- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
Arthritis is one of the most prevalent chronic health conditions in the population. Knowing the prevalence of arthritis is important for public health surveillance and the planning and prioritization of population-based initiatives for the control of arthritis, especially the promotion of exercise and physical activity and arthritis self-management (1–3). Population estimates of the prevalence of chronic conditions are usually obtained through self-reported data in population health surveys where respondents are asked whether they have been told by a health care professional if they have a specific condition; arthritis is no exception. However, as pointed out by others (4, 5), this definition may exclude respondents who did not consult a health care professional for their symptoms, who misunderstood or forgot the diagnosis, or who did not realize that their symptoms relate to arthritis.
Individuals with joint symptoms represent a substantial and relatively unstudied subgroup of the population. From 1997–2007, the Behavioral Risk Factor Surveillance System (BRFSS) survey in the US incorporated questions to identify people who have joint symptoms (4, 6). While estimates for the prevalence of arthritis for the US were in the range of 18–36%, the BRFSS showed that an additional 10–21% of the population reported chronic joint symptoms (CJS) (7). Because CJS have been considered an indication of possible arthritis (4, 7), previous studies have been mainly concerned with attempting to establish whether individuals reporting these problems have arthritis or not (8–11). The uncertainty as to whether this is the case has led to a US recommendation to focus on those explicitly reporting arthritis for surveillance purposes and as a target for education and control initiatives (8). Less attention has been paid to the impact of these reported joint symptoms on health outcomes such as pain and quality of life, and on health service use.
Using data from a recent Canadian population health survey, which included questions about joint symptoms posed to individuals not reporting arthritis, the objectives of this study were to examine whether the predictors of reporting arthritis are similar to those of reporting joint symptoms, and to identify the risks of reporting health outcomes and health service use by comparing individuals with arthritis, CJS, and sporadic joint symptoms (SJS) with those with no arthritis or joint symptoms (NJS).
Significance & Innovations
In addition to people who report arthritis, a substantial proportion of the population also reports joint symptoms, a large proportion of whom has chronic joint symptoms.
Chronic joint symptoms have a large impact on pain, which is only slightly less than that of arthritis. In addition, when compared to those reporting arthritis, people with these symptoms reported similar health outcomes (physical activity, activity limitations, poor/fair self-rated health, and poor/fair self-rated mental health) and health care use (visits to primary care physicians, specialists, physiotherapists, and chiropractors, and inpatient stays).
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
Arthritis and joint symptoms in the past year were reported by 16.0% and 21.7% of the Canadian population age ≥15 years, respectively. The group reporting joint symptoms was comprised of an estimated 10.1% of the population who reported CJS (that is, pain that persisted most days for at least 1 month in the past 12 months) and 11.6% of the population who reported SJS. Overall, 26.1% reported either arthritis or CJS.
The sociodemographic and lifestyle characteristics of the respondents in the 4 groups examined are shown in Table 1. The most notable difference between the arthritis and CJS groups was the age distribution; a large proportion of individuals with arthritis were ages ≥65 years (42.4%), while the majority of those of those with CJS and SJS were younger than 65 years (87.0% and 86.2%, respectively). Arthritis and CJS were more common among women (61.3% and 51.7%, respectively). Individuals with arthritis reported lower socioeconomic status (low education or lower household income) more frequently than those in the other groups. A substantial proportion of individuals with arthritis (58.1%) or joint symptoms (52.4% of those with CJS and 50.6% of those with SJS) were overweight or obese; the proportion was much lower for the NJS group (41.1%). There appeared to be a gradient in the proportion of individuals reporting comorbidities from highest to lowest for those with arthritis (78.2%), CJS (66.9%), SJS (56.6%), and NJS (39.5%).
