Full text of DOI:10.1207/S15327558IJBM0704_07

INTERNATIONAL JOURNAL OF BEHAVIORAL MEDICINE, 7(4), 372–392
Copyright © 2000, Lawrence Erlbaum Associates, Inc.
Occupational and Nonoccupational Risk
Indicators for Incident and Chronic Low
Back Pain in a Sample of the Swedish
General Population During a 4-Year
Period: An Influence of Depression?
Carina Bildt, Lars Alfredsson, Hans Michélsen,
Laura Punnett, Eva Vingård, Margareta Torgén,
Arne Öhman, and Åsa Kilbom
The objectives of this study were to examine the relation between occupational and
nonoccupational conditions and both incident (IBLP) and chronic low back pain
(CLBP), as well as depression, among women and men. Data from a 4-year follow-up
of a study group consisting of 420 participants of both genders from the general
Swedish population were analysed. Occupational risk indicators were predictors for
both ILBP and CLBP. Depression was not found to be a predictor for either IBLP or
CLBP among women, but was a predictor for CLBP among men. Among women,
depression had some risk indicators in common with ILBP and appeared to be a con-
current outcome rather than a risk indicator. Low back pain in 1993 was not a predictor
for depression in 1997 in either women or men. Occupational conditions are of rele-
vance in relation to both incident and CLBP. Nonoccupational conditions were stronger
predictors for CLBP than for ILBP but had some predictive value also for ILBP.
Key words: longitudinal, pain, causal and maintaining factors, coping
Carina Bildt, Gender and Work, National Institute for Working Life, Stockholm, Sweden, and Sec-
tion of Psychology, Department of Clinical Neuroscience, Karolinska Institute, Sweden; Lars
Alfredsson, Institute of Environmental Medicine, Karolinska Institute, Sweden; Hans Michélsen, De-
partment of Education, Stockholm University, Sweden; Laura Punnett, Department of Work Environ-
ment, University of Massachusetts–Lowell, Lowell, MA, USA; Eva Vingård, Section of Personal Acci-
dent Prevention, Department of Clinical Neuroscience, Karolinska Institute, Sweden; Margareta
Torgén, Gender and Work, National Institute for Working Life, Stockholm, Sweden; Arne Öhman, Sec-
tion of Psychology, Department of Clinical Neuroscience, Karolinska Institute, Sweden; Åsa Kilbom,
Gender and Work, National Institute for Working Life, Stockholm, Sweden.
Correspondence concerning this article should be sent to Carina Bildt, National Institute for Working
Life, S–112 79 Stockholm, Sweden. E-mail carina.bildt@niwl.se

LOWER BACK PAIN AND DEPRESSION
373
In the general population, low back pain is a common problem. The lifetime inci-
dence of low back pain in the Danish general population was found to be 61% to
80% among women and men in different occupations (Biering-Sørensen, 1983),
and in an earlier population study the lifetime incidence rates of low back pain were
reported to be 51% to 80% (Abrahamson, Terespolsky, & Brook, 1965). Pain in the
lower back often induces constricted movements and can seriously afflict daily life
for stricken individuals. Persistent pain also influences the psychological well-be-
ing of the individual (Esbjörnsson, 1991).
OCCUPATIONAL CONDITIONS AND LOW BACK PAIN
Psychosocial working conditions include the individual’s perception of the con-
tents and organisation of work, as well as of the social relations at work.
Psychosocial factors that have been found to be reliably associated with low back
pain are monotonous work, time pressure, poor job satisfaction, and lack of control
in the working situation (Frank, Kerr, Brooker, DeMaio, Maetzel, Shannon,
Sullivan, Norman, & Wills, 1996a; Vingård, 1999). Several hypotheses about how
psychosocial factors may affect the musculoskeletal system have been proposed.
Psychosocial occupational factors have been assumed to be involved in the devel-
opment of low back pain directly or by increasing the effect of stress or mechanical
load on tissues (Bernard, 1997; Theorell, 1996). Several possible pathways for this
have been suggested. A high level of stress, for example, will lead to tense muscles,
resulting in more vulnerable muscle tissues (Bernard, 1997; Waddell, Newton,
Henderson, Somerville, & Main, 1993). Initial episodes of low back pain elicited
by a mechanical load trigger may induce a chronic dysfunction of the nervous sys-
tem, with associated chronic pain processes (Bernard, 1997). Thus, associations
found between psychosocial occupational factors and low back pain may arise ei-
ther through a causal or an effect-modifying relation.
The physical work environment, that is, mechanical load, is the most studied
part of the working conditions in relation to low back pain. Physical risk factors re-
liably associated with low back pain are heavy lifting, awkward trunk postures, ve-
hicle driving, and whole-body vibrations (Frank et al., 1996a). Physical
occupational factors may result in low back pain through several mechanisms. As-
sociations found between physical occupational factors and low back pain proba-
bly arise through a causal relation.
Nonoccupational Conditions and Low Back Pain
Demanding nonoccupational conditions include the actual psychosocial and physi-
cal demands on the individual from children, elderly relatives, and household tasks.
The family chores are, in general, more burdensome for women than for men

