G Model

NS-2434; No. of Pages 12

International Journal of Nursing Studies xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Nursing Studies

journal homepage: www.elsevier.com/ijns

The mediating role of demand and control in the relationship

between leadership behaviour and employee distress:
A cross-sectional study
Caroline Lornudd a,*, Susanne Tafvelin b, Ulrica von Thiele Schwarz a,c,
David Bergman a
a
Department of Learning, Informatics, Management, and Ethics, Karolinska Institute, Sweden
b
Department of Psychology, Umeå University, Sweden
c
Department of Psychology, Stockholm University, Sweden

A R T I C L E I N F O A B S T R A C T

Article history: Background: The relationship between leadership and employee distress is well
Received 18 February 2014 established, however, the processes involved in this relationship remain largely unclear.
Received in revised form 4 June 2014 For a stretched nursing workforce, understanding in what ways leadership may influence
Accepted 8 August 2014
employee distress is particularly important.
Objectives: To examine possible mediating effects of the work environment factors
Keywords:
demand and control in the relationship between leadership behaviour in change,
Leadership behaviour
production, and employee orientation and employee distress.
Employee distress
Demand-control Design: Cross-sectional study design.
Mediation Settings: The study was conducted at a large county council in Sweden providing both
Healthcare institutional and non-institutional care.
Participants: A random sample of 1249 employees (primarily nurses, but also a wide range
of other healthcare professionals and administrative staff), who had a healthcare manager
that was about to enter a leadership development programme (n = 171), responded to a
web-based questionnaire. The response rate was 62%.
Methods: The employees rated their healthcare managers’ behaviour in change,
production, and employee orientation, as well as their own perceptions of level of
demand, control (subdivided into decision authority and skill discretion), and five distress
outcomes. Multilevel analysis was performed.
Results: The mediators demand, decision authority, and skill discretion were significant
predictors of all five distress outcomes for all three leadership orientations. In eight of 15
regressions, the mediators fully explained the relationships between leadership
orientations and outcomes. Four of five relationships with distress outcomes were fully
mediated for change-oriented leadership, whereas two of five outcomes were fully
mediated for production- and employee-oriented leadership. In all three leadership
orientations, the relationship between the mediator skill discretion and the distress
measure disengagement were particularly strong, with B-coefficients ( .44, p < .001)
twice as high as for any of the other relationships.
Conclusions: It seems that the way that employees perceive healthcare managers’ change-
oriented behaviour, and how that aspect is related to employee distress, is primarily
explained by perception of demand and control. Furthermore, regardless of leadership

* Corresponding author. Tel.: +46 852483788.

E-mail address: Caroline.lornudd@ki.se (C. Lornudd).

http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
0020-7489/ß 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
G Model
NS-2434; No. of Pages 12

2 C. Lornudd et al. / International Journal of Nursing Studies xxx (2014) xxx–xxx

behaviour orientation, how employees perceive their opportunity to use specific job
skills plays an important role in the interplay between perception of healthcare
managers’ behaviour and disengagement.
ß 2014 Elsevier Ltd. All rights reserved.

What is already known about the topic? 2010; Sellgren et al., 2008) and moral distress (de Veer
et al., 2013).
 Leadership behaviour characterised by support and However, the processes involved in the relationship
consideration is positively associated with employee between leadership and employee well-being remain
well-being, and assessments of the relationship between largely unclear (Gregersen et al., 2011; Skakon et al.,
production-/task-oriented or change-oriented leader- 2010). It is plausible that these can be clarified by examining
ship behaviours and employee well-being have provided the various ways in which leadership can potentially
mixed results. influence well-being in employees. The impact of the leader
 The Job Demand-Control model is a well-supported may be exerted directly on employee well-being, or
theory of how the work environment affects physiologi- indirectly in the sense that the leader influences the work
cal and psychological health of employees. environment, which in turn affects the employees (Skakon
 Employee perception of control has been shown to et al., 2010; Tafvelin et al., 2011). In recent years, the indirect
mediate in the relationship between supervisor support effects of leadership have gained increasing interest through
and employee job satisfaction. investigations of mediating variables (e.g. Arnold et al.,
2007; Munir et al., 2012; Nielsen et al., 2008a). Nonetheless,
What this paper adds these studies have focused on transformational leadership,
and thus less is known about the processes that link
 The results suggest employee perceptions of demand and alternative leadership models to employee well-being.
control as mediators in the influential process between Three-dimensional leadership models including produc-
leadership behaviour and employees’ distress, in partic- tion-, employee-, and change-oriented behaviours have
ular concerning change-oriented leadership behaviour. been shown to predict various effectiveness outcomes that
 The findings highlight the interplay between leadership are well in line with more dominating leadership models,
behaviour, employee disengagement and skill discretion. such as transformational and transactional leadership
 Support is provided for a positive relationship between (Michel et al., 2011). Transformational leadership, in
well-being of healthcare employees and change- and combination with certain aspects of transactional leader-
production-oriented leadership behaviour, respectively. ship, has been positively associated with a multitude of
 The knowledge base of the processes involved in the employee well-being outcomes, and this model has
relationship between leadership and employee distress emerged to occupy a strong position in leadership research
is expanded by analysis of change-, production-, and during the last two decades. However, analyses have also
employee-oriented leadership behaviours. indicated that transformational and transactional leader-
ship, in contrast to three-dimensional leadership models,
1. Introduction omit leader behaviours of particular relevance to task
performance (Michel et al., 2011). In the present study we
The psychosocial environment of the workplace com- examined the well-validated model of change-, production-,
prises a multitude of factors that can be beneficial or and employee-oriented leadership behaviours, and the
detrimental to employee health, and one of the crucial potential mediating effect of the two work environment
factors determining work-related well-being among factors job demand and control (Karasek, 1979) on the
employees is related to the quality of leadership. In a relationship with employee well-being.
recent review of 30 years of research, Skakon et al. (2010)
concluded that leadership (e.g. leaders’ behaviours and 1.1. Change-, production-, and employee-oriented leadership
leadership style) was associated with employee well- and employee well-being
being. Leadership has been correlated with an array of
employee outcome measures, such as ischaemic heart In the 1990s, the long-lived conceptual distinction of two
disease (Nyberg et al., 2009), stress (Westerlund et al., broad categories of leadership behaviour – production/task
2010), work climate (Malloy and Penprase, 2010), and sick orientation and employee/relation orientation – was
leave and disability pension (Kuoppala et al., 2008). In the extended by adding a third category, change orientation
field of nursing management, particular interest has been (Ekvall and Arvonen, 1991; Yukl, 1999). Change orientation
centred on the association between leadership and staff encompasses behaviours such as creating visions, encour-
well-being to address the problems of a stretched nursing aging innovative thinking, and driving change. The change-,
workforce and an international shortage of nurses (Cum- production-, and employee-oriented leadership behaviours
mings et al., 2010). In this field, leadership has been shown model (the CPE model) has been validated in both the
to be related to well-being outcomes such as burnout United States and Scandinavia in a number of studies of
(Kanste et al., 2007), job satisfaction (Cummings et al., samples covering different branches and hierarchical levels

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
G Model
NS-2434; No. of Pages 12