Table 1. Sociodemographic and lifestyle characteristics of Canadian people in 2008*
|Age and sex|| || || |
| 15–44 years||13.5 (3.1–23.9)||41.2 (39.2–43.2)#||49.2 (47.4–51.0)#||61.8 (61.2–62.4)#|
| 45–64 years||44.1 (31.2–57.0)||45.8 (43.8–47.8)||37.0 (35.2–38.8)||28.4 (27.8–29.0)**|
| ≥65 years||42.4 (31.0–53.8)||13.0 (12.0–14.0)#||13.8 (12.8–14.8)#||9.8 (9.4–10.2)#|
| Women||61.3 (60.1–62.5)||51.7 (49.7–53.7)#||48.2 (46.6–49.8)#||48.4 (48.0–48.8)#|
|SES|| || || || |
| Education (secondary or less)||42.3 (40.1–44.5)||33.0 (27.8–38.2)**||30.8 (27.0–34.6)#||32.5 (30.3–34.7)#|
| Household income (<$39,999)||35.3 (33.0–37.6)||24.0 (21.8–26.2)#||20.2 (17.8–22.6)#||19.8 (17.6–22.0)#|
|Ethnicity|| || || || |
| Asian||6.3 (5.1–7.5)||6.3 (5.1–7.5)||8.0 (6.6–9.4)||13.3 (12.5–14.1)#|
| Aboriginal||3.7 (3.1–4.3)||3.7 (2.9–4.5)||3.0 (2.4–3.6)||3.1 (2.9–3.3)|
|Lifestyle|| || || || |
| Overweight (25.0–29.9 kg/m2)||34.2 (32.8–35.6)||33.4 (31.4–35.4)||32.6 (30.8–34.4)||28.7 (27.9–29.5)#|
| Obese (≥30.0 kg/m2)||23.9 (22.5–25.3)||19.0 (17.4–20.6)**||18.0 (16.6–19.4)**||12.4 (11.8–13.0)#|
| Current smokers||21.1 (19.9–22.3)||27.4 (25.5–29.3)#||22.0 (19.9–24.1)||20.4 (19.8–21.0)|
|Number of other chronic conditions|| || || || |
| None||18.8 (17.5–20.2)||33.1 (31.0–35.1)#||43.4 (41.5–45.2)#||60.5 (59.6–61.3)#|
| 1–2||53.8 (52.3–55.3)||53.1 (51.0–55.2)||49.6 (47.7–51.4)**||35.7 (34.9–36.6)#|
| ≥3||27.4 (26.1–28.7)||13.8 (12.5–15.1)#||7.1 (6.3–7.9)#||3.8 (3.5–4.1)#|
The results of the log Poisson regression models examining predictors of reporting arthritis, CJS, and SJS compared to the NJS group confirmed increasing risks with older age for arthritis, and showed that arthritis and CJS were more likely to be reported by women (Table 2). However, the contribution of ethnicity (Aboriginal), overweight/obesity, and smoking status was similar for those with arthritis and CJS and higher than that for SJS. For example, individuals with arthritis and CJS were more likely to also report being obese than those with SJS or NJS. Low socioeconomic status was associated with an increased likelihood of reporting arthritis but not joint symptoms. Having ≥1 comorbidities was also associated with reporting of all 3 arthritis or joint symptom groups.
Table 2. Results from separate log Poisson regression models for the predictors of arthritis, CJS, and SJS compared to those with NJS for Canadian people in 2008*
|Age groups, years (ref.15–24)|| || || |
| 25–34||2.29 (1.66–2.92)||1.21 (1.02–1.40)†||0.93 (0.8–1.06)†|
| 35–44||5.01 (3.76–6.26)||1.67 (1.41–1.93)†||1.22 (1.07–1.37)†|
| 45–54||9.97 (7.56–12.38)||2.31 (1.96–2.66)†||1.32 (1.15–1.49)†|
| 55–64||14.63 (11.1–18.16)||2.20 (1.87–2.53)†||1.52 (1.33–1.71)†|
| 65–74||17.22 (13.11–21.33)||1.60 (1.35–1.85)†||1.44 (1.23–1.65)†|
| ≥75||19.83 (15.04–24.62)||1.50 (1.23–1.77)†||1.34 (1.13–1.55)†|
|Women (ref. men)||1.39 (1.31–1.47)||1.15 (1.06–1.24)†||1.04 (0.97–1.11)†|
|Education (ref. greater than secondary)|| || || |
| Less than secondary||1.00 (0.95–1.05)||1.00 (0.90–1.10)||0.95 (0.86–1.04)|
| Secondary||0.99 (0.91–1.07)||0.99 (0.88–1.10)||0.88 (0.80–0.96)|
|Household income (ref. ≥$60,000)|| || || |
| <$20,000||1.08 (1.00–1.16)||0.97 (0.84–1.10)||0.89 (0.77–1.01)|
| $20,000–$39,999||1.12 (1.04–1.2)||1.08 (0.96–1.20)||0.94 (0.86–1.02)|
| $40,000–$59,999||1.05 (0.98–1.12)||1.02 (0.92–1.12)||1.03 (0.94–1.12)|
| Not stated||0.96 (0.87–1.05)||0.89 (0.77–1.01)||0.89 (0.89–0.89)|
|Ethnicity (ref. white)|| || || |
| Asian||0.77 (0.64–0.90)||0.63 (0.51–0.75)||0.72 (0.61–0.83)†|
| Aboriginal||1.27 (1.12–1.42)||1.10 (0.92–1.28)||0.97 (0.83–1.11)|
| Other||0.77 (0.59–1.02)||0.92 (0.72–1.12)||0.85 (0.70–1.00)|
|BMI (ref. <25.0 kg/m2)|| || || |
