Facebook YouTube Twitter Linkedin
Facebook YouTube Twitter Linkedin


13 October 2021

Fatigue and episodic memory in relapsing-remitting multiple sclerosis: A web-based CANTAB study

Clinical Scientist, Luke Allen, presented on fatigue and episodic memory in relapsing-remitting multiple sclerosis at ECTRIMS (European Committee for Treatment and Research in Multiple Sclerosis).

Read on for the key findings and full poster.


Fatigue affects over 75% of people with MS and impacts quality of life and functional status. Fatigue is also associated with disturbances in mood and cognition, yet these interactions remain underexplored, particularly in real-world data gathering contexts.

In this study, online recruitment platform prolific was utilised to collect objective cognitive, self-reported fatigue, mood, and other measurements in PWMS at home via web-based testing. The goal of this study was to ascertain which, if any, cognitive measures of episodic and working memory are most relevant to fatigue in MS.



  • Participants with a diagnosis of relapsing-remitting MS were recruited via prolific. Objective cognitive, and self-reported measures including fatigue, mood, and disability severity were obtained.
  • Relationships among cognitive and self-report measures were explored using correlation and regression analyses.
  • Participants were classified into fatigued or non-fatigued, compared, and entered into an exploratory machine learning framework along with all outcome measures from Spatial Working Memory (SWM) and Paired Associates Learning (PAL) memory tasks.
  • Random-forest-derived feature importance was used to identify cognitive features most relevant to fatigue, and general linear models were used to assess relationships between measures and group comparisons of fatigued and non-fatigued subjects (with covariates age, sex, education, disability severity, and mood).



  • Measures of feature importance identified two outcome measures from PAL which were highly relevant to fatigue (total attempts; PALTA, and mean errors to success; PALMETS; Figure 1). After adjusting for multiple comparisons, only PALTA was significantly associated with MFIS scores (Figure 2).
  • Fatigued participants exhibited significantly greater PALTA scores (t=3.2, p=0.003, d=1.1), adjusting for covariates (Figure 1), than non-fatigued subjects.
  • Further regression analysis, controlling for mood, disease duration, and disability, revealed that PALTA was significantly associated with fatigue scores.
  • Fatigue was strongly associated with lower mood, and greater disability severity, perceived deficits, and anxiety.
  • PALTA did not significantly differ between severely disabled and moderately/non-disabled participants, nor was it significantly correlated with disability severity.


  • RRMS-related fatigue is significantly associated with poorer episodic memory performance but in terms of the number of attempts to complete the task rather than the number of errors made. This suggests elevated levels of fatigue were associated with repeated mistakes and impaired learning.
  • Associations with mood, anxiety, perceived deficits, and disability confirm the known burden of, and complexity of interactions among, MS symptomatology.
  • The present study confirms the suitability and sensitivity of web-based testing to assess objective measures of cognitive function and their association with self-reported and clinically relevant outcomes in neurological disorders such as MS.
  • Objective measures of cognition hold value for the development of digital tools and biomarkers to assess and characterise fatigue and fatiguability.

View poster

Contact us 

Tags : cognition | cantab | cognitive testing | cognitive science | fatigue | web-based testing

Author portrait

Luke Allen - Clinical Scientist