C. Lornudd et al. / International Journal of Nursing Studies xxx (2014) xxx–xxx 3

(Ekvall and Arvonen, 1991; Yukl, 1999; Yukl et al., 2002). styles (i.e. the definition of them as task-focused leadership)
Even though the CPE model primarily has been evaluated in showed some superficial similarities to production orienta-
the private industrial sector, there is support for the tion, their operationalisation set them apart from it. For
relevance of the model in the healthcare sector. Sellgren example, dissonant leadership style was measured by the
et al. (2008) found that nurse managers with a leadership absence of emotional intelligence, and management-by-
profile characterised by high ratings in all three behaviour exception was measured according to the degree to which
orientation categories also received the highest employee the leader paid attention to mistakes. This makes the
job satisfaction rates. Furthermore, in a qualitative case interpretation of the findings in relation to production
study of leadership behaviours in four successful organisa- orientation less than straightforward. In summary, it seems
tions, including one hospital, it was observed that the that low production-oriented leadership behaviour does not
common leadership behaviours that were identified corre- benefit the health of employees at certain levels in certain
sponded to the behaviour orientations in the CPE model sectors, and in healthcare services, a high level of production
(Larsson and Vinberg, 2010). Notwithstanding the men- orientation can be related to employee distress if it is not
tioned observations, the scientific support for an association associated with employee-oriented leadership behaviour.
between the CPE model and employee well-being is limited Employee orientation refers to leadership behaviours
(Larsson and Vinberg, 2010), although information on this that demonstrate consideration for employees (i.e. show-
subject is increasing. In the few studies that have assessed ing regard, being friendly, and relying on subordinates).
the relationship between change-oriented leadership and Overall, there appears to be consistent support for a
employee well-being, change-orientation primarily has positive association between employee-oriented leader-
been shown to have a positive influence on employee ship behaviour and employee well-being. In an interna-
outcomes. Among blue-collar workers, change-oriented tional review and meta-analysis of 109 articles, Kuoppala
leadership has been found to have a weak positive et al. (2008) noted a moderate relationship between
correlation with mental fatigue (Arvonen, 1995), but on higher employee-oriented leadership styles (considerate,
the other hand a positive association with job satisfaction supportive, or transformational style) and better job well-
(Arvonen, 1995; Skogstad and Einarsen, 1999). In healthcare being, less sick leave, and lower disability pension.
services, low scores in change orientation have been Furthermore, Cummings et al. (2010) found that nurses
related to high levels of musculoskeletal pain in employees reported better health, less anxiety, and less emotional
(Fjell et al., 2007). Moreover, team performance and team exhaustion and stress when they had a relationally focused
satisfaction among healthcare professionals have been leader (transformational, considerate, supportive, or reso-
reported to be positively correlated with change-oriented nant). This is supported by a study showing that leadership
leadership (Gil et al., 2005). characterised by low employee orientation (e.g. autocratic,
Production orientation includes behaviours related to malevolent, or self-centred) is associated with poor mental
structure (i.e. bringing order, planning, and following up), health, low vitality, and behavioural stress among employ-
and the connection between this orientation and employee ees (Nyberg et al., 2011).
well-being points in different directions: a low leadership
score in production orientation has been moderately 1.2. Mediators between leadership behaviour and employee
associated with job dissatisfaction among blue-collar well-being
workers, although this relationship was not observed
among white-collar workers or managers (Arvonen, 1995). In recent years, an increasing number of studies have
Hence employees at different organisational levels may assessed mediating factors between perception of leader-
differ regarding the way they are affected by low ship and employee well-being, in particular considering
production orientation in their leaders. In addition, the transformational leadership, and several of these investi-
importance of production-oriented leadership may vary gations have been conducted in the healthcare sector. The
between sectors. Low scores in production orientation relationship between transformational leadership and
have been found to be associated with musculoskeletal healthcare workers’ well-being has been found to be
pain in domestic catering services but not in home and mediated by factors such as employee perception of work-
healthcare services (Fjell et al., 2007). Duxbury et al. life conflicts (Munir et al., 2012), meaningful work (Arnold
(1984) examined the relationship between head nurses’ et al., 2007), positive climate for innovation (Tafvelin et al.,
leadership styles and nurse staff burnout and job satisfac- 2011), role clarity, meaningfulness and opportunity to
tion, and the results of that study showed that a develop (Nielsen et al., 2008a,b), employee involvement
production-oriented leadership style (‘‘initiating struc- and influence (Nielsen et al., 2008b), and team- and self-
ture’’) alone was not related to neither burnout nor job efficacy (Nielsen et al., 2009). Mediating factors in other
satisfaction. Nevertheless, a combination of high produc- leadership styles or leadership behaviour orientations
tion orientation and low employee orientation was have not been studied to the same extent, and, to our
associated with lower levels of job satisfaction and more knowledge, the CPE model remains to be investigated.
extensive staff burnout. In contrast, Cummings et al.
(2010) reviewed the literature on leadership styles and 1.3. Job demand and control as mediators
nursing workforce outcomes and found that task-focused
leadership styles (e.g. dissonant, management-by-excep- The Job Demand-Control (JDC) model (Karasek, 1979;
tion and laissez-faire) were related to lower nurse Karasek and Theorell, 1990) is one of the most influential
job satisfaction. However, although these leadership theories about the effects of the work environment on

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
G Model
NS-2434; No. of Pages 12

4 C. Lornudd et al. / International Journal of Nursing Studies xxx (2014) xxx–xxx

employee well-being. This model includes the psychoso- is production-oriented (i.e. plans carefully, gives explicit
cial factors demand, referring to the psychological instructions, and follows work closely) can offer a
demands perceived by an employee (primarily in terms beneficial level of structure in an ever-changing setting.
of workload), and control (often called decision latitude This might in turn enhance employees’ sense of control and
and sub-divided into decision authority and skill discre- provide clearer demands, and thus be associated with less
tion), representing the subjective amount of operating distress. However, it is also plausible to assume that a
freedom an employee has to handle the mentioned healthcare manager showing a very pronounced produc-
demands, and the perceived opportunity to use specific tion-orientation can be associated with employees that
job skills at work. In the late 1980s, the original JDC model experience less control (e.g. a smaller subjective amount of
was expanded to include social support, JDC(S) (Johnson operating freedom), and hence be related to more distress.
and Hall, 1988; Johnson et al., 1989) as a work-coping Considering the contradictory nature of the relationship
resource dimension that can reduce the effects of job between production-oriented leadership behaviour and
strain. The combination high demand and low control in employee outcomes, we refrain from proposing any
the JDC model is postulated to lead to a high-strain work direction of this particular association:
environment, which is called the strain hypothesis. This
hypothesis has been consistently supported in relation to Hypothesis 1. Production-oriented leadership behaviour
psychological well-being measures in research reviews is related to employee distress, and this relationship is
covering almost four decades (Hausser et al., 2010; Van der mediated by the perceived level of demand and control.
Doef and Maes, 1999). Among nurses, support for the strain Leadership influence is less explicit in the JDC model
hypothesis has been gained in relation to a variety of than in the JDC(S) model. In the latter, social support
psychological well-being outcomes, such as mental health comprises instrumental support (e.g. assistance with tasks
(Amick et al., 1998), psychological distress (Bourbonnais or information) or socio-emotional support from a
et al., 1998), sleep quality (Winwood and Lushington, manager or co-workers. A meta-analytic review has shown
2006), and job dissatisfaction and emotional exhaustion a negative relationship between managerial support and
(de Jonge et al., 2000). demands (Luchman and Gonzalez-Morales, 2013), and
In the literature, there seems to be little consistency among nurses a higher degree of managerial support has
regarding how well-being is operationalised, hence been found to be related to decreased job stress (Hall,
measures can have either positive or negative directions. 2007). It seems logical to assume that there is a strong
In relation to the JDC model, this can be exemplified by positive correlation between healthcare managers that
well-being, which is sometimes defined as job satisfaction provide socio-emotional support and those that are
and at other times as (lack of) job dissatisfaction (Hausser perceived as employee-oriented. We suggest that health-
et al., 2010). There is a strong correlation between the care managers showing pronounced employee orientation
presence of distress and the absence of well-being (Abbott might be regarded by the employees as a resource for
et al., 2006), and it has been suggested that psychological dealing with challenging demands (which could also affect
well-being and distress can be measured as a single the sense of control). In line with that, we hypothesise that
construct representing the opposite sides of a continuum employee-oriented healthcare managers are associated
(Uher and Goodman, 2010). However, in relation to well- with employee distress by indirectly influencing employ-
being, the central premise of the strain hypothesis focuses ees’ perception of demands. Based on the previous studies
on the dimensions of demand and control and how they reporting a positive relationship between employee-
are associated with negative employee outcomes oriented leadership and employee well-being outcome
(‘‘strain’’) (Van der Doef and Maes, 1999). We therefore measures, we propose that there is a negative association
decided to measure well-being in terms of psychological with employee distress:
distress, using the outcome measures exhaustion,
disengagement, depression, sleep disturbances, and self- Hypothesis 2. Employee-oriented leadership behaviour is
rated ill health. negatively related to employee distress, and this relation-
Job control has been shown to mediate in the ship is mediated by the perceived level of demand and
relationship between supervisor support and job satisfac- control.
tion (Moyle, 1998). Additional dimensions of leadership
behaviour remain to be investigated, thus in the present To our knowledge, the relationship between leaders’
study we examined the mediating effects of the demand change orientation and employee perception of demand
and control dimensions of the JDC model on the relation- and control has not yet been examined. Previous
ship between leadership behaviour (change, production, research has suggested that organisational change is
and employee orientation) and employee distress in a associated with an increased risk of health problems
healthcare organisation. (Dahl, 2011) and changes (positive or negative) in
During the last decades, healthcare has been the object demand-control perceptions among employees (de
of several reforms and changes in many Western countries Lange et al., 2002). However, considering that change-
(Bejerot and Hasselbladh, 2003; Olson et al., 1998). It has oriented leadership behaviour has primarily been found
been proposed that ongoing organisational changes are to be positively associated with employee well-being, we
associated with a decreased level of employee perceptions propose that there may be a negative relationship
of control (Fugate et al., 2002). Applying the JDC between change-oriented leadership and employee
dimensions, it is possible that a healthcare manager that distress:

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
G Model
NS-2434; No. of Pages 12

C. Lornudd et al. / International Journal of Nursing Studies xxx (2014) xxx–xxx 5

Hypothesis 3. Change-oriented leadership behaviour is list order of all subordinates for each manager, the first 15
negatively related to employee distress, and this relation- subordinates on the list were selected.
ship is mediated by the perceived level of demand and The subordinates were invited via e-mail to respond to
control. a web-based survey including the Change, Production, and
Employee (CPE) leadership behaviour instrument and the
2. Methods Webb-QPS questionnaire. The response rate was 64%
(n = 1276). However, the number of respondents with
The study ‘‘Effects of leadership development in complete data for this study was somewhat lower
healthcare’’ was conducted within the Healthcare Provi- (n = 1249; 62%), as well as the number of rated healthcare
sion Stockholm County (SLSO) healthcare system. SLSO managers (n = 171). The 1249 respondents had a mean age
runs nine organisations that provide institutional and non- of 47.6 (range 20–70) years, and 87% were women. The
institutional care, and serve a catchment area with a majority of the respondents were nurses (36%), phy-
population of 1.6 million. The study was approved by the siotherapists/occupational therapists (16%), and adminis-
Regional Ethical Review Board. The present study uses trative personnel (14%); the rest indicated having a health-
baseline data from the study ‘‘Effects of leadership related (22%) or managerial (3%) position, or classified
development in healthcare’’, which is partly the same themselves as ‘‘others’’ (7%). Two percent did not respond
data used in another study examining the relationship to the item concerning professional position.
between healthcare managers’ self-rated personality, and
external ratings of leadership behaviour (Bergman et al., 2.2. Measures
2014).
2.2.1. Leadership orientation
2.1. Procedure and participants We measured leadership orientation in three different
subscales considering orientation towards change, pro-
In September 2010, January 2011, and September 2011, duction, and employee from the CPE questionnaire. The
the SLSO invited all their healthcare managers (n = 589) to questionnaire is based on the CPE model of leadership
participate in a two-year leadership development pro- behaviour, which was developed using data on 4000
gramme. Those who agreed to take part (n = 204) Swedish managers and 60 000 of their subordinates
represented 35% of all healthcare managers in the (Arvonen, 2002). Leadership behaviour was rated by the
organisation; 87% of them were women, and the mean subordinates on a 6-point scale ranging from ‘‘Does not
age was 51 years (range 33–64 years). The participating agree at all’’ (1) to ‘‘Agrees fully’’ (6). A confirmatory factor
healthcare managers worked in the following areas: analysis (CFA) of the full scale (24 items) showed a poor fit
primary care (46.2%), adult psychiatry (26.0%), dependency to the data, and another CFA performed to assess
disorder care (5.2%), child and adolescent psychiatry distinctiveness of the subscales in the short (15-item)
(4.6%), rehabilitation and functional assistance (3.5%), version of the CPE instrument (Sverke et al., 1999) also
geriatrics (6.9%), central administration (4.6%), and others indicated relatively poor fit to the data. However, after we
(2.9%). The SLSO healthcare managers that did not removed two items (‘‘Is very exacting about plan being
participate in the programme (n = 385) had a mean age followed’’ in the production subscale and ‘‘Allows his/her
of 55 years, and 76% were women. subordinates to decide’’ in the employee orientation
The healthcare managers and their subordinates subscale) from the short version, the factor structure
received written information about the study before the showed an acceptable fit: x2 (62) = 887.29, p < .000,
onset of the programme. Participation in the study was CFI = .94, TLI = .92, RMSEA = .10, 95% CI .09–.11. The
voluntary and informed consent was obtained. Confidenti- reliability of the subscales and included items were as
ality and anonymity was guaranteed. Immediately before follows: change .93 (‘‘Shares thoughts and plans about the
enrolment in the leadership programme, data were future’’, ‘‘Consistently pushes for development’’, ‘‘Initiates
collected from a random sample of up to 15 of each new projects’’, ‘‘Experiments with new ways of doing
healthcare manager’s subordinates (mean number of things’’, and ‘‘Offers ideas about new and different ways of
subordinates per group 7.3, range 1–15). The reliability doing things’’); production .87 (‘‘Makes a point of following
of a 360-degree instrument is positively correlated to the rules and principles’’, ‘‘Follows and controls work closely’’,
number of raters included (Hensel et al., 2010). Hensel et ‘‘Plans carefully’’, ‘‘Gives clear instructions’’); employee .92
al. (2010) concluded that ten raters corresponded to a (‘‘Relies on his/her subordinates’’, ‘‘Is considerate’’, ‘‘Is just
satisfactory level of reliability of 0.7 (Nunnally, 1978). In in treating subordinates’’, and ‘‘Shows respect for other
our naturalistic study, each healthcare manager had a people’’).
different number of subordinates due to the organisation
of the healthcare setting. An advantage of our investigation 2.2.2. Distress
is that we chose up to 15 subordinates in order to reach as To measure distress we used the Webb-QPS instrument
high mean number of subordinate raters as possible. For (Hasson et al., 2008), a short web-based version of QPS
each healthcare manager, the selection of subordinates Nordic (Dallner et al., 2000) which is a well-validated
were randomised from a list of all subordinates. The instrument for assessing health and psychosocial aspects
randomisations were made by use of Microsoft Excel 2010 in employees in Sweden. In addition to demographic data,
based on the alphabetic order of the last names of the we assessed distress in the subscales disengagement (five
subordinates. After a computerised randomisation of the items), exhaustion (five items), depression (six items), and

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
G Model
NS-2434; No. of Pages 12

