The Journal of Pain, Vol 10, No 4 (April), 2009: pp 408-415

Available online at www.sciencedirect.com

Acceptance, Mindfulness, and Values-Based Action May

Counteract Fear and Avoidance of Emotions in Chronic Pain:
An Analysis of Anxiety Sensitivity

Lance M. McCracken and Edmund Keogh

Centre for Pain Services, Royal National Hospital for Rheumatic Diseases, and Centre for Pain Research,
University of Bath, Bath, United Kingdom.

Abstract: People often respond with distress and avoidance to their own negative experiences,
such as the physical, cognitive, and emotional aspects of depression or anxiety. When people with
chronic pain respond this way, their overall level of distress may increase, they may struggle to avoid
their emotional experiences, and their daily functioning may decrease. The purpose of this study was
to examine the role of anxiety sensitivity (AS), or ‘‘fear of anxiety,’’ in relation to these processes. It
was predicted that those persons with chronic pain who report higher AS will also report higher
emotional distress and greater disability caused by chronic pain. A second purpose was to examine
whether therapeutic processes designed to reduce emotional avoidance, namely, acceptance, mind-
fulness, and values, could be demonstrated to reduce the role of AS in relation to this distress and
disability based on a statistical model including these variables. Subjects were 125 consecutive adult
patients (64.8% women) seeking services from a specialty pain service in the United Kingdom. All
patients completed a standard set of measures of AS, acceptance of pain, mindfulness, and
values-based action, as well as measures of pain, disability, and emotional functioning, at their initial
consultation, and these data formed the basis for the current study. In correlation and regression
analyses, AS was associated with greater pain, disability, and distress. In regression analyses, the
3 proposed therapeutic processes reduced the average variance accounted for by AS in patient func-
tioning from DR2 = .21 to DR2 = .048. This means that when the 3 therapeutic variables are taken into
account statistically, AS alone retained relatively little association with patient functioning. These re-
sults suggest that AS may amplify the impact of emotional distress on patient functioning in chronic
pain and that processes of acceptance, mindfulness, and values-based action may reduce this effect.
Perspective: Humans can fear and struggle to avoid their own emotional experiences, even when
these cannot harm them. Data presented here show individuals with chronic pain have more distress
and disability when they manifest more fear of anxiety symptoms, and behavior patterns of ‘‘accep-
tance’’ and ‘‘mindfulness’’ may reduce this effect.
ª 2009 by the American Pain Society
Key words: Chronic pain, anxiety, avoidance, acceptance, mindfulness.

S
uffering and disability from chronic pain can arise of behavior entail what is referred to as psychological in-
from ineffectual avoidance of unwanted physical flexibility.6 In daily life, psychological inflexibility results
or emotional feelings, a lack of realistic contact in recurrent patterns of failing to persist in, or change,
with moment-to-moment experience, and lack of guid- behavior as required for healthy functioning.6 Its role
ance by values and goals over actions.4 Along with cogni- in chronic pain has been supported by a number of re-
tive processes that amplify distress and narrow cent studies.16-19
behavioral options (ie, ‘‘cognitive fusion’’) these qualities A key process in psychological inflexibility is experien-
tial avoidance, a behavior pattern in which individuals
Received May 6, 2008; Revised September 10, 2008; Accepted September
attempt to control or limit their contact with emotional,
30, 2008. physical, and cognitive experiences they evaluate as
Address reprint requests to Dr Lance M. McCracken, Centre for Pain Ser- undesirable,6 such as pain, fatigue, depression, anxiety,
vices, RNHRD, Bath BA1 1RL UK. E-mail: Lance.McCracken@RNHRD.nhs.uk
1526-5900/$36.00 anger, or painful memories. Naturally, if a person with
ª 2009 by the American Pain Society chronic pain is unwilling to have undesirable experi-
doi:10.1016/j.jpain.2008.09.015 ences, he or she will not engage in activities that include

