Predictors of cognitive fatigue and fatigability in multiple sclerosis.

BACKGROUND

Persons with multiple sclerosis (PwMS) commonly experience cognitive fatigue (CF), defined as a decrease in cognitive performance with sustained activity, yet CF remains understudied. Further, the relationship between subjective CF and objective CF, or cognitive fatiguability, has not been fully elucidated in previous studies. Understanding the predictors of cognitive fatigue may scaffold the development of interventions that target this symptom. The objective of this prospective study was to evaluate the extent to which depression, anxiety, information processing speed, and sleep quality predict subjective and objective CF.

METHODS

PwMS were recruited from one academic MS clinic in London (ON) Canada. Objective CF was measured by the Paced Auditory Serial Addition Test (PASAT), where performance on the last third of the PASAT is compared to performance on the 1st third, a validated measurement of objective CF. Subjective CF was measured by the cognitive component of the Modified Fatigue Impact Scale (MFIS). Additionally, depression, anxiety, information processing speed, and sleep quality data was collected. All assessments took place on the same day. Pearson's r was calculated to examine the relationship among all continuous outcome measures while linear regression analyses were used to examine predictors of subjective and objective CF.

RESULTS

The sample consisted of 53 subjects who were mostly female (37; 69.8%) with a mean age of 44.2 years; the majority (47; 88.7%) had relapsing MS. Objective CF and subjective CF were not significantly related (r = - 0.16). Further, there was no statistically significant predictors of objective CF noted. In contrast, subjective CF demonstrated a statistically significant relationship with the Symbol Digit Modalities Test (SDMT; r = - 0.29, p = .05), Hospital Anxiety and Depression Scale (HADS), depression subscale (r = 0.61, p < .001), HADS anxiety subscale (r = 0.54, p < .001), and sleep quality (r = 0.33, p = .02). Additionally, all variables predicted subjective CF, R = 0.384 [F (6, 40) = 5.783, p = .0002]. In particular, anxiety significantly predicted subjective CF when controlling for depression, information speed, and sleep quality.

CONCLUSION

This study demonstrated that subjective CF is significantly predicted by anxiety, and strongly influenced by information processing impairment and depression. Addressing underlying affective factors, such as anxiety or depression, may help alleviate perceived or subjective CF among PwMS, thus improving their function and quality of life. Further studies with a larger sample size or longitudinal follow up may help define predictors of objective CF.