Abstract
Background: Multiple sclerosis (MS) is a degenerative neurological disease that affects the central nervous system. Studies estimate a prevalence rate of cognitive impairment in between 43% and 70% of the MS population over the course of the disease (Chiaravalloti & Deluca, 2008). Processing speed (PS) is considered the most or one of the most affected cognitive domains in MS. However, there is a need for more prospective longitudinal studies of cognitive development in people with MS (Sumowski et al., 2018). Therefore, the goals of this study were to 1) Investigate if baseline PS could predict cognitive decline in people with MS and 2) Investigate if baseline PS could predict cognitive dysfunction in MS patients. Methods: This study was done as a part of the Oslo Longitudinal MS cohort project. One of the authors contributed with data collection. A group of 76 people with newly diagnosed MS using a longitidunal prospective design were tested with a battery consisting of the Symbol digit modalities test, the Paced auditory serial addition test, the Color-word interference test, the Controlled oral word association test, the Brief visual memory test, the California verbal learning test second edition and some subtests of the Wechscler abbreviated scale of intelligence. We took two statistical approaches, a linear regression-based model and a linear mixed-effects based model. We chose to use baseline PS as our predictor with variables like other cognitive domains, BDI, FSS, IQ, EDSS & cognitive reserve included in our models when attempting to predict later cognitive outcome. Results: Over the first two follow-ups, there was a trend of increased performance on neuropsychological tests. On the third follow-up, there was a drop in performance relative to the previous follow-up. We found a significant association with a small to medium effect size (d =-0.519) between baseline PS and cognitive dysfunction. We could not find this when controlling for some known confounding variables. Further, we investigated whether a dysfunction in PS at baseline could predict later cognitive dysfunction. There was a significant and moderate effect size (d = 0.738) for this relationship. We could not find a statistically significant association between baseline PS and later cognitive decline. 2 Discussion & conclusion: Baseline PS could not predict cognitive decline. However, baseline PS was significantly related later cognitive dysfunction. We interpret this finding as that performance on cognitive tests are relatively stable over time. EF emerges as a better predictor compared to PS in relation to later cognitive outcome. The pattern of cognitive test performance deviates from earlier MS research litterature. We argue that this is due to our group being an example of the modern MS patient, as practice effects alone cannot fully explain the trend. Additionally, the participant group is highly educated. Further, the defining characteristic of those fulfilling the criteria for decline was an increase in BDI score from T1 to T4 and lower EF T-score at T1. We could not find additional baseline data that can be used for prediction of later cognitive outcome.