Job satisfaction and intention to quit the job

Abstract

Background
Negative psychosocial work conditions may influence the motivation of employees to adhere to their job.AimsTo elucidate the perception of psychosocial work conditions among Danish hospital employees who would quit their job if economically possible and those who would not.MethodsA cross-sectional questionnaire study of hospital employees. The questionnaire gave information on elements of the psychosocial work environment (job demands, job influence, job support, management quality, exposure to bullying), general health status, sick-leave during the preceding year, life style (leisure time physical activity, alcohol intake and smoking habits), age, sex and profession.ResultsThere were 1809 participants with a response rate of 65%. About a quarter (26%) reported that they would quit their job if economically possible; this rose to 40% among the 17% who considered their health mediocre or bad. In a final logistic regression model, six factors were identified as independently associated with the wish to quit or not: self-assessed health status, meaningfulness of the job, quality of collaboration among colleagues, age, trustworthiness of closest superior(s) and exposure to bullying. Based on these factors it was possible to identify groups with fewer than 15% wishing to quit, and similarly, groups where 50% or more would quit if this was economically possible.Conclusions
Psychosocial work conditions, in particular meaningfulness of the job, were independently associated with intention to quit the job if economically possible and relevant within different job categories.

Full text

Occupational Medicine 2013;63:96–102
Advance Access publication 30 January 2013 doi:10.1093/occmed/kqs233
© The Author 2013. Published by Oxford University Press on behalf of the Society of Occupational Medicine.
All rights reserved. For Permissions, please email: journals.permissions@oup.com
Job satisfaction and intention to quit thejob
P.Suadicani, J. P.Bonde, K.Olesen and F.Gyntelberg
Department of Occupational and Environmental Medicine, Bispebjerg University Hospital, Copenhagen 2400, Denmark.
Correspondence to: P. Suadicani, Department of Occupational and Environmental Medicine, Bispebjerg University Hospital,
Bispebjerg Bakke 23, Copenhagen 2400, Denmark. Tel: +45 35 31 60 67; fax: +45 35 31 60 70; e-mail: PSUA0001@bbh.regionh.dk
Background Negative psychosocial work conditions may inuence the motivation of employees to adhere to their
job.
Aims To elucidate the perception of psychosocial work conditions among Danish hospital employees who
would quit their job if economically possible and those who would not.
Methods A cross-sectional questionnaire study of hospital employees. The questionnaire gave information on
elements of the psychosocial work environment (job demands, job inuence, job support, manage-
ment quality, exposure to bullying), general health status, sick-leave during the preceding year, life
style (leisure time physical activity, alcohol intake and smoking habits), age, sex and profession.
Results There were 1809 participants with a response rate of 65%. About a quarter (26%) reported that they
would quit their job if economically possible; this rose to 40% among the 17% who considered their
health mediocre or bad. In a nal logistic regression model, six factors were identied as indepen-
dently associated with the wish to quit or not: self-assessed health status, meaningfulness of the job,
quality of collaboration among colleagues, age, trustworthiness of closest superior(s) and exposure
to bullying. Based on these factors it was possible to identify groups with fewer than 15% wishing to
quit, and similarly, groups where 50% or more would quit if this was economically possible.
Conclusions Psychosocial work conditions, in particular meaningfulness of the job, were independently associated
with intention to quit the job if economically possible and relevant within different job categories.
Key words Economy; health status; psychosocial work quality; retirement intention.
Introduction
With ageing populations in developed countries, due to
an increase in life expectancy and low birth rates, the
ratio of young people to old people (the dependency
ratio) is gradually decreasing. Thus the proportion of
gainfully employed people relative to those receiving
money from social security systems and retirement funds
is also decreasing as discussed in a 2008 paper on early
retirement intentions among Belgian nurses [1]. This
may over time be a threat to the very structure of society,
and despite the presently high youth unemployment, the
tendency in European and other western societies has
been to raise the statutory retirement age, and to oth-
erwise encourage people to postpone their age of with-
drawal from the labour market.