6 C. Lornudd et al. / International Journal of Nursing Studies xxx (2014) xxx–xxx

sleep disturbances (five items), and these dimensions were outcome variable, which was used as a baseline against
measured on a 4-point Likert scale ranging from ‘‘Never/ which the contribution of predictors could be compared,
Not at all’’ (1) to ‘‘Always/Agrees fully’’ (4). Also, self-rated and then we assessed an unconditional model for each
health was measured by a single question on a 5-point outcome by allowing the intercept term to vary by leader
scale ranging from ‘‘Good’’ (1) to ‘‘Bad’’ (5), and this item (i.e. the level 2 variable). Thereafter, control variables (age
was reversed in the analysis to form the outcome variable and tenure) and predictor (leadership orientation) were
self-rated ill health. The reliability of the Webb-QPS entered into the equation (step 1). Finally, the proposed
subscales were .76 (disengagement), .78 (exhaustion), mediators were added (step 2). Following the procedure
.87 (depression), .74 (sleep disturbances). for multilevel modelling outlined by Heck et al. (2010),
after allowing intercept to vary by leader, random slope
2.2.3. Mediators models were investigated.
The subscales demand (four items measuring percep-
tions of workload and demanding decisions), decision 3. Results
authority (seven items measuring perceptions of role
clarity and influence of work content), and skill discretion Means, standard deviations, and correlations for all study
(three items measuring opportunity to use skills at work variables are presented in Table 1. We found significant
and meaningfulness of work) constituted our mediators. correlations between the three leadership orientations and
Decision authority and skill discretion constitute the two all distress variables and the proposed mediators, and this
aspects of control in the JDC model. The subscales were merited further investigation of the three hypotheses
included in the Webb-QPS (Hasson et al., 2008) and they outlined in the introduction. Intra-class correlation (ICC)
were rated on a 5-point Likert scale ranging from ‘‘Very was computed based on the unconditional model. The
seldom’’ (1) to ‘‘Very often’’ (5). The reliability of the second level of analyses (group level) explained 6–8% of the
subscales were .74 (demand), .75 (decision authority), and variance for all but one distress variable, self-rated ill health,
.76 (skill discretion). for which the rate was only 3.5%. Inasmuch as ICC indicated
that data were nested for four out of five variables,
2.2.4. Statistical analyses multilevel analysis was deemed appropriate.
The missing data analysis uncovered a low percentage Tables 2–4 show the results of the multilevel analysis.
of missing data in the predictor, mediator, and outcome After controlling for the group effect and demographic
variables (< 6%). The Expectation-Maximisation (EM) predictors (age and tenure), each of the predictors
algorithm (Little and Rubin, 1987) was used when (leadership orientations) showed a significant negative
replacing missing values. To test the relationships between relationship with the distress outcomes. Thus our analyses
leadership and distress outcomes, we performed a supported an initial effect of the investigated leadership
multilevel analysis using SPSS 20, and, prior to this, all orientations on the distress outcomes of interest. After
predictors were grand mean centred (Hofmann and Gavin, these initial analyses, we followed the approach to
1998; McKee et al., 2011). The three leadership scales were mediation described by Baron and Kenny (1986) by
used as individual level predictors, as were the mediators assessing whether our three leadership orientations
and the outcomes. No variables were aggregated or used as (change, production, and employee) were associated with
group-level predictors. However, multilevel analysis was each of the three mediators (job demand, decision
applied to take into account and control for the nested data authority, and skill discretion). A mixed model analysis
structure. We first constructed a null model for each controlling for age and tenure demonstrated that all

Table 1
Correlations and mean values.

1 2 3 4 5 6 7 8 9 10 11 12 13

1. Age
2 Tenure .453**
3. Employee .050 .054
4. Change .050 .034 .826**
5. Production .014 .008 .749** .785**
6. Disengagement .059* .036 .310** .252** .249**
7. Exhaustion .040 .084** .254** .214** .199** .470**
8. Depression .081** .049 .238** .182** .211** .443** .604**
9. Sleep disturbance .093** .062* .205** .178** .205** .321** .613** .542**
10. Self-rated ill health .055 .060* .166** .120** .153** .289** .522** .545** .458**
11. Job demands .020 .010 .130** .107** .092** .056* .335** .207** .234** .136**
12. Decision authority .035 .055 .436** .373** .393** .395** .413** .386** .359** .250** .229**
13. Skill discretion .045 .013 .208** .181** .191** .585** .236** .253** .185** .191** .165** .377**

Mean 47.6 7.7 4.44 4.37 4.32 1.87 2.27 1.52 1.96 1.64 3.49 3.68 4.36
Standard deviation 9.6 6.0 1.03 0.99 0.96 0.51 0.59 0.50 0.48 0.81 0.72 0.60 0.60

Note. Listwise N for correlations = 1249.

* p < .05.
** p < .01.

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
G Model
NS-2434; No. of Pages 12

C. Lornudd et al. / International Journal of Nursing Studies xxx (2014) xxx–xxx 7

Table 2
Effects of employee-oriented leadership behaviour on employee distress.

Model Disengagement Exhaustion Depression Sleep disturbance Self-rated ill health

a
Null ( 2LL ) 1865.06 2236.48 1797.80 1710.18 3022.37
Unconditional ( 2LLa) 1842.87 2216.51 1783.59 1696.59 3017.94
Step 1 ( 2LLa) 1762.52 2159.47 1730.32 1665.43 3005.20
Step 2 ( 2LLa) 1260.63 1879.53 1559.54 1500.76 2932.81

Model coefficients step 1 (Unstandardised Beta)

Age .01*** .00 .00* .00* .00
Tenure .01* .01* .00 .00 .00
Leadership: employee .13*** .14*** .12*** .09*** .13***

Model coefficients step 2

Age .003* .00 .00 .01*** .00
Tenure .005* .01* .00 .00 .01
Leadership: employee .05*** .03* .04* .01 .04
Job demands .07*** .24*** .12*** .13*** .14***
Decision authority .11*** .25*** .20*** .21*** .20***
Skill discretion .44*** .18*** .14*** .09*** .21***

Note. N = 1249, no. of groups = 171.

a
2 Log Likelihood.
* p < .05 (2-tailed).
*** p < .001 (2-tailed).

Table 3
Effects of change-oriented leadership behaviour on employee distress.

Model Disengagement Exhaustion Depression^ Sleep disturbance Self-rated ill health^

a
Null (-2LL ) 1865.06 2236.48 1797.80 1710.18 3022.37
Unconditional ( 2LLa) 1842.87 2216.51 1783.59 1696.59 3017.94
Step 1 ( 2LLa) 1784.08 2183.11 1747.13 1678.92 3001.38
Step 2 ( 2LLa) 1270.89 1881.24 1548.39 1501.04 2917.20

Model coefficients step 1 (Unstandardised Beta)

Age .01*** .00 .00* .00* .00
Tenure .01* .01* .00 .00 .00
Leadership: change .13*** .12*** .09*** .08*** .10**

Model coefficients step 2

Age .00* .00 .00 .01*** .00
Tenure .01* .01* .00 .00 .00
Leadership: change .05*** .02 .02 .01 .01
Job demands .07*** .24*** .11*** .13*** .14***
Decision authority .12*** .26*** .22*** .22*** .21***
Skill discretion .45*** .18*** .15*** .10*** .21***

Note. N = 1249, no. of groups = 171.

a
2 Log Likelihood.
* p < .05 (2-tailed).
*** p < .001 (2-tailed).
^
Random slope model.

leadership orientations were significantly related to all outcomes and all three leadership orientations. Adding
potential mediators. Also, all three leadership orientations the mediators reduced the unstandardised regression
were negatively related to distress outcome measures, and coefficient (B) for all leadership orientations. The most
the mediators were significant predictors of all distress substantial reduction of B was found in employee-oriented
outcomes for all leadership orientations. leadership, in relation to the outcome variable exhaustion
For all but three outcomes, we found that allowing the (reduction from B .14, p < .001 to B .03, p < .05). In eight
slope to vary did not significantly improve the model, and out of 15 regressions, the mediators fully explained the
only fixed effects were estimated for these outcomes. relationships between leadership orientations and out-
Random slope models were estimated in the analysis comes. For change-oriented leadership, four out of five
relating change-oriented leadership to depression, and in relationships were fully mediated by job demand, decision
the analyses relating change-oriented leadership and authority, and skill discretion, whereas two out of five
production-oriented leadership to self-rated ill health. outcomes were fully mediated for production- and
In the final step, the mediators were added to the employee-oriented leadership. Considering the distress
models (see Tables 2–4), which substantially reduced the outcomes, the relationships between self-rated ill health
model fit measure 2 Log Likelihood ( 2LL) for all and all leadership orientations were fully mediated by job

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
G Model
NS-2434; No. of Pages 12

8 C. Lornudd et al. / International Journal of Nursing Studies xxx (2014) xxx–xxx

Table 4
Effects of production-oriented leadership behaviour on employee distress.