408
McCracken and Keogh 409
these experiences, and this is likely to restrict overall collected between January and October 2006. Mean
functioning.17 age was 46.6 years (SD = 13.2). Most patients were mar-
There are a number of factors that may contribute to ried, 60.5%, followed by single, 21.8%, divorced,
experiential avoidance. One such factor that has recently 10.5%, cohabiting, 3.2%, and other, 4.0%. Average edu-
emerged in relation to chronic pain is anxiety sensitivity. cation completed was 12.5 years (SD = 2.4). Circumstances
Anxiety sensitivity (AS) can be defined as a person’s fear of pain onset were reported as unknown, 31.5%, acci-
of their own anxiety-related symptoms and is regarded dent at work, 21.0%, other accident, 14.5%, following
as a predisposing factor in the development of certain an illness, 9.7%, following a surgical operation, 6.5%,
anxiety-related disorders.22 AS has been studied in the or other, 16.8%. Most frequent diagnoses included
context of pain11 and shown to play a role in both tran- non-specific musculoskeletal pain, 35.4%, fibromyalgia,
sient laboratory pain12,13 and pain-related distress in 30.2%, failed lumbar surgery, 12.9%, complex regional
chronic pain sufferers.2,3,24,27 pain syndromes, 6.0%, or other, 15.5%. The median dura-
AS has not typically been considered in relation to tion of pain for the sample was 96.0 months (range 7.0 to
broader aspects of physical and psychosocial functioning 576.0). Only 16.0% of patients were in employment ei-
in patients with chronic pain or in relation to potential ther full or part time at the time of assessment. The local
treatment process variables. This is surprising, given research ethics committee approved the methods of this
that fear of one’s own emotional experiences would be study, and all patients provided written consent for their
expected to lead to avoidance of such experiences and data to be used in research.
situations that evoke them, and both distressing emo-
tions and avoidance appear as cardinal features of
chronic pain. This appears particularly true in those peo- Measures and Procedure
ple who experience relatively greater suffering and Measures at initial assessment were mailed to patients’
disability. homes, with instructions, for their first appointment.
Potential treatment process variables of interest in re- Standard background information regarding age, mari-
lation to AS include acceptance of pain, mindfulness, and tal status, work status, circumstances of pain onset, med-
values-based action. These entail, in turn, willingness to ication and health care use, and 0 to 10 numerical pain
have pain without struggling with it, present focused ratings were collected on a short self-report form. Med-
and nonreactive awareness, and actions under the influ- ication was quantified as the number of separate classes
ence of what the person holds as important. ‘‘Process’’ of pain medication currently being taken as reported in
here simply means that controlling elements of these the patient’s current medical record. Health care use in-
variables are potentially manipulable in treatment and cluded the patients’ report of the number of times they
may have the potential for reducing the influence of had seen their GP related to their pain in the past 6
AS. These processes are specified in a general treatment months. These measures of medication and health care
approach called Acceptance and Commitment Therapy use have been used frequently in previous studies and
(ACT)7 and in Contextual Cognitive Behavioral Therapy, obtain significant correlations with other measures of
a treatment approach for chronic pain based on ACT.14 functioning, supporting their validity.16,18 Patients also
As they are particularly designed to counteract processes completed the 7 standardized measures of daily function
of experiential avoidance and psychological inflexibility, and potential psychological influences described below.
they may be well suited to address the adverse effects All forms were reviewed at the consultation visit, to
engendered in AS. check for and assure completeness, yielding a high rate
The purpose of this study was to examine the role of AS of full complete data (less than 10% provide no data or
in relation to the functioning of persons seeking treat- data regarded as unusable because it includes mostly
ment for chronic pain. It was predicted that AS would missing responses).
be associated with higher levels of distress and lower The Anxiety Sensitivity Index (ASI)21 is a 16-item mea-
levels of daily functioning in this group. We also planned sure of fear of anxiety symptoms. Each item is rated on
to examine the role of 3 therapeutic process variables a 5-point scale from ‘‘very little’’ to ‘‘very much.’’ It has
(acceptance of pain, mindfulness, and values-based ac- been used in a number of studies of patients with chronic
tion) to investigate whether they appear to ‘‘buffer’’ or pain.2,3 The ASI contains separate factors related to phys-
partially nullify the role of AS in relation to patient func- ical (8 items), social (4 items), and mental (4 items) con-
tioning. It should be noted that this study involved data cerns that were also examined as a part of the present
