Content uploaded by Ashley E. Nixon
Author content
All content in this area was uploaded by Ashley E. Nixon on Aug 13, 2014
Content may be subject to copyright.
PLEASE SCROLL DOWN FOR ARTICLE
This article was downloaded by:
[University Of South Florida]
On:
27 May 2011
Access details:
Access Details: [subscription number 936075832]
Publisher
Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-
41 Mortimer Street, London W1T 3JH, UK
Work & Stress
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713697904
Can work make you sick? A meta-analysis of the relationships between job
stressors and physical symptoms
Ashley E. Nixona; Joseph J. Mazzolab; Jeremy Bauera; Jeremy R. Kruegerc; Paul E. Spectora
a Department of Psychology, University of South Florida, Tampa, FL, USA b Department of Psychology,
University of Tulsa, Tulsa, OK, USA c College of Medicine, University of South Florida, Tampa, FL,
USA
Online publication date: 15 April 2011
To cite this Article Nixon, Ashley E. , Mazzola, Joseph J. , Bauer, Jeremy , Krueger, Jeremy R. and Spector, Paul E.(2011)
'Can work make you sick? A meta-analysis of the relationships between job stressors and physical symptoms', Work &
Stress, 25: 1, 1 — 22
To link to this Article: DOI: 10.1080/02678373.2011.569175
URL: http://dx.doi.org/10.1080/02678373.2011.569175
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
Can work make you sick? A meta-analysis of the relationships between
job stressors and physical symptoms
Ashley E. Nixon
a
*, Joseph J. Mazzola
b
, Jeremy Bauer
a
, Jeremy R. Krueger
c
and
Paul E. Spector
a
a
Department of Psychology, University of South Florida, Tampa, FL, USA;
b
Department of
Psychology, University of Tulsa, Tulsa, OK, USA;
c
College of Medicine, University of South
Florida, Tampa, FL, USA
A meta-analysis of 79 studies reporting cross-sectional and longitudinal relationships between
physical symptoms and various occupational stressors was conducted. Stressors were
organizational constraints, interpersonal conflict, role conflict, role ambiguity, workload,
work hours, and lack of control. The relationships between stressors and eight physical
symptoms were quantitatively summarized and contrasted, for both individual symptoms and
composite symptom scales. All of the occupational stressors were significantly related to
physical symptoms in cross-sectional analyses, and the effect sizes of these relationships varied
both by the stressor and the individual symptom examined. The longitudinal relationships
were similar to the cross-sectional results, and provided some evidence of temporal consistency
of the occupational stressorphysical symptom relationship. Organizational constraints and
interpersonal conflict had the strongest relationships with symptoms in both the cross-
sectional and longitudinal analyses. Gastrointestinal problems and sleep disturbances were
significantly related to more stressors than other symptoms examined. These findings show
that it is important to examine physical symptoms, as they are related to a wide range of job
stressors and these relationships prevail over time. Potential underlying mechanisms, including
the immediacy of physiological reactions to stressors, participants’ attributions concerning
stressorphysical symptom relationships, and the possible multidimensional nature of
symptoms, are proposed and discussed.
Keywords: stress; physical symptoms; organizational constraints; interpersonal conflict; role
stressors; workload; work hours; meta-analysis; work-related stress
Introduction
Employee health and well-being have gained increasing societal attention, driven by
both rising worker compensation claims and the considerable personal, organiza-
tional, and medical costs associated with stress-related illness (Smith, Karsh,
Carayon, & Conway, 2003). For example, the annual costs of employee stress,
including costs for missed wages due to absenteeism and reduced productivity and
health care costs, have been estimated to be $200350 billion in the United States,
$64.866.1 billion in the United Kingdom, and $232 billion in Japan (Miree, 2007).
*Corresponding author. Email: aenixon@mail.usf.edu
Work & Stress
Vol. 25, No. 1, JanuaryMarch 2011, 122
ISSN 0267-8373 print/ISSN 1464-5335 online
#2011 Taylor & Francis
DOI: 10.1080/02678373.2011.569175
http://www.informaworld.com
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
Researchers have conducted a great deal of research using a variety of health
indicators, but few quantitative reviews are available that integrate this literature. A
notable exception is Spector and Jex (1998), who conducted a small-scale (2 to 8
studies across variables) meta-analysis in order to provide validation evidence for
their physical symptoms scale by showing positive relationships of symptoms with
role ambiguity, role conflict, and weekly working hours, and a negative relationship
with job autonomy. However, this meta-analysis had limitations: it was confined to a
few stressors, included only studies using the authors’own symptom scale, looked
only at an overall symptom score, analysed a small number of studies, and only
investigated cross-sectional designs. Although those finding are certainly suggestive,
it is important to see whether they hold for specific symptoms, across diverse
indictors of physical symptoms, and in longitudinal designs that enable more
confident conclusions to be drawn about possible antecedents of symptoms.
The present meta-analysis addresses this gap in the literature by presenting a
large-scale review of the association between specific health complaints and various
indicators of job stress in both cross-sectional and longitudinal designs.
While most studies report results for composite symptom scores, it is important
to look at individual symptoms to see if they each relate to stressors in a similar way.
For example, does the perception of organizational constraints relate similarly to a
variety of symptoms, or does it relate more to headaches than to eye strains or
stomach distress? Darr and Johns (2008), looking at individual symptoms, found
that absenteeism had a stronger relationship with fatigue (r.32) than a composite
of other physical symptoms (r.17). These findings underscore the need to examine
the relationships between stressors and individual symptoms in addition to
composite scores. We believe that this study makes an important contribution to
the literature in that it is probably the first review to address the unique relationships
between specific symptoms and various job stressors.
Stress and physical symptoms
There are several theoretical models that attempt to explain the occupational
stressorstrain relationship, and many of them contain similar variables and
elements (e.g., French & Raven, 1962; Frese & Zapf, 1988; Spector, 1988). The
process that underlies such theoretical frameworks posits a stimulus-response process
in which job stressors (aspects of the work environment) lead to strains (psycholo-
gical, physical, and behavioural reactions), mediated by perceptions of the
environment. Emotional responses, such as anxiety or frustration, are often the
most immediate psychological strain responses that are associated with physiological
changes in the body (Spector, 1988).
The neurochemical responses of the body to stress have been examined
extensively, and while there remain many interactions between these neurochemicals
and the body that need further investigation, the major components of the process
are believed to be well identified and understood. Current research suggests that the
hypothalamic-pituitary-adrenocortical (HPA) axis and the sympathetic-adrenome-
dullary (SAM) systems are heavily involved in the relationship between stressors and
health (Dienstbier, 1989; Frankenhaeuser, 1991; Taylor, 1999). When an employee
encounters an event that is evaluated as harmful or threatening, the hypothalamus is
activated and reacts in two ways. First, it activates the HPA system, which is best
2A.E. Nixon et al.
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
understood through the three stages of the general adaptation syndrome: alarm,
resistance, and exhaustion (Selye, 1936, 1976). During this process, the hypothalamus
releases corticotrophin-releasing factor (CRF) into the nervous system. CRF, in
turn, stimulates the pituitary gland, causing it to secrete growth hormone, prolactin,
and adrenocorticotropic hormone (ACTH). ACTH affects the adrenal cortex, which
releases corticosteroids. Corticosteroids help the body return to a neutral state after
reacting to a stressor. Enzymes that act on opioid receptors, such as beta-endorphin
and enkephalin, are also released during the HPA system response to a stressor. They
are associated with immune system disruption (Cohen, Kessler, & Gordon, 1995).
When the hypothalamus is activated, it in turn activates the SAM system by
setting off one of the earliest responses to stress through the sympathetic nervous
system, a reaction that is commonly referred to as the fight-or-flight response
(Cannon, 1932). The activation of the sympathetic arousal in turn stimulates the
adrenal medulla, which secretes catecholamines, specifically epinephrine and
norepinephrine. The secretion of these hormones leads to increased pulse rate,
increased blood pressure, and sweating, among other physiological reactions. Over
time such reactions can result in the experience of physical symptoms, such as
stomach distress, headache, backache, and other musculoskeletal pain. We will next
discuss specific mechanisms leading to such individual symptoms.