374
BILDT ET AL.
(Josephson, 1999; Lundberg, Mårdberg, & Frankenhauser, 1994). Other factors in-
cluded in nonoccupational conditions are poor social support (instrumental as well
as emotional), conflicts with family members or friends, lack of time for personal
interests, and lack of time for physical and psychological recuperation after work.
Nonoccupational conditions like these have seldom been analysed as potential risk
factors for low back pain, but other nonoccupational factors, such as poor coping
strategies, have been analysed and seem to have a stronger maintaining effect than
occupational factors on chronic low back pain (CLBP; Frank, Kerr, Brooker,
DeMaio, Maetzel, Shannon, Sullivan, Norman, & Wills, 1996b). Nonoccupational
factors, such as physical load outside work and poor social support from friends and
family, are likely to affect the muscles in a way similar to the occupational factors
(Theorell, 1996). Thus, associations found between nonoccupational conditions
andlowbackpainmayariseeitherthroughacausaloraneffect-modifyingrelation.
INCIDENT AND CHRONIC LOW BACK PAIN
It has been proposed that the factors causing the onset of low back pain may be dif-
ferent from the factors that maintain it, that is, cause it to persist after onset (Burton,
Tillotson, Symonds, Burke, & Mathewson, 1996; Frank et al., 1996a, 1996b).
Nonoccupational factors seem to have a stronger maintaining effect than occupa-
tional factors on CLBP (Frank et al., 1996b). In agreement with clinical observa-
tions, personality traits and psychological disorders (especially depression) have
been suggested as maintaining factors for CLBP (Esbjörnsson, 1991; Gentry,
Shows, & Thomas, 1974). Often only nonoccupational factors, such as personality
traits and psychological disorders, have been studied in relation to CLBP, factors
that seldom have been studied in relation to incident low back pain (ILBP).
Persistent pain has been shown to influence the mood, personality functioning,
and adaptive capacity of individuals (Dionne, von Korff, et al., 1998, Esbjörnsson,
1991). Anxiety and worries can often be seen in persons afflicted with acute pain,
and depression is common among persons with long-lasting pain (Esbjörnsson,
1991). Depression has been proposed as a contributing cause for chronic pain syn-
dromes, through increased sensitivity to pain because of influence on the hormonal
activity (Kilbom, Armstrong, Buckle, Fine, Hagberg, Haring-Sweeney, Martin,
Punnet, Silverstein, Sjögaard, Theorell, & Viikari-Juntara, 1996). Thus, depres-
sion may be related to low back pain as a result of the pain or as a contributory
cause for the low back pain, or it may be unrelated but influenced (perhaps caused)
by similar conditions on and off work.
RISK FACTORS FOR DEPRESSION
Because depression may be a parallel outcome to low back pain and therefore may
be influenced by conditions similar to those related to low back pain, the known re-

LOWER BACK PAIN AND DEPRESSION
375
lations between occupational and nonoccupational factors and depression are of
relevance for this article. Both occupational and nonoccupational factors have been
examined as potential risk factors for depression, occupational factors mostly as
potential risk factors for depression among men and nonoccupational factors
mostly as potential risk factors for depression among women (Bildt Thorbjörnsson,
1998). However, occupational factors predicting depression in longitudinal studies
include time pressure, high mental demands, job stress, shift work, few opportuni-
ties to influence working situation, heavy workload, and piecework (Braun & Hol-
lander, 1988; Cooper & Melhuish, 1984, Cooper, Watts, et al., 1987; Eskelinen,
Toikkanen, et al., 1990; Eskelinen, Toikkanen, et al., 1991; Estryn-Behar,
Kaminski, et al., 1990; Goldenhar, Swanson, Hurrell, Ruder, & Deddens, 1998;
Lennon, 1995; Lundberg et al., 1994). These risk factors are very similar to the risk
factors presented for low back pain earlier in this article.
In this study, occupational and nonoccupational conditions were analysed in re-
lation to both incident and CLBP, as was depression. The analysis addressed the
following questions:
1. Are occupational conditions more strongly related to incident than to
chronic low back pain and nonoccupational conditions more strongly re-
lated to chronic than to incident low back pain?
2. Is depression a predictor for low back pain?
3. Are low back pain and depression influenced by similar occupational and
nonoccupational conditions?
4. Is low back pain a predictor for depression?
METHOD
Study Group
The participants were examined on three different occasions (see Figure 1).
In 1969 a survey of approximately 32,000 women and men, ages 18 to 65 years
and living in the county of Stockholm—the REBUS study—was undertaken (first
box to the left in Figure 1). The purpose was to investigate (a) the need for medical
and social services, (b) differences among subgroups of the population in their ac-
tual needs for services, and (c) the steps taken so far to meet these needs (Bygren,
1974). To select the group of participants needed to fulfill the aims of the study in
1969, 32,186 persons from 18 to 65 years of age were selected randomly in an
age-stratified manner where the number of participants selected from the youngest
age groups was enhanced (Theobald, Bygren, Carsten, Hauffman, & Engfeldt,
1998). The enhancement was done to make it possible to study disorders that occur
less often among younger people. Out of these 32,000 participants, 3,064 were se-
lected (based on their responses to a questionnaire with questions about their