collected at 1 point in time and therefore examined indi- study.26 The total score for the ASI ranges from 0 to 64.
vidual differences in the ‘‘treatment process’’ variables Previously published Cronbach’s a values for the physical,
but not change in the processes over time. social, and mental concerns scales are .89, .62, and .85, re-
spectively.25 Since there are few (if any) published values
within chronic pain groups, we also calculated values
Methods based on the current data, which were .85, .61, and .79,
and so comparable with previously published values.
Participants The British Columbia-Major Depression Inventory (BC-
Participants for this study were 125 consecutive adult MDI)9 is a 19-item self-report measure of depression
patients (64.8% women), assessed on a specialty pain modeled after the Diagnostic and Statistical Manual of
management unit in southwest England. All data were Mental Disorders (4th Edition; DSM-IV) criteria for major
410 Anxiety Sensitivity and Chronic Pain
1
depression. Items 1 to 16 are symptoms of depression. to chronic pain. Patients rate each item on a scale from
Patients are asked to indicate symptoms they have had 0 (never) to 5 (always) indicating how often they do or
in the past 2 weeks and rate each 1 on a 1 to 5 scale of experience each of the actions or thoughts described.
severity, from 1 (very mild problem) to 5 (very severe The total score for the PASS-20 ranges from 0 to 100.
problem). Three additional items ask patients to rate The scale has demonstrated good internal consistency re-
the impact of the endorsed symptoms on their lives, in liability, strong correlations with the original subscales
areas of work or school, family, and social life. Only the and with measures of patient functioning, and a con-
symptom severity score was used in the present study. firmed factor structure.15,23
This score ranges from 0 to 80. Scores from the BC-MDI The Sickness Impact Profile (SIP)4 is a 136-item measure
have demonstrated adequate reliability (a = .83 in pa- of the disability in relation to health problems. It con-
tients with chronic medical problems) and validity.8,9 tains 12 categories of functioning that are combined to
The Chronic Pain Acceptance Questionnaire (CPAQ)18 form composite scores including scores for physical and
is a 20-item inventory designed to measure acceptance psychosocial aspects of disability. Patients are asked to
of pain. The CPAQ includes 2 subscales: activity engage- endorse statements that describe problems they are hav-
ment and pain willingness. It generally assesses the pa- ing with functioning in relation to their health. All of the
tient’s tendency to perform activities with pain present scores from the SIP range from 0 to 1. The temporal con-
and the relative absence of attempts to control or avoid sistency reliability of the SIP total score is very good at r =
pain. Patients rate each item on a scale of 0 (never true) .92, and the composite scores have demonstrated good
to 6 (always true). The total score for the CPAQ ranges validity.4
from 0 to 120. The CPAQ scales have achieved internal
consistency values of .78 to .82 supporting reliability,
and significant correlations with measures of avoid- Analyses
ance, distress, and daily functioning, supporting their Preliminary analyses were done to examine means and
validity.18 standard deviations for the primary variables of the
The Chronic Pain Values Inventory (CPVI)19 is a 12-item study. The ASI scores were also examined in correlations
measure of importance and success in living according to and mean comparisons to understand their relations
values developed for use with patients with chronic pain. with patient background variables. In these analyses
This inventory was originally developed based on a values and throughout, the pain duration variable was log-
clarification exercise from ACT.7 The values domains in- transformed due to an extreme positive skew in these
cluded are family, intimate relations, friends, work, data. A series of correlations were calculated to examine
health, and growth or learning. For each domain patients relations between the ASI score and acceptance, mind-
are asked to make 2 ratings, 1 reflecting the importance fulness, and values scores as well as relations of all of
of their values in each domain and the second reflecting these with the measures of pain and patient functioning.
their success at living according to their values in each do- The primary analyses involved hierarchical multiple
main. Each item is rated on a 0 to 5 scale indicating their regression. Two separate sets of 6 multiple regression
degree of importance or success. Previous study has analyses were conducted. The criterion variables in
shown that the ‘‘importance’’ items from the CPVI do each case included depression, pain-related anxiety,
not demonstrate adequate variability for use in psycho- physical disability, psychosocial disability, pain medica-
metric analyses.19 The mean success rating was used in tion, and GP visits for pain. For each set of analyses, pa-
the present study as a measure of values-based action, tient background variables were tested for entry and
