Are cognitive and behavioural factors associated with knee pain? A systematic review

Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]

Contents lists available at ScienceDirect

Seminars in Arthritis and Rheumatism journal homepage: www.elsevier.com/locate/semarthrit

Are cognitive and behavioural factors associated with knee pain? A systematic review Donna M. Urquhart, BPhysio (Hons), PhD1, Pyae P. Phyomaung, MBBS, MPH1, Julia Dubowitz, MBBS, Sanduni Fernando, BBMedSc (Hons), MBBS, Anita E. Wluka, MBBS, PhD, FRACP, Paul Raajmaakers, MSc, Yuanyuan Wang, MD, PhD, Flavia M. Cicuttini, MBBS, FRACP, PhDn Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, Alfred Hospital, Commercial Rd, Melbourne 3004, Victoria, Australia

a r t i c l e in fo

Keywords: Cognitive factors Behavioural factors Knee pain Osteoarthritis Systematic review Psychosocial factors

a b s t r a c t Objectives: Although studies that have examined the relationship between cognitive and behavioural factors and knee pain report conflicting results, no systematic review has been performed to summarise the evidence. The aim of this systematic review was to examine the relationship between cognitive and behavioural factors and pain at the knee. Methods: Electronic searches of MEDLINE, EMBASE and PsycINFO were performed to identify relevant studies published up to April 2014 using MeSH terms and keywords. Studies that met a set of predefined criteria were included. Coping, self-efficacy, somatising, pain catastrophising and helplessness were grouped together as “cognitive factors,” while kinesiophobia and pain-related fear-avoidance were considered “behavioural factors.” Two independent reviewers extracted the data and assessed the methodological quality of the selected studies. Due to the heterogeneity of the studies, a best-evidence synthesis was performed. Results: A total of 14 studies were included in the review, of which nine examined cognitive factors, one investigated behavioural factors and four studied both cognitive and behavioural factors. Eight of 14 studies were of high quality. The best-evidence synthesis showed moderate evidence for a relationship between cognitive factors and knee pain and limited evidence for no association between the behavioural factors and knee pain. Conclusion: This review found evidence for a relationship between cognitive factors, but not behavioural factors, and knee pain. These findings will need to be confirmed with high-quality longitudinal studies, but the data suggest that cognitive factors may be important to target in the management of knee pain. & 2014 Elsevier Inc. All rights reserved.

Introduction Knee pain is a common clinical symptom, with almost half of those aged 50 years and over reporting pain at the knee, and 25% of them experiencing symptoms of a chronic nature [1]. Moreover, one in six individuals with knee pain will consult a doctor in the course of a year and one-third of them will have severe pain and disability [2]. The main underlying cause of knee pain is

Sources of support: D.U. and A.W. were supported by NHMRC Fellowships, Australia, (1011975 and 545876, respectively). n Corresponding author. E-mail addresses: fl[email protected], [email protected] (F.M. Cicuttini). 1 Joint first authors. http://dx.doi.org/10.1016/j.semarthrit.2014.07.005 0049-0172/& 2014 Elsevier Inc. All rights reserved.

osteoarthritis, which imposes a significant health care burden. However, even in the setting of knee osteoarthritis, it is becoming increasingly evident that structural changes alone do not account for all knee pain [3]. Rather, knee pain is a complex interaction of factors, including structural damage, peripheral and central pain processing mechanisms, culture, sex and psychosocial factors [4]. The role of cognitive and behavioural factors in chronic musculoskeletal pain has been investigated over several decades. Behavioural concepts were first described by Fordyce [5] who highlighted a distinction between the original cause of pain and pain behaviours, which were defined as reports of or displays of pain that may persist for longer than normal expected healing time. The concept of cognitive constructs, which include painrelated thoughts and beliefs, developed and was combined with behaviour theories to form the cognitive behavioural model [6].

2

D.M. Urquhart et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]

The model holds the assumption that chronic pain is multifactorial and is influenced by a variety of physical and psychosocial factors, including the way individuals react to their pain (their behaviour), which is a product of their thoughts and beliefs (cognitions). Cognitive and behavioural factors have been shown to be predictors of pain and disability in a variety of musculoskeletal conditions, including chronic low back pain [7] and neck pain [8]. It has also been suggested that cognitive and behavioural factors are involved not only in the pathogenesis of acute to chronic musculoskeletal pain but in the progression of chronicity of pain [9]. While two systematic reviews have specifically examined one or two psychosocial factors, along with a number of demographic, physical and patient-related factors, as prognostic factors for knee pain [10,11], no systematic review has specifically focused on examining the relationship between cognitive and behavioural factors and pain at the knee. Moreover, a recent comparative study found that patients with knee OA who received training in coping compared to arthritis education or standard care had lower levels of pain, physical disability and pain behaviour at six months follow-up [12]. Understanding the role of psychosocial factors in knee pain is important in the development of non-pharmacological management strategies for the management of this debilitating condition. The aim of this review was to systematically review the literature to determine whether cognitive (coping, self-efficacy, somatization, pain catastrophising and helplessness) and behavioural (kinesiophobia and pain-related fear-avoidance) factors are associated with knee pain.

Methods A systematic review was conducted according to the 2009 PRISMA statement [13].

Data sources and search strategy Electronic searches of MEDLINE, EMBASE and PsycINFO were performed to identify studies that examined the relationship between psychosocial factors and knee pain up to April 2014 using the following MeSH terms: “knee pain” and “knee osteoarthritis” and the following keywords: “knee,” “osteoarthritis,” “pain,” “psychosocial,” “psychosomatic,” “psychological” and “psychophysiologic.” The search was limited to human studies of adults, published in the English language.