376
BILDT ET AL.
FIGURE 1 The baseline examination and the two follow-ups in the REBUS Study; the num-
ber of eligible particpants, the number of participants, and the type of information collected on
the different occasions.
health and possible handicap, and which resulted in division of the participants ac-
cording to their assumed health) to participate in the study. Of these, 2,579 actually
came to participate in the REBUS study in 1969. This final selection was based on
the four groups, and a group-stratified selection was made with enhancement of
participants from the group with suspected very poor health and, to a lesser degree,
from the group with suspected less poor health (approximately 30% and 25%, re-
spectively, of the participants belonged to these groups). These enhancements
were done to make it possible to study disorders that occur less often among the
general population. All 2,579 participants underwent a medical examination, and
medical diagnoses were given whenever appropriate. For a musculoskeletal diag-
nosis, symptoms and signs as well as consequences for daily living were required.
A follow-up study was conducted in 1993 (upper middle box in Figure 1). It was
based on the youngest subgroup, which was between 18 and 34 years of age in 1969
(Bildt-Thorbjörnsson,Alfredsson,etal.,1998).Participantsinvitedtothere-exami-
nation were below 59 years of age in 1993, had not had a musculoskeletal diagnosis
or other diagnoses in 1969 or social circumstances that were thought to decrease
their chances of maintaining employment, and were still living in Sweden. Official
registers were used to trace all eligible participants, and only 22 participants could
not be reached with the request for participation. Of 783 eligible participants, 484
(62%) finally participated in the re-examination. Each participant came to the Insti-
tute for a 1-day examination, including interviews and self-administered question-
naires about physical and psychosocial work conditions and musculoskeletal and
mental problems. They also underwent laboratory tests of physical capacity and a
musculoskeletal examination. The main aim in 1993 was to identify risk indicators
for musculoskeletal disorders; a second aim was to identify risk indicators for poor
mentalhealth.Inthe1993follow-up,backpainduring1970to1992wasstudiedret-
rospectively.Pastandpresentworkandfamilyconditionswerefoundtoberelatedto
musculoskeletal and mental health (Bildt-Thorbjörnsson, Alfredsson, et al., 1998;

LOWER BACK PAIN AND DEPRESSION
377
Bildt-Thorbjörnsson, Alfredsson, Fredriksson, Köster, Michélsen, Vingård,
Torgém, & Kilbom, 2000; Bildt-Thorbjörnsson & Kilbom, 1998; Fredriksson,
Alfredsson, et al., 1998).
In 1997, the participants were again approached and asked to participate in a
second follow-up, which was completely based on self-administered question-
naire data (upper right box in Figure 1). The level of participation was high. Al-
most 87% (88% and 85% among women and men, respectively) of the 484
participants from 1993 participated in the follow-up, resulting in 222 women and
198 men, then between 46 and 63 years of age, in a broad range of occupations.
The main aim in 1997 was to examine the predictive value of the physical and
psychosocial working conditions in 1993 for musculoskeletal disorders in 1997; a
second aim was to examine the predictive value for poor mental health. The partic-
ipants were included in the study after they had been fully informed about all parts
of the study and had given their informed consent to participate. The study was re-
viewed and approved by the Ethics Committee of Human Research at the
Karolinska Institute, Stockholm, Sweden.
The eligible study group in 1993 differed very little from the general population
in Stockholm, regarding income and socioeconomic conditions. Larger differ-
ences were found in the comparison with the general population in Sweden, proba-
bly because the industrial structure in Stockholm differs from the rest of Sweden.
In Stockholm, there is less manufacturing industry and more high-tech industry
than in the rest of Sweden.
Data Collection Procedure
The information about occupational and nonoccupational conditions analysed in
this study was chosen because it has been shown to be related to low back pain or
depression or both in different empirical studies.
Information about occupational factors in 1993 was collected by a self-admin-
istered questionnaire (see Table 1). The responses to the questions about occupa-
tional conditions were dichotomised, and the cut-off points are described in the
appendix. The participants who did not work in 1993 were considered as unex-
posed to all occupational factors.
Information about the quality of social contacts and coping strategies was col-
lected in an interview. The participant’s ability to maintain social relations with
friends and relatives was estimated from an interview by a psychologist. The par-
ticipant was asked to describe his or her relations with friends and relatives, both at
present and in the past. The social ability was categorised according to its quality,
expressed as good, rather good, rather superficial, superficial, or meagre. The
interrater agreement (based on duplicate analyses of tape recordings from 24 inter-
views, percentage agreement) was 85% (Bildt-Thorbjörnsson, et al., 1999). The
participant’s coping strategies in troublesome situations were evaluated in the

378
BILDT ET AL.
TABLE 1
Exposure and Outcome Data Collected in 1993
Occupational factors
Physical Conditions
Sedentary work
Whole body vibrations
Work with hands below knee-level
Bent and twisted body postures
Lifting 5–15 kg
High perceived workload
Psychosocial Conditions
Low occupational pride
Job strain
Few possibilities for on-the-job development
No education at employer’s expense
General Conditions
Full-time work
Overtime work
Shift work
Temporary employment
Nonoccupational factors
Individual Factors
Low frequency of social contacts
Poor quality of social contacts
Smoking
High perceived physical load
Inadequate coping strategies
Earlier Health Problems
Depression 1993
LBP 1993
same interview. Each participant was asked to describe one or several difficult sit-
uations they had experienced, and also to describe how they tried to cope, and to
solve the problems they experienced. The choice of coping strategies was categor-
ised according to its quality, expressed as adequate and mature, adequate but less
mature, adequate but weak, depressive, or schizoid/paranoid. The interrater agree-
ment (based on duplicate analyses of tape recordings from 24 interviews, percent-
age agreement) was 75%. The nonoccupational factors about which information
was collected are listed in Table 1, and the cut-off point for the individual factors is
described in the appendix.
Low back pain during 1970 to 1992 was defined on the basis of the participant
having sought medical consultation and treatment from a doctor, physiotherapist,
or chiropractor because of low back pain at any time during that period. Low back

