Abstract

Multiple sclerosis (MS) is a chronic, immune-mediated disease of the central nervous system in which the immune system attacks the myelin sheaths of central axons, producing disseminated lesions that disrupt neural conduction in the brain, spinal cord, and optic nerves. It interests cognitive psychology because roughly half of those affected develop measurable cognitive impairment, most prominently a slowing of information-processing speed, alongside deficits in episodic memory, complex attention, and executive function that appear early and progress independently of physical disability. This article treats the cognitive profile of MS as a case study in how distributed white-matter damage degrades the timing and coordination of cognition rather than erasing discrete stores of knowledge.

Keywords: multiple sclerosis, demyelination, processing speed, cognitive impairment, white matter

Multiple sclerosis was first delineated as a distinct clinico-pathological entity by Jean-Martin Charcot in the 1860s, who linked the disseminated hardened plaques found at autopsy to a characteristic triad of clinical signs. Modern accounts define it as an inflammatory, demyelinating, and neurodegenerative disease in which focal immune attack strips the myelin from central axons and, over time, transects the axons themselves (Compston & Coles, 2008; Reich et al., 2018). It is among the most common causes of non-traumatic neurological disability in young adults, with prevalence rising worldwide to an estimated 2.8 million people (Walton et al., 2020). For cognitive psychology, MS is important less as a lesion catalogue than as a natural model of *disconnection*: because its lesions fall in white-matter tracts that carry signals between grey-matter regions, MS degrades the speed and synchrony of communication across cognitive networks, and the resulting profile — a disproportionate slowing of mental operations — illuminates how much of cognition depends on conduction timing rather than on the integrity of any single processing centre (Chiaravalloti & DeLuca, 2008; Benedict et al., 2020).

Key Takeaways
  • Multiple sclerosis is an immune-mediated attack on central myelin, producing lesions disseminated in space and time throughout the brain, spinal cord, and optic nerves.
  • Demyelination slows or blocks the saltatory conduction of nerve impulses; cumulative axonal transection drives the irreversible disability of later disease.
  • Cognitive impairment affects roughly 40 to 65 percent of people with MS, with slowed information-processing speed the most characteristic and earliest deficit.
  • The cognitive profile reflects disconnection of white-matter tracts rather than focal loss of a knowledge store, so recognition often outperforms free recall and speed suffers before accuracy.
  • Diagnosis rests on demonstrating dissemination in space and time by the McDonald criteria; disease-modifying therapies now substantially reduce relapse rates.

What Multiple Sclerosis Is

Multiple sclerosis is defined by the coincidence of three processes in the central nervous system: inflammation, demyelination, and neurodegeneration (Reich et al., 2018). Autoreactive lymphocytes cross the blood–brain barrier and mount an attack on components of the myelin sheath — the lipid-rich insulating wrapping, formed by oligodendrocytes, that enables rapid nerve conduction. The attack leaves focal areas of myelin loss, called plaques or lesions, scattered through the white matter and, as later work established, through grey matter as well. The disease takes its name from these lesions: *multiple* because they are disseminated across the neuraxis, *sclerosis* because the healed plaque is a hardened, gliotic scar.

The clinical consequence of a lesion depends entirely on where it falls. A plaque in the optic nerve produces optic neuritis and visual loss; one in the spinal cord produces weakness, sensory disturbance, or bladder dysfunction; one in the cerebellar pathways produces ataxia and tremor. Because lesions accumulate at different sites and different times, the hallmark of MS — and the basis of its diagnosis — is dissemination *in space* (lesions in more than one central location) and *in time* (lesions arising on more than one occasion), a principle formalised in the McDonald diagnostic criteria and their subsequent revisions (Thompson et al., 2018). Beyond the focal plaque, a diffuse, low-grade neurodegeneration proceeds throughout the normal-appearing white matter and cortex, and it is this component, rather than the dramatic relapse, that best predicts long-term disability (Filippi et al., 2018).

Figure 1

Saltatory Conduction in a Healthy and a Demyelinated Axon

A healthy myelinated axon conducting rapidly by leaping between nodes, above a demyelinated axon in which the impulse slows and stalls at a bare segment. Two horizontal axons are drawn. The upper axon carries evenly spaced myelin segments with gaps at the nodes of Ranvier, and arcs show the impulse jumping quickly from node to node. The lower axon has one myelin segment stripped away; over the bare region the impulse arrow is dashed and marked slowed or blocked, illustrating degraded conduction after demyelination. Healthy myelinated axon Demyelinated axon slowed / blocked
Note. In the healthy axon the impulse leaps rapidly between nodes of Ranvier (green arcs). Where myelin is stripped away (dashed segment), current leaks across the bare membrane and the impulse slows or fails to propagate. Original schematic.

Types of Multiple Sclerosis

The Medical Subject Headings (MeSH) classification, under which MS is indexed as descriptor D009103 within the parent category of central-nervous-system demyelinating autoimmune diseases, recognises two narrower descriptors that correspond to the disease's principal courses. These MeSH subtypes are an *indexing* classification — a scheme for organising the literature — not a mechanistic taxonomy of distinct diseases, and the modern clinical phenotypes overlap them rather than mapping one-to-one; the categories are also not mutually exclusive, since a relapsing course commonly evolves into a progressive one in the same person (Lublin et al., 2014).

SubtypeIn brief
Multiple Sclerosis, Relapsing-RemittingThe most common presentation, accounting for roughly 85 percent of new diagnoses. Discrete relapses of neurological symptoms are followed by partial or complete recovery, with clinical stability between attacks.
Multiple Sclerosis, Chronic ProgressiveA grouping for courses marked by steady accumulation of disability rather than discrete relapses, encompassing the primary-progressive form (progressive from onset) and the secondary-progressive form into which most relapsing-remitting disease eventually converts.

The 2013 revision of the clinical-course definitions refined these headings into the phenotypes used in practice — clinically isolