Effects of spouse-assisted coping skills training and exercise training

in patients with osteoarthritic knee pain: a randomized controlled study

Francis J. Keefea,*, James Blumenthala, Donald Baucomb, Glenn Affleckc, Robert Waugha,David S. Caldwella, Pat Beauprea, Susmita Kashikar-Zucka,

Katherine Wrighta, Jennifer Egerta, John Lefebvred

aDuke University Medical Center, Box 3129, Durham, NC 27710, USAbUniversity of North Carolina at Chapel Hill, Chapel Hill, NC, USA

cUniversity of Connecticut, Storrs, CT, USAdWofford College, Spartanburg, SC, USA

Received 6 June 2003; received in revised form 21 January 2004; accepted 4 March 2004

Abstract

This study tested the separate and combined effects of spouse-assisted pain coping skills training (SA-CST) and exercise training (ET) in a

sample of patients having persistent osteoarthritic knee pain. Seventy-two married osteoarthritis (OA) patients with persistent knee pain and

their spouses were randomly assigned to: SA-CST alone, SA-CST plus ET, ET alone, or standard care (SC). Patients in SA-CST alone, together

with their spouses, attended 12 weekly, 2-h group sessions for training in pain coping and couples skills. Patients in SA-CST þ ET received

spouse-assisted coping skills training and attended 12-weeks supervised ET. Patients in the ET alone condition received just an exercise

program. Data analyses revealed: (1) physical fitness and strength: the SA-CST þ ET and ET alone groups had significant improvements in

physical fitness compared to SA-CST alone and patients in SA-CST þ ET and ET alone had significant improvements in leg flexion and

extension compared to SA-CST alone and SC, (2) pain coping: patients in SA-CST þ ET and SA-CST alone groups had significant

improvements in coping attempts compared to ET alone or SC and spouses in SA-CST þ ET rated their partners as showing significant

improvements in coping attempts compared to ET alone or SC, and (3) self-efficacy: patients in SA-CST þ ET reported significant

improvements in self-efficacy and their spouses rated them as showing significant improvements in self-efficacy compared to ET alone or SC.

Patients receiving SA-CST þ ET who showed increased self-efficacy were more likely to have improvements in psychological disability. An

intervention that combines spouse-assisted coping skills training and exercise training can improve physical fitness, strength, pain coping, and

self-efficacy in patients suffering from pain due to osteoarthritis.

q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.

Keywords: Pain coping skills; Osteoarthritis; Spouse-assisted coping skills training

Osteoarthritis (OA) is the most common rheumatic disease

affecting up to 70% of older adults (Felson et al., 1987;

Hochberg et al., 1989; Lawrence, 1960; Lawrence et al.,

1966). Medical treatments (e.g. aspirin, nonsteroidal anti-

inflammatory drugs, steroid injections) benefit some OA

patients but may have significant side effects (gastrointes-

tinal changes, potential for accelerating the disease) that

limit their long-term use (Schnitzer, 1993). With recog-

nition of the limitations of medical treatment has increased

interest in self-management approaches to OA (Brandt,

2000; Buckwalter et al., 2001; Creamer, 2000; Felson et al.,

2000; Hochberg et al., 1995; Parker et al., 1993).

Over the past 15 years, two approaches to the self-

management of OA pain have been developed. The first

approach, developed by health psychologists and arthritis

educators, is based on biopsychosocial theories of pain

(Keefe et al., 2002). The goal of this approach is to enhance

patients’ abilities to cope with pain and other arthritis

symptoms through systematic training in pain coping skills.

An example of this approach is a study testing the efficacy of

a spouse-assisted pain coping skills training protocol for

OA (Keefe et al., 1996a,b, 1999). In this protocol,

patients attended treatment sessions with their spouses

0304-3959/$20.00 q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.

doi:10.1016/j.pain.2004.03.022

Pain 110 (2004) 539–549

www.elsevier.com/locate/pain

* Corresponding author. Tel.: þ1-919-668-2806; fax: þ1-919-668-2850.

E-mail address: keefe003@mc.duke.edu (F.J. Keefe).

and the training not only included instruction in pain coping

skills, but also instruction in couples skills designed to

supplement and reinforce learned pain coping skills. Results

revealed a consistent pattern in which patients who received

spouse-assisted coping skills training had the best outcomes,

those in the conventional pain coping skills training

condition (patients treated on their own) the next best

outcomes, and patients in an arthritis education-social

support control condition (patients and spouses attending

informational/support group sessions) had the poorest

outcomes (Keefe et al., 1996a,b, 1999). This study suggests

that spouse-assisted coping skills may be particularly

beneficial for patients having OA.

The second approach to OA self-management, developed

by specialists in physical medicine and rehabilitation,

involves exercise training (van Baar et al., 1999; Chard

and Dieppe, 2001; Clyman, 2001; Petrella, 2000). The

primary rationale for exercise training is that it reduces the

effect of deconditioning seen in many patients with

persistent OA pain and increases muscular strength to

provide greater support for the affected knee. Recent studies

have shown that isometric exercise training focused on

strengthening the quadriceps muscles (Gur et al., 2002;

Petrella and Bartha, 2000), supervised walking (Toda,

2001), or general aerobic conditioning (van Baar et al.,

2001; Penninx et al., 2001) can lead to improved

functioning in OA patients.

The present study compared the separate and combined

effects of spouse-assisted coping skills training and exercise

for OA. Prior studies have shown that arthritis patients vary

considerably in how well they respond to these treatments,

with patients showing increases in self-efficacy over the

course of treatment experiencing much greater improve-

ments (van Baar et al., 2001; Keefe et al., 1996a,b). Thus, a

secondary goal was to examine how changes in self-efficacy

occurring over treatment relate to the separate and

combined effects of spouse-assisted coping skills training

and exercise interventions.

