Unravelling the biological, environmental, physical and psychosocial factors impinging on outcomes in early rheumatoid arthritis

Objective:
The aim was to study the course of disease activity and disability in early rheumatoid arthritis (RA) and the relationship between the two over 24 months. Baseline predictors of adverse outcome were sought as potential tools for targeting future therapy in RA.
Methods:
Yorkshire Early Arthritis Register (YEAR) recruited patients with a clinician-made diagnosis of RA and symptoms of ≤24 months. Change in disease activity, measured using disease activity score from counts of 28 joints (DAS28) and disability, measured using the disability index component of the health assessment questionnaire (HAQ-DI), after 6 and 12 months, were outcomes in linear regression models. In order to take advantage of the longitudinal data and examine change in DAS28 and HAQ-DI over time, latent growth curves were applied to the data. Latent class growth analyses (LCGA) identified trajectories of change in DAS28 and HAQ-DI over 2 years. Multinomial logistic regression identified predictors of trajectory group membership. Finally, a dual trajectory analysis explored the relationship between DAS28 and HAQ-DI trajectories. Missing data were handled using multiple imputation and maximum likelihood estimation.
Results:
Data from 1416 YEAR cases were included. Baseline fatigue visual analogue score (VAS) consistently predicted worse outcome including lesser change in DAS28 and HAQ-DI after 6 and 12 months and adverse DAS28 and HAQ-DI trajectories. There were 2 DAS28 and 2 HAQ-DI trajectories and 4 dual trajectory groups: half of patients followed the most favourable, low DAS28/low HAQ-DI, trajectory and only 1% followed a high DAS28/low HAQ-DI trajectory. High DAS28/high HAQ-DI and low DAS28/high HAQ-DI trajectories were more likely for females, cases from more deprived socio-economic areas, and those reporting greater fatigue at baseline. The high DAS28/high HAQ-DI trajectory was more likely in cases with higher baseline DAS28, but baseline DAS28 did not predict low DAS28/high HAQ-DI trajectory group membership. Membership of the low DAS28/high HAQ-DI group was more likely in cases with greater contribution of subjective components to baseline DAS28 .
Conclusion:
Baseline fatigue consistently predicted adverse DAS28 and HAQ-DI over 2 years. In some cases, there was persistent disability and disease activity, whilst another group had greater disability despite less disease activity. Compared to those with low disease activity and disability, membership of these 2 groups was more likely with greater baseline fatigue. Further research into the drivers of fatigue in RA may help target therapies and limit disability in RA.