Inclusion and exclusion criteria Studies were included if they examined cognitive factors, including coping, self-efficacy, somatisation, pain catastrophising and helplessness and/or behavioural factors, including kinesiophobia (pain-related fear of movement) and pain-related fearavoidance, in relation to knee pain outcomes. A number of the studies examined the constructs of anxiety or depression; however, these have been systematically reviewed in a previous study [14]. Studies on knee pain were included whether or not knee osteoarthritis was specified. The following were the exclusion criteria: (1) studies that did not separate knee pain from pain in other regions, such as the hip and back; (2) studies investigating the reverse outcome (i.e., the effect of pain on psychosocial health); (3) studies that did not focus on pain at the knee as the outcome; (4) study participants who had inflammatory arthritis or other associated medical conditions affecting joints; (5) study populations who had undergone knee surgery and (6) studies examining depression and anxiety as psychosocial factors.

Data extraction Data on the characteristics of the included studies were extracted, including (1) study design, study population, number of participants, mean age, percentage of female participants, definition of OA and previous knee injury; (2) methods of assessment of cognitive (coping, self-efficacy, somatisation, pain catastrophising and helplessness) and behavioural (kinesiophobia and pain-related fear-avoidance) factors; (3) outcome measures assessment of knee pain and (4) study results (Tables 3–5). Methodological quality assessment The methodological quality of each study was assessed independently by two reviewers (J.D. and S.F.) using the standard criteria adapted from the study by Lievense et al. [15]. Scores were compared and a consensus score was obtained for each study (Table 1). Some of the standard criteria pertain specifically to either case–control or cohort studies and therefore could not be applied to every study. Only relevant criteria were included in calculations of the total and percentage mean quality score for each study. Any study that obtained a score above the mean of 68.3% was considered to be of high quality. Data synthesis Due to heterogeneity in methodology between the studies, the decision was made to use a best-evidence synthesis to summarise the data [15] (Table 2). Studies were ranked according to their design, with cohort studies considered to be of a higher level of Table 1 Criteria used to assess the methodological quality of selected cohort and crosssectional studies Item

Study population 1 2 3 4 5 Assessment of risk factor 6 7 8 Assessment of outcome 9 10 11

Criteria

Study type

Selection before disease was present or at uniform point Cases and controls were drawn from the same population Participation rate Z80% for cases/cohort Participation rate Z80% for controls Sufficient description of baseline characteristics

CH/CC/CS

Psychosocial assessment was blinded Psychosocial factors were measured identical for cases and controls Psychosocial factors were assessed prior to the outcome

CH/CC/CS CC

OA was assessed identical in studied population Presence of OA/pain was assessed reproducibly Presence of OA/pain was assessed according to standard definitions

CH/CC/CS

Study design 12 Prospective design was used 13 Follow-up time Ztwo years 14 Withdrawals r20% Analysis and data presentation 15 Appropriate analysis techniques were used 16 Adjusted for at least age and sex

CC CH/CC/CS CC CH/CC/CS

CH/CC/CS

CH/CC/CS CH/CC/CS

CH/CC/CS CH CH

CH/CC/CS CH/CC/CS

CH—applicable to cohort studies; CC—applicable to case–control studies; CS—applicable to cross-sectional studies; OA—osteoarthritis.

D.M. Urquhart et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]] Table 2 Criteria list for determining the level of evidence for best-evidence synthesis, adapted from Lievense et al. [15] Level of evidence

Criteria for inclusion in best-evidence synthesis

Strong evidence

Generally consistent findings in multiple high-quality cohort studies Generally consistent findings in 1 high-quality cohort study and 4 2 highquality case–control studies 43 high-quality case–control studies Generally consistent findings in single cohort study 1 or 2 case–control studies or multiple cross-sectional studies Inconsistent findings in o 75% of the trials No studies could be found

Moderate evidence

Limited evidence

Conflicting evidence No evidence

evidence than case–control and cross-sectional design. Since not all criteria were appropriate for all study designs, only relevant items contributed to the total score for each study. The level of evidence of the studies was determined in conjunction with the quality score calculated for each study.

Results Identification and selection of the literature A total of 754 studies were identified from our electronic database search and 52 were potentially eligible for selection after excluding case reports, review articles, studies involving surgical interventions and studies that examined cognitive and behavioural

3

factors as dependent variables. Of the 52 studies, 27 studies were excluded as they examined psychosocial factors that were not cognitive or behavioural. A further 11 studies were excluded as they examined an economic evaluation of a self-management programme [16], the pain experience [17], ethnicity [18], musculoskeletal pain (Refs.), knee pain that was not grouped separately from other musculoskeletal pain [19], walking speed [20], OA of various joints except the knee [21–23] and prediction of somatisation [24]. Of the two remaining studies, one was a validation study and the other a literature review [25] (Fig). Characteristics of included studies A total of 14 studies were included (Table 3). Of them, 10 were cross-sectional studies [26–35], one was a nested case–control study [36], two were longitudinal studies [37,38] and one was a quasi-experimental study [39]. Seven studies [27,29–31,35–37] examined knee pain in general, while seven studies [26,28,32– 34,38,39] examined knee pain in patients with a diagnosis of OA. Of the studies included in this review, two were undertaken in the United Kingdom [35,36], five in the United States of America [26,28,33,34,37], two in The Netherlands [32,38] and one each in Japan [31], Taiwan [39], Sri Lanka [30], Spain [27] and New Zealand [29]. Participants were recruited from general practitioner's surgeries or community health care clinics in four studies [32,35,38,39], various occupational groups or office workers in three studies [29–31], a previous longitudinal study in two studies [26,36], community-based individuals in two studies [28,33], outpatient clinic in two studies [27,34] and physiotherapy clinic in a single study [37]. The mean age of participants ranged from 31.2 to 69.3 years. The percentage of females in the studies varied from 32% to 88%.

Fig. Flow diagram of included and excluded studies according to the PRISMA statement.

