1. Top of page
  2. Introduction
  3. Statement of the Problem
  4. Review of the Literature and Evidence
  5. Summary
  6. Acknowledgements

Arthritis and rheumatic conditions (i.e., arthritis) are responsible for major health care expenditures and disability burdens. The impact of arthritis is not restrained by national boundaries. It is one of the most prevalent chronic conditions and is a leading cause of disability in Australia (1), Canada (2, 3), Europe (4), the United Kingdom (5), and the United States (6, 7), affecting an estimated 3 million Australians, 6 million Canadians, 8 million in the UK, almost 43 million people in the US, and 103 million across Europe. With the aging of the baby boomers, these numbers and the associated disabilities will quickly escalate. By 2020 in the US alone, arthritis is projected to affect 60 million people, and the activities of 12 million people may be limited by arthritis (6).

The growing magnitude of people affected by arthritis motivates the need to review what is known about its national costs to identify areas where current information is lacking. In addition, it is important to determine targets for public health efforts that will reduce the costs of and burden from arthritis. This knowledge will facilitate planning research agendas that support informed public policy decisions.

Statement of the Problem

  1. Top of page
  2. Introduction
  3. Statement of the Problem
  4. Review of the Literature and Evidence
  5. Summary
  6. Acknowledgements

The present study describes the costs of arthritis at national levels and provides a perspective for a public health response to contain these costs. First, we review national studies to describe its monetary and nonmonetary costs. Findings from national population-based samples are reviewed for information related to the monetary cost, health care utilization, daily activity loss, and disability experienced by people with arthritis. Many of the studies reviewed, even among recent publications, are not based on contemporary data. This study augments this literature with new information about the economic cost, health care utilization, and disability related to arthritis from more recent US national probability samples, the 1993–1998 Asset of Health Dynamics Among the Oldest Old (AHEAD) (8) and the 1996 Medical Expenditure Panel Survey (MEPS) (9), which is a subsample of the 1995 National Health Interview Survey (NHIS).

Second, the costs of arthritis are considered from a public health point of view. Because it is well known that there are high economic and social costs related to the high prevalence of disability associated with arthritis (6, 7, 10–12), it is important to identify those people with arthritis who are at greatest risk of becoming disabled or deteriorating further. For this purpose, we describe the magnitude of functional limitations among people with symptomatic arthritis and identify those at greatest risk of functional deterioration, using the longitudinal experience of the AHEAD cohort.

Review of the Literature and Evidence

  1. Top of page
  2. Introduction
  3. Statement of the Problem
  4. Review of the Literature and Evidence
  5. Summary
  6. Acknowledgements

Findings based on national probability samples have important methodologic advantages. Specifically, valid analyses from a national probability sample permit inferences about a nation as a whole to provide estimates of the national impact of arthritis. In addition, frequent systematic federal surveys, which have common components such as the NHIS, allow comparisons of measured characteristics over time. Two caveats must be considered for findings based on national studies. First, most national surveys rely on self reported diagnoses of rheumatic diseases. Although these self reports lack diagnostic accuracy, they are relevant from a public policy perspective because many persons with rheumatic disease do not see a health care provider for their symptoms; thus, to measure the full burden of the disease, one must often rely on self reported data. Second, the definition of rheumatic diseases varies across studies. The most commonly used definition of arthritis is based on the International Classification of Diseases, Ninth Revision codes defined by the National Arthritis Data Work (NADW) group (6), but other definitions are used in the literature. Also, arthritis ascertained from interviews generally captures symptomatic disease, but the nature of the questions used to ascertain disease varies and can influence findings. The Appendix provides information on the ascertainment of arthritis reported by each of the reviewed studies.

National studies on the economic costs of arthritis.

Studies evaluating the economic costs of illnesses are plentiful in the medical literature, but most are limited to a selected clinical population or geographic area. Those reporting the costs of arthritis are no exception (13–15). Table 1 summarizes recent studies that estimate the economic costs of arthritis based on national data sets. This table begins with the most recent US study (16) cited in the NADW review article (17). Earlier international studies on the costs of arthritis are succinctly summarized elsewhere (18–20). The economic costs of arthritis in Table 1 are presented in terms of direct costs, for which medical payments are made; indirect costs, which include lost productivity and lost resources unrelated to medical care delivery; and total cost, the sum of direct plus indirect costs. All studies summarized used a prevalence-based approach to estimate the direct costs incurred for health services. Measurement of indirect costs is a matter of some debate among economists. Estimation of indirect costs reported below consistently included productivity loss, and most, but not all, used lifetime costs to account for lost productivity due to premature death. Lost utility due to increased pain or reduced consumption of leisure pursuits is not estimated or included in these studies. All costs are expressed in 2000 US dollars. Adjustment for inflation was based on standard monetary practices, using the core Consumer Price Index (CPI) (e.g., Australia, Canada, France, US) or the core Retail Price Index (e.g., UK) from the respective country. These adjustments are conservative because the health care component has risen faster than other economic components for most countries.

Table 1. Economic cost of arthritis based on national databases, reported in 2000 US dollars
AuthorDatabase yearsDiseaseCountryAge, yearsAnnual national direct cost, $ billionAnnual national indirect cost, $ billionAnnual national total cost, $ billion*
  • *

    Direct plus indirect cost may not sum to total cost due to rounding error.

  • 1989 National Health Interview Survey, 1988 National Hospital Discharge Survey, 1985 National Ambulatory Care Survey, 1980 National Medical Care Utilization and Expenditure Survey, 1985 National Nursing Home Survey, and other sources.

  • 1989–1991 National Health Interview Survey (NHIS).

  • §

    1981–1982 General Practice Morbidity Survey, Office of Population Censuses, 1986 Survey of Disability in Great Britain, and other sources.

  • National studies from L'Institut National de la Statistique et des Études Économiques (INSEE), Centre de Recherche d'Etude (CREDES), Institut National d'Etudes d'emographiques (INED), and other sources.

  • #

    1995 National Health Interview Survey, 1995 National Hospital Discharge Survey, 1995 National Ambulatory Care Survey, 1992 National Hospital Ambulatory Medical Care Survey, 1995 National Survey of Ambulatory Surgery, 1996 National Home Health and Hospice Survey, 1987 National Medical Expenditure Survey, 1995 National Nursing Home Survey, and other sources.

  • **

    1995 National Health Survey and other sources including hospital morbidity data and casemix data.

  • ††

    Indirect costs estimated at triple the direct costs.