Individual symptoms
Our meta-analysis will summarize the findings regarding eight types of individual
symptom: Backache, headache, eyestrain, sleep disturbance, dizziness, fatigue,
appetite, and gastrointestinal problems. We chose these symptoms because they
have been found to be associated with stress and because for each one there was a
sufficient number of cross-sectional studies available for a meta-analysis to be
conducted (k]5).
Backache. Musculoskeletal pains that originate in the upper back, lower back,
shoulders, and neck generally fall into the overarching category of backache. It has
been shown that tenderness in muscles and soft tissues such as tendons and ligaments
are markedly influenced by the impact of daily stressors (Manne & Zautra, 1989;
Zautra et al., 1998). A longitudinal study in individuals with the chronic
musculoskeletal pain diseases, specifically Fibromyalgia and arthritis, found that
exposure to stressors and negative mood over one week was related to increased
reports of pain intensity over the ensuing weeks (Davis, Zautra, & Reich, 2001;
Zautra, Johnson, & Davis, 2005). This relationship between stressors and muscu-
loskeletal pain is believed to arise from the low cortisol and elevated prolactin levels
present during times of stress, which may increase overall pain sensitivity by down-
regulation of the immune system and increased inflammation (Huyser & Parker,
1999; McLean et al., 2005; McLean et al., 2006).
Headache. Headache is pain experienced in the upper half and back of the head,
arising from a variety of mechanisms. Tension-type headaches in particular are
associated with exposure to stressors. The mechanism underlying such a headache is
likely multifactorial and mostly neurological in nature (Spierings, Ranke, &
Work & Stress 3
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
Honkoop, 2001). It is widely believed that pain-signalling pathways in the brain can
become overly sensitive to painful stimuli when stress is experienced (Bendtsen,
2003). Once the brain is overly sensitive to pain, even the slightest twinge may
increase nerve excitability and be translated into pain signals, leading to the increased
experience of headaches. This heightened nerve sensitivity leads to increased muscle
activity, which facilitates muscular tension, thus leading to the co-occurrence of
stress-induced back and body aches with headaches.
Eye strain. Eye strain, or asthenopia, refers to itchy, sore, or heavy eyes, as well as
blurred or double vision. Eye strains are considered one of several types of
musculoskeletal disorder (MSD) originating from muscular tension associated with
occupational stressors (Gura, 2002). As with backaches and headaches, an
individual’s response to occupational stressors produces a biochemical reaction
that promotes inflammation and increases sensitivity to pain sensation in and around
the eyes. Furthermore, workplace activities such as using a computer or performing
visually intense tasks promote the muscles of the inner eye to tighten and fatigue.
These effects combine with a heightened sensitivity to pain and are cumulative.
Sleep disturbance. Work-related stress is the most frequently reported cause of sleep-
related difficulties (Ertel, Karestan, & Berkman, 2008; Rau, Georgiades, Fredrikson,
Lemne, & de Faire, 2001; Roth & Ancoli-Israel, 1999) and 10% to 40% of the
working-age population (2060 years) in Sweden reported that they experienced
difficulty sleeping (Linton & Bryngelsson, 2000). Sleep difficulties are negatively
related to workplace productivity and positively related to the use of physical and
psychological health services and sick leave (Chevalier, Souques, Coing, Dab, &
Lambrozo, 1999; Jacquinet-Salord, Lang, Fouriaud, Nicoulet, & Bingham, 1993;
Stoller, 1994). In a self-report study, work demands were positively related to the
development and chronicity of sleeping problems (Jansson & Linton, 2006). Some
theories emphasize that blood ACTH and cortisol concentrations are increased in
response to acute stress by the release of corticotropin-releasing hormone (CRH)
from the hypothalamic-pituitary-adrenal axis (Aguilera, 1994). These hormones have
been shown to consequently reduce surges in night-time melatonin levels, which may
provoke insomnia and sleep disturbance.
Dizziness. The stress process can lead to feelings of dizziness or light-headedness for
some individuals. The primary mechanisms for this reaction are through changes in
metabolic rate, such as blood pressure and heart rate that are associated with
stressors (Sparacino, 1982). Dizziness may also be related to hyperventilation.
An increase in breathing rate mediated by the sympathetic nervous system during a
stressful time, even if it is minute, can alter the chemistry including the acid/base
levels in the body. This in turn may disrupt neural responses to balance and
coordination through the cerebellum and eighth nerve, which control balance. It has
been shown that nonspecific dizziness due to hyperventilation gradually increases
over several minutes and eventually resolves once the stressor is removed.
Fatigue. Fatigue, a feeling of lack of energy or tiredness, is one of the most common
symptoms reported by individuals who experience stressful situations (Dittner,
Wessely, & Brown, 2004; Franssen, Bultmann, Kant, & van Amelsvoort, 2003).
4A.E. Nixon et al.
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
Individuals who report fatigue often report other overlapping symptoms including
muscle pains, sore throats, inattentiveness, headaches, and unrefreshing sleep
(Fukuda et al., 1994). Stress-mediated fatigue is typically reported in the absence
of other illnesses and is directly correlated with a triggering event, even persisting well
beyond the period of stress. Several endocrine and metabolic dysfunctions have been
postulated to generate fatigue symptoms, yet the physiological mechanisms of fatigue
remain unclear. Studies do suggest a relationship with undersecretion of CRH and
cortisol (Demitrack et al., 1991), while others alternatively show no alteration in the
pituitary-adrenal axis (Young et al., 1998). Interestingly, recent genetic studies have
proposed that certain sequences of DNA that control stress through the hypotha-
lamic-pituitary-adrenal axis and the sympathetic nervous system vary among people
(Goertzel et al., 2006). On the individual basis, response to the hormones and
chemical messengers released by stress are interpreted uniquely (Smith, White,
Aslakson, Vollmer-Conna, & Rajeevan, 2006). Thus, people may have a genetic
predisposition to develop stress-mediated fatigue. This condition, no matter what the
physiological basis, is often chronic and disabling, leading to employees’decreased
overall health and well-being as well as a decreased ability to meet the job
requirements (Wagner, 1997).
Appetite. Exposure to stressful events has been associated with increased or
decreased appetite (Kandiah, Yake, & Willett, 2008). There are several mechanisms
through which this reaction can occur. For instance, brain-signalling pathways
respond to emotional stressors by rapidly increasing the production of hormones
from the melanocortinergic pathways, namely ACTH and melanocyte-stimulating
hormone (MSH). These in turn have been shown to mediate stress-induced anorexia,
leading to a loss of appetite (Liu et al., 2007). Alternatively, the sympathetic nervous
system mediates release of the amino acid ghrelin from the stomach. Ghrelin, in turn,
creates a hunger response that stimulates mealtime initiation. The normal ghrelin
levels present before a meal are increased by acute stressors, thus creating a need-for-
food state in the body and an increase in appetite (Ochi et al., 2008). While these
mechanisms may function in the immediate stress response, stress-induced cortisol
production tends to exert its effect over longer periods of time. Chronically high
levels of cortisol and ghrelin cause increased body fat stores. These newly formed fat
deposits in turn release the signalling peptide leptin, which decreases appetite
through neurogenic pathways. Leptin is released following the exhaustion phase of
the physiological stress process. This process may account for the recent finding that
21% of respondents reported a significantly lower appetite after a stressful event
(Kandiah et al., 2008). Research on changes in appetite in the occupational stress
and health literature primarily focus on loss of appetite (rather than increase in
appetite), which will be examined in this meta-analysis.
Gastrointestinal problems. There are several specific symptoms that fall into the
category of gastrointestinal problems, including nausea, acid indigestion or heart-
burn, abdominal and stomach cramps (excluding those that are due to menstrua-
tion), and irritable bowel syndrome. There are several physiological mechanisms
through which the stress process can affect gastrointestinal process. For instance,
during the acute stress response, the body produces increased amounts of ACTH.
