Mental health challenges and work engagement: The results from a cross-sectional study of Norwegian priests

Abstract Priests’ professional demands are many and varied. To experience work engagement, it is likely particularly important to have adequate job resources when dealing with Mental Health Challenges (MHC). Hence, we sought to compare the occurrence of work engagement in priests with and without MHC and to study the associations between work resources and work engagement. Eight hundred and four priests completed the Hospital Anxiety and Depression Scale, the Utrecht Work Engagement Scale, and self-reported measurements of job demands and job resources. A significant percentage of the priests (71.4%) were highly engaged in their jobs, particularly those without MHC. Work engagement was associated with job resources and not with job demands as described in the job demand resource theory. The high level of work engagement among priests should be considered a positive asset for the Norwegian Church. Thus, the Church should preserve this engagement by maintaining a working environment with adequate job resources. Given the high incidence of MHC among priests, it may be beneficial to focus specifically on the job resources needed for this group to maintain work engagement.

Abstract: Priests' professional demands are many and varied. To experience work engagement, it is likely particularly important to have adequate job resources when dealing with Mental Health Challenges (MHC). Hence, we sought to compare the occurrence of work engagement in priests with and without MHC and to study the associations between work resources and work engagement. Eight hundred and four priests completed the Hospital Anxiety and Depression Scale, the Utrecht Work Engagement Scale, and self-reported measurements of job demands and job resources. A significant percentage of the priests (71.4%) were highly engaged in their jobs, particularly those without MHC. Work engagement was associated with job resources and not with job demands as described in the job demand resource theory. The high level of work engagement among priests should be considered a positive asset for the Norwegian Church. Thus, the Church should preserve this engagement by maintaining a working environment with adequate job resources. Given the high incidence of MHC among priests, it may be beneficial to focus specifically on the job resources needed for this group to maintain work engagement.
ABOUT THE AUTHOR Bjørn Lau is a licensed psychologist with a Ph.D. in Psychology. Lau has worked as a visiting professor of organizational psychology at the University of Oslo, and as a clinical psychologist and an executive organizational psychologist. He currently serves as Head of the Department of Psychology at the University of Oslo. Lau's research has been conducted in the areas of work and organizational psychology, especially in the fields of occupational health psychology and clinical occupational psychology. His research has been aimed at organizational and psychosocial working conditions of importance to health, well-being, and motivation. The conditions studied include management, shift work, restructuring, bullying, as well as prominent models like the demand-control model and the effort-reward imbalance model. Another direction in his research has been to examine the importance of working conditions for returning to work among employees on sick leave due to musculoskeletal disorders and mental disorders.

PUBLIC INTEREST STATEMENT
The priest profession is diverse and demanding. This suggests the need for an optimal balance between work-related demands and available job resources. Nevertheless, little research has been done among priests on such matters. In a previous study, several Norwegian priests reported mental health problems related to anxiety and depression. Therefore, this study examined work engagement among priests with and without these health challenges. Approximately twothirds of the priests in the Norwegian Church attended the survey. A large proportion of them reported being highly engaged in their work, especially those without mental health challenges. Furthermore, work engagement was associated with job resources, among all priests. The high level of work engagement among priests should be considered a definite advantage for the Norwegian Church. The Church should uphold this engagement by ensuring a working environment with adequate job resources.

