.28, 95% CI [−0.39, −0.16], p < 0.001). Workload exhibited a smaller association with MCS (r = −0.26, 95% CI [−0.37, −0.14], p < 0.001). The composite stress score correlated with MCS at r = −0.35, 95% CI [−0.45, −0.24], p < 0.001.

TABLE 3.

Pearson correlations between stressor domains and SF‐12 MCS (N = 257).

Stressor domain	r with MCS	95% CI	p
Nurse–physician conflict	−0.33	[−0.43, −0.22]	< 0.001
Lack of staff support	−0.30	[−0.41, −0.19]	< 0.001
Conflict with other nurses	−0.28	[−0.39, −0.16]	< 0.001
Workload	−0.26	[−0.37, −0.14]	< 0.001
Uncertainty concerning treatment	−0.25	[−0.36, −0.13]	< 0.001
Inadequate preparation	−0.24	[−0.35, −0.12]	< 0.001
Death and dying	−0.22	[−0.34, −0.10]	< 0.001

Note: All correlations were significant at p < 0.001 after Holm–Bonferroni correction.

Abbreviation: MCS, Mental Component Summary.

3.6. Group Differences in Mental HRQoL and Occupational Stress

Group comparisons, together with gender as a pre‐specified covariate of interest, are summarized in Table 4. Independent‐samples t‐tests used Welch's correction where Levene's test indicated variance heterogeneity.

TABLE 4.

Group differences in mental HRQoL (MCS) by gender, position and work area.

Comparison	Group A, mean (SD)	Group B, mean (SD)	d	p
Gender: Male (A) vs. Female (B)	52.64 (8.66)	49.18 (10.26)	0.37	0.022
Position: Manager (A) vs. Staff nurse (B)	54.62 (7.44)	49.07 (10.20)	0.61	0.002
Work area: IP ward (A) vs. Critical care (B)	51.38 (9.01)	47.26 (11.28)	0.41	0.004

Note: Higher MCS scores indicate better mental HRQoL.

Abbreviations: CC, critical care; IP, inpatient; MCS, Mental Component Summary.

3.6.1. Gender

Male nurses reported reliably higher MCS scores than female nurses (M = 52.64, SD = 8.66 vs. M = 49.18, SD = 10.26; t(220) = 2.31, p = 0.022, d = 0.37, 95% CI [0.05, 0.69]).

3.6.2. Professional Position (Test of Proposition 2)

Consistent with Proposition 2, nurse managers reported reliably better mental HRQoL than staff nurses (MCS: M = 54.62, SD = 7.44 vs. M = 49.07, SD = 10.20; t(255) = 3.13, p = 0.002, d = 0.61, 95% CI [0.25, 0.97]), a medium‐magnitude effect.

3.6.3. Work Area (Test of Proposition 3)

Consistent with Proposition 3, a one‐way ANOVA across the three work‐area categories revealed reliable differences in MCS (F(2, 254) = 5.82, p = 0.003, η ² _p = 0.044, 90% CI [0.010, 0.088]). Games–Howell post hoc contrasts indicated that critical care nurses reported lower MCS than inpatient ward nurses (M = 47.26, SD = 11.28 vs. M = 51.38, SD = 9.01; mean difference = −4.12, 95% CI [−7.04, −1.20], p = 0.004, d = 0.41). All contrasts remained reliable after Holm–Bonferroni correction.

3.7. Hierarchical Regression Predicting Mental HRQoL

A four‐step hierarchical regression analysis was conducted to examine predictors of mental health–related quality of life (MCS). Demographic variables (age, gender and experience) were entered in Step 1, role variables (position and work area) in Step 2, the seven stressor domains in Step 3, and three interaction terms (Gender × Total Stress, Position × Total Stress, and Work Area × Total Stress) in Step 4. As shown in Table 5, the final model explained 34% of the variance in MCS (adjusted R ² = 0.30, F(9, 247) = 12.89, p < 0.001).