| Overweight (25.0–29.9 kg/m2)||1.16 (1.09–1.23)||1.16 (1.06–1.26)||1.16 (1.07–1.25)|
| Obese (≥30.0 kg/m2)||1.36 (1.27–1.45)||1.20 (1.09–1.31)||1.27 (1.15–1.39)†|
|Smoking (ref. never smoked)|| || || |
| Current smoker||1.30 (1.21–1.39)||1.43 (1.28–1.58)||1.18 (1.07–1.29)|
| Former smoker||1.14 (1.07–1.21)||1.26 (1.15–1.37)||1.17 (1.08–1.26)|
|Number of other chronic conditions (ref. none)|| || || |
| 1–2||2.17 (2.00–2.37)||2.01 (1.84–2.21)||1.53 (1.41–1.65)†|
| ≥3||3.15 (2.89–3.42)||3.15 (2.80–3.55)||1.66 (1.45–1.90)†|
The percent distributions of the health and health service use outcomes examined are shown in Table 3 for each of the groups of interest. Similar proportions of individuals reporting being physically inactive were found for the arthritis and CJS groups. A markedly greater proportion of individuals with arthritis and CJS reported pain that limits at least a few activities, sometimes or often having activity limitations, or poor/fair general health than those with SJS or NJS. The proportions with these outcomes were somewhat higher for the arthritis than the CJS group. Generally, there was a gradient in the proportion reporting poor health outcomes (being physically inactive, poor/fair self-rated health, poor/fair self-rated mental health, pain that limits activities, and activity limitations) or increased use of health services (visits to family doctors, visits to specialists, inpatient stays, visits to physiotherapists, and visits to chiropractors) from highest to lowest for those with arthritis, CJS, SJS, and NJS.
Table 3. Health outcomes and health service use among Canadian individuals ages ≥15 years in 2008*
|Health outcomes|| || || || |
| Physical activity (inactive)||56.7 (55.1–58.3)||52.3 (50.3–54.3)#||47.5 (45.7–49.3)**||47.1 (46.3–47.9)**|
| Pain limits activities||39.8 (38.0–41.6)||27.7 (25.9–29.5)**||6.2 (5.4–7.0)**||4.6 (4.2–5.0)**|
| Activity limitations (sometimes or often)||29.7 (28.5–30.9)||18.7 (17.5–19.9)**||7.1 (6.3–7.9)**||4.8 (4.2–5.4)**|
| Self-rated health (fair or poor)||29.7 (28.3–31.1)||19.6 (18.0–21.2)**||9.4 (8.2–10.6)**||6.5 (6.1–6.9)**|
| Self-rated mental health (fair or poor)||8.8 (7.8–9.8)||8.7 (7.5–9.9)||5.1 (4.1–6.1)**||3.5 (3.1–3.9)**|
|Health service use|| || || || |
| Visits to primary care physicians (≥4 visits)||45.4 (43.8–47.0)||33.9 (32.1–35.7)**||23.8 (22.4–25.2)**||19.2 (18.6–19.8)**|
| Visits to specialists††||43.6 (42.2–45.0)||39.4 (37.4–41.4)**||31.4 (29.8–33.0)**||24.1 (23.3–24.9)**|
| Inpatient overnight††||12.8 (12.0–13.6)||11.2 (10.0–12.4)**||6.4 (5.6–7.2)**||6.6 (6.2–7.0)**|
| Visits to physiotherapists††||14.4 (13.4–15.4)||17.9 (16.5–19.3)**||9.1 (7.9–10.3)**||6.3 (5.9–6.7)**|
| Visits to chiropractors††||13.3 (12.3–14.3)||16.5 (14.9–18.1)#||12.3 (11.3–13.3)||10.1 (9.7–10.5)**|
Table 4 shows the results of the log Poisson regression models examining predictors of the health and health service use outcomes. After adjusting for sociodemographic and lifestyle characteristics, as well as comorbidities, there was still a gradient of poor health outcomes and increased health service utilization with the PRs from highest to lowest for those with arthritis, CJS, and SJS. There were no differences in risks, as shown by the overlapping 95% CIs for the PRs of the arthritis and CJS groups for all the outcomes studied, except for pain that limits activities. Nevertheless, for the latter, the PRs for both arthritis and CJS were substantial (5.07 [95% CI 4.48–5.73] and 4.11 [95% CI 3.65–4.43], respectively). PRs of all outcomes were significantly lower for the SJS group than the CJS group. However, for the SJS group, the PRs for most of the outcomes were significantly higher than for NJS.