Model Disengagement Exhaustion Depression Sleep disturbance Self-rated ill health^

a
Null ( 2LL ) 1865.06 2236.48 1797.80 1710.18 3022.37
Unconditional ( 2LLa) 1842.87 2216.51 1783.59 1696.59 3017.94
Step 1 ( 2LLa) 1789.04 2187.00 1749.74 1662.22 3008.04
Step 2 ( 2LLa) 1274.68 1882.52 1561.90 1497.19 2911.72

Model coefficients step 1 (Unstandardised Beta)

Age .00** .00 .00* .00** .00
Tenure .01* .01* .00 .00 .00
Leadership: Production .13*** .12*** .11*** .10*** .13***

Model coefficients step 2

Age .00* .00 .00 .01*** .00*
Tenure .01* .01** .00 .00 .00
Leadership: production .04** .02*** .03 .03* .05
Job demands .07*** .24*** .12*** .13*** .13***
Decision authority .12*** .27*** .21*** .20*** .20***
Skill discretion .45*** .18*** .15*** .09*** .20***

Note. N = 1249, no. of groups = 171.

a
2 Log Likelihood.
* p < .05 (2-tailed).
*** p < .001 (2-tailed).
^
Random slope model.

demand, decision authority, and skill discretion. For change-oriented leadership showed a weak positive
disengagement, on the other hand, the mediation was relationship with mental fatigue but was not correlated
only partial in all regressions. Both full and partial with symptoms of depression, anxiety or insomnia.
mediation were found for exhaustion, depression, and However, the sample we studied and that analysed by
sleep disturbances. Arvonen (1995) represented disparate organisations
(healthcare service and process industry) that might have
4. Discussion reacted differently to change-oriented leadership, and it is
also possible there are dissimilarities in how the practice of
We assessed the mediating effect of the dimensions of change-oriented leadership affects employees. A health-
the well-validated JDC model on the relationship between care service might include more tasks that a change-
leadership behaviour (change, production, and employee oriented healthcare manager can allocate among employ-
orientation) and employee distress, and the results ees in a flexible way (e.g. administrative tasks) and thereby
provided support for our three hypotheses. In short, for lower the level of distress.
each leadership behaviour orientation, job demand and A second conclusion drawn from our results is that the
control partly or fully mediated the relationship with relationship between the mediator skill discretion and the
employee distress. Also, all leadership behaviour orienta- outcome measure disengagement can be regarded as
tions were negatively related to employee distress. Hence, noteworthy, considering that the B-coefficient for this
overall, employee perception of demand and control in the relationship, irrespective of the leadership orientation
work environment was found to account for a substantial being analysed, was twice as high as noted for any other
part of the relationship between perceptions of leadership associations between mediators and distress outcomes.
and employee distress. Skill discretion refers to the subjective meaningfulness of
Three main conclusions can be drawn from our the work content and to what extent the employees can
findings. First, considering change, production, and em- make use of their competence and experience positive
ployee behaviour orientations in healthcare managers, it challenges. Meaningfulness has previously been shown to
was change orientation that was most consistently found mediate between transformational leadership and well-
to be fully mediated by demand and control in relation to being of healthcare workers (Arnold et al., 2007; Nielsen
the outcomes of employee distress (with the exception of et al., 2008a,b), and skill under-utilisation has been
disengagement). Thus it seems that perception of change associated with human service burnout (Leiter, 1990).
orientation in a healthcare manager, and how this is According to the job demands-resources (JD-R) model
related to employee distress, is primarily explained by (Demerouti et al., 2001), disengagement and exhaustion are
employee perception of demand and control. Furthermore, the core components of the burnout syndrome. Our results
the negative relationship between change orientation and indicate that, regardless of leadership behaviour orienta-
employee distress supports the findings of previous tion, the way that employees perceive their opportunity to
studies showing higher scores in change-oriented leader- use specific job skills at work plays an important role in
ship behaviour in relation to positive employee outcomes the interplay between how they perceive healthcare
(Fjell et al., 2007; Gil et al., 2005; Skogstad and Einarsen, managers’ behaviour and their own disengagement. In
1999). Our findings contradict the results reported by addition, it is of interest to note that neither disengagement
Arvonen (1995) indicating that, in blue-collar workers, nor exhaustion is fully mediated, which suggests

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
G Model
NS-2434; No. of Pages 12

C. Lornudd et al. / International Journal of Nursing Studies xxx (2014) xxx–xxx 9

that healthcare managers’ behaviours are also directly captures leadership behaviour that is important for
associated with this particular employee distress outcome. enhancing a sense of structure and predictability among
Compared to depression, sleep disturbances, or self-rated ill healthcare personnel. We recommend that the significance
health, the burnout-related constructs of disengagement of production-oriented leadership in relation to employee
and exhaustion exhibit a specific connection with working distress in healthcare services be further investigated.
life, which implies that there is a direct association between Our results also support a negative association between
leadership behaviour and disengagement or exhaustion. change-oriented leadership behaviour in healthcare man-
However, an alternative explanation for such a direct agers and employee distress. This may seem somewhat
relationship is that an effect is exerted by other mediating surprising, considering that a relationship has been
factors that were not examined in the present study. demonstrated between organisational change and in-
The third conclusion emerging from our study is that, in creased risk of employee health problems (Dahl, 2011).
contrast to disengagement and exhaustion, self-rated ill However, Syrek et al. (2013) found that transformational
health appears to be fully mediated by perceptions of leadership – a leadership style highly characterised by
demand and control. This observation is in contrast to the behaviours that encourage change in employees –
findings of Westerlund et al. (2010), which demonstrated moderates the positive association between time pressure
an association between leadership and age-relative self- and employee exhaustion, as indicated by the observation
rated health even after controlling for JDC(S). The that this relationship was weaker when there was a higher
contradiction between the two studies might be explained level of transformational leadership. Accordingly, a highly
by differences in the assessments of leadership. Wester- change-oriented healthcare manager might be more
lund et al. used a measure of attentive managerial compatible with the continuously changing conditions
leadership (AML), which covered a broader spectrum of that prevail in healthcare services today, for example, by
psychosocial work environment, for example, the atmo- being able to reframe stressful organisational changes into
sphere in the working unit and general perceptions of fair something that is meaningful and manageable for employ-
division of labour and result appraisals. Since these factors ees. In this sense, a healthcare manager’s change orienta-
themselves may be related to work stress symptoms tion might moderate the effects of organisational change
(Elovainio et al., 2006), the unique contribution of the on employee distress.
leadership becomes less clear. Also, by using the dimen- The second contribution of our study is that the
sions of the JDC model, rather than the JDC(S) model – dimensions of the JDC model are identified as mediators
which includes superior support – we were able to in the influential process between leadership and employ-
concentrate measurement of leadership on one variable, ee distress. How should this relationship be understood?
which we believe facilitated the interpretation of our One possible interpretation is that a leader indirectly
results. Our findings show that leadership behaviour in influences the level of distress by directly affecting the
change, production, and employee orientation did not have work environment – the influence of demand and control
a direct association with self-rated ill health. of the work environment on employee distress are well-
validated. Yet it could also be argued that in the present
4.1. Theoretical implications study, the level of perceived demand and control might be
affected by some unmeasured situational variable, such as
The scientific contributions of our findings are twofold. objective workload. According to the JDC model, employ-
First, we provide support for a relationship between the ees in medical units with a higher workload are more likely
CPE model and employee distress, which thus far has been to rate themselves as more distressed. In addition, there is
largely uncertain. The negative association between research suggesting that employees’ feelings and behav-
healthcare managers’ production orientation and employ- iour have an impact on how they are treated by their
ee distress is intriguing in this context. Production leaders (Horn et al., 2004). Hence another way of
orientation deviates from the well-studied transforma- interpreting our results is to assume that the level of
tional leadership style, and it also comprises behaviours distress experienced by employees can influence the
that are usually not included in transactional leadership, behaviours of these individuals, which in turn can affect
such as short-term planning and clarifying roles (Michel how they are treated by their leader. To our knowledge,
et al., 2011). Our results modify previous observations in demand and control have not previously been investigated
the review by Cummings et al. (2010), which suggest that as mediators between leadership and employee distress,
leadership styles focused on task are related to lower job neither in terms of the CPE model nor in relation to
satisfaction in the nursing workforce. According to this transformational or transactional leadership. Despite
review, the styles focused on task included, among other several potential interpretations of our findings, the
styles, being emotionally out of touch with employees present study gives a more specific understanding of
(dissonant leadership) or concentrating attention on the ways that leadership behaviour might be related to
failures either before or after they have occurred (man- employee well-being and distress. Concerning change-
agement-by-exception), and both of these leadership oriented leadership, which was found to be fully mediated,
styles were associated with emotional exhaustion. Not- it seems that perceptions of demand and control can better
withstanding, the opposite is suggested by the present explain variation in how employee distress is related
results concerning leadership styles focused on task to perceptions of this leadership orientation than to
measured as production orientation. It is possible that production- or employee-oriented leadership behaviour.
production orientation measured as in the CPE model It is plausible to propose that, compared to production- or