TABLE 5.

Hierarchical regression model summary (outcome: MCS).

Step	Predictors entered	R 2	ΔR 2	F‐change	p (F‐change)
1	Demographics (age, gender, experience)	0.03	0.03	2.60	0.054
2	+ Role variables (position, work area)	0.17	0.14	20.13	< 0.001
3	+ Seven stressor domains	0.32	0.15	7.85	< 0.001
4	+ Interaction terms (3)	0.34	0.02	2.47	0.064

The demographic variables accounted for 3% of the variance (R ² = 0.03, p = 0.054). Adding role variables significantly increased the explained variance to 17% (ΔR ² = 0.14, p < 0.001). The stressor domains contributed a further 15% of explained variance (ΔR ² = 0.15, p < 0.001), whereas the interaction terms added a non‐significant 2% (ΔR ² = 0.02, p = 0.064).

In the final model (Table 6), Total Stress emerged as the strongest predictor of lower MCS (β = −0.28, p = 0.003). Female gender (β = −0.17, p = 0.026), staff‐nurse position (β = −0.14, p = 0.021) and critical‐care work area (β = −0.11, p = 0.047) were also significantly associated with lower MCS. None of the interaction terms reached statistical significance, although the Work Area × Total Stress interaction approached significance (p = 0.060). Figure 2 presents the standardized regression coefficients and 95% confidence intervals for the final model.

TABLE 6.

Final‐model (Step 4) standardized coefficients predicting MCS.

Predictor	β	p	Interpretation
Intercept	66.12	—	—
Female (vs. male)	−0.17	0.026	Reliable negative association
Staff nurse (vs. manager)	−0.14	0.021	Reliable negative association
Critical care (vs. other)	−0.11	0.047	Reliable negative association
Total Stress (composite)	−0.28	0.003	Strongest unique predictor
Gender × Total Stress	—	0.500	Non‐significant interaction
Position × Total Stress	—	0.890	Non‐significant interaction
Work Area × Total Stress	—	0.060	Borderline; under‐powered

Note: Final model: adjusted R ² = 0.30, F(9, 247) = 12.89, p < 0.001. Total model R ² = 0.34.

Abbreviation: MCS, Mental Component Summary.

FIGURE 2.

Forest plot of standardized β coefficients from the final hierarchical regression model specification.

4. Discussion

This study examined occupational stressor domains and mental health–related quality of life (HRQoL) among 257 hospital nurses in the Northern Border Region of Saudi Arabia, contributing multi‐hospital, domain‐specific evidence from a region substantially underrepresented in the nursing occupational health literature. The findings reveal a consistent and statistically reliable inverse association between occupational stress and mental HRQoL, with workload representing the most frequently endorsed stressor and nurse–physician conflict demonstrating the strongest bivariate association with impaired mental health (r = −0.33, 95% CI [−0.43, −0.22], p < 0.001). These patterns are broadly consistent with international evidence (Sarafis et al. 2016; Babapour et al. 2022; Yan et al. 2022) and extend its application to the distinctive organizational and cultural conditions of Saudi hospital nursing. The discussion is deliberately cautious in its causal language, reflecting the exploratory, cross‐sectional nature of the design.

4.1. Workload as the Dominant Stressor and Its Structural Context

The prominence of workload as the most highly rated stressor is consistent with the JD‐R model's prediction that challenge demands placed on nurses may exceed available organizational resources (Bakker and Demerouti 2017). In the Saudi context, this finding reflects structural realities documented by Alluhidan et al. (2020), who identified workforce shortages, high dependence on expatriate nurses, and rapid healthcare service expansion under Vision 2030 as key drivers of intensified demand on individual nurses. Comparable patterns have been reported in prior Saudi nursing studies (Aserri et al. 2021; Alharbi and Alshehry 2019; Kaushik et al. 2021).

