Materials and Methods
- Top of page
- Materials and Methods
- AUTHOR CONTRIBUTIONS
The study was based on cross-sectional data collected from June to October 2009 in Sint Maartenskliniek, Nijmegen (The Netherlands), a specialized hospital for orthopedics, rheumatology, and rehabilitation. Men and women (age ≥18 years) received a questionnaire within 14 days after they visited an orthopedic surgeon for a new episode of complaints due to hip or knee OA. Radiographs of the index joint were taken from all patients, as this is standard procedure. Patients who self-reported concurrent rheumatoid arthritis, fibromyalgia, or psoriatic arthritis were excluded. The Institutional Review Board of the University Medical Centre, Nijmegen, approved the study.
Sociodemographic characteristics were assessed by standardized questions. All posteroanterior fixed flexion knee and pelvic radiographs (5) were scored for the index joint by a trained and experienced rheumatologist according to the Kellgren/Lawrence (K/L) grading scale (6). Validated and reliable questionnaires were used to assess the following disease-related measures: general functioning (range 0–100) and pain (range 0–100; Short Form 36 [SF-36] Health Survey, where lower scores represent unfavorable outcomes) (7); stiffness (range 0–100), pain (range 0–100), and functioning (range 0–100) of the index joint (Western Ontario and McMaster Universities Osteoarthritis Index [WOMAC] subscale scores were transformed to a 0–100 scale, i.e., 100 − [actual raw score × 100/possible raw score range], where lower scores represent unfavorable outcomes); fatigue (range 8–56, cutoff score is ≥35, severely fatigued; Checklist Individual Strength [CIS]) (8); self-efficacy (range 10–40; General Self-Efficacy Scale, where lower scores represent unfavorable outcomes) (9); kinesiophobia (range 17–68; Tampa Scale for Kinesiophobia, where lower scores represent favorable outcomes) (10); anxiety (range 0–21) and depression (range 0–21, cutoff score at ≥8, depressive symptoms; Hospital Anxiety and Depression Scale [HADS], where lower scores represent favorable outcomes) (11); and health-related quality of life (range 0–1; EuroQol 5-domain [EQ-5D], where lower scores represent unfavorable outcomes) (12).
Information about health service utilization was obtained by asking straightforward questions (e.g., “Have you ever been in contact with a physiotherapist for your knee or hip complaints?”). We assessed contact with a general practitioner, physiotherapist, and psychologist. We also assessed analgesic use (i.e., acetaminophen, nonsteroidal antiinflammatory drugs [NSAIDs], and opioids) and the use of complimentary/alternative care (e.g., acupuncture, healers, dietary supplements, etc.).
Finally, the number of painful joint sites (i.e., feet, ankles, knees, hips, lumbar spine, thoracic spine, cervical spine, shoulders, elbows, wrists, and hands) was assessed by the following 2 straightforward questions: “Do you experience symptoms from your (left/right) joint?” and “If so, did you experience symptoms of your (left/right) joint for more than half of the time during the last month?” A joint was counted as “painful” in the case of a positive answer to the second question. Joint-pain comorbidity was defined as pain perceived in a joint, other than the index joint, for more than half of the days in the preceding month.
A total of 384 participants were needed to sufficiently power the study, assuming a 50% prevalence rate of joint-pain comorbidity, an error margin of 5%, and 95% confidence limits. Descriptive statistics were used to describe the amount of missing data and the study population. Frequencies and percentages were used to report the prevalence of joint-pain comorbidities. Based on complete case data, sociodemographic data between the group with and without joint-pain comorbidities were compared using t-test and chi-square statistics. To compare disease-related characteristics of participants with and without joint-pain comorbidities, we performed analyses of covariance and logistic regression when appropriate, adjusting for age, sex, education, medical comorbidity, and index joint (i.e., hip or knee). Subgroup analyses were performed for the isolated hip and knee groups.
- Top of page
- Materials and Methods
- AUTHOR CONTRIBUTIONS
A total of 401 people completed and returned the questionnaire (82% response rate). The nonresponder group (of whom 25% were contacted by telephone) did not differ significantly from the responder group on demographic variables. Three hundred forty-four participants (86%) had <3 missing data points. The study sample was 58% women with a mean ± SD age of 58 ± 13 years and a mean ± SD body mass index of 27 ± 5 kg/m2. A total of 284 participants were referred for knee OA and 117 for hip OA; 73 participants (18%) reported both hip and knee complaints. Seventy-six percent of the participants had a K/L grade of ≥2.