  • ‡‡

    1996 Medical Expenditure Panel Survey data on actual medical expenditures.

Rice 1992 (16)1980, 1985, 1988–1989ArthritisUSAll18.661.179.7
Yelin 1996 (25)1989–1991Rheumatoid arthritisUS18–645.510.215.7
McIntosh 1996 (24)1981–1982, 1986§Rheumatoid arthritisUnited Kingdom≥
Levy et al 1993 (23)1985–1990OsteoarthritisFranceAll0.90.51.4
Rice 1999 (21)1987, 1995, 1996#ArthritisUSAll24.668.893.4
Access Economics 2001 (1)1993–1995**ArthritisAustraliaAll1.33.9††5.3
Yelin et al 2001 (22) with additional analyses by authors1996‡‡ArthritisUS≥1642.682.2124.8

A comparison of the economic costs of arthritis over time is provided by 2 studies by Rice (16, 21) that estimated the costs in the US using similar definitions and accounting methods applied to national survey information from the 1980s (1980–1989 data) and the 1990s (1987–1996 data). These data from the general population include people residing in the community or in institutional settings. Total costs due to arthritis from these estimates represented 1.5–2% of the Gross National Product (GNP) from the years those costs were calculated. However, direct costs of arthritis have risen faster than indirect costs. In the 1990s, the $24.6 billion annual direct costs represented approximately 26% (compared to 23% in the 1980s) of the total costs, with the remainder of $68.8 billion representing indirect costs due to lost wages. Direct costs of arthritis rose faster than the medical component of the CPI over the same period (70% versus 59%, unadjusted for inflation) (21). The greater direct costs in the 1990s partly reflect treatment advances that include joint replacement surgery and more effective, but expensive, pharmaceuticals.

Recent information on the costs of arthritis was obtained from 1996 MEPS data (analysis by authors). The total costs due to arthritis represent 2.4% of the 1996 GNP, based on methods described by Yelin et al (22). In 2000 dollars, this translates to $124.8 billion. The direct costs estimate of $42.6 billion represents one-third of the estimated total costs due to arthritis. Analyses indicated that direct cost estimates ranged from $17.0 billion to $57.6 billion, depending on adjustment for demographic, functional status, and other high-cost medical conditions. The estimate of $42.6 billion adjusts for differences in the presence of selected diseases, demographic characteristics, and the number of chronic disease conditions present among those with and without arthritis. Direct cost estimates represent the additional health care expenditures due to arthritis. However, the total expenditures for all medical care (office, emergency room, inpatient, home health, prescriptions, other) for any reason among people with arthritis were $140 billion in 2000 dollars, representing almost 2.8% of the 1996 GNP.

The economic costs of arthritis estimates from the Rice studies (16, 21) used a fairly restrictive definition of arthritis, but cost estimates based on MEPS data used a definition comparable to that of NADW, which additionally included other rheumatic conditions (see Appendix). Under the broader arthritis definition, the MEPS medical expenditure estimate is conservative because it is based on a community-dwelling national sample, which probably underestimates the prevalence of arthritis in the US. The high expenditures based on the MEPS represent a significant commitment of resources on medical care for persons with arthritis.

The costs of arthritis in Australia (1) were estimated using national data from 1993–1995. The total costs of arthritis represented about 1.6% of their 1994 GNP, or almost $5.3 billion in US 2000 dollars. Direct costs of medical care account for about one-fourth of the total costs with the remaining $3.9 billion due to indirect costs of lost earnings and additional needs.

The costs of osteoarthritis (OA) in France (23) were estimated using national data from 1985–1990. The total 1991 costs of OA in France represented about 0.1% of GNP, which translates to almost $1.4 billion in US 2000 dollars, of which almost two-thirds was due to the direct costs of medical care. However, indirect costs in this study may be underestimated to the extent it underestimated lost productivity from those no longer in the labor force, which comprises most of the indirect costs from the other studies cited.

The costs of rheumatoid arthritis (RA) in the UK (24) were estimated using 1981–1986 data based on the entire population. The total 1986 costs of RA in the UK represented about 1.2% of the GNP, which translates to approximately $2.3 billion in US 2000 dollars. Almost half of the total costs among persons with RA were due to the direct costs of medical care, with the remaining $1.2 billion due to indirect costs of lost wages.

The costs of RA (25) in the US were estimated using national 1989–1991 data, based on the NHIS, a sample of the noninstituionalized population. The total costs of RA, which represent 0.4% of the 1990 GNP, translates to almost $16 billion in US dollars from 2000. More than one-third of the total costs among persons with RA were due to the direct costs of medical care, with the remaining $10.2 billion due to indirect costs of lost wages. Because these data do not include institutionalized persons and are restricted in age (18–64 years), both the prevalence and costs of RA in the US are probably underestimated.

These national studies did not directly compare the costs of different rheumatologic conditions. Studies using clinical populations suggest that the per capita costs of RA are greater than those of OA (26). However, this comparison is not yet confirmed in national surveys, which requires cost calculations from the same national data set using the same accounting method applied to different conditions.

Regardless of the definition or the accounting method, the economic impact of arthritis is substantial and rising. Two decades ago, its economic costs in the US were estimated at about 1% of the GNP (18). More recent estimates now indicate that costs among people with arthritis represent 1.5–2.5% of the GNP. The costs of RA alone represent as much as 1% of the GNP. Furthermore, total US health care expenditures among people with arthritis (whether or not the expenditures are related to arthritis) approaches 3% of the GNP. Given its rapidly increasing prevalence due to growing numbers of older adults, future economic costs due to arthritis are only poised to accelerate.

National studies on health care utilization for persons with arthritis.

The impact of arthritis on the health care system is evaluated from national studies that enumerate the use of medical services. These studies primarily utilize 2 types of sources: national samples of individuals and national samples of health care service providers.

Table 2 presents data on the average annual utilization of physician and hospital services, based on self reported interview data from community-dwelling samples. Three sets of utilization findings are based on NHIS interview data from 1976 (27), 1984–1986 (18), and 1995 (analysis by authors taken from 1996 MEPS segment). Although there are some differences in the ascertainment of arthritis (see Appendix), these NHIS data provide a perspective of the demand on the US health care system over time by people with arthritis. The average use of physician and hospital services among people with arthritis was fairly stable over the 1980s and 1990s. Annually, persons with arthritis reported an average of 8 physician visits per person for all conditions, and 0.2–0.3 hospitalizations that lasted an average of 7–8 total days. In the US, persons reporting OA and RA annually averaged 9 and 11–12 physician visits per person, respectively; there was an average of 0.3 hospitalizations lasting on average 8–9 days for both conditions. All Table 2 estimates of the amount of health care used are conservative because they are based on community-dwelling samples, which omit people in institutions, who typically have higher medical utilization.

Table 2. Average self reported health care utilization based on national community dwelling samples*
AuthorSampleDiseaseCountryAge, yearsAnnual number of physician visits/personAnnual hospital admissions/ personAnnual number of hospital days/person
  • *

    NHIS = National Health Interview Survey; RA = rheumatoid arthritis; OA = osteoarthritis; MEPS = Medical Expenditure Panel Survey.

  • Among persons hospitalized.