4

Table 3 Characteristics of studies References

Study population

Cross-sectional studies Creamer et al. Recruited from the Baltimore [26] Longitudinal Study of Aging; community-based individuals 440 years of age

No. of participants (% women)

374 (32)

Age, mean 7 SD (range)

Previous hip or knee injury

Diagnosis of osteoarthritis

Pain assessment

Psychosocial factor assessment

Quality score

Men: 63.8 7 0.80 Women: 62.8 7 1.08

NA

K/L score

NHANES

54

32 (10)

Excluded

NA

VAS

VAS AIMS Self-reported knee pain lasting for more than a day in the month before the survey Knee pain: CUPID Questionnaire

Self-efficacy: ASES (pain subscale) Helplessness: helplessness subscale of RAI Pain catastrophising: pain catastrophising scale Self-efficacy: SelfEfficacy Scale Self-efficacy: SelfEfficacy Scale Somatising tendency: BSI

64

73

Recruited from hospital outpatients clinics in orthopaedic surgery during a one-year period

97 (83)

Marks [28]

Community-recruited urbandwelling North American adults Randomly selected nurses, postal workers and office workers using computers Subjects identified by random sampling across four occupational groups Nurses, office workers, sales/ marketing personnel and transportation operatives Participants presenting to their GPs with knee OA Subjects with knee OA recruited through a university medical centre, flyers posted in the community and advertisements in local newspapers

100 (82)

67.9 (12.5)

NA

443 (NA)

NA (range: 20–59)

NA

Clinical criteria of the ACR NA

852 (58)

31 (range 16–59)

NA

NA

NA (range: 19–64)

NA

NA

Self-reported knee pain in the past month and past year

Somatising: BSI

82

112 (88.4)

69.3 7 8.1

174 (82)

57.7 7 9.8

NA

Clinical criteria of the ACR

AIMS

Pain coping: PCI Beliefs: FABQ Pain catastrophising: CSQ Pain-related fear: Tampa Scale of Kinesiophobia Self-efficacy: ASES

64

58.7 7 9.2

Clinical criteria of the ACR Clinical criteria of the ACR

Severity of pain: VAS

106 (77)

Excluded if pathology explained the complaint NA

48 7 1.8

NA

NA

Pain in the knee lasting Z one day during past 12 months

Somatising BSI

70

Cases: 66.3 7 9.2 Controls: 64.6 7 8.2

Previous knee surgery n (%): 26 (9.1)

NA

Pain intensity: Chronic Pain Grade Pain extent: areas of pain experienced in previous month shaded on wholebody manikin Night pain: single item on WOMAC

Coping and appraisal: 1-item CSQ

78

Harcombe et al. (2010) [29] Warnakulasuriya et al. [30] Matsudaira et al. [31] van Baar et al. [32] Somers et al. [33]

Pells (2008) [34]

Palmer (2007) [35]

Subjects with knee OA recruited through the Rheumatology, Orthopaedic Surgery and Pain Management clinics Subjects were selected from registers of five GP clinics

Nested case–control studies Peat (2009) [36] Recruited from the Clinical Assessment Study of the Knee

2290 (32)

1798 (979)

285 (55)

AIMS

64 72

70

64

D.M. Urquhart et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]

Domenech et al. [27]

NHANES—National Health and Nutritional Examination Survey; PFS—Physical Functioning Score; WOMAC—Western Ontario and McMaster University Osteoarthritis Index; PCI—Pain Coping Inventory; CSQ—Coping Strategies Questionnaire; FABQ—Fear-Avoidance Beliefs Questionnaire; ASES—Arthritis Self-efficacy Scale; RAI—Rheumatology Attitudes Index; CUPID—Cultural and Psychosocial Influences on Disability Questionnaire; BSI—Brief Symptoms Inventory; NPRS—Numerical Pain Rating Scale; SOPA—Survey of Pain Attitudes; VAS—Visual Analogue Scale; OA—osteoarthritis; K/L scale—Kellgren and Lawrence Atlas of Standard Radiographs of Arthritis; AIMS—Arthritis Impact Measurement Scales; ACR—American College of Rheumatology; NA—not available.

69 Self-efficacy: ASES Self-ASE Pain Scale Pain Beliefs SOPA-35 Clinical criteria of the ACR I: 67 7 10.05 C: 68 7 11.2 217 (70) Quasi-experimental studies Wu (2011) (knee Randomly selected patients from OA) [39] Community Health Centres

68.5 7 8.9 119 (84.0) Recruited from a randomised control trial with the diagnosis of knee OA Steultjens (2001) [38]

Excluded if previous knee replacement surgery

Coping: PCI Fear-avoidance beliefs: FABQ Overall pain in the last week: VAS (0–100) Clinical criteria of the ACR

69

Fear-avoidance beliefs: FABQ Pain intensity: 11-point NPRS Median 26 (25th– 75th percentile 22–38) Longitudinal studies Piva (2009) [37] Subjects aged between 12 and 50 years referred to physiotherapy with a primary diagnosis of PFPS

74 (53)

Excluded if previous patellar dislocation, knee surgery over the last two years, ligamentous knee injury or laxity and internal knee derangement NA

NA

64

D.M. Urquhart et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]