Table 4. Results from separate log Poisson regression models for health and health service use outcomes among the Canadian household population age ≥15 years in 2008*
|Health outcomes|| || || |
| Physical activity (ref. active/moderately active)||1.06 (1.02–1.11)||1.05 (1.01–1.11)||0.99 (0.95–1.03)|
| Pain prevents activities (ref. no pain/pain prevents no activities)||5.07 (4.48–5.73)||4.11 (3.65–4.43)‡||1.14 (0.97–1.34)‡|
| Activity limitations (ref. never)||2.64 (2.47–2.82)||2.50 (2.33–2.69)||1.60 (1.47–1.74)‡|
| Self-rated health (ref. excellent/very good/good)||1.95 (1.74–2.19)||2.03 (1.80–2.29)||1.19 (1.03–1.38)‡|
| Self-rated mental health (ref. excellent/very good/good)||1.61 (1.35–1.91)||1.76 (1.49–2.07)||1.23 (1.00–1.51)|
|Health service use outcomes|| || || |
| Visits to primary care physicians (ref. <4)||1.51 (1.42–1.62)||1.45 (1.35–1.56)||1.14 (1.06–1.22)‡|
| Visits to specialists (ref. none)||1.36 (1.28–1.44)||1.43 (1.34–1.52)||1.20 (1.13–1.28)|
| Inpatient stays (ref. none)||1.15 (1.04–1.28)||1.34 (1.18–1.53)||0.86 (0.75–1.00)‡|
| Visits to physiotherapists (ref. none)||2.20 (1.97–2.47)||2.55 (2.28–2.85)||1.36 (1.17–1.58)‡|
| Visits to chiropractors (ref. none)||1.21 (1.09–1.33)||1.33 (1.19–1.49)||1.06 (0.96–1.17)|
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
Almost two-fifths of the respondents (37.7%, >10 million people) reported either arthritis or joint symptoms in a nationally representative health survey of the Canadian population, with 26.1% of the population (7 million people) reporting arthritis or CJS (symptoms on most days in the previous month). Although people with CJS tended to be much younger, a major finding from this study is the similarity of the impact of arthritis and CJS on a range of health-related outcomes, including the use of the health care system, after adjustment for age, sex, ethnicity, socioeconomic status, and lifestyle characteristics.
Previous studies have established the impact of arthritis on activity limitations, reporting fair or poor self-rated overall and mental health and increased health service use (19–21), but studies comparing outcomes associated with arthritis and with CJS have been limited in scope. All studies have found an increased risk of activity limitation in those with joint symptoms compared to those without joint symptoms (6, 8, 9, 13, 20). There is less agreement when the risks for activity limitation are compared for individuals with joint symptoms and with arthritis. Studies from the US generally show a lower risk for those with joint symptoms (8, 9), whereas our study as well as the study by Busija et al show that those with CJS were at a similar risk of activity limitation to the arthritis group (13). A study of the BRFSS population showed increased risks of CJS for reporting only fair or poor health or physical inactivity (10). The new findings from our study are the similarities in outcome for a range of indicators of health service use for arthritis and CJS, and that the impact of CJS on pain that prevents activities is only slightly less than that of arthritis.