Kramer et al 1983 (27)1976 NHISRAUSAll11.50.3 
  OA  10.90.3 
Felts and Yelin 1989 (18)1984–1986 NHISArthritisUSAll7.80.38.2
Analysis by authors1996 MEPS (1995 NHIS subsample)ArthritisUSAll8.20.27.3

Table 3 presents estimates of total national utilization in the US based on national health services surveys of the entire population. Two reports by Rice that use the same definition of arthritis (but more restrictive than the NADW definition) applied to National Hospital Discharge Survey data from 1988 (16) and 1995 (21) show fairly stable hospital use across the surveys of about 2.6–2.9 million days of hospital care for persons hospitalized for arthritis. The 1999 Centers for Disease Control and Prevention (CDC) report (7), which applied the NADW definition of arthritis to the 1997 National Hospital Discharge Survey data, estimates that arthritis is the first listed hospital discharge diagnosis annually for 744,000 hospitalizations, lasting a total of 3.8 million days. This represents 2.4% of all hospital discharges and 2.4% of all days of care in that year. The same CDC study (7) also reports 44 million ambulatory care visits attributed to arthritis, of which almost 39 million were to physician offices. Studies summarized in Tables 2 and 3 document that people with arthritis, on the individual and national level, are substantial users of health care services.

Table 3. Total health care utilization based on US health service samples*
AuthorSampleDiseaseCountryAge, yearsTotal annual physician visitsTotal annual hospital admissionsTotal annual hospital days
  • *

    NHDS = National Hospital Discharge Survey; CDC = Centers for Disease Control and Prevention; NAMCS = National Ambulatory Medical Care Survey; NHAMCS = National Hospital Ambulatory Medical Care Survey.

Rice 1992 (16)1988 NHDSArthritisUSAll  2.9 million
Rice 1999 (21)1995 NHDSArthritisUSAll  2.6 million
CDC 1999 (7)1997 NHDSArthritisUSAll 744,000 (first listed discharge diagnosis)3.8 million
 1997 NAMCS, 1997 NHAMCSArthritis All44 million ambulatory (primary diagnosis)  

National studies on lost work, activity, and bed days for persons with arthritis.

The impact of arthritis on daily life is chronicled in Table 4, which summarizes reports on lost workdays, restricted activity days, and days in bed. The number of lost days was estimated from 1976 (27) and 1984–1986 NHIS (18) data. These findings show fairly stable work loss over the mid-1970s and 1980s, averaging 0.1–0.3 days over a 2-week period for persons in the US with RA, OA, or arthritis. Work loss, restricted to labor force participants, averaged almost 1 day over 2 weeks among all persons with arthritis, and approximately one-half day over 2 weeks among persons with RA and OA. Time loss of activity days is far more substantial. Persons with arthritis reported an average of 2 days in which they experienced restricted activity in the prior 2 weeks. A study that focused on people over age 65 years using the 1984 Supplement on Aging (SOA) to the NHIS (28) indicates that older people with arthritis also reported similar rates of restricted activity days. Persons with RA and OA reported 2–3 restricted activity days in a 2-week period. An Australian study based on 1995 data (1) indicates that about 1% of their workforce with arthritis lost work (or school) days, and almost 8% reported days of reduced activity in a 2-week period. A Canadian study (2) based on 1990 data reported that 1.3% of adults experienced restricted activity days in the previous 2 weeks that were attributed to arthritis. The time lost due to days in bed was almost 1 day per 2 weeks, with slightly higher rates reported by people with RA. These studies document that people with arthritis experience substantial time loss related to work and activities.

Table 4. Average restricted activity, bed, and work loss based on national community-dwelling samples*
AuthorSampleAge, yearsDiseaseCountryLost days workLost work days per labor participantRestricted activity daysBed days
With diseaseOverall prevalence
  • *

    NHIS = National Health Interview Survey; RA = rheumatoid arthritis; OA = osteoarthritis; SOA = Supplement on Aging.

Kramer et al 1983 (27)1976 NHISAllRAUS0.2/2 weeks 3.4/2 weeks 1.6/2 weeks
   OA 0.3/2 weeks 2.2/2 weeks 1.0/2 weeks
Felts and Yelin 1989 (18)1984–1986AllArthritisUS0.1/2 weeks0.8/2 weeks2.0/2 weeks 0.8/2 weeks
 NHIS RA 0.3/2 weeks0.5/2 weeks2.4/2 weeks 1.1/2 weeks
   OA 0.1/2 weeks0.5/2 weeks2.5/2 weeks 0.8/2 weeks
Kosorok et al 1992 (28)1984 SOA of NHIS≥65ArthritisUS  1.5/2 weeks  
Badley et al 1994 (2)1990 Ontario Health Survey≥16ArthritisCanada   1.3%/2 weeks due to arthritis 
Access Economics 2001 (1)1995 National Health Survey15–64ArthritisAustralia0.9%/2 weeks 7.8%/2 weeks  

National studies on disability for persons with arthritis.

The International Classification of Impairments Disabilities and Handicaps defines disability as “Any restriction or lack (resulting from an impairment) of ability to perform an activity in the manner or within the range considered normal for a human being.” The American Disabilities Act (ADA) defines disability as a limitation in one's ability to carry out major activities of daily life (29). Many disability measures are consistent with these definitions, some of which are summarized in Tables 5 and 6. This includes disabilities in the areas of work, mobility, activities, and basic tasks related to activities of daily living (ADL), as reported by people with arthritis from national samples.

Table 5. Average disability based on national samples*
AuthorSampleAge, yearsDiseaseCountryWork disability, %Loss of mobility, %Activity limitationComment
With disease, %Overall prevalence, %
  • *

    SSSDW = Social Security Survey of Disability and Work; RA = rheumatoid arthritis; OA = osteoarthritis; NHIS = National Health Interview Survey; HALS = Health and Activity Limitation Survey; NPHS = National Population Health Survey.

Pincus et al 1989 (30)1978 SSSDW ArthritisUS54    
   RA 72    
  18–64OA 53    
Felts and Yelin 1989 (18)1984–1986 NHISAllArthritisUS  19 Unable to do major activity
   RA   29  
   OA   25  
Reynolds et al 1992 (32)1986–1987 HALS≥16ArthritisCanada 18  Mobility = inability to walk 400 m, climb stairs, stand long periods
Yelin 1992 (33)1987 NHIS11–64ArthritisUS  59.73.8Activity limitation attributed to arthritis
CDC 1994 (6)1989–1991 NHISAllArthritisUS   2.8 USActivity limitation attributed to arthritis
Yelin 1995 (31)1991–1993 NHISAllRAUS64    
Badley and Wang 1998 (3)1994 NPHS≥15ArthritisCanada   2.3Activity restriction due to arthritis
CDC 2001 (34)1994–1995 NHISAllArthritisUS   2.9 USActivity limitation attributed to arthritis
Table 6. Disability reports using national samples of community dwelling older adults*
AuthorSampleAge, yearsDiseaseCountryWork disability, %Loss of mobility, %ADL limitation, %Comment
  • *

    ADL = activities of daily living; SOA = Supplement on Aging; NHIS = National Health Interview Survey; LSOA = Longitudinal Study on Aging; HRS = Health and Retirement Survey; AHEAD = Asset and Health Dynamics Among the Oldest Old.