5

Four studies excluded participants with previous injury, surgery or diagnosed knee pathology. Assessment of cognitive factors Cognitive factors were examined using a variety of methods. Coping was assessed using two different instruments, the Pain Coping Inventory (PCI) [32,38] and the Coping Strategies Questionnaire (CSQ) [36]. Self-efficacy was assessed in five studies [26,28,29,34,39] using the Arthritis Self-Efficacy Scale, somatisation was assessed in three studies [30,31,35] with the Brief Systems Inventory (BSI) and pain catastrophising was assessed in three studies [27,33,36] using the catastrophising subscale of CSQ. Helplessness was assessed in a single study using the Rheumatology Attitudes Index (RAI) [26]. Assessment of behavioural factors Kinesiophobia was assessed in two studies [27,33] using the Tampa Scale of Kinesiophobia. Pain-related fear-avoidance was assessed with the Fear-Avoidance Beliefs Questionnaire in three studies [32,37,38]. Diagnosis of OA in study participants In the seven studies assessing knee pain in patients with OA, two methods were used to assess OA. Six studies [28,32–34,38,39] used the American College of Rheumatology clinical criteria, while one study [26] used the Kellgren–Lawrence radiological score. Assessment of pain A range of scales was implemented to assess pain (Table 3). These included the National Health and Nutritional Examination Survey (NHANES) [26]; the CUPID Questionnaire [30]; Visual Analogue Scale (VAS) [27,28,32,38]; Arthritis Impact Measurement Scales (AIMS) [28,33,34]; Western Ontario and McMaster Universities (WOMAC)[36]; 11-point numerical Pain Rating Scale [37] and Survey of Pain Attitudes (SOPA-35) [39]. Methodological quality assessment The mean methodological quality score of the included studies was 68.3% with a range of 53.9–82.3%. Eight of the 14 studies were considered to be of high quality, as their quality score was above the mean. An analysis of the scoring criteria and their associated quality scores revealed that most studies scored well on selection of participants with disease at uniform point (criteria 1), identical assessment of outcome (criteria 9), sufficient description of baseline characteristics (criteria 5), analysis technique (criteria 15) and adjustment for age and sex (criteria 16). However, a number of studies scored poorly on blinded assessment of cognitive and behavioural factors (criteria 6), assessment of the risk factor prior to outcome (criteria 8) and reproducible assessment of outcome (criteria 10). Only two cohort studies and one quasi-experimental study used prospective designs. Cognitive factors The associations between cognitive factors and knee pain are summarised in Table 4. Coping The relationship between coping and knee pain was examined in a cohort, a nested case–control study and a cross-sectional study. Of these, the cohort and nested case–control study were of high quality. The cohort study [38] concluded that pain transformation from pain coping inventory was found to be a risk factor for knee pain. The nested case–control study [36] found that substantial deterioration of knee pain among cases was

6

Table 4 Studies examining the relationship between cognitive factors and knee pain References

Coping Peat (2009) (knee pain) [36]

van Baar et al. (1998) (knee OA) [32]

Assessment of psychosocial factor

Assessment of pain

Results

Conclusion

Quality score

Nested case–control Cases: mild pain intensity (o50/100) knee pain at study entry and became severe (change characteristic pain intensity Z70/100) at 18 months follow-up. Control: random sample of participants from the same cohort who had mild knee pain at study entry, which was not severe at 18 months follow-up Cohort

Coping and appraisal: 1-item Coping Strategies Questionnaire

Characteristic pain intensity: Chronic Pain Grade

Higher scores among cases than controls Coping and appraisal—Mean difference (95% CI) between cases and controls at 18 months: 11 (2–16%)

Substantial deterioration of knee pain was accompanied by an increase in perceived inability to cope

78

Pain Coping Inventory (PCI)

Overall pain in the last week VAS (0–100)

Pain transformation was found to be a risk factor for pain in knee OA.

69

Cross-sectional

Pain Coping Inventory (PCI)

Severity of pain: VAS

Knee pain patients with high levels of retreating experienced more pain. Pain coping was associated with knee pain in OA patients.

64

There was no statistically significant association between knee pain and selfefficacy.

72

Coping self-statement—Mean difference (95% CI) between cases and controls at 18 months: 12 (0–21%) Pain Catastrophising—Mean difference (95% CI) between cases and controls at 18 months: 37 (23–49%) Multivariate analysis—Knee OA: pain transformation comprises four items that focus on distracting attention from pain, β ¼ 0.206 (p ¼ 0.003) Correlation with pain: resting (consists of five items that assess the level to which patients avoid physical activity when in pain) r ¼ 0.22; NS Pain transformation r ¼ 0.34; p o 0.05 Lowering demands (items of this subscale assess the extent to which patients lower the demands of their activities r ¼ 0.11; NS PCI : multivariate analysis Pain (Knee OA)

 Retreating r ¼ 0.36; p r 0.01  Other forms of coping: distraction, reducing demands and worrying were not significant in multivariate analysis. Correlation with pain

      Self-efficacy Harcombe et al. (2009) (knee pain) [29]

Cross-sectional study

Self-efficacy: general SelfEfficacy Scale

Standardised Nordic Questionnaires for musculoskeletal symptoms Brief Symptom Inventory diagram showing the area of the body

Retreating r ¼ 0.39; p r 0.01. Pain transformation r ¼ 0.27; p r 0.01. Distraction, r ¼ 0.25; p r 0.01. Reducing demands r ¼ 0.21; p r 0.05. Worrying r ¼ 0.36; p r 0.01. Resting R ¼ 0.20; p r 0.05.

Logistic regression with adjustment for occupation, age, sex and body mass index—Knee Pain and selfefficacy: OR ¼ 1.01 (95% CI: 0.95–1.08); p ¼ 0.648

D.M. Urquhart et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]

Steultjens (2001) (knee OA) [38]

Study design

Cross-sectional

Arthritis Self-Efficacy Scale (Pain Subscale)

Pain Severity (WOMAC, Visual Analogue Scale and McGill Pain Questionnaire (MPQ))

Marks (knee OA) [28]

Cross-sectional

Arthritis Self-Efficacy Scale

Pain: VAS AIMS

Pells (2008) (knee OA) [34]

Cross-sectional

Arthritis Self-Efficacy Scale

Pain: AIMS

Wu (2011) (knee OA) [39]

Quasi-experimental study

Arthritis Self-efficacy Scale (ASE)

ASE Pain Scale Survey of Pain Attitudes (SOPA-35)

Cross-sectional

Somatising tendency: Brief System Inventory

CUPID Questionnaire

Somatising and knee pain: OR(95% CI) ¼ 1.8 (1.3, 2.6)

Cross-sectional

Somatising: Brief Symptom Inventory

Self-reported knee pain in past month or in the past year

Cross-sectional

Somatising: BSI

Pain in the knee lasting 4 ¼ 1 day during past 12 months

Somatising and disabling knee pain: OR(95% CI) ¼ 3.3 (2.2–4.9) Somatising and any knee pain: OR(95% CI) ¼ 2.6 (1.9–3.6) Univariate analysis Somatising and new-onset knee pain: OR(95% CI) ¼ 1.6 (1.0–2.7) (Multivariate Analysis—Not reported) Multivariate analysis Somatising and persistence of knee pain: OR (95% CI) ¼ 3.2 (1.8–5.7), adjusted for age, sex, general practice, frequency and associated disability of pain.

