t [ A Psychometric Investigation of Fear-

advertisement
[
research report
]
Steven Z. George, PT, PhD1 • Carolina Valencia, PT2 • Jason M. Beneciuk, PT, DPT, FAAOMPT2
A Psychometric Investigation of FearAvoidance Model Measures in Patients
With Chronic Low Back Pain
T
he fear-avoidance model of musculoskeletal pain (FAM)
has been highlighted as a psychological model of potential
importance for rehabilitation.33 The FAM proposes that during
a musculoskeletal pain episode, anxiety, pain-related fear, and
pain catastrophizing interact to determine whether an individual
will resume normal activities (low psychological distress) or will
t STUDY DESIGN: Validity and test-retest reliability of questionnaires related to the fear-avoidance
model (FAM).
t OBJECTIVE: To investigate test-retest reliability,
construct redundancy, and criterion validity for 4
commonly used FAM measures.
t BACKGROUND: Few studies have reported
psychometric properties for more than 2 FAM
measures within the same cohort, making it difficult
to determine which specific measures should be
implemented in outpatient physical therapy settings.
t METHODS: Fifty-three consecutive patients
(mean age, 44.3  18.5 years) with chronic low
back pain participated in this study. Data were collected with validated measures for FAM constructs,
including the Fear-Avoidance Beliefs Questionnaire
(FABQ), Fear of Pain Questionnaire (FPQ), Tampa
Scale for Kinesiophobia, and Pain Catastrophizing
Scale. Validated measures were used to investigate
criterion validity of the FAM measures, including
the Patient Health Questionnaire for depression,
the numerical rating scale for pain intensity, the
Physical Impairment Scale for physical impairment, and the Oswestry Disability Questionnaire
for self-report of disability. Test-retest reliability of the FAM measures was determined with
intraclass correlation coefficients (ICC2,1) for total
questionnaire scores at baseline and 48 hours
later. Construct redundancy was determined with
Pearson correlation coefficients for FAM measures.
Criterion validity was assessed by 4 separate mul-
tiple regression models that included age, sex, and
employment status as covariates. Depression, pain
intensity, physical impairment, and disability were
the dependent variables for these analyses.
t RESULTS: Test-retest ICC coefficients ranged
from 0.90 to 0.96 for all FAM questionnaires. The
FAM measures were significantly correlated with
each other, with the only exception being that the
FPQ was not correlated with the FABQ work scale.
In the multiple regression models, the Pain Catastrophizing Scale contributed additional variance
to depression. The FABQ physical activity scale
contributed additional variance to pain intensity
and disability. The FABQ work scale contributed
additional variance to physical impairment and
disability. No other FAM measures contributed to
these regression models.
t CONCLUSION: These data suggest that 4
commonly used FAM measures have similar testretest reliability, with varying amounts of construct
redundancy. The criterion validity analyses suggest
that measurement of fear-avoidance constructs
for patients seeking outpatient physical therapy
with chronic low back pain should include the Pain
Catastrophizing Scale and the FABQ. J Orthop
Sports Phys Ther 2010;40(4):197-205. doi:10.2519/
jospt.2010.3298
t KEY WORDS: chronic pain, lumbar spine, pain
catastrophizing
avoid normal activities due to anticipation of increased pain and/or reinjury
(high psychological distress).33,35,60 The
FAM suggests that high psychological distress will be associated with poor
clinical outcomes, potentially resulting
in depressive symptoms, elevated pain
intensity, greater physical impairments,
and continued disability.33,35,60
Evidence supporting validity of the
FAM can be found in the clinical studies of low back pain (LBP). Longitudinal
studies have indicated that elevated FAM
measures were predictive of poor outcomes for individuals with LBP.2,12,14,44,46,52
Furthermore, effective FAM treatment
strategies have been reported, including
patient education,3,5 graded exercise,13,36
and graded exposure.6,20,21,37,58,59 While
portions of the FAM are well supported
in the literature, there are unresolved
questions and issues affecting its application in clinical settings.33
One unresolved issue is related to
measurement of FAM constructs in outpatient physical therapy settings, which
is the focus of this particular study. The
FAM is associated with multiple psychological constructs, and there are
several available measurement tools.
Examples include the Fear-Avoidance
Beliefs Questionnaire (FABQ),61 Fear of
Pain Questionnaire (FPQ),40 ����������
Tampa Scale for Kinesiophobia (TSK),64 and Pain
Catastrophizing Scale (PCS).49 Typically,
psychometric studies incorporate 1 or 2
Associate Professor and Assistant Department Chair, Department of Physical Therapy, University of Florida, Gainesville, FL; Associate Professor, Center for Pain Research and
Behavioral Health, University of Florida, Gainesville, FL. 2 Rehabilitation Science Doctoral Program Student, Department of Physical Therapy, University of Florida, Gainesville,
FL. Steven Z. George (principal investigator) and Carolina Valencia received support from the NIH/NIAMS grant AR051128, and Jason M. Beneciuk received support from the
National Institutes of Health T-32 Neural Plasticity Research Training Fellowship (grant HD043730) while preparing this manuscript. This study was approved by the University of
Florida Institutional Review Board. Address correspondence to Dr Steven Z. George, PO Box 100154, Gainesville, FL 32615. E-mail: szgeorge@phhp.ufl.edu
1 journal of orthopaedic & sports physical therapy | volume 40 | number 4 | april 2010 |
02 George.indd 197
197
3/23/10 5:19:35 PM
[
of these measures in studies of chronic
pain,4,7,23,24,26,28,53,57 with investigations
using multiple FAM measures underreported in the physical therapy literature.
Few studies have investigated multiple
FAM measures in the same outpatient
physical therapy cohort for stability
(test-retest reliability), association with
each other (construct redundancy), and
association with clinical measures (criterion validity).
FAM measures may be very highly
correlated with each other, and the resulting redundancy would suggest that
1 or 2 measures are sufficient to capture
constructs of interest. This scenario was
recently indicated in a study of patients
with LBP that reported very high correlations (r = 0.86) between culturally
adapted versions of the TSK and FABQ.7
In our own research involving patients
with LBP and shoulder pain, we have
seen statistically significant but lowermagnitude correlations between FAM
measures (r0.6),16,18,19 suggesting that
separate measurement may be warranted.
