OBJECTIVE

To assess patient's self-reported global assessment of health (PGA) in relation to inflammation, pain and disability in patients with rheumatoid arthritis (RA).

METHODS

Data were obtained from the Better Anti-Rheumatic PharmacOTherapy early RA cohort, in which patients were scheduled for follow-up of 15 years. Longitudinal PGA trajectories were identified using hierarchical agglomerative clustering procedure with Wards methods, based on PGA assessments during the first 2 years. Multivariate linear regression and mixed models were used to evaluate associations between PGA, inflammation, pain and disability.

RESULTS

Totally 2238 patients were included in 1991-2006. Three PGA trajectories were identified, low, medium and high, including 34%, 48% and 18% of the patients, respectively. The similar PGA patterns were shown for the inclusion periods before and after 1999. The patients in the low PGA group were less often women, had lower body mass index, lower levels of inflammatory variables, visual analogue scale (VAS) pain and Health Assessment Questionnaire (HAQ) scores at baseline than patients in the higher PGA groups. After adjustments, smoking and antibodies to cyclic citrullinated peptide status differentiated between being in low and high PGA groups. During the first year, all PGA groups showed improved outcomes, most pronounced in low PGA group. Fewer patients in high PGA group achieved remission, independent of remission criteria used. The outcomes were consistently different between the PGA groups from 6 months and onwards. The PGA levels from baseline over 15 years were best explained by VAS pain, followed by HAQ score, Disease Activity Score on 28 joints-3 and tender joint count, adjusted R up to 77%, 41%, 27% and 26%, respectively.

CONCLUSION

Persistently higher PGA levels in RA were mostly related to pain and disability.