Introduction
Mental health challenges (MHC), such as anxiety and depression, are prevalent in the adult population in western countries (World Health Organization, 2017). The annual prevalence of MHC is estimated to be 17.6 percent in the general working population, with a lifetime prevalence of 29.2 percent (Steel et al., 2014). Such disorders can be challenging to cope with when working and are associated with reduced work ability (Jansson & Gunnarsson, 2018) and job performance (Clark, DiBenedetti, & Perez, 2016;Jones, Latreille, & Sloane, 2016). Moreover, a large proportion of long-term sickness absence (Henderson, Harvey, Øverland, Mykletun, & Hotopf, 2011) and new disability cases are associated with anxiety and depression (Knudsen, Øverland, Hotopf, & Mykletun, 2012).
In a previous study of a representative sample of Norwegian ministers, high rates of anxiety problems were found, especially among female priests (Lau, 2018). Nearly one-third of the female priests had a score indicating a need for further clinical follow-up for anxiety problems, while the corresponding figure among male priests was approximately 20 percent. These numbers are significantly higher than those found in the Norwegian population (Bjelland et al., 2009;Grav, Stordal, Romild, & Hellzen, 2012;Sterud, Hem, Ekeberg, & Lau, 2008;Stordal et al., 2001) and higher than others have found in studies of priests (Proeschold-Bell et al., 2013). A significantly lower prevalence of symptoms of depression was found (9.3 percent) than of anxiety among Norwegian priests. This occurrence was at the general population level (Grav et al., 2012;Stordal et al., 2001).
In Norway, Evangelical Lutheran Christianity accounts for approximately 70 percent of the population. Most people who confess the Lutheran Evangelical faith are members of the Norwegian Church. The Norwegian Church was Norway's official church until 2016 when the church was subordinated directly to the state. Priests' professional demands and tasks are summarized in six key roles: pastor, teacher, preacher, ritual leader, organizer and administrator (Carroll & McMillan, 2006). In addition, clergy are responsible for managing the role of the church in the community, keeping church operating costs low (Kuhne & Donaldson, 1995), helping members of the congregation with psychological challenges (Wang, Berglund, & Kessler, 2003) and balancing their own work demands with family and spousal expectations (Morris & Blanton, 1998;Lee & Iverson-Gilbert, 2003). Such a work situation could be perceived as even more demanding for employees with MHC, as they also have to cope with their own health challenges related to anxiety and depression. Therefore, in the priestly profession, it is probably especially important to have adequate work resources when dealing with MHC.
The job demands-resources model (JD-R model) is a model that considers job demands against job resources. The starting point for the model is that all aspects of a job can be classified as job demands or job resources (Demerouti, Nachreiner, Bakker, & Schaufeli, 2001). High job demands are considered to create a health-deteriorating process, supported by studies showing that high job demands predict burnout and depression, long-term sick leave and lower job performance (Bakker, Demerouti, de Boer, & Schaufeli, 2003;Bakker, Demerouti, & Verbeke, 2004;Hakanen, Schaufeli, & Ahola, 2008;Simbula, 2010). Job resources, on the other hand, are physical, mental, social or organizational aspects that are functional in achieving work objectives. Job resources have been shown to buffer the adverse effects of high job demands on experienced burdens (Bakker, Demerouti, & Euwema, 2005;Bakker, van Veldhoven, & Xanthopoulou, 2010;. They can also stimulate personal growth, learning and development through a motivational process (Bakker, 2011; and have been shown to predict work engagement, organizational commitment, low absenteeism and extra-role performance (Bakker, Demerouti, De Boer, & Schaufeli, 2003;Bakker, Demerouti, & Sanz-Vergel, 2014;Bakker et al., 2004;Hakanen et al., 2008;Simbula, 2010). Because job resources are especially useful in situations with high job demands that challenge employees to learn and engage in new behaviors, motivation is substantially influenced by job resources when job demands are high (Bakker, Hakanen, Demerouti, & Xanthopoulou, 2007;Hakanen, Bakker, & Demerouti, 2005).