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
G Model
NS-2434; No. of Pages 12

10 C. Lornudd et al. / International Journal of Nursing Studies xxx (2014) xxx–xxx

employee-oriented leadership, change-oriented leader- Another common rater effect that is particularly relevant
ship is associated with a higher degree of objective when assessing leadership is the tendency for subordinates
changes in a medical unit. Similar to research results to give higher ratings to a manager that is well known or
indicating an association between actual changes in an liked. For example, the relationship between employee
organisation and changes in demand-control perceptions satisfaction and employee-oriented leadership has been
(de Lange et al., 2002), it is possible that change-oriented shown to be influenced by this type of measurement error
leadership can be extensively involved in the demand- (Schriesheim et al., 1979), although such an effect was not
control aspects of the work environment. found for the relationship between the same outcome
measure (employee satisfaction) and ‘‘initiating structure’’
4.2. Practical implications leadership behaviour (similar to production orientation) in
the cited investigation. Nevertheless, we cannot exclude the
This study has several implications for practitioners. possibility of this type of common rater effect on the
Healthcare managers would benefit from being aware of the relationships found in the present study. Also, the health-
association between how their leadership behaviour is care managers that were rated by their subordinates had
perceived and how demand and control in the work accepted an invitation to participate in a leadership
environment are perceived. This suggests that a continuous programme, which means that they may have represented
dialogue about perceived level of demand and control can be a biased sample by being more change-oriented and
an essential tool to help these healthcare managers make therefore more interested in development than the health-
individual adjustments to prevent distress among subordi- care managers who declined to attend the programme. It is
nates. Our findings imply that this might be particularly also possible that the healthcare managers who did not take
important for healthcare managers who are rated high in part in the programme were more production oriented, and
change orientation. The concrete leadership behaviours therefore down-prioritised the programme. Further re-
applied in the present study might prove useful in search in this area could examine other theoretically
workplace interventions. Also, considering that the way motivated mediating variables in order to unravel the
that employees perceive their opportunity to use specific job processes between leadership behaviour and employee
skills at work seems to play an essential role in the interplay distress. It should be emphasised that the term ‘‘full
between leadership behaviour in general and the distress mediation’’ does not imply that the relationship between
outcome measure disengagement, and also considering that that particular leadership behaviour and employee
skill under-utilisation has been associated with employee distress has been thoroughly elucidated. Indeed, as Rucker
burnout (Leiter, 1990), it might be beneficial to pay special et al. (2011) noted, use of this term may have an inhibiting
attention to this particular aspect of demand and control in effect on theory development, because additional med-
analyses and development of work environments. More- iators may not be investigated.
over, based on our results, employees in healthcare services In conclusion, we examined the mediating effects of the
may experience a lower level of distress if they have a dimensions demand and control of the JDC model on the
manager who maintains a certain level of structure by being relationship between healthcare employees’ distress and
production-oriented, for example, by giving clear instruc- the way these individuals perceive change-, production-,
tions, following work closely, and planning carefully. and employee-oriented leadership behaviour in their
healthcare managers. The results showed that all leader-
4.3. Methodological considerations and future directions ship orientations were negatively related to employee
distress, and all of the detected relationships were found to
When interpreting the results, a number of limitations be partly or fully mediated by demand-control. Therefore,
must be taken into account. Obviously, the cross-sectional we suggest that the dimensions of the JDC model are
design of the study restricts the possibility of explaining mediators in the influential process between leadership
the causal direction of any of the discerned relationships. and employee distress. Change orientation proved to be
However, based on a review by de Lange et al. (2003), it is the leadership behaviour that was most often fully
plausible to assume a normal causal effect of the JDC model mediated by demand-control in relation to employee
on psychological distress. There is also a risk of common distress. Thus it seems that employees’ perception of
rater effects, because the assessment used employee change orientation in a healthcare manager, and how that
ratings rather than objective measures; such effects can aspect is related to employee distress, is primarily
lead to measurement error and thereby deflate or inflate explained by the way that employees perceive demand
the relationships that are detected. A potential rater effect and control. Furthermore, it appears that, regardless of
is negative affectivity (NA), which was not controlled for in leadership behaviour orientation, how employees perceive
this study. Some researchers have suggested that NA can the opportunity to use specific job skills play an essential
influence self-ratings and therefore also have an impact on role in the interplay between employee perceptions of
the strength of a relationship between variables (Brief leadership behaviour and disengagement.
et al., 1988). However, others have argued against
controlling for this aspect (Spector et al., 2000), primarily Acknowledgements
pointing out that if a measure of a job stressor is considered
to reflect employee perceptions, rather than an objective We wish to express deepest gratitude for the financial
work stressor, NA may be interlinked with the construct of support from AFA Försäkring, grant no. 100073, which
interest and hence should not be controlled for. made this study and analyses possible. We are grateful for

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
G Model
NS-2434; No. of Pages 12

