The effectiveness of graded activity in patients with non-specific low-back pain: A systematic review

Abstract

Non-specific low-back pain (LBP) is considered a major health and economic problem in Western society. Nowadays a common used intervention on non-specific LBP is graded activity (GA). Graded Activity developed by Lindström et al., consisted of four parts: (i) measurements of functional capacity; (ii) a work-place visit; (iii) back school education and (iv) an individual, sub-maximal, gradually increased exercise program with an operant-conditioning behavioural approach as described by Fordyce et al.
To evaluate the effectiveness of GA in adults with non-specific LBP on pain, disabilities and return to work.
An extensive literature search of PubMed, Embase, CINAHL and The Cochrane Library was conducted in July 2011.
Randomized controlled trials (RCTs) evaluating the effect of GA in patients with non-specific LBP were eligible. Methodological quality of the studies was assessed according to the PEDro scale. A best-evidence synthesis was conducted according to van Peppen et al. to interpret the outcomes of the included studies.
Ten articles were included in this systematic review; these articles described five RCTs (680 patients). The best-evidence synthesis revealed that there was no or insufficient evidence for a positive effect of GA on pain, disabilities and return to work in patients with non-specific LBP.
Currently there is no or insufficient evidence that GA results in better outcomes of patients with non-specific LBP.

Full text


Disability & Rehabilitation
2012
34
13
1070
1076
© 2012 Informa UK, Ltd.
10.3109/09638288.2011.631682
0963-8288
1464-5165
Disability & Rehabilitation, 2012; 34(13): 1070–1076
© 2012 Informa UK, Ltd.
ISSN 0963-8288 print/ISSN 1464-5165 online
DOI: 10.3109/09638288.2011.631682
Background: Non-specific low-back pain (LBP) is considered
a major health and economic problem in Western society.
Nowadays a common used intervention on non-specific LBP is
graded activity (GA). Graded Activity developed by Lindström
et al., consisted of four parts: (i) measurements of functional
capacity; (ii) a work-place visit; (iii) back school education and (iv)
an individual, sub-maximal, gradually increased exercise program
with an operant-conditioning behavioural approach as described
by Fordyce et al. Objective: To evaluate the effectiveness of GA
in adults with non-specific LBP on pain, disabilities and return
to work. Data sources: An extensive literature search of PubMed,
Embase, CINAHL and The Cochrane Library was conducted
in July 2011. Review Methods: Randomized controlled trials
(RCTs) evaluating the effect of GA in patients with non-specific
LBP were eligible. Methodological quality of the studies was
assessed according to the PEDro scale. A best-evidence synthesis
was conducted according to van Peppen et al. to interpret the
outcomes of the included studies. Results: Ten articles were
included in this systematic review; these articles described five
RCTs (680 patients). The best-evidence synthesis revealed that
there was no or insufficient evidence for a positive effect of GA on
pain, disabilities and return to work in patients with non-specific
LBP. Conclusion: Currently there is no or insufficient evidence that
GA results in better outcomes of patients with non-specific LBP.
Keywords: Behavioural graded activity, graded activity,
low-back pain, physiotherapy, systematic review
Introduction
Background
Low-back pain (LBP) is a major health problem and eco-
nomic problem in Western society [1–4]. LBP has impact
on disability, sickness absence and work disability [5]. In
the Netherlands, the point prevalence of LBP was 26.9% in
1998 [1]. Most of the time LBP is benign and self-limiting [6]
and can be considered as non-specic LBP since no specic
musculoskeletal pathology is found [3,4,6,7].
Nowadays a common used intervention on non-specic
LBP is graded activity (GA [8,9]).
e rationale for the GA approach is that disabilities result-
ing from the presence of pain are not only inuenced by muscu-
loskeletal pathology only, but also by patient beliefs, attitudes,
illness behaviour and psychological distress [10]. erefore,
GA is characterized by operant conditioning as described
by Fordyce et al. [11]. Main features of operant conditioning
are: positive reinforcement of healthy behaviour, consequent
withdrawal of attention towards pain behaviour and its time-
contingent instead of pain-contingent management.
GA developed by Lindström [8] consisted of four parts:
(i) measurements of functional capacity; (ii) a work-place
visit; (iii) back school education and (iv) an individual, sub-
maximal, gradually increased exercise program. An operant-
conditioning behavioural approach was applied as described
REVIEW ARTICLE
The effectiveness of graded activity in patients with non-specific
low-back pain: a systematic review
R. N. van der Giessen1, C. M. Speksnijder1,2,3 & P. J. M. Helders1,3
1Department of Physiotherapy Science, Clinical Health Sciences, Utrecht University, Utrecht, The Netherlands, 2Department
of Oral and Maxillofacial Surgery and Special Dental Care, University Medical Center Utrecht, Utrecht, The Netherlands,
and 3Department of Physical Therapy & Exercise Physiology, University Hospital for Children and Youth ‘Het Wilhelmina
Kinderziekenhuis’, University Medical Center Utrecht, Utrecht, The Netherlands
Correspondence: Dr. C. M. Speksnijder, UMC Utrecht, G05.122, P.O. Box 85090, 3508 AB Utrecht. Tel: +31 88 7568040.