the extent to which patients see their behavior as guided retained if significant in a first block. Pain intensity was
by their values. This score ranges from 0 to 5. This scale has entered next, before the primary predictors of interest.
achieved a reliability coefficient of .82 and significant cor- In the first set of regressions the 3 ASI subscale scores
relations with related constructs such as avoidance and were entered simultaneously to test their unique and
acceptance.19 combined relations with the criterion variables. In the
The Mindful Attention Awareness Scale (MAAS)5 is second set of regressions acceptance, mindfulness, and
a 15-item measure of mindfulness. The item content re- values-based action scores were entered after pain,
flects the opposite of the construct of mindfulness, or and then the 3 ASI scores after that. This second set
‘‘mindlessness,’’ and thus endorsing each item at a lower of regressions was to examine the relative contribution
frequency is taken to mean a higher level of mindfulness. of acceptance, mindfulness, and values-based action, in
Each item is rated on a scale from 1 (almost always) to 6 comparison to the ASI scores, and to see if the ASI
(almost never), and an average calculated. Hence, the scores retained any significant relationship to the crite-
mean item rating for the MAAS ranges from 1 to 6. Scores rion variables after these were taken into account in the
from the MAAS achieve reliability levels above .80, corre- prediction equation.
late with measures of emotional distress and physical
symptoms, distinguish individuals with mindfulness
training and practice from those without, and correlate Results
with measures of self-awareness.5
The Pain Anxiety Symptoms Scale (PASS-20)15 is a 20- Preliminary Analyses
item version of the original 40-item PASS.20 It assesses The mean total ASI score in the current sample was
fear, avoidance, and other anxiety responses in relation 23.2 (SD = 12.1). Ranges, means, and standard deviations
McCracken and Keogh 411
Table 1.Range, Means, and Standard sample. The total score and 3 subscales from the ASI
Deviations for Primary Variables are included as well as scores for acceptance, mindful-
ness, and values-based action, and measures of pain, dis-
RANGE OF STANDARD tress, disability, and healthcare use. Generally, the ASI
SCALE SCORES MEAN DEVIATION
scores were moderately negatively correlated with ac-
Anxiety Sensitivity Total (ASI) 1-57 23.2 12.1 ceptance, mindfulness and values-based action, and pos-
ASI Physical concerns 0-28 11.0 6.9 itively correlated with measures of patient distress,
ASI Social concerns 0-12 4.8 2.9 disability, and healthcare use. The ASI scores achieved
ASI Mental concerns 0-20 7.4 4.0 higher correlations with measures of emotional distress
Acceptance of pain (CPAQ) 5-106 45.3 20.1 and disability and smaller correlations with pain inten-
Mindfulness (MAAS) 1.2-6 4.1 .89
sity, and medication use.
Values-based action (CPVI) 0-4.8 1.7 1.1
Pain intensity (0-10) 2-10 6.8 2.0
Depression (BC-MDI) 0-66 28.1 14.5 Regression Analyses
Pain anxiety (PASS) 8-100 49.3 19.9
Table 3 shows the results from the first set of regression
Physical disability (SIP) .02-.68 .25 .16
Psychsoc. disability (SIP) .00-.86 .29 .18 analyses of the ASI scores. Pain was a significant predictor
No. of meds (classes analgesics) 0-7 2.9 1.5 in each of the 6 equations, DR2 ranged from .047 to .13.
GP visits for pain (last 6 mo) 0-20 5.0 4.7 Patient background variables, on the other hand, played
little role. The 3 ASI variables accounted for significant
n = 125; except GP visits, n = 121, and MAAS, n = 122. variance in 5 of 6 equations; significant DR2 values
ranged from .13 in the equations for physical disability
to .33 in the equation for depression. The mean propor-
for all of the primary study variables are included in tion of variance accounted for by the ASI scores overall
Table 1. was .21. In each case the significant regression coeffi-
In correlation analyses, none of the ASI total or sub- cients were in the predicted direction, higher scores on
scale scores were significantly correlated with age, years components of anxiety sensitivity were associated with
of education, or duration of pain, all r < .17, P > .05. In t greater distress, disability, and healthcare use. From
tests, there were no differences between men versus among the ASI scores, mental concerns was a significant
women, married or cohabitating versus living alone, un- unique predictor in 4 equations, physical concerns in 2,
known onset circumstances versus accident, or working and social concerns in none.
to any degree versus not working due to pain, on any Table 4 shows the results from the alternate regression
of the ASI scores, all t < 1.60, P > .10. analyses where first acceptance, mindfulness, and
values-based action were entered, and then the ASI
scores were entered after that. As expected, again,
Correlation Analyses pain was a significant predictor in each equation and pa-
Table 2 includes results of correlation analyses of all tient background variables played little role. The block
primary and secondary measures administered to the of predictors including acceptance, mindfulness, and