From a global policy perspective, this finding resonates with the WHO's (2022) identification of excessive workload as a primary psychosocial risk factor in healthcare settings, the ILO's (2016) recognition of understaffing as a preventable occupational hazard, and the EU‐OSHA's (2022) emphasis on workload management as a core component of organizational‐level stress prevention strategies. NIOSH (2020) similarly highlights that inadequate staffing levels in acute‐care settings constitute a modifiable determinant of nurse occupational ill‐health. These converging institutional positions reinforce the imperative for evidence‐based acuity‐sensitive staffing reviews in Saudi hospitals, particularly as the healthcare system absorbs the service expansion driven by Vision 2030. Recent large‐scale US evidence similarly indicates that hospital nurse understaffing persists as a systemic problem with demonstrable consequences for both patient outcomes and nurse well‐being (Lasater et al. 2024), suggesting that the workload stressor documented here is a manifestation of a global structural challenge. In the same line, the International Council of Nurses (2021) has warned that unsustainable workloads and inadequate well‐being support are principal drivers of nurse attrition, while the WHO (2020) identifies investment in decent nursing employment and leadership as the central policy lever for workforce sustainability. The workload burden documented in the present sample therefore constitutes region‐specific empirical support for precisely these global policy priorities, and it illustrates how an internationally endorsed agenda for psychosocial risk management translates into the operational realities of Saudi hospitals undergoing rapid, Vision 2030–driven service expansion. The WHO/ILO joint estimates (WHO and ILO 2021) have established long working hours, the proximal behavioural expression of unmanaged workload, as the occupational risk factor carrying the largest documented disease burden, with 745,000 deaths from ischemic heart disease and stroke attributable to exposure of ≥ 55 h per week in 2016; the present finding therefore elevates workload from a sample‐level nuisance variable to a measurable hazard within the parameters of global occupational disease surveillance. The NIOSH Total Worker Health program (NIOSH, n.d.) reaches a convergent conclusion through an intervention lens, holding that organizational work‐design strategies, rather than individual‐level stress management, constitute the appropriate locus of workload control.

4.2. The Nurse–Physician Conflict Paradox: Hindrance Demands and Mental Health

The most theoretically consequential finding is the dissociation between stressor frequency and stressor impact. Nurse–physician conflict was among the least frequently endorsed stressors (M = 2.17), yet it showed the strongest inverse association with MCS (r = −0.33). Within the JD‐R framework, Crawford et al. (2010) distinguish between challenge demands and hindrance demands; interpersonal conflict with physicians constitutes a prototypical hindrance demand that obstructs nurses' capacity to perform their role and cultivates an interpersonal climate characterized by threat rather than challenge. A recent authoritative synthesis confirms that hindrance stressors are more consistently and strongly linked to strain and burnout than challenge stressors (Podsakoff et al. 2023), and the present findings add nursing‐specific evidence from a Saudi context. From an occupational health standpoint, NIOSH (2020) classifies such organizational and interpersonal stressors as modifiable hazards that should be addressed through organizational‐level controls rather than individual coping alone, a position that reinforces the primacy of systemic remedies over individual resilience‐oriented approaches.

In line with the ILO's (2016) and EU‐OSHA's (2022) frameworks for workplace psychosocial risk management, interprofessional conflict represents a category of psychosocial hazard that is both preventable and amenable to organizational intervention. The WHO (2021) Global Patient Safety Action Plan further notes that poor communication and hierarchy‐driven conflict in healthcare settings are significant contributors to adverse events and workforce psychological distress. A comparable interpretation applies to lack of staff support (r = −0.30), which represents resource losses within the COR framework (Hobfoll et al. 2018) and functions as a hindrance demand by creating conditions under which nurses face escalating demands without adequate interpersonal or supervisory resources. These institutional positions collectively frame nurse–physician conflict and deficient peer support not as interpersonal idiosyncrasies but as recognized psychosocial hazards warranting the same systematic hierarchy of controls applied to physical occupational risks, elimination and substitution of the hazard where possible, followed by organizational and, only lastly, individual‐level measures (EU‐OSHA 2022; ILO 2016; NIOSH 2020).