E-mail: C.M.Speksnijder@umcutrecht.nl
(Accepted October 2011)
Graded Activity in non-specic low-back pain
Non-specic low-back pain is not only a physical •
problem but can be inuenced by patient’s beliefs,
psychological distress and illness behaviour.
In clinical practise the use of Graded Activity (GA) •
can be recommend when a discrepancy between mus-
culoskeletal functioning and disabilities are present in
a patient, for instance the presence of kinesiophobia.
ere is no or insucient evidence that GA results in •
better outcomes than usual care.
Implications for Rehabilitation
Disabil Rehabil Downloaded from informahealthcare.com by University Library Utrecht on 06/11/12
For personal use only.

Graded activity in non-specic low-back pain 1071
©  Informa UK, Ltd.
by Fordyce et al. [11], based on the results of the tests and the
demands from the patient’s work.
Changing cognitive factors, behavioral factors and physi-
cal activity by GA may result in defeating non-specic LBP.
Physiotherapy generally aims to aect multiple outcome
domains according to the International Classication of
Functioning, Disability and Health (ICF) classication [12].
erefore limitation of bodily function, disability in daily life
and participation have to be studied [12].
In 2001, a systematic review was published with a focus on
behavioural treatments, including GA, in specic and non-
specic chronic LBP [10]. is review concluded that evidence
for GA was limited, but was based upon only one good quality
trial by Lindström et al. [8,13]. Two Cochrane reviews have been
published and evaluated the eectiveness of GA on LBP [14,15].
e review by Ostelo et al. [14] on behavioral treatments for
chronic LBP concluded: GA is more eective than usual care
(UC), for early return to work and reduced long-term sick leave.
e evidence was limited and based upon one trial; furthermore,
this review was strictly focused on patients with chronic LBP.
e second Cochrane review presented the evidence of exercise
therapy treatments for non-specic LBP [15]. is review pro-
vided moderate evidence for the eectiveness of GA on patients
with sub-acute LBP [15]. In recent years, new clinical trials have
been published evaluating the eectiveness of GA. e eec-
tiveness of GA on acute, sub-acute and chronic LBP has not yet
been determined [10,14,15]. us a synopsis of evidence for
GA is missing. To our knowledge, a systematic review about the
eectiveness of GA in patients with acute, sub-acute or chronic
non-specic LBP has not been published. erefore the objec-
tive of this systematic review was to assess the ecacy of GA as
treatment for non-specic acute, sub-acute and chronic LBP in
adults, on pain, disabilities and rate of return to work.
Methodology
Search strategy
A literature search of PubMed (1950 - 17 July 2011), Embase
(<1950 - 17 July 2011) CINAHL (1982 - 17 July 2011) and
e Cochrane Library (1800 - 17 July 2011) was conducted.
For identifying non-specic LBP patients, the medical subject
headings (MeSH) ‘Back Pain’ and ‘Low-Back Pain’ were used.
A MeSH term for non-specic LBP was absent. Multiple text
words describing GA were used to identify the intervention,
combined with the MeSH ‘Behavior erapy’ and ‘Physical
erapy Modalities’. e Boolean word ‘AND’ connected the
intervention and the patient group. A sensitive lter (sensitivity
99.5% and precision 19.3%) was added for identifying random-
ized clinical trials (RCTs [16]). RCTs were included to possibly
obtain higher quality studies. A hand search was added of the
reference lists of the articles found. e highly sensitive search
strategy for retrieval of reports of RCTs was run in conjunc-
tion with a specic search strategy for non-specic LBP, using
patient, intervention, control, outcome (PICO) elements [17].
Selection criteria
All published RCTs into measuring the eect of GA in patients
with non-specic LBP were eligible for inclusion in this
review. No restrictions were applied to any specic language.
Studies should involve adults (age over 18 years). e type
of intervention included had to be GA, behavioural graded
activity (BGA), or operant conditioning in combination with
time-contingent GA.
Selection of studies
Titles of all identied studies were assessed by two reviewers
(one of whom (R.N.G.) is an author) to determine whether
the studies were pertinent to our research question, according
to the inclusion and exclusion criteria. Aer title selection,
the two reviewers assessed the abstracts to determine whether
the selected studies were eligible for this systematic review.
When there was uncertainty regarding the eligibility of the
paper from reading title and abstract, the full text version of
the paper was retrieved and re-evaluated. e full text version
of all papers that met the inclusion criteria were retrieved for
quality assessment and data extraction.
