Prevalence and Incidence

Prevalence and incidence
The proportion of people in the general population who experience osteoarthritis is a useful measure of its
impact. For an intermittent episodic problem such as osteoarthritis, prevalence needs to be measured across
a defined period of time. Regular national data, based on self-reports, are now available about its prevalence
through the National Health Surveys conducted by the Australian Bureau of Statistics. However, no national data
based on case definition by physical examination or radiological evidence are available.
The incidence of osteoarthritis can be modelled using the prevalence data and other epidemiological parameters
(AIHW: Mathers & Penm 1999). There are no direct sources for this information.
Self-reported prevalence
In the 2001 NHS, about 75 out of 1,000 Australians reported osteoarthritis. This equates to around
1.4 million people. This estimate is based on the NHS question: whether the survey respondent ‘currently has
osteoarthritis’. The NHS survey assumes all reported cases of osteoarthritis to be long term (i.e. conditions that
have lasted at least six months, or that are likely to last six months or more). The prevalence of osteoarthritis
increases with age: relatively few people at younger ages report having it. By age 55, however, the prevalence
rises sharply (ABS 2002).
Osteoarthritis is reported more frequently by females than males (92 compared with 57 per 1,000 persons in
2001). The difference persists across all ages. In 2001, the prevalence was 331 per 1,000, among females
aged 65–74, rising to 374 per 1,000 among those aged 75 and over. Comparable rates among males that year
were 186 and 236 per 1,000, respectively (Figure 3.2).
Several studies have reported a crossover in osteoarthritis prevalence between the two sexes around the
age of 45. Males are affected more commonly below age 45, whereas above age 45 females are affected not
only more frequently but also more severely (Kelsey & Hochberg 1988). No such crossover was noted in the
NHS self-reports.
Other regional/jurisdictional sources confirm the high prevalence of osteoarthritis in Australia.
• In an omnibus survey of the South Australian population, Hill et al. (1999) estimated the prevalence of
osteoarthritis among those aged 15 and above to be around 86 per 1,000 persons (51 per 1,000 males and
111 per 1,000 females). The prevalence increased with age, rising above 261 per 1,000 among those aged
70 and above.
• A study in North Sydney estimated the prevalence of osteoarthritis to be around 79 per 1,000 persons
(March et al. 1998).
• Symptomatic osteoarthritis was also reported by more than one-quarter of persons aged 60 and above in the Dubbo Osteoporosis Study (Jones et al. 1995).
There is no regular time series available on the prevalence of osteoarthritis in Australia. The NHS indicates
that the prevalence of osteoarthritis has risen from 69 per 1,000 persons in 1995 to 75 per 1,000 persons in
2001. The Survey of Disability, Ageing and Carers (SDAC) categorises osteoarthritis together with other forms of
arthritis. No comparative information on that account is therefore available.
Radiological evidence
The radiological evidence suggests much higher prevalence of osteoarthritis than the self-reports. Osteoarthritisrelated
changes were noted on x-ray in more than 50% of persons over the age of 65, and almost universally in
those after age 85 in North Sydney (March 1997).
The radiographic evidence is based on the presence of osteophytes, joint space narrowing, subchondral cysts
and bone remodelling, with the severity of the condition, graded from none (0) through doubtful (1), minimal (2)
and moderate (3) to severe (4). One of the problems with this case definition is that many people with positive
x-ray findings report no pain or disability (Lawrence et al. 1966). Conversely, some individuals report pain but
show no radiological evidence. In addition, primary sources of data on osteoarthritis are based on radiographs of
only a few joints in each person (McDuffie et al. 1987).
Incidence
Direct estimation of the incidence of osteoarthritis is difficult. The Australian Burden of Disease Study, using
DISMOD software to model epidemiological parameters, estimated the incidence of radiological osteoarthritis in
Australia to be 2.9 per 1,000 females and 1.7 per 1,000 males (AIHW: Mathers & Penm 1999). This translates
to some 27,000 new cases annually. The incidence increases with age. It is highest among females between the
ages of 65 and 74 (14 per 1,000) and among males aged 75 and over (9 per 1,000).
To date, no prospective population-based study has been undertaken in Australia to estimate the incidence of
osteoarthritis. More recent longitudinal surveys in the United Kingdom suggest that the incidence may be higher,
with 20–30 per 1,000 females aged 50 to 60 developing new radiological knee, hip and spinal osteoarthritis
each year.
Estimating the prevalence and incidence of osteoarthritis is complicated by a variety of factors. The estimates
may vary depending upon the number of joints studied, the age and sex of the respondents, and the reporting
method used (physical examination, x-ray, self-report). The correspondence between the radiological evidence,
clinical features and self-assessment is also variable.