Leading the way to mental health: the role of mental health-specific leadership and stigmatizing attitudes for intra- and interpersonal mental health outcomes

MHSL stems from the broader concept of health-specific leadership, which focuses on leader behaviors aimed at improving health outcomes, such as informing employees about health topics (Adler et al., 2014; Gurt et al., 2011). Health-oriented leadership goes further than health-promoting leadership (i.e. general leader behaviors that support employee health, such as providing positive feedback and fostering a supportive work environment; Yao et al., 2021), by actively integrating health awareness into everyday interactions between leaders and employees, focusing on promoting well-being through these exchanges (Franke et al., 2014). Some studies linked these broader constructs to mental-health-related processes, such as disclosure intentions (Pischel and Felfe, 2023) and psychological safety (Esmaeilbeigi et al., 2025). This work, however, often treats mental health as just one aspect of overall health. We argue for the importance of focusing on mental health specifically, because mental health brings delicate challenges that general health does not, such as employees' reluctance to draw attention to it specifically (Pischel and Felfe, 2023). Leadership content that explicitly addresses mental health, rather than general health, may therefore more likely signal that the topic is safe to raise. Yet leadership research rarely focused on this specifically (see for exception: Paganin et al., 2023), and it remains unclear whether leaders' explicit focus on mental health adds value for employee mental health.

Building on health-specific leadership literature (Adler et al., 2014; Gurt et al., 2011), we define MHSL as leader behaviors reflecting clear commitment to supporting employee mental well-being. MHSL comprises (1) task-related behaviors (e.g. setting mental health goals, providing feedback) and (2) relationship-related behaviors (e.g. fostering open dialogue about mental health). Thus, mental health-specific leaders openly discuss and address employee mental health, take responsibility for it and actively involve employees in improvement efforts.

Although related to other constructs, MHSL is conceptually distinct in focus and scope. For example, psychological safety describes a team climate where employees feel safe to be vulnerable (Edmondson, 1999), while MHSL positions leaders as key agents fostering that safety. Leader–member exchange highlights trust and respect in leader–follower relationships (Graen and Uhl-Bien, 1995), while MHSL directs these qualities specifically toward recognizing and supporting employees' mental health needs. Ethical leadership emphasizes values such as fairness and care (Brown and Treviño, 2006), whereas MHSL translates these values into concrete mental health-focused behaviors. MHSL thus uniquely centers mental health within daily leadership practice.

Workplace mental health varies from mental illness to flourishing (Kelloway et al., 2023). We focus on burnout and work engagement as opposite ends of this spectrum. Burnout is a psychological syndrome marked by exhaustion, reduced cognitive and emotional control, and mental distancing (Schaufeli et al., 2002). Work engagement is “a positive, fulfilling work-related state of mind characterized by vigor, dedication, and absorption” (Schaufeli et al., 2002, p. 74).

To develop our predictions about MHSL's relationship with burnout and work engagement, we draw on the JD–R model (Demerouti et al., 2001), which distinguishes two pathways. In the health-impairment pathway, leaders can function as a demand, for example by imposing excessive workload, which contributes to burnout. In the motivational pathway, leaders can function as a resource, for example by offering support, which fosters work engagement. Prior research showed that constructive leadership may function as a resource that relates to followers' work engagement (Pletzer et al., 2024). MHSL can be considered a mental-health-specific form of constructive leadership. Rather than demonstrating general care, leaders demonstrate care specifically for followers' mental health. Drawing on the role of leadership in the JD–R model, we propose that MHSL supports employees' intrapersonal mental health through both pathways. MHSL behaviors, such as clearly communicating expectations, identifying early signs of overload and encouraging openness about help-seeking, can help employees set boundaries and mitigate stressors (Milligan-Saville et al., 2017). Simultaneously, MHSL fosters trust, demonstrates empathy and provides tailored, practical support. These behaviors can foster psychological safety (i.e. the belief that one can speak up about challenges, including mental health concerns, without fear of negative consequences; Edmondson, 1999). Such safety supports early help-seeking and reduces burnout risks (Edmondson, 1999).