Methodological study quality
e methodological quality of each included study was deter-
mined using the (PEDro) score, that is specically developed
to rate the methodological quality of RCTs evaluating physio-
therapy interventions [18]. e PEDro scale contains 11 items
for internal (criteria 2–9), statistical (criteria 10 and 11) and
external (criterion 1) validity (Tab l e I). Each item is rated yes or
Table I. Methodological quality according to the PEDro scale.
George [35] Lindström [8,13] Smeets [36–38] Staal [31,32] Hlobil [34] Steenstra [33]
Eligibility criteria 1 1 1 1 1
Random allocation 1 1 1 1 1
Concealed allocation 1 1 1 1 1
Baseline comparability 1 1 1 1 0
Blind subject 0 0 0 0 0
Blind clinician 0 0 0 0 0
Blind assessor 1 1 1 1 1
Adequate follow-up 0 1 1 1 1
Intention-to-treat analysis 1 0 1 1 1
Between-group analysis 1 1 1 1 1
Point estimates and variability 1 1 1 1 1
Total score 7/10 7/10 8/10 8/10 7/10
Disabil Rehabil Downloaded from informahealthcare.com by University Library Utrecht on 06/11/12
For personal use only.

1072 R. N. van der Giessen et al.
Disability & Rehabilitation
no. All ‘yes’ scores were summed resulting in a maximum score
of 10 points. e reliability of the PEDro scale is assessed ‘fair’
to ‘good’ (intraclass correlation coecient (ICC) = 0.68 [19]).
Cohen’s κ was determined as an estimate of agreement between
the two assessors for methodological quality of the studies [20].
When indicated, a third independent reviewer was consulted.
Data extraction
A standardized protocol by van Tulder et al. [21] was used for
data extraction.
Qualitative analyses
Pooling of studies was not possible due to the heterogeneity in
interventions applied, control treatments, outcome measures
and follow-up time. erefore, a best-evidence synthesis was
applied. e criteria set out by van Peppen et al. [22] based
on the methodological quality score of the PEDro scale was
used and modied for RCTs only. Consequently, studies were
categorized into ve levels of evidence: strong evidence, mod-
erate evidence, limited evidence, indicative ndings and no or
insucient evidence (Table II).
Results
Search strategy
e specic literature search in the dierent databases revealed
155 citations in total. Removal of duplicates le 103 articles.
When assessed for eligibility, 85 articles were ineligible. Aer
reviewing 18 articles on abstract and full text, eight articles were
excluded as they did not meet the inclusion criteria. Four articles
did not describe GA [23–26], one article focused on economic
aspects of GA [27], one article was a study protocol [28], one
article was not an RCT [29] and one article did describe neck
pain patients [30]. In conclusion, 10 articles were identied that
met our inclusion criteria [8,13,31–38]. e ow diagram of
article retrieval and selection is presented in Figure 1.
Description of studies
e 10 included articles consisted of ve dierent studies: (i)
George et al. [35] reported on one study, (ii) two articles by
Lindström et al. [8,13] were written about the same study,
but reported on dierent outcome measures, (iii) three arti-
cles by Smeets et al. [36–38] were part of a larger study, con-
sisting direct post-treatment results and long-term results,
(iv) three articles by Staal et al. [31,32] and Hlobil et al. [34]
reported on a similar study and (iv) Steenstra et al. [33]
reported on one study. Study characteristics are presented
in Table III.
Table II. Best-evidence synthesis by van Peppen et al. [22].
Level of evidence Criteria
Strong evidence Provided by statistically signicant ndings in outcome measures in
at least two high-quality RCTs, with PEDro scores of at least 4 pointsa
Moderate evidence Provided by statistically signicant ndings in outcome measures in
at least one high-quality RCT and
at least one low-quality RCT (≤ /3 points on PEDro) or one high-quality CCTa
Limited evidence Provided by statistically signicant ndings in outcome measures in
at least one high-quality RCT or
at least two high-quality CCTsa (in the absence of high-quality RCTs)
Indicative ndings Provided by statistically signicant ndings in outcome measures in at least
one high-quality CCT or low-quality RCTsa (in the absence of high-quality RCTs), or
two studies of a non-experimental nature with sucient quality (in absence of RCTs and CCTs)a
No or insucient
evidence
In the case that results of eligible studies do not meet the criteria for one of the above stated levels of evidence, or
in the case of conicting (statistically signicant positive and statistically signicant negative) results among RCTs and CCTs, or
in the case of no eligible studies
aIf the number of studies that show evidence is <50% of the total number of studies found within the same category of methodological quality and study design (RCT, CCT or
non-experimental studies), no evidence will be classied.