This sympathetic nervous system mediator delays the stomach emptying phase of
Work & Stress 5
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
digestion, potentially leading to cramping and aches. Parodoxically, chronic stress
actually leads to increased amounts of the stomach enzyme ghrelin that in turn
increases hunger/appetite (Ochi et al., 2008).
The stress process also worsens the symptoms of heartburn by reducing the
stomach’s threshold level for pain. This in turn increases the perceived severity of
symptoms, such as burning ulcers and indigestion (Farre´et al., 2007; Fass et al.,
2008). In much the same way, stress responses in the nerve fibres in the
gastrointestinal track lead to a lower tolerance for the stomach’s natural ability to
expand. This leads to increased colonic muscle contractions, which may result in
irritable bowel symptoms or diarrhoea. These symptoms are thought to be secondary
to overactive production and release of corticotropin releasing factor (CRF), a
peptide that mediates stress responses (Keck & Holsboer, 2001).
The current study
We decided to conduct a meta-analysis, the goals of which were four-fold. First, we
quantitatively summarized findings across studies that showed relationships between
people’s reports (perceptions) of specific occupational stressors and physical
symptoms. This was to provide some evidence concerning the magnitude of effects
that speak to the importance of studying stressorsymptom relationships. We report
results for variable combinations where there were sufficient numbers of studies
available. Second, we contrasted the magnitude of correlations across different
stressors to show if the potential effects of some stressors were larger than others.
Such findings can potentially help focus intervention attempts toward those stressors
that are likely to have the biggest impact on health. Third, we compared the
magnitude of correlations across different symptoms within each stressor to show if
the potential effects of stressors varied by the particular symptom. Finally, we
compared results from studies using cross-sectional designs with results from studies
using longitudinal designs to help rule out the possibility that observed relationships
might be due to occasion factors (i.e., shared biases produced by transitory
conditions such as mood), and to show that the potential effects of stressors might
persist over time. Although the nature of all of these studies does not allow for
confident causal conclusions to be drawn, demonstrating relationships both
contemporaneously and over time is a first step.
The studies we meta-analysed assessed physical symptoms by means of employee
self-reports. The nature of these symptoms is such that it would be difficult to use
other than self-reports, since do not normally exhibit physical signs that could be
subjected to a medical test. However, such symptoms can serve as important
indicators of health, as they have been shown to predict a variety of long- and short-
term health outcomes (Benyamini & Idler, 1999; Ferraro & Kelley-Moore, 2001;
Idler & Benyamini, 1997), including mortality (Idler & Benyamini, 1997).
Method
Literature search
In order to locate relevant studies for this meta-analysis, we conducted an electronic
search of bibliographies using the PsycInfo (Psychological Abstracts), ABI/Inform,
6A.E. Nixon et al.
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
Medline, and Dissertation Abstract International databases. We searched for
published studies up to 2009 that examined the relationship between occupational
stressors and physical health outcomes. Occupational stressors examined in this
review include (lack of) control or autonomy at work, interpersonal conflict,
organizational constraints, role conflict, role ambiguity, workload, and work hours.
Stressors were included that were included in at least five studies that we could locate,
which was the minimum number for a meta-analysis. The following search terms
were utilized: role conflict, role ambiguity, workload, work hours, constraints,
interpersonal conflict, relationship conflict, conflict at work, autonomy, control, well-
being, physical symptoms, symptoms, health, ill-health, and illness. Additionally,
database searches were conducted for two prominent symptom scales in stress
research: the Physical Symptom Index (PSI; Spector & Jex, 1998) and the
Occupational Stress Indicator (OSI; Cooper, Sloan, & Williams, 1988).
Manual searches for the most popular outlets for occupational stress research
were conducted, and included three conferences for the years 2002 through 2009:
these were the Work, Stress, & Health conferences,American Psychological
Association conventions, and the conferences of the Society for Industrial/Organiza-
tional Psychology. We manually searched six journals for the years 1996 through
2009: Work & Stress, Journal of Occupational Health Psychology, Journal of
Organizational Behavior, Journal of Occupational and Organizational Psychology,
Stress & Health, and Anxiety, Stress, & Coping. Additionally, we emailed relevant
electronic mailing lists (listservs) to request unpublished and new work being
conducted. Finally, several prominent authors in this area were contacted in order to
obtain additional samples of published and unpublished work. Given that multiple
search methods were used, we came across similar articles by the same author using
the same or overlapping datasets. Overlapping data only occurred when conference
presentations, dissertations, or unpublished data sets were used, which were then
later published in a peer-reviewed journal. If two versions of the same study reported
some variables in one but not the other, both studies were used. In the case of
overlapping data sets, only one was included in our analyses and the duplicate was
discarded.
Inclusion criteria
Our meta-analysis examined the relationships between a variety of occupational
stressors and physical health outcomes. The studies that were considered for the
meta-analysis had to meet several criteria: (a) the level of analysis had to be the
individual employee as opposed to a group, team, or organization; (b) the study had
to use one or more measures of occupational stressor; (c) the study had to use one or
more measures of physical symptoms; (d) the study had to report a correlation
coefficient, or other statistic that could be converted into a correlation coefficient;
(e) the study must have been conducted with working populations reporting about
naturally occurring job settings, thus excluding laboratory samples or scenario
studies; (f) the data must have been self-report by the employee experiencing the
stressor only and not reported from other sources; and (g) longitudinal data must
have reported the relationship between occupational stressor variables measured at
time one and physical symptoms measured subsequently.
Work & Stress 7
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
Stressor variables
Interpersonal conflict. Interpersonal conflict refers to negative interactions with
others in the workplace, which can range from momentary disagreements to heated
arguments and bullying. Interpersonal conflict can develop as a result of a
competitive work environment and rude or aggressive behaviour of coworkers,
supervisors, and customers. Interpersonal conflict has also been examined by
separating conflict with supervisors from conflict with coworkers, although the
strength of the associations between interpersonal conflict and employee physical
symptoms did not vary significantly across the supervisors and coworkers in the few
studies that compared them (Bruk-Lee & Spector, 2006; Frone, 2000; Lubbers,
Loughlin, & Zweig, 2005). Therefore, these two sources of conflict were combined by
averaging correlations across sources. Interpersonal conflict included measures of
direct and indirect conflict at work, verbal and physical abuse, and bullying. In the
studies that we surveyed interpersonal conflict was measured with a wide variety of
scales, the most commonly used of which was the Interpersonal Conflict at Work
Scale (ICAWS; Spector & Jex, 1998).
Lack of control. Job control, or job autonomy, is a manifestation of perceived control
in the workplace. High levels of job autonomy indicate that an employee has control
over how or when his or her tasks are performed (Jex, 2002). Studies in this meta-
analysis used measures of both the presence of and lack of control. However, the
relationships examined in this study were based on lack of control, and for any study
reporting relationships between control and symptoms we reversed the sign of the
correlation. The most common measures for control were the Job Content
Questionnaire (JCQ; Karasek, 1985), the Job Diagnostic Survey (JDS; Hackman
& Oldham, 1975, 1976; Idaszak & Drasgow, 1987), and the Leiden Quality of Work
Questionnaire (LQWQ, van der Doef & Maes, 1999).
Organizational constraints. Organizational constraints are situations or environmen-
tal hurdles that impede employees from completing their job duties and hinder job
performance. Organizational constraints can refer to elements of the job, such as not
having job-related information, having limited time and materials, or not having the
necessary job-related authority to be able to complete one’s task (Peters & O’Connor,
1988). All organizational constraints studies used in this meta-analysis measured
them with either the Organizational Constraints Scale (Spector & Jex, 1998) or the
Job Effectiveness Survey (Eulberg, 1984).
Role ambiguity. Role ambiguity occurs when job-related information is not made
clear, causing an employee to be unsure of his or her job duties and requirements.
Role ambiguity can refer to performance expectations or standards, tasks or duties,
scheduling, and other aspects of the work environment. The most common measure
that we found for role ambiguity was developed by Rizzo, House, and Lirtzman
(1970).
Role conflict. Role conflict occurs when different members of the organization give
employees inconsistent or conflicting information concerning their job demands.
Conflicting or inconsistent information could come from multiple individuals or a
8A.E. Nixon et al.
