Mortality in rheumatoid arthritis: relationship to single and composite measures of disease activity

Rheumatoid arthritis (RA) is a heterogeneous disease characterized by a variable course of remissions and relapses. Single measures of disease activity at only one point in time may not reflect the overall control of disease activity.

The aim was to determine (i) the predictive value of 20 baseline demographic and disease variables on mortality, and (ii) the relationship between serial measures of the Stoke index (SI; a validated index of disease activity in RA) and mortality in RA.

Mortality in 309 RA patients followed up for a median of 14 yr was analysed retrospectively. The standardized mortality ratio (SMR) was calculated for all causes of death. The predictive values of baseline and time-integrated variables were assessed using multivariate Cox proportional hazards regression analysis.

The SMR was 1.65. At baseline, only nodules, erosions, RA latex titre, white cell count and globulin level were predictive of mortality after correction for age, sex and disease duration. Using a stepwise Cox proportional hazards regression model, the most powerful predictors of mortality were age, nodules and RA latex titre. Individual measures of disease activity and the SI at baseline were not predictive of mortality. However, the mean level of the SI over 12 months was related to mortality (P=0.039).

At baseline, the demographic and disease variables most significantly related to mortality in RA are age, nodules and RA latex titre. Individual measures of disease activity at a single point in time are poor predictors of mortality in RA. However, measurement of the mean level of disease activity over time using the composite SI has a significant relationship with mortality. A high level of sustained inflammation appears to be an important predictor of premature death.