Three interpretive caveats warrant explicit acknowledgment. First, the present cross‐sectional design cannot establish the temporal ordering of these associations; it is equally plausible that nurses with poorer baseline mental health perceive interpersonal interactions as more conflictual, consistent with a reverse‐causation pathway documented in affective science (Podsakoff et al. 2023). Second, the culturally layered authority structures of Saudi hospitals, where a largely expatriate nursing workforce interacts with predominantly Saudi national physicians, may amplify the hindrance character of interprofessional friction; recent qualitative evidence from Riyadh tertiary hospitals suggests that language asymmetries and status differentials compound the psychological cost of nurse–physician communication breakdowns (AlSaeed et al. 2024). Third, not all recent evidence concurs: Werke and Weret (2023), working in a sub‐Saharan African context, identified workload as the single dominant predictor of psychological distress, underscoring the context‐dependence of the present findings. Furthermore, the normative weight of the ILO Violence and Harassment Convention, 2019 (no. 190) (ILO 2019), the first international treaty to recognize the right of all workers to a world of work free from violence and harassment, extends the interpretation of the present findings beyond interpersonal friction: workplace conduct that undermines dignity, including interprofessional intimidation, now falls within a binding international normative framework, and the primary burden of prevention rests with the employer as duty holder.

4.3. Subgroup Differences: Descriptive Patterns and Inferential Limits

Female nurses reported modestly but reliably lower MCS scores than male nurses (49.18 vs. 52.64; d = 0.37). This pattern is consistent with international evidence showing that female nurses often report greater stress burden than males, partly attributed to the co‐occurrence of professional demands with domestic caregiving responsibilities (Sugawara et al. 2017; Aldhafeeri et al. 2025). Within the Saudi context, the intersection of occupational demands with cultural norms regarding women's domestic and family roles may place female nurses at elevated risk for cumulative role overload (AlAzzam et al. 2017; Alshutwi 2016). The present study did not directly measure work–family conflict, domestic caregiving load or external social support; the finding of lower MCS among female nurses should therefore be treated as a hypothesis‐generating observation rather than a confirmed explanation. This pattern carries direct policy relevance: the WHO (2022) guidelines on mental health at work and the ILO (2016) both identify the unequal distribution of domestic and caregiving responsibilities as a structural psychosocial risk for working women, and the International Council of Nurses (2021) has called for gender‐responsive workforce policies, including flexible scheduling and confidential access to well‐being support, measures that align closely with the culturally responsive institutional supports proposed in Section 4.5.

The better mental health reported by nurse managers compared with staff nurses (MCS: 54.62 vs. 49.07; d = 0.61) is consistent with job design theory's prediction that greater decision latitude, role authority and organizational control buffer psychological strain (Bakker and Demerouti 2017). The non‐significant Position × Total Stress interaction term (β = 0.01, p = 0.890) indicates that professional position did not significantly modify the stress–MCS association; the more defensible interpretation is therefore that staff nurses occupy a more stressful professional position and report lower mental health, rather than that they experience a categorically different stress–health pathway. Critical care nurses reported poorer mental health and higher overall stress than inpatient ward nurses. This is consistent with evidence that high‐acuity nursing environments generate sustained physiological vigilance, repeated exposure to acute patient deterioration, rapid decision‐making under uncertainty and high emotional labour demands (Acebedo‐Urdiales et al. 2018; Olaleye et al. 2022). Viewed through the lenses of the JD‐R and COR frameworks, and consistent with the WHO (2022) emphasis on organizational resources such as supervisor support, role clarity and decision latitude as protective factors for healthcare workers' mental health, this pattern underscores that strengthening job resources, not merely reducing demands, represents an evidence‐aligned strategy for protecting the mental well‐being of the nursing workforce, as further elaborated in the policy implications below.