Although the literature on the subject of retirement
intention specically is relatively limited, it has been
shown repeatedly that adverse psychosocial workloads
are associated with an increased wish for early retire-
ment [1–10]. In order to identify which occupational
psychosocial factors may be the most important in motiv-
ating people to continue working or the opposite, at least
two conditions may play a major role: health status and
economic necessity. These conditions may heavily inu-
ence the interpretation of studies trying to elucidate the
association of psychosocial work conditions and actual
retirement.
In a recent study on job satisfaction at a Danish
hospital, questions were asked covering occupational
psychosocial elements and factors related to individual
life style (smoking habits, alcohol intake and physical
activity). In the context of the present paper, two
important pieces of information were available: study
participants were able to report on their health status on
a graded scale, and study participants were asked if they
would quit their job if it was economically possible.
We hypothesized that elements of the perceived psy-
chosocial work environment, irrespective of job category,
would be associated with the intention to quit one’s work
if it was economically possible and that the association
could not be explained by health status or sex.
by guest on January 9, 2016http://occmed.oxfordjournals.org/Downloaded from

P. SUADICANI ET AL.: JOB SATISFACTION AND INTENTION TO QUIT THE JOB 97
Methods
Data collection was performed by use of electronic
or paper questionnaires. All employees belong-
ing to the permanent staff in Bispebjerg University
Hospital (BBH) were invited to participate in a survey.
Permanent employment was dened as having the sal-
ary paid monthly. The survey was conducted during the
period from 13 January 2009 to 16 March 2009. If the
employee had not responded within a week, a reminder
was sent. We have no information on how many remind-
ers were sent, since we did not receive the database until
after it was established. However, the authors were fully
informed on how employees had been strongly encour-
aged to participate by the hospital administration prior
to the survey. Access to the full database was allowed
only to the company UNI-C, the Danish IT Centre for
Education and Research. For the analyses, this organ-
ization passed on the database in anonymized form to
the Department of Occupational and Environmental
Medicine at BBH. The questionnaire gave information
on elements of the psychosocial work environment (job
demands, job inuence, job support, management qual-
ity, exposure to bullying), general health status, sick leave
during the preceding year, life style (leisure time physical
activity, alcohol intake and smoking habits), age, gender
and profession of the participants.
The question on intention to quit the job was phrased
as a statement: ‘If Ihad the economic opportunity to do so,
Iwould quit my job.’ Answer options were ‘yes’ and ‘no’.
According to Danish law, ethical approval is not
required for an anonymized questionnairestudy.
Multiple logistic regression analyses were used when
analysing the association of single items with the inten-
tion to quit the job if economically possible, taking into
account age and/or sex (Tables 1–3). For the nal ana-
lysis (Table 4), multiple logistic regression analysis was
used with stepwise backward elimination, allowing vari-
ables with P<0.10, the program default, to remain in
the model. For all analyses we used SPSS for Windows
version 18 (SPSS for Windows, Release 18.0 2009.
Chicago, IL, USA: SPSS Inc.). To ascertain goodness-
of-t we used the Hosmer-Lemeshow test [11].
Results
Altogether, 2802 members of the permanent staff were
drawn from the salary administration database; 1809
individuals returned completed questionnaires, corre-
sponding to an overall response rate of 65%. Analysing
available information on responders and non-responders
showed small differences with respect to age and gen-
der, but craftsmen and unskilled workers and nurses’
aides had a lower response rate than other job categories
(~47%); 1782 employees (99%) had answered the ques-
tion on whether they would quit their job, and 26% (469)
employees reported that they would quit their job if it
was economically possible to do so. The proportions of
employees in the different job categories who would quit
their job if economically possible were as follows: phy-
sicians (17%), administrative personnel (25%), nurses
(23%), nurses’ aides (36%), therapists and technicians
(31%) and craftsmen and unskilled workers (35%).
These rates did not change materially with adjustment
for age andsex.
Table1 shows socio-demographic, lifestyle and health
characteristics of those who would quit their work if it
was economically possible and those who would not.