In addition to direct support, MHSL encourages employees to actively participate in mental health initiatives. Such involvement can itself function as a resource by reinforcing a sense of purpose and influence over one's work environment. These effects are especially likely when leaders are perceived as caring, responsive and knowledgeable, which relates to reduced burnout and increased engagement (Decuypere and Schaufeli, 2020; Montano et al., 2017). Moreover, leaders who proactively develop mental health literacy contribute to a culture where mental health resources are more visible and accessible (Dimoff et al., 2016). Insights from leadership intervention studies support MHSL's potential: Brief feedback training reduced emotional exhaustion (Eastburg et al., 1994), while enhanced coaching skills improved engagement (Peláez Zuberbuhler et al., 2020). Thus:

Because leadership affects both individuals and teams (Burke et al., 2006), we expect MHSL to also influence interpersonal outcomes. Leaders foster supportive environments of trust, care and collaboration through high-quality relationships that guide team interaction (Kim et al., 2021). For example, Paganin et al. (2023) found that MHSL reduced interpersonal conflict at work.

We examine two key interpersonal outcomes of effective team functioning: (1) perceived team support, reflecting employees' experience of a social environment characterized by warmth, trust, reliability and safety (West et al., 2004) and (2) health-related extra-role performance, involving voluntarily helping colleagues facing mental or physical health challenges (Bowling, 2010). We focus on health-related extra-role performance because it illustrates how MHSL can foster compassion and inclusion through everyday interactions. Although MHSL emphasizes mental health, we adopt a broader view for health-related extra-role performance. Namely, employees may not know whether a colleague's struggles are mental or physical. Even so, support in either case may strengthen team cohesion.

MHSL may foster team support and health-related extra-role performance in two interconnected ways. First, by modeling supportive behaviors, such as checking in with team members, encouraging dialogue and offering help, leaders signal that mental health is valued, reinforcing norms of care and openness (Bandura, 1986; Nielsen and Daniels, 2012). Over time, these norms may foster psychological safety and encourage employees to engage in behaviors beyond their formal role expectations (Edmondson, 1999). Second, MHSL may also improve how team members interact. By creating a safe space to talk about mental health and encouraging support during difficult times, MHSL can foster reciprocal behaviors among team members. These interactions can build mutual respect, cooperation and proactive problem-solving over time (Cropanzano and Mitchell, 2005). Therefore:

The strength of MHSL's relationship with intra- and interpersonal outcomes may depend on employee characteristics, particularly employees' stigmatizing attitudes toward mental health. Stigmatizing attitudes refer to negative evaluative beliefs about individuals with mental health issues, such as viewing them as incompetent, unreliable or unable to cope with job demands (Link and Phelan, 2001).

Regarding intrapersonal mental health outcomes, we expect stigmatizing attitudes to undermine MHSL's beneficial role in reducing burnout and enhancing work engagement. Research on stereotypes shows that individuals categorize themselves and others based on salient characteristics, such as mental health status, which are shaped by persistent cultural stereotypes (Turner, 1999). Mental health issues are often seen as signs of weakness or unprofessionalism (Abdullah and Brown, 2011). Once internalized, these stereotypes influence how individuals process information. People tend to reject perspectives that challenge their beliefs and selectively attend to cues that confirm existing views (Howard and Rothbart, 1980). Therefore, employees with strong stigmatizing attitudes may feel discomfort or defensiveness toward MHSL behaviors emphasizing mental health openness, as these behaviors conflict with their self-image. This can trigger cognitive and emotional dissonance, which may lead to disengagement from MHSL initiatives, avoiding discussions about mental health or downplaying their own difficulties. In turn, such resistance may undermine MHSL's benefits for reducing burnout.

