Ample evidence shows that rheumatoid arthritis (RA) is associated with a shortened life expectancy, with indicators of severe disease and rheumatoid factor (RF) as risk factors.1,–,3 Early and aggressive treatment strategies over the past decade have resulted in more favourable functional and radiological outcomes, probably also having an impact on mortality. Methotrexate treatment has been associated with decreased preterm deaths.4 Some studies have found no difference in mortality between patients and the general population.5 6

To examine the present status in Finland, we used the national registry of the Social Insurance Institution, covering all residents, to identify all incident cases of RA with an index day between 1 January 2000 and 31 December 2007. The case definition was eligibility—for the first time during the lifetime—for special reimbursement for disease-modifying antirheumatic drugs because of RA (ICD-10 code M05 or M06), which was based on an examination performed by a clinic of rheumatology or by a rheumatologist. Death certificate data were obtained from the National Population Registry until the end of 2008, and mortalities were calculated.

During the 8-year period, 14 878 new patients with RA aged over 16 years were identified; 9319 (63%) had an ICD-10 code of M05—that is, RF+ RA, while 5559 (37%) had the code of M06—that is, RF– RA. A total of 10 119 (68%) were female (62% RF+), mean age 55.8 (SD 15.8) years, and 4759 (32%) were male (64% RF+), mean age 57.5 (SD 13.9) years.

The 8-year survival rates were 89.6% (95% CI 88.6 to 90.4) in women and 81.8% (95% CI 80.4 to 83.4) in men (figure 1). The age-adjusted standardised mortality ratio (SMR) was 0.95 (95% CI 0.88 to 1.03) in women and 0.98 (95% CI 0.90 to 1.07) in men. In the entire cohort, age- and sex-adjusted SMR was 0.97 (95% CI 0.91 to 1.02).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Survival of incident patients with rheumatoid arthritis. The Kaplan–Meier curve indicates the estimated percentages (with 95% CI) of patients alive after the index day. The number of patients alive at time points 0, 2, 4, 6 and 8 years and on whom the estimates are based are also shown. The upper numbers indicate female subjects, the lower ones male subjects.

RF+ patients had a higher death rate than RF– subjects; the age-adjusted HR (95% CI) was 1.48 (1.24 to 1.75) in women and 1.34 (1.11 to 1.62) in men. Age- and sex-adjusted SMRs were 1.08 (95% CI 1.01 to 1.16) in RF+ patients and 0.77 (95% CI 0.69 to 0.85) in RF– patients. Figure 2 shows SMRs by age group.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Sex- and age-adjusted standardised mortality ratio with 95% CI by age group according to rheumatoid factor (RF).

Our results—although speculative and only hypothesis-generating until confirmed—suggest that new patients with RA have on average no increased risk for death. Our follow-up period, however, is short and differences in mortality between patients and the general population may become apparent with longer duration of the disease.7 The presence of RF remains a risk factor for mortality in RA.3 8 The RF– patients had reduced mortality compared with the general population.

A key factor in epidemiology is the case definition. We have no data on fulfilment of the American College of Rheumatology (ACR) classification criteria for our patients. Our case definition—eligibility for special reimbursement of antirheumatic drugs—involves the clinical diagnosis of RA and a rheumatologist-assessed need to start antirheumatic drug treatment. The incidence of RA in our study is somewhat higher than in earlier reports in some regions of Finland using the ACR criteria.9 10 This suggests inclusion of patients with milder and earlier RA than previously. Patients with other longstanding idiopathic inflammatory arthritides, including undifferentiated arthritis, are also eligible for special reimbursement of antirheumatic drugs in Finland, but these are assigned other ICD codes than those for patients with RA.

The good prognosis of our patients may reflect the impact of effective drug treatment.