Those who wanted to quit (quitters) were slightly older
Table1. Socio-demographic, lifestyle and health characteristics of those who would quit their job if economically possible and those who
would not
Would quit Would not quit Age and/or sex-adjusted
n=469 n=1313 OR (95% CI)
Socio-demographic factors
Age 44.9 ± 11.1 42.6 ± 11.3 1.21 (1.10–1.34)
Male sex, % 28 24 1.13 (0.89–1.45)
Lifestyle factors
Current smokers, % 28 26 0.87 (0.66–1.15)
Alcohol, beverages/week, men only 9.3 ± 8.6 7.2 ± 6.9 1.03 (1.01–1.06)
Alcohol, beverages/week, women only 3.8 ± 4.7 3.1 ± 4.4 0.97 (0.94–0.99)
Low leisure time physical activity <4 h/week, % 75 67 1.48 (1.15–1.90)
Health factors
Health considered mediocre/bad, % 27 14 2.25 ( 1.73–2.92)
Sick leave, number of days preceding year 13.2 ± 24.5 9.9 ± 20.6 1.006 (1.001–1.01)
BMI, men only 25.6 ± 3.6 25.1 ± 3.3 1.04 (0.98–1.11)
BMI, women only 24.1 ± 4.1 23.6 ± 4.1 1.03 (0.99–1.06)
OR, odds ratio.
Values presented are mean (standard deviation) or frequency in per cent.
by guest on January 9, 2016http://occmed.oxfordjournals.org/Downloaded from

98 OCCUPATIONAL MEDICINE
than those who would not quit (non quitters). No signi-
cant sex differences were observed.
With respect to life style factors no major differences
were observed between potential quitters and non-quit-
ters, although a statistically signicant difference was
found with respect to leisure time physical activity, with
less physical activity reported and also a slightly higher
consumption of alcohol among potential quitters.
Self-assessed health was signicantly rated as mediocre
or bad more often among quitters, who also had a higher
sickness absence in the preceding year. Finally, potential
quitters had a higher BMI (0.5 units) than non-quitters,
a difference statistically signicant amongwomen.
Table 2 shows ‘psychosocial work characteristics’
of those who would quit their job if economically pos-
sible and those who would not. With respect to work
demands small differences were observed between quit-
ters and non-quitters. Only one item was clearly differ-
ent—whether job demands included coming up with
good ideas always or often; the quitters reported that this
was less frequently necessary, probably reecting low
inuence.
With respect to work control, quitters reported signi-
cantly lower decision latitude than non-quitters.
Regarding work support, three items were clearly sig-
nicantly different between quitters and non quitters:
closest superior being good at solving conicts, good col-
laboration among colleagues and having been exposed
to bullying.
Table 3 shows how ‘quality of management’ was
assessed among those who would quit their job if
economically possible and those who would not. With
respect to assessing quality of management, management
quality was rated lower by quitters than non-quitters
with respect to all items included in thisTable.
Table 4 shows determinants of intention to quit if
economically possible. The multiple logistic regression
analysis included factors from Tables 1–3, which were
signicantly associated with the outcome (would quit/
would not) following logistic regression analysis. Among
Table2. Psychosocial work characteristics of those who would quit their job if economically possible and those who would not
Would quit Would not quit Age and sex-adjusted
n=469 n=1313 OR (95% CI)
Work demands
Quantitative
Is it necessary to work very fast? Always/often, % 63 58 1.17 (0.94–1.46)
Is your work irregular and piling up? Always/often, % 40 33 1.28 (1.02–1.60)
How often are you unable to nish your work tasks? Always/often, % 28 25 1.15 (0.91–1.47)
Cognitive
Does your job demand you to get good ideas? Always/often, % 59 67 0.70 (0.56–0.88)
Emotional
Does your work bring you into emotional situations? Always/often, % 19 14 1.37 (1.03–1.82)
Are you emotionally touched by your work? Always/often, % 18 14 1.32 (0.99–1.77)
Does your work require that you hide your feelings? Always/often, % 22 17 1.32 (1.00–1.73)
Work control
Decision latitude
Do you have much inuence on decisions at your work? Always/often, % 39 49 0.63 (0.50–0.78)
Do you have inuence on your amount of work? Always/often, % 13 18 0.67 (0.49–0.92)
Do you have inuence on your work tasks? Always/often, % 32 42 0.61 (0.49–0.77)
Do you consider your job meaningful? To a high or very high degree, % 71 84 0.46 (0.35–0.59)
Work support
Is your closest superior good at solving conicts? To a high or very high
degree, %
29 38 0.68 (0.54–0.86)
How often do you receive support from other co-workers? Always/often, % 62 66 0.88 (0.70–1.11)
How often do you receive support from your closest superior? Always/often, % 37 44 0.79 (0.63–0.99)
How often do you speak with your colleagues about how well you perform your
work? Always/often, %
19 24 0.80 (0.62–1.05)
How often do you speak with your superiors about how well you perform
your job? Always/often, %
12 17 0.73 (0.54–1.01)
Is the collaboration among colleagues at your work place good? To a high or
very high degree, %
61 73 0.59 (0.47–0.74)
Have you been exposed to bullying within the last year? % yes 12 6 1.86 (1.28–2.69)
Answer options were modied (dichotomized) from questionnaire options:
Frequency of exposure: (1) Always; (2) Often; (3) Sometimes; (4) Rarely; (5) Never/nearly never.