LOWER BACK PAIN AND DEPRESSION
379
pain in 1993 or in 1997 was defined in terms of medical consultation and treatment
by a doctor, physiotherapist, or chiropractor because of low back pain (question-
naire) or pain in the lower back more than 7 consecutive days during the 12-month
period preceding the examination (data from interview). In this study we defined
the recently developed low back pain as “incident” rather than “acute” because the
time period about which information concerning low back pain in 1997 covered a
12-month period. Cases of ILBP had had low back pain in 1997 but not during
1970 to 1996. Cases of CLBP had had low back pain both in 1993 and in 1997, as
well as pain in the lower back at some time during 1994 to 1996. Healthy partici-
pants had not had low back pain during 1970 to 1997.
Information concerning symptoms of depression during the 12 months preced-
ing the examination in 1993 was obtained from an interview by a psychologist.
The questions posed to the participant during the interview were based on the Di-
agnostic and Statistical Manual of Mental Disorders (third edition, revised;
DSM–III–R; American Psychiatric Association, 1987) and included questions
about lack of interests, lack of energy, sleeping problems, restlessness, poor
self-esteem, poor ability to concentrate, worries, and thoughts of suicide. A diag-
nosis of major depression required at least five symptoms, according to the
DSM–III–R. In this study, subclinical depression was defined as at least two de-
pressive symptoms at any time during the past year. The interrater agreement of
the depression estimation (based on duplicate analyses of tape recordings from 24
interviews, percentage agreement) was 90% (Bildt-Thorbjörnsson et al., 1999). In
1997, interviews based on the DSM–III–R. were not made, because the follow-up
was questionnaire based. Therefore a standardised questionnaire including similar
types of items was used to collect information about depression. Depression was
defined as high values on both of two subscales from the Swedish version of the
Nottingham Life Quality Questionnaire that measure emotional reactions and iso-
lation (Hunt & Wiklund, 1987; Wiklund, 1992). The two subscales include ques-
tions about worries, sleeping problems, lack of control, lonesomeness, feeling
oneself to be a burden to people, and thinking that life is not worth living. Different
symptoms were given different weights, depending on the importance they have
been given by large samples of participants (Wiklund, 1992). The 75th percentile
was used as the cut-off point for depression.
Data Analysis
Both bivariate and multivariate analyses were performed. The participants with
ILBP were compared, as to their exposure, with the healthy participants, as were
the participants with CLBP. The exposure variables that were shown in the age-ad-
justed analyses to be associated with the outcome (where the confidence interval
did not include one) were included, one by one, in the multivariate analyses where
age already had been included.

380
BILDT ET AL.
In the bivariate analyses, age group (46 to 55 and 56 to 63 years) adjusted odds
ratios (ORs) for ILBP and CLBP and for depression associated with exposure to
different occupational and individual factors in 1993 were calculated separately
for women and men (using the module PROC FREQ in the SAS statistical soft-
ware; SAS, 1989). The precision of the point estimates was estimated by
test-based 95% confidence intervals (Miettinen, 1976). In the analyses of associa-
tion between different occupational and nonoccupational factors in 1993 and de-
pression in 1997, the participants with depression in 1993 were excluded.
In the multivariate analyses, ORs for ILBP, CLBP, and depression, adjusted si-
multaneously for age group (46 to 55 and 56 to 63 years), and multiple occupa-
tional and individual factors, were also calculated separately for women and men
(modified Cox regression, using the module PROC PHREG in the SAS statistical
software; SAS 1989). If two factors were highly correlated with each other and
therefore could not be included in the same model, the factor that had the largest
influence on the other included factor parameters was kept in the model. If a factor
had a marked influence on the other included factors, it was kept in the model even
if it did not provide a significant contribution (p ≤ .05).
Missing Participants
REBUS participants who did not wish to participate in 1997 were compared to the
participants regarding low back pain, depression, and occupational and
nonoccupational factors in 1993. The differences in proportions of exposures and
outcomes in 1993 between the two groups were calculated with 95% confidence in-
tervals, using the program CIA (Gardner & Altman, 1989).
RESULTS
The proportion of participants reporting exposure to potential occupational risk
factors in 1993 differed between the genders (see Table 2). More men than women
reported whole-body vibrations, full-time work, and overtime work, and more
women than men reported job strain and few possibilities for development in their
work.
Since 1993, 27 women and 16 men had changed their workplaces. The majority
of them (25 and 13, respectively) had changed to a new job within the same type of
occupation.
The prevalence of ILBP in 1997 was 14% among women and 15% among men,
and the prevalence of CLBP was 13% among women and 10% among men in 1997.
Nine percent of the women and 10% of the men were subclinically depressed in
1997. Twenty percent of the women with ILBP and 4% of the women with CLBP
were depressed. The corresponding proportions among men were 10% and 11%.