Figure 1. Flow diagram of article retrieval and selection. RCT,
randomized clinical trial; LBP, low back pain; GA, graded activity.
Disabil Rehabil Downloaded from informahealthcare.com by University Library Utrecht on 06/11/12
For personal use only.

Graded activity in non-specic low-back pain 1073
©  Informa UK, Ltd.
Table III. Summary of study characteristics.
Trial NPopulationaExperimental & Control groupbOutcome Follow-up Author’s conclusion Notes
George
et al. [35]
108 Age: 38 Graded
exposure (GA)
Standard care:
Could contain
exercise,
manipulation,
stabilization.
ODQ, NRS 1 month No dierences in reduction of disability, pain intensity and
physical impairments between groups. (χ2 and ANOVA
statistics)
GA (mean nr.
appointments 5.4
(1.9)).
M/F: 39/69 6 months
Duration: 1–24 weeks (TBC) (mean nr.
appointments 6.2
(2.1)).
LBP with or without radiation,
proximal or distal below the knee.
Lindström
et al. [8]
103 Age: 40 GA UC (concisely
described).
Rate of return to
work.
12 months 1 year results: GA was focused on
return to work with a
workplace visit..
M/F: 71/32 24 months GA returned signicant earlier to work (χ2 4.7 p = 0.03). Log
ratio likelihood ratio test male participant in GA returned to
work signicant earlier than women (χ2 6.1, p= 0.01).
Lindström
et al. [13]
Duration: 6 weeks Average sick
leave.
Patients were industrial blue
collar workers
No power analysis.
2 year results:
An average sick leave duration between groups aer 1 year was a
signicant dierence of p = 0.03 was found in favour of GA with
men. No dierence for return to work was found between groups
for female patients, post intervention and 1 year follow-up.
Smeets
et al. [36]
223 Age: 42 GA Waiting list
controls.
RDQ Post-
treatment
6 month
Post-treatment: GA has a signicant positive eect on pain and
RDQ, (p < 0.01), compared to waiting list controls.
Duration of all
interventions was 10
weeks, 3 times a week.
Smeets
et al. [37]
M/F: 93/79 APT: Group
sessions of
aerobic and
strength.
training
Back Pain 12 months No statistically signicant dierences between APT and GA
were found on any outcome measure (data not presented in
article).
Power analysis was
not intended to be
sucient to compare
APT. versus GA.
Duration: >12 weeks 100-mm VAS
Smeets
et al. [38]
Patients with LBP resulting in
disability (RDQ > 3).
PRI
Staal
et al. [31]
134 Age: 38 GA, 1 hour
sessions two
times a week.
UC Return to work,
sick leave due
to LBP
3 months Number of days absent from work because of LBP: Up to 50
days aer randomization: Hazard ratio was 1.0 (95% CI 0.6 to
1.8; p > 0.2). (p = 0.99).
Aer 50 days: Hazard ratio was 1.9 (95% CI 1.2–3.2; p = 0.009),
in favour of GA.
6 months
M/F: 126/8 12 months
Staal
et al. [32] Duration: 6–13 weeks RDQ
KLM workers with LBP
Hlobil [34] Pain intensity 11
point VAS. RDQ: and pain intensity 11 point VAS. No signicant eects
between groups.
Recurrences of sick leave due to non-specic LBP over the 1
year follow-up period was not statistically signicant (p= 0.75).
Total number of sick days; no statistical dierence between
groups was found (p = 0.38).
Steenstra
et al. [33]
112 Age: 42 GA UC Return to work. 3 months Median time until lasting return to work diered signicant
Kaplan-Meier survival calculation (p < 0.01) between GA
(139 days) and UC (111 days) in favour of the UC group, at 3
months. Aer 6 months a signicant (p = 0.03) eect in favour
of the UC group.
M/F: 43/69 6 months
Duration: >8 weeks Roland-Morris
Disability-24
questionnaire.
LBP (ICD-10 codes: M54.5,
M54.4, M54.3, M54.1, M54.8,
and M54.9)
Pain intensity 10
point VAS. No signicant dierences were found between groups at 3
months or 6 months in pain intensity and disability. Only at 6
months pain improvement was statistically signicant in favour
2–6 weeks sick leave due to LBP
aAge: mean age in years; M/F, Male/Female.
bOnly the distinctive therapies are described.
APT, active physical treatment; GA, graded activity; LBP, low-back pain; ODQ, Oswestry Disability Questionnaire; NRS, Numerical Rating Scale for pain intensity; PRI, pain rating index of the McGill Pain Questionnaire; RDQ, Roland Dis-
ability Questionnaire; UC, usual care, guideline of Dutch College of Occupational Practitioners; VAS, visual analogue scale.
Disabil Rehabil Downloaded from informahealthcare.com by University Library Utrecht on 06/11/12
For personal use only.