Degree of exposure: (1) To a very high degree; (2) To a high degree; (3) Partly; (4) Little; (5) Very little.
by guest on January 9, 2016http://occmed.oxfordjournals.org/Downloaded from

P. SUADICANI ET AL.: JOB SATISFACTION AND INTENTION TO QUIT THE JOB 99
the 10 factors included in this nal analysis, six remained
in the model following backward elimination of variables.
Taking into account also job categories in the analysis
had no material inuence on the strength of covariates
for intention to quit (not shown). These six factors may,
when added up, be used as an index for determination of
intention to quit the work if economically possible and,
correspondingly, lack of such an intention. The associ-
ation of this index with intention to quit work is illus-
trated in Figures1a,b and 2.
As seen in Figure 1a, including all study partici-
pants among those without adverse psychosocial con-
ditions, fewer than 15% wished to quit, and among
those with four or more adverse conditions, ~50%
wished toquit.
Figure1b shows the result of a similar analysis bro-
ken down according to profession. The largest two
groups with the same basic education within each
group were the nurses and the doctors, so for the
purpose of illustration we present nurses, doctors
and others (comprising nurse’s aides, administrators,
craftsmen, unskilled workers and allied health profes-
sionals (radiographers, laboratory workers, etc.)). The
ability of the identied factors to distinguish between
employees wishing to quit their job and others was
quite similar within even smaller subgroups of job cat-
egories (not shown).
Figure 2 demonstrates the association between psy-
chosocial work conditions index and intention to quit
the job among study participants with self-reported good
health and among those reporting a suboptimal or even
bad health. A graded trend was found for both groups
with a generally more pronounced wish to quit if possible
among those in suboptimal health. Even with a cluster-
ing of good psychosocial work quality (and young age,
i.e. <50years) ~30% in this group would quit compared
with <15% among employees in good health. More than
50% would quit if they had a clustering of three to ve of
the adverse factors, including the age.
Table3. Assessed quality of management among those who would quit their job if economically possible and those who would not
Would quit Would not quit Age and sex-adjusted
n=469 n=1313 OR (95% CI)
Does your closest superior keep his/her promises? To a high or very
high degree, %
46 59 0.57 (0.46–0.71)
Are you treated as an equal independent of your job function, sex,
age, ethnic background, etc.? To a high or very high degree, %
74 85 0.55 (0.42–0.71)
Are the employees treated with respect by their superiors? To a
high or very high degree, %
53 66 0.59 (0.47–0.73)
Do you consider the superiors mentioned below competent to lead within their eld of responsibility? To a high or very high degree, %
Your closest superior 54 67 0.64 (0.51–0.80)
The superior for your area/function 49 65 0.65 (0.53–0.81)
Your department management 53 67 0.59 (0.47–0.73)
Hospital managementa44 61 0.51 (0.38–0.67)
Do you consider the superiors mentioned below good at communicating with the employees? To a high or very high degree, %
Your closest superior 46 58 0.64 (0.52–0.80)
The superior for your area/function 40 52 0.63 (0.50–0.78)
Your department management 37 47 0.66 (0.53–0.82)
Hospital managementa28 39 0.61 (0.45–0.83)
Answer options were modied (dichotomized) from questionnaire options:
Degree of exposure: (1)To a very high degree; (2)To a high degree; (3)Partly; (4)Little; (5)Very little.
aSubjects reporting ‘don’t know’ (~35%) were excluded from the analysis.
