The Performance Speed Test (PST) is an electronic implementation of the symbol digit modalities test. The PST has been longitudinally administered to a multi-institutional multiple sclerosis (MS) population as part of the MS PATHS (Partners Advancing Technology and Health Solutions) initiative.

Data were extracted from the MS PATHS database. Inclusion required a diagnosis of MS and ≥1 PST evaluation. Correlations were determined between numeric predictors and PST scores using Spearman’s Correlation Coefficient. Analysis of variance with planned contrasts was used for categorical variables. Linear regression predicting PST was employed adjusting for important covariates.

We identified 9,004 participants (15,000 unique visits). The mean age was 47 years and the mean disease duration was 11.74 years. Age most strongly correlated with PST performance (r = -0.47, p<0.001) with years from diagnosis also negatively correlating with PST scores (r = -0.33, p<0.001). PST scores positively correlated with education years (r = 0.27, p<0.001). Participants with private insurance had mean PST scores that were 9.9 points higher than those with public insurance (p<0.001). Participants with relapsing-remitting MS (RRMS) had mean scores that were 2 points higher than those with progressive MS (p<0.001) and those with relapsing onset MS had mean scores 1.23 points higher than those with progressive onset MS (p<0.001). The regression model accounted for 43.1% of PST variance and all predictors were significant. PST score decreased by 0.43 points for each year of age (p<0.001). Having primary progressive MS reduced scores by 10.26 points relative to RRMS (p<0.001). A significant interaction revealed that scores dropped at a faster rate per year for those with progressive MS. 

Multiple predictors of PST performance were identified. Age and MS phenotype exerted an especially strong influence.