LOWER BACK PAIN AND DEPRESSION
381
TABLE 2
Proportions of Subjects (in Percentage) Reporting the Different Occupational
and Individual Factors in 1993 by Gender
Womena
Menb
Occupational factors
Physical conditions
Sedentary work
31
29
15
41
39
34
37
Whole body vibrations
Work with hands below knee-level
Bent and twisted body postures
Lifting 5–15 kg
High perceived workload
General conditions
45
18
35
44
31
Full-time work
Overtime work
Shift work
66
31
20
6
86
44
18
2
Temporary employment
Psychosocial conditions
Low occupational pride
Job strain
Few possibilities for on-the-job development
No education at employer’s expense
Non-occupational factors
Individual factors
19
14
23
51
16
7
13
53
Low frequency of social contacts
Poor quality of social contacts
Smoking
Lack of physical exercise
High perceived load
Inadequate coping strategies
Earlier health problems
Depression 1993
30
21
26
11
27
16
39
26
24
14
22
19
9
37
10
35
LBP 1993
Age distribution in 1997
46–55 years
56–63 years
33
67
38
62
Note. LBP = lower back pain.
an = 222. bn = 198.
Bivariate Analyses: Women
In the age-adjusted analyses, no physical factors were associated with ILBP among
women (see Table 3). One physical factor associated with CLBP among women
was high perceived workload. Shift work was associated with both ILBP and CLBP
and temporary employment with ILBP. Psychosocial factors like low occupational
pride, few possibilities for development in one’s work, job strain, and no education

382

383

384
BILDT ET AL.
at employer’s expense were associated with ILBP, and few possibilities for job de-
velopment and no education at employer’s expense with CLBP. None of the
nonoccupational factors examined were associated with ILBP, but poor quality of
social contacts was associated with CLBP. Depression in 1993 was not related to
either ILBP or CLBP among women.
Work with hands below knee level and bent or twisted body postures were re-
lated to depression among women (see Table 3). Also shift work, lack of education
at employer’s expense, and job strain were associated with depression. Inadequate
coping strategies was also associated with depression, as were poor quality of social
contacts and high perceived physical load outside work.
Bivariate Analyses: Men
Several physical factors were associated with ILBP among men (see Table 4).
Whole-body vibrations, heavy lifting, and high perceived workload were associ-
ated with an increased risk, and sedentary work with a decreased risk, for ILBP.
Work with hands below knee level and heavy lifting were associated with CLBP.
Full-time work was associated with a decreased risk for both ILBP and CLBP.
None of the psychosocial factors was reliably related to ILBP or CLBP among men.
Among men, low frequency of social contacts outside work was associated with
ILBP, and poor quality of social contacts and inadequate coping strategies with
CLBP. Depression in 1993 was found to be associated with CLBP among men.
Few occupational factors were related to depression among men (see Table 4).
Inadequate coping strategies was associated with depression among men, as was
being a smoker.
Multivariate Analysis: Women
When several occupational and nonoccupational factors from 1993 and age were
considered in the same analyses, a small number of factors remained related to
ILBP and CLBP among women (see Table 5). Temporary employment and job
strain were associated with ILBP, and high perceived workload, shift work, and
few possibilities of development on the job were associated with CLBP.
Whensimilaranalysesweremadeinrelationtodepression,onlyjobstrainandin-
adequatecopingstrategieswererelatedtodepressionamongwomen(seeTable5).
Multivariate Analyses: Men
The factors that remained in the multivariate model for ILBP among men were lift-
ing and poor quality of social contacts outside work, and for CLBP, work with
hands below knee level and inadequate coping strategies (see Table 5).

385

386

LOWER BACK PAIN AND DEPRESSION
387
TABLE 5
Adjusted Multivariate Estimatesa of Associations Between Occupational and Individual
Factors in 1993 and Chronic and Incident LBP and Depression in 1997
OR
CI
Women
Incident low back pain
Temporary employment
Job strain
2.8
2.3
1.2, 3.7
1.1, 4.7
Chronic low back pain
High perceived work load
Shift work
Few possibilities for on-the-job development
Depression
1.9
2.3
2.3
1.1, 3.6
1.1, 4.7
1.1, 4.8
Job strain
Inadequate coping strategies
2.0
3.0
1.0, 4.1
1.6, 5.7
Men
Incident low back pain
Lifting 5–15 kg
4.1
2.0
2.1, 8.1
1.1, 4.0
Poor quality of social contacts
Chronic low back pain
Work with hands below knee-level
Inadequate coping strategies
Depression
4.7
3.9
1.9, 12.1
1.5, 10.1
Shift work
Smoking
Inadequate coping strategies
2.1
1.9
3.4
1.1, 4.2
1.0, 3.8
1.8, 6.7
Note. LBP = lower back pain; OR = odds ratio; CI = confidence interval.
aAge is included in the analyses.
Among men, shift work, smoking, and inadequate coping strategies were related
to depression (see Table 5).
Lost-to-Follow-Up Group
The lost-to-follow-up group in this study consisted of only 13% of the study group
from 1993 (30 women and 34 men). More women within the study group than
within the lost-to-follow-up group were temporarily employed in 1993 (9% in the
study group and 0% in the lost-to-follow-up group). Among the male participants,
there were more smokers in the study group than in the lost-to-follow-up group (24
% and 18%). No other significant differences between the participants and the
dropouts were found regarding occupational and nonoccupational conditions, or
regarding low back pain or depression in 1993.

