Disability and depression in rheumatoid arthritis. A multi‐trait, multi‐method investigation

In a sample of 107 patients with classic or definite rheumatoid arthritis (RA), we examined the convergent and divergent validity of measures of disability and depression. Scores on the self-report Disability index of the Health Assessment Questionnaire were highly correlated with physical therapist and spouse ratings of disability. Although Health Assessment Questionnaire disability scores were significantly correlated with self-reported and interviewer-assessed ratings of depression, these correlations were significantly smaller. Factor analyses of the Beck Depression Inventory (BDI) and rheumatologist ratings of BDI items indicated that this measure is highly contaminated by the inclusion of items reflecting RA disease severity. A dysphoric mood subcomponent of the BDI may be a more valid measure of depression in RA populations. Although depression and disability are clearly positively correlated in RA patients, depression scales that include somatic items are likely to yield an overestimate of the association. Finally, self-reported pain intensity was more clearly related to disability and reported recent disease activity than to depression.