1074 R. N. van der Giessen et al.
Disability & Rehabilitation
Population characteristics
Each study included more than 100 patients, to a maximum
of 223. In total, 680 patients participated in these studies
when corrected for overlapping patients. e mean age of
the participants in the included studies was 38–42 years. e
duration of complaints diered between studies from 1 to a
maximum of 24 weeks. e percentage of female participation
in the studies ranged from 5.9% [31,32,34] to 64% [35]. e
study of Lindström et al. [8,13] made an analysis between the
eects of male and female patients.
Study interventions
All studies described a time-contingent management and
reported the use of GA as described in the selection criteria of
this review. Consequently every study used operant condition-
ing, measured functional capacity, provided back school edu-
cational and an individual sub-maximal gradually increasing
exercise programme. e durations and the frequencies of the
GA intervention diered. One hour sessions were provided
two times a week by Staal et al. [31,32,34]. e length of this
programme varied between 6 and 13 weeks. e intervention
by Smeets et al. [36–38] consisted of a 10 week programme
with three sessions a week. Lindström et al. [8,13] did not
set a specic duration or frequencies; the GA treatment was
adjusted to the patient’s demands and needs [8,13].
e ‘interventions’ given to the control group varied
between the included studies. ese ‘interventions’ con-
sisted of treatment according to guidelines of occupational
medicine in Staal et al. [31,32,34] and Steenstra et al. [33],
an active physical treatment in George et al. [35] and Smeets
et al. [36–38]. Smeets et al. [36–38] also used waiting list con-
trols as a control group (Table III).
Methodological quality
All studies were of good quality according to the PEDro scale
(Table I) [19]. Cohen’s κ was 0.71 and can be considered as
‘good’ agreement between assessors [20].
Power analysis
George et al. [35] reported that post hoc calculations of
eect sizes indicated that lager sample sizes were required
to detect statistically signicant group dierences. e stud-
ies of Steenstra et al. [33], Staal et al. [32–34] and Smeets
et al. [36–38] mentioned a power analysis, but did not per-
form a power analysis or a post hoc calculation. e study by
Lindström et al. [8,13] study did not mention or calculate a
power analysis.
Qualitative analyses for pain, disabilities and rate of
return to work
George et al. [35], Smeets et al. [36–38], Staal et al. [31,32,34]
and Steenstra et al. [33] reported on the eect of GA on pain
(Table III). George et al. [35]. used a numerical rating scale
and Smeets et al. [36–38], Staal et al. [31,32,34] and Steenstra
et al. [33] used a visual analogue scale (VAS) to measure
pain (Table III). All of these studies, with the exception of
Steenstra [33] reported no statistical signicant dierences
between the control group and GA on the outcome measure
pain. Steenstra’s study [33] reported a signicant dierence at
a 6 months follow-up in favour of the UC control group. In
conclusion, the best-evidence synthesis showed either no or
insucient evidence for the eect of GA on pain in relation
to UC.
George et al. [35], Smeets et al. [36–38], Staal et al. [31,32,34].
and Steenstra et al. [33]. reported on disability due to non-
specic LBP. No statistical signicant dierences were found
between GA and UC. According to the BES, there is no or
insucient evidence for the eect of GA on disabilities.
Lindström et al. [8,13], Staal et al. [31,32,34] and Steenstra
et al. [33] reported on return to work-rate. Lindström et al. [8,13].
reported a signicant dierence (p = 0.03) between the GA
and UC group in favour of GA. However, Steenstra et al. [33].
reported 3 and 6 months statistical signicant ndings (p < 0.01,
p = 0.03) in favour to UC. In addition, Staal et al. [31] reported
a statistical signicant positive eect (p = 0.009) in favour to the
UC group 50 days aer randomization. In conclusion, there is
conicting evidence for the return to work-rate. According to
the BES, there is no or insucient evidence for the eect of GA
on return to work-rate.
Discussion
is systematic literature review presents the found evidence
of the eectiveness of GA in patients with non-specic LBP
when compared to usual care. e best-evidence synthesis
established that there is no or insucient evidence for a posi-
tive eect of GA on pain, disabilities and the return to work-
rate. In this review, 10 articles were included, consisting of ve
dierent RCTs describing the eects of GA [8,13,31–38]. e
included studies were, according to the PEDro scale, of good
methodological quality (Table I). However, some important
comments about the methodological quality can be made. All
studies lacked blinding of participants as well as clinicians.
e studies of Lindström, et al. [8,13], Staal et al. [31,32,34],
Smeets et al. [36–38] did not report power analyses. e stud-
ies of Lindström et al. [8,13], Steenstra et al. [33] and Staal
et al. [31,32,34] did not report the exact UC policy.
e studies diered in study population, administered GA
and UC, as well as in duration of intervention. Consequently,
a meta-analysis could not be performed. erefore the best-
evidence synthesis was a good replacing option [22]. Most
included studies [31–38] used pain as an outcome measure.
is conicts with the rationale of GA. e focus of GA is to
gradually improve activities of patients, so pain is not suit-
able as an outcome measure for the eect of GA. Moreover,
measuring pain may inuence the focus of the patient to pain,
instead of a focus towards function.