Verbrugge et al 1991 (35)1984 SOA of NHIS≥55ArthritisUS 3517Mobility Loss = difficulty walking ¼ mile
Yelin 1992 (33)1984 LSOA≥70ArthritisUS  36 
Yelin 1995 (31)1992 HRS51–61ArthritisUS41   
Dunlop et al 2001 (36) with additional analysis by authors1993 AHEAD≥70ArthritisUS 65.648Mobility Loss = difficulty walking several blocks or climbing stairs

Table 5 presents findings, primarily from national community-dwelling samples, on disability measures in the areas of work, mobility, and activities. The prevalence of work disability among people with arthritis, based on loss of employment, was estimated from the 1978 Social Security Survey of Disability and Work (30). On the whole, more than half of the people reporting some form of symptomatic arthritis stopped working as of the study year. The group with arthritis was divided into those with symptoms consistent with RA and with OA. More than half of those with symptomatic OA and almost three-fourths of those with RA reported work disability. A later study (31) based on 1991–1993 NHIS data again showed a high proportion of people with RA (64%) were not working. These findings show that work disability is common among people with arthritis, affecting more than half who report symptomatic disease.

The prevalence of loss of mobility among people with arthritis was estimated using data from the 1986–1987 Health and Activity Limitation Survey (HALS) of people living in Canada, a national sample that included health-related institutions. Mobility loss was reported by 18% of Canadians with arthritis, based on self reported inability to walk 400 meters, climb stairs, or stand for long periods (32). This limited information from the 1980s indicates that mobility loss imposes a substantial burden on people with arthritis.

The prevalence of activity limitations was estimated by several national studies shown in Table 5, but different definitions were applied. Activity limitation, defined by inability to perform the person's major activity, was reported by 19% of people with arthritis, estimated from 1984–1986 NHIS data (18). Among persons with RA and OA, 29% and 25%, respectively, reported that they were completely unable to do their major activity. Activity limitation attributed to arthritis was reported by almost 60% of working-aged adults with arthritis, based on the 1987 NHIS, which represented almost 4% of all working-aged adults that year (33). Among the general US population, the prevalence of activity limitations attributed to arthritis was almost 3% in the last decade (estimates from the 1989–1991 NHIS [6] and 1994–1995 NHIS [34] data). This represents almost 8 million persons in the US with activity limitations attributed to arthritis as of 1997 (34). More than 2% of Canadian adults reported activity limitations at home, school, work, leisure, or other activities that were related to arthritis (estimated from 1994 National Population Health Survey data) (3), representing 595,000 Canadians in 1994. These studies indicate that almost one-fifth of people with arthritis are unable to perform their major activity, and even higher rates are reported by people with OA and RA. Furthermore, arthritis-related limitations are experienced by 3–4% of all people in the US and by more than 2% of all Canadians.

Table 6 presents findings from national samples of older adults living in the community on disability measures in the areas of work, mobility, and ADL limitations. The prevalence of work disability, based on the decision to stop working, among people with arthritis approaching retirement, was estimated from the 1992 Health and Retirement Study (HRS) (31). More than 40% of 8.78 million people ages 51–61 reporting arthritis in 1992 reported work disability. Work disability rates as high as 70% were found among people with arthritis who additionally reported activity limitations (31).

The prevalence of loss of mobility was estimated by 2 national studies of older adults, using different age groups and definitions of mobility. Data from the 1984 Supplement on Aging showed that 35% of people aged 55 years or older reporting arthritis had difficulty walking a quarter mile (35). Data from the 1993 AHEAD, using a cohort of people aged 70 years or older, found almost two-thirds of people reporting arthritis had mobility difficulty in walking several blocks or climbing stairs (36). Regardless of the ages and definition used, these studies show that one-third to two-thirds of older adults with arthritis report mobility limitations.

Limitations in performing basic functional tasks related to ADL tasks are particularly serious, because these limitations threaten the ability of older adults to live independently. The prevalence of ADL limitations was estimated by 2 national studies of older adults. Data from the 1984 SOA and its subset, the 1984 Longitudinal Study on Aging, provide ADL information. Among people with arthritis, almost one-fifth of people older than 55 years (35) and more than one-third older than 70 years (33) reported ADL limitation. Data from the 1993 AHEAD study showed that almost half of older adults older than 70 years reported difficulty with 1 or more ADL tasks (36). Although there are differences between the 2 studies in the assessment of ADL limitations and the ascertainment of self reported arthritis (see Appendix), these findings indicate that arthritis poses a serious functional burden for the older population.

All studies listed in Tables 5 and 6, with the exception of the study by Reynolds and colleagues (32) were based on national community-dwelling samples, which omitted institutionalized people (such as people in nursing homes) whose inclusion would likely result in higher disability estimates. Thus, the findings summarized in Tables 5 and 6 are likely to underestimate the rates of disability in the overall population. In addition, these studies demonstrate the diverse nature of documented disabilities.

Despite some methodologic differences, findings from national studies presented in Tables 5 and 6 present a consistent message: There is substantial disability associated with arthritis. Among people with arthritis, 40–70% reported work disability; almost 20% reported loss of mobility, with those rates doubling among older adults; 19–25% were unable to do their major activity, and ADL limitations were reported by one-third to half of the people older than 70 years.

Based on the high levels of disability among people with arthritis, it is not surprising that the largest portion of economic costs from arthritis, as reviewed earlier, is due to indirect costs resulting from lost productivity. This is consistent with other findings where arthritis is listed first in impact on limitations, but far down the rankings in terms of impact on medical utilization, when compared to other chronic conditions (37). Although arthritis is costly in economic terms, it is more expensive in personal and social terms due to interference with daily function.

Risk factors for functional deterioration among people with arthritis.

Finally, we address the costs of arthritis from a public health point of view. Recognizing that high economic and societal costs of arthritis are concentrated among those people who experience functional problems, we evaluate data from a cohort of adults with symptomatic arthritis to determine who is at greatest risk for functional deterioration. For this purpose, longitudinal 1995–1998 data from AHEAD (8) were analyzed. These analyses focus on 2,828 white and minority (Hispanic and African American) older adults with symptomatic arthritis whose functional abilities were ascertained at both 1995 and 1998 interviews.

Symptomatic arthritis was based on an affirmative response to the question, “Have you ever had or has a doctor ever told that you have arthritis or rheumatism?” or a self-report of joint replacement that was not associated with a hip fracture.