Similarly, stigmatizing attitudes may weaken MHSL's benefits for work engagement. According to person–environment fit theory, value clashes between employees and leaders can diminish engagement (Kristof-Brown et al., 2005). MHSL aims to foster trust, emotional safety and support to encourage full employee investment (Adler et al., 2014). However, employees who hold stigmatizing attitudes toward mental health may see these behaviors as conflicting with their values or workplace norms that prioritize resilience or task focus over emotional connection. This misalignment reduces MHSL's relevance and motivational effect. Thus, MHSL is less likely to enhance work engagement among employees with stronger stigmatizing attitudes.

Regarding interpersonal outcomes, we expect strong stigmatizing attitudes to weaken the relationship between MHSL and both perceived team support and health-related extra-role performance, through two related mechanisms. Stigmatizing attitudes involve viewing individuals with mental health challenges as fundamentally different from oneself, reflecting an underlying bias. Central to this is drawing a boundary between “us” and “them”: Once someone is labeled through a stigmatizing lens, they are seen as categorically different, rather than as full member of one's own group, creating social distance (Corrigan et al., 2001; Link and Phelan, 2001). This matters because people generally feel more likely to identify with, connect with and support those they see as their in-group (Balliet et al., 2014). Beyond this social distance, a leader who models empathy and support for colleagues with mental health challenges typically invites reciprocation: Employees come to see mutual support as a team norm, and respond by supporting others themselves (Emerson, 1976). For employees who hold stigmatizing attitudes, both processes are disrupted. The social distance they feel toward colleagues experiencing mental health problems makes MHSL's emphasis on empathy and support feel less relevant, or even uncomfortable, while their conflicting beliefs weaken the reciprocal response that MHSL would otherwise invite. Together, these employees are less likely to experience the team climate MHSL fosters as warm and trustworthy and to feel supported by it themselves, and are less likely to take to heart the norm of mutual support that MHSL establishes, dampening their motivation to offer help, show initiative or contribute to an inclusive team atmosphere.

Participants were recruited from a Dutch public organization as part of a European mental health project H-WORK, with ethical approval obtained from the Ethics Review Board of University of Amsterdam (2021-WOP-13113). Senior managers distributed the online questionnaire link among project team employees. An information letter outlined the study's purpose, confidentiality and voluntary participation; participants then provided digital informed consent. To reduce social desirability bias, given the sensitivity of mental health topics and the distribution via managers, we assured participant anonymity. Results were reported only in aggregate form, with supervisors having no access to individual data. To further emphasize research independence, the university-hosted Qualtrics survey prominently displayed our university logo.

The organization has a matrix structure, with employees reporting to both formal supervisors and project managers. We focused on project managers as leaders given their central role in employees' daily work. Of the 417 employees who began the questionnaire, 267 completed it. We excluded 33 project managers (to focus on employee perceptions) and 24 participants who did not indicate a team. The final sample included 210 employees (79% male, 21% female) across 53 teams. Age was measured categorically: 0.5% were <25, 13.3% were 25–34, 24.3% were 35–44, 34.3% were 45–54, 23.8% were 55–64 and 3.8% were ≥65. Tenure was >10 years for 49.5%. Educational levels were predominantly Bachelor's (61.9%) and Master's (27.6%).

The survey was in Dutch. When needed, items were adapted and translated from English, using translation back-translation methods. Appendix 1 provides all measures. Table 1 displays Cronbach's alphas. MHSL was assessed with seven items adapted from the Health-Specific Leadership Scale (Gurt et al., 2011), rewording “health” to “mental health and well-being.” Stigmatizing attitudes were measured with a five-item subscale from the Opening Minds Survey for Workplace Attitudes (Dobson and Szeto, 2021). Burnout was assessed with the BAT-12 (Hadžibajramović et al., 2022) and work engagement with the UWES-9 (Schaufeli et al., 2006). Perceived team support was measured with a three-item subscale from the Team Performance Inventory (West et al., 2004), and health-related extra-role performance with a single item from the Organizational Citizenship Behavior Scale (Smith et al., 1983). Although multi-item scales are often preferred, single-item measures can be suitable for narrow, clearly defined constructs (Matthews et al., 2022). Given our focus on specific support behaviors toward unwell colleagues, a single-item measure was deemed appropriate.