C. Lornudd et al. / International Journal of Nursing Studies xxx (2014) xxx–xxx 11

invaluable advice from Professor Christer Sandahl, Kar- de Veer, A.J.E., Francke, A.L., Struijs, A., Willems, D.L., 2013. Determinants
of moral distress in daily nursing practice: a cross sectional correla-
olinska Institute, and for support from Mikael Ohrling, CEO tional questionnaire survey. Int. J. Nurs. Stud. 50 (1), 100–108.
at Stockholm County Council (SCC). We would like to thank Demerouti, E., Bakker, A.B., Nachreiner, F., Schaufeli, W.B., 2001. The job
the participating respondents at SCC for taking their time demands-resources model of burnout. J. Appl. Psychol. 86 (3),
499–512.
to answer the questionnaires, and Therese Wahlström, Duxbury, M.L., Armstrong, G.D., Drew, D.J., Henly, S.J., 1984. Head nurse
Karolinska Institute, for help with data collection. leadership-style with staff nurse burnout and job-satisfaction in
Conflict of interest: Dr Bergman was partly financed by neonatal intensive-care units. Nurs. Res. 33 (2), 97–101.
Ekvall, Arvonen, J., 1991. Change-centered leadership: an extension of the
the organization where the study ‘‘Effects of leadership two-dimensional model. Scand. J. Manage. 7 (1), 17–26.
development in healthcare’’ took place, for coordination of Elovainio, M., Kivimaki, M., Puttonen, S., Lindholm, H., Pohjonen, T.,
data collection. Sinervo, T., 2006. Organisational injustice and impaired cardiovascu-
lar regulation among female employees. Occup. Environ. Med. 63 (2),
Funding: Funding was provided by AFA Försäkring,
141–144.
grant no 100073. The funder had no involvement in study Fjell, Y., Osterberg, M., Alexanderson, K., Karlqvist, L., Bildt, C., 2007.
design, data collection and analysis, decision to publish, or Appraised leadership styles, psychosocial work factors, and muscu-
preparation of the manuscript. loskeletal pain among public employees. Int. Arch. Occup. Environ.
Health 81 (1), 19–30.
Ethical approval: Regional Ethical Review Board, reg. no. Fugate, M., Kinicki, A.J., Scheck, C.L., 2002. Coping with an organizational
2010/979-31/5. merger over four stages. Personnel Psychol. 55 (4), 905–928.
Gil, F., Rico, R., Alcover, C.M., Barrasa, A., 2005. Change-oriented leader-
ship, satisfaction and performance in work groups: effects of
References team climate and group potency. J. Managerial Psychol. 20 (3/4),
312–328.
Abbott, R.A., Ploubidis, G.B., Huppert, F.A., Kuh, D., Wandsworth, M.E.J., Gregersen, S., Kuhnert, S., Zimber, A., Nienhaus, A., 2011. Leadership
Croudace, T.J., 2006. Psychometric evaluation and predictive validity behaviour and health - current research state. Gesundheitswesen
of Ryff’s psychological well-being items in a UK birth cohort sample of 73 (1), 3–12.
women. Health Qual. Life Outcomes 4 (1), 76. Hall, D.S., 2007. The relationship between supervisor support and registered
Amick III, B.C., Kawachi, I., Coakley, E.H., Lerner, D., Levine, S., Colditz, G.A., nurse outcomes in nursing care units. Nurs. Adm. Q. 31 (1), 68–80.
1998. Relationship of job strain and iso-strain to health status in a Hasson, D., Gustavsson, P., Sandahl, C., 2008. Webb-QPS – Utveckling och
cohort of women in the United States. Scand. J. Work Environ. Health prövning av en webbaserad metod för mätning av psykosocial arbets-
24 (1), 54–61. miljö och hälsa. (Webb-QPS – Development and test of a web-based
Arnold, K.A., Turner, N., Barling, J., Kelloway, E.K., McKee, M.C., 2007. method for assessment of psycho-social work environment and
Transformational leadership and psychological well-being: the me- health (Report C 2008-1)).Karolinska Institute, Stockholm.
diating role of meaningful work. J. Occup. Health Psychol. 12 (3), 193– Hausser, J.A., Mojzisch, A., Niesel, M., Schulz-Hardt, S., 2010. Ten years on:
203. a review of recent research on the Job Demand-Control (-Support)
Arvonen, J., 1995. Ledarskap och medarbetarhälsa – fallstudie i svensk model and psychological well-being. Work Stress 24 (1), 1–35.
processindustri [Leadership behavior and co-worker health. A study Heck, R.H., Thomas, S.L., Tabata, L.N., 2010. Multilevel and Longitudinal
in process industry]. (Report No. SR083). Stockholm: Stockholm Modeling with IBM SPSS. Taylor and Francis Group, Routledge, New
University. York.
Arvonen, J., 2002. Change production and employees – an integrated Hensel, R., Meijers, F., van der Leeden, R., Kessels, J., 2010. 360 degree
model of leadership. (Doctoral dissertation)Stockholm University, feedback: how many raters are needed for reliable ratings on the
Stockholm. capacity to develop competences, with personal qualities as devel-
Baron, R.M., Kenny, D.A., 1986. The Moderator Mediator Variable Distinc- opmental goals? Int. J. Hum. Resource Manage. 21 (15), 2813–2830.
tion in Social Psychological-Research – Conceptual, Strategic, Hofmann, D.A., Gavin, M.B., 1998. Centering decisions in hierarchical
and Statistical Considerations. J. Pers. Soc. Psychol. 51 (6), linear models: implications for research in organizations. J. Manage.
1173–1182. 24 (5), 623–641.
Bejerot, E., Hasselbladh, H., 2003. Nya kontroll-och maktrelationer inom Horn, J.E., Taris, T.W., Schaufeli, W.B., Schreurs, P.J., 2004. The structure of
sjukvården. (New control- and power relations within healthcare). occupational well-being: a study among Dutch teachers. J. Occupa-
Arbetsmarknad & arbetsliv 9 (2), 107–127. tional Organizational Psychol. 77 (3), 365–375.
Bergman, D., Lornudd, C., Sjöberg, L., von Thiele Schwarz, U., 2014. Leader Johnson, J.V., Hall, E.M., 1988. Job strain work place social support, and
personality and 360-degree assessments of leader behavior. Scand. J. cardiovascular-disease - a cross-sectional study of a random sample
of Psychol. 55, 389–397. of the Swedish working population. Am. J. Public Health 78 (10),
Bourbonnais, R., Comeau, M., Vezina, M., Dion, G., 1998. Job strain, 1336–1342.
psychological distress, and burnout in nurses. Am. J. Ind. Med. 34 Johnson, J.V., Hall, E.M., Theorell, T., 1989. Combined effects of job strain
(1), 20–28. and social-isolation on cardiovascular-disease morbidity and mortal-
Brief, A.P., Burke, M.J., Robinson, B.S., George, J.M., Webster, J., 1988. ity in a random sample of the Swedish male working population.
Should negative affectivity remain an unmeasured variable in the Scand. J. Work Environ. Health 15 (4), 271–279.
study of job stress. J. Appl. Psychol. 73 (2), 193–198. Kanste, O., Kyngas, H., Nikkila, J., 2007. The relationship between multi-
Cummings, G.G., MacGregor, T., Davey, M., Lee, H., Wong, C.A., Lo, E., dimensional leadership and burnout among nursing staff. J. Nurs.
Stafford, E., 2010. Leadership styles and outcome patterns for the Manag. 15, 731–739.
nursing workforce and work environment: a systematic review. Int. Karasek, R.A., 1979. Job demands, job decision latitude, and mental strain
J. Nurs. Stud. 47 (3), 363–385. – implications for job redesign. Adm. Sci. Q. 24 (2), 285–308.
Dahl, M.S., 2011. Organizational change and employee stress. Manage. Sci. Karasek, R.A., Theorell, T., 1990. Healthy Work: Stress, Productivity and
57 (2), 240–256. the Reconstruction of Working Life. Basic Books, New York.
Dallner, M., Elo, A.-L., Gamberale, F., Hottinen, V., Knardahl, S., Lindström, Kuoppala, J., Lamminpaa, A., Liira, J., Vainio, H., 2008. Leadership, job well-
K., Skogstad, A., Örhede, E., 2000. Validation of the General Nordic being, and health effects - a systematic review and a meta-analysis. J.
Questionnaire (QPS Nordic) for Psychological and Social Factors at Occup. Environ. Med. 50 (8), 904–915.
Work. Nordic Council of Ministers, Copenhagen. Larsson, J., Vinberg, S., 2010. Leadership behaviour in successful organisa-
de Jonge, J., Dollard, M.F., Dormann, C., Le Blanc, P.M., Houtman, I.L.D., tions: universal or situation-dependent? Total Quality Manage.
2000. The demand-control model: specific demands, specific control, Business Excellence 21 (3), 317–334.
and well-defined groups. Int. J. Stress Manage. 7 (4), 269–287. Leiter, M.P., 1990. The impact of family resources. Control Coping, and
de Lange, A.H., Taris, T.W., Kompier, M.A.J., Houtman, I.L., Bongers, P.M., Skill Utilization on the Development of Burnout - a Longitudinal-
2003. The very best of the millennium: longitudinal research and the Study. Hum. Relat. 43 (11), 1067–1083.
demand-control-(support) model. J. Occup. Health Psychol. 8 (4), Little, R.J.A., Rubin, D.B., 1987. Statistical Analysis with Missing Data. John
282–305. Wiley and Sons, New York.
de Lange, A.H., Taris, T.W., Kompier, M.A.J., Houtman, I.L.D., Bongers, P.M., Luchman, J.N., Gonzalez-Morales, M.G., 2013. Demands, control, and
2002. Effects of stable and changing demand-control histories on support: a meta-analytic review of work characteristics interrelation-
worker health. Scand. J. Work Environ. Health 28 (2), 94–108. ships. J. Occup. Health Psychol. 18 (1), 37–52.