Functional limitations are defined in terms of higher-level instrumental activities of daily living (IADL) and basic ADL tasks. The IADL tasks included preparing hot meals, shopping for groceries, making telephone calls, taking medications, and managing money. The ADL tasks included walking across a room, dressing, bathing, eating, using the toilet, and transferring from a bed. A limitation in a specific IADL or ADL task is ascertained from affirmative responses to having difficulty with the task due to health or memory problems or use of equipment for the tasks of walking or getting out of bed. Function limitation in 1995 and 1998 was classified as either none; mild = IADL only; moderate = 1–2 ADL; severe = 3+ ADL limitations. Functional deterioration was identified by progression to a higher (worse) level of function limitation. Note that these analyses omit 1993 data because interview questions to ascertain functional limitations were changed beginning with the 1995 AHEAD interview.

AHEAD oversampled Hispanics and African Americans. The cohort was divided into Hispanic and non-Hispanics; non-Hispanics were further divided into African-Americans, and whites. For simplicity, these 3 mutually exclusive groups are referred to as Hispanics, non-Hispanic whites, and African Americans.

The relative impact of demographic (ethnicity/race, sex, age, marital status), self reported comorbid conditions (diabetes, cancer, hypertension, heart disease, lung disease, obesity, stroke, and psychiatric problems), economic access (income, wealth, education, medical insurance), and health behavior (exercise, smoking, alcohol, weight gain or loss) risk factors on 3-year functional deterioration was evaluated. Logistic regression was used to estimate the odds of 3-year functional deterioration by applying Taylor series methods with between-cluster robust estimation to adjust for the complex sampling design using SUDAAN software (38, 39). Adjustment for potential bias due to nonresponse handled respondents as an additional sampling stage to obtain sampling weights for 1995 and 1998 AHEAD respondents, using standard sampling methodology (40).

Functional limitations among older adults with arthritis.

Table 7 shows characteristics of 2,828 older adults with arthritis in 1995. This arthritis cohort was primarily non-Hispanic white (84%), female (69%), not married (56%), and had a mean age of 79 years. The vast majority (87%) reported comorbid conditions. Regarding economic access factors, about half reported income less than $15,000 (47%), assets less than $50,000 (54%), a high school education (57%), and almost all had health insurance through Medicare (98%). These older adults were largely nonsmokers (93%), who had consumed alcohol (51%), but did not exercise vigorously (71%) on a regular basis (at least 3 times per week). Weight loss (28%) was more common than weight gain (11%) in the 2 years prior to 1995.

Table 7. Characteristics of 1995–1998 arthritis cohort from Asset of Health Dynamics Among the Oldest Old (AHEAD) study*
Characteristics1995 Arthritis cohort n = 2,828
Population, %±95% CI
  • *

    95% CI = 95% confidence interval. Obesity is defined as body mass index (BMI) ≥ 27.8 for men or BMI ≥ 27.3 for women.

Demographic factors  
 African American11.552.53
 Non-Hispanic white83.753.19
 Average age, years79.030.25
Comorbid conditions  
 Heart disease36.681.90
 Lung disease11.631.47
Economic access  
 Education 12 years or more56.502.69
 Annual family income $15,000 or more52.682.24
 Net assets $50,000 or more45.892.96
 Health insurance  
  Medicare only22.151.90
  Medicare and Medicaid11.421.74
  Other health insurance1.400.57
  No health insurance0.330.27
  Medicare and Private/Government64.702.85
Health behaviors  
 No regular vigorous exercise71.372.39
 Current smoker6.670.87
 Ever use alcohol51.262.65
 Weight gain11.161.35
 Weight loss28.181.80

Figure 1 shows the distribution of functional limitations by ethnic/racial group of this older cohort with arthritis. Older minorities (African Americans and Hispanics) were more likely to report IADL or ADL functional limitations than non-Hispanic whites. Of 2,828 older adults with arthritis, 56% of non-Hispanic whites, 48% of Hispanics, and 41% of African Americans were free of such functional limitations.

thumbnail image

Figure 1. Functional limitations among older adults with arthritis by ethnic group. IADL = instrumental activities of daily living; ADL = activities of living.

Download figure to PowerPoint

Functional deterioration among older adults with arthritis.

Of the 2,828 older adults with arthritis, 2,428 were at risk of functional decline (<3 ADL limitations in 1995). Figure 2 shows the frequencies at which this older cohort experienced functional deterioration. Overall, 28% of those at risk (37% Hispanics, 32% African Americans, 27% non-Hispanic whites) had functional decline over 3 years. Substantial rates of functional transition have also been observed from the HRS, a companion study to the AHEAD, which surveyed a younger cohort (aged 51–61 years). The HRS analyses found 12% of that cohort with musculoskeletal conditions had incident disability over 2 years (41).

thumbnail image

Figure 2. Three-year functional deterioration by ethnic group.

Download figure to PowerPoint

Table 8 shows adjusted odds ratios (OR) for minority groups compared with non-Hispanic whites resulting from staged logistic regressions that hierarchically adjust for demographics (Model 1), comorbid conditions (Model 2), economic access factors (Model 3), and health behaviors (Model 4). Hispanics consistently had the greatest odds of functional deterioration among the 3 ethnic groups across all models. Compared with non-Hispanic whites, Hispanics had significantly greater odds of deteriorating function, controlling for demographics and comorbidity (Model 2 adjusted OR = 1.5, 95% confidence interval [95% CI] 1.1–2.0). Further adjustment for economic access (Model 3) and behavior factors (Model 4) reduced this disparity associated with Hispanics.

Table 8. Adjusted odds of 3-year functional decline by ethnicity/race (n = 2,428)*
 Adjusted odds ratios
Model 1 OR (95% CI)Model 2 OR (95% CI)Model 3 OR (95% CI)Model 4 OR (95% CI)
  • *

    Bolded odds ratios (OR) have associated 95% confidence interval (95% CI) that excludes one, indicating statistical significance at a nominal alpha = 0.05 level. Model 1 = OR adjusted for demographic factors (gender, age, marital status); Model 2 = OR adjusted for demographic factors plus comorbid conditions (diabetes, cancer, hypertension, heart disease, lung disease, obesity, stroke, and psychiatric problems); Model 3 = OR adjusted for demographic factors plus comorbid conditions plus economic access (income, wealth, education, medical insurance); Model 4 = OR adjusted for demographic factors plus comorbid conditions plus economic access plus behavioral factors (exercise, smoking, alcohol, weight gain or loss).

Hispanic1.56 (1.16–2.10)1.47 (1.07–2.02)1.11 (0.78, 1.57)1.08 (0.75, 1.56)
African American1.24 (0.92, 1.67)1.22 (0.91, 1.65)0.99 (0.73, 1.35)0.97 (0.72, 1.30)
Non-Hispanic white (reference)    

Table 9 shows adjusted ORs for all risk factors from the final (Model 4) multiple logistic regression analysis. Significant risk factors for functional deterioration include older age, comorbidies (diabetes, hypertension, lung disease, psychiatric disease), low education, weight loss, and lack of vigorous exercise. Because frail individuals tend to be older and are characterized by chronic conditions and weight loss, these characteristics are consistent with geriatric literature on functional limitation risk factors (42, 43). Low education may be a surrogate for poor health, which puts a person at risk for functional decline (44–46). Alternatively, low education may reflect a lack of knowledge regarding available medical services and preventive care. People who are unaware of available interventions would not benefit from potential help.