1. Methods

1.1. Participants

Participants and their spouses were recruited from

rheumatology clinics and advertisements placed in news-

papers. Patients underwent a physical examination by a

board-certified rheumatologist (DSC). Patients were

excluded if they had comorbid medical conditions that

could affect their health status over the course of the trial (e.g.

a recent myocardial infarction), an abnormal cardiac

response to exercise (e.g. exercise-induced ventricular

tachycardia, abnormal blood pressure response), or other

known organic disease that would contraindicate safe

participation in the study (e.g. chronic obstructive

pulmonary disease, congestive heart failure, or cancer).

Of the individuals who volunteered for the study, 30 were

disqualified during the initial evaluation process due to the

following: cardiac problems ðn ¼ 7Þ; high blood pressure

ðn ¼ 2Þ; multiple medical problems ðn ¼ 8Þ; a positive stress

test ðn ¼ 4Þ; osteopororis ðn ¼ 2Þ; fibromyalgia ðn ¼ 1Þ;

severe depression ðn ¼ 1Þ; death of their spouse ðn ¼ 1Þ; lack

of agreement for participation from their primary physician

ðn ¼ 1Þ; not having osteoarthritis ðn ¼ 1Þ; and joining an

exercise club ðn ¼ 1Þ: Sixteen of the patients meeting criteria

for study entry withdrew prior to treatment due to inability to

make sessions at the scheduled times ðn ¼ 13Þ and other

reasons ðn ¼ 3Þ: The final study sample consisted of 72

married patients having persistent knee pain due to OA

and who were diagnosed as having OA of the knees and their

respective spouses. Over the course of the study, five subjects

dropped out of the study, two from the spouse-assisted

coping skills training condition, one from the spouse-assisted

coping skills training plus exercise training condition, and

two from the standard care control condition.

1.2. Procedure

Informed consent was obtained and patients were

provided with an Arthritis Foundation pamphlet providing

basic information about OA. The protocol was approved by

the Institutional Review Board at Duke Medical Center.

Patients completed a baseline evaluation (see below) and

were then randomly assigned to one of four conditions: (1)

spouse-assisted CST (SA-CST), (2) spouse-assisted CST

plus exercise training (SA-CST þ ET), (3) exercise training

alone (ET), or (4) standard care (SC). Following the

treatment program, participants completed a second assess-

ment battery identical to the baseline evaluation.

1.3. Assessment measures

Aerobic fitness and strength measures. Participants

underwent bicycle ergometry and strength testing before

and after the 12-week treatment program. Each participant

performed a maximal effort bicycle ergometry exercise test

following an initial practice test on a Corival 400 cycle

ergometer. The graded exercise protocol consisted of 3-min

stages starting at 150 kpm and increasing 150 kpm at each

stage. Participants maintained a pedaling rate of 50 rpm and

exercised until exhaustion. This is a standard exercise

testing protocol that has been used in our previous research

with older individuals (Blumenthal et al., 1991). Heart rates

were recorded every minute. Blood pressure was measured

automatically with an automated blood pressure monitor.

Respiratory and oxygen consumption measurements were

obtained and measurements of VO2K were obtained every

15 s. Participants were retested with an identical protocol at

the end of the 12-week treatment phase of the study.

Improvements in aerobic fitness were determined by

F.J. Keefe et al. / Pain 110 (2004) 539–549540

comparisonsof pre-and post-training peak VO2K(in mlkg/min).

To assess muscle strength, each participant performed a set

of maximal effort leg extensions, leg flexions, and bicep

curls following an initial practice test. This is a standardized

procedure for documenting strength that has been used in

our previous work (Blumenthal et al., 1988).

Pain coping. To assess pain coping, patients completed

the Coping Strategies Questionnaire (CSQ) (Rosenstiel and

Keefe, 1983). The CSQ is a 44-item instrument that asks

patients to report on the frequency that they use seven

coping strategies when they feel pain (coping self-

statements, praying or hoping, ignoring pain sensations,

reinterpreting pain sensations, increasing behavioral activi-

ties, catastrophizing, and diverting attention). Factor

analytic studies have identified two coping factors that

explain the majority of variance in arthritis patients’

responses on the CSQ: (a) Coping Attempts and (b) Pain

Control and Rational Thinking (Keefe et al., 1987a,b;

Parker et al., 1989). The spouse version of the Coping

Strategies Questionnaire was used to evaluate spouses’

perceptions of the patient’s coping strategies (Keefe et al.,

1996a,b). This instrument is identical to the patient version

with the exception that spouses are asked to rate how

frequently they believed their partners were using each of

the coping strategies to cope or deal with pain.

Self efficacy. The Arthritis Self-Efficacy Scale (Lorig

et al., 1989) was used to assess self-efficacy. This instrument

includes three subscales that assess self-efficacy for control

of three important areas: (1) pain management, (2) physical

function, and (3) other arthritis symptoms. Scores on these

subscales were combined to form a composite self-efficacy

measure (total self-efficacy) (Keefe et al., 1996a,b, 1999).

Prior studies have supported both the reliability and validity

of the Arthritis Self-Efficacy Scale (Lorig et al., 1989; Lorig

and Gonzalez, 1992; Schiaffino et al., 1991).

A spouse version of the Arthritis Self-Efficacy Scale was

used to evaluate spouses’ perceptions of the patients’ self-

efficacy. This instrument, used in prior studies (Keefe et al.,

1996a,b, 1997, 1999), was identical to the patient version

with the exception the spouses were asked to rate how

confident they were that their partners could engage in

behaviors to control pain, physical function, and other

arthritis symptoms. Alpha coefficients were conducted on

data collected on spouse total self-efficacy at pre-treatment

and post-treatment and were found to be high (alpha (pre-

treatment) ¼ 0.92; alpha (post-treatment) ¼ 0.95) support-

ing the internal consistency of this measure. The validity of

the spouse version of this measure is supported by prior

research in osteoarthritis patients that showed that spouses’

scores on self-efficacy are significantly related to patient

ratings of pain intensity and reports of use of pain coping

strategies (Keefe et al., 1997).

