Mental health nurses working in psychiatric hospitals and community mental health centers play a vital role in delivering comprehensive care to individuals with psychiatric disorders. Owing to the nature of psychiatric symptoms, these nurses are frequently exposed to patients' selfharm, physical aggression, and verbal abuse [1]. They are also required to rely on emotional presence and interpersonal engagement as key therapeutic tools, making emotional labor a central and demanding aspect of their work [2]. Consequently, the prevention of emotional exhaustion among psychiatric nurses has been widely acknowledged as a critical priority [3]. Against this backdrop, the concept of professional quality of life (ProQOL) is increasingly used to explain emotional responses and outcomes among professionals engaged in caregiving work [4].

ProQOL captures the affective consequences of helping work rather than overall life status [5]. In the ProQOL model, the rewarding side is compassion satisfaction, whereas the strain side is compassion fatigue, which comprises burnout and secondary traumatic stress [6]. Burnout is the result of the depletion of emotional and physical resources owing to work demands, whereas secondary traumatic stress involves stress reactions and functional impairment caused by indirect exposure to traumatic experiences [6]. Unlike general quality of life, which encompasses physical, mental, social, and environmental do-mains, ProQOL focuses specifically on the emotional impact of caregiving work and is therefore particularly relevant to nursing professionals who are repeatedly exposed to others' suffering [5,6].

According to the ProQOL framework [6], professional well-being in caregiving occupations has been discussed in relation to personal characteristics, client-related factors, and work-related environmental conditions. Among these, the work environment is a well-established correlate of nurses' performance and well-being [7]. The nursing practice environment includes participation in organizational affairs, quality-related foundations of care, manager leadership, staffing and resource adequacy, and nurse-physician relations [8]. Favorable work environments have been associated with greater job satisfaction, better patient outcomes, and lower turnover [7]. In psychiatric settings, these organizational conditions may be especially important, as nurses routinely work in emotionally demanding and safety-sensitive environments [9].

Another salient condition in psychiatric nursing is exposure to patient violence. Violence from patients is a major client-related occupational stressor and has been linked to depressive symptoms, post-traumatic stress symptoms, reduced job satisfaction, and turnover intention [4,9]. Meanwhile, nurses exposed to similar work conditions do not necessarily report the same level of professional distress, suggesting that external occupational conditions alone may not fully explain the variation in ProQOL [4]. This observation highlights the relevance of internal psychological characteristics that may be associated with how nurses respond to demanding or threatening clinical situations. However, although previous studies have identified work environment and violence exposure as important correlates of nurses' well-being, fewer studies have examined how these factors relate to Pro-QOL alongside specific internal traits such as mental toughness and self-compassion [4]. In addition, conceptual precision remains limited, particularly regarding which internal factors may be differentially associated with positive and negative aspects of ProQOL [10].

Mental toughness was included as a personal resource because psychiatric nurses often need to function under continuing interpersonal and safety-related strain. The construct originated in sport psychology but has been extended to other high-pressure occupational contexts [11]. Rather than equating it with resilience, this study treats mental toughness as a related but separate construct characterized by challenge appraisal, persistence, perceived control, and confidence [12]. Whereas resilience centers on recovery after adversity, mental toughness emphasizes effective functioning while stressors are still unfolding [13]. For mental health nurses, this capacity may be relevant during patient aggression, clinical crises, and emotionally demanding therapeutic encounters [4,14].

Self-compassion may also serve as a relevant intrapersonal factor. Self-compassion describes a tendency to respond to personal difficulties with warmth, understanding, and acceptance, rather than with severe self-blame or criticism [15]. According to Neff's model, this construct consists of six elements: self-kindness, self-judgment, common humanity, isolation, mindfulness, and over-identification [15]. More recent work has suggested that these elements can be organized into two conceptually related but distinct dimensions. The first reflects the adaptive aspects of self-compassion, including self-kindness, common humanity, and mindfulness, whereas the second captures maladaptive responses such as self-judgment, isolation, and over-identification, which are frequently described as self-coldness [16]. This distinction is important because the positive and negative components do not appear to function merely as opposite poles of a single continuum; instead, they may show different patterns of association with positive and negative indicators of psychological adjustment [16].