Table 9. Risk factor-adjusted odds ratio for 3-year functional decline (n = 2,428)*
Risk factorsModel 4 results
Adjusted OR95% CI
  • *

    Model 4 odds ratio (OR) adjusted for demographic plus comorbid conditions plus economic access plus behavioral factors.

  • Bolded OR, have associated 95% confidence intervals (95% CI) that exclude one, indicating statistical significance at a nominal alpha = 0.05 level.

 African American0.970.72–1.30
 Non-Hispanic white  (reference)  
 Age 80 years or older1.821.44–2.29
 Not married0.890.66–1.19
Comorbid conditions  
 Heart condition1.090.84–1.41
 Lung disease1.521.13–2.06
Economic access  
 Education <12 years1.471.15–1.88
  12 years1.230.95–1.59
  13 or more (reference)  
 Income <$7,5001.200.81–1.77
$15,000 or more (reference)  
 Assets <$1,0001.130.76–1.69
$50,000 or more (reference)  
 Health Insurance  
  Medicare only1.030.80–1.33
  Medicare and Medicaid1.050.70–1.57
  Other health insurance1.200.41–3.57
  No health insurance0.580.13–2.58
Medicare and private/government (reference)  
Health behaviors  
 No vigorous exercise1.561.25–1.93
 Smoke now1.140.74–1.77
 Alcohol ever0.850.71–1.01
 Weight gain1.330.90–1.96
 Weight loss1.321.10–1.59
 Weight same (reference)  

The strong association of lack of vigorous exercise with subsequent functional deterioration (adjusted OR = 1.6, 95% CI 1.3–1.9) is particularly important from a public health point of view because this risk factor is amenable to intervention. To provide a perspective on the potential benefit from exercise, we estimated the relative risk of functional progression, using Model 4, for a typical older person with arthritis under 2 risk factor profiles. A low risk profile included no significant detrimental risk factors (e.g., non-Hispanic white person with no comorbidies, high education, no weight change). A high risk profile included all significant risk factors (e.g., an older Hispanic with comorbidies, low education, and weight loss). Under the low risk profile, the relative risk (exercise versus no exercise) is 0.68, which represents a reduction from 16% to 11% of those who would functionally deteriorate with exercise. Under the high risk profile, the relative risk is 0.88, which represents a reduction from 75% to 65% of those who would functionally deteriorate with exercise. This indicates that benefit of exercise provides an absolute risk reduction of 5–10% in functional deterioration among older people with arthritis. Recognizing this potential benefit requires that the person be capable of exercise; a limitation of these findings from the AHEAD study is that people were not asked whether or not they were able to exercise. However, the risk group used for this analysis was limited to people who initially reported none to mild (0–2 ADL) functional limitations, which bases the findings on people who are likely to be able to exercise.


  1. Top of page
  2. Introduction
  3. Statement of the Problem
  4. Review of the Literature and Evidence
  5. Summary
  6. Acknowledgements

This literature from national studies on the costs of arthritis documents a burden of great magnitude, showing that arthritis exacts substantial national tolls. Economic costs of arthritis represent 1.5–2.5% of the GNP and total health care expenditures among those with arthritis approach 3% of the GNP. People with arthritis are substantial users of health care services, averaging 9–10 physician visits and 0.2–0.3 hospital admissions annually. In the US alone, it was the primary listed diagnosis for 44 million ambulatory visits and 3.8 million hospital days in 1997. Finally, people with arthritis experience significant time loss; more than half with work disabilities and as many as 60% with activity limitations attributable to arthritis.

However, the relevance of these findings on the costs of arthritis for policy purposes is limited because the most recent results, based on national data, are largely from the early 1990s. The latest estimates of economic costs are based on 1987–1996 data sets. Recent findings relating medical utilization on the individual level used 1995 data and on the national level used 1997 data. Even the most recent 1997 medical utilization estimates were based on US data that predated the effects of the US 1997 Balanced Budget Act on US health care (47, 48). Documentation of work and activity restrictions utilized data from 1990 or earlier. Findings on disability among people with arthritis analyzed data prior to 1995. The most recent estimates of the costs and burden of arthritis are generally based on national data sets that are 5–10 years old. This deficiency points to the need for contemporary estimates that reflect current treatments and current health care systems.

It is also notable that national population-based studies reviewed almost exclusively evaluate the US and Canadian experience, with a heavy predominance of US data. This demonstrates a need for current information from the international community related to the impact of arthritis at national levels, to promote public policies that are responsive to the needs of all people with arthritis.

Finally, to facilitate a public health response to reduce the high costs and burden of arthritis, risk factors that predicted functional deterioration among older people with arthritis were identified. Older Hispanic adults with arthritis were at greatest risk of functional deterioration. However, this disparity is modified by health behaviors, which are amenable to intervention. Specifically, vigorous exercise could provide an absolute reduction of 5–10% in the progression of functional deterioration among older adults with arthritis. Public health policies, health education, and prevention efforts to maintain functional abilities in people with arthritis should target Hispanics, particularly those with less education. Prevention should include vigorous exercise and medical intervention for health needs. In older adults, weight maintenance should also be promoted.

Implications for making physical activity recommendations.

This evidence supports the promotion of physical activity to reduce the progression of functional limitations in older adults with arthritis. Those who would most benefit from physical activity are those having detrimental risk factors, including Hispanic ethnicity, the very old, low educational attainment, weight loss, and comorbid conditions (diabetes, hypertension, lung disease, psychiatric problems). Because physical inactivity is associated with greater medical costs among adults with arthritis, particularly for people with functional difficulties (49), the potential to reduce functional decline by motivating inactive people to exercise could substantially curtail the growing costs due to arthritis.

Suggestions for future research.

First, future studies are needed that use current national data sets to document the economic costs and burden of arthritis. It is important that policy makers have relevant, up-to-date information to address the impact of arthritis at the individual and national level. Second, population-based studies on the costs and burden of arthritis are needed from the international community. Because most population-based studies focus on the US and Canadian experience, and because arthritis is not restrained by political boundaries, studies from other international communities are needed to promote public policies that are responsive to the needs of all people with arthritis.


  1. Top of page
  2. Introduction
  3. Statement of the Problem
  4. Review of the Literature and Evidence
  5. Summary
  6. Acknowledgements

The authors thank Sandee Feldman and Linda Perloff for library reference help, Orit Almagor, MS, and Miriam Cisternas for programming support, Robert Dunlop for currency information, and Barbara Brown and Lisa Hanagan, RN, for language translation. Functional deterioration analyses used public release data from The Asset of Health Dynamics Among the Oldest Old (AHEAD), sponsored by the National Institute of Aging and conducted by the University of Michigan.


