Predictive Models of Cognitive Fatigue in Multiple Sclerosis

Objective Cognitive fatigue (CF) can be defined as decreased performance with sustained cognitive effort. The present study examined the interrelatedness of disease severity, fatigue, depression, and sleep quality in order to evaluate their predictive roles of CF in MS. Four theoretical models examining these variables were assessed. Methods Fifty-eight individuals with a diagnosis of MS were recruited. CF was measured by examining last third versus first third performance on the Paced Auditory Serial Addition Test (PASAT). The PASAT and self-report measures of fatigue, depression, and sleep quality were administered. Path analysis was used to evaluate each of the models. Results CF was correlated only with depression (r = .362, p = .006) and sleep quality (r = .433, p = .001). Sleep quality was the greatest significant independent predictor of CF ( = .433, t((1,55)) = 3.53, p < .001), accounting for 17.3% of the total variance. The best fitting model showed sleep quality as the largest contributor to CF; however, depression played a smaller predictive role. Furthermore, depression emerged as the strongest predictor of sleep quality and fatigue. Disease severity weakly predicted depression. Conclusions Sleep quality is the most significant predictor of CF in MS. As such, sleep quality may be a treatable cause of CF. Sleep quality itself, however, accounted for only 17.3% of the variance in CF suggesting that other variables which were not formally assessed in this sample (e.g., anxiety, etc.) may also play a predictive role. Follow-up studies should evaluate how results may differ with a larger sample size.