Presently, the focus of physiotherapy and occupational
medicine has shied from treating impairments to focus-
sing on function [39–41]. Nowadays, therapies are focussing
on activities and participation according to the ICF [12].
Consequently, during UC, behavioural aspects are most
probably advocated, like a positive reinforcement of healthy
behaviour or a time-contingent management. is could, to
some extent, explain the small dierences between UC and
GA in the included studies.
Disabil Rehabil Downloaded from informahealthcare.com by University Library Utrecht on 06/11/12
For personal use only.

Graded activity in non-specic low-back pain 1075
©  Informa UK, Ltd.
Lindström et al. [8,13] showed a statistical signicant
eect of GA for male patients on return to work-rate, but not
a signicant eect for female patients. Other studies reported
no statistical signicant gender dierences between the inter-
vention and control group for pain, disabilities and return to
work-rate [31–38]. In these studies, however, subgroup analy-
ses for gender were not performed.
Other systematic reviews have been published on the eec-
tiveness of GA [10,14,15]. All these reviews present limited
evidence in favour of GA. Our systematic review presents no
or insucient evidence of GA in patients with LBP. is dier-
ence can be claried; our systematic review specically focuses
on a GA intervention. We did not exclude RCT’s based on the
duration (acute, sub-acute and or chronic) of non-specic LBP
in the study populations. e focus of published systematic
reviews was on behavioural treatments in chronic LBP [10,14]
or exercise therapy treatments in sub-acute LBP [15]. For this
reason, we could include more RCT’s. Furthermore, we have
included recently published RCT’s. Hence, the possible dier-
ences in the included RCT’s and thus the outcomes.
e eectiveness of GA has been evaluated by a RCT in
patients aer rst time lumbar disc surgery [42]. is study
reports no statistical dierences between UC and GA on func-
tional status, pain, pain catastrophizing, fear of movement,
range of motion or return to work-rate aer intervention and
aer 1-year-follow up. e results of our study provided the
same results on pain, disabilities and return to work-rate in non-
specic LBP. e authors conclude that treatment principles as
implemented in GA may not apply to this group [42]. We concur,
the inuences of patient beliefs, attitudes, illness behaviour and
psychological distress and musculoskeletal pathology can result
in the presence of pain and disabilities. Our systematic review
could not report whether the study populations had self-limiting
beliefs, attitudes, illness behaviour and or psychological distress,
as these factors were not objectied in the included studies.
Future studies should investigate underlying mechanisms
of behavioural treatment in non-specic LBP patients. It
would be interesting to examine the eect of GA in patients
with objectied self-limiting beliefs, attitudes, illness behav-
iour or psychological distress [10]. Changing cognitive factors,
behavioural factors and physical activity with GA, may result
in defeating non-specic LBP. is would test the rationale for
the GA approach.
In clinical practise, we can recommend to use GA when
a discrepancy between musculoskeletal functioning and dis-
abilities are present in a patient (for instance the presence
of kinesiophobia). If self-limiting inuences from beliefs,
attitudes, illness behaviour and/or psychological distress in
everyday living are present, GA can be used [10]. However,
when implementing GA, it is recommended to use outcome
measures which evaluate the rationale of GA. Measuring per-
formance or disabilities focuses the patient towards function-
ing and not towards pain. e studies included in this review
used the Oswestry Disability Questionnaire, Roland Disability
Questionnaire or the Roland-Morris Disability questionnaire,
but other instruments focusing on functioning can be used.
Further research on instruments measuring performance
or disabilities in relation to GA can be performed.
In conclusion, there is no or insucient evidence that GA
results in better outcomes than UC. Non-specic LBP is not
only a physical problem, but can be inuenced by patient’s
beliefs, psychological distress and illness behaviour. If one or
more of these factors are present within a patient, it can be
recommended to implement GA.
Acknowledgements
e authors would like to thank J.M. Westeneng for selecting
and assessing the methodological quality of the studies and
suggestions. Furthermore our thanks go to E.M. Buining, for
assisting in the literature search.
Declaration of Interest: e authors report no conicts of
interest.
Reference
1. van Tulder MW, Koes BW, Bouter LM. A cost-of-illness study of back
pain in e Netherlands. Pain 1995;62:233–240.
2. van Tulder M. Conservative non-pharmacological treatment for
chronic low back pain. BMJ 2008;337:a974.
