Cognitive dysfunction is common in multiple sclerosis (MS), yet few studies have examined effects of treatment on neuropsychological (NP) performance. To evaluate the effects of interferon β‐1a (IFNβ‐1a, 30 μg administered intramuscularly once weekly [Avonex]) on cognitive function, a Comprehensive NP Battery was administered at baseline and week 104 to relapsing MS patients in the phase III study, 166 of whom completed both assessments. A Brief NP Battery was also administered at 6‐month intervals. The primary NP outcome measure was 2‐year change on the Comprehensive NP Battery, grouped into domains of information processing and learning/memory (set A), visuospatial abilities and problem solving (set B), and verbal abilities and attention span (set C). NP effects were most pronounced in cognitive domains vulnerable to MS: IFNβ‐1a had a significant beneficial effect on the set A composite, with a favorable trend evident on set B. Secondary outcome analyses revealed significant between‐group differences in slopes for Brief NP Battery performance and time to sustained deterioration in a Paced Auditory Serial Addition Test processing rate, favoring the IFNβ‐1a group. These results support and extend previous observations of significant beneficial effects of IFNβ‐1a for relapsing MS. Ann Neurol 2000;48:885–892

This study examined the relationship between processing speed (PS) and working memory (WM), as measured by performance on an n-back task, in relapsing-remitting multiple sclerosis (RRMS) patients. Simple PS was defined as reaction time (RT) on the 0-back task and complex PS was defined as RT on both the 1-back and 2-back tasks. Participants were administered all three n-back tasks (0-, 1-, and 2-back). Total correct responses, total dyads, and RTs were recorded. As expected, RT for all participants slowed as WM load increased. MS patients had slower RTs than controls across all tasks, and the difference between groups for RT was greatest during the 2-back task. When RT for simple PS (0-back) was parsed from the 1- and 2-back tasks, MS patients still showed impaired complex PS compared to controls. MS patients also made significantly fewer total correct responses and had fewer dyads than controls only on the 2-back task. These findings suggest that both WM and PS deficits are present in RRMS, and that as WM demand increases (from 1- to 2-back) both PS and WM deficits become more prominent.

Working memory (WM) deficits are common in multiple sclerosis (MS). The Paced Auditory Serial Addition Test (PASAT) is used frequently to measure WM in clinical settings. The n-back paradigm is used often in experimental studies of WM. One unique component of the n-back task is that it provides a measure of reaction time (RT), an additional behavioral index of processing speed and task difficulty. Despite the use of both tasks to measure WM, their common variance has not been documented. We tested 32 MS patients and 20 controls; performance measures were obtained for both tasks. Compared with controls, MS patients generally had poorer performance on both the PASAT and n-back task. MS patients also had slower RTs on the n-back than controls and showed more slowing than controls as a function of WM load. Correlational analyses showed a high correspondence between performance measures on the PASAT and n-back. Principal components analysis pointed to a common feature of the PASAT, n-back, and specific other neuropsychological measures, that is, processing speed. Although the PASAT and n-back were shown to have a significant amount of shared variance, each test has specific advantages and disadvantages for use in clinical populations.