Given age and gender differences in well-being (Inglehart, 2002), we considered these as potential control variables. Following Becker et al. (2016), correlations with outcomes were examined to assess the need for inclusion in the analyses.

To test our hypotheses, we conducted multilevel path analyses in Mplus 8.8, with employees (Level 1) nested within teams (Level 2), using maximum likelihood estimation with robust standard errors (MLR). This approach decomposes variance into a within-team component (individual level) and a between-team component (team level), allowing to examine relationships between MHSL and the outcomes at both the individual and team level.

We estimated eight models (one per outcome variable) with four including only main effects and four additionally including the MHSL × stigmatizing attitudes interaction. Separate models were preferred over a multivariate specification due to instability in the latter, including negative variance estimates at the team level. The limited number of clusters (k = 53) and small average cluster size (M = 3.96) likely prevented reliable parameter estimation in the multivariate model.

MHSL and stigmatizing attitudes were standardized prior to analysis. MHSL and the four outcomes were modeled at both levels. Because stigmatizing attitudes reflect an individual-level construct, it was standardized and specified as a within-level variable. The interaction term was computed as the product of the two standardized variables and included as an individual-level predictor. We explicitly distinguish effects significant at the conventional 0.05 level from those significant only at the more lenient 0.10 level; the latter are reported for transparency but interpreted with additional caution. Significant interactions were further examined using simple slopes (MODEL CONSTRAINT function in Mplus).

Table 1 displays descriptives and correlations. Because age related to MHSL (r = 0.17, p = 0.015) and stigmatizing attitudes (r = 0.18, p = 0.008), and gender related to burnout (r = 0.16, p = 0.024), we included these as individual-level covariates in subsequent analyses.

In the sampled population, team size ranged from five to 16 members. In our sample, we identified 53 distinct teams, with 1–9 respondents per team (M = 3.96; SD = 2.34). Intraclass correlations (Table 1) show that 2.7%–21.8% of the variance in the study variables resided at the team level, with the remaining 78.2%–97.3% at the individual level.

Figure 1 displays the theoretical model with key individual-level results (see Appendix 2 for full results). Hypothesis 1a, predicting a negative association between MHSL and burnout, was supported at the individual level at the 0.10 level (B = −0.08, p = 0.064), but not at the 0.05 level ^[1]. At the team level, the MHSL-burnout relationship was not significant (B = 0.26, p = 0.190). Hypothesis 1b, predicting a positive association between MHSL and work engagement, was supported at the individual level (B = 0.16, p = 0.043), but not at the team level (B = −0.52, p = 0.176). Hypothesis 2a, predicting a positive association between MHSL and perceived team support, was supported at the individual level (B = 0.26, p < 0.001), but not at the team level (B = 0.37, p = 0.679). Hypothesis 2b, predicting a positive association between MHSL and health-related extra-role performance, was supported at the individual level (B = 0.16, p = 0.025), but not at the team level (B = −0.06, p = 0.865).

Hypothesis 3, proposing that stigmatizing attitudes moderate MHSL's relationships with burnout (3a) and work engagement (3b), was not supported, as neither interaction was significant at the individual level.

Hypothesis 4a predicted that stigmatizing attitudes weaken the positive link between MHSL and perceived team support. The individual-level MHSL × stigmatizing attitudes interaction was not significant at the.05 level, but was significant at the 0.10 level (B = −0.08, p = 0.067). Simple slopes analysis suggested a stronger positive relationship between MHSL and team support at low stigmatizing attitudes (M − 1SD; B = 0.34, p < 0.001) compared to high stigmatizing attitudes (M + 1SD; B = 0.17, p = 0.031), though both slopes were significantly positive (see Figure 2).