Marital adjustment. The Dyadic Adjustment scale

(Spanier, 1976) was used to assess marital adjustment.

This 32-item instrument was administered to patients. This

measure is one of the most widely used marital adjustment

measures and shows strong evidence of internal consist-

ency, test–retest reliability, and validity (Gotmman and

Krakoff, 1989).

Pain and psychological disability. The Arthritis Impact

Measurement Scales (AIMS) was used to assess pain and

psychological disability (Meenan et al., 1980). The AIMS

pain scale includes four items that assess the severity of

pain, frequency of severe pain, duration of morning

stiffness, and frequency of pain in multiple joints. The

AIMS psychological disability scale includes two individ-

ual scales (anxiety and depression). The AIMS is widely

used in studies of rheumatic disease patients and prior

research has provided strong support for its reliability and

validity (Kazis et al., 1983; Meenan et al., 1980; Meenan

et al., 1982).

1.3.1. Treatment conditions

1.3.1.1. General aspects. Patients in the spouse-assisted

CST conditions (SA-CST plus ET or SA-CST alone) met in

small groups (three to five couples) for 12 weekly, 2-h group

sessions. These group sessions were conducted by PhD level

psychologists who had been trained in spouse-assisted CST

and who followed a detailed treatment manual. The sessions

were audiotaped and the audiotapes were reviewed and

discussed in weekly supervision sessions. Patients in the

exercise training conditions (SA-CST plus ET, ET alone)

attended 60 min exercise sessions in small groups (three to

five patients) three times a week for 12 weeks. The exercise

sessions were run by BA level or above exercise physiol-

ogists who followed a treatment manual. Thus, the total

number of hours of treatment exposure was 24 for the

SA-CST alone condition, 26 h for the ET alone condition,

and 50 h for the SA-CST plus ET condition. Attendance by

patients and spouses at the treatment sessions was good

(range ¼ 85–92%). There were no significant differences in

patient or spouse attendance among the treatment conditions.

1.3.1.2. Spouse-assisted CST alone condition. Patients in the

SA-CST alone condition attended 12 weekly, 2-h group

sessions along with their spouses. The training was similar

to that used in our study of spouse-assisted CST (Keefe et al.,

1996a,b). Couples were provided with a rationale that

emphasized that: (1) pain is a complex experience, which as

the Gate Control Theory (Melzack and Wall, 1965)

suggests, can be influenced by thoughts, feelings, and

behaviors; (2) patients and their spouses can acquire and

maintain skills for managing pain through frequent practice;

and (3) because OA is a couples issue that affects each

partner and their relationship, involving the spouse in

training can be quite helpful.

The training sessions emphasized active learning. Thus,

participants were not only instructed in coping skills, but

also were encouraged to practice the skills alone and with

their partners both in the group and also at home. The group

leaders provided feedback and suggestions to enhance the

efficacy of skills practice. Each session involved two major

F.J. Keefe et al. / Pain 110 (2004) 539–549 541

components: (1) training in pain coping skills, and (2)

training in couples skills designed to supplement and

reinforce the patient’s pain coping skills. During the pain

coping skills component, patients were encouraged to

develop a menu of pain coping skills. To assist them,

systematic training was provided in attention diversion

skills (relaxation, imagery, and distraction), activity-based

skills (activity-rest cycling, pleasant activity scheduling),

and cognitive coping strategies (cognitive restructuring and

self-instructional methods for dealing with severe pain.)

During the couples skills training component, each couple

was encouraged to develop a menu of couples skills. To

achieve this goal, training was provided in wide range of

couples skills including communication skills, behavioral

rehearsal, mutual goal setting, joint home practice, and in

vivo practice.

1.3.1.3. Exercise training alone condition. Patients in the ET

alone condition attended three supervised group exercise

sessions per week for 12 consecutive weeks. Their spouses

did not attend the exercise sessions. A detailed description of

the ET protocol used in this study is provided elsewhere

(Keefe et al., 1996b.) Briefly, the exercise program included:

(1) cardiopulmonary endurance training, (2) strength train-

ing, and (3) flexibility/range of motion training. The data

gathered during the bicycle ergometry assessment and

maximum strength test were used to establish a baseline

level and to develop an individualized training program.

Patients participated in 30 min of aerobic training three days

per week. The intensity of aerobic training started at 50–70%

of heart rate reserve and gradually increased to 70–85% over

12 weeks. The aerobic training sessions included a warm up,

low intensity biking or walking, 30 min of continuous

aerobic activity (walking, biking, or water aerobics)

performed at or above patient’s prescribed training range,

and a cool down period. Patients also participated in 30 min

of strength training two days per week.

1.3.1.4. Spouse-assisted CST plus exercise training. Patients

in this condition attended 12 weekly 2-h SA-CST group

sessions along with their spouses and participated in the ET

intervention. Each weekly SA-CST session was held just

prior to one of the weekly ET sessions. The pain coping

skills and couples skills training methods used in this

condition were similar to those used in the SA-CST

condition. The major differences were: (1) more emphasis

was placed on the role that pain coping skills and couples

skills can play during exercise, and (2) the spouse attended

one exercise session with the patient each week where their

role was to coach the patient in applying pain coping skills

during exercise.

1.3.1.5. Standard care condition. The SC condition was

designed to serve as a routine treatment control. Patients

assigned to this condition continued to receive their routine

care. Neither they nor their spouses attended coping skills

training sessions or exercise sessions. These patients and

their spouses completed all measures at time intervals

corresponding to the beginning and end of the treatment

period.

2. Results

Data analysis was carried out in three steps: (1) an

evaluation of pre-treatment differences among the four

groups, (2) an evaluation of post-treatment differences

among the four groups, and (3) analysis of how changes in

self-efficacy related to improvements in the outcome

measures.