Another potential scenario is that FAM
measures have differing associations with
clinical measures, indicating specificity of
criterion validity. We have observed this
scenario in our own studies. In patients
with LBP, fear of pain was associated with
first-pulse thermal pain sensitivity, while
pain catastrophizing was associated with
temporal summation.19 In subjects with
shoulder pain, fear of pain was predictive of induced pain sensitivity, while pain
catastrophizing was predictive of clinical
pain intensity.16
Additional measurement focused
studies could assist with clinical implementation of the FAM in physical therapy
settings. Therefore, the purpose of this
study was to investigate 4 commonly used
FAM measures in a cohort of patients with
chronic LBP, who were seeking treatment
at outpatient physical therapy clinics. The
goals of this study were 3-fold. First, we
investigated 48-hour test-retest reliability and calculated stability estimates. The
novelty for the current study was that the
same group of patients would complete
research report
4 FAM measures, whereas prior studies
only focused on reliability of 1 or 2 FAM
measures.4,7,23,24,26,28,53,57 Second, we investigated construct redundancy by determining the amount of shared variance
among the 4 FAM measures. Third, we
investigated criterion validity in a multivariate setting by determining FAM
measures’ association with depression,
pain intensity, physical impairment, and
disability.
METHODS
Overview
T
his study was approved by the
University of Florida Institutional
Review Board, and all subjects
provided informed consent before study
participation was confirmed. The cohort
consisted of patients with chronic LBP,
because they would be clinically stable
during the 48-hour test-retest period.
Subjects considered for this study were
recruited during routine clinical visits for
physical therapy.
Subjects
A sample of convenience was recruited
from patients seeking treatment for
LBP at University of Florida-affiliated
outpatient physical therapy clinics. Inclusion criteria were patients aged between 15 and 60 years who had chronic
LBP (greater than 3 months in duration) with or without radiating symptoms. Patients had to have the ability
to read and speak English, because numerous self-report questionnaires were
used. Exclusion criteria were patients
with acute/subacute LBP (less than 3
months in duration), signs of nerve root
compression, lumbar spinal stenosis,
and postlumbar spine surgery. Patients
were also excluded for pregnancy, osteoporosis, and spinal disorders related to
metastatic disease, visceral disease, or
fracture. Eligible patients completed a
standard questionnaire to collect demographic information that included age,
sex, employment status, and prior history of LBP.
]
Measures
FAM Measures Validated questionnaires
commonly used to assess FAM constructs
were used in this study. The FABQ was
used to quantify fear-avoidance beliefs
specific to LBP.61 The FABQ contains 2
scales, a 7-item FABQ work scale (FABQW; range, 0-42) and a 4-item FABQ
physical activity scale (FABQ-PA; range,
0-24). Higher scores indicate higher
levels of fear-avoidance beliefs for both
FABQ scales. The FPQ was used to assess fear of pain to specific activities and
events.1,40,41 A shortened 9-item version
of the FPQ was used (FPQ-9), giving
the scale a total range of 9 to 45, with
higher scores indicating higher fear of
pain. The TSK was used to assess fear
of movement (kinesiophobia) and reinjury.22,43,64 We used a validated 11-item
version (TSK-11), giving the scale a range
of 11 to 44, with higher scores indicating
higher fear of movement and reinjury.64
The PCS was used to assess the extent of
catastrophic cognitions a patient reports
due to LBP.8,49,56 The PCS is a 13-item
scale, with a total range of 0 to 52, and
higher scores are associated with higher
amounts of pain catastrophizing.
Clinical Measures The 4 clinical measures used in this study were consistent
with outcome domains recommended for
chronic pain clinical trials.10 Depression
was assessed as part of the emotional functioning domain.10 Symptoms of depression
were assessed with the Patient Health
Questionnaire (PHQ-9),32 a measure that
has demonstrated validity in different settings.9,27 The 9 items on this questionnaire
come directly from signs and symptoms of
major depression. Each item is rated from
0 (not at all) to 3 (nearly every day), giving the scale a potential range of 0 to 27,32
with higher numbers indicating higher depressive symptoms. The total score PHQ9 score was reported for the purposes of
this study. To account for the pain intensity domain, subjects rated their present
pain intensity using a numerical rating
scale (NRS), ranging from 0 (no pain) to
10 (worst pain imaginable).30
To account for the physical function
198 | april 2010 | volume 40 | number 4 | journal of orthopaedic & sports physical therapy
02 George.indd 198
3/23/10 5:19:36 PM
domain, 2 different measures were completed. First, a physical impairment scale
(PIS) was performed by a physical therapist to quantify physical impairment due
to LBP.62 The PIS consists of 7 different
physical examination procedures performed by the patient, and performance
for each procedure is scored as a negative
(0) or positive (1) for presence of impairment. The PIS has a total range of 0 to 7,
and higher scores indicate higher levels
of physical impairment due to LBP. Second, self-report of disability was assessed
with the modified Oswestry Disability
Questionnaire (ODQ).11 The modified
ODQ has 10 items that assess how LBP
affects common daily activities such as
sitting, standing, and lifting. The ODQ
has a range of 0 (no disability due to back
pain) to 100 (completely disabled due to
back pain), with higher scores indicating
higher disability from LBP.
Sample Size
The goal was to recruit 50 subjects, as this
was the minimum number that would allow for planned test-retest and regression
analyses.31 Specifically, 50 subjects allow
for adequate stability estimates and multivariate regression models to be built with
5 to 10 predictor variables.31 If these guidelines are followed, the planned regression
models will not be overfit and provide
reasonable R2 estimates. This sample size
and resulting range of predictor variables
was adequate for the maximum number
of variables that would be considered in
the criterion validity analysis.
Procedures
Patients meeting eligibility criteria and
providing informed consent completed
questionnaires and then underwent physical examination by a physical therapist
(C.V. or G.Z.) to collect the physical impairment data (PIS). Subjects were then
given a second set of questionnaires to
take home with instructions to complete
the packet 48 hours later and return them
to the study physical therapist (C.V.) by
using the provided, prepaid envelope. The
PIS was not completed a second time.