Nested case–control

Coping and appraisal: pain catastrophising subscale CSQ Pain catastrophising scale

Characteristic pain intensity: Chronic Pain Grade

Somatisation Warnakulasuriya et al. (knee pain) [30] Matsudaira et al. (knee pain) [31]

Palmer (2007) (knee pain) [35]

Pain catastrophising Peat (2009) (knee pain) [36] Domenech et al. (knee pain) [27]

Cross-sectional

Visual Analogue Scale (VAS)

Pain Catastrophising: mean difference (95% CI) between cases and controls at 18 months: 37 (23–49%) VAS score—mean (SD) High level of catastrophising (Z24) ¼ 7.7 (1.5) Low level of catastrophising (r 24) ¼ 6.5 (1.8) p ¼ 0.001

A moderate correlation was found between self-efficacy and pain (measured by the WOMAC), but with adjustment in regression models, this relationship did not remain.

54

There was a correlation between self-efficacy for pain control and knee pain in OA patients.

64

Self-efficacy for pain control was the only factor significantly associated with pain. Self-efficacy for physical function and other symptoms of arthritis were not associated with pain (p Z 0.17). There was a significant improvement in self-efficacy among the intervention group compared with the control group.

64

There was a strong correlation between somatising and knee pain. There was a strong relationship between somatising and disabling knee pain and any knee pain.

64

New onset and persistence of knee pain were significantly associated with tendency to somatise.

70

Deterioration of knee pain was accompanied by an increase in perceived inability to cope Pain catastrophising was significantly associated with pain levels after controlling for confounding variables.

78

69

82

D.M. Urquhart et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]

Correlation with self-efficacy: univariate analysis WOMAC: r ¼  0.48, p o 0.001 VAS: r ¼  0.28, p o 0.05 MPQ : r ¼  0.23, p ¼ NS With a stepwise regression model, which adjusted age, race, BMI, sex, disease duration, years of formal education, depression, anxiety and helplessness, for each measure of pain, self-efficacy did not remain in the model. Correlation between self-efficacy for pain control and VAS pain scale: r ¼  0.286; p ¼ 0.004. Efficacy for pain control and AIMS: r ¼  0.299; p ¼ 0.002 Multiple regression analysis was not performed. Correlation between self-efficacy for pain control and AIMS pain scale: r ¼  0.50; p o 0.01. Multiple regression for AIMS pain scale after controlling for demographic and medical characteristics: self-efficacy for pain control β ¼  0.28 (p ¼ 0.01); R2 ¼ 0.35 Baseline to post-intervention (week 4) ASE pain: mean (SD) Intervention (I) ¼ 1.70 (1.57), Control (C) ¼  0.22 (9.50); p ¼ 0.251 Baseline to follow-up (week 12) ASE pain I ¼ 2.25 (8.9), C ¼ 1.44 (12.38); p ¼ o0.001

Creamer et al.(knee OA) [26]

73

7

D.M. Urquhart et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]

54 Increased helplessness is associated with increased knee pain severity in people with knee OA. Pain Severity (WOMAC, Visual Analogue Scale and McGill Pain Questionnaire (MPQ)) Rheumatology Attitudes Index (RAI)— Helplessness subscale

Correlation between pain and helplessness: WOMAC: r ¼ 0.43 (p o 0.001) MPQ: r ¼ 0.45(p o 0.001). VAS: r ¼ 0.40 (p o 0.001) Stepwise regression analysis: WOMAC: R2 ¼ 0.44; p ¼ 0.02 MPQ: R2 ¼ 0.39; p ¼ 0.02 VAS: R2 ¼ 0.53; p ¼ 0.008

70 Pain catastrophising was significantly associated with pain severity. Pain severity: Arthritis Impact Measurement Scales Pain catastrophising subscale of Coping Strategies Questionnaires

Regression analysis for Pain After controlling for sex, age, anxiety, depression and kinesiophobia, p ¼ o0.05 Pain catastrophising associated with severity of pain: β ¼ 0.35, t ¼ 3.15, p ¼ 0.002

Quality score Assessment of pain Assessment of psychosocial factor

Results

Conclusion

8

accompanied by an increase in perceived inability to cope compared to controls. The cross-sectional study [32] found that patients who displayed a high level of retreating (coping style) experienced more knee pain. Self-efficacy The relationship between self-efficacy and knee pain was examined in five studies. Of these, four were cross-sectional [26,28,29,34] and one was quasi-experimental [39]. The quasiexperimental study [39] and one of the cross-sectional studies [29] were found to be of high quality. While the quasi-experimental study found a significant improvement in self-efficacy in the intervention group who experienced a reduction in knee pain [39], the high-quality cross-sectional study that examined various occupational groups did not find an association between selfefficacy and the presence of self-reported knee pain [29]. Two lowquality cross-sectional studies concluded that self-efficacy for pain control was significantly associated with pain, while the other cross-sectional study [26] did not find a significant association. Somatisation Three cross-sectional studies examined the association between somatisation and knee pain. While two of the studies were of high quality, all three studies showed that somatising tendency was associated with knee pain. While Warnakulasuriya et al. [30] and Matsudaira et al. [31] found an association between somatisation and the presence of self-reported pain in various occupational groups, Palmer et al. [35] examined participants with and without OA and found a relationship between somatisation and both new-onset and persistent pain. Pain catastrophising The relationship between pain catastrophising and knee pain was investigated by a nested case–control [36] and two crosssectional studies [27,33]. Both were found to be of high quality. The nested case–control study [36] found that cases had a significant deterioration of knee pain that was accompanied by an increase in pain catastrophising compared to controls. Similarly, both cross-sectional studies [27,33] found that pain catastrophising was associated with knee pain severity in OA patients. Helplessness The association between helplessness and knee pain was examined in a single low-quality cross-sectional study [26]. This study found that helplessness was associated with self-reported knee pain severity. Behavioural factors

