How "soft" are soft neurological signs? The relationship of subjective neuropsychiatric complaints to cognitive function in systemic lupus erythematosus.

OBJECTIVE: As part of a longitudinal study of cognitive function in systemic lupus erythematosus (SLE), we documented the range and frequency of subjective neurologic and/or psychiatric (NP) complaints in Never-NP-SLE patients, and related these to cognitive function, using the latter as a primary indicator of nervous system involvement. METHODS: Thirty patients with SLE who did not have major neurologic and psychiatric involvement underwent baseline and followup neuropsychological testing roughly 5 years apart. Within 0-13 months prior to retesting, each patient completed a 42 item questionnaire recording NP symptoms. RESULTS: The group as a whole endorsed 26% of symptoms. Fourteen patients labelled high endorsers (> 35% of items) endorsed, on average, 42% of symptoms. There was a significant association between higher item endorsement and lower cognitive function (r = -0.46, p < 0.02) and significantly poorer cognitive performance in the high compared to low endorser groups (t = -3.07, p < 0.005). In addition, a subset of 8 items was endorsed at least twice as often by SLE patients as by patients with rheumatoid arthritis (n = 12) or healthy controls (n = 10). CONCLUSION: These results suggest that "minor" NP symptoms and, in particular, a small subset of subjective complaints may be sufficient to raise suspicion of subclinical nervous system involvement in the absence of clinically evident NP-SLE.