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003
G Model
NS-2434; No. of Pages 12

12 C. Lornudd et al. / International Journal of Nursing Studies xxx (2014) xxx–xxx

Malloy, T., Penprase, B., 2010. Nursing leadership style and psychosocial Sellgren, S.F., Ekvall, G., Tomson, G., 2008. Leadership behaviour of nurse
work environment. J. Nurs. Manag. 18 (6), 715–725. managers in relation to job satisfaction and work climate. J. Nurs.
McKee, M.C., Driscoll, C., Kelloway, K.E., Kelley, E., 2011. Exploring lin- Manag. 16 (5), 578–587.
kages among transformational leadership, workplace spirituality and Skakon, J., Nielsen, K., Borg, V., Guzman, J., 2010. Are leaders’ well-being,
well-being in health care workers. J. Manage Spirituality Religion 8 behaviours and style associated with the affective well-being of their
(3), 233–255. employees?. A systematic review of three decades of research. Work
Michel, Lyons, B.D., Cho, J.W., 2011. Is the full-range model of leadership Stress 24 (2), 107–139.
really a full-range model of effective leader behavior? J. Leadership Skogstad, A., Einarsen, S., 1999. The importance of a change-centred
Organizational Stud. 18 (4), 493–507. leadership style in four organizational cultures. Scand. J. Manage.
Moyle, P., 1998. Longitudinal influences of managerial support on em- 15, 289–306.
ployee well-being. Work Stress 12 (1), 29–49. Spector, P.E., Zapf, D., Chen, P.Y., Frese, M., 2000. Why negative affectivi-
Munir, F., Nielsen, K., Garde, A.H., Albertsen, K., Carneiro, I.G., 2012. ty should not be controlled in job stress research: don’t throw out
Mediating the effects of work-life conflict between transformational the baby with the bath water. J. Organizational Behav. 21 (1), 79–95.
leadership and health-care workers’ job satisfaction and psychologi- Sverke, M., Arvonen, J., Lindell, M., 1999. Assessing Change-, Production-,
cal wellbeing. J. Nurs. Manag. 20 (4), 512–521. and Employee-oriented Leadership: Cross-cultural Comparisons of
Nielsen, K., Randall, R., Yarker, J., Brenner, S.O., 2008. The effects of Measurement Properties (Report No. 861) Stockholm University,
transformational leadership on followers’ perceived work character- Stockholm.
istics and psychological well-being: a longitudinal study. Work Stress Syrek, C.J., Apostel, E., Antoni, C.H., 2013. Stress in highly demanding it
22 (1), 16–32. jobs: transformational leadership moderates the impact of time
Nielsen, K., Yarker, J., Brenner, S.O., Randall, R., Borg, V., 2008. The pressure on exhaustion and work-life balance. J. Occup. Health Psy-
importance of transformational leadership style for the well-being chol. 18 (3), 252–261.
of employees working with older people. J. Adv. Nurs. 63 (5), Tafvelin, S., Armelius, K., Westerberg, K., 2011. Toward understanding the
465–475. direct and indirect effects of transformational leadership on well-
Nielsen, K., Yarker, J., Randall, R., Munir, F., 2009. The mediating effects of being: a longitudinal study. J. Leadership Organizational Stud. 18 (4),
team and self-efficacy on the relationship between transformational 480–492.
leadership, and job satisfaction and psychological well-being in Uher, R., Goodman, R., 2010. The everyday feeling questionnaire: the
healthcare professionals: a cross-sectional questionnaire survey. structure and validation of a measure of general psychological well-
Int. J. Nurs. Stud. 46 (9), 1236–1244. being and distress. Soc. Psychiatry Psychiatr. Epidemiol. 45 (3), 413–423.
Nyberg, A., Alfredsson, L., Theorell, T., Westerlund, H., Vahtera, J., Kivi- Van der Doef, M., Maes, S., 1999. The Job Demand-Control (-Support)
maki, M., 2009. Managerial leadership and ischaemic heart disease model and psychological well-being: a review of 20 years of empirical
among employees: the Swedish WOLF study. Occup. Environ. Med. 66 research. Work Stress 13 (2), 87–114.
(1), 51–55. Westerlund, H., Nyberg, A., Bernin, P., Hyde, M., Oxenstierna, G., Jappinen,
Nyberg, A., Holmberg, I., Bernin, P., Alderling, M., Akerblom, S., Widerszal- P., Theorell, T., 2010. Managerial leadership is associated with em-
Bazyl, M., Theorell, T., 2011. Destructive managerial leadership and ployee stress, health, and sickness absence independently of the
psychological well-being among employees in Swedish, Polish, and demand-control-support model. WORK: J. Prevent. Assessment Reha-
Italian hotels. WORK: J. Prevent. Assessment Rehabil. 39 (3), 267–281. bil. 37 (1), 71–79.
Olson, O., Guthrie, J., Humphrey, C., 1998. Global Warning: Debating Winwood, P.C., Lushington, K., 2006. Disentangling the effects of psycho-
International Developments in New Public Financial Management. logical and physical work demands on sleep, recovery and maladap-
Cappelen Akademisk Folag. tive chronic stress outcomes within a large sample of Australian
Rucker, D.D., Preacher, K.J., Tormala, Z.L., Petty, R.E., 2011. Mediation nurses. J. Adv. Nurs. 56 (6), 679–689.
analysis in social psychology: current practices and new recommen- Yukl, G., 1999. An evaluative essay on current conceptions of effective
dations. Social Personal. Psychol. Compass 5/6, 359–371. leadership. Eur. J. Work Organizational Psychol. 8, 33–48.
Schriesheim, C.A., Kinicki, A.J., Schriesheim, J.F., 1979. Effect of leniency Yukl, G., Gordon, A., Taber, T., 2002. A hierarchical taxonomy of leadership
on leader-behavior descriptions. Organ. Behav. Hum. Perform. 23 (1), behavior: integrating a half century of behavior research. J. Leader-
1–29. ship Organizational Stud. 9 (1), 15–32.

Please cite this article in press as: Lornudd, C., et al., The mediating role of demand and control in the relationship
between leadership behaviour and employee distress:
?A cross-sectional study. Int. J. Nurs. Stud. (2014), http://dx.doi.org/10.1016/j.ijnurstu.2014.08.003