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
single person within the organization. The most common measure for role conflict in
the studies that we surveyed was that of Rizzo et al. (1970).
Workload. Workload refers to the amount of work an employee is required to
complete in a given amount of time, along with the effort it takes to complete it.
Workload includes both a quantitative component, such as the amount of work one
is expected to complete, and a qualitative component, or the mental effort required
to complete the tasks. The most common workload measure used was the
Quantitative Workload Inventory (Spector & Jex, 1998).
Work hours. Work hours refer to the amount of time an employee works per unit of
time on average. Work hours were measured in either total number of hours worked
per week or average number of hours worked per day/week.
Physical symptoms. Physical symptoms were necessarily measured via self-report. In
this meta-analysis, analyses were conducted for both total aggregate symptoms and
individual symptoms, specifically backache, headache, eye strain, sleep disturbance,
dizziness, fatigue, appetite loss, and gastrointestinal problems. The majority of
studies only indicated symptoms at the aggregate level, but when individual symptom
data were available, the study was utilized in both total and individual symptoms
analyses. The most commonly utilized physical health/symptom measure was the PSI
(Spector & Jex, 1998).
A complete list of measures for all independent and dependent variables analysed in
this meta-analysis can be obtained from the first author.
Statistical methods
For our analyses, we used the Rosenthal (1991) approach to meta-analysis. The
Rosenthal approach computes descriptive statistics based on the sample of
correlations that are similar to the bare-bones approach outlined by Hunter and
Schmidt (1990). The primary differences between these approaches is that the Hunter
and Schmidt approach recommends estimating how much observed variance in
correlations can be attributed to artefacts such as sampling error or unreliability of
measures by applying adjustment formulas. These formulas are also used to estimate
the underlying theoretical population parameters by allowing for the adjustment of
the observed mean correlations. These adjustments can be problematic and
inaccurate (e.g., James, Demaree, & Mulaik, 1986; Rosenthal, 1991; Spector &
Levine, 1987), particularly when analysing a relatively small number of studies per
variable, as is the case with some of our variable combinations in this meta-analysis.
Additionally, the most frequent correction applied in the Hunter and Schmidt
approach is unreliability in the criterion. Our studies did not routinely report internal
consistencies for total symptom scores, and no measures of reliability were reported
for single-item measures of individual symptoms. Data from longitudinal studies
were dealt with separately from cross-sectional studies.
The coding process included identifying the relevant correlations and sample
sizes for studies that met our inclusion criteria. When the sample size was reported as
a range, a conservative approach was employed by taking the minimum of the range.
Work & Stress 9
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
Coding of relevant information from manuscripts was conducted by the authors,
with each article being coding by two coders. Any discrepancies found were clarified
by an additional coder to ensure accurate information was recorded. For the 95%
confidence intervals used to draw conclusions about the significance of the mean
correlations, we multiplied the standard deviation of the mean by 1.96 and then
added and subtracted the result from the mean correlation as recommended by Lee
(1989).
Results
Our literature search returned a total of 510 published journal articles, dissertations,
and unpublished papers/data that concerned occupational stressors and health
outcomes. Studies were screened according to the inclusion criteria described below.
We found 79 studies that fitted the inclusion criteria; 64 were journal articles, 7 were
theses or dissertations, 4 were conference presentations, and 4 were unpublished data
sources. For cross-sectional relationships between occupational stressors and
composite physical symptoms, the literature search yielded 313 correlations. The
number of samples per variable for the cross-sectional analyses ranged from 25 to 92
for composite symptoms analyses and 3 to 40 for the analysis conducted at the
individual symptom level. The number of samples per variable for longitudinal
analyses ranged from 2 to 7.
A list of references for articles included in the meta-analysis is available upon
request from the first author.
Composite physical symptoms and occupational stressors
The relationships between the overall physical symptoms scale scores and each of the
occupational stressors are presented in Table 1. More complete meta-analytical
results for all analyses are available upon request from the first author. All of the
occupational stressors examined in this study were positively related to composite
symptom scores. The confidence intervals for each of the occupational stressors
Table 1. Meta-analysis of the relationships between composite physical symptoms and
occupational stressors.
Cross-sectional analyses Longitudinal analyses
Occupational stressor k N wr SD k N wr SD
Interpersonal conflict 25 10,215 .22* .11 3 1,483 .13* .17
Lack of control 58 32,645 .07* .15 4 2,455 .14* .15
Organizational constraints 34 8,212 .33* .09 1 110 .18
Role ambiguity 33 13,556 .15* .11 4 726 .17* .11
Role conflict 26 4,880 .27* .13 4 499 .10* .13
Work hours 39 11,354 .09* .11 2 496 .06 .04
Workload 92 36,610 .22* .14 7 3,057 .16* .13
Note: knumber of correlation; Ncombined sample size; wrweighted effect size; SDstandard
deviation.
*95% confidence interval excludes 0.
10 A.E. Nixon et al.
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
excluded zero, indicating that the relationships with composite scores from physical
symptom inventories are significantly different from zero.
The meta-analytical results indicated that we can classify the stressors into three
groups based on strength of relationship with physical symptoms, using z-tests for
independent correlations to determine significant differences in effect sizes.
Organizational constraints had the strongest mean weighted correlation (wr) with
overall physical symptoms; wr .33. This relationship was significantly stronger
than all of the other relationships between occupational stressors and physical
symptoms. The next largest mean correlation was for role conflict (wr.27), which
was significantly larger than for the remaining variables. Next was interpersonal
conflict (wr.22), and workload (wr.22), both of which were significantly larger
than for the remaining variables. Next was role ambiguity (wr .15), which was
significantly different from that of the other variables. Finally, work hours (wr.09)
and lack of control (wr .07) had significant relationships with the composite
physical symptoms. These last relationships did not differ from one another,
although they were significantly smaller than other occupational stressors relation-
ships with physical symptoms, as demonstrated by the exclusive confidence intervals.
Longitudinal relationships between composite physical symptoms and occupational
stressors
All of the occupational stressors had been examined in relation to composite physical
symptoms in a longitudinal context (see Table 1). Time frames across studies varied
considerably from one month to more than a year, and given the limited numbers of
studies, time frames were not the same across stressors. As can be seen in Table 1, six
of eight mean correlations were significantly different from zero (all but interpersonal
conflict, which had a small sample size [N110], and working hours). The
magnitudes of the relationships between three of the longitudinal mean correlations
(organizational constraints, role conflict, and work hours) were not significantly
different from the cross-sectional results based on z-tests. Three of the longitudinal
relationships (interpersonal conflict, role conflict, and workload) were significantly
smaller, and one longitudinal relationship (lack of control) was significantly larger in
magnitude when compared to cross-sectional relationships. Additionally, the
relationship between work hours and physical symptoms was found to be non-
significant in the longitudinal context and significant in the cross-sectional context.
Individual physical symptoms and occupational stressors
The individual symptom analysis revealed that the occupational stressors did
differentially relate to individual strains (see Table 2ac). Interpersonal conflict,
organizational constraints, and workload were significantly related to every physical
symptom, although there was a different pattern in relative strength of relationship
across these stressors. Interpersonal conflict had a significantly larger mean
correlation with sleep disturbances than with dizziness, headaches, or fatigue.
Organizational constraints had significantly larger mean correlations with fatigue
and gastrointestinal problems than the other symptoms. Workload had a larger mean
correlation with fatigue, followed by its relationship with eye strain, which were both
significantly larger than the relationship with the other symptoms. Workload had
Work & Stress 11
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
Table 2a. Meta-analysis of the relationships between occupational stressors and individual
physiological symptoms: Backache, headache and eye strain.