In nursing contexts, these adaptive and maladaptive aspects of self-related responding may help explain how nurses regulate emotional strain and maintain psychological adjustment. Positive self-compassion may support compassion satisfaction by facilitating balanced awareness of distress, self-kindness after difficult clinical encounters, and a sense that occupational difficulties are part of a shared human and professional experience. By contrast, self-coldness may be associated with compassion fatigue through self-critical rumination, perceived isolation, and emotional over-identification. When nurses interpret patient aggression, therapeutic setbacks, or emotional exhaustion as personal failure, self-coldness may prolong negative affect, intensify shame or helplessness, and reduce adaptive recovery following stressful encounters. Over-identification may also blur the psychological boundary between patients' distress and nurses' own emotional responses, thereby increasing vulnerability to secondary traumatic stress and burnout. Thus, self-coldness is not merely the absence of positive self-compassion but may represent a maladaptive self-relational process that maintains or amplifies emotional strain in caregiving work.

Prior studies have shown that mindfulness may reduce burnout, whereas feelings of isolation may exacerbate it [17]. Other evidence suggests that self-compassion facilitates emotional regulation, supports meaning-making, and promotes recovery following distressing experiences [18]. For mental health nurses, who often encounter high emotional demands, self-compassion may support positive professional well-being, whereas self-coldness may increase vulnerability to compassion fatigue. Despite growing interest in self-compassion in nursing research, relatively few studies have explicitly examined the distinct roles of self-compassion and self-coldness in relation to ProQOL. Clarifying these potentially differentiated associations may provide a more nuanced understanding of how nurses' internal psychological processes relate to both the positive and negative dimensions of professional well-being.

Taken together, prior literature suggests that both external occupational conditions and internal psychological characteristics are relevant to ProQOL. From a Job Demands-Resources perspective (JD-R) [19], work environment and exposure to violence are considered important external occupational conditions, whereas mental toughness and self-compassion-related processes may represent internal psychological resources relevant to professional well-being. In this framework, mental toughness and selfcompassion may function as adaptive psychological resources, whereas self-coldness may reflect a maladaptive self-relational tendency that increases vulnerability to emotional strain. However, the extent to which these factors are differentially associated with compassion satisfaction and compassion fatigue remains unclear.

Accordingly, this study examined the associations of work environment, exposure to patient violence, mental toughness, self-compassion, and self-coldness with Pro-QOL among mental health nurses (Figure 1). Based on the ProQOL framework, work environment and exposure to patient violence were conceptualized as key external occupational conditions, whereas mental toughness, self-compassion, and self-coldness were treated as internal psychological factors relevant to professional well-being. In particular, the study aimed to determine whether these internal factors demonstrate distinct associations with compassion satisfaction and compassion fatigue. It also intended to examine whether the associations of external occupational conditions with ProQOL are better understood when these internal psychological factors are considered.

This study examined the direct and indirect associations of work environment, exposure to patient violence, mental toughness, and the two components of self-compassion with ProQOL among psychiatric and community mental health nurses. Although the original ProQOL framework primarily emphasizes direct associations among these factors [6], the present findings suggest that these relationships may be more comprehensively understood when internal psychological factors are considered. In particular, mental toughness, self-compassion, and selfcoldness demonstrated distinct associations with ProQOL, highlighting the combined importance of external occupational conditions and internal psychological processes in professional well-being. These findings are broadly consistent with both the ProQOL framework and the Job Demands-Resources theory.