2.1. Pre-treatment comparisons among groups

Table 1 displays the unadjusted pre-treatment means

(and SDs) for each of the measures for patients in each of the

treatment groups. To examine pre-treatment differences

among these groups a series of analyses of covariance was

conducted. Covariates in these analyses were identical to

those used in our prior outcome studies of osteoarthritis

patients and included age, gender, and obesity status. These

three variables were used as covariates because they have

been found in our prior studies (Keefe et al., 1990a,b) to be

important in explaining pain and disability in patients

having osteoarthritis of the knees.

No group differences were found on any of the pre-

treatment measures except for the AIMS Pain measure

(Fð3; 67Þ ¼ 4:52; P ¼ 0:006). Post hoc analyses using

Tukey’s LSD revealed that patients in the spouse-assisted

CST alone and spouse-assisted CST plus exercise con-

ditions had higher pre-treatment levels of pain than those in

the two other conditions. Because of this difference, each

participant’s pre-treatment score on the AIMS Pain measure

was used as an additional covariate in all subsequent

outcome analyses.

2.2. Post-treatment differences among the groups

Table 1 presents the unadjusted post-treatment group

means (and SDs) for each of the measures. A series of

ANCOVAs was conducted to examine the effects of

treatment on these measures. In each ANCOVA, partici-

pants’ pre-treatment score on the relevant measure was used

as a covariate.

Effects of treatment on aerobic fitness and muscle

strength. An ANCOVA for Peak VO2K revealed a

significant post-treatment group (Fð3; 53Þ ¼ 4:67;

P ¼ 0:006) (Fig. 1). Paired comparisons between treatment

group means were made using Tukey’s LSD. These

analyses revealed that patients in the SA-CST plus ET

group showed significantly higher post-treatment Peak

VO2K than patients in the SC condition. Patients in the

SA-CST plus ET group and patients in the ET alone group

F.J. Keefe et al. / Pain 110 (2004) 539–549542

showed significantly higher post-treatment Peak VO2K than

patients in the SA-CST alone condition.

ANCOVAs were also conducted to examine post-treat-

ment group differences on the measures of muscle strength.

Significant group differences were found for the leg extension

(Fð3; 44Þ ¼ 7:02; P ¼ 0:001), leg flexion (Fð3; 49Þ ¼ 9:00;

P , 0:001), and bicep curl measures (Fð3; 46Þ ¼ 3:93;

P ¼ 0:014). Paired comparisons between treatment group

means were made using Tukey’s LSD. As can be seen in

Table 2, patients in the SA-CST plus ET group and patients in

the ET alone group showed significantly higher post-

treatment scores on leg extension and leg flexion than those

in the SA-CST alone and SC conditions. As can be seen in

Table 2, participants in the ET alone condition had

significantly higher bicep curl values compared to those in

the SC and SA-CST alone conditions and participants in

Table 1

Demographic data and pre- and post-treatment means of measures for patients in each of the treatment groups exercise

Variable Treatment group

Spouse-assisted CST Spouse-Assisted

CST þ exercise

Exercise Standard

care control

Age 60.00 (12.15) 60.20 (9.09) 60.25 (8.74) 57.56 (14.27)

Sex

Male 9 7 10 7

Female 9 13 6 11

Pre Post Pre Post Pre Post Pre Post

Peak VO2K 20.36

(6.77)

20.40

(7.21)

20.42

(5.48)

24.03

(5.88)

21.37

(5.74)

24.35

(5.97)

21.86

(6.16)

21.94

(7.99)

Coping factors

Coping attempts 69.36

(26.33)

78.18

(23.96)

55.19

(32.27)

73.76

(25.78)

49.35

(26.61)

47.44

(20.29)

57.22

(16.98)

51.01

(21.16)

Pain control and

rational thinking

1.17

(4.89)

3.56

(3.28)

1.39

(3.68)

3.81

(2.71)

0.72

(4.54)

1.50

(4.67)

3.39

(2.25)

2.62

(3.57)

Self-efficacy 201.93

(45.69)

234.13

(37.43)

196.68

(41.68)

238.71

(31.61)

215.42

(36.95)

220.46

(44.66)

224.08

(39.75)

224.17

(54.26)

Dyadic adjustment

scale

112.67

(15.55)

111.70

(18.04)

112.49

(14.51)

111.54

(12.04)

108.54

(21.34)

110.24

(23.41)

121.11

(17.37)

116.81

(18.48)

AIMS

Pain 5.44

(1.88)

4.00

(1.56)

5.20

(1.20)

4.26

(1.45)

3.91

(1.64)

3.19

(1.85)

3.91

(1.73)

4.03

(2.08)

Psychological

disability

2.83

(1.64)

2.38

(1.38)

2.57

(1.14)

2.21

(1.21)

2.36

(1.22)

1.88

(0.87)

1.85

(0.33)

1.80

(1.04)

Fig. 1. Adjusted post-treatment means for peak VO2K (with effect size) by treatment condition.

F.J. Keefe et al. / Pain 110 (2004) 539–549 543

Table 2

Adjusted post-treatment means (with effect size) for treatment condition

CA, coping attempts; PCRT, pain control and rational thinking.

F.J. Keefe et al. / Pain 110 (2004) 539–549544

the SA-CST plus ET group had significantly higher bicep curl

values than participants in the SC.

Effects of treatment on pain coping. An ANCOVA

revealed significant between group differences for patients’

scores on the Coping Attempts factor (Fð3; 57Þ ¼ 6:18;

P ¼ 0:001) of the CSQ. As can be seen in Table 2, pairwise

comparisons among the means revealed that patients in the

SA-CST alone condition and the SA-CST plus ET condition

had significantly higher post-treatment scores on Coping

Attempts than patients in the ET alone and SC conditions.

The ANCOVA for the Pain Control and Rational Thinking

factor of the CSQ also showed a significant group effect

(Fð3; 57Þ ¼ 3:00; P ¼ 0:038). As shown in Table 2,

participants in the SA-CST plus ET condition showed

significantly higher post-treatment scores on Pain Control

and Rational Thinking compared to participants in the ET

alone and SC conditions.