Table4. Covariates of intention to quit work if economically
possible versus no intention
Odds ratio (95% CI)
Self-assessed good health versus not 0.49 (0.37–0.65)
Considers work meaningful (to a high/very
high degree versus less)
0.51 (0.39–0.68)
Collaboration among colleagues at the
work place is good (to a high/very high
degree versus less)
0.73 (0.57–0.94)
Age, years
≤30 1a
31–40 1.26 (0.87–1.81)
41–49 1.29 (0.88–1.88)
50+ 1.70 (1.19–2.43)
Closest superior(s) keeps promises (to a
high/very high degree versus less)
0.75 (0.59–0.95)
Exposed to bullying preceding year, no
versus yes
0.68 (0.46–1.02)
Multiple logistic regression analysis including factors from Tables 2–4, which
were signicantly associated with the outcome following logistic regression
analysis.; variables are presented according to strength of association with the
outcome following multivariable adjustment.
Excluded from the nal model, P > 0.10: gender; inuence on work tasks; equal
treatment independent of job function, gender, age, ethnic background, etc.;
communication from closest superior.
aReference category.
by guest on January 9, 2016http://occmed.oxfordjournals.org/Downloaded from

100 OCCUPATIONAL MEDICINE
Discussion
As hypothesized, exposure to perceived psychosocial risk
factors at work was clearly associated with the intention
Figure 1. Prevalence of intention to quit the job if economically pos-
sible according to clustering of covariates identied in Table 4: (1)
self-assessed health; (2) meaningful work; (3) collaboration among
colleagues; (4) age (below or above 50); (5) trustworthiness of closest
superior(s); (6) exposure to bullying.
to quit the job, if this intention was not moderated by eco-
nomic considerations; the associations could only in part
be attributed to health status and gender differences. In
fact, differences between male and female employees were
generally small and not statistically signicant in any of the
analyses. Furthermore, adverse psychosocial exposures
were relevant determinants of intention to quit irrespective
of job category. The psychosocial factors were quite strong
relative potential determinants or covariates of the inten-
tion to quit the job as seen in Table4. When these determi-
nants clustered, the absolute ‘risk’ of reporting an intention
to quit work, increased from approximately 15% among
those without risk factors to more than half of the employ-
ees among those who had ve to six of the covariates. With
respect to the physician group, only 1 of 67 employees
would quit working if none of the adverse conditions were
present. The number of physicians who clustered four to
six adverse conditions was small (n=12). However, eight
(65%) would quit working if economically possible.
Among quantitative work demands (Table 2) only
work assessed as irregular and piling up was associated
with the intention to quit. Among cognitive and emo-
tional work demands all single items were associated with
the intention to quit. Questionnaire items assessing work
control were all signicantly associated with intention
to quit, with a particularly strong association found for
perceived meaningfulness of the job. In contrast, ques-
tionnaire items on work support varied in their strength
of association with intention to quit. Statistically, in this
category, the strongest associations with intention to quit
were found for the questions: ‘is your closest superior
good at solving conicts?’ ‘is the collaboration among col-
leagues good?’ and ‘have you been exposed to bullying?’
All questionnaire items assessing quality of manage-
ment were strongly, and statistically signicantly, associ-
ated with the intention to quit (Table 3). High quality
management was associated with a reduction of nearly
50% of reporting an intention toquit.
The nal model (Table4) may be regarded as the iden-
tication of the most relevant adverse psychosocial work
conditions associated with the intention to quit work
in this study. Apart from age, four occupational factors
were independently associated with intention to quit and
nearly as strongly associated herewith as self-assessed
health status. As illustrated in the gures, on the basis of
these few factors, it was possible to identify groups with
a high inclination to quit if economically possible. In an
additional analysis (not shown) we looked at full-time
employees and part-time employees separately. The clus-
tering of these same factors as shown in Figure1a were
associated with intention to quit in both groups. Among
full-time employees (72% of the population), 11% would
quit if they had no covariates increasing to 52% among
those with ve to six covariates. Among part-time work-
ers the corresponding gures were 24% increasing to
approximately50%.