TABLE 1
Baseline Summary for Subjects
With Chronic Low Back Pain
Variable
Age, y (%)
Value*
44.3 (18.5)
Sex, n (%)
Male
10 (20%)
Female
43 (80%)
Race, n (%)
Caucasian
36 (67%)
African American
16 (31%)
Other
1 (2%)
Employment status
Employed
30 (56%)
Not employed
23 (44%)
FABQ-PA (potential range, 0-24)
13.6 (6.2)
FABQ-W (potential range, 0-42)
13.9 (12.2)
FPQ-9 (potential range, 9-45)
22.9 (7.2)
TSK-11 (potential range, 11-44)
22.6 (6.5)
PCS (potential range, 0-52)
15.0 (12.4)
PHQ-9 (potential range, 0-27)
6.3 (5.7)
NRS (potential range, 0-10)
3.6 (2.4)
PIS (potential range, 0-7)
3.4 (2.2)
ODQ (potential range, 0-100)
27.4 (17.8)
Abbreviations: FABQ-PA, Fear-Avoidance Beliefs Questionnaire physical activity scale; FABQ-W,
Fear-Avoidance Beliefs Questionnaire work scale; FPQ-9, Fear of Pain Questionnaire (9-items); NRS,
numerical rating scale for pain intensity; ODQ, Oswestry Disability Questionnaire; PCS, Pain Catastrophizing Scale; PHQ-9, Patient Health Questionnaire (9-items); PIS, Physical Impairment Scale;
TSK-11, Tampa Scale Kinesiophobia (11 items).
* Values are mean (SD) except where otherwise indicated.
Data Analysis
Descriptive analyses were generated and
reported in the appropriate metric for
continuous and categorical variables.
Test-Retest Reliability Reliability analyses for the 4 FAM measures included
intraclass correlation coefficient (ICC2,1)
for total questionnaire scores obtained
at baseline and 48 hours later. These
results were reported with a 95% confidence interval (CI). From these data,
the standard error of measurement
(SEM) was calculated for each measure
using a previously described method:
SD  1 – (test-retest reliability coefficient)42,47,48 The minimal detectable
change (MDC95) was also calculated using a previously described method (1.96
 SEM  2) to provide a responsiveness
metric.65
Construct Redundancy Construct redundancy analyses included calculation of
Pearson r correlation coefficients among
the 4 FAM measures.
Criterion Validity Criterion validity was
assessed by 4 separate multiple regression models. These models tested FAM
contributions to recommended pain
clinical trial outcome domains, including
emotional function, pain intensity, and
physical function.10 The regression models consisted of a hierarchical portion to
account for age, sex, and employment
status. The regression models included
a stepwise portion to determine which
specific FAM measure had the strongest
statistical association with the dependent
variable of interest. Stepwise regression
was warranted because we had no a priori
hypothesis for FAM measure order of entry, and, as per the purpose of this study,
we wanted to create parsimonious models. Variance inflation factor was reported
with the final models to provide an esti-
journal of orthopaedic & sports physical therapy | volume 40 | number 4 | april 2010 |
02 George.indd 199
199
3/23/10 5:19:37 PM
[
mate of multicollinearity. The 4 separate
regression models had depression (PHQ9), pain intensity (NRS), physical impairment (PIS), and disability (ODQ) as the
dependent variables. In each model, age,
sex, and employment status were included as covariates in the first block of the
model, followed by stepwise consideration of the FAM measures (FABQ-PA,
FABQ-W, PCS, FPQ-9, and TSK).
research report
D
FAM VariableICC2,1
Construct Redundancy
Pearson correlation coefficients are reported in TABLE 3. As expected, the FAM
measures were significantly correlated
with each other, with the only exception being an absence of association between the FPQ-9 and FABQ-W (r = 0.04,
P.05). The variance shared between
FAM measures ranged from 9% (FPQ-9
and TSK-11) to 48% (PCS and TSK-11).
Criterion Validity
The final regression model for depression
scores is reported in TABLE 4. In the hierarchical block, age, sex, and employment
95% CISEM
MDC95
FABQ-PA
0.90
0.82-0.94
2.0
5.4
FABQ-W
0.96
0.92-0.98
2.4
6.8
FPQ-9
0.92
0.85-0.95
2.0
5.6
TSK-11
0.93
0.87-0.96
1.7
4.8
PCS
0.93
0.88-0.96
3.3
9.1
Abbreviations: FABQ-PA, Fear-Avoidance Beliefs Questionnaire physical activity scale; FABQ-W,
Fear-Avoidance Beliefs Questionnaire work scale; FPQ-9, Fear of Pain Questionnaire (9-items); ICC,
intraclass correlation coefficient; MDC, minimal detectable change; PCS, Pain Catastrophizing Scale;
SEM, standard error of measurement; TSK-11, Tampa Scale Kinesiophobia (11-items).
Association Among
Fear-Avoidance Model Measures*
TABLE 3
FABQ-PA
FABQ-W
FPQ-9
TSK-11
PCS
0.45†
0.35‡
0.62†
0.64†
0.04
0.38†
0.45†
0.30‡
0.36†
FABQ-W
FPQ-9
TSK-11
0.69†
Abbreviations: FABQ-PA, Fear-Avoidance Beliefs Questionnaire physical activity scale; FABQ-W, FearAvoidance Beliefs Questionnaire work scale; FPQ-9, Fear of Pain Questionnaire (9-items); PCS, Pain
Catastrophizing Scale; TSK-11, Tampa Scale Kinesiophobia (11 items).
* Pearson correlation coefficients are reported in the table.
†
P.01.
‡
P.05.
Reliability
Reliability results are reported in TABLE 2,
with ICC coefficients ranging from 0.90
to 0.97 for all questionnaires. Stability estimates (eg, SEM and MDC95) are also reported in TABLE 2 for each questionnaire.
The MDC estimates provide a responsiveness metric by indicating how much
change is necessary in a particular questionnaire to exceed measurement error.
Forty-Eight-Hour Test-Retest
Reliability and Stability Estimates
for Fear-Avoidance Model Measures
TABLE 2
RESULTS
escriptive statistics for baseline measures of this cohort are
reported in TABLE 1. Eighty-seven
percent (46/53) of the subjects completed
the 48-hour reliability assessment, with
no differences in key variables for those
that completed and those that did not
complete the reliability follow-up. The
data from those completing the 48-hour
assessment were used for the reliability
analyses, while baseline data for all 53
subjects were used for all other analyses.
]
TABLE 4
Prediction of Depression
With Fear-Avoidance Model Measures*
Beta
P-Value
Age
.24
.02
1.1
Sex†
–.06
.51
1.0
Employment status‡
–.09
.36
1.0
.66
.01
1.2
PCS
VIF
Abbreviations: PCS, Pain Catastrophizing Scale; PHQ-9, Patient Health Questionnaire (9-items).