An ANCOVA also revealed significant between group

differences for spouses’ ratings of patients’ coping on the

Coping Attempts factor of the CSQ-S (Fð3; 51Þ ¼ 2:91;

P ¼ 0:043). Table 2 shows the adjusted post-treatment

means for spouses’ ratings of patients’ Coping Attempts.

Pairwise comparisons among the adjusted means revealed

that spouses in the SA-CST plus ET group rated their

partners as scoring significantly higher post-treatment on

Coping Attempts than patients in the ET alone and SC

conditions. Spouses in the SA-CST alone condition also

rated their partners as scoring significantly higher post-

treatment on Coping Attempts than patients in the SC

condition. The ANCOVA for the spouses’ ratings of

patients’ Pain Control and Rational Thinking was also

significant (Fð3; 54Þ ¼ 3:83; P ¼ 0:015). Table 2 shows

spouses in the SA-CST alone and SA-CST plus ET groups

rated their partners as scoring significantly higher post-

treatment on the Pain Control and Rational Thinking factor

than spouses in the ET alone condition.

Effects of treatment on self-efficacy measures. An

ANCOVA conducted to analyze the effects of treatment

on patient’s ratings of overall self-efficacy revealed a

significant treatment group effect (Fð3; 57Þ ¼ 4:37;

P ¼ 0:008). Fig. 2 displays the adjusted post-treatment

means for self-efficacy. As can be seen, patients in the

SA-CST plus ET condition and the SA-CST alone condition

showed significantly higher levels of self-efficacy than

patients in the SC condition. Patients in the SA-CST plus ET

condition also tended to report significantly higher levels of

self-efficacy than patients in the ET alone condition.

An ANCOVA conducted to analyze the effects of

treatment on spouse’s ratings of the patients overall self-

efficacy was also significant (Fð3; 55Þ ¼ 3:47; P ¼ 0:022).

As can be seen in Fig. 3, spouses in the SA-CST plus ET

condition and the SA-CST alone condition rated their

partners as showing significantly higher levels of total self-

efficacy than spouses of patients in the ET alone and SC.

Effects of treatment on marital adjustment. The effect of

treatment on marital adjustment was not significant

(Fð3; 56Þ ¼ 0:158; P ¼ 0:92).

Effects of treatment on pain and psychological disability.

The effects of treatment on AIMS Pain (Fð3; 58Þ ¼ 0:80;

P ¼ 0:50) or Psychological Disability (Fð3; 57Þ ¼ 0:33;

P ¼ 0:80) measures were not significant.

2.3. Relationship of changes in self-efficacy to outcome

A series of correlational analyses was conducted to

examine how changes in patients’ ratings of self-efficacy

related to changes in the outcome measures. Change scores

for total self-efficacy were calculated by subtracting pre-

treatment from post-treatment values. For patients in the

SA-CST plus ET condition, the results of the correlations

revealed that patients who showed increases in self-efficacy

were more likely to have improvements in psychological

Fig. 2. Adjusted post-treatment means for patient coping attempts (with effect size) by treatment condition.

F.J. Keefe et al. / Pain 110 (2004) 539–549 545

disability (r ¼ 20:64; P ¼ 0:003) and in scores on the pain

control and rational thinking factor of the Coping Strategies

Questionnaire (r ¼ 0:48; P ¼ 0:038). For patients in the ET

alone condition, increases in self-efficacy were significantly

related to increases in peak VO2K (r ¼ 0:58; P ¼ 0:029).

Finally, for patients in the SC condition, increases in self-

efficacy were related to decreases in physical disability

(r ¼ 20:53; P ¼ 0:04), and increases in scores on the pain

control and rational thinking factor of the Coping Strategies

Questionnaire (r ¼ 0:52; P ¼ 0:047).

3. Discussion

To our knowledge, no prior controlled studies of OA

patients have directly compared the effects of spouse-

assisted pain coping skills training and exercise interven-

tions. Our results show that each of these interventions

appear to have very specific benefits. Spouse-assisted

training, either alone or in combination with exercise

training, was found to produce improvements in coping

and self-efficacy, whereas exercise training, either alone or

in combination with spouse-assisted coping skills training

produced improvements in physical fitness and muscle

strength. Clinicians working with OA patients should be

aware that spouse-assisted CST and exercise have specific

benefits and that, when the primary clinical goal is to

improve self-efficacy and pain coping, spouse-assisted

CST may be particularly beneficial. Whereas, when the

primary clinical goal is to improve physical fitness and

muscle strength, exercise training may be particularly

beneficial.

The findings regarding specificity of treatment effects

have potentially important implications for multidisciplin-

ary or multicomponent treatment programs for patients

suffering from persistent pain. Our results suggest that when

one combines coping skills training and exercise interven-

tions, then improvements can be obtained across a much

broader range of outcomes (i.e. in coping, self-efficacy,

physical fitness, and muscle strength) than can be achieved

through either intervention alone.

The results of the present study agree with prior studies

in showing that exercise interventions can produce signifi-

cant improvements in physical fitness and muscle strength

in patients suffering from OA (van Baar et al., 2001; Gur

et al., 2002; Penninx et al., 2001; Petrella and Bartha, 2000).

These findings are important for several reasons. First, many

OA patients respond to knee pain by decreasing their

physical activity, a response that can lead to an overly

sedentary lifestyle, weight gain, and muscle weakness all of

which, in turn, can increase pain and disability (Felson and

Chaisson, 1997). Second, there is a growing consensus that,

by increasing and maintaining physical fitness and muscle

strength, exercise interventions may play a role in prevent-

ing disability in OA (AGS Panel on Exercise and

Osteoarthritis, 2001). Finally, there is growing recognition

that strategies for maintaining physical fitness and muscle

strength are very important in maintaining independence

and quality of life in older adults (Brill et al., 1998;

Fiatarone et al., 1990).