3. Ozguler A, Leclerc A, Landre MF, Pietri-Taleb F, Niedhammer I.
Individual and occupational determinants of low back pain accord-
ing to various denitions of low back pain. J Epidemiol Community
Health 2000;54:215–220.
4. Grith LE, Hogg-Johnson S, Cole DC, Krause N, Hayden J, Burdorf
A, Leclerc A, et al.; Meta-Analysis of Pain in the Lower Back and Work
Exposures (MAPLE) Collaborative Group. Low-back pain denitions
in occupational studies were categorized for a meta-analysis using
Delphi consensus methods. J Clin Epidemiol 2007;60:625–633.
5. Picavet HS, Schouten JS. Musculoskeletal pain in the Netherlands:
Prevalences, consequences and risk groups, the DMC(3)-study. Pain
2003;102:167–178.
6. Urquhart DM, Hoving JL, Assendel WW, Roland M, van Tulder MW.
Antidepressants for non-specic low back pain. Cochrane Database
Syst Rev 2008:CD001703.
7. Cochrane back group, [Internet] Wiley InterScience [cited; December
2008]; Available at: http://mrw.interscience.wiley.com/cochrane/
clabout/articles/BACK/frame.html
8. Lindström I, Ohlund C, Eek C, Wallin L, Peterson LE, Fordyce WE,
Nachemson AL. e eect of graded activity on patients with subacute
low back pain: A randomized prospective clinical study with an oper-
ant-conditioning behavioral approach. Phys er 1992;72:279–90;
discussion 291.
9. Fordyce WE, Fowler RS Jr, Lehmann JF, Delateur BJ, Sand PL,
Trieschmann RB. Operant conditioning in the treatment of chronic
pain. Arch Phys Med Rehabil 1973;54:399–408.
10. van Tulder MW, Ostelo R, Vlaeyen JW, Linton SJ, Morley SJ, Assendel
WJ. Behavioral treatment for chronic low back pain: A systematic
review within the framework of the Cochrane Back Review Group.
Spine 2001;26:270–281.
11. Fordyce WE. Behavioral methods for chronic pain and illness. St
Louis: CV Mosby Co; 1976.
12. World Health Organisation [Internet]. International classication of
functioning: [cited; December 2008]; Available at: www.who.icf.int/
classications/icf/site/icemplate.cfm
13. Lindström I, Ohlund C, Eek C, Wallin L, Peterson LE, Nachemson
A. Mobility, strength, and tness aer a graded activity program for
patients with subacute low back pain. A randomized prospective clini-
cal study with a behavioral therapy approach. Spine 1992;17:641–652.
14. Ostelo RW, van Tulder MW, Vlaeyen JW, et al. Behavioural treatment
for chronic low-back pain. Cochrane Database of Systematic Reviews
2005, Art.1 No.: CD002014.
15. Hayden J, van Tulder MW, Malmivaara A, Koes BW. Exercise therapy
for treatment of non specic low back pain. Cochrane Database of
Systematic Reviews 2008, Art. No.:CD000335.
16. Robinson KA, Dickersin K. Development of a highly sensitive search
strategy for the retrieval of reports of controlled trials using PubMed.
Int J Epidemiol 2002;31:150–153.
Disabil Rehabil Downloaded from informahealthcare.com by University Library Utrecht on 06/11/12
For personal use only.

1076 R. N. van der Giessen et al.
Disability & Rehabilitation
17. Stone P: Popping the (PICO) question in research and evidence based
practice. Appl Nurs Res 16;2008:197–198.
18. Sherrington C, Herbert RD, Maher CG, Moseley AM. PEDro. A data-
base of randomized trials and systematic reviews in physiotherapy.
Man er 2000;5:223–226.
19. Maher CG, Sherrington C, Herbert RD, Moseley AM, Elkins M.
Reliability of the PEDro scale for rating quality of randomized con-
trolled trials. Phys er 2003;83:713–721.
20. Altman DG. Practical statistics for medical research. London, England:
Chapman and Hall; 1991, pp 403–404.
21. van Tulder MW, Touray T, Furlan AD, Solway S, Bouter LM; Cochrane
Back Review Group. Muscle relaxants for nonspecic low back pain:
A systematic review within the framework of the cochrane collabora-
tion. Spine 2003;28:1978–1992.
22. Van Peppen RP, Kwakkel G, Wood-Dauphinee S, Hendriks HJ, Van der
Wees PJ, Dekker J. e impact of physical therapy on functional outcomes
aer stroke: What’s the evidence? Clin Rehabil 2004;18:833–862.
23. Vlaeyen JW, Haazen IW, Schuerman JA, Kole-Snijders AM, van Eek
H. Behavioural rehabilitation of chronic low back pain: Comparison
of an operant treatment, an operant-cognitive treatment and an oper-
ant-respondent treatment. Br J Clin Psychol 1995;34 (Pt 1):95–118.