Figure 2. Prevalence of intention to quit the job if economically pos-
sible according to self-assessed health status and clustering of other
covariates identied in Table 4: (1) meaningful work; (2) collaboration
among colleagues; (3) age (below or above 50); (4) trustworthiness of
closest superior(s); (5) exposure to bullying.
by guest on January 9, 2016http://occmed.oxfordjournals.org/Downloaded from

P. SUADICANI ET AL.: JOB SATISFACTION AND INTENTION TO QUIT THE JOB 101
Assuming that the identied factors are universally
valid among working employees, the factors might be used
as a tool for monitoring the quality of the psychosocial
working environment. Because this method is simple
and cheap, such monitoring could easily be applied on
a regular basis in different workplaces. Furthermore,
several of the factors identied in this study as potential
determinants of intention to quit work have previously
been found relevant also as predictors of sickness absence
[12,13].
A cross-sectional study cannot establish causation
or the direction of associations between exposures and
outcomes, i.e. reverse causality may exist. Employees
answering that they wanted to quit their job may have
certain personality traits or may be negative towards
the work place in general. Accordingly, such employ-
ees may respond more negatively to questions about
their psychosocial work conditions. Thus, there is a
risk of circular argumentation when trying to establish
the direction of causality. However, even so, the asso-
ciations found are relevant for the employees, because
they reect their perception of the psychosocial work
environment. If reverse causation explains much of the
associations found, an improvement of the psychosocial
work environment may have a limited effect. However,
the Finnish longitudinal study by von Bonsdorff et al.
[4] indicated that low satisfaction and negative percep-
tions about work was less pronounced among employees
still working 11years later compared with those retired
either because of disability or because of old age. This,
of course, does not disprove the possibility of reverse
causality in ourstudy.
Another methodological problem in observational
studies may be bias, in particular related to selection,
measurement and confounding. In this study some
selection bias may have inuenced the results because
of non-respondents. The overall response rate was ~65%
but among some groups as low as 47%. The ability of the
identied factors to discriminate between groups with a
high or a low proportion of employees who would quit
their work was, however, strong, irrespective of profes-
sion. However, it cannot be excluded that some selection
bias has inuenced the association between the outcome
and the covariates. At least if the response rate, high or
low, was itself associated with intention to quit the job.
Whether this is the case is unknown.
All information related to psychosocial work condi-
tions, life style and health status were based on self-report
with the inherent risk of some misclassication. Since a
vast majority of Danish hospital employees are able to
give answers to a questionnaire, and that the associations
found between work conditions and intention to quit
were comparable between highly educated study partici-
pants and others, measurement bias must be considered
a minor problem when assessing the perceived psycho-
social work conditions as covariates of intention toquit.
In order to minimize confounding bias and residual
confounding, multivariate analysis was used, taking
into account the correlation between work condi-
tions, the impact of lifestyle and health conditions.
Multivariate analysis in a cross-sectional study, like the
present one, was considered the best method for con-
founder control, since a number of dependent variables
were analysed.
Considering the high impact of health and personal
economic situation for actual retirement, these factors
may blunt the relationship between psychosocial work
environment factors and retirement in observational
studies using actual retirement as an endpoint. So a study
like this using intention to quit one’s job without hindrance
by economic considerations as an endpoint may have
the advantage that the impact of this major confounder
is minimized. However, results from studies on de facto
retirement may identify a broader spectrum of ‘risk
factors’ for retirement, occupational as well as others. For
example, it has been shown that being a current smoker
is a strong predictor of retirement [14]. In the present
study, the intention to quit was practically the same
among smokers and non-smokers. So the hypothetical
question on intention to quit used in this study may be a
cheap and easily applied method as a measurement tool
for not only identifying individuals who would like to
quit their job but also evaluating the psychosocial quality
of a given work or workplace. In addition, a study like
this may identify which of the many psychosocial factors
included have the strongest association with the wish of
quitting one’s job if economically possible.
In the literature addressing determinants of intention
to quit the job, the idea of identifying a few core deter-
minants capable of distinguishing between groups likely
to leave their job and those not so inclined to do so is, to
our knowledge, lacking.