* Final model for depression scores (PHQ-9): R2 = 0.60, adjusted R2 = 0.56, P.01.
†
Sex coding: 0 = male, 1 = female.
‡
Employment status coding: 0 = employed, 1 = unemployed.
status accounted for 23% of the variance
(P.01) in depression scores. In the stepwise portion, the PCS contributed an additional 37% (P.01) variance in depression
scores. No other FAM measures contributed to the regression model for depression.
The final regression model for pain
intensity ratings is reported in TABLE 5. In
the hierarchical block, age, sex, and em-
ployment status accounted for 22% of the
variance (P.01) in pain intensity ratings.
In the stepwise portion, the FABQ-PA
contributed an additional 23% (P.01)
variance in pain intensity ratings. No
other FAM measures contributed to the
regression model for pain intensity.
The final regression model for physical
impairment is reported in TABLE 6. In the
200 | april 2010 | volume 40 | number 4 | journal of orthopaedic & sports physical therapy
02 George.indd 200
3/23/10 5:19:38 PM
TABLE 5
Prediction of Pain Intensity With
Fear-Avoidance Model Measures*
Beta
P-Value
VIF
Age
.20
.09
1.1
Sex –.02
.89
1.0
Employment status‡
–.22
.06
1.1
.51
.01
1.1
†
FABQ-PA
Abbreviations: FABQ-PA, Fear-Avoidance Beliefs Questionnaire physical activity scale; NRS, numeric
rating scale.
* Final model for pain intensity ratings (NRS): R2 = 0.45, adjusted R2 = 0.40, P.01.
†
Sex coding: 0 = male, 1 = female.
‡
Employment status coding: 0 = employed, 1 = unemployed.
TABLE 6
Prediction of Physical Impairment
With Fear-avoidance Model Measures*
Beta
P-Value
Age
.35
.01
1.1
Sex†
.13
.23
1.0
–.03
.79
1.3
.43
.01
1.4
Employment status‡
FABQ-W
VIF
Abbreviations: FABQ-W, Fear-Avoidance Beliefs Questionnaire work scale; PIS, Physical Impairment
Scale.
* Final model for physical impairment scale scores (PIS): R2 = 0.44, adjusted R2 = 0.40, P.01.
†
Sex coding: 0 = male, 1 = female.
‡
Employment status coding: 0 = employed, 1 = unemployed.
TABLE 7
Prediction of Disability With
Fear-Avoidance Model Measures*
Beta
P-Value
Age
.32
.01
VIF
1.2
Sex†
–.03
.72
1.0
Employment status‡
.06
.57
1.2
FABQ-PA
.36
.01
1.4
FABQ-W
.38
.01
1.8
Abbreviations: FABQ-PA, Fear-Avoidance Beliefs Questionnaire physical activity scale; FABQ-W, FearAvoidance Beliefs Questionnaire work scale; ODQ, Oswestry Disability Questionnaire.
* Final model for disability scores (ODQ ): R2 = 0.63, adjusted R2 = 0.59, P.01.
†
Sex coding: 0 = male, 1 = female.
‡
Employment status coding: 0 = employed, 1 = unemployed.
hierarchical block, age, sex, and employment status accounted for 31% of the variance (P.01) in physical impairment. In
the stepwise portion, the FABQ-W contributed an additional 13% (P.01) variance in
physical impairment scores. No other FAM
measures contributed to the regression
model for physical impairment.
The final regression model for disabili-
ty is reported in TABLE 7. In the hierarchical
block, age, sex, and employment status accounted for 32% of the variance (P.01) in
disability scores. In the stepwise portion,
the FABQ-PA and FABQ-W contributed
additional 23% (P.01) and 8% (P.01)
variance in disability scores, respectively.
No other FAM measures contributed to
the regression model for disability.
DISCUSSION
T
his study investigated testretest reliability, construct redundancy, and criterion validity of
commonly used FAM measures. A novel
aspect of this study was that 4 FAM measures were used in the same cohort of
patients with chronic LBP seeking outpatient physical therapy treatment. Our
results suggested that FAM measures
had similar levels of test-retest stability
and shared low to moderate amounts of
variance with each other. Multivariate
models, including age, sex, and employment status were used to investigate criterion validity. In these models, the PCS
contributed additional variance to depression, the FABQ-PA contributed additional variance to pain intensity, and the
FABQ-W contributed additional variance
to physical impairment. Both the FABQPA and FABQ-W contributed additional
variance to self-report of disability.
It was an expected finding that the selected FAM measures demonstrated high
levels of test-retest reliability, because we
used previously validated measures. Collectively, these results suggest that use of
a particular FAM measure in clinical settings should not be solely based on reliability estimates, as they were all similarly
high (all ICCs, 0.89). Our data were
somewhat unique in that stability was
calculated in the same cohort of patients
and included estimates of measurement
error (SEM and MDC). Prior reliability
studies focused on 1 or 2 measures and
used different patient cohorts, making
direct comparisons between FAM measures difficult.4,7,23,24,26,28,53,57 The previously cited studies also tended to focus on
reporting factor structure and reliability
coefficients, which do not provide a direct metric for measurement error. Measurement error estimates were included
in the study from Grotle et al,23 who reported higher MDCs for the FABQ-PA
(9 points) and FABQ-W (12 points) than
we reported in the current study. Differences in these measurement error
estimates could be because our cohort
journal of orthopaedic & sports physical therapy | volume 40 | number 4 | april 2010 |
02 George.indd 201
201
3/23/10 5:19:38 PM
[
consisted of patients with chronic LBP,
while the Grotle et al23 study investigated
responsiveness of the FABQ in individuals with acute LBP. The data reported in
our study provide estimates of measurement error that can be used to determine
responsiveness for FAM measures. This
information may be helpful in interpreting magnitude of change in studies
that have highlighted the importance of
reducing pain-related fear and/or pain
catastrophizing.15,21,29,45,50,63
Similarly, the construct redundancy
analyses did not provide definitive direction for clinical use of FAM measures.