Physiological symptoms
Backache Headache Eye strain
Occupational stressor k N wr SD k N wr SD k N wr SD
Interpersonal conflict 6 2,853 .19* .04 5 844 .12* .16 5 845 .19* .10
Lack of control 13 5,556 .13* .08 12 3,522 .07* .10 6 867 .01 .11
Organizational
constraints
29 7,226 .16* .10 29 7,326 .18* .10 29 7,327 .19* .08
Role ambiguity 5 971 .05 .16 4 644 .02 .13 4 644 .00 .09
Role conflict 4 487 .16* .10 4 487 .04 .10 4 487 .13* .05
Work hours 30 7,894 .01 .08 29 7,567 .03 .10 29 7,568 .08* .07
Workload 40 11,086 .12* .08 39 10,643 .14* .11 33 7,989 .20* .09
Table 2b. Meta-analysis of the relationships between occupational stressors and individual
physiological symptoms: Sleep disturbances, dizziness and fatigue.
Physiological symptoms
Sleep disturbances Dizziness Fatigue
Occupational stressor k N wr SD k N wr SD k N wr SD
Interpersonal conflict 5 845 .22* .06 4 710 .11* .12 3 351 .09* .02
Lack of control 12 3,522 .13* .12 6 867 .06 .08 6 1,621 .08 .10
Organizational constraints 29 7,326 .17* .10 29 7,326 .17* .11 29 7,327 .25* .08
Role ambiguity 4 644 .04 .20 4 644 .02 .07 3 508 .13* .11
Role conflict 4 487 .13* .15 4 487 .10* .07 3 351 .09 .02
Work hours 29 7,568 .05* .09 29 7,567 .02 .11 30 8,458 .09* .09
Workload 39 10,644 .14* .09 33 7,988 .10* .10 33 8,744 .31* .09
Table 2c. Meta-analysis of the relationships between occupational stressors and individual
physiological symptoms: Appetite and gastrointestinal problems.
Physiological symptoms
Appetite Gastrointestinal problems
Occupational stressor k N wr SD k N wr SD
Interpersonal conflict 5 845 .16* .07 6 2,852 .19* .06
Lack of control 6 867 .03 .10 12 3,522 .09* .08
Organizational constraints 29 7,327 .18* .09 29 7,325 .26* .09
Role ambiguity 4 644 -.04 .11 4 644 .06 .13
Role conflict 4 487 .10 .15 4 487 .17* .03
Work hours 29 7,568 .04* .10 29 7,565 .05* .09
Workload 33 7,989 .11* .10 39 10,642 .13* .09
Note: knumber of correlations; Ncombined sample size; wrweighted effect size; SD standard
deviation.
*95% confidence interval excludes 0.
12 A.E. Nixon et al.
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
significantly larger mean correlations with appetite loss and dizziness than headaches
and sleep disturbances, based on results from z-tests.
Role conflict and work hours were each significantly related to five of the eight
physical symptoms examined. Role conflict had positive relationships with back-
aches, dizziness, eye strain, gastrointestinal problems, and sleep disturbances. Role
conflict had a significantly larger mean correlation with gastrointestinal problems
than with headaches, based on results from z-tests. Work hours had positive
relationships with appetite, eye strain, fatigue, gastrointestinal problems, and sleep
disturbances. Work hours had a larger mean correlation with fatigue than all other
symptoms except eyestrain. Work hours had a larger mean correlation with eyestrain
than backaches, headaches, dizziness, and appetite loss. Additionally, work hours
had a significantly smaller relationship with backaches than all other symptoms with
the exception dizziness. Lack of control had positive relationships with four of the
physical symptom variables, including backaches, gastrointestinal problems, head-
aches, and sleep disturbances. Backaches and sleep disturbance had larger mean
correlations with lack of control than headaches and appetite loss. Additionally, lack
of control had a smaller mean correlation with eye strain than backaches, headaches,
sleep disturbances, fatigue, and gastrointestinal problems. Finally, role ambiguity
was only significantly related to fatigue; this relationship was significantly larger than
the relationships between role ambiguity and eyestrain or appetite, based on z-tests.
None of the physical symptoms related to all of the occupational stressors.
However, gastrointestinal problems and sleep disturbances were significantly related
to more occupational stressors than any of the other physical symptoms, each having
significant relationships with six of the seven occupational stressors examined.
Backaches, eye strains, and fatigue were each related to five occupational stressors.
Changes in appetite, dizziness, and headaches showed the fewest significant
relationships with occupational stressors, each relating to four occupational stressors.
Discussion
Physical symptoms as defined in our study are manifestations of physical strain that
are likely to be responses to environmental stressors at work. Such strains are
particularly important given that they are predictive of future morbidity and
mortality as well as absenteeism (Benyamini & Idler, 1999; Darr & Johns, 2008;
Ferraro & Kelley-Moore, 2001; Idler & Benyamini, 1997). The goal of this meta-
analysis was to quantitatively summarize the cross-sectional and longitudinal
relationships between occupational stressors and both composite and individual
self-reported physical symptoms. These relationships had not been as thoroughly
examined previously in a meta-analysis. Results demonstrate that occupational
stressors were related to physical symptoms both cross-sectionally and over time.
Additionally, certain physical symptoms, including gastrointestinal problems and
sleep disturbances, were related to more occupational stressors than other symptoms,
such as dizziness or headaches. In the following section, we will discuss the
relationship patterns identified in these analyses mechanisms that may help
explain these relationship patterns, as well as relevant research and practical
implications and opportunities.
Work & Stress 13
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
Composite symptoms and occupational stressors
All of the occupational stressors included in this meta-analysis were significantly and
positively related to composite physical symptoms in the analysis of cross-sectional
studies, although relationships with physical symptoms were stronger for organiza-
tional constraints, role conflict, interpersonal conflict, and workload than for role
ambiguity, work hours, and lack of control. Longitudinal analyses, while hindered by
the limited number of available studies (k2 to 7), provide some evidence for
temporal consistency of the stressorphysical symptom relationships. Importantly,
the effect sizes of the longitudinal relationships between some of the occupational
stressors were not significantly different from the cross-sectional results. Specifically,
longitudinal relationships between interpersonal conflict, workload, and working
hours were positive and not significantly different from the corresponding cross-
sectional relationships. It should be kept in mind that the sample sizes for
some stressors in the longitudinal studies was rather limited, thus not providing
adequate statistical power. For example, the largest correlation with the longitudinal
studies was for organizational constraints (r.18), which was non-significant,
although only based on a sample size of 110.
Our results show that three of the four stressors (organizational constraints,
interpersonal conflict, and workload) relating most strongly to overall symptoms are
the same that have consistently emerged as most salient in qualitative studies of
stressful work incidents (e.g., Keenan & Newton, 1985; Liu, Spector, & Shi, 2007;
Narayanan, Menon, & Spector, 1999). Perhaps these stressors are reported more
frequently in qualitative research because they have a stronger impact on strains,
including physical symptoms, and thus the stressors are more salient to employees.
Thus, employees might judge the severity and impact of a stressor by the
physiological reaction that follows the stressor. On a different note, this pattern of
meta-analytic findings has implications for the ways in which occupational stressors
are conceptualized. Specifically, our results show that both chronic stressors (i.e.,
organizational constraints, role conflict) and acute stressors (i.e., interpersonal
conflict) can have significant relationships with employees’health; therefore further
examination into both types of stressor is essential. Additionally, research addressing
the underlying mechanisms for acute and chronic stressors is warranted, as these
chronic and acute stressors may impact on physical symptoms through unique
underlying mechanisms. Finally, this pattern of results can help guide intervention
and restructuring efforts by organizational leaders concerned with the physical health
of their employees. Reducing organizational constraints and role conflict may be
effective and accessible ways to reduce employees’physical symptoms that might
play a role in more serious health problems.
Immediacy of physiological reactions and specific symptoms
The individual symptom analyses revealed differences and patterns across relation-
ships between occupational stressors and physical symptoms. We propose that the
immediacy of physiological reactions represents an underlying mechanism that is
useful in understanding many of these diverse relationships. Taking all of the
occupational stressorindividual symptom relationships into account, gastrointest-
inal problems and sleep disturbances were significantly related to more occupational
14 A.E. Nixon et al.
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
stressors than the other physical symptoms. One reason that individuals may
frequently report these physical symptoms is that they are the by-product of
immediate physiological reactions. Experiences of increased levels of physical and
mental stressors initiate an immediate physiological reaction involving elevated
ACTH and sympathetic nervous system agonists, which create the rapid response of
slowing digestion and causing an upset stomach. The time period for this reaction to
occur is minimal, leading to the almost instantaneous perception of the physical
symptom, and an awareness that a stress reaction has taken place. The fact that many
people report gastrointestinal reactions is probably secondary to this strain being one
of the first noticed and easily relatable symptoms to stressors.