The results of this paper indicate that it may be poss-
ible to screen the overall quality of the psychosocial
working environment by applying the following four
questions: ‘Do you consider your job meaningful?’, ‘Is
the collaboration among colleagues at your work place
good?’, ‘Does your closest superior keep his/her prom-
ises?’ and ‘Have you been exposed to bullying within the
last year?’
Key points
•In a large population of work active hospital
employees approximately one in four would quit
their job if this was economically possible.
•Intention to quit the job was clearly associated with
job category but not with the sex of the employee.
•Perceived suboptimal health and dissatisfaction
with the psychosocial work quality were strongly
associated with the intention to quit.
by guest on January 9, 2016http://occmed.oxfordjournals.org/Downloaded from

102 OCCUPATIONAL MEDICINE
Conicts of interest
None declared.
References
1. Boumans NP, de Jong AH, Vanderlinden L. Determinants
of early retirement intentions among Belgian nurses. J Adv
Nurs 2008;63:64–74.
2. von Bonsdorff ME, Vanhala S, Seitsamo J, Janhonen M,
Husman P. Employee well-being, early-retirement inten-
tions, and company performance. J Occup Environ Med
2010;52:1255–1261.
3. Volanen SM, Suominen S, Lahelma E, Koskenvuo K,
Koskenvuo M, Silventoinen K. Sense of coherence and
intentions to retire early among Finnish women and men.
BMC Public Health 2010;10:22.
4. von Bonsdorff ME, Huuhtanen P, Tuomi K, Seitsamo J.
Predictors of employees’ early retirement intentions: an
11-year longitudinal study. Occup Med (Lond) 2010;60:
94–100.
5. Brett TD, Arnold-Reed DE, Hince DA, Wood IK,
Moorhead RG. Retirement intentions of general prac-
titioners aged 45–65 years. Med J Aust 2009;191:
75–77.
6. Harkonmäki K, Martikainen P, Lahelma E et al.
Intentions to retire, life dissatisfaction and the subse-
quent risk of disability retirement. Scand J Public Health
2009;37:252–259.
7. Heponiemi T, Kouvonen A, Vänskä J etal. Health, psycho-
social factors and retirement intentions among Finnish
physicians. Occup Med (Lond) 2008;58:406–412.
8. Blakeley JA, Ribeiro VE. Early retirement among registered
nurses: contributing factors. J Nurs Manag 2008;16:29–37.
9. Siegrist J, Wahrendorf M, von dem Knesebeck O, Jürges H,
Börsch-Supan A. Quality of work, well-being, and intended
early retirement of older employees: baseline results from
the SHARE Study. Eur J Public Health 2007;17:62–68.
10. Harkonmäki K, Rahkonen O, Martikainen P, Silventoinen
K, Lahelma E. Associations of SF-36 mental health func-
tioning and work and family related factors with inten-
tions to retire early among employees. Occup Environ Med
2006;63:558–563.
11. Hosmer DW, Lemeshow S. Applied Logistic Regression. New
York: Wiley-Interscience Publication, John Wiley & Sons,
1989.
12. Nielsen ML, Rugulies R, Smith-Hansen L, Christensen
KB, Kristensen TS. Psychosocial work environment and
registered absence from work: estimating the etiologic frac-
tion. Am J Ind Med 2006;49:187–196.
13. Nielsen ML, Rugulies R, Christensen KB, Smith-Hansen
L, Bjørner JB, Kristensen TS. Impact of the psychosocial
work environment on registered absence from work: Atwo-
year longitudinal study using the IPAW cohort. Work Stress
2004;18:323–335.
14. Alber tsen K, Lund T, Christensen KB, Kristensen TS,
Villadsen E. Predictors of disability pension over a
10-year period for men and women. Scand J Public Health
2007;35:78–85.
1
EXCELLENCE IN EVIDENCE
1
Quality & speed
Advance Access publication from Oxford Journals
Articles published online ahead of print are available to read and cite
with Advance Access.
www.oxfordjournals.org
by guest on January 9, 2016http://occmed.oxfordjournals.org/Downloaded from