Statistically significant associations were
reported between most FAM measures,
and they shared a minimum of 9% and
a maximum of 48% variance, suggesting
that overlap of these particular measures
is not an obvious problem. FAM constructs have a close theoretical association with each other,33,60 and the ability
to separately measure these constructs
from each other and other psychological constructs has been questioned in the
literature.25,51 The construct redundancy
results from this study address some
of these theoretical issues. Our results
provide clinical support to the theory
that catastrophizing and fear of pain
are separate constructs. In this sample,
maximum overlap was noted for the PCS
and FABQ-PA (r = 0.64) and the PCS and
TSK-11 (r = 0.69). Fear of pain is theorized to have multiple components, and
our study included measures of several
of these components. Interestingly, the
lowest association among FAM measures
was not between fear-of-pain and paincatastrophizing measures. Instead, the
lowest correlation was between 2 fearof-pain measures, the FPQ-9 and the
FABQ-W (r = 0.04). This specific finding
suggests that fear of specific activities is
not associated with fear and beliefs about
performing work activities while experiencing LBP. Collectively, these construct
redundancy results suggest that convergence is common among FAM measures,
but the amount of convergence potentially varies greatly. Therefore, separation of
research report
these constructs by clinical measurement
may be dependent on the specific questionnaire used.
The multivariate regression analyses involving chronic pain outcome domains provided novel information on
criterion validity for FAM measures in
outpatient physical therapy settings. In
these analyses, we included age, sex, and
employment status as covariates, then
allowed for stepwise consideration of
FAM measures. Stepwise regression was
used because we had no a priori hypotheses about specific questionnaires and
wanted to create parsimonious models
to guide clinical recommendations. The
first model investigated emotional functioning, measured by depression in this
study. The PCS contributed additional
variance after considering age, sex, and
employment status. Pain catastrophizing
has been linked to depression both theoretically51,54 and empirically,25,29,55 so its
association with depression was expected. It was somewhat surprising that none
of the fear-of-pain measures contributed
to depression, given that depression
and anxiety are known to be very highly
correlated.38
The other regression models investigated pain intensity, physical impairment, and self-report of disability. In
these analyses, the FABQ was consistently the only FAM measure entered into
the multivariate regression model after
age, sex, and employment status were
considered. Specifically, the FABQ-PA
contributed additional variance in pain
intensity and disability, while the FABQW contributed additional variance to
physical impairment and disability. The
association of fear of pain with pain and
disability is not a novel finding in itself.
However, it is a novel finding that neither the other fear measures (TSK-11
and FPQ-9) nor the pain-catastrophizing
measure (PCS) contributed to these regression models.
These analyses have potential implications for those measuring FAM constructs in outpatient physical therapy
settings. There are many measurement
]
options available to assess FAM constructs, and our results indicated that
these measures performed in a similar
fashion in test-retest reliability. That is,
all of the selected FAM measures had
stability estimates consistent with use in
clinical settings for patients with chronic
LBP. Furthermore, all FAM measures
were susceptible to varying extent of
construct redundancy with each other.
The lone exception to the construct redundancy trend was for the FPQ-9, a
measure that is not often used clinically,
based on our experiences. Therefore,
our clinical recommendations are derived from the criterion validity analyses.
These results suggested that assessment
of the FAM for patients with chronic LBP
should include the PCS and the FABQ.
Specifically, the PCS should be used for
those interested in determining the fearavoidance influence on emotional function, while the FABQ should be used for
determining the fear-avoidance on pain
intensity and physical function.
Several limitations should be considered when interpreting the results
of this study. First, the validity analyses
were cross-sectional, and our data cannot be used to make conclusions about
the predictive validity of these measures.
Future studies that incorporate longitudinal designs will determine which FAM
measures are predictive of outcomes of
interest. Second, this study recruited
patients who were already considered
to have chronic LBP, so these data cannot be used to make conclusions about
the development of chronic LBP. Future
studies that recruit an inception cohort
of patients with acute LBP will determine if elevated pain-related fear and
pain catastrophizing are precursors to
the development of chronic LBP. Third,
this study had a relatively small sample
size. The sample size was sufficient to investigate our stated purposes. However,
there is a chance of type II error with our
sample size and it should be noted that
we are potentially underestimating the
influence of the other FAM measures.
Type II error does not seem to be a con-
202 | april 2010 | volume 40 | number 4 | journal of orthopaedic & sports physical therapy
02 George.indd 202
3/23/10 5:19:39 PM
cern with our models for depression and
disability, as the variance explained was
60% and 63%, respectively. However,
the models for pain intensity and physical impairment had only 45% and 44%
variance explained, respectively. We acknowledge that these regression models
in particular may have included other
FAM measures if a larger sample was recruited. Future prospective studies with
larger sample sizes will be able to draw
more definitive conclusions about measurement of pain intensity and physical
impairment. Fourth, the associations
reported in this study are specific to the
clinical measures used in this study, and
the same influences cannot be assumed
for other measures of depression, pain intensity, physical impairment, or disability. Fifth, while we included 4 commonly
used FAM measures in this study, this
was not a comprehensive investigation
of FAM measures. Most notably, we did
not include the Pain Anxiety Symptoms
Scale.39 We also did not include recently
proposed measures that assess fear related to specific activities, such as the Fear
of Daily Activities Questionnaire17 and
the Photograph Series of Daily Activities
short electronic version. 34
CONCLUSION
T
his study recruited subjects
with chronic LBP who were seeking outpatient physical therapy
interventions, to investigate test-retest
reliability, construct redundancy, and
criterion validity for 4 commonly used
FAM measures. The results suggest that
FAM measures had similar levels of stability and shared varying amounts of
variance with each other. Multivariate
models controlled for age, sex, and employment status indicated that only the
PCS contributed additional variance to
depression. The FABQ contributed additional variance to pain intensity, physical
impairment, and disability in the multivariate models. Therefore, these data
suggest that the PCS and FABQ should
be used to assess FAM constructs for pa-
tients with chronic LBP seeking outpatient physical therapy. t
KEY POINTS
FINDINGS: Four commonly used FAM
measures had similar test-retest stability
estimates and shared low to moderate
amounts of variance with each other.
The PCS was associated with depression, and the FABQ was associated with
pain intensity, physical impairment, and
disability.
IMPLICATIONS: Assessment of FAM constructs for patients with chronic LBP in
physical therapy settings should include
the PCS for emotional function and the
FABQ for pain intensity and physical
functioning.
CAUTION: The results of the present study
cannot be used to make clinical decisions about the predictive validity of any
of these FAM measures.
ACKNOWLEDGeMENTS: Giorgio Zeppieri, Jr as-
sisted with data collection.
references
1. A
lbaret MC, Munoz Sastre MT, Cottencin A,
Mullet E. The Fear of Pain questionnaire: factor
structure in samples of young, middle-aged
and elderly European people. Eur J Pain.