In much the same way, sleep patterns may vary in the immediate setting as well.
As the body enters a stressful situation, its sympathetic nervous system activity is
elevated. The hormones generated by the ‘‘fight or flight’’ response may cause lasting
tension and eventually disrupt sleep. In fact, the very mechanisms needed to gain
sleep are the opposing mechanisms driven by the ‘‘fight or flight’’ response; they
specifically include blood pressure, heart rate, and respiratory rate. A disruption of
these mechanisms may lead to ineffective sleep, which is often an immediate outcome
and one that is, like gastrointestinal problems, easily attributable to stressors.
A promising area for future research could be examining the role that mental
rumination, or replaying and dwelling on a stressful event, may play in sleep
disruption by increasing anxiety and stress reactions (Cropley, Dijk, & Stanley,
2006).
Many of the other symptoms, including backaches, headaches, eye strain, and
loss of appetite, represent physical reactions that generally develop over time.
Stressful events are often related to increased complaints of pains and strains weeks
after the initial incident (Davis et al., 2001; Manne & Zautra, 1989; Zautra et al.,
1998; Zautra et al., 2005). When individuals experience stressors, the body’s pain
threshold is decreased while the pain signalling pathways become increasingly
sensitive (Huyser & Parker, 1999; McLean et al., 2005; McLean et al., 2006). Thus,
the nervous system becomes overly sensitive to pain signals, such that even the
slightest insult to the system may cause symptoms such as headaches, back pains, or
eye strains.
Elusively, transient and subtle symptoms may not be as easily explained by the
immediacy of physiological responses. These symptoms, such as dizziness, are easier
to miss or otherwise explain away. Dizziness passes quickly, thus individuals may
take little note of this reaction, while backache remains more pervasive over long
periods of time, making it more likely that individuals will be concerned about back
pain and recall it when completing a symptom checklist. This alternative hypothesis
should be evaluated through future research.
The immediacy of physiological responses in reaction to stressors reinforces the
need for longitudinal research examining physical reactions to occupational
stressors. Currently, to our knowledge individual symptoms related to occupational
stressors have not been examined in a longitudinal research setting. This is an
important next step for this area of research, as the relationships between these
variables may vary over time. In particular, time series designs would be useful in
examining the development of chronic physical symptoms, as well as changes such as
habituation or further progression to chronic illnesses.
Work & Stress 15
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
Fatigue
Fatigue is somewhat unique among the individual symptoms that were examined in
this study. Unlike the other seven symptoms, fatigue is a condition that can be
associated with other symptoms, including muscle pains, headaches, and unrefresh-
ing sleep. Consequently, the relationships between fatigue and occupational stressors
may be explained through additional means than the other physical symptoms. One
potential reason that fatigue is a very frequently reported strain is that it is the most
easily recognizable and often disabling consequence of chronic stress (Dittner et al.,
2004; Franssen et al., 2003). Alternatively, the multidimensional nature of fatigue
may itself account for the frequency with which it is reported. If individuals report
fatigue when any combination of symptoms is present, it might receive more frequent
endorsements than a uni-dimensional symptom, such as back pain. In the future,
researchers should consider studying reports of fatigue in more depth, to gain a
better understanding of how occupational stressors impact on this constellation of
symptoms.
Limitations
Despite an extensive search of both published and unpublished literature in an
attempt to include as many effect sizes as possible given our inclusion criteria, a few
of the findings reported here are based on a small number of studies and, therefore,
require more careful consideration. This is particularly relevant when interpreting
the longitudinal results and the findings associated with several occupational
stressors and individual physical symptoms. Given the inability to draw causal
conclusions from cross-sectional data, it is important that researchers conduct more
longitudinal research in this area. As more longitudinal research is conducted, the
nature of longitudinal relationships should be examined in comparison to the cross-
sectional relationships to search for emerging differences.
Another limitation particular to the individual symptom analyses is that many
had low incidence rates, thereby attenuating the magnitude of possible correlations.
One reason that individual symptoms are often combined into composite indices is
to increase the range and variance of the measure, thus increasing the likelihood of
finding significant relationships. Our results suggest that relationships of individual
symptoms can vary across stressors, which argues against combining symptoms into
composites. At the very least one needs to be cautious not to interpret composite
scores to mean that all components of the composite are necessarily related to
stressors (or other variables) or that the magnitude of relationships are equivalent
across components.
There are also sources of potential bias that plague meta-analytic work, and this
project is no exception. First, there is always the danger than these results suffer from
publication bias that produces the ‘‘file drawer’’ problem in which only significant
results are published. It is possible that our results over-estimate the size of stressor
symptom relationships due to this limitation. However, it should be noted that many
of the studies in our meta-analysis included measures of psychological strains that
generally show stronger stressorstrain relationships. It seems unlikely that an
author would fail to publish a study just because there were small or nonsignificant
relationships with physical symptoms, as psychological strains would likely show
16 A.E. Nixon et al.
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
significant results. Furthermore, in an attempt to address these sources of bias, we
included conference presentations, dissertations, and unpublished studies that would
not be subject to the same publication biases, unless it is the case that authors are
reluctant to share null results. Nevertheless, as with any meta-analysis, we cannot be
certain that we did not have a biased sample of studies.
Furthermore, the studies included in our meta-analysis were entirely self-report
with all measures coming from a survey of employees. Thus we can say that
perceptions of stressors relates to reported symptoms. What cannot be determined
from such designs is the extent to which these relationships generalize to actual
environmental exposure. In other words, is it the amount of the stressor (e.g., number
of constraints) or merely the perception? This is important as it determines whether
interventions should focus on the environment or the individual. If the former, then a
focus on changing the environment would likely be most important, for example, by
shifting workloads to relieve overwork, or hiring more staff to reduce overtime hours.
On the other hand if the environment itself is unimportant, a focus on the individual,
perhaps through stress management training focused on cognitive restructuring
should be the target of intervention.
Another limitation to single-source designs is whether observed correlations are
distorted by shared biases among variables. For example, are people who are high on
neuroticism predisposed to be negative and complain, thus giving high ratings on
stressor and symptom scales? One possible mechanism by which biases could arise is
attributions: that is, the way in which an individual attributes a cause to an effect.
Individuals’schemata of strains may consist of more immediate reactions, such as
abdominal and stomach cramps and fatigue, as opposed to long-term reactions, such
as eye strain. It may be that while responding to a survey about stressors, this strain
schema is activated, leading individuals to report more immediate physical strains.
This mechanism could also potentially explain why dizziness is less often reported as
a physical symptom associated with occupational stressors. While dizziness is an
immediate reaction to stress, in our findings it did not demonstrate a similar pattern
of relationships with occupational stressors as did the other immediate physical
symptoms, perhaps because dizziness is not primarily attributed to the physiological
stress process (Kasper et al., 2005).
The attributional process could potentially inflate relationships between
stressors and strain. For instance, individuals who have reported many symptoms
that they believe relate to stressors might then inflate reports of stressors, or
individuals who report frequent and intense experience of occupational stressors
might then inflate symptom reports. In the future it would be interesting to
investigate how schemata about the impact of occupational stressors on physical
health might impact on self-reported physical symptom data. The lack of research
concerning the role of attributional processes in which individuals engage when
responding to stressor and strain surveys is a substantial hindrance in the
development and advancement of this field, and one that critically needs to be
addressed in future research.
Conclusions
We believe that this study advances knowledge and provides a platform for
researchers to better understand the relationships between occupational stressors
Work & Stress 17
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
and employees’physical health, as well as provide pointers to essential next steps in
this research area. All of the occupational stressors examined were significantly
related to composite ratings of physical symptoms, with organizational constraints,
role conflict, interpersonal conflict, and workload having the strongest relationships
with physical symptoms. Studies that examined longitudinal relationships between
these occupational stressors and physical symptoms provide some evidence for the
temporal consistency of symptoms, and help control for transient occasion factors
that could distort observed relationships.