Specifically, work environment and mental toughness were each associated with higher compassion satisfaction and lower compassion fatigue, and mental toughness accounted for significant indirect associations between work environment and both ProQOL components. This pattern is consistent with previous research emphasizing the relevance of psychological resilience in high-stress occupations [3]. One plausible explanation is that supportive work environments may help nurses maintain confidence, persistence, and emotional control, thereby contributing to more favorable professional well-being. Meanwhile, the present findings suggest that organizational support alone may not be sufficient if nurses lack adequate internal coping resources. Accordingly, efforts to improve nurses' professional well-being may benefit from combining environmental improvement with strategies that strengthen psychological resilience. Given that the present data are crosssectional, the temporal direction of these associations should be interpreted cautiously.

Self-compassion and self-coldness exhibited distinct patterns of association with ProQOL. In the present study, the term "self-compassion" refers specifically to the positive component of the two-factor model, whereas the overall construct measured by the self-compassion scale includes self-coldness. This distinction is important because the same terminology is used for both the overall construct and its positive component, which may otherwise lead to conceptual ambiguity. Self-compassion was associated with higher compassion satisfaction but not with compassion fatigue, whereas self-coldness was pos-itively associated with compassion fatigue but not with compassion satisfaction. This pattern supports the twofactor view that the positive and negative components of self-compassion are not merely opposite poles of a single continuum but are differentially linked to positive and negative aspects of occupational well-being. In the present context, self-compassion may be more closely related to meaning, fulfillment, and emotional balance in caregiving, whereas self-coldness may be more closely related to self-critical, isolating, and over-identifying responses that amplify emotional depletion. Individuals with high selfcoldness often exhibit emotional rigidity, harsh self-evaluation, and difficulties in adaptive emotional regulation [16,17]. Accordingly, interventions for mental health nurses may need to combine strategies that cultivate self-compassion with approaches that directly target self-critical and maladaptive self-relational patterns associated with self-coldness.

Although exposure to patient violence was not directly associated with ProQOL in the final model, a small indirect association with compassion satisfaction through self-compassion was observed. This finding should not be interpreted as indicating that patient violence is unimportant for nurses' professional well-being. Rather, it may reflect the distinction between objective exposure to aggression and the subjective appraisal and processing of such events. In this study, violence exposure was assessed using severity-weighted indicators that reflected multiple forms of patient aggression, including verbal hostility, destructive or threatening behavior involving objects, behaviors in which patients harmed or threatened to harm themselves, and aggressive acts directed at other individuals. However, this measure did not directly assess how nurses appraised these incidents, such as perceived threat, fear, controllability, emotional impact, coping capacity, perceived blame, or the availability of organizational and peer support after the event. ProQOL may therefore be more closely related to how violent incidents are interpreted, emotionally processed, and supported than to objective exposure severity alone.

The characteristics of the present sample may also have contributed to this pattern. Participants were currently employed mental health nurses with at least one year of clinical experience, and more than half had five or more years of nursing experience. Therefore, nurses who had left psychiatric or community mental health settings after severe or poorly supported exposure to violence were not represented. This possible retention effect may have attenuated the direct association between violence exposure and ProQOL. In addition, for nurses who remain in mental health settings despite repeated exposure to aggression, self-compassion may become particularly salient as a self-regulatory resource following violent incidents, because such events can evoke fear, helplessness, self-blame, or perceived professional failure. The small indirect association between violence exposure and compassion satisfaction through self-compassion may therefore suggest that self-relational processes are relevant to how nurses recover from aggression while maintaining meaning in their caregiving role. However, this finding should be interpreted cautiously and does not imply that exposure to violence has a beneficial effect. Future studies should assess both objective exposure and subjective appraisal, including perceived threat, coping responses, post-incident support, and recovery processes, to clarify how patient violence is associated with ProQOL.

This study offers empirical support for a more integrated understanding of ProQOL in mental health nursing, highlighting the relevance of both organizational conditions and individual psychological resources. In clinical settings, where compassion is central to patient care, promoting compassion satisfaction may be associated not only with provider well-being but also with care quality [5]. In addition to efforts to reduce self-coldness-related selfcritical response patterns, institutions may benefit from supporting the development and application of self-compassion-based strategies.