One of the most interesting findings of this study was

how changes in self-efficacy related to treatment outcome.

Patients who showed improvements in self-efficacy over the

course of exercise training were much more likely to show

increases in physical fitness. Improvements in self-efficacy

were also related to decreases in psychological disability

and improvements in Pain Control and Rational Thinking in

patients receiving the combination of spouse-assisted CST

and exercise training. Taken together, these findings

underscore the importance of self-efficacy in understanding

adjustment to arthritis and add to a growing literature

showing that changes in self-efficacy are associated with

outcomes following behavioral treatments for arthritis.

The results of this study found that an intervention that

combines spouse-assisted training in pain coping skills and

exercise training can improve physical fitness and muscle

Fig. 3. Adjusted post-treatment means for patient pain control and rational thinking (with effect size) by treatment condition.

F.J. Keefe et al. / Pain 110 (2004) 539–549546

strength and enhance pain coping and self-efficacy in

patients with OA of the knees. The present study is the first

to systematically involve spouses of patients undergoing

exercise training in an intervention designed to enhance

pain coping skills. Interestingly, spouses who participated in

this intervention reported significant increases in their

confidence (self-efficacy) that the patient could cope with

arthritis. They also reported that their partners showed

significant improvements in coping, both in the frequency of

coping attempts and in the perceived effectiveness of

coping.

Very few studies have systematically tested the effects of

involving spouses in coping skills training interventions for

patients with OA pain. In fact, the vast majority of pain

coping skills training studies conducted with persistent pain

conditions have focused on training the patient alone. There

is growing recognition that coping with painful diseases

such as OA takes place in a social context (Carr, 1999;

Reisine, 1995) and that many patients may prefer a

communal approach that involves them and their partner

in joint coping efforts (Sullivan et al., 2001). Patients with

OA engage in verbal and non-verbal behaviors commu-

nicating that pain is being experienced (Beaupre et al.,

1997) and these behaviors often provide a means of eliciting

spousal attention as well as instrumental and emotional

support (Sullivan et al., 2001). Couples-based training

programs that systematically engage partners in pain

management and provide opportunities for partners to

reinforce pain coping efforts can be helpful in meeting

patient’s interpersonal needs. As such, they represent a

potentially valuable addition to our treatment armementar-

ium. The findings of the current study agree with those

obtained in our prior work with OA in demonstrating that

spouse-assisted CST is effective in improve pain coping and

self-efficacy (Keefe et al., 1996a,b, 1999). Future studies

need to examine whether spouse-assisted CST is effective

for other pain conditions. Spouse- and partner-based

interventions may be particularly beneficial and appropriate

in clinical situations where the demands of pain on spouses

and caregivers are quite high (e.g. advanced cancer, brain

tumor patients, or in end-of-life care).

This study did not find evidence of treatment effects on

self-report measures of physical disability and psychologi-

cal disability. The fact that the exercise training produced

significant improvements in an objective measure of

physical fitness, but failed to show a significant effect on a

self-report measure of physical disability was surprising. It

is possible that the physical disability measure was less

sensitive to changes produced by the exercise training

intervention. Prior studies by our research team have shown

that interventions consisting of either coping skills training

alone (Keefe et al., 1990a) or SA-CST (Keefe et al., 1996a,b)

can produce significant improvements in psychological

disability in patients having OA of the knees. The present

study, however, did not find evidence SA-CST alone or

SA-CST plus ET produce changes in psychological

disability. One possible explanation for this discrepancy in

that the patients in this study were, prior to treatment,

experiencing somewhat lower levels of psychological

disability than patients in some of our prior studies.

Nevertheless, these SA-CST interventions tested in this

study (SA-CST alone, SA-CST plus ET) did produce

improvements in psychological outcomes such as self-

efficacy and pain coping. Considered overall, the findings of

this study suggest that additional research is needed to

determine the precise impact of SA-CST alone, ET, or SA-

CST plus ET on measures of physical disability and

psychological disability.

In terms of limitations, it should be noted that the patients

and spouses who participated were all volunteers with

generally high levels of marital satisfaction. The results, thus,

may not generalize to patients and spouses who have

significant marital problems. Future research needs to

examine the separate and combined effects of spouse-

assisted CST and exercise in couples who are more maritally

distressed. Based on research conducted in the marital

therapy area, it is possible that spouse-assisted interventions

would lead to even greater improvements in couples having

significant marital distress (Baucom et al., 1998).

The relatively small sample size of this study ðN ¼ 72Þ is

a limitation. This may have limited the ability of this study

to detect treatment effects that were potentially important,

but more moderate in magnitude. With a larger sample, for

example, we might have had sufficient power to detect

treatment effects on pain. Prior research conducted in our

lab has demonstrated that spouse-assisted pain coping skills

training can produce significant improvements in pain in

patients having OA of the knees (Keefe et al., 1996a,b).

Replication of the current study with a larger study sample

is warranted. However, the effect sizes in this study ranged

from 0.17 to 0.44 and thus generally fell in the range Cohen

(1988) considers as small (0.20) to medium (0.50)

suggesting that the treatment effects, in general, were not

as robust as other studies.

For purposes of experimental rigor and safety, volunteers

for this study were carefully screened. Of those who

volunteered, 30 were disqualified during the initial evalu-

ation process, mostly due to medical problems that

precluded their safe participation in exercise. This raises

concerns regarding the generalizability of the interventions

and findings to the broad population of patients with OA,

many of whom have medical problems and co-morbidities.

Careful medical screening and clearance, such as conducted

in this study, is considered to be the standard of care for

older adults who are candidates for formal exercise training.

Although persons who fail such screening may not be

appropriate for exercise interventions, they benefit from

other lifestyle interventions that reduce OA pain (e.g.

weight loss interventions). Furthermore, most of those

disqualified in this study could have safely participated in a

spouse-assisted coping skills training intervention so long as

this intervention was not combined with exercise training.