Based on our findings, we conclude that physical symptoms are a vitally
important strain to continue studying in occupational health research, not just
because of the possible severity of outcomes, which can be highly related to morbidly
and mortality, but also because physical symptoms are affected by a wide range of
occupational stressors and these relationships can prevail over time.
By examining the relationships between occupational stressors and specific
physical symptoms (rather than composite symptoms), we identified that there were
differences across these relationships. Furthermore, a novel feature of this study is
the distinction that was drawn between immediate and delayed physiological
responses to stressors. This immediacy of physiological responses was proposed as
a mechanism to explain differences in relationships between stressors and specific
physical symptoms. We have also drawn attention to the importance of under-
standing how individuals’attributions about the nature of physical responses to
stressors may relate to stressorphysical symptom research. Future research
examining the diverse relationships between occupational stressors and physical
symptoms, as well as potential underlying mechanisms, may lead to changes in the
research methodology currently used to study physical symptoms, such that they may
be better examined by multidimensional scales, rather than the uni-dimensional
scales currently employed.
Acknowledgements
This research was supported in part by the occupational health psychology program of the
Sunshine Education and Research Center at the University of South Florida. The Sunshine
Education and Research Center is supported by Training Grant No. T42-CCT412874 from the
Centers for Disease Control and Prevention/National Institute for Occupational Safety and
Health.
References
Note: A complete list of references for articles included in the meta-analysis is available upon
request from the first author.
Aguilera, G. (1994). Regulation of pituitary ACTH secretion during chronic stress. Frontiers in
Neuroendocrinoloy,15(4), 321350.
Bendtsen, L. (2003). Central and peripheral sensitization in tension-type headache. Current
Pain Headache Reports,7, 460465.
Benyamini, Y., & Idler, E.L. (1999). Community studies reporting association between self-
rated health and mortality: Additional studies, 1995 to 1998. Research on Aging,21, 392
401.
18 A.E. Nixon et al.
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
Bruk-Lee, V., & Spector, P.E. (2006). The social stressorscounterproductive work behaviors
link: Are conflicts with supervisors and coworkers the same? Journal of Occupational Health
Psychology,11, 145156.
Cannon, W.B. (1932). The wisdom of the body (2nd ed.). New York: Norton.
Chevalier, A., Souques, M., Coing, F., Dab, W., & Lambrozo, J. (1999). Absenteeism and
mortality of workers exposed to electromagnetic fields in the French electricity company.
Journal of Occupational Medicine,49(8), 517524.
Cohen, S., Kessler, R.C., & Gordon, L.U. (1995). Conceptualizing stress and its relation to
disease. In S. Cohen, R.C. Kessler, & L.U. Gordon (Eds.), Measuring stress: A guide for
health and social scientists (pp. 326). New York: Oxford University Press.
Cooper, C.L., Sloan, S.J., & Williams, S. (1988). Occupational stress indicator. Windsor, UK:
NFERNelson.
Cropley, M., Dijk, D., & Stanley, N. (2006). Job strain, work rumination, and sleep in school
teachers. European Journal of Work and Organizational Psychology,15(2), 181196.
Darr, W., & Johns, G. (2008). Work strain, health, and absenteeism: A meta-analysis. Journal
of Occupational Health Psychology,13, 293318.
Davis, M.C., Zautra, A.J., & Reich, J.W. (2001). Vulnerability to stress among women in chronic
pain from fibromyalgia and osteoarthritis. Annals of Behavioral Medicine,23, 215226.
Demitrack, M., Dale, J., Straus, S., Laue, L., Listwak, S., Kruesi, M. et al. (1991). Evidence for
impaired activation of the hypothalamic-pituitary-adrenal axis in patients with chronic
fatigue syndrome. Journal of Clinical Endocrinology and Metabolism,73, 12241234.
Dienstbier, R.A. (1989). Arousal and physiological toughness: Implications for mental and
physical health. Psychological Review,96,84100.
Dittner, A.J., Wessely, S.C., & Brown, R.G. (2004). The assessment of fatigue A practical
guide for clinicians and researchers. Journal of Psychosomatic Research,56, 157170.
Ertel, K.A., Karestan, K., & Berkman, L.F. (2008). Incorporating home demands into models
of job strain: Findings from the work, family, and health network. Journal of Occupational
and Environmental Medicine,50, 1244 1252.
Eulberg, J.R. (1984). Job effectiveness survey. Unpublished manuscript available from J.R.
Eulberg, Hay Management Consultants, Dallas, TX.
Farre´, R., De Vos, R., Geboes, K., Verbecke, K., Vanden, B.P., Depoortere, I. et al. (2007).
Critical role of stress in increased oesophageal mucosa permeability and dilated intercellular
spaces. Gut,56, 11911197.
Fass, R., Naliboff, B.D., Fass, S.S., Peleg, N., Wendel, C., Malagon, I.B., & Mayer, E.A.
(2008). The effect of auditory stress of perception of intraesophageal acid in patients with
gastroesophageal reflux disease. Gastroenterology,134, 696705.
Ferraro, K.F., & Kelley-Moore, J.A. (2001). Self-rated health and mortality among black and
white adults: Examining the dynamic evaluation thesis. The Journals of Gerontology,56,
195205.
Frankenhaeuser, M. (1991). The psychophysiology of workload, stress, and health: Compar-
ison between the sexes. Annals of Behavioral Medicine,13, 197204.
Franssen, P.M., Bultmann, U., Kant, I., & van Amelsvoort, L.G. (2003). The association
between chronic diseases and fatigue in the working population. Journal of Psychosomatic
Research,54, 339344.
French, J.R., & Raven, B. (1962). The bases of social power. In D. Cartwright (Ed.), Group
dynamics: Research and theory (pp. 607623). Evanston, IL: Row Peterson.
Frese, M., & Zapf, D. (1988). Methodological issues in the study of work stress: Objective vs.
subjective measurement of work stress and the question of longitudinal studies. In C.L.
Cooper & R. Payne (Eds.), Causes, coping and consequences of stress at work (pp. 375411).
Chichester: Wiley.
Frone, M.R. (2000). Work-family conflict and employee psychiatric disorders: The national
comorbidity survey. Journal of Applied Psychology,85, 888895.
Fukuda, K., Straus, S., Hickie, I., Sharpe, M.C., Dobbins, J.G., & Komaroff, A. (1994).
Chronic fatigue syndrome: A comprehensive approach to its definition and study.
Work & Stress 19
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
International Chronic Fatigue Syndrome Study Group. Annals of Internal Medicine,121,
953959.
Goertzel, B.N., Pennachin, C., de Souza, C.L., Maloney, E.M., Jones, J.F., & Gurbaxani, B.
(2006). Allostatic load is associated with symptoms in chronic fatigue syndrome patients.
Pharmacogenomics,7, 485494.
Gura, S.T. (2002). Yoga for stress reduction and injury prevention at work. Work: Journal of
Prevention, Assessment & Rehabilitation,19,37.
Hackman, J.R., & Oldham, G.R. (1975). Development of the Job Diagnostic Survey. Journal
of Applied Psychology,60, 159170.
Hackman, J.R., & Oldham, G.R. (1976). Motivation through design of work: Test of a theory.
Organizational Behavior and Human Performance,16, 250279.
Hunter, J.E., & Schmidt, F.L. (1990). Methods of meta-analysis: Correcting error and bias in
research findings. Newbury Park, CA: Sage.
Huyser, B.A., & Parker, J.C. (1999). Negative affect and pain in arthritis. Rheumatic Disease
Clinics of North America,25, 105121.
Idaszak, J.R., & Drasgow, F. (1987). A revision of the Job Diagnostic Survey: Elimination of a
measurement artifact. Journal of Applied Psychology,72,6974.
Idler, E.L., & Benyamini, Y. (1997). Self-rated health and mortality: A review of twenty-seven
community studies. Journal of Health Social Behavior,38,2137.