  1. Top of page
  2. Introduction
  3. Statement of the Problem
  4. Review of the Literature and Evidence
  5. Summary
  6. Acknowledgements
  • 1
    Access Economics. The prevalence, cost, and disease burden of arthritis in Australia. Prepared for the Arthritis Foundation of Australia. Accessed July, 2001. URL:
  • 2
    Badley EM, Rasooly I, Webster GK. Relative importance of musculoskeletal disorders as a cause of chronic health problems, disability, and health care utilization: findings from the 1990 Ontario Health Survey. J Rheumatol 1994; 21: 50514.
  • 3
    Badley EM, Wang PP. Arthritis and the aging population: projections of arthritis prevalence in Canada 1991 to 2031. J Rheumatol 1998; 25: 13844.
  • 4
    Jones R. Manifesto for the third millennium. Patient's Network 2000; 5: 23.
  • 5
    Arthritis Research Campaign Web site. Accessed July 2001. URL:
  • 6
    Centers for Disease Control and Prevention. Arthritis prevalence and activity limitations, United States, 1990. MMWR Morbid Mortal Wkly Rep 1994; 43: 4338.
  • 7
    Centers for Disease Control and Prevention. Impact of arthritis and other rheumatic conditions on the health-care system. MMWR Morbid Mortal Wkly Rep 1999; 48: 34953.
  • 8
    University of Michigan. Asset and health dynamics among the oldest old study. Accessed January, 2002. URL:∼hrswww/center/center.html.
  • 9
    Agency for Healthcare Research and Quality. Medical expenditure panel survey household component full year files. Accessed July, 2001. URL:
  • 10
    Ferrucci L, Guralnik JM, Pahor M, Corti MC, Havlik RJ. Hospital diagnoses, Medicare charges, and nursing home admissions in the year when older persons become severely disabled. JAMA 1997; 277: 72834.
  • 11
    Guralnik JM, Simonsick EM. Physical disability in older Americans. J Gerontol 1993; 48: 310.
  • 12
    Manton KG, Corder LS, Stallard E. Estimates of change in chronic disability and institutional incidence and prevalence rates in the U. S. elderly population from the 1982, 1984, and 1989 National Long Term Care Survey. J Gerontol 1993; 48: S15366.
  • 13
    Pugner KM, Scott DI, Holmes JW, Hieke K. The costs of rheumatoid arthritis: an international long-term view. Semin Arthritis Rheum 2000; 29: 30520.
  • 14
    Clarke AE, Esdaile JM, Bloch DA, Lacaille D, Danoff DS, Fries JF. A Canadian study of the total medical costs for patients with systemic lupus erythematosus and the predictors of costs. Arthritis Rheum 1993; 36: 154859.
  • 15
    Clarke AE, Petri MA, Manzi S, Isenberg DA, Gordon C, Senecal JL, et al. An international perspective on the well being and health care costs for patients with systemic lupus erythematosus. Tri-Nation Study Group. J Rheumatol 1999; 26: 150011.
  • 16
    Rice D. Costs of musculoskeletal conditions. In: PraemerA, FurnerSE, RiceD, editors. Musculoskeletal conditions in the United States. Rosemont, IL: American Academy of Orthopedic Surgeons; 1992. p. 14569.
  • 17
    Yelin E, Callahan LF. The economic cost and social and psychological impact of musculoskeletal conditions. National Arthritis Data Work Groups. Arthritis Rheum 1995; 38: 135162.
  • 18
    Felts W, Yelin E. The economic impact of the rheumatic diseases in the United States. J Rheumatol 1989; 16: 86784.
  • 19
    March LM, Bachmeier CJ. Economics of osteoarthritis: a global perspective. Baillieres Clin Rheumatol 1997; 11: 81734.
  • 20
    Lubeck DP. The economic impact of arthritis. Arthritis Care Res 1995; 8: 30410.
  • 21
    Rice D. The economic burden of musculoskeletal conditions, 1995. In: PraemerA, FurnerS, RiceD, editors. Musculoskeletal conditions in the United States. Rosemont, IL: American Academy of Orthopedic Surgeons; 1999. p. 13962.
  • 22
    Yelin E, Herrndorf A, Trupin L, Sonneborn D. A national study of medical care expenditures for musculoskeletal conditions: the impact of health insurance and managed care. Arthritis Rheum 2001; 44: 11609.
  • 23
    Levy E, Ferme A, Perocheau D, Bono I. [Socioeconomic costs of osteoarthritis in France]. Rev Rhum Ed Fr 1993; 60: 63S7S.
  • 24
    McIntosh E. The cost of rheumatoid arthritis. Br J Rheumatol 1996; 35: 78190.
  • 25
    Yelin E. The costs of rheumatoid arthritis: absolute, incremental, and marginal estimates. J Rheumatol Suppl 1996; 44: 4751.
  • 26
    Gabriel SE, Crowson CS, Campion ME, O'Fallon WM. Direct medical costs unique to people with arthritis. J Rheumatol 1997; 24: 71925.
  • 27
    Kramer JS, Yelin EH, Epstein WV. Social and economic impacts of four musculoskeletal conditions: a study using national community-based data. Arthritis Rheum 1983; 26: 9017.
  • 28
    Kosorok MR, Omenn GS, Diehr P, Koepsell TD, Patrick DL. Restricted activity days among older adults. Am J Public Health 1992; 82: 12637.
  • 29
    United States Congress. Americans with Disabilities Act of 1990. Vol Public Law No. 101–336; 1990.
  • 30
    Pincus T, Mitchell JM, Burkhauser RV. Substantial work disability and earnings losses in individuals less than age 65 with osteoarthritis: comparisons with rheumatoid arthritis. J Clin Epidemiol 1989; 42: 44957.
  • 31
    Yelin EH. Musculoskeletal conditions and employment. Arthritis Care Res 1995; 8: 3117.
  • 32
    Reynolds DL, Chambers LW, Badley EM, et al. Physical disability among Canadians reporting musculoskeletal diseases. J Rheumatol 1992; 19: 102030.
  • 33
    Yelin E. Arthritis: the cumulative impact of a common chronic condition. Arthritis Rheum 1992; 35: 48997.
  • 34
    Centers for Disease Control and Prevention. Prevalence of Arthritis — United States, 1997. MMWR Morbid Mortal Wkly Rep 2001; 50: 3346.
  • 35
    Verbrugge LM, Lepkowski JM, Konkol LL. Levels of disability among U.S. adults with arthritis. J Gerontol 1991; 46: S7183.
  • 36
    Dunlop DD, Manheim LM, Song J, Chang RW. Arthritis prevalence and activity limitations in older adults. Arthritis Rheum 2001; 44: 21221.
  • 37
    Verbrugge LM, Patrick DL. Seven chronic conditions: their impact on US adults' activity levels and use of medical services. Am J Public Health 1995; 85: 17382.
  • 38
    Williams R. A note on variance estimation for cluster-correlated data. Biometrics 2000; 56: 2189.
  • 39
    Shah BV, Barnwell BG, Bieler GS. SUDAAN user's manual release 7.5. Research Triangle Park, NC: Research Triangle Institute; 1997.
  • 40
    United States Bureau of the Census. The current populations survey: a report on methodology. Washington, DC: U.S. Government Printing Office; 1963. Technical Paper No 7.
  • 41
    Yelin EH, Trupin LS, Sebesta DS. Transitions in employment, morbidity, and disability among persons ages 51–61 with musculoskeletal and non-musculoskeletal conditions in the US, 1992–1994. Arthritis Rheum 1999; 42: 76979.
  • 42
    Payette H, Coulombe C, Boutier V, Gray-Donald K. Nutrition risk factors for institutionalization in a free-living functionally dependent elderly population. J Clin Epidemiol 2000; 53: 57987.
  • 43
    Chin APMJ, Dekker JM, Feskens EJ, Schouten EG, Kromhout D. How to select a frail elderly population? A comparison of three working definitions. J Clin Epidemiol 1999; 52: 101521.
  • 44
    Crystal S, Shea D. Cumulative advantage, cumulative disadvantage, and inequality among elderly people. Gerontologist 1990; 30: 43743.
  • 45
    Mutchler JE, Burr JA. Racial differences in health and health care service utilization in later life: the effect of socioeconomic status. J Health Soc Behav 1991; 32: 34256.
  • 46
    Kington RS, Smith JP. Socioeconomic status and racial and ethnic differences in functional status associated with chronic diseases. Am J Public Health 1997; 87: 80510.
  • 47
    Ginzberg E, Ostow M. Managed care: a look back and a look ahead. N Engl J Med 1997; 336: 101820.
  • 48
    Manheim LM. Managed care and arthritis care: patients and providers under the new medical care organizations. Arthritis Care Res 1995; 8: 298303.
  • 49
    Wang G, Helmick CG, Macera C, Zhang P, Pratt M. Inactivity-associated medical costs among US adults with arthritis. Arthritis Rheum 2001; 45: 43945.