388
BILDT ET AL.
DISCUSSION
Four questions about possible relations between depression, low back pain, and oc-
cupational and nonoccupational conditions were addressed in this study.
Occupational Conditions Related to ILBP and
Nonoccupational Conditions Related to CLBP
Occupational factors were expected to be identified as stronger predictors for ILBP
than for CLBP, and nonoccupational factors as stronger predictors for CLBP than
for ILBP.
In the age-adjusted analyses of occupational factors, there were some differ-
ences between IBLP and CLBP regarding the risk indicator pattern. However, the
main impression was that it was the strength of the observed ORs rather than the
pattern of associations that differed.
In this study, physical factors at work were associated with ILBP or CLBP
among men. Even though psychosocial factors at work were related to both ILBP
and CLBP, physical factors seemed to be the more important factors among men.
The opposite was true for women. This may suggest that there are different mecha-
nisms for the development of low back pain between the two genders. Many of the
psychosocial factors under study can be seen as unspecific demanding factors (i.e.,
factors that relate to several outcomes), rather than specific. Also, many of the fac-
tors from family and leisure time are of the same character. If an individual is ex-
posed to several unspecific demanding factors, sudden exposure to one specific
factor—for example, heavy lifting—may trigger low back pain, although it is the
exposure to the many unspecific factors that has made the individual vulnerable.
Socioeconomic status has been shown to be important in this context (Hallqvist,
1998). A reasonable hypothesis is that gender is another important factor. Alterna-
tively, we may not have collected the relevant information about physical working
conditions for women. Heavy lifting was found to predict both ILBP and CLBP
among men, and additional analysis showed that heavy lifting mainly occurred
within the manufacturing, service, and transportation sectors. Lack of possibilities
for development at work was related to ILBP among women and was observed
mainly in the commerce, service, and transportation sector. High perceived work-
load, related to CLBP, occurred mainly in the health care, commerce, manufacturing,
and service sector.
Depression as a Predictor for Low Back Pain
DepressionwasnotfoundtobeapredictorforeitherILBP orCLBP amongwomen,
butwasapredictorforCLBPamongmen.However,wheninadequatecopingstrate-
gies was included in the multivariate analysis of CLBP among men, depression did

LOWER BACK PAIN AND DEPRESSION
389
notremaininthemodel.Thiswasduetoinadequatecopingstrategiesactingasacon-
founder, that is, it was a risk indicator for CLBP as well as being correlated with de-
pression in 1993. Among women, depression had some risk indicators in common
with ILBP and appeared to be a concurrent outcome rather than a risk indicator.
The relatively low number of participants in this study may be the reason we do
not find support for the hypothesis about a causal relation between depression and
ILBP or CLBP among women, in spite of the longitudinal design of the study. An-
other possible explanation for this may be the study design. Earlier reports of per-
sonality traits or depression as risk factors for CLBP, or other chronic pain
syndromes, have mainly been based on cross-sectional analyses. In this study, no
association between low back pain and depression could be found. Effects of pain
on personality and psychological health may be mistaken as predictors for concur-
rent pain if the design of the study is cross-sectional. When the data from 1997 in
this study was re-analysed with a cross-sectional design, the OR for ILBP in rela-
tion to depression among women was 1.9 (CI = 1.0, 3.4) and for depression in rela-
tion to ILBP 2.5 (CI = 1.0, 6.2). In 1997, no such associations was found among
men, but when depression 1993 was analysed as a potential risk factor for low back
pain in 1993, the OR was 1.6 (CI = 1.0, 2.6).
Low Back Pain and Depression as Influenced by Similar
Conditions
Because of the similarities in risk indicator pattern between ILBP and depression,
especially among women, in the bivariate analyses these outcomes may result from
similar working conditions, which also could explain the lack of support for the
causal hypothesis. For the men in the study, such similarities were not found to the
same extent, which is somewhat intriguing.
Low Back Pain as a Predictor for Depression
Low back pain in 1993 was not a predictor for depression in 1997 in either women
or men. The major portion of the participants we defined as cases of CLBP were not
on long-term or permanent sick leave because of their low back pain, even though
they had suffered from pain for several years. This might be the main explanation
for the lack of association between low back pain and subsequent depression in this
study. It might also be that the chronic case group in this study has included cases
that were recurrences rather than truly chronic (continuous) pain.
Methodological Considerations
The longitudinal design of the study prevented misclassification of exposure be-
cause of ill health at the time that the exposure information was collected, at least