4.4. Cultural, Religious and Organizational Context of the Findings

The unique cultural‐religious, organizational and workforce contextual features of Saudi hospital nursing are likely to have materially shaped the patterns observed in this study, and their influence warrants explicit discussion. The Saudi nursing workforce is characterized by significant cultural heterogeneity: a predominantly expatriate nursing staff, drawn from South and Southeast Asia, Sub‐Saharan Africa and the wider Arab world, operates within an organizational environment framed by Saudi cultural norms, Islamic values and a hierarchical authority structure. This cultural plurality may intensify certain stressor domains in ways that differ substantially from more culturally homogeneous Western healthcare settings.

Religious practice is an important aspect of daily life in Saudi Arabia and may influence nurses' workplace experiences. Evidence suggests that supportive spiritual environments can promote work engagement and psychological well‐being among nurses, highlighting the potential role of spirituality as a coping resource in healthcare settings (Cruz et al. 2022). The predominantly expatriate composition of the Saudi nursing workforce introduces additional occupational challenges. International evidence indicates that migrant healthcare workers may experience psychosocial risks related to job insecurity, limited social support, cultural adaptation and communication barriers, all of which can affect occupational well‐being (ILO 2016). Furthermore, studies conducted in Saudi healthcare settings have identified language differences, communication barriers, workload pressures and interpersonal conflict as factors that may hinder effective collaboration and contribute to workplace stress (Aljuaid and Alkarani 2023; Barnawi and Barnawi 2023). These considerations acquire normative force through the WHO Global Code of Practice on the International Recruitment of Health Personnel (WHO 2010), which commits destination countries such as Saudi Arabia to ensuring that internationally recruited health personnel work in conditions that safeguard their rights and well‐being; the elevated relational stressors observed here, particularly within a predominantly expatriate nursing workforce, constitute precisely the working conditions the Code was designed to protect.

Organizational factors may have contributed to the observed associations in the present study. Evidence from Saudi healthcare settings suggests that nurses often practice within hierarchical organizational structures where autonomy and participation in decision‐making may be constrained, potentially increasing the burden of workplace stressors and limiting opportunities for conflict resolution and professional support (Alenezi 2026; Tumala 2025). Leadership practices, organizational culture and power dynamics have likewise been identified as important influences on nurses' work engagement, professional autonomy and workplace experiences in Saudi hospitals (Alluhaybi et al. 2024; Alenezi 2025). Furthermore, the predominantly expatriate nursing workforce introduces additional challenges related to language, cultural differences, communication barriers and adaptation to local healthcare contexts, all of which may influence occupational well‐being and interprofessional relationships (Alahmmari 2017; Aljuaid and Alkarani 2023; Alharazi et al. 2025). These contextual factors suggest that interventions targeting the organizational and relational environment of Saudi hospitals should be culturally responsive and sensitive to workforce diversity. A limitation of the present study is that organizational climate, cultural identity, expatriate status and religious engagement were not measured directly. Future research should incorporate these variables to better understand their independent contributions to nurses' mental health–related quality of life. These observations converge with the EU‐OSHA (2022) position that psychosocial risk prevention is most effective when it targets organizational structures, worker participation and leadership practices rather than individual characteristics alone, and with the NIOSH (2020) prevention logic that organizational‐level controls should precede individual‐level coping supports. Incorporating such organizational variables, and the voice of nurses themselves, in the form of their own proposals for improvement, into future designs would allow the field to move from documenting psychosocial hazards to co‐producing, with the nursing workforce, the remedies evaluated against them.

4.5. Clinical and Organizational Implications

Given the exploratory and cross‐sectional nature of the present findings, the recommendations below are offered as directional guidance rather than evidence‐based prescriptions.