2004;8:273-281. http://dx.doi.org/10.1016/j.
ejpain.2003.09.005
2. Al-Obaidi SM, Beattie P, Al-Zoabi B, Al-Wekeel
S. The relationship of anticipated pain and fear
avoidance beliefs to outcome in patients with
chronic low back pain who are not receiving
workers’ compensation. Spine (Phila Pa 1976).
2005;30:1051-1057.
3. Burton AK, Waddell G, Tillotson KM, Summerton
N. Information and advice to patients with back
pain can have a positive effect. A randomized
controlled trial of a novel educational booklet in
primary care. Spine. 1999;24:2484-2491.
4. Cleland JA, Fritz JM, Childs JD. Psychometric
properties of the Fear-Avoidance Beliefs Questionnaire and Tampa Scale of Kinesiophobia in
patients with neck pain. Am J Phys Med Rehabil.
2008;87:109-117. http://dx.doi.org/10.1097/
PHM.0b013e31815b61f1
5. Coudeyre E, Tubach F, Rannou F, et al. Effect of a
simple information booklet on pain persistence
after an acute episode of low back pain: a nonrandomized trial in a primary care setting. PLoS
One. 2007;2:e706. http://dx.doi.org/10.1371/
journal.pone.0000706
6. d e Jong JR, Vlaeyen JW, Onghena P, Goossens
ME, Geilen M, Mulder H. Fear of movement/(re)
injury in chronic low back pain: education or
exposure in vivo as mediator to fear reduction?
Clin J Pain. 2005;21:9-17; discussion 69-72.
7. de Souza FS, Marinho Cda S, Siqueira FB,
Maher CG, Costa LO. Psychometric testing confirms that the Brazilian-Portuguese
adaptations, the original versions of the
Fear-Avoidance Beliefs Questionnaire, and the
Tampa Scale of Kinesiophobia have similar
measurement properties. Spine (Phila Pa 1976).
2008;33:1028-1033. http://dx.doi.org/10.1097/
BRS.0b013e31816c8329
8. D’Eon JL, Harris CA, Ellis JA. Testing factorial validity and gender invariance of the pain catastrophizing scale. J Behav Med. 2004;27:361-372.
9. Duffy FF, Chung H, Trivedi M, Rae DS, Regier DA,
Katzelnick DJ. Systematic use of patient-rated
depression severity monitoring: is it helpful and
feasible in clinical psychiatry? Psychiatr Serv.
2008;59:1148-1154. http://dx.doi.org/10.1176/
appi.ps.59.10.1148
10. Dworkin RH, Turk DC, Wyrwich KW, et al. Interpreting the clinical importance of treatment outcomes in chronic pain clinical trials: IMMPACT
recommendations. J Pain. 2008;9:105-121.
http://dx.doi.org/10.1016/j.jpain.2007.09.005
11. Fairbank JC, Pynsent PB. The Oswestry Disability Index. Spine (Phila Pa 1976). 2000;25:29402952; discussion 2952.
12. Fritz JM, George SZ, Delitto A. The role of fearavoidance beliefs in acute low back pain: relationships with current and future disability and
work status. Pain. 2001;94:7-15.
13. George SZ, Fritz JM, Bialosky JE, Donald DA. The
effect of a fear-avoidance-based physical therapy intervention for patients with acute low back
pain: results of a randomized clinical trial. Spine
(Phila Pa 1976). 2003;28:2551-2560. http://
dx.doi.org/10.1097/01.BRS.0000096677.84605.
A2
14. George SZ, Fritz JM, Childs JD. Investigation of
elevated fear-avoidance beliefs for patients with
low back pain: a secondary analysis involving
patients enrolled in physical therapy clinical trials. J Orthop Sports Phys Ther. 2008;38:50-58.
http://dx.doi.org/10.2519/jospt.2008.2647
15. George SZ, Fritz JM, McNeil DW. Fear-avoidance
beliefs as measured by the fear-avoidance
beliefs questionnaire: change in fear-avoidance
beliefs questionnaire is predictive of change in
self-report of disability and pain intensity for
patients with acute low back pain. Clin J Pain.
2006;22:197-203.
16. George SZ, Hirsh AT. Psychologic influence on
experimental pain sensitivity and clinical pain
intensity for patients with shoulder pain. J Pain.
2009;10:293-299. http://dx.doi.org/10.1016/j.
jpain.2008.09.004
17. George SZ, Valencia C, Zeppieri G, Jr., Robinson
ME. Development of a self-report measure
of fearful activities for patients with low back
pain: the fear of daily activities questionnaire.
Phys Ther. 2009;89:969-979. http://dx.doi.
journal of orthopaedic & sports physical therapy | volume 40 | number 4 | april 2010 |
02 George.indd 203
203
3/23/10 5:19:40 PM
[
org/10.2522/ptj.20090032
18. G
eorge SZ, Wittmer VT, Fillingim RB, Robinson
ME. Fear-avoidance beliefs and temporal summation of evoked thermal pain influence selfreport of disability in patients with chronic low
back pain. J Occup Rehabil. 2006;16:95-108.
http://dx.doi.org/10.1007/s10926-005-9007-y
19. George SZ, Wittmer VT, Fillingim RB, Robinson
ME. Sex and pain-related psychological variables are associated with thermal pain sensitivity for patients with chronic low back pain. J
Pain. 2007;8:2-10. http://dx.doi.org/10.1016/j.
jpain.2006.05.009
20. George SZ, Zeppieri G. Physical therapy utilization of graded exposure for patients with
low back pain. J Orthop Sports Phys Ther.
2009;39:496-505. http://dx.doi.org/10.2519/
jospt.2009.2983
21. George SZ, Zeppieri G, Jr., Cere AL, et al. A
randomized trial of behavioral physical therapy
interventions for acute and sub-acute low back
pain (NCT00373867). Pain. 2008;140:145-157.
http://dx.doi.org/10.1016/j.pain.2008.07.029
22. Goubert L, Crombez G, Van Damme S, Vlaeyen
JW, Bijttebier P, Roelofs J. Confirmatory factor
analysis of the Tampa Scale for Kinesiophobia:
invariant two-factor model across low back pain
patients and fibromyalgia patients. Clin J Pain.
2004;20:103-110.