The present findings have several practical implications. Improving the work environment, particularly in areas such as staffing adequacy, leadership support, and interprofessional collaboration, through regular assessments of nurses' perceptions may be an important factor in supporting employee well-being [7]. Additionally, because mental toughness and self-compassion are modifiable characteristics, psychological skills training and selfmanagement programs may be incorporated into continuing education for nurses.

There is growing evidence that interventions incorporating components such as improved communication, assertiveness training, mindfulness practices, psychoeducation, and structured supervision are beneficial for mental health nurses. Compassion-focused therapy and mindfulness-based self-compassion programs have also been proposed as promising approaches [30]. However, empirical evidence supporting the application of these interventions in nursing populations remains limited. Therefore, prospective and intervention-based studies are required to evaluate the feasibility, effectiveness, and scalability of such programs.

Moreover, the differentiated pathways identified in this study suggest that self-compassion training may primarily enhance compassion satisfaction, while complementary strategies that directly address self-critical cognitive patterns may be needed to reduce compassion fatigue. Further research is required to elucidate the distinct mechanisms through which these emotional resources operate. From a policy perspective, there is an increasing need to adopt a preventive occupational health model that addresses both systemic stressors (e.g., patient aggression) and individual vulnerabilities (e.g., susceptibility to burnout). Therefore, nursing education curricula may benefit from incorporating structured training in self-compassion, mental toughness, and reflective practice to better prepare nurses for emotionally demanding clinical environments. Furthermore, previous research has predominantly focused on identifying protective factors that prevent burnout. Future studies should explore how to actively cultivate positive psychological states such as flourishing, meaning, and engagement in mental healthcare settings.

This study had several limitations. First, the survey measured workplace conditions, violence exposure, psychological resources, and ProQOL at a single time point. Therefore, the model cannot determine whether these factors preceded the ProQOL outcomes or reflected them.

Second, all variables were based on participants' reports. The estimates may therefore have been influenced by recall error, socially desirable responding, or commonmethod variance. Subsequent studies should combine questionnaire data with independent indicators, such as staffing records, workplace-violence incident reports, supervisor or peer ratings, and documentation of post-incident support.

Third, the MTQ-18 demonstrated limitations in its factor structure in this sample. Although the total score showed good internal consistency and was retained as a global indicator, findings related to mental toughness should not be interpreted at the subdimension level. Rather, they should be understood as reflecting overall mental toughness. Future research should further examine the construct validity, measurement invariance, and culturally appropriate interpretation of the MTQ-18 in Korean nursing populations.

Fourth, although SEM provided a more appropriate framework for modeling shared variance among correlated indicators, the substantial overlap among some observed subdimensions suggests that these facets should not be interpreted as fully independent predictors. Additionally, because the revised structural model included theory-guided but not fully prespecified modifications, the final model should be regarded as provisional and should be replicated in independent samples.

Finally, the generalizability of the findings is limited owing to the use of convenience sampling and the regional focus on Korea. Nevertheless, the core constructs are conceptually applicable across diverse healthcare settings. While the numerical estimates should be interpreted cautiously, the relational patterns identified in this study may be transferable to mental health nursing populations in other countries, particularly in environments characterized by high emotional labor and exposure to violence. Future studies should build on this model by applying it to diverse clinical settings and exploring additional psychological and contextual factors.

This study suggests that both environmental conditions and internal psychological traits jointly associated with mental health nurses' professional quality of life. A positive nursing work environment and higher levels of mental toughness were associated with greater compassion satisfaction and reduced compassion fatigue. The two-factor structure of self-compassion revealed different pathways to emotional well-being. These findings emphasize the importance of integrating structural improvements with individualized psychological support to sustain a resilient and compassionate workforce in psychiatric settings. As emotional labor is a core component of mental health nursing, future interventions should reduce stressors while actively fostering strengths such as mental toughness and self-compassion. Promoting a dual focus on external support and internal capacity building may ultimately enhance care quality while safeguarding providers' emotional sustainability.