Hypothesis 4b predicted that stigmatizing attitudes weaken the positive link between MHSL and health-related extra-role performance. The individual-level MHSL × stigmatizing attitudes interaction was significant (B = 0.17, p = 0.004). However, in contrast to Hypothesis 4b, simple slopes analysis showed that MHSL was positively related to health-related extra-role performance at high stigmatizing attitudes (M + 1SD; B = 0.36, p = 0.001), but not at low stigmatizing attitudes (M − 1SD; B = 0.01, p = 0.886) (see Figure 2).

This study examined how MHSL relates to intra- and interpersonal employee well-being, and whether these relationships are moderated by stigmatizing attitudes. At the individual level, MHSL related to intrapersonal outcomes (i.e. tentatively negatively to burnout and positively to work engagement) and interpersonal outcomes (i.e. positively to team support and health-related extra-role performance). At the team level, MHSL was not related to these outcomes. Stigmatizing attitudes moderated the relationships with interpersonal, but not intrapersonal, outcomes. Specifically, high stigmatizing attitudes tentatively weakened the positive MHSL-perceived team support relationship, and unexpectedly strengthened the MHSL-health-related extra-role performance relationship.

This study makes several theoretical contributions. First, it advances health-specific leadership research by defining and empirically examining MHSL. The findings for intrapersonal outcomes may reflect the nature of MHSL, which was operationalized primarily as fostering team resources and a supportive climate, rather than providing direct, individualized support to employees in distress. Drawing on the JD–R model (Demerouti et al., 2001), such resource-enhancing behaviors may function as a job resource that directly fuels work engagement by enhancing employees' sense of belonging, psychological safety and motivation. In contrast, burnout, and particularly emotional exhaustion, is more strongly driven by job demands (Demerouti et al., 2001). Because MHSL primarily adds resources rather than demands, its relationship with burnout may be less pronounced in a general workforce sample. This is consistent with recent work in a general workforce sample suggesting that resource-enhancing leadership behaviors tend to play a stronger role in fostering positive work-related states than in mitigating more distal strain outcomes such as burnout (Sijbom and Grutterink, 2026). However, the weaker findings for burnout do not necessarily preclude the strain-reducing potential of MHSL. Rather, this effect may more likely emerge among employees experiencing mental health challenges, for whom MHSL may extend beyond resource provision to reduce stigma, facilitate disclosure and provide more targeted support. This aligns with evidence that leader openness to mental health discussions promotes well-being in this population (Pischel and Felfe, 2023). Together, these findings suggest that the theoretical scope of MHSL with respect to strain outcomes is perhaps more circumscribed than the initial framework assumed, and that further theoretical development is needed to specify for which employees MHSL matters most.

Second, our study contributes to team leadership research (e.g. Burke et al., 2006) by linking MHSL to perceived team support and health-related extra-role performance at the individual level. Consistent with social cognitive theory (Bandura, 1986), leaders who model supportive, open behaviors may encourage similar responses among team members, fostering reciprocal support and mutual care within teams (Cropanzano and Mitchell, 2005). Notably, these findings emerged at the individual rather than the team level, suggesting that MHSL is interpreted and enacted idiosyncratically rather than as a uniform team-level phenomenon.

Third, our study extends stigma research by examining the perspective of stigmatizers. Stigmatizing attitudes did not uniformly moderate MHSL's associations across outcomes. For perceived team support, the interaction reached significance only at the 0.10 level, so this finding should be interpreted with caution. Tentatively, employees with stronger stigmatizing attitudes seemed to benefit less from MHSL, possibly because such attitudes promote out-group categorization and reduce trust (Turner, 1999). Self-stigma may further reduce help-seeking and social engagement (Prizeman et al., 2023; Vogel et al., 2006), limiting MHSL's effectiveness. In contrast, stigmatizing attitudes strengthened the association between MHSL and health-related extra-role performance. Employees with low stigmatizing attitudes showed no association, suggesting they may engage in such behaviors through other motivational pathways. This divergence may reflect a fundamental distinction between perceptual and behavioral outcomes: while stigmatizing attitudes may distort how employees perceive the team climate, behavior is also shaped by social norms. Leaders may establish such norms by modeling inclusive behaviors (Bandura, 1986), thereby creating normative pressure on employees, particularly those with stigmatizing attitudes, to align their behavior accordingly. Future research should examine these mechanisms and assess whether these effects persist over time.