390
BILDT ET AL.
when concerning ILBP and depression. The CLBP cases, of course, had low back
pain already in 1993, which may have influenced their reports of working condi-
tions. The information about psychosocial working conditions, physical working
conditions, nonoccupational factors, and low back pain were collected in separate
interviews or questionnaires, in the hope that it would reduce this source of
misclassification.
Differences in characteristics between the dropouts and the participants may
cause serious bias in an epidemiological study. To be a source of bias, differences
in exposure conditions among participants and dropouts—and such are often
found—must be related to the studied outcome (Criqui, 1979). Such a systematic
bias may lead to under- or overestimation of the ratio estimates found in epidemio-
logical studies. The eligible participants who did not participate in the follow-up in
1997 differed in a few aspects from the participants, but we have no reason to be-
lieve that these differences have affected our results and conclusions. When, in an-
other context, the consequences from these differences were calculated, it was
found that the impact on the ORs were moderate (Bildt-Thorbjörnsson, 1999).
When changes occurred, the ORs increased or decreased only 0.1 or 0.2.
The proportion of subclinically depressed CLBP cases in this study agrees well
with earlier studies (Sullivan, Reesor, Mikail, & Fisher, 1992). Eleven percent of
the women and 20% of the men who were depressed in 1997 had also been de-
pressed in 1993. The criteria for defining depression in 1993 and 1997 differed due
to the different measurement instruments. The questions in the two instruments
were very similar, so we do not think that the use of two different instruments un-
dermines our conclusions. The chosen cut-off point of 75th percentile in the two
subscales of the Nottingham questionnaire resulted in levels well above the mean
value for both women and men in the English reference material (Hunt, McEwen,
& McKenna, 1984). The proportion of depressed participants in 1993 and 1997
was similar, which supports the assumption that the two instruments gave similar
conclusions.
REFERENCES
Abrahamson, J., Terespolsky, L., & Brook, J. (1965). Cornell medical index as a health measure in epi-
demiological studies. A test of the validity of a health questionnaire. British Journal of Preventive
Social Medicine, 19, 103–110.
American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd
ed., rev.). Washington, DC: Author.
Bernard, B. (Ed.). (1997). Musculoskeletal disorders and workplace factors. A critical review of
epidemiologic evidence for work-related musculoskeletal disorders of the neck, upper extremity,
and low back. Cincinnati, OH: U.S. Department of Health and Human Services.
Biering-Sørensen, F. (1983). A prospective study of low back pain in a general population. Scandina-
vian Journal for Rehabilitation Medicine, 15, 71–79.

LOWER BACK PAIN AND DEPRESSION
391
Bildt-Thorbjörnsson, C. (1998). Job stress among women. In Å. Kilbom, K. Messing, & C.
Bildt-Thorbjörnsson (Eds.), Women’s health at work (p. 321). Stockholm: National Institute for
Working Life.
Bildt-Thorbjörnsson, C. (1999). A quarter century perspective on low back pain: A longitudinal study.
Master’s thesis, Arbete och Hälsa, 8.
Bildt-Thorbjörnsson, C., Alfredsson, L., Fredriksson, K., Michélsen, H., Punnett, L., Vingård, E.,
Torgén, M., & Kilbom, Å. (2000). Physical and psychosocial factors related to low back pain during
a 24-year period: A nested case control analysis. Spine, 25, 369–374.
Bildt-Thorbjörnsson, C., & Kilbom, Å. (1998, August). Psychosocial and physical risk factors for de-
pression and reduced mental well-being among women in a longitudinal study: Interaction effects?
PaperpresentedattheFifthInternationalCongressofBehaviorMedicine, Copenhagen, Denmark.
Bildt-Thorbjörnsson, C., Michélsen, H., & Kilbom, Å. (1999). Method for retrospective collecting of
work-related psychosocial risk factors for musculoskeletal disorders: Reliability aggregation. Jour-
nal of Occupational Health Psychology, 4, 193–206.
Bildt-Thorbjörnsson, C. O., Alfredsson, L., Fredriksson, K., Köster, M., Michélsen, H., Vingård, E.,
Torgén, M., & Kilbom, Å. (1998). Psychosocial and physical risk factors associated with low back
pain: A 24-year follow-up among women and men in a broad range of occupations. Occupational
and Environmental Medicine, 55, 84–90.
Braun, S., & Hollander, R. (1988). Work and depression among women in the federal republic of Ger-
many. Women & Health, 14, 3–26.
Burton, K., Tillotson, M., Symonds, T., Burke, C., & Mathhewson, T. (1996). Occupational risk factors
for the first-onset and subsequent course of low back trouble. Spine, 21, 2612–2620.
Bygren, L. O. (1974). Met and unmet needs for medical and social services. Scandinavian Journal for
Social Medicine, (Suppl. 8), 1–134.
Cooper, C., & Melhuish, A. (1984). Executive stress and health. Differences between men and women.
Journal of Occupational Medicine, 26, 99–104.
Cooper, C., Watts, J., & Kelly, M.. (1987). Job satisfaction, mental health, and job stressors among gen-
eral dental practitioners in the UK. British Dentists, 162, 77–81.
Criqui, M. (1979). Response bias and risk ratios in epidemiologic studies. American Journal of Epide-
miology, 4, 394–399.
Esbjörnsson, E. (1991). Acuteandchronicpain—psychologicalaspects. InS. Andersson, S. Arnér, etal.
(Eds.), Acute pain (pp. 63–75). Lund, Studentlitteratur.
Eskelinen, L., Toikkanen, J., et al. (1990). Work-related stress symptoms of ageing employees in munic-
ipal occupation. Scandinavian Journal of Work and Environmental Health, 17(Suppl. 1), 87–93.
Eskelinen, L., Toikkanen, J., Tuomi, K., Mauno, I., Nygård, C-L., Illmarinen, J. (1991). Symptoms of
mental and physical stress in different categories on municipal work. Scandinavian Journal of Work
and Environmental Health, 17(Suppl. 1), 82–86.
Estryn-Behar, M., Kaminski, M., Peigne, E., Bonnet, N., Vaichere, E., Gozlan, C., Azoulay, S., Giorgi,
M. (1990). Stress at work and mental health status among female hospital workers. British Journal
of Industrial Medicine, 4, 20–28.
Frank, J. W., Kerr, M. S., Brooker, A-S., DeMaio, S. E., Maetzel, A., Shannon, H. S., Sullivan, T. J.,
Norman, R. W., & Wills, R. P. (1996a). Disability resulting from occupational low back pain: Part 1.
What do we know about primary prevention? A review of the scientific evidence on prevention be-
fore disability begins. Spine, 21, 2908–2917.
Frank, J. W., Kerr, M. S., Brooker, A-S., DeMaio, S. E., Maetzel, A., Shannon, H. S., Sullivan, T. J.,
Norman, R. W., & Wills, R. P. (1996b). Disability resulting from occupational low back pain: Part 2.
What do we know about secondary prevention? A review of the scientific evidence on prevention af-
ter disability begins? Spine, 21, 2918–2929.