Cross-sectional Helplessness Creamer et al. (knee OA) [26]

Cross-sectional Somers et al. (knee OA) [33]

References

Table 4 (continued )

Study design

The associations between behavioural factors and knee pain are summarised in Table 5. Kinesiophobia (pain-related fear of movement) The association between kinesiophobia and knee pain was assessed by two high-quality cross-sectional studies [27,33]. Both studies found that pain-related fear of movement was not significantly associated with knee pain in OA patients. Pain-related fear-avoidance The association between fear-avoidance beliefs and knee pain was examined in three studies: two were cohort studies [37,38] and one was cross-sectional study [32]. Of those, only one cohort study [38] was of high quality. That study [38] found no significant association between fear-avoidance and knee pain. The lowquality cohort study found that pain-related fear-avoidance was

Table 5 Studies examining the relationship between behavioural factors and knee pain References

Assessment of psychosocial factor

Assessment of pain

Results

Conclusion

Quality Score

Cross-sectional

Tampa Scale of Kinesiophobia Tampa Scale of Kinesiophobia

Pain-related fear of movement related to pain severity: β ¼ 0.05, t ¼ 0.50, p ¼ 0.62 VAS score—mean (SD) High level of kinesiophobia (Z40) ¼ 7.2 (1.6) Low level of kinesiophobia (r 40) ¼ 5.8 (2.2) T test comparison; p ¼ 0.002 Regression analysis for pain After controlling for sex, age, anxiety, depression and catastrophising—not significant. The value of p was not reported.

Pain-related fear was not significantly associated with pain severity. Even though kinesiophobia showed a correlation with current knee pain intensity, the logistic regression modelling was not statistically significant after controlling for confounding variables.

70

Cross-sectional

Pain severity: Arthritis Impact Measurement Scales Visual Analogue Scale (VAS)

11-Point Numerical Pain Rating Scale (NPRS)

Pain-related fear-avoidance association with change in pain severity (NPRS). Forward regression model: FABs-PA: R2 ¼ 0.33, p ¼ 0.000 FABs-W: R2 ¼ 0.43, p ¼ 0.010 Correlation with pain (knee OA) r ¼ 0.2; p ¼ r0.01. Fear-avoidance Multiple regression analysis—Knee pain after adjusted for muscle strength, range of motion, age, sex, medication, age, sex, duration of complaints and psychological well-being ¼ NS, actual p value was not reported. Correlation with Pain and fear-avoidance beliefs knee OA: r ¼  0.08; p ¼ NS Multiple regression after adjusting for age, sex, BMI, baseline pain level, duration of pain and coping strategies was not significant (p value was not reported).

After controlling for age, sex, height and weight, fear-avoidance both about work and about physical activity were found to be predictors of change in pain in patients with patella–femoral pain syndrome. The association between fear-avoidance beliefs and knee pain was not statistically significant.

64

Pain-related fear-avoidance Piva (2009) (knee Longitudinal pain) [37]

van Baar (knee OA) [32]

Cross-sectional

Steultjens (2001) (knee OA) [38]

Cohort

Fear-Avoidance Beliefs Questionnaire FABQ-W for work FABQ-PA for physical activity Fear-Avoidance Beliefs Questionnaire

Fear-Avoidance Beliefs Questionnaire

Severity of pain: Visual Analogue Scale

Overall pain in the last week Visual Analogue Scale (0– 100)

There was no significant relationship between fearavoidance beliefs and pain.

73

64

69

D.M. Urquhart et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]

Kinesiophobia Somers et al. (knee OA) [33] Domenech et al. (knee pain) [27]

Study design

9

10

D.M. Urquhart et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]

associated with knee pain in a population unselected for knee OA [37]. Best-evidence synthesis In the best-evidence synthesis, studies were considered high quality if they had a quality score above the mean quality score of 68.3%. Given the different constructs examined, factors were grouped broadly as “cognitive” or “behavioural.” For the cognitive factors, 8 of 15 studies were identified as high quality; one was a longitudinal study, one was a nested case–control study, one was quasi-experimental study and the remaining studies were cross-sectional. With the exception of one cross-sectional study, all high-quality studies showed an association between cognitive factors and knee pain. Taken together, these findings suggest that there is moderate evidence for a relationship between cognitive factors and knee pain. With respect to the behavioural factors, both a high-quality cohort and two high-quality cross-sectional studies found no association between behavioural factors and knee pain, indicating limited evidence for no relationship. We found no differences in the results between studies that examined participants with knee pain who were selected or unselected for OA. Of the eight high-quality studies that found a relationship between cognitive factors and knee pain, four studies examined populations with knee OA and four studies investigated those unselected for OA. Similarly, of the high-quality studies that found no association between behavioural factors and knee pain, two investigated participants with knee OA and one study examined individuals unselected for knee OA.