23. Grotle M, Brox JI, Vollestad NK. Reliability, validity and responsiveness of the
fear-avoidance beliefs questionnaire: methodological aspects of the Norwegian version. J
Rehabil Med. 2006;38:346-353. http://dx.doi.
org/10.1080/16501970600722403
24. Haugen AJ, Grovle L, Keller A, Grotle M.
Cross-cultural adaptation and validation of
the Norwegian version of the Tampa scale
for kinesiophobia. Spine (Phila Pa 1976).
2008;33:E595-601. http://dx.doi.org/10.1097/
BRS.0b013e31817c6c4b
25. Hirsh AT, George SZ, Riley JL, 3rd, Robinson ME.
An evaluation of the measurement of pain catastrophizing by the coping strategies questionnaire. Eur J Pain. 2007;11:75-81. http://dx.doi.
org/10.1016/j.ejpain.2005.12.010
26. Holm I, Friis A, Storheim K, Brox JI. Measuring self-reported functional status and pain in
patients with chronic low back pain by postal
questionnaires: a reliability study. Spine (Phila
Pa 1976). 2003;28:828-833.
27. Huang FY, Chung H, Kroenke K, Delucchi KL,
Spitzer RL. Using the Patient Health Questionnaire-9 to measure depression among racially
and ethnically diverse primary care patients. J
Gen Intern Med. 2006;21:547-552. http://dx.doi.
org/10.1111/j.1525-1497.2006.00409.x
28. Jacob T, Baras M, Zeev A, Epstein L. Low back
pain: reliability of a set of pain measurement
tools. Arch Phys Med Rehabil. 2001;82:735-742.
http://dx.doi.org/10.1053/apmr.2001.22623
29. Jensen MP, Turner JA, Romano JM. Changes in beliefs, catastrophizing, and coping are associated
with improvement in multidisciplinary pain treatment. J Consult Clin Psychol. 2001;69:655-662.
research report
30. J ensen MP, Turner JA, Romano JM, Fisher LD.
Comparative reliability and validity of chronic
pain intensity measures. Pain. 1999;83:157-162.
31. Kleinbaum DG, Kupper LL, Muller KE, Nizam A.
Applied Regression Analysis and Other Multivariable Methods. 3rd ed. Belmont, CA: Thomson
Brooks-Cole; 1997.
32. Kroenke K, Spitzer RL. The PHQ-9: a new
depression diagnostic and severity measure.
Psych Annals. 2002;32:509-521.
33. Leeuw M, Goossens ME, Linton SJ, Crombez
G, Boersma K, Vlaeyen JW. The fear-avoidance
model of musculoskeletal pain: current state of
scientific evidence. J Behav Med. 2007;30:77-94.
http://dx.doi.org/10.1007/s10865-006-9085-0
34. Leeuw M, Goossens ME, van Breukelen GJ,
Boersma K, Vlaeyen JW. Measuring perceived
harmfulness of physical activities in patients
with chronic low back pain: the Photograph
Series of Daily Activities--short electronic version. J Pain. 2007;8:840-849. http://dx.doi.
org/10.1016/j.jpain.2007.05.013
35. Lethem J, Slade PD, Troup JD, Bentley G. Outline
of a Fear-Avoidance Model of exaggerated pain
perception--I. Behav Res Ther. 1983;21:401-408.
36. Lindstrom I, Ohlund C, Eek C, et al. The effect
of graded activity on patients with subacute low
back pain: a randomized prospective clinical
study with an operant-conditioning behavioral
approach. Phys Ther. 1992;72:279-290; discussion 291-273.
37. Linton SJ, Boersma K, Jansson M, Overmeer
T, Lindblom K, Vlaeyen JW. A randomized controlled trial of exposure in vivo for patients with
spinal pain reporting fear of work-related activities. Eur J Pain. 2008;12:722-730. http://dx.doi.
org/10.1016/j.ejpain.2007.11.001
38. Lowe B, Spitzer RL, Williams JB, Mussell M,
Schellberg D, Kroenke K. Depression, anxiety
and somatization in primary care: syndrome
overlap and functional impairment. Gen Hosp
Psychiatry. 2008;30:191-199. http://dx.doi.
org/10.1016/j.genhosppsych.2008.01.001
39. McCracken LM, Zayfert C, Gross RT. The Pain
Anxiety Symptoms Scale: development and validation of a scale to measure fear of pain. Pain.
1992;50:67-73.
40. McNeil DW, Rainwater AJ, 3rd. Development of
the Fear of Pain Questionnaire--III. J Behav Med.
1998;21:389-410.
41. Osman A, Breitenstein JL, Barrios FX, Gutierrez
PM, Kopper BA. The Fear of Pain QuestionnaireIII: further reliability and validity with nonclinical
samples. J Behav Med. 2002;25:155-173.
42. Roebroeck ME, Harlaar J, Lankhorst GJ. The application of generalizability theory to reliability
assessment: an illustration using isometric
force measurements. Phys Ther. 1993;73:386395; discussion 396-401.
43. Roelofs J, Goubert L, Peters ML, Vlaeyen JW,
Crombez G. The Tampa Scale for Kinesiophobia:
further examination of psychometric properties
in patients with chronic low back pain and fibromyalgia. Eur J Pain. 2004;8:495-502. http://
dx.doi.org/10.1016/j.ejpain.2003.11.016
]
44. S
evereijns R, Vlaeyen JW, van den Hout MA,
Weber WE. Pain catastrophizing predicts pain
intensity, disability, and psychological distress
independent of the level of physical impairment.
Clin J Pain. 2001;17:165-172.
45. Smeets RJ, Vlaeyen JW, Kester AD, Knottnerus
JA. Reduction of pain catastrophizing mediates
the outcome of both physical and cognitivebehavioral treatment in chronic low back
pain. J Pain. 2006;7:261-271. http://dx.doi.
org/10.1016/j.jpain.2005.10.011
46. Staerkle R, Mannion AF, Elfering A, et al.
Longitudinal validation of the fear-avoidance
beliefs questionnaire (FABQ) in a Swiss-German
sample of low back pain patients. Eur Spine J.
2004;13:332-340. http://dx.doi.org/10.1007/
s00586-003-0663-3
47. Stratford PW, Binkley FM, Riddle DL. Health
status measures: strategies and analytic methods for assessing change scores. Phys Ther.
1996;76:1109-1123.
48. Stratford PW, Goldsmith CH. Use of the standard
error as a reliability index of interest: an applied
example using elbow flexor strength data. Phys
Ther. 1997;77:745-750.