These recommendations are aligned with international occupational health frameworks advanced by the WHO (2022), ILO (2016), EU‐OSHA (2022), and NIOSH (2020) and are intended to inform nursing practice, nursing leadership, nursing education and health policy within Saudi Arabia's evolving healthcare landscape.

4.5.1. Tied to Workload ( r  = −0.26)

Hospitals should implement acuity‐sensitive staffing reviews and systematically monitor unit‐level demand relative to staffing capacity. Nursing leaders and hospital administrators should adopt evidence‐based nurse‐to‐patient ratio guidelines informed by international standards, and advocate for their formal incorporation into the Saudi National Transformation Program policy instruments. The framing of this recommendation is reinforced by the WHO (2016) Workforce 2030 strategy, which positions decent working conditions and workload planning among the primary levers of health‐workforce sustainability.

4.5.2. Tied to Nurse–Physician Conflict ( r  = −0.33)

Hospitals should invest in structured interprofessional communication mechanisms, including joint clinical huddles, clear escalation pathways and conflict–resolution training for both nurses and physicians (Powell et al. 2019; Radu 2023; Cullati et al. 2019). Nursing education programs should integrate interprofessional communication and conflict competencies into pre‐registration curricula, and continuing professional development programs should incorporate culturally and linguistically responsive communication training, given the multinational workforce composition of Saudi hospitals (AlSaeed et al. 2024; Alsadaan et al. 2021). Such measures operationalize the employer duties articulated in ILO Convention No. 190 (ILO 2019) and the organizational‐intervention priorities of the WHO (2022) guidelines, which recommend structured manager training and participatory organizational change as first‐line strategies for reducing interpersonal workplace adversity.

4.5.3. Tied to Position‐Based Difference in MCS ( d  = 0.61)

Nurse managers and healthcare organizations share responsibility for protecting staff nurse mental health. Formal mentorship programs, psychologically informed supervision, and institutional pathways for reporting workload concerns are needed (Wu et al. 2021; Ohue and Menta 2024). Nursing leadership curricula should address the creation of psychologically safe ward environments.

4.5.4. Tied to Critical‐Care Subgroup Difference

Targeted debriefing protocols after high‐intensity clinical events, protected recovery time, and competency development programs are needed for critical care nurses (Olaleye et al. 2022; Acebedo‐Urdiales et al. 2018). Hospital policy should mandate structured post‐incident psychological support for critical care teams, and health policy frameworks should recognize critical care nursing as a specialist area with distinct occupational health risk profiles requiring differentiated monitoring and intervention. This tiered response reflects the integrated protection–promotion logic of NIOSH Total Worker Health (NIOSH, n.d.) and the WHO (2022) recommendation that high‐strain clinical environments receive prioritized organizational support.

4.5.5. Tied to Gender‐Based Difference ( d  = 0.37)

For female nurses, culturally responsive institutional support should include flexible scheduling, confidential access to occupational well‐being services, and managerial recognition of cumulative dual‐role demands (Al Mutair et al. 2022; Alotaibi et al. 2016; Aldhafeeri et al. 2025).

4.6. Strengths and Limitations

Strengths include multi‐site recruitment from three hospitals, assessment of stress across seven distinct domains, use of a validated and norm‐referenced mental‐health outcome measure (SF‐12 MCS), preliminary psychometric evaluation of the adapted Nursing Stress Scale via exploratory and confirmatory factor analysis, and provision of multi‐hospital data from an understudied Saudi region.

Seven principal limitations should be considered in interpreting the findings. (1) The cross‐sectional design precludes any inference of temporal precedence or causal direction; specifically, reverse pathways whereby poorer baseline mental health amplifies the perception of interpersonal friction cannot be ruled out. (2) Stratified quota sampling from three hospitals in one regional health system limits generalizability. (3) Self‐report questionnaires administered at a single time point create susceptibility to common method bias. (4) Hospital‐level clustering was not formally accounted for in primary regression models. (5) The power analysis was calibrated for main effects rather than interaction effects; the non‐significant interaction terms may reflect insufficient statistical power. (6) Social desirability bias is a plausible threat to validity, particularly given workplace survey administration. (7) The SF‐12 MCS was scored using US‐based population norms, and measurement invariance between the US normative sample and Arabic‐speaking Saudi nurses was not formally tested (Haddad et al. 2021).