In this study, we wanted to examine factors related to the motivational process in the jobdemands resource model, where particularly work engagement is considered as an outcome of available job resources at work. Work engagement is defined as a persistent and positive affectivemotivational state in employees characterized by vigor, dedication, and absorption (Schaufeli, Salanova, González-romá, & Bakker, 2002). As far as we know, none of the few existing studies of work engagement among priests, nuns and religious workers have compared their level of work engagement with other occupational populations or the general population (Ariza-Montes, Leal-Rodríguez, Ramírez-Sobrino, & Molina-Sánchez, 2019; Ariza-Montes, Molina-Sánchez, Ramirez-Sobrino, & Giorgi, 2018;Phoseka, 2017). However, by comparing the average values reported with the norms for the instrument used in these studies, the Utrecht Work Engagement Scale (Schaufeli & Bakker, 2004;Schaufeli et al., 2002), it turns out that priests have comparably higher work engagement than the general population. Hence, we expected Norwegian priests to have a higher level of work engagement than anticipated by the norm data. As a follow-up to this, we wanted to examine whether the background factors, such as age, gender, and marital status, were related to work engagement among the clergy. To the best of our knowledge, such associations have not been reported in the research literature. This led us to test an exploratory hypothesis as to the direction of such possible associations. Thus we had the following hypotheses: Hypothesis 1a: The level of work engagement among priests in the Norwegian Church is higher compared to the general population, as specified by the norms for The Utrecht Work Engagement Scale.
Hypothesis 1b: The level of work engagement among Norwegian priest varies according to gender, age, and marital status.
However, it has been found that MHC, such as symptoms of anxiety and depression, are associated with lower levels of work engagement (Innstrand, Langballe, & Falkum, 2012). Hence, we wish to examine the level of work engagement among priests with a high level of anxiety and depression symptoms compared to priests without these health challenges. Because depression symptoms are associated with decreased mood, less ability to enjoy and experience interests and negative rumination, we assume that priests with symptoms of depression are less able to engage in work. Anxiety problems are correspondingly expressed through emotional anxiety, physiological symptoms caused by physiological activation and worrying thoughts. As with depression problems, we assume that anxiety problems interfere with work engagement. Thus we have the following hypothesis: Hypothesis 2: Priests with MHC, that is, with scores above the clinical level in subscales of anxiety and depression (one or both), are expected to report lower levels of work engagement than priests without such health challenges.
A number of studies have identified associations between work resources and work engagement, while associations between hindrance job demands and work engagement are rarely detected (Lesener, Gusy, & Wolter, 2019). We wanted to test whether we could find the same pattern among the priests of the Norwegian Church. However, few studies have examined whether job resources for people with MHC are associated with work engagement at the same level as for those without such health issues. We believe these associations to be stronger among MHC priests because they need more resources to handle work and become engaged as they also have to cope with the challenges of anxiety and depression. Consequently, we want to examine whether job resources are positively linked to work engagement in the same way for priests with MHC as for those without such symptoms. Thus, we wanted to test the following hypotheses: Hypothesis 3: Among priests, work engagement is associated with job resources and not with job demands.
Hypothesis 4: Job resources are expected to be more positively associated with work engagement among priests with MHC than among priests without such health challenges.