49. Sullivan M, Bishop SR, Pivik J. The pain catastrophizing scale: development and validation.
Psych Assess. 1995;7:524-532.
50. Sullivan MJ, Stanish WD. Psychologically based
occupational rehabilitation: the Pain-Disability
Prevention Program. Clin J Pain. 2003;19:97104.
51. Sullivan MJ, Thorn B, Haythornthwaite JA, et
al. Theoretical perspectives on the relation
between catastrophizing and pain. Clin J Pain.
2001;17:52-64.
52. Swinkels-Meewisse EJ, Swinkels RA, Verbeek
AL, Vlaeyen JW, Oostendorp RA. Psychometric
properties of the Tampa Scale for kinesiophobia
and the fear-avoidance beliefs questionnaire in
acute low back pain. Man Ther. 2003;8:29-36.
53. Swinkels-Meewisse IE, Roelofs J, Schouten
EG, Verbeek AL, Oostendorp RA, Vlaeyen
JW. Fear of movement/(re)injury predicting
chronic disabling low back pain: a prospective
inception cohort study. Spine (Phila Pa 1976).
2006;31:658-664. http://dx.doi.org/10.1097/01.
brs.0000203709.65384.9d
54. Turner JA, Aaron LA. Pain-related catastrophizing: what is it? Clin J Pain. 2001;17:65-71.
55. Turner JA, Dworkin SF, Mancl L, Huggins KH,
Truelove EL. The roles of beliefs, catastrophizing, and coping in the functioning of patients
with temporomandibular disorders. Pain.
2001;92:41-51.
56. Van Damme S, Crombez G, Bijttebier P, Goubert
L, Van Houdenhove B. A confirmatory factor
analysis of the Pain Catastrophizing Scale: invariant factor structure across clinical and nonclinical populations. Pain. 2002;96:319-324.
57. van Wijk AJ, Hoogstraten J. Dutch translation of the Fear of Pain Questionnaire: factor
structure, reliability and validity. Eur J Pain.
2006;10:479-486. http://dx.doi.org/10.1016/j.
ejpain.2005.06.008
204 | april 2010 | volume 40 | number 4 | journal of orthopaedic & sports physical therapy
02 George.indd 204
3/23/10 5:19:41 PM
58. V
laeyen JW, de Jong J, Geilen M, Heuts PH,
van Breukelen G. Graded exposure in vivo in
the treatment of pain-related fear: a replicated
single-case experimental design in four patients
with chronic low back pain. Behav Res Ther.
2001;39:151-166.
59. Vlaeyen JW, de Jong J, Geilen M, Heuts PH, van
Breukelen G. The treatment of fear of movement/(re)injury in chronic low back pain: further
evidence on the effectiveness of exposure in
vivo. Clin J Pain. 2002;18:251-261.
60. Vlaeyen JW, Linton SJ. Fear-avoidance and its
consequences in chronic musculoskeletal pain:
a state of the art. Pain. 2000;85:317-332.
61. W
addell G, Newton M, Henderson I, Somerville
D, Main CJ. A Fear-Avoidance Beliefs Questionnaire (FABQ) and the role of fear-avoidance
beliefs in chronic low back pain and disability.
Pain. 1993;52:157-168.
62. Waddell G, Somerville D, Henderson I, Newton
M. Objective clinical evaluation of physical impairment in chronic low back pain. Spine (Phila
Pa 1976). 1992;17:617-628.
63. Wideman TH, Adams H, Sullivan MJ. A prospective sequential analysis of the fear-avoidance
model of pain. Pain. 2009;145:45-51. http://
dx.doi.org/10.1016/j.pain.2009.04.022
64. Woby SR, Roach NK, Urmston M, Watson
PJ. Psychometric properties of the TSK-11: a
shortened version of the Tampa Scale for Kinesiophobia. Pain. 2005;117:137-144. http://dx.doi.
org/10.1016/j.pain.2005.05.029
65. Wyrwich KW, Tierney WM, Wolinsky FD. Further
evidence supporting an SEM-based criterion for
identifying meaningful intra-individual changes
in health-related quality of life. J Clin Epidemiol.
1999;52:861-873.
@
more information
www.jospt.org
PUBLISH Your Manuscript in a Journal With International Reach
JOSPT offers authors of accepted papers an international audience. The
Journal is currently distributed to the members of the following
organizations as a member benefit:
• APTA's Orthopaedic and Sports Physical Therapy Sections
• Sports Physiotherapy Australia (SPA) Titled Members
• Physio Austria (PA) Sports Group
• Belgische Vereniging van Manueel Therapeuten-Association Belge des
Thérapeutes Manuels (BVMT-ABTM)
• Comitê de Fisioterapia Esportiva do Estado do Rio de Janeiro
(COFEERJ)
• Canadian Physiotherapy Association (CPA) Orthopaedic Division
• Suomen Ortopedisen Manuaalisen Terapian Yhdistys ry (SOMTY)
• German Federal Association of Manual Therapists (DFAMT)
• Hellenic Scientific Society of Physiotherapy (HSSPT) Sports
Injury Section
• Chartered Physiotherapists in Sports and Exercise Medicine (CPSEM)
and Chartered Physiotherapists in Manipulative Therapy (CPMT)
of the Irish Society of Chartered Physiotherapists (ISCP)
• Israeli Physiotherapy Society (IPTS)
• Gruppo di Terapi Manuale (GTM), a special interest group
of Associazione Italiana Fisioterapisti (AIFI)
• Nederlandse Vereniging Voor Manuele Therapie (NVMT)
• New Zealand Sports and Orthopaedic Physiotherapy Association
• Portuguese Sports Physiotherapy Group (PSPG) of the Portuguese
Association of Physiotherapists
• Orthopaedic Manipulative Physiotherapy Group (OMPTG) of the
South African Society of Physiotherapy (SASP)
• Swiss Sports Physiotherapy Association (SSPA)
• Association of Turkish Sports Physiotherapists (ATSP)
In addition, JOSPT reaches students and faculty, physical therapists and
physicians at more than 1,400 institutions in the United States and around
the world. We invite you to review our Information for and Instructions
to Authors at www.jospt.org and submit your manuscript for peer review at
http://mc.manuscriptcentral.com/jospt.
journal of orthopaedic & sports physical therapy | volume 40 | number 4 | april 2010 |
02 George.indd 205
205
3/23/10 5:19:43 PM
Download