An eighth limitation relates to the absence of direct measures of cultural, religious and expatriate status variables in the survey instrument. As discussed in Section 4.4, these contextual factors may constitute important independent predictors and moderators of nurse mental HRQoL in Saudi hospital settings. Their omission limits the interpretive reach of the present findings and should be addressed in future research. In addition, the NSS does not capture several hazard categories central to contemporary international frameworks, including effort–reward imbalance, organizational justice and violence and harassment as defined by ILO Convention no. 190 (ILO 2019); its coverage of the psychosocial risk space, although adequate against the ILO (2016) taxonomy, remains partial when evaluated against the fuller hazard inventory recommended by EU‐OSHA (2022) and the WHO (2022) surveillance agenda.

4.7. Future Research Directions

Five specific directions would substantially advance the evidence base. (1) Prospective longitudinal designs with at least two measurement waves separated by 6–12 months to examine whether baseline occupational stress predicts change in MCS over time. (2) Direct measurement of work–family conflict, psychological capital, organizational climate quality and interprofessional relationship quality as potential mediators and moderators. (3) Larger, multi‐region probability‐based samples (minimum approximately 400–500 nurses) to test confirmatory moderation hypotheses with adequate statistical power and pre‐registration of directional expectations. (4) Qualitative and mixed‐methods research to illuminate how nurses in Saudi Arabia interpret and manage occupational stress in culturally specific ways. (5) Formal measurement‐invariance testing of the SF‐12 MCS across Saudi nursing subpopulations (national vs. expatriate, male vs. female, critical care vs. ward).

A sixth direction, identified in response to the finding that the final model explained only 34% of MCS variance, is the systematic incorporation of individual‐level psychological resources, including resilience, coping, self‐efficacy and emotional intelligence, alongside organizational climate variables and expatriate‐specific psychosocial risk factors, in future explanatory models of nurse mental HRQoL (Podsakoff et al. 2023; Hobfoll et al. 2018; Aldhafeeri et al. 2025). Methodologically, future surveillance should emulate the longitudinal, indicator‐based monitoring model of the EU‐OSHA European Survey of Enterprises on New and Emerging Risks (EU‐OSHA 2024), enabling trend detection in psychosocial hazards across Saudi health facilities, and should embrace the integrated hazard–well‐being measurement logic of NIOSH Total Worker Health (NIOSH, n.d.) so that organizational and individual levels of intervention are evaluated jointly rather than in isolation.

5. Conclusions

This multi‐hospital cross‐sectional study demonstrated that occupational stress is inversely associated with mental HRQoL among nurses in the Northern Border Region of Saudi Arabia. Nurse–physician conflict and lack of staff support showed stronger associations with mental health than workload alone, highlighting the importance of the relational and organizational work environment. Female nurses, staff nurses and critical care nurses reported comparatively poorer mental health outcomes, indicating groups that may benefit from targeted support. Importantly, these findings align with, and provide region‐specific empirical support for, the positions of the leading international occupational health and nursing organizations: the WHO (2022) guidelines on mental health at work identify excessive workload, deficient support and interpersonal conflict as priority psychosocial risks requiring organizational‐level intervention; the ILO (2016) classifies understaffing and inadequate social support as preventable collective hazards; EU‐OSHA (2022) situates these hazards within participatory, organization‐level risk assessment; NIOSH (2020) designates organizational stress prevention as the most effective tier of intervention; and the International Council of Nurses (2021) affirms that safeguarding nurses' mental health is a precondition for patient safety and health‐system resilience.