Procedure and data collection
In the spring of 2012, all employees of the Norwegian Church were invited by e-mail to participate in a web survey aimed at examining working conditions, health, and work engagement. Central and local employers prepared electronic lists of the e-mail addresses of all the employees in the Norwegian Church. In addition to the invitation, this email included a written description of the project and an explanation of how to take part in the study. The collection of the data was completed in spring 2012.

Participants
The response rates were difficult to determine because we only obtained the e-mail addresses of those employed in the Norwegian Church without information on employees' category, gender, age, or marital status. However, our best estimate is that approximately 60 percent of priests in the Norwegian Church participated in the survey based on a comparison of the number of respondents with those registered in different job categories in the employer payroll and personnel systems.

Background variables
Age, gender and marital status were self-reported. The respondents were mainly male (72.6 percent), reflecting the population of the profession. A substantial proportion (56.7 percent) were over 50 years of age. Eighty-five percent lived in a relationship (married or cohabited), while 15 percent lived without a partner (single, separated, divorced or widowed). For more details, see (Lau, 2018).

Job demands
The participants responded to the Norwegian version (Lau, 2008) of the standardized, selfadministered Effort-Reward Imbalance Questionnaire (ERI-Q) (Siegrist et al., 2004) and The Nordic Questionnaire on Psychological and Social Factors at Work (QPSNordic), (Wännström, Peterson, Åsberg, Nygren, & Gustavsson, 2009). These instruments contain scales that measure distinct types of job demands, defined in the JD-R model as physical, psychological, social, or organizational aspects of the job, that require sustained physical and/or psychological effort or skills. ERI-Q has a scale that measures job demands referred to as Effort. The effort scale questions asked if the items described a typical work experience and, if so, if the conditions caused stress. The options for response included 1 = "does not apply"; 2 = "does apply, but not strained"; 3 = "does apply and somewhat strained"; 4 = "does apply and strained"; and 5 = "does apply and very strained." The Effort scale was calculated by averaging the scale item scores. QPSNordic measures four types of job demands; Quantitative demands, Decision demands, Learning demands, and Role conflict. For all questions, a five-point response scale was used, ranging from 1 (very seldom or never) to 5 (very often or always). The mean values for the items in each subscale were calculated. Cronbach's alpha values and descriptive statistics for the scales that measured the job demands are shown in Table 2.

Job resources
Job resources are defined in the JD-R model as physical, mental, social or organizational aspects that are functional in achieving work objectives. In this study they were also measured with scales taken from ERI-Q (Lau, 2008) and QPSNordic (Wännström et al., 2009). From ERI-Q, three reward scales were used; Esteem, Job promotion, and Job security. The items measuring reward were similarly framed as the Effort scale in ERI-Q, although with reverse coding. Thus, the lower the overall reward score is, the higher the subjective distress rating due to low rewards. The reward scale scores were calculated by averaging the scale item scores. From QPS-Nordic, we used five scales that measure various forms of job resources; Role clarity, Control over decisions, Control of work pacing, Support from superiors, and Support from colleagues. A five-point response scale ranging from 1 (very seldom or never) to 5 (very often or always) was used for all questions. The mean values for the items were calculated for each subscale. Table 1 shows Cronbach's alpha values and descriptive statistics for the scales that measured the job resources.

Symptoms of anxiety and depression
The Hospital Anxiety and Depression Scale (HADS) (Zigmond & Snaith, 1983), a survey with 14 items scored from zero to three, was used to measure anxiety and depression symptoms. A principal component analysis supported a two-factor solution; the Cronbach's alphas were 0.83 and 0.76 for the anxiety and depression scales, respectively. As a result, a subscale for anxiety (HAD-A) and a subscale for depression (HAD-D) were established with seven items each. Forms were included if at least 5 of the 7 items had been completed for each subscale. The values were calculated by multiplying the mean of the valid answers by 7. Because the best balance between sensitivity and specificity for HADS as a screening instrument is most frequently reported with a cut-off score of 8 (Bjelland, Dahl, Haug, & Neckelmann, 2002), this value was used for the present study. In our sample, 7.4 percent had both a high HAD-A score and a high HAD-D score. Furthermore, 16 percent had a high score on only the HAD-A subscale and 2 percent on only the HAD-D subscale. The remaining 74.6 percent of the sample had neither high HAD-A nor HAD-D scores. Based on these results, a new dichotomous variable called Mental Health Challenges (MHC) was created. Participants with high scores on the HAD-A and HAD-A subscales (only one or both) were merged into one category (MHC+), while priests who did not have high scores on any of the scales were placed in the second category (MHC-).

Work engagement
Work engagement is defined as a positive, fulfilling work-related state of mind. The Utrecht Work Engagement Scale (UWES) (Schaufeli et al., 2002) was used to measure the levels of engagement at work. The UWES consists of three scales: vigor (6 items), dedication (five items) and absorption (six items). All items are scored on a 7-point frequency rating scale ranging from 0 (never) to 6 (always). A principal component analysis did not support a three-factor solution for this instrument. Therefore, as recommended in the UWES manual (Schaufeli & Bakker, 2004), a one-factor solution with a Cronbach's alpha coefficient of 0.95 was used. Then, we created a dichotomous version of this scale to examine the number of priests with high levels of work engagement. We used the 75 th percentile value 4.67, described in the UWES manual as the cut-off for "high" levels of work engagement (Schaufeli & Bakker, 2004). In our sample, as many as 71.4 percents had UWES scores above this cut-off.