Correlations Between Measures of Anxiety Sensitivity, Acceptance, Mindfulness, Values-

Table 2.
Based Action, and Emotional and Daily Functioning (n = 125)
MEASURES 1 2 3 4 5 6 7 8 9 10 11 12 13

1. ASIa Total
2. ASI Physical .93z
3. ASI Social .78z .58z
4. ASI Mental .82z .65z .53z
5. Pain Accept (CPAQ) .48z .48z .30z .41z
6. Mindfulness (MAAS)b .46z .42z .27y .50z .39z
7. Values (CPVI) .30z .30z .15 .27z .62z .27y
8. Pain intensity (0-10) .22* .20* .25y .09 .32z .31z .21*
9. Depression (BC-MDI) .55z .53z .35z .57z .69z .55z .57z .30z
10. Pain anxiety (PASS) .58z .57z .36z .51z .73z .53z .44z .35z .66z
11. Physical disabil (SIP) .42z .37z .37z .36z .51z .40z .43z .36z .61z .57z
12. Psychsoc disabil (SIP) .55z .48z .33z .60z .64z .58z .52z .27y .78z .59z .62z
13. Number meds .23* .16 .22* .23* .31z .30z .36z .22* .36z .25y .44z .34z
14. GP visits for painb .29z .17 .30z .35z .35z .32z .28y .22* .34z .31z .47z .46z .34z

Abbreviations: ASI, Anxiety Sensitivity Index. It yields scores for physical, social, and mental concerns and a total.
a
The numbers across the top row refer to the variables listed in rows 1 to 13.
b
n = 121 for GP visits; n = 122 for MAAS (all other, n = 125).
* P < .05,
y
P < .01,
z
P < .001.
412 Anxiety Sensitivity and Chronic Pain
Hierarchical Multiple Regression
Table 3. the significant unique predictor of functioning among
Analyses of Relations Between Components of the ASI scores.
Anxiety Sensitivity and Patient Functioning
DR2 Discussion
BLOCK PREDICTORS b (FINAL) (BLOCK) TOTAL R2
Results of this study show that AS and its components
Depression are associated with greater pain, emotional distress, and
1. Pain intensity .23y .092z disability in patients with chronic pain. From among the
2. Physical concerns .26y .33z 3 components, concern over mental experiences of anxi-
Social concerns .090 ety appears as the strongest predictor of depression, dis-
Mental concerns .43z .42z ability, and GP visits related to pain. Results also show
Pain-related anxiety that processes of acceptance of pain, mindfulness, and
1. Gender .12 .035*
values-based action reduce, but do not eliminate, the ex-
2. Education .046 .033*
tent to which the components of AS predict patient func-
3. Pain Intensity .26z .11z
4. Physical concerns .34z .26z tioning. The current results are similar to those from
Social concerns .048 other studies of AS in chronic pain, which show that anx-
Mental concerns .27y .44z iety sensitivity is associated with higher level of reported
Physical disability pain.2,3 However, the current results extend these find-
1. Pain intensity .29z .13z ings by showing that AS is also associated with greater
2. Physical concerns .10 .13z depression, pain-related anxiety, disability, and GP visits
Social concerns .13 related to pain.
Mental concerns .20 .26z It appears that if patients with chronic pain experience
Psychosocial disability
concern about their experiences of pain, and are fearful
1. Pain intensity .21y .072y
about such experiences, they may suffer greater overall
2. Physical concerns .15 .34z
Social concerns .093
distress, and a greater impact on daily functioning. AS
Mental concerns .53z .42z could then be conceptualized as a type of distress ampli-
No. of pain fier, or a process that contributes additional adverse
medications psychological meaning and influences to emotional ex-
1. Pain intensity .19* .048* periences.
2. Physical concerns .065 .050 Although AS generally was found to be related to
Social concerns .11 a range of pain-related outcomes, this was not found
Mental concerns .19 .099* for all 3 subcomponents. Indeed, regression analysis re-
GP visits related vealed that the physical and mental concerns, but not
to paina
the social concerns components of AS, were significant
1. Pain intensity .19* .047*
predictors. In addition, it seems that the mental concerns
2. Physical concerns .21 .14z
Social concerns .17 scale was the most consistent predictor, and particularly
Mental concerns .38z .18z related to pain-related disability. This is interesting since
it might have been expected that the physical concerns
a
n = 121 for GP visits (all other, n = 125). component of AS would be the most strongly related
* P < .05, to pain. Instead, AS physical concerns was only found
y
P < .01,
z
P < .001. to be the strongest predictor for 1 variable, namely
pain-related anxiety. With respect to the lack of impact
that the social concerns component of the ASI had, this
values-based action contributed significantly to the could be partially due to the relatively lower reliability
prediction equation for each of the 6 criterion variables; that we, as well as others, have observed with this sub-
significant DR2 values ranged from .13 in the equation scale. Together these results suggest that the fear of
for pain-related GP visits to .51 in the equation for de- mental events, and possibly physical concerns also, may
pression. The mean proportion of variance accounted be particularly important in chronic pain.
for by this block overall was .33. In each case significant Although AS has been primarily conceptualized as
regression coefficients for acceptance, mindfulness, a trait-like predisposing factor for the development of
and values based action were in the predicted direction, anxiety disorders, it is possible to broaden this conceptu-
indicating a positive relationship with better function- alization somewhat. We have considered it in relation to
ing. The 3 ASI scores entered in the last block contributed a process called experiential avoidance, which is adopted
significantly to the prediction equations in 4 instances of from the model of Acceptance and Commitment Ther-
6; significant DR2 values ranged from .040 to .081 apy.6 AS may be considered as part of a more general ten-
(although 1 of these yielded a unpredicted result, a sig- dency to respond in a distressed and avoidant way
nificant negative regression coefficient between physi- toward one’s own experiences of emotions. If so, then
cal concerns and pain-related GP visits). The mean what the ASI might be tapping is a behavior pattern
proportion of variance accounted for by the ASI scores that generalizes to experiences of other emotional expe-
in this set of analyses was .048. Again, it was the mental riences, such as anger, depression, fear, frustration, and
concerns component of the ASI that was most frequently perhaps others such as guilt, shame, or embarrassment.
McCracken and Keogh 413
Table 4.Hierarchical Multiple Regression Analyses of Relations Between Components of Anxiety
Sensitivity and Patient Functioning After Taking Into Account Relations of Acceptance,
Mindfulness, and Values-Related Processes With Patient Functioning
DR2 TOTAL
BLOCK PREDICTORS b (FINAL) (BLOCK) R2