  1. Top of page
  2. Introduction
  3. Statement of the Problem
  4. Review of the Literature and Evidence
  5. Summary
  6. Acknowledgements
Table  . Definitions of Rheumatic Diseases*
  • *

    NADW = National Arthritis Data Workgroup; ICD-9 = International Classification of Diseases, Ninth Revision; MEPS = Medical Expenditure Panel Survey; NHIS = National Health Interview Survey; AHEAD = Asset and Health Dynamics Among the Oldest Old; CDC = Centers for Disease Control and Prevention; ICDA = International Classification of Diseases; OMS = l'Organisation mondiale de la Santé; ICD-9-CM = ICD-9, Clinical Modification.

Analysis by authors using 1996 MEPS sample from 1995 NHISArthritisNADW classification: ICD-9 Codes: 095.6, 095.7, 098.5, 099.3, 136.1, 274, 277.2, 287.0, 344.6, 353.0, 354.0, 355.5, 357.1, 390, 391, 437.4, 443.0, 446, 447.6, 696.0, 710–716, 719.0, 719.2–719.9, 720–721, 725–727, 728.6–728.9, 728.6–728.9, 729.0–729.1, and 729.4.
Analysis by authors using 1995–1998 AHEADArthritisSelf report of ever having or ever being told by a doctor that person has arthritis or rheumatism, arthritis-related physician visit, or joint replacement not related to a hip fracture.
Badley et al 1994 (2)ArthritisSelf reported medical conditions coded as Statistics Canada Musculoskeletal Impairment Supplemental Coding Scheme: VA00–VA19; VJ00–12, 14, 16–19; VK00–12, 14, 16–19; VM00–12, 14, 18, 19; VN00–12, 14, 16–19; VS00–12, 14, 16, 18, 19; VT00–12, 14, 16–19; VU00–12, 14, 16–19; 710.0–719.9, 725.0–729.9, 274.0–274.9, V43.6.
Badley and Wang 1998 (3)ArthritisAffirmative response to “Arthritis or rheumatism” included in a list of health conditions and presented with the probing question: Do you have any of the following long-term conditions that have been diagnosed by a health professional?”
CDC 1994 (6)ArthritisNADW ICD-9-code classification of arthritis.
CDC 1999 (7)ArthritisNADW ICD-9-code classification of arthritis.
CDC 2001 (34)ArthritisNADW ICD-9-code classification of arthritis.
Dunlop et al 2001 (36)ArthritisSelf reported physician related visit for arthritis or rheumatism within 12 months or joint replacement not related to a hip fracture.
Felts and Yelin 1989 (18)ArthritisSelf reported medical conditions coded as ICD-9: 710.0–719.9.
 Rheumatoid arthritisSelf reported medical conditions coded as ICD-9: 714.0–714.2.
 OsteoarthritisSelf reported medical conditions coded as ICD-9: 715.0–715.9.
Kosorok et al 1992 (28)ArthritisAffirmative response to “During the last 12 months, did you have arthritis or any kind of rheumatism?”
Kramer et al 1983 (27)Rheumatoid arthritisSelf reported medical conditions coded as ICDA: 7122–7125,
 OsteoarthritisSelf-reported medical conditions coded as ICDA: 7130–7139
Levy et al 1993 (23)ArthritisOMS classification of disease codes 715,721
McIntosh 1996 (24)Rheumatoid arthritisNot further defined.
Pincus et al 1989 (30)ArthritisSelf reported arthritis diagnosis and arthritis symptoms.
 Rheumatoid arthritisSelf reported symptoms consistent with symmetric polyarthritis.
 OsteoarthritisSelf reported symptoms consistent with asymmetric oligoarthritis.
Reynolds et al 1992 (32)ArthritisAlgorithm based on responses to 12 activity-specific questions and one nonspecific activity question used to ascertain arthritis/rheumatism.
Rice 1992 (16)ArthritisICD-9-CM codes 098.5, 274.0, 711–716, 720.
Rice 1999 (21)ArthritisICD-9-CM codes: 098.5, 274.0, 711–716, 720.
Verbrugge et al 1991 (35)ArthritisSelf reported medical conditions coded as ICD-9 711.6, 0, 9; 712.b, 8, 9; 714–716; 720.0, 721.
Yelin 1992 (33)ArthritisSelf reported symptoms and diagnoses (respondents asked if a physician told them the specific names of their medical conditions).
Yelin 1995 (31)Rheumatoid arthritisSymptoms consistent with reported diagnosis of RA.
Yelin et al 2001 (22) with additional analyses by authorsArthritisICD-9-CM codes 274, 710–716, 719–721, 725–729.