The findings have important implications for nursing practice, education, leadership and health policy. Interventions that strengthen interprofessional communication, conflict resolution, peer support, self‐care, mentorship, psychologically safe work environments and acuity‐sensitive staffing may improve nurses' mental well‐being. The results also support the integration of workforce well‐being initiatives and systematic psychosocial risk management into healthcare policy and organizational practice. The findings translate into actionable directions across three policy levels. With respect to workers' policy, Saudi occupational health regulations should formally recognize psychosocial risk factors, excessive workload, deficient staff support and interprofessional conflict as reportable occupational hazards subject to systematic assessment and control, consistent with the ILO (2016) framework and the EU‐OSHA (2022) emphasis on participatory risk assessment. With respect to health policy, national workforce strategies should institutionalize acuity‐sensitive staffing standards, structured interprofessional communication mechanisms and routine surveillance of nurses' mental health in line with the WHO (2022) guidelines on mental health at work, the WHO (2021) Global Patient Safety Action Plan, and the NIOSH (2020) organizational‐level approach to stress prevention; embedding these measures within Vision 2030 health‐sector transformation instruments would endow them with regulatory force. With respect to social policy, culturally responsive supports addressing the dual professional and family responsibilities reported by female nurses, together with safeguards against language‐ and status‐related disadvantage among migrant nurses, respond to the International Council of Nurses' (2021) call for workforce well‐being as a shared societal responsibility and to the sociocultural profile of the Saudi nursing workforce characterized in this study.

Although occupational stress explained a substantial proportion of variance in mental HRQoL (34%), additional individual, organizational and contextual factors should be examined in future longitudinal and multi‐region studies.

Overall, protecting nurses' mental health should be recognized as a workforce priority and a prerequisite for sustaining high‐quality patient care and healthcare system performance in Saudi Arabia. In sum, this study demonstrates that internationally endorsed frameworks for the prevention of occupational stress are directly applicable to the Saudi hospital context, and that their systematic adoption offers a credible pathway towards protecting the mental health of the nursing workforce and, by extension, the quality and safety of patient care. The convergence of the present findings with the WHO/ILO (2021) quantification of the global disease burden attributable to psychosocial occupational exposures underscores that workload‐related stress in nursing is not an idiosyncratic regional inconvenience but a measurable component of the global occupational disease profile, and its control a shared obligation under the international frameworks adopted by WHO, ILO, EU‐OSHA, NIOSH and the International Council of Nurses.

Author Contributions

Fathia Ahmed Mersal: conceptualization, investigation, writing – original draft, writing – review and editing, methodology, formal analysis, software, data curation, supervision. Ibrahim Naif Alenezi: methodology, investigation, writing – original draft, writing – review and editing, data curation, software, visualization. Fathia Gamal Al Said: investigation, writing – original draft, writing – review and editing, software, data curation, conceptualization, methodology. Faisal Khalaf Alanazi: investigation, writing – original draft, writing – review and editing, resources. Rasmia Abd El Sattar Ali: writing – original draft, writing – review and editing, visualization, supervision, data curation, software. Majid Ali Alotni: investigation, writing – original draft, writing – review and editing, data curation, software, validation. Basema Alharbi: writing – original draft, writing – review and editing, validation, software, visualization, data curation. Ebtesam A. Elsayed: investigation, writing – original draft, writing – review and editing, software, methodology, data curation.

Funding

This research received a grant from the Deanship of Scientific Research at Northern Border University, Arar, Saudi Arabia, for funding this research through project number ‘NBU‐FFR‐2026‐3326‐09’.

Ethics Statement

Ethical approval for all study procedures was obtained from the relevant institutional review authority before commencement of data collection, in full accordance with the Declaration of Helsinki and applicable national research governance standards.

Consent

Written informed consent was obtained from all participants.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Appendix S1: nop270869‐sup‐0001‐Supinfo1.docx.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.