Depression
1. Pain intensity .048 .087z
2. Acceptance of pain .32z .51z
Mindfulnessa .22y
Values-based action .21y
3. Physical concerns .072 .048y
Social concerns .022
Mental concerns .23* .65z
Pain-related anxiety
1. Gender .01 .040*
2. Duration of pain .079 .034*
3. Pain Intensity .068 .11z
4. Acceptance of pain .54z .45z
Mindfulness .19y
Values-based action .059
5. Physical concerns .13 .036y
Social concerns .049
Mental concerns .16 .66z
Physical disability
1. Pain intensity .14 .11z
2. Acceptance of pain .22* .23z
Mindfulness .13
Values-based action .18
3. Physical concerns .056 .040
Social concerns .17
Mental concerns .12 .38z
Psychosocial disability
1. Pain intensity .040 .078y
2. Acceptance of pain .32z .49z
Mindfulness .26z
Values-based action .18*
3. Physical concerns .040 .052y
Social concerns .021
Mental concerns .31z .63z
No. of pain medications
1. Pain intensity .069 .044*
2. Acceptance of pain .030 .15z
Mindfulness .19
Values-based action .31y
3. Physical concerns .20 .030
Social concerns .18
Mental concerns .077 .22z
a
GP visits related to pain
1. Pain intensity .086 .047*
2. Acceptance of pain .16 .13y
Mindfulness .14
Values-based action .13
3. Physical concernsb .33y .081y
Social concerns .21
Mental concerns .28* .26z