Discussion This systematic review found moderate evidence for a relationship between cognitive factors (self-efficacy, coping, somatising, helplessness and pain catastrophising) and knee pain, but limited evidence for no association between behavioural factors (kinesiophobia and fear-avoidance beliefs) and knee pain. These findings highlight the importance of a biopsychosocial approach to the management of knee pain and suggest that treatment strategies targeted at psychological functioning may be important in a comprehensive management plan. We found an association between reduced cognitive functioning, specific to coping, self-efficacy, somatising, pain catastrophising and helplessness, and the presence or deterioration of pain at the knee. Eight of 15 studies identified from our search were of high quality, and all the high-quality studies, with the exception of one, showed a significant inverse relationship between cognitive function and knee pain. This is consistent with previous systematic reviews that have reported somatisation and catastrophising to be predictors of chronicity in low back and neck pain, respectively [7,8]. Our findings may be explained by the biopsychosocial model, which suggests that structural changes at the tissue level alone cannot explain pain syndromes and that pain is influenced by a variety of factors, including biological, psychological and social factors [10]. Moreover, it is believed that chronic pain is a product of an individual's cognitions, whereby irrational thoughts and beliefs that pain is a signal of impending threat to the body lead to pain-related fear and disability [40]. In this review, we found limited evidence for no relationship between the behavioural factors, kinesiophobia and fear-avoidance, and knee pain. However, our conclusions were significantly limited by the small number of high-quality studies identified and the lack of longitudinal data available. Only two cohort studies were identified; one study was of high quality and found no

significant association between pain-related fear-avoidance and knee pain [38], while the other cohort study of low quality found a significant relationship [37]. Further investigation is needed to understand the role of behavioural factors in the development and persistence of knee pain. While behavioural factors may not be risk factors for knee pain or its chronicity, there is evidence suggesting that cognitive and behavioural factors are linked to each other along the pathway of development of chronic pain [7,9], with reduced cognitive processes having the potential to lead to maladaptive behavioural factors in the progression to chronic pain [9,41]. Moreover, there is evidence to suggest that fear-avoidance beliefs may be risk factors for the progression from acute to chronic low back pain in primary care [42]. This review has several limitations. We were not able to perform a meta-analysis to summarise our results due to the heterogeneity of the studies included in this review, and therefore, a best-evidence synthesis was performed. Our review was limited by the lack of a high-quality cohort investigating behavioural factors as risk factors for knee pain in both populations selected and unselected for OA. The majority of studies in this review were case–control or cross-sectional studies, which significantly reduced the quality of the evidence. Given the complexity associated with investigating the role of psychosocial factors in knee pain, we have not included all psychosocial factors in this review. We have recently published a systematic review examining the role of depression, anxiety and general mental health in knee pain [14]. Although fear-avoidance consists of both cognitive and behavioural elements, for the purpose of this review, we classified this factor as primarily behavioural. This systematic review supports the role of nonpharmacological strategies, including cognitive behavioural therapy, in the management of knee pain. A comparative study [12] has investigated the efficacy of the cognitive behavioural approach compared to arthritis education or standard care in patients with knee OA. Although the study was on a small number of participants (n ¼ 30, each group) and a RCT design was not used, they found that those who received training in coping had lower levels of pain, physical disability and pain behaviour at six months follow-up compared to controls. Future research, particularly in the form of RCTs, is needed to examine the effectiveness of a cognitive behavioural approach for reducing pain in the treatment of knee pain.

Conclusion Overall, this systematic review found an association between cognitive, but not behavioural factors, and knee pain. This review highlights the importance of a biopsychosocial approach to the management of knee pain to optimise clinical outcomes. Moreover, there is a need for further high-quality longitudinal studies if we are to clearly understand the role of cognitive and behavioural factors in the management of this condition. References [1] Jinks C, Jordan K, Ong BN, Croft P. A brief screening tool for knee pain in primary care (KNEST). 2. Results from a survey in the general population aged 50 and over. Rheumatology (Oxford) 2004;43:55–61. [2] Peat G, McCarney R, Croft P. Knee pain and osteoarthritis in older adults: a review of community burden and current use of primary health care. Ann Rheum Dis 2001;60:91–7. [3] Symmons DP. Knee pain in older adults: the latest musculoskeletal "epidemic". Ann Rheum Dis 2001;60:89–90. [4] Creamer P, Hochberg MC. Why does osteoarthritis of the knee hurt—sometimes? Br J Rheumatol 1997;36:726–8. [5] Fordyce WE. Chronic pain: the behavioral perspective. In Chronic pain: a reformulation of the cognitive–behavioural model. Behav Res Ther 2001;7:39.

D.M. Urquhart et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]] [6] Sharp TJ. Chronic pain: a reformulation of the cognitive–behavioural model. Behav Res Ther 2001;39:787–800. [7] Pincus T, Burton AK, Vogel S, Field AP. A systematic review of psychological factors as predictors of chronicity/disability in prospective cohorts of low back pain. Spine (Phila Pa 1976) 2002;27:E109–20. [8] Christensen JO, Knardahl S. Work and neck pain: a prospective study of psychological, social, and mechanical risk factors. Pain 2010;151:162–73. [9] Innes SI. Psychosocial factors and their role in chronic pain: A brief review of development and current status. Chiropr Osteopat 2005;13:6. [10] van Dijk GM, Dekker J, Veenhof C, van den Ende CH. Course of functional status and pain in osteoarthritis of the hip or knee: a systematic review of the literature. Arthritis Rheum 2006;55:779–85. [11] Blagojevic M, Jinks C, Jeffery A, Jordan KP. Risk factors for onset of osteoarthritis of the knee in older adults: a systematic review and meta-analysis. Osteoarthritis Cartilage 2010;18:24–33. [12] Keefe FJ, Caldwell DS, Williams DA, Gil KM, Mitchell D, Robertson D, et al. Pain coping skills training in the management of osteoarthritic kneepain: a comparative study. Behav Ther 1990;21:49–62. [13] Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. Br Med J 2009;339:b2700. [14] Phyomaung PP, Dubowitz J, Cicuttini FM, Fernando S, Wluka AE, Raaijmaakers P, et al. Are depression, anxiety and poor mental health risk factors for knee pain? A systematic review. BMC Musculoskelet Disord 2014;15:10. [15] Lievense A, Bierma-Zeinstra S, Verhagen A, Verhaar J, Koes B. Influence of work on the development of osteoarthritis of the hip: a systematic review. J Rheumatol 2001;28:2520–8. [16] Hurley MV, Walsh NE, Mitchell HL, Pimm TJ, Williamson E, Jones RH, et al. Economic evaluation of a rehabilitation program integrating exercise, selfmanagement, and active coping strategies for chronic knee pain. Arthritis Rheum 2007;57:1220–9. [17] Hawker GA, Stewart L, French MR, Cibere J, Jordan JM, March L, et al. Understanding the pain experience in hip and knee osteoarthritis—an OARSI/ OMERACT initiative. Osteoarthritis Cartilage 2008;16:415–22. [18] Ibrahim SA, Burant CJ, Mercer MB, Siminoff LA, Kwoh CK. Older patients' perceptions of quality of chronic knee or hip pain: differences by ethnicity and relationship to clinical variables. J Gerontol A Biol Sci Med Sci 2003;58:M472–7. [19] Murphy SL, Strasburg DM, Lyden AK, Smith DM, Koliba JF, Dadabhoy DP, et al. Effects of activity strategy training on pain and physical activity in older adults with knee or hip osteoarthritis: a pilot study. Arthritis Rheum 2008;59:1480–7. [20] Tiedemann A, Sherrington C, Lord SR. Physiological and psychological predictors of walking speed in older community-dwelling people. Gerontology 2005;51:390–5. [21] Gignac MA, Backman CL, Davis AM, Lacaille D, Cao X, Badley EM. Social role participation and the life course in healthy adults and individuals with osteoarthritis: are we overlooking the impact on the middle-aged? Soc Sci Med 2013;81:87–93. [22] Heuts PH, Vlaeyen JW, Roelofs J, de Bie RA, Aretz K, van Weel C, et al. Pain-related fear and daily functioning in patients with osteoarthritis. Pain 2004;110:228–35. [23] Thumboo J, Chew LH, Lewin-Koh SC. Socioeconomic and psychosocial factors influence pain or physical function in Asian patients with knee or hip osteoarthritis. Ann Rheum Dis 2002;61:1017–20.