While our study provides valuable insights into MHSL, it also has limitations. First, its correlational design prevents causal inferences and limits assessment of MHSL's long-term effects. Future research could use longitudinal designs to examine how MHSL evolves and impacts intrapersonal and interpersonal outcomes over time.

Second, the reliance on employee perceptions to measure MHSL risks common-method bias and may inflate observed relationships (Podsakoff et al., 2003). Employees may also be unaware of certain supportive leader behaviors occurring behind the scenes, limiting a full capture of MHSL. To capture MHSL more comprehensively, future research could also employ leader self-reports and objective indicators like employee involvement in mental health initiatives or resource access.

Third, the team-level relationships of MHSL were nonsignificant for all outcomes, potentially caused by power constraints at the team level. Moreover, the relatively small team sizes prevented us from examining within-team variability in perceptions of MHSL. Future research should collect data from larger numbers of teams with more respondents per team. This would allow examination of how team-level MHSL as well as its dispersion relates to collective outcomes such as team climate, psychological safety or shared perceptions of team support.

Our findings offer several practical implications that can inform organizational policy and leadership training frameworks aimed at creating mentally healthy, inclusive workplaces. First, our findings highlight the importance of leadership development that equips leaders with mental health-specific skills within training frameworks. Although MHSL interventions remain limited, evidence suggests that mental health training for leaders can foster work engagement and supportive team environments (Stuber et al., 2021). A critical starting point is improving leaders' mental health literacy and reducing stigma. Mental health leadership training programs (Dimoff et al., 2016) may use role-playing and scenario-based exercises to help leaders recognize distress, engage in supportive conversations and refer employees to appropriate resources. Such interventions may be particularly valuable in high-stigma, male-dominated industries (e.g. construction, agriculture, manufacturing), where mental health challenges are prevalent (Roche et al., 2016).

Second, our findings show that MHSL appears to operate primarily through individual-level interactions rather than uniform team processes. Leaders should therefore prioritize tailored one-on-one conversations that address employee well-being. Simultaneously, leaders can normalize mental health discussions in team settings by acknowledging stressors and modeling openness, for instance by sharing coping strategies (Dimoff and Kelloway, 2019; Ito et al., 2025; Milligan-Saville et al., 2017). In high-risk contexts such as healthcare, leaders may further support well-being by acknowledging emotional demands, encouraging brief team debriefings after difficult cases and creating space for staff to discuss stress. Such practices can enhance psychological safety and promote early help-seeking (Edmondson, 1999; Pischel and Felfe, 2023).

Third, our findings tentatively suggest that stigmatizing attitudes may limit MHSL's association with team support. Organizations should therefore complement leadership development policies with anti-stigma initiatives. Programs like Mental Health First Aid can reduce misconceptions and foster supportive attitudes (Hadlaczky et al., 2014). Embedding mental health into onboarding, regular team check-ins and leadership communication can further normalize the topic and encourage disclosure (Pischel and Felfe, 2023).

Finally, MHSL appears especially effective in boosting health-related extra-role performance in teams where mental health stigma is more prevalent. This suggests that leaders can foster prosocial behaviors even in high-stigma contexts by modeling supportive mental health behaviors. Combining leadership development with stigma reduction may therefore be promising in fostering an environment that encourages health-related extra-role performance.

Our study highlights the role of MHSL in project teams and shows that it links to both intrapersonal (work engagement) and interpersonal outcomes (team support and health-related extra-role performance). Stigmatizing attitudes shaped these interpersonal relationships, strengthening the association with health-related extra-role performance and, more tentatively, weakening the association with team support. These findings deepen the theoretical understanding of how leadership behaviors are linked to supportive behaviors in teams and offer valuable practical insights for organizations aiming to build mentally healthy, inclusive workplaces.

The supplementary material for this article can be found online.