F.J. Keefe et al. / Pain 110 (2004) 539–549 547

Thus, the SA-CST intervention protocol used in this study

may be more easily adapted to treating a general population

of OA patients, than the exercise training intervention we

investigated.

Taken as a whole, the results of this study indicate that an

intervention combining spouse-assisted coping skills train-

ing and exercise can improve physical fitness, pain coping,

and self-efficacy in patients suffering from OA of the knees.

These results are interesting and support the need for further

research.

Acknowledgements

This research was supported by National Institute of

Arthritis and Musculoskeletal Diseases Grant No. AR-

35270.

References

American Geriatrics Society, Exercise prescription for older adults with

osteoarthritis pain: consensus practice recommendations. J Am Geriatr

Soc 2001;49:808–23.

van Baar ME, Assendelft WJ, Dekker J, Oostendorp RA, Bijlsma JW.

Effectiveness of exercise therapy in patients with osteoarthritis of the

hip or knee: a systematic review of randomized clinical trials. Arthritis

Rheum 1999;42:1361–9.

van Baar ME, Dekker J, Oostendorp RA, Voorn TB, Bijlsma JW.

Effectiveness of exercise in patients with osteoarthritis of the hip or

knee: nine months follow-up. Ann Rheum Dis 2001;60:1123–30.

Baucom DH, Shoham V, Mueser KT, Daiuto AD, Stickle TR. Empirically

supported couples and family therapies for adult problems. J Consult

Clin Psychol 1998;66:53–88.

Beaupre P, Keefe FJ, Lester N, Affleck G, Frederickson B, Caldwell DS. A

computer-assisted observational method for assessing spouses’ judg-

ments of osteoarthritis patients’ pain. Psychol Health Med 1997;2:

99–108.

Blumenthal JA, Emery CF, Walsh MA, Cox DR. Exercise training in

healthy Type A middle aged men: effects on behavioral and

cardiovascular responses. Psychosom Med 1988;50:418–33.

Blumenthal JA, Emery CF, Madden DJ, Schniebolk S, Walsh-Riddle M,

Goerge LK, Mckee D, Higginbotham MB, Cobb FRy, Coleman E. Long

term effects of exercise on psychological functioning in older men and

women. J Geront: Psychol Sci 1991;46:352–61.

Brandt KD. The role of analgesics in the management of osteoarthritis pain.

Am J Ther 2000;7:75–90.

Brill PA, Jensen RL, Koltyn KF, Morgan LA, Morrow JR, Keller MJ,

Jackson AW. The feasibility of conducting a group-based progressive

strength training program in residents of a multi-level care facility. Act

Adapt Aging 1998;22:53–63.

Buckwalter JA, Stanish WD, Rosier RN, Schenck RC, Dennis DA, Coutts

RD. The increasing need for nonoperative treatment of patients with

osteoarthritis. Clin Orthop Rel Res 2001;385:36–45.

Carr AJ. Beyond disability: measuring the social and personal conse-

quences of osteoarthritis. Osteoarthritis Cartilage 1999;7:230–8.

Chard J, Dieppe P. The case for nonpharmacologic therapy of osteoarthritis.

Curr Rheumatol Rep 2001;3:251–7.

Clyman B. Exercise in the treatment of osteoarthritis. Curr Rheumatol Rep

2001;3:520–3.

Cohen J. Statistical power analysis for the behavioral sciences. Hillsdale,

NJ: Lawrence Erlbaum; 1988.

Creamer P. Osteoarthritis pain and its treatment. Curr Opin Rheumatol

2000;12:450–5.

Felson DT, Chaisson CE. Understanding the relationship between body

weight and osteoarthritis. Baillieres Clin Rheumatol 1997;11:

671–81.

Felson DT, Naimark A, Anderson J, Kazis L, Castelli W, Meenan RF. The

prevalence of knee osteoarthritis in the elderly: the Framingham

osteoarthritis study. Arthritis Rheum 1987;30:914–8.

Felson DT, Lawrence RC, Hochberg M, C , McAlindon T, Dieppe PA,

Minor MA, Blair SN, Berman BM, Fries JF, Weinberger M, Lorig KR,

Jacobs JJ, Goldberg V. Osteoarthritis: new insights. Part 2. Treatment

approaches. Ann Int Med 2000;133:726–31.

Fiatarone MA, Marks EC, Ryan ND, Meredith CN, Lipsitz LA, Evans WJ.

High-intensity strength training in nonagenarians. J Am Med Assoc

1990;263:3029–34.

Gottman JM, Krakoff LJ. Marital interaction and satisfaction: a longitudinal

view. J Consult Clin Psychol 1989;57:47–52.

Gur H, Cakin N, Akova B, Okay E, Kucukoglu S. Concentric versus

combined concentric-eccentric isokinetic training: Effects on functional

capacity and symptoms in patients with osteoarthrois of the knee. Arch

Phys Med Rehabil 2002;83:308–16.

Hochberg MC, Lawrence RC, Everett EF, Cornoni-Huntly J. Epidemio-

logic association of pain in osteoarthritis of the knee: data from the

National Health and Nutrition Examination Survey and the National

Health and Nutrition Examination. I. Epidemiologic follow-up survey.

Semin Arthritis Rheum 1989;18:4–9.

Hochberg MC, Altman RD, Brandt KD, Clark BM, Dieppe PA, Griffin MR,

Moskowitz RW, Schnitzer TJ. Guidelines for the medical management

of osteoarthritis. Part II. Osteoarthritis of the knee. American College of

Rheumatology. Arthritis Rheum 1995;38:1541–6.

Kazis LE, Meenan RF, Anderson J. Pain in the rheumatic diseases:

investigations of a key health status component. Arthritis Rheum 1983;

26:1017–22.

Keefe FJ, Caldwell DS, Queen KT, Gil KM, Martinez S, Crisson JE, Ogden

W, Nunley J. Pain coping strategies in osteoarthritis patients. J Consult

Clin Psychol 1987a;55:208–12.