11

[24] Howard KJ, Ellis HB, Wang J, Von Der Gruen JK, Bucholz R. Evaluating the effects of somatization disorder for patients with severe end-stage lowerextremity osteoarthritis. J Appl Biobehav Res 2012;17:79–93. [25] Macfarlane GJ, Pallewatte N, Paudyal P, Blyth FM, Coggon D, Crombez G, et al. Evaluation of work-related psychosocial factors and regional musculoskeletal pain: results from a EULAR Task Force. Ann Rheum Dis 2009;68:885–91. [26] Creamer P, Lethbridge-Cejku M, Hochberg MC. Determinants of pain severity in knee osteoarthritis: effect of demographic and psychosocial variables using 3 pain measures. J Rheumatol 1999;26(8):1785–92. [27] Domenech J, Sanchis-Alfonso V, Lopez L, Espejo B. Influence of kinesiophobia and catastrophizing on pain and disability in anterior knee pain patients. Knee Surg Sports Traumatol Arthrosc 2013;21:1562–8. [28] Marks R. Physical and psychological correlates of disability among a cohort of individuals with knee osteoarthritis. Can J Aging 2007;26:367–77. [29] Harcombe H, McBride D, Derrett S, Gray A. Physical and psychosocial risk factors for musculoskeletal disorders in New Zealand nurses, postal workers and office workers. Inj Prev 2010;16(2):96–100. [30] Warnakulasuriya SS, Peiris-John RJ, Coggon D, Ntani G, Sathiakumar N, Wickremasinghe AR. Musculoskeletal pain in four occupational populations in Sri Lanka. Occup Med (Lond) 2012;62:269–72. [31] Matsudaira K, Palmer KT, Reading I, Hirai M, Yoshimura N, Coggon D. Prevalence and correlates of regional pain and associated disability in Japanese workers. Occup Environ Med 2011;68:191–6. [32] van Baar ME, Dekker J, Lemmens JA, Oostendorp RA, Bijlsma JW. Pain and disability in patients with osteoarthritis of hip or knee: the relationship with articular, kinesiological, and psychological characteristics. J Rheumatol 1998;25(1):125–33. [33] Somers TJ, Keefe FJ, Pells JJ, Dixon KE, Waters SJ, Riordan PA, et al. Pain catastrophizing and pain-related fear in osteoarthritis patients: relationships to pain and disability. J Pain Symptom Manage 2009;37:863–72. [34] Pells JJ, Shelby RA, Keefe FJ, Dixon KE, Blumenthal JA, Lacaille L, et al. Arthritis self-efficacy and self-efficacy for resisting eating: relationships to pain, disability, and eating behavior in overweight and obese individuals with osteoarthritic knee pain. Pain 2008;136:340–7. [35] Palmer KT, Reading I, Calnan M, Linaker C, Coggon D. Does knee pain in the community behave like a regional pain syndrome? Prospective cohort study of incidence and persistence. Ann Rheum Dis 2007;66:1190–4. [36] Peat G, Thomas E. When knee pain becomes severe: a nested case-control analysis in community-dwelling older adults. J Pain 2009;10(8):798–808. [37] Piva SR, Fitzgerald GK, Wisniewski S, Delitto A. Predictors of pain and function outcome after rehabilitation in patients with patellofemoral pain syndrome. J Rehabil Med 2009;41(8):604–12. [38] Steultjens MP, Dekker J, Bijlsma JW. Coping, pain, and disability in osteoarthritis: a longitudinal study. J Rheumatol 2001;28:1068–72. [39] Wu SF, Kao MJ, Wu MP, Tsai MW, Chang WW. Effects of an osteoarthritis selfmanagement programme. J Adv Nurs 2011;67:1491–501. [40] Vlaeyen JW, Linton SJ. Fear-avoidance and its consequences in chronic musculoskeletal pain: a state of the art. Pain 2000;85:317–32. [41] Keefe FJ, Rumble ME, Scipio CD, Giordano LA, Perri LM. Psychological aspects of persistent pain: current state of the science. J Pain 2004;5:195–211. [42] Ramond A, Bouton C, Richard I, Roquelaure Y, Baufreton C, Legrand E, et al. Psychosocial risk factors for chronic low back pain in primary care—a systematic review. Fam Pract 2011;28:12–21.