Methods of analyses
The first hypothesis (Hypothesis 1a), that the priest would have a higher level of work engagement than the ordinary working population was tested by calculating a t-test based on the mean value and the standard deviation given in the norm material for the Utrecht Work Engagement Scale and the values observed among the priests in this study.
Analyses of variance were also conducted to test Hypothesis 1b, whether the average work engagement values differed by gender, age or marital status. Correspondingly, Pearson's chisquare tests were used to determine whether the occurrence of high levels of work engagement differed by gender, age group or marital status.
To test Hypothesis 2, whether priests with MHC had lower average work engagement than priests without MHC, an analysis of variance was conducted. In this analysis, the UWES score was entered as the dependent variable, and the dichotome MHC was entered as the independent variable. Differences in the frequency of high UWES values between the two MHC groups were tested using Pearson's chi-square test.
A series of Pearson correlations were calculated between the psychosocial work components and work engagement in order to test Hypothesis 3, that work engagement would be associated with job resources and not with job demands. These analyses were done separately for high MHC and low MHC priests.
To test Hypothesis 4, that job resources would be more strongly associated with work engagement among priests with MHC than among priests without MHC; separate Pearson correlation analyzes between the individual job resource variables and work engagement was conducted for priests with MHC and without MHC. Subsequently, these correlations were compared based on Fisher r-to-z transformation.
The data were analyzed using SPSS for Windows (version 25, IBM Corp., Armonk, NY).

Ethics and approvals
Informed consent was appropriately obtained from the participants. The study was conducted following the World Medical Association Declaration of Helsinki and with permission from the Norwegian Centre for Research Data.

Results
The first hypothesis (Hypothesis 1a) was that the priests in this study would have a higher level of work engagement than the one given in the UWES manual (Schaufeli & Bakker, 2004). As shown in Table 1, the average among priests was 5.17 (SD = 1.04), while the norm material based on 9,679 respondents, declares an average value of 3.74 (SD = 1.17). A t-test showed that the priests had a significantly higher mean value than the specified value in the manual (t = −33.7; df = 10489; p < 0.0001). Thus, we found support for Hypothesis 1a. The support for this hypothesis is also illustrated by the fact that when applying the UWES user manual's 75th percentile to make a dichotomous version of this scale, we found that 71.4 percent of the clergy in our sample had scores above the cut-off.
We found only partial support for Hypothesis 1b that the level of work engagement among Norwegian priest varies according to gender, age, and marital status. As shown in Table 1, no gender differences or differences according to marital status were found for either the prevalence rate of high work engagement or for average values. However, there was significantly higher work engagement related to age, both concerning the number of employees with high scores (χ2 = 44.88) and average values (F = 13.06), where priests over fifty years of age had notably high levels of work engagement. Based on these results, we adjusted for age in the subsequent analyses.
Hypothesis 2 that priests with MHC, were expected to report lower levels of work engagement than priests without such health challenges was supported. As shown in Table 1, the priests with MHC had significantly and moderately lower average UWES values than priests without MHC (F = 69.26), and fewer priests with MHC reported a high level of work engagement (52.7 percent vs. 77.8 percent, χ2 = 46.58).
The third hypothesis was that work engagement would be associated with job resources and not with job demands. Table 2 lists the scales used to measure job characteristics which, according to the JD-R model, can be referred to as job demands and job resources, respectively. The Cronbach's alpha values show that most scales had an acceptable fit. A visible exception was the only scale consisting of two items; Job-security. Besides, the scales measuring Decision demands and Learning demands had somewhat modest alpha values. Table 3, Hypothesis 3 was mainly supported as work engagement was generally associated with job resources and not job demands. Except for the decision demands and role conflict scales, no significant correlations between job demands and work engagement were found. These two correlations were found only among priests without MHC, and both associations were small. In contrast, virtually all the job resource variables were associated with work engagement for priests with and without MHC, respectively.

