JOINT EFFECTS: WATER EXERCISE AND MANUAL THERAPIES FOR ARTHRITIS MELAINIE CAMERON THESIS SUBMITTED TO SATISFY THE REQUIREMENT FOR THE DOCTOR OF PHILOSOPHY SCHOOL OF HUMAN MOVEMENT, RECREATION AND PERFORMANCE VICTORIA UNIVERSITY 20 JANUARY 2005 FTS THESIS 616.72206 CAM 30001008090203 Cameron, Melainie Joint effects : water exercise and manual therapies for arthritis ABSTRACT This thesis was an investigation of the relative effects of manual therapies and warm-water exercise on the health-related quality of life (HRQOL) of people with common arthritides (rheumatoid arthritis [RA] or osteoarthritis [OA]). The thesis comprised a pilot study and three linked clinical studies. The three clinical studies were randomised, unblinded, controlled, clinical trials of three or four groups with repeated measures. The mixed design allowed comparison between and within groups. The intervention component of each clinical study lasted for 10 weeks (concurrent with one Victorian school term). Each clinical study included adjunctive therapy and control (usual care) groups, to which participants were randomly allocated. Participation in each of the studies was voluntary. Participants were free to withdraw from any study at any time. Participants who were unable or unwilling to participate in the intervention groups were re-allocated to the control group or withdrew from the studies. The Pilot Study (see Chapter 3) concerned validation of the data collection tools prior to use in the clinical studies. A battery of standardised, validated questionnaires was used for data collection in each of the clinical studies. Measures of generic quality of life (SF-36; Ware & Sherboume, 1992), pain (Short Form McGill Pain Questionnaire, SF- MPQ; Melzack, 1987), arthritis-specific health status (AIMS2; Meenan, Mason, Anderson, Guccione, & Kazis, 1992) and social support (Medical Outcomes Study Social Support Survey, MOS-SS; Sherboume & Stewart, 1991) were collected at baseline, week 5 and week 9 of each of the 10-week trials, and at 2 weeks and 14 weeks after completion of the interventions in Study 3 (i.e., weeks 12 and 24). Analyses of covariance (ANCOVA) were used to assess group differences for the SF-36, AIMS2, SF-MPQ, and MOS-SS measures. Pre-intervention (i.e., week 1, baseline) scores for each HRQOL subscale and total social support at week 9, were used as 11 covariates to control for initial differences between groups and any social support afforded by the interventions. In Study 1 (people with OA) improvements in HRQOL were consistently observed in the joint mobilisation group, and on many HRQOL domains, these improvements were associated with large to very large effect sizes. Participants in the massage group improved only moderately compared with the control group across the same measures. Results differed according to disease profiles. Several participants with RA assigned to manual therapy groups reported worsening pain and withdrew from Study 2. Results from Study 2 were inconclusive, and hampered by small sample sizes. Reasons for, and lessons arising from, the failure of Study 2 are discussed in detail (see Chapter 5). Results from Studies 1 and 2 informed the design of Study 3, which did not include a massage group or any people with RA. In Study 3, 22 adults with an average of 15 years of osteoarthritis were randomly assigned to usual care (control; n = 4), joint mobilization (n = 4), warm-water exercise {n = 8), or combined joint mobilization and warm-water exercise {n - 6) groups. At week 9 participants in the intervention groups reported better HRQOL across most subscales than participants in the control (usual care) group. The combined therapies group outperformed the control and single therapy groups on the mobility, household tasks, arthritis pain, mood, and satisfaction subscales of the AIMS2, as well as the sensory pain, total pain, and present pain index components of the SF-MPQ, and the physical role limitations, bodily pain, general health, social function, and health transition subscales of the SF-36. Many of the improvements in HRQOL reported at week 9 were maintained at week 12 and week 24 (2 and 14 weeks post intervention). Repeated measures analyses of covariance (ANCOVA), using baseline measures and week 9 social support scores as covariates, revealed that large to very large effects (improvements) on the arthritis pain (r| = .25), mood {y\ = .35), and satisfaction {r\ = .21) Ill subscales of the AIMS2 could be attributed to participation in the combined therapies. The same pattern was evident for the sensory pain (ri^ = .29), total pain (r|^ = .23), and present pain index (T] = .37) components of the SF-MPQ, and the physical role limitations (r| = .26), bodily pain (r\ = .18), social function (r| = .33), and health transition (r| = .28) subscales of the SF-36. Differences in social support, and medication use, across time and between groups were negligible, and do not account for the reported improvements. The results are interpreted, as recommended by Kazis, Anderson, and Meenan (1989), in terms of clinically meaningful effect sizes, rather than statistical significance, due to the small sample size and the increased probability of Type II errors. Omnibus effect sizes are reported as r| . Eta squared (TJ ) represents the amount of variance in a variable accounted for by group membership (Tabachnick & Fidell, 2001), and is therefore a relevant measure of effect size because it explains the strength of association between treatments and the variables measured. Univariate effect sizes (Cohen's d) are reported as estimates of the magnitude of change between and within groups for each intervention and each HRQOL domaia Overall the results indicate that the combination of joint mobilisation and warm- water exercise appears to be more effective than either therapy in isolation for improving quality of life in people with OA. The usefulness of combined therapies needs to be balanced against the financial costs of the same. The persistence of HRQOL improvements at 2 weeks post therapy suggests that fortnightly therapy is worthy of investigation, an approach that would make combined therapies more affordable. IV TABLE OF CONTENTS ABSTRACT TABLE OF CONTENTS IV DECLARATION OF AUTHORSHIP xi ACKNOWLEDGEMENTS xii DISCLOSURE xiii LIST OF TABLES AND FIGURES xiv LIST OF APPENDICES xvi LIST OF PUBLICATIONS ARISING FROM THESIS xvii CHAPTER 1: INTRODUCTION 1 Treatments for Arthritis 2 Health-Related Quality of Life 5 Why Do this Research? 6 Significance of the Research 6 Contribution to Knowledge 7 CHAPTER 2: REVIEW OF LITERATURE 8 Overview of Arthritis 8 Epidemiology 8 Incidence 8 Prevalence 9 Types of Arthritides 10 Diagnoses and Diagnostic Criteria 14 Treatment and Management of Arthritides 19 Usual Medical Care 19 Monitoring Arthritis 22 Joint Counts 23 Serology 24 Imaging 25 Osteoarthritis 25 Rheumatoid arthritis 26 Magnetic resonance imaging 27 Self-report Measures 28 Improvement Criteria 29 Allied Medical, Complementary, and Alternative Therapies 32 Active Therapies: Exercise 33 Passive Therapies: Manual Therapy 36 Physical and Manual Therapies: Possible Mechanisms of Action 39 Health-Related Quality of Life 40 Impairment, Disability, and Handicap 42 Psychological Considerations 44 Psychological Constructs Important in Arthritis Care 45 Working alliance 45 Locus of control 47 Self-efficacy 49 Stages of change 49 Life events. Disease onset. Diagnosis, and Psychological Health 52 Pain 54 Pain Behaviour 56 The Sick Role: Joint Protection as a Case Study 57 Depression 59 Anxiety 60 Loss and Grief 61 VI Loss of Bodily Function 62 Loss of Social Role 62 Economic Loss 64 Loss of Social Support 65 Body Image 66 Psychological Interventions in Physical and Manual Therapies 67 Measurement and Statistical Issues 69 Measuring Health-Related Quality of Life 69 Statistical Concerns 70 Limitations of these Studies 72 CHAPTER 3: PILOT STUDY 73 Introduction 73 Method 75 Participants 75 Measures 75 AIMS2 75 SF-MPQ 77 SF-36 79 MOS-SS 80 Medication Use Survey 81 Procedures 81 82 Data Analysis 82 Results 87 Discussion 87 Participants' Feedback 89 Recall vu Health-Related Quality of Life 89 Impact of Arthritis: AIMS2 90 General Health-Related Quality of Life: SF-36 91 Pain 92 How This Study Informed Subsequent Studies 93 Conclusions 94 CHAPTER 4: STUDY 1 96 Introduction 96 Method 97 Participants 97 Measures 97 Compliance data 98 Procedures 98 Data Analysis 100 Results 102 Missing data 102 Floor and ceiling effects 103 Analysis of the Covariates 103 Medication use 104 Social support 104 Changes Within and Between Groups 105 Within-Groups Analyses 105 Between-Groups Analyses 108 Discussion 114 Health-Related Quality of Life 114 Covariate Measures 115 Vlll Medication use 115 Social support 115 Impact of Arthritis: AIMS2 115 General Health-Related Quality of Life: SF-36 117 Pain 119 Social Functioning: Disparity Between Instruments 120 Clinical Implications 121 Costs and Benefits 122 Clinical Cautions and Limitations 124 Recommendations for Future Research and Conclusions 126 How this Study Informed Study 3 126 Future Research 126 Conclusions 126 CHAPTER 5: STUDY 2 128 Introduction 128 Method 129 Participants 130 Measures 130 131 Compliance data 131 Procedures 133 Issues Arising from the Research Design 137 Data Analysis and Results 137 Data Analysis 138 Issues Arising from the Data Analysis and Results 140 Recommendations for Future Research and Conclusions 142 CHAPTER 6: STUDY 3