a
n = 121 for GP visits; n = 122 for MAAS (all other, n = 125).
b
The reverse sign of the regression coefficient for physical concerns in the equation for GP visits suggested a possible suppressor effect in the regression model. When
this variable was left out of the analyses the contribution from the block of variables related to AS no longer met significance and none of the regression coefficients
were significant.
* P < .05,
y
P < .01,
z
P < .001.
414 Anxiety Sensitivity and Chronic Pain
7
If AS does reflect a general tendency toward emo- include ACT, Contextual Cognitive Behavioral Ther-
tional/experiential avoidance, the current findings are apy,14 and mindfulness-based approaches.10 These types
of particular interest for treatment development. The of therapies, in part, focus not so much on changing
same model that proposes experiential avoidance as what people think or feel but on altering the meaning
a process of psychopathology also proposes a number derived and the influences exerted in emotion and be-
of therapeutic processes including acceptance, mindful- havior from what is thought or felt. By the use of such
ness, and values-based action. With multiple regression methods as exposure, paradox, metaphor, and aware-
we attempted to test the theory behind this model: ness training, these approaches may, for example, aim
that experiential avoidance contributes to suffering to reduce avoidance of anxious experiences, rather
and restricted functioning and that the proposed thera- than trying to eliminate the anxious experiences them-
peutic processes may reduce the impact of experiential selves.
avoidance. We demonstrated that when acceptance, The current methods are limited in a number of impor-
mindfulness, and values-based action are considered in tant ways. All data were collected at one point in time, and
relation to patient functioning along with AS, the vari- there was no experimental manipulation. Therefore, clear
ance in functioning for which they can account limits functional or causal relations between the processes and
(but does not completely eliminate) the predictive role behavior patterns studied here cannot be presumed.
played by AS, from a mean of 21% of variance to about Also, there may be influences such as shared method var-
5%. This provides support for the notion that these ther- iance, pain or emotional state variables, or other influ-
apeutic processes may reduce some of the restrictive in- ences that lead respondents to attempt to report on
fluences that emotional experiences can acquire in their circumstances in a consistent fashion, that may in-
relation to patient functioning. flate the size of the relationships that emerge in the
It is interesting that the ASI is able to examine mean- data. The sample is relatively small and from a specific
ings and influences born by physical, social, and mental healthcare context. Both the reliability and generality of
contexts of distress and that these are to some extent results will require further investigation. In our data the
separate. This reminds us that, in a sense, there is no social concerns subscale of the ASI had a relatively low in-
such thing as anxiety per se. Anxiety is a hypothetical ternal consistency value. This may have placed a limit on
construct, or a way of speaking about a cluster of psycho- the extent to which this variable could show relations
logical experiences and related behavior patterns. In the with relevant measures of daily functioning. This facet of
same sense neither anxiety nor AS per se are able to cause AS in particular may require further investigation. Finally,
or exert any influence on aspects of patient behavior or AS was used to sample a broader domain of emotional ex-
functioning. Underneath these terms are interrelated ex- perience. It is worth investigating the extent to which
periences and behavior patterns, that have been given there are differences in experiential avoidance across dif-
their relations with each other via the individual’s his- fering emotional experiences, and across different indi-
tory, and there are, as well, potentially manipulable situ- viduals, in different situations, and at different times.
ational and contextual elements that when manipulated In summary, processes of AS predict greater pain, dis-
in practice can alter these interrelations. tress, disability, and healthcare use in patients with
In the past it was not always clear whether AS would chronic pain. We infer from these data that AS may
form a link with a specific approach to treatment, or if have a relationship with processes of experiential avoid-
it might promote the development of specific methods ance. Acceptance, mindfulness, and values-based action
for modifying AS in chronic pain. The current results sug- predict patient distress and disability in ways that appear
gest that the contextual behavioral and cognitive thera- to reduce the relations of AS with these same variables.
pies, that include processes of acceptance, mindfulness, These results support the role of therapeutic processes
and values, may be particularly well designed to under- from contextual cognitive-behavioral approaches in
mine processes of AS as they relate to suffering and reducing the suffering and restricted functioning of
disability in chronic pain. These contextual therapies patients with chronic pain.