As shown in
In general, the fourth hypothesis that there should be stronger correlations between job resources and work engagement among priests with MHC compared to priests without such health challenges was not supported. As shown in Table 3, only the z-value for Esteem indicated such a difference, while none of the other correlations were significantly different from each other.

Discussion
In line with Hypothesis 1a, it was found that priests have a higher level of work engagement than the general working population. The exploratory hypothesis that work engagement levels would vary with background factors such as gender, age, and marital status (Hypothesis 1b) was partly supported, as the elderly priest had higher levels of work engagement. Priests with mental health challenges had lower levels of work engagement than priests without these health challenges. This finding was consistent with Hypothesis 2. The findings mainly supported the third hypothesis that work engagement would be associated with job resources and not with job demands. However, the fourth hypothesis that there should be stronger correlations between job resources and work engagement among priests with MHC compared to priests without such health challenges was not supported.
There may be several reasons for the high level of work engagement among the priests. Several of the priests have probably chosen their profession based on a religious vocation. Consequently, in a job that matches their values and beliefs, it would be natural to experience a high level of work engagement. At the same time, the priesthood is characterized by several demanding tasks (Morris & Blanton, 1998;Carroll & McMillan, 2006;Kuhne & Donaldson, 1995;Lee & Iverson-Gilbert, 2003;Wang et al., 2003). However, if these demands are perceived as challenging, this is a situation known to be associated with a high level of work engagement (Searle & Auton, 2015). Nevertheless, it can not be concluded whether the high work engagement found among the priest in this study is related to values and religiosity or specific work challenging tasks in their professional practice.
In line with Hypothesis 1b, an effect of age on work engagement was found, including high prevalence and high average UWES values, among priests over 50 years old. There may be various reasons for these results. First, if the most engaged priests remain in the clergy, they may indicate a healthy worker effect. Nor can we rule out the fact that work experience and confidence, accumulated over time, contribute to work engagement. A third possibility is that this represents a generational difference in which younger priests are not as engaged in their work as the generation older than 50.
The second hypothesis that priests with mental health challenges would report lower levels of work engagement than priests without such health challenges was supported. This hypothesis was based on the assumption that anxiety and depression symptoms may make it more difficult to engage in work. Because depression is associated with decreased mood, less ability to enjoy and experience interests, and negative rumination, priests with depression symptoms may be less able to engage in work with vigor, dedication, and absorption. Correspondingly, anxiety expressed through emotional anxiety, autonomous physiological activation, and worrying thoughts was assumed to interfere with work engagement negatively. However, it should be noted that although priests who scored high on anxiety and depression had less work engagement, the results showed that 52.7 percent of them still had high levels of work engagement. This number is higher than that found in the norm material, as described in the manual for the UWES (Schaufeli & Bakker, 2004). This indicates that there is a high level of work engagement in Norwegian priests in general, while also indicating that there is a lower level of work engagement among priests with anxiety and depression problems.
The third hypothesis, that work engagement would be positively associated with job resources and not with job demands, was mainly supported. Among priests with MHC, none of the job demands were associated with the UWES. However, among priests without MHC, we found a low positive correlation between decision demands and the UWES value. This may imply that decision demands are perceived as challenging (Searle & Auton, 2015), which creates a higher degree of work engagement. Assuming that these associations reflect a causal relationship, the development could also go the other way, where engaged priests create a job situation involving several decision-making demands through job crafting. Job crafting is defined as proactive changes that employees make in their job demands and job resources (Tims, Bakker, & Derks, 2012) and is associated with favorable changes in the work environment and indirectly with increased work engagement (Tims, Bakker, & Derks, 2013).