392
BILDT ET AL.
Fredriksson, K., Alfredsson, L., Köster, M., Thorbjörnsson, C., Toomingas, A., Torgén, M., & Kilbom,
Å. (1998). Risk factors for musculoskeletal disorders—a 24 year follow up. Journal of Occupa-
tional and Environmental Medicine, 56, 59–66.
Gardner, M., & Altman, D. (1989). Statistics with confidence—confidence intervals and statistical
guidelines. London, British medical journal.
Gentry, D., Shows, D., & Thomas, M. (1974). Chronic low back pain: A psychological profile.
Psychosomatics, 1574–1577.
Goldenhar, L., Swanson, N., Hurrell, J., Ruder, A., & Deddens, J. (1998). Stressors and adverse out-
comes for female construction workers. Journal of Occupational Health Psychology, 3, 19–32.
Hallqvist, J. (1998). Socioeconomic differences in myocardial infraction risk: Epidemiological analy-
ses of causes and mechanisms. Stockholm: Department for Social Medicine, Karolinska Institute.
Hunt, S., McEwen, J., & McKenna, S. (1984). Perceived health: Age and sex comparisons in a commu-
nity. Journal of Epidemiologic Community Health, 38, 156–160.
Hunt, S., & Wiklund, I. (1987). Cross-cultural variation in the weighting of health statements; a compar-
ison of English and Swedish valuations. Health Policy, 8, 227–235.
Josephson, M. (1999). Work factors and musculoskeletal disorders—an epidemiological approach fo-
cusingonfemalenursingpersonnel. Stockholm:DepartmentofPublicHealth, KarolinskaInstitute.
Kilbom, Å., Armstrong, T., Buckle, P., Fine, L., Hagberg, M., Haring-Sweeney, M., Martin, B., Punnet,
L., Silverstein, B., Sjögaard, G., Theorell, T., & Viikari-Juntara, E. (1996). Musculoskeletal disor-
ders: Work-related risk factors and prevention. Journal of Occupational and Environmental Medi-
cine, 2, 239–246.
Lennon, M. C. (1995). Work conditions as explanations for the relation between socioeconomic status,
gender and psychological disorders. Epidemiologic Reviews, 17, 120–127.
Lundberg, U., Mårdberg, B., & Frankenhauser, M. (1994). The total workload of male and female white
collar workers as related to age, occupational level, and number of children. Scandinavian Journal
of Psychology, 35, 315–327.
Miettinen, O. (1976). Estimability and estimation in case-referent studies. American Journal of
Epidemiol, 103, 226–235.
SAS Institute. (1989). SAS/STAT user’s guide (Version 6, 4th ed.). Cary, NC: Author.
Sullivan, M. J. L., Reesor, K., Mikail, S., & Fisher, R. (1992). The treatment of depression in chronic low
back pain; Review and recommendations. Pain, 50, 5–13.
Theobald, H., Bygren, L.-O., Carstensen, J., Hauffman, M., & Engfeld, P. (1998). Effects of assessment
of needs for medical and social services on long-term mortality: A randomized controlled study. In-
ternational Journal of Epidemiology, 27, 194–198.
Theorell, T. (1996). Possible mechanisms behind the relationship between the demand-control-support
model and disorders of the locomotor system. In S. D. Moon & S. L. Sauter (Eds.), Beyond
biomechanics. Psychosocial aspects of musculoskeletal disorders (pp. 65–74). London: Taylor &
Francis.
Waddell, G., Newton, M., Henderson, I., Somerville, D., & Main, C. (1993). A Fear-Avoidance Beliefs
Questionnarie(FABQ)andtheroleoffear-avoidancebeliefsinchroniclowbackpainanddisability.
Pain, 52, 157–168.
Wiklund, I. (1992). Svensk version av Nottingham Health Profile—ett frågeformulër som mëter
hëlsorelaterad livskvalitet [Swedish version of Nottingham Health Profile—a questionnaire for
measuring health related quality of life]. Göteborg, Germany: Göteborgs Universitet.
Vingård, E. (1999). Lowbackpainanddisordersandoccupationalfactors. Manuscriptinpreparation.