24. Nicholas MK, Wilson PH, Goyen J. Operant-behavioural and cogni-
tive-behavioural treatment for chronic low back pain. Behav Res er
1991;29:225–238.
25. van den Hout JH, Vlaeyen JW, Heuts PH, Zijlema JH, Wijnen JA.
Secondary prevention of work-related disability in nonspecic low
back pain: Does problem-solving therapy help? A randomized clinical
trial. Clin J Pain 2003;19:87–96.
26. Woods MP, Asmundson GJ. Evaluating the ecacy of graded in vivo
exposure for the treatment of fear in patients with chronic back pain:
A randomized controlled clinical trial. Pain 2008;136:271–280.
27. Hlobil H, Uegaki K, Staal JB, de Bruyne MC, Smid T, van Mechelen W.
Substantial sick-leave costs savings due to a graded activity interven-
tion for workers with non-specic sub-acute low back pain. Eur Spine
J 2007;16:919–924.
28. Macedo LG, Latimer J, Maher CG, Hodges PW, Nicholas M, Tonkin
L, McAuley JH, Staord R. Motor control or graded activity exer-
cises for chronic low back pain? A randomised controlled trial. BMC
Musculoskelet Disord 2008;9:65.
29. Staal JB, Hlobil H, van Tulder MW, Köke AJ, Smid T, van Mechelen
W. Return-to-work interventions for low back pain: A descriptive
review of contents and concepts of working mechanisms. Sports Med
2002;32:251–267.
30. Linton SJ, Boersma K, Jansson M, Svärd L, Botvalde M. e eects
of cognitive-behavioral and physical therapy preventive interventions
on pain-related sick leave: a randomized controlled trial. Clin J Pain
2005;21:109–119.
31. Staal JB, Hlobil H, Twisk JW, Smid T, Köke AJ, van Mechelen W.
Graded activity for low back pain in occupational health care: A ran-
domized, controlled trial. Ann Intern Med 2004;140:77–84.
32. Staal JB, Hlobil H, Köke AJ, Twisk JW, Smid T, van Mechelen W.
Graded activity for workers with low back pain: Who benets most
and how does it work? Arthritis Rheum 2008;59:642–649.
33. Steenstra IA, Anema JR, Bongers PM, de Vet HC, Knol DL, van
Mechelen W. e eectiveness of graded activity for low back pain in
occupational healthcare. Occup Environ Med 2006;63:718–725.
34. Hlobil H, Staal JB, Twisk J, Köke A, Ariëns G, Smid T, van Mechelen W.
e eects of a graded activity intervention for low back pain in occupa-
tional health on sick leave, functional status and pain: 12-month results
of a randomized controlled trial. J Occup Rehabil 2005;15:569–580.
35. George SZ, Zeppieri G Jr, Cere AL, Cere MR, Borut MS, Hodges MJ,
Reed DM, et al. A randomized trial of behavioral physical therapy
interventions for acute and sub-acute low back pain (NCT00373867).
Pain 2008;140:145–157.
36. Smeets RJ, Vlaeyen JW, Hidding A, Kester AD, van der Heijden GJ,
Knottnerus JA. Chronic low back pain: Physical training, graded activ-
ity with problem solving training, or both? e one-year post-treatment
results of a randomized controlled trial. Pain 2008;134:263–276.
37. Smeets RJ, Vlaeyen JW, Kester AD, Knottnerus JA. Reduction of pain
catastrophizing mediates the outcome of both physical and cognitive-
behavioral treatment in chronic low back pain. J Pain 2006;7:261–271.
38. Smeets RJEM, Vlaeyen JWS, Hidding A, Kester ADM, van der
Heijden GJMG, Van Geel ACM, Knottnerus JA: Active rehabilitation
for chronic low back pain: Cognitive-behavioral, physical, or both?
First direct post-treatment results from a randomized controlled trial.
BMC Musuloskel Disord 2006;7:16.
39. Waddell G. 1987 Volvo award in clinical sciences. A new clinical
model for the treatment of low-back pain. Spine 1987;12:632–644.
40. Weiner BK. Spine update: e biopsychosocial model and spine care.
Spine 2008;33:219–223.
41. Engel GL. e need for a new medical model: A challenge for bio-
medicine. Science 1977;196:129–136.
42. Ostelo RW, de Vet HC, Vlaeyen JW, Kerckhos MR, Berfelo WM,
Wolters PM, van den Brandt PA. Behavioral graded activity following
rst-time lumbar disc surgery: 1-year results of a randomized clinical
trial. Spine 2003;28:1757–1765.
Disabil Rehabil Downloaded from informahealthcare.com by University Library Utrecht on 06/11/12
For personal use only.