Interestingly, we did not find an association between decision demands and engagement among priests with MHC. In other words, priests with MHC either do not consider decisionmaking a challenge or their work engagement does not lead them to take on more decision demands. However, due to this study's cross-sectional design, we cannot confirm such a causal relationship over time.
In line with what we expected in the third hypothesis, job resources were associated with work engagement among priests both with and without MHC. In the correlation analyses, we found that most of the job resources, such as role clarity, control over decisions, control of work pacing, support from superiors and colleagues, and recognition and career opportunities were associated with work engagement. This is consistent with studies in other occupational populations Demerouti et al., 2001;Hakanen et al., 2005Hakanen et al., , 2008, which have found job resources to be associated with work engagement without finding a corresponding pattern for job demands.
If there is a causal connection, the question is raised of how these job resources affect work engagement among priests. As initially mentioned, priests have many and varied roles (Carroll & McMillan, 2006). Because role clarity provides information on what to do, priests with high scores on this scale will likely be more focused at work. Furthermore, having control over decisions and work pacing allows one to prioritize tasks and decide how and when to perform them. By choosing their working methods, priests may experience continuous learning and gain competence through the experiences they have in the execution of their work. Instrumental support from colleagues and managers may lead to a superior performance of the work, resulting in an enjoyment of mastering the work. Likewise, recognition can provide mastery of work if experienced as a form of positive feedback, while career opportunities can be a rewarding experience of being empowered to use one's skills and education while practicing work.
As described in self-determination theory, basic psychological needs such as autonomy and competence are prerequisites for autonomous motivation. Therefore, the fulfillment of such basic psychological needs may explain, in accordance with this theory, how job resources are translated into work engagement (Van den Broeck, Vansteenkiste, De Witte, & Lens, 2008). However, because this is not a prospective longitudinal study, we do not know whether these factors in the working environment cause work engagement. It may be that engaged priests create a working environment characterized by both clarity and control of crucial decisions through job crafting (Tims et al., 2012).
The fourth hypothesis, that job resources were expected to be more positively associated with work engagement among priests with MHC than among priests without such health challenges, was in general not supported. There may be several reasons why this hypothesis was not supported. The starting point for this hypothesis was that the presence of MHC would be perceived as an additional demand that had to be mastered in order to carry out the work as a priest and that the job resources would, therefore, be of greater importance for the work engagement of this group. However, both anxiety and depression symptoms are considered to counteract the subjective experience of work engagement, so job resources may, for that reason, have a limited impact on work engagement for this group.
Admittedly, we found that recognition was stronger associated with work engagement among priests with MHC. This finding may indicate that priests with MHC need, in particular, recognition for experiencing work involvement, possibly because they are unable to recognize the work they do because of rumination or worrying thoughts, and therefore are more dependent on the recognition of others.
The high level of work engagement among priests should be considered a definite asset for the Church. Thus, as an employer, the Church should ensure the preservation of this engagement by maintaining a working environment with adequate job resources. The factors identified in this study, may be a good starting point. Systematic work environment surveys, interviews and follow-up through ongoing processes with a focus on the work environment can be interventions that ensure continued work engagement. Given the high incidence of MHC among priests, it may be beneficial to focus specifically on the job resources needed for this group to maintain work engagement.

Limitations and strengths
The strength of this study was its reliance on a large-scale survey of Norwegian priests with a reasonable response rate and using valid and reliable measurements. However, it was based on a cross-sectional design. Therefore, the associations observed cannot be used to deduce causal relationships but only to generate well-supported hypotheses. If we assume a causal relationship, job resources may lead to increased work involvement. Engaged priests, however, could also shape their working environment to reflect more resources through job crafting. Moreover, since individual resources, such as work-related self-efficacy or personality, were not included in this study, confounding influences from such factors cannot be excluded.