Keefe FJ, Caldwell DS, Queen KT, Gil KM, Martinez S, Crisson JE, Ogdan

W, Nunley J. Osteoarthritic knee pain: a behavioral analysis. Pain

1987b;28:309–21.

Keefe FJ, Caldwell DS, Williams DA, Gil KM, Mitchell D, Robertson C,

Martinez S, Nunley J, Beckham JC, Crisson JE, Helms M. Pain coping

skills training in the management of osteoarthritic knee pain: a

comparative study. Behav Ther 1990a;21:49–62.

Keefe FJ, Caldwell DS, Williams DA, Gil KM, Mitchell D, Robertson C,

Martinez S, Nunley J, Beckham JC, Crisson JE, Helms M. Pain coping

skills training in the management of osteoarthritic knee pain: follow-up

results. Behav Ther 1990b;21:435–48.

Keefe FJ, Caldwell DS, Baucom D, Salley A, Robinson E, Timmons K,

Beaupre P, Weisberg J, Helms M. Spouse-assisted coping skills training

in the management of osteoarthritis knee pain. Arthritis Care Res 1996;

9:279–91.

Keefe FJ, Kashikar-Suck S, Opiteck J, Hage E, Dalrymple L, Blumenthal J.

Pain in arthritis and musculoskeletal disorders: The role of coping skills

training and exercise interventions. J Orthop Sports Phys Ther 1996;24:

279–90.

Keefe FJ, Kashikar-Zuck S, Robinson E, Salley A, Beaupre P, Caldwell D,

Baucom D, Haythornthwaite J. Pain coping strategies that predict

patients’ and spouses’ ratings of patients’ self-efficacy. Pain 1997;73:

191–9.

Keefe FJ, Caldwell DS, Baucom D, Salley A, Robinson E, Timmons K,

Beaupre P, Weisberg J, Helms M. Spouse-assisted coping skills training

in the management of knee pain in osteoarthritis: Long-term followup

results. Arthritis Care Res 1999;12:101–11.

Keefe FJ, Smith SJ, Buffington ALH, Gibson J, Studts J, Caldwell DS.

Recent advances and future directions in the biopsychosocial assess-

ment and treatment of arthritis. J Consult Clin Psychol 2002;70:

640–55.

F.J. Keefe et al. / Pain 110 (2004) 539–549548

Lawrence JS. Generalized osteoarthritis in a population sample. Am J

Epidemiol 1960;90:381–9.

Lawrence JS, Bremer JM, Bier F. Osteoarthritis: prevalence in the

population and relationship between symptoms and x-ray changes.

Ann Rheum Dis 1966;25:1–24.

Lorig K, Gonzalez V. The integration of theory with practice: a 12 year case

study. Health Educ Q 1992;19:335–68.

Lorig K, Chastain RL, Ung E, Shoor S, Holman HR. Development and

evaluation of a scale to measure perceived self-efficacy in people with

arthritis. Arthritis Rheum 1989;32:37–45.

Meenan RF, Gertman PM, Mason JH. Measuring health status in arthritis:

the arthritis impact measurement scales. Arthritis Rheum 1980;23:

146–52.

Meenan RF, Gertman PM, Mason JH, Dunait R. The arthritis impact

measurement scales: further investigations of a health status measure.

Arthritis Rheum 1982;23:1048–53.

Melzack R, Wall P. Pain mechanisms: a new theory. Science 1965;50:971–9.

Parker JC, Smarr KL, Buescher KL, Phillips LR, Frank RG, Beck NC,

Anderson SK, Walker SE. Pain control and rational thinking:

Implications for rheumatoid arthritis. Arthritis Rheum 1989;32:

984–90.

Parker JC, Bradley LA, Devellis RM, Gerber LH, Holman HR, Keefe FJ,

Lawrence TS, Liang MH, Lorig KR, Nicassio PM, Revenson TA,

Rogers MP, Wallston KA, Wilson MG, Wolfe F. Biopsychosocial

contributions to the management of arthritis disability: proceedings

from an NIDRR-sponsored conference. Arthritis Rheum 1993;36:

885–9.

Penninx BW, Messier SP, Rejeski WJ, Williamson JD, DiBari M,

Cavazzinei C, Applegate WB, Pahor M. Physical exercise and the

prevention of disability in activities of daily living in older persons with

osteoarthritis. Arch Int Med 2001;161:2309–16.

Petrella RJ. Is exercise effective treatment for osteoarthritis of the knee? Br

J Sports Med 2000;34:326–31.

Petrella RJ, Bartha C. Home based exercise therapy for older patients with

knee osteoarthritis: a randomized clinical trial. J Rheum 2000;27:

2215–21.

Reisine ST. Arthritis and the family. Arthritis Care Res 1995;8:265–71.

Rosenstiel AK, Keefe FJ. The use of coping strategies in chronic low back

pain patients: relationship to patient characteristics and current

adjustment. Pain 1983;17:33–44.

Schiaffino KM, Revenson TA, Gibofsky A. Assessing the impact of self-

efficacy beliefs on adaptation to rheumatoid arthritis. Arthritis Care Res

1991;4:150–7.

Schnitzer TJ. Osteoarthritis treatment update. Postgrad Med 1993;93:

89–95.

Spanier GB. Measuring dyadic adjustment: new scales for assessing the

quality of marital and similar dyads. J Marriage Fam 1976;38:15–27.

Sullivan MJL, Thorn B, Haythornthwaite J, Keefe FJ, Martin M, Bradley L,

Lefebvre JC. Theoretical perspectives on the relation between

catastrophizing and pain. J Clin Pain 2001;17:52–64.

Toda Y. The effect of energy restriction, walking, and exercise on lower

extremity lean body mass in obese women with osteoarthritis of the

knee. J Orthop Sci 2001;6:148–54.

F.J. Keefe et al. / Pain 110 (2004) 539–549 549