Fifty-eight percent (n = 231) of the participants reported symptoms in more than 1 joint for more than half of the time during the previous month (Table 1). The median number of affected joints was 2 (interquartile range 1–4) joints. Peripheral regions most and least often affected in our total sample, other than hip (34%) and knee (69%), were the shoulders (16%) and elbow joints (6%), respectively. The lumbar spine (27%) was most often reported as painful when considering the axial joints. The number of patients with knee OA reporting comorbid hip pain was significantly lower than the number of individuals with hip OA reporting comorbid knee pain (14% and 33%, respectively). Also, pain in the lumbar spine region was more frequent in the patients referred for hip OA than those referred for knee OA (23% and 37%, respectively).
Table 1. Prevalence of joint-pain comorbidities (pain or aching on more than half the days in the preceding month) by location*
|Pain location||Knee group (n = 284)||Hip group (n = 117)|
|Both sides||Single side||Total||Both sides||Single side||Total|
|Spine pain|| || || || || || |
| Lumbar||–||66 (23)||66 (23)||–||43 (37)||43 (37)|
| Thoracic||–||18 (6)||18 (6)||–||12 (10)||12 (10)|
| Neck||–||49 (17)||49 (17)||–||18 (15)||18 (15)|
|Extremity pain|| || || || || || |
| Shoulder||25 (9)||22 (8)||47 (17)||10 (9)||7 (7)||17 (16)|
| Elbow||7 (2)||13 (5)||20 (7)||2 (2)||3 (3)||5 (5)|
| Wrist||14 (5)||12 (4)||26 (9)||4 (4)||5 (5)||9 (9)|
| Hand||23 (8)||16 (6)||39 (14)||6 (6)||4 (4)||10 (10)|
| Hip||15 (5)||26 (9)||41 (14)||31 (26)||64 (55)||95 (81)|
| Knee||81 (29)||160 (56)||241 (85)||10 (9)||26 (24)||36 (33)|
| Ankle||9 (3)||22 (8)||31 (11)||5 (5)||7 (7)||12 (12)|
| Foot||18 (6)||21 (7)||39 (14)||7 (7)||6 (6)||13 (13)|
Participants with painful joints other than the index joints were more likely to be women, less educated, and have more medical comorbidities (Table 2). Individuals with joint pain elsewhere showed significantly more depressive symptoms (HADS depression score ≥8, 35% versus 12%; P < 0.001) and were more severely fatigued (CIS fatigue score ≥35, 56% versus 31%; P < 0.001) than those without joint-pain comorbidities (odds ratio [OR] 3.56; 95% confidence interval [95% CI] 2.00, 6.35 and OR 2.34; 95% CI 1.49, 3.70, respectively) (Table 3). Moreover, pain (SF-36 and WOMAC), symptoms (WOMAC), functioning (SF-36 and WOMAC), quality of life (EQ-5D), and the value of own health (EQ-5D) were significantly worse in individuals with joint-pain comorbidities. No statistically significant differences between groups were found in radiographic severity (P = 0.791), kinesiophobia (P = 0.727), duration of complaints (P = 0.089), and self-efficacy (P = 0.099).
Table 2. Patient and disease characteristics of individuals with hip and/or knee OA with and without joint-pain comorbidities*
|Demographics||No JPC (n = 170)||≥1 JPC (n = 231)||Mean difference or OR (95% CI)||P|
|Age, mean ± SD years||57.7 ± 13.7||58.4 ± 12.2||−0.72 (−3.27, 1.83)||0.579|
|Women||87 (51)||144 (62)|| ||0.025†|
|BMI, mean ± SD kg/m2||27.1 ± 4.5||27.3 ± 4.9||0.34 (−0.65, 1.34)||0.496|
|Medical comorbidity‡|| || || || |
| Diabetes mellitus||9 (5)||19 (8)||0.83 (0.34, 2.00)||0.674|
| Cardiovascular disease||13 (8)||34 (15)||0.45 (0.21, 0.96)||0.040†|
| Respiratory disease||6 (4)||21 (9)||0.31 (0.11, 0.87)||0.026†|
| Kidney/liver disease||2 (1)||2 (1)||0.71 (0.10, 5.12)||0.732|
| Gastrointestinal complaints||8 (5)||24 (11)||0.39 (0.16, 0.94)||0.036†|
|Education, secondary school||155 (95)||191 (87)|| ||0.036†|
|Additionally insured||146 (88)||202 (92)|| ||0.207|
|Income, >€40,000 per year||55 (36)||55 (29)|| ||0.118|
|Currently employed||87 (53)||94 (43)|| ||0.062|
Table 3. Disease characteristics of individuals with hip and/or knee OA with and without JPC*
| ||No JPC (n = 170)||≥1 JPC (n = 231)||Mean difference (95% CI)||OR (95% CI)|
|Health- and disease-related status†|| || || || |
| Fatigued, CIS score ≥35||51 (31)||126 (56)||–||2.34 (1.49, 3.70)‡|
| Depressive symptoms, HADS ≥8||20 (12)||79 (35)||–||3.56 (2.00, 6.35)‡|
| Anxiety, mean ± SD§||4.3 ± 3.4||5.9 ± 4.1||1.44 (0.64, 2.23)‡||–|
| Kinesiophobia, mean ± SD¶||38.6 ± 7.8||39.3 ± 8.1||0.30 (−1.37, 1.96)||–|
| Self-efficacy, mean ± SD#||32.8 ± 4.9||31.4 ± 5.6||−0.91 (−2.00, 0.17)||–|
| Quality of life, mean ± SD**||0.71 ± 0.17||0.63 ± 0.19||−0.06 (−0.10, −0.02)‡||–|
| Value own health, mean ± SD**||6.1 ± 1.7||5.8 ± 1.6||−0.64 (−0.99, −0.30)‡||–|
| SF-36|| || || || |
| Bodily pain#||45.8 ± 20.5||35.8 ± 18.4||−8.80 (−12.85, −4.75)‡||–|
| Physical functioning#||48.2 ± 23.1||36.9 ± 20.4||−9.17 (−13.60, −4.74)‡||–|
|OA and symptoms†|| || || || |
| Complaints duration, ≥5 years||88 (55)||139 (66)||–||1.47 (0.94, 2.29)|
| K/L grade ≥2||129 (76)||174 (76)||–||0.93 (0.57, 1.54)|
| WOMAC††|| || || || |
| Pain, mean ± SD#||61.0 ± 21.6||49.7 ± 20.0||−9.82 (−14.17, −5.48)‡||–|
| Stiffness, mean ± SD#||52.9 ± 21.5||40.6 ± 21.1||−11.73 (−16.19, −7.26)‡||–|
| Function, mean ± SD#||60.2 ± 22.0||48.8 ± 20.3||−9.75 (14.29, −5.21)‡||–|
|Health care utilization†|| || || || |
| General practitioner contact (≥1)||157 (93)||209 (92)||–||0.81 (0.37, 1.78)|
| Physiotherapy contact (≥1)||126 (75)||180 (79)||–||1.26 (0.75, 2.12)|
| CAM|| || || || |
| Contact||19 (11)||42 (18)||–||1.57 (0.84, 2.95)|
| Supplement use||50 (29)||93 (40)||–||1.65 (1.04, 2.63)‡|
| Acetaminophen use||98 (58)||155 (68)||–||1.32 (0.85, 2.06)|
| NSAID use||92 (54)||146 (64)||–||1.60 (1.03, 2.49)‡|
| Opioids use||18 (11)||55 (24)||–||2.29 (1.25, 4.16)‡|
NSAIDs (54% versus 64%; P = 0.038), opioids (11% versus 24%; P = 0.010), and supplements use (29% versus 40%; P = 0.019) were significantly higher in the group that reported joint-pain comorbidities than those who did not (Table 3): OR 1.60 (95% CI 1.03, 2.49), OR 2.21 (95% CI 1.21, 4.05), and OR 1.75 (95% CI 1.10, 2.79), respectively. No statistically significant differences were found in the number of patients that contacted a physiotherapist (75% versus 79%; P = 0.397) or a general practitioner (93% versus 92%; P = 0.618) for their symptoms.
We repeated all analyses for the hip (n = 117) and knee group (n = 284) separately. Analyses in the knee group yielded results similar to the primary analyses. For the hip group, we did find some differences from the primary analyses. Compared with the primary analyses, no significant difference between hip OA patients with and without joint-pain comorbidities was found in the use of NSAIDs, opioids, and supplements, as well as pain (SF-36), functioning (WOMAC and SF-36), and quality of life (EQ-5D). However, although not significant, the directions of the differences were similar to the primary analyses, except for NSAID use.
- Top of page
- Materials and Methods
- AUTHOR CONTRIBUTIONS
In this study, we showed that the majority of people referred for hip or knee OA experience joint pain elsewhere. The physical and psychological burden of symptoms related to knee or hip OA is significantly higher in individuals with painful joint sites elsewhere than in individuals who did not report joint-pain comorbidity, especially in the group of patients with knee OA. Moreover, the use of NSAIDs, opioids, and supplements was also considerably higher in the group with joint-pain comorbidities, whereas no differences in general practitioner and physiotherapy contacts were observed. Therefore, we identified a relevant subgroup of patients possibly requiring more intensive and complex treatment interventions.
In contrast to previous studies (3, 13), the prevalence of joint-pain comorbidities in our study (58%) was lower. Dissimilarities in prevalence might be explained by differences in the method of determining joint-pain comorbidities. We defined a joint site as painful when a participant reported pain for more than half of the days in the preceding month, whereas Croft et al defined a joint site as painful in case of pain in one or more day in the last month (13). Differences in the prevalence of painful joint sites between our study and that by Suri et al (3) could be explained by the fact that they use a manikin to determine joint-pain comorbidities. Although both methods (written questions and manikin) give valid results, the use of a manikin is known to result in a higher prevalence of musculoskeletal pain (14).
Individuals with hip or knee OA with joint-pain comorbidities are more fatigued, experience lower quality of life, more bodily pain, deprived functioning, and might even be at risk for depressive symptoms compared with people without joint-pain comorbidities. The magnitude of these differences is substantial and can be considered as clinically relevant. For instance, the absolute number of participants who had a fatigue score >35 points (8), indicating severe levels of fatigue, was 23% higher in participants with joint-pain comorbidities than in participants without joint-pain comorbidities. Moreover, patients with pain elsewhere use more and stronger analgesics than patients without joint-pain comorbidities and are more likely to use dietary supplements of controversial efficacy. In conclusion, pain elsewhere in hip or knee OA is associated with compromised health status and more (unfavorable) forms of medication and supplement use. We therefore concur with the recommendation that in both research and clinical practice, joint-pain comorbidities should be accounted for in hip and knee OA (4). In our opinion, differences in health status and quality of life between individuals with and without joint-pain comorbidities cannot be attributed to differences in the progression of OA, as disease duration and K/L grading were comparable in both groups. Our results do not allow interpretation of the causal relationships between joint-pain comorbidities and health status, due to the cross-sectional nature of the study.
Our data showed that the proportion of patients who contacted health care providers was similar between the 2 groups, whereas stronger medication use and dietary supplement intake were more prevalent in the group with joint-pain comorbidities. Considering the substantially higher burden of disease in multiple domains in the subgroup of patients with joint-pain comorbidities (e.g., more medical comorbidities, depressed mood, anxiety, and severe fatigue), it is conceivable that these patients may require, next to medication, more complex and intensive treatment modalities than patients without joint-pain comorbidities. Our results might be a starting point to tailor treatment for this category of patients.
A possible limitation of our study is that we did not collect data on the nature of joint-pain comorbidities other than the index joint. Therefore, it is unknown to what extent those complaints are related to OA (e.g., generalized OA) or possibly to other diagnoses. Further research should focus on the cause and nature of joint-pain comorbidities in knee or hip OA. Another limitation of our study is that we recruited individuals from one center, a large specialized hospital in The Netherlands, which might hamper the generalizability of our findings. However, as patient and disease characteristics of our sample resemble those of the Osteoarthritis Initiative incidence subcohort (15), we believe that our results are generalizable to patients with hip or knee OA referred to secondary care. A final limitation of our study is that we did not power the study specifically for subgroup analyses. Due to the lack of power, it remains unclear whether the differences in results between the hip and knee subgroups reflect true differences. Nevertheless, considering that we are the first to investigate joint-pain comorbidities in hip OA, we think that additional research is warranted.
In conclusion, we generated empirical data confirming the expectation that people with joint-pain comorbidities represent a clinically relevant and relatively large subgroup. In this subgroup of patients, the burden of disease is considerably higher, medical comorbidities are more prevalent, and medication use is significantly higher. Because our study does not allow inferences about the nature and causality of joint-pain comorbidities in OA, further research is warranted.
- Top of page
- Materials and Methods
- AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Hoogeboom had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Hoogeboom, den Broeder, Swierstra, die Bie, van den Ende.
Acquisition of data. Hoogeboom, den Broeder, Swierstra.
Analysis and interpretation of data. Hoogeboom, die Bie, van den Ende.