Jacquinet-Salord, M.C., Lang, T., Fouriand, C., Nicoulet, I., & Bingham, A. (1993). Sleeping
tablet consumption, self-reported quality of sleep, and working conditions. Group of
Occupational Physicians of APSAT. Journal of Epidemiology & Community Health,47,6468.
James, L.R., Demaree, R.G., & Mulaik, S.A. (1986). A note on validity generalization
procedures. Journal of Applied Psychology,71, 440450.
Jansson, M., & Linton, S.J. (2006). Psychosocial work stressors in the development and
maintenance of insomnia: A prospective study. Journal of Occupational Health Psychology,
11, 241248.
Jex, S.M. (2002). Organizational psychology: A scientist-practitioner approach. New York:
Wiley.
Kandiah, J., Yake, M., & Willett, H. (2008). Effects of stress on eating practices among adults.
Family and Consumer Sciences Research Journal,37,2738.
Karasek, R.A. (1985). Job content questionnaire and user’s guide. Lowell, MA: University of
Massachusetts Lowell, Department of Work Environment.
Kasper, D.L., Braunwald, E., Fauci, A.S., Hauser, S.L., Longo, D.L., & Jameson, J.L. (2005).
Harrison’s principles of internal medicine (16th ed.). New York: McGraw-Hill.
Keck, M.E., & Holsboer, F. (2001). Hyperactivity of CRH neuronal circuits as a target for
therapeutic interventions in affective disorders. Peptides,22, 835844.
Keenan, A., & Newton, T.J. (1985). Stressful events, stressors and psychological strains in
young professional engineers. Journal of Occupational Behavior,6, 151156.
Lee, P.M. (1989). Bayesian statistics: An introduction. New York: Halsed Press.
Linton, S.J., & Bryngelsson, I.L. (2000). Insomnia and its relationship to work and health in a
working-age population. Journal of Occupational Rehabilitation,10(2), 169182.
Liu, J., Garza, J.C., Van Thi, T.H., Henschel, J., Zhang, W., & Lu, X.Y. (2007). The
melanocortinergic pathway is rapidly recruited by acute emotional stress and contributes to
stress-induced anorexia and anxiety-like behavior. Endocrinology,148, 55315540.
Liu, C., Spector, P.E., & Shi, L. (2007). Cross-national job stress: A quantitative and
qualitative study. Journal of Organizational Behavior,28(2), 209239.
Lubbers, R., Loughlin, C., & Zweig, D. (2005). Common Pathways to health and
performance Young workers’job self-efficacy and affect. Journal of Vocational Behavior,
67, 199214.
Manne, S.L., & Zautra, A.J. (1989). Spouse criticism and support: Their association with
coping and psychological adjustment among women with rheumatoid arthritis. Journal of
Personality and Social Psychology,56, 608617.
20 A.E. Nixon et al.
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
McLean, S.A., Williams, D.A., Harris, R.E., Kop, W.J., Groner, K.H., Ambrose, K. et al.
(2005). Momentary relationship between cortisol secretion and symptoms in patients with
fibromyalgia. Arthritis & Rheumatism,52(11), 36603669.
McLean, S.A., Williams, D.A., Stein, P.K., Harris, R.E., Lyden, A.K., Whalen, G. et al.
(2006). Cerebrospinal fluid corticotropin-releasing factor concentration is associated with
pain but not fatigue symptoms in patients with fibromyalgia. Neuropsychopharmacology,31,
27762782.
Miree, L.F. (2007). Financial implications of employee job stress. Research presented at the
Annual Student/Faculty Research Conference, American University of Bulgaria.
Narayanan, L., Menon, S., & Spector, P.E. (1999). Stress in the workplace: A comparison of
gender and occupations. Journal of Organizational Behavior,20(1), 6373.
Ochi, M., Tominaga, K., Tanaka, F., Tanigawa, T., Shiba, M., Watanabe, T. et al. (2008).
Effect of chronic stress on gastric emptying and plasma ghrelin levels in rats. Life Science,
82, 862868.
Peters, L., & O’Connor, E. (1988). Measuring work obstacles: Procedures, issues, and
implications. In F.D. Schoorman & B. Schneider (Eds.), Facilitating work effectiveness
(pp. 105124). Lexington, MA: Lexington Books.
Rau, R., Georgiades, A., Fredrikson, M., Lemne, C., & de Faire U. (2001). Psychosocial work
characteristics and perceived control in relation to cardiovascular rewind at night. Journal of
Occupational Health Psychology, 6, 171181.
Rizzo, J.R., House, R.J., & Lirtzman, S.I. (1970). Role conflict and ambiguity in complex
organizations. Administrative Science Quarterly,14, 150163.
Rosenthal, R. (1991). Meta-analytic procedures for social research. Newbury Park, CA: Sage.
Roth, T., & Ancoli-Israel, S. (1999). Daytime consequences and correlates of insomnia in the
United States: Results of the 1991 National Sleep Foundation Survey. II. Sleep,22, 354
358.
Selye, H. (1936). Thymus and adrenals in response of the organism to injuries and
intoxications. British Journal of Experimental Pathology,17, 234241.
Selye, H. (1976). The stress of life (Rev. ed.). New York: McGraw-Hill.
Smith, M.J., Karsh, B., Carayon, P., & Conway, F.T. (2003). Controlling occupational safety
and health hazards. In J.C. Quick & L.E. Tetrick (Eds.), Handbook of occupational health
psychology (pp. 163189). Washington, DC: American Psychological Association.
Smith, A.K., White, P.D., Aslakson, E., Vollmer-Conna, U., & Rajeevan, M.S. (2006).
Polymorphisms in genes regulating the HPA axis associated with empirically delineated
classes of unexplained chronic fatigue. Pharmacogenomics,7, 387394.
Sparacino, J. (1982). Blood pressure, stress and mental health. Nursing Research,31(2), 8994.
Spector, P.E. (1998). A control model of the job stress process. In C.L. Copper (Ed.), Theories
of Organizational Stress (pp. 153169). London: Oxford University Press.
Spector, P.E., & Jex, S.M. (1998). Development of four self-report measures of job stressors
and strain: Interpersonal conflict at work scale, organizational constraints scale, quantita-
tive workload inventory, and physical symptoms inventory. Journal of Occupational Health
Psychology,3, 356367.
Spector, P.E., & Levine, E.L. (1987). Meta-analysis for integrating study outcomes: A Monte
Carlo study of its susceptibility to type I and type II errors. Journal of Applied Psychology,
72,39.
Spierings, E.L., Ranke, A.H., & Honkoop, P.C. (2001). Precipitating and aggravating factors
of migraine versus tension-type headache. Headache,41, 554558.
Stoller, M. (1994). Economic effects of insomnia. Clinical Therapy,16, 873897.
Taylor, S.E. (1999). What is stress? In S.E. Taylor (Ed.), Health psychology (pp. 168201). New
York: McGraw-Hill.
van der Doef, M., & Maes, S. (1999). The Leiden quality of work questionnaire: Its
construction, factor structure, and psychometric qualities. Psychological Reports,85, 954
962.
Work & Stress 21
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
Wagner, E. (1997). Managed care, and chronic illness: Health services research needs. Health
Services Research,32, 702 714.
Young, A.H., Sharpe, M., Clements, A., Dowling, B., Hawton, K.E., & Cowen, P.J. (1998).
Basal activity of the hypothalamic-pituitary-adrenal axis in patients with the chronic fatigue
syndrome (neurasthenia). Biological Psychiatry,43, 236237.
Zautra, A.J., Hoffman, J.M., Matt, K.S., Yocum, D., Potter, P.T., Castro, W.L. et al. (1998).
An examination of individual differences in the relationship between interpersonal stress
and disease activity among women with rheumatoid arthritis. Arthritis Care & Research,11,
271279.
Zautra, A.J., Johnson, L.M., & Davis, M.C. (2005). Positive affect as a source of resilience for
women in chronic pain. Journal of Consulting and Clinical Psychology,73, 212220.
22 A.E. Nixon et al.
Downloaded By: [University Of South Florida] At: 16:09 27 May 2011