References 5. Brown KW, Ryan RM: The benefits of being present:

Mindfulness and its role in psychological well-being. J Pers
and Soc Psychol 84:822-848, 2003
1. American Psychiatric Association: Diagnostic and Statisti-
cal Manual of Mental Disorder, (4th ed.). Washington. DC, 6. Hayes SC, Luoma JB, Bond FW, Masuda A, Lillis J: Accep-
American Psychiatric Association, 1994 tance and commitment therapy: Model, processes and out-
comes. Behav Res Ther 44:1-25, 2006
2. Asmundson GJG, Norton GR: Anxiety sensitivity in pa-
tients with physically unexplained chronic back pain: A pre- 7. Hayes SC, Strosahl K, Wilson KG: Acceptance and commit-
liminary report. Behav Res Ther 33:771-777, 1995 ment therapy: An experiential approach to behavior change.
New York, Guilford, 1999
3. Asmundson GJG, Taylor S: Role of anxiety sensitivity in pain- 8. Iverson GL: Psychometric properties of the British Colum-
related fear and avoidance. J Behav Med 19:577-586, 1996 bia Major Depression Inventory. Can Psychology 42:49, 2001
4. Bergner M, Bobbitt RA, Carter WB, Gilson BS: The Sickness 9. Iverson GL, Remick R: Diagnostic accuracy of the British
Impact Profile: Development and final revision of a health Columbia Major Depression Inventory. Psychol Rep 95:
status measure. Med Care 19:787-805, 1981 1241-1247, 2004
McCracken and Keogh 415
10. Kabat-Zinn J: Full catastrophe living: Using the wisdom 19. McCracken LM, Yang S-Y: The role of values in a contex-
of your body and mind to face stress, pain, and illness, tual cognitive-behavioral analysis of chronic pain. Pain 123:
New York, Dell Publishing, 1990 137-145, 2006
11. Keogh E, Asmundson GJG: Negative affectivity, cata- 20. McCracken LM, Zayfert C, Gross RT: The Pain Anxiety
strophizing and anxiety sensitivity. In: Asmundson GJG, Symptoms Scale: Development and validation of a scale to
Vlaeyen JW, Crombez G (eds): Understanding and Treating measure fear of pain. Pain 50:67-73, 1992
Fear of Pain, Oxford, Oxford University Press, pp 91-115, 2004
21. Peterson RA, Reiss S: Anxiety sensitivity index manual,
12. Keogh E, Cochrane M: Anxiety sensitivity, cognitive bias Palos Heights, IL, International Diagnostic Systems, 1987
and the experience of pain. J Pain 3:320-329, 2002
22. Reiss S: Theoretical perspectives on the fear of anxiety.
13. Keogh E, Mansoor L: Investigating the effects of anxiety Clin Psychol Rev 7:585-596, 1987
sensitivity and coping strategy on the perception of cold
pressor pain in healthy women. Eur J Pain 5:11-25, 2001 23. Roelofs J, McCracken LM, Peters ML, Crombez G, van
Breukelen G, Vlaeyen JWS: Psychometric evaluation of the
14. McCracken LM: Contextual Cognitive-Behavioral Ther- Pain Anxiety Symptoms Scale (PASS) in chronic pain patients.
apy for chronic pain, progress in pain research and manage- J Behav Med 27:167-183, 2004
ment, vol. 33, Seattle, IASP Press, 2005
24. Stewart SH, Asmundson GJG: Anxiety sensitivity and its
15. McCracken LM, Dhingra L: A short version of the Pain impact on pain experiences and conditions: A state of the
Anxiety Symptoms Scale (PASS-20): Preliminary develop- art. Cogn Behav Ther 35:185-188, 2006
ment and validity. Pain Res Manag 7:45-50, 2002
25. Zinbarg R, Barlow DH: Structure of anxiety and the anx-
16. McCracken LM, Gauntlett-Gilbert, Vowles K: The role of iety disorders: A hierarchical model. J Abnorm Psychol 105:
mindfulness in a contextual cognitive-behavioral analysis of 181-193, 1996
chronic pain-related suffering and disability. Pain 131:63-69,
2007 26. Zinbarg R, Barlow DH, Brown TA: The hierarchical struc-
ture and general factor saturation of the Anxiety Sensitivity
17. McCracken LM, Samuel VM: The role of avoidance, pac- Index: Evidence and implications. Psychol Assess 35:277-284,
ing, and other activity patterns in chronic pain. Pain 130: 1997
119-125, 2007
27. Zvolensky MJ, Goodie JL, McNeil DW, Sperry JA,
18. McCracken LM, Vowles KE, Eccleston C: Acceptance of Sorrell JT: Anxiety sensitivity in the prediction of pain-
chronic pain: Component analysis and a revised assessment related fear and anxiety in a heterogeneous chronic pain
method. Pain 107:159-166, 2004 population. Behav Res Ther 39:683-696, 2001