It is unclear what the CJS group represents. This group may represent people in the early stages of arthritis who have not yet been diagnosed, particularly since they were more likely to be of a younger age. Findings from a supplement to the 2008 CCHS, the arthritis module of the Survey on Living with Chronic Diseases in Canada (SLCDC), showed a difference of 3 years between the reported onset of joint symptoms and the subsequent diagnosis of arthritis (Tosevsky, et al: unpublished observations). The CJS group may also represent individuals with conditions affecting the joint structures that would fit within the general category of soft tissue disorders. The definition of CJS also varies in the literature based on the duration of the symptoms. As in our study, most studies from population surveys asked questions about symptoms in a 12-month period (6, 8, 13); however, a recent study using data from the 2002 BRFSS defined CJS based on a modified question about symptoms that began >3 months ago (8). We decided to keep the terminology consistent with the majority of the literature. Given the high impact on both the individual and health care system, further research is needed to understand the characteristics of CJS. Some indication of the extent and sites of joint involvement in those reporting joint symptoms would be useful to help understand the nature of these symptoms and to plan appropriate interventions.
The overall proportion of people with arthritis or joint symptoms in this study is lower than the estimate of ∼50% for the Australian population ages ≥18 years (13) and 41% for the US population ages ≥18 years (8). The proportion with arthritis or CJS (26.1%) is also slightly lower than the estimate for the US for a similar combination (8). However, the prevalence estimates for CJS and SJS found here are similar to the estimates by Busija et al for CJS alone and SJS alone (13). It should be noted that the format and order of the questions differed somewhat between the surveys, which may have affected prevalence estimates as well as the proportion with the different outcomes studied.
Our findings that arthritis was more prevalent among women, those of older age, and those with low socioeconomic status, high BMI, and current and former smoking status are consistent with the findings from previous studies (14, 16, 22–24), and the lower prevalence in Asians and higher prevalence in Aboriginals are compatible with earlier findings from the Canadian population (23). Notably, the predictors for CJS in this population other than age were similar to those for arthritis. Our finding that those with joint symptoms were younger is similar to findings from previous studies (8, 9, 13). In contrast to the study by Busija et al (13), we found women were more likely to report CJS, whereas in US studies, which did not differentiate CJS and SJS, men were slightly more likely to report joint symptoms (8, 9, 13).
A substantial proportion of the population also reported SJS. While the impact in this group was less overall than for CJS, the risks of the adverse outcomes studied here were significantly higher than for the population with NJS. Longitudinal studies are needed to determine whether those with SJS go on to develop CJS, and whether those with CJS are at higher risk for later arthritis.
The validity of the question on joint symptoms as an indicator of possible arthritis in US studies has been examined, with conflicting results (8–10, 25). Using data from Chicago residents ages ≥45 years, Feinglass et al (9) suggested that given the similarities between those with joint symptoms and arthritis, joint symptoms should be used to identify individuals with arthritis in population surveys. However, a more recent US study using data from 4 states found that only a very small proportion of those with joint symptoms who had seen a doctor had received a diagnosis compatible with arthritis (8). In this study, joint symptoms were defined as symptoms of any pain, aching, or stiffness in or around a joint experienced in the past 30 days that had begun more than 3 months ago. The questions did not ask whether the symptoms were experienced on most days, as was the case for the CJS group in our study. In view of the younger age and lesser activity limitations in those with CJS compared to those reporting arthritis, the US study concluded that those with CJS “would be unlikely amenable to medical and public health interventions.” Nonetheless, given the similarity in risks for health outcomes and health service use for the arthritis and CJS groups in our study, we suggest that this conclusion deserves reexamination.
Our study has some strengths and limitations. The questions pertaining to joint symptoms were asked only to participants without arthritis, which did not allow us to estimate the occurrence of both joint symptoms and arthritis. However, the analysis of the arthritis module of the SLCDC suggested that over 95% of individuals with arthritis reported joint symptoms (unpublished observations). Also, it is not clear whether the order in which the questions were asked could have affected the reporting of arthritis or joint symptoms. The strengths of this study include that the data came from a nationally representative population health survey with a large sample size and a range of variables, which means we could examine several health outcomes.
This study found that in addition to people who report arthritis, a substantial proportion of the population also reports joint symptoms, a large proportion of whom has CJS. Attention needs to be paid to the nature of these joint symptoms to determine the extent to which they may be an indication of early arthritis or other arthritis-related conditions. A major finding of this study is the similarities in impact on health and health service use between individuals reporting arthritis with those reporting CJS. Whether or not people reporting CJS have arthritis, at the very least, the findings in our study suggest a need for interventions and advice to manage these symptoms to control pain and activity limitation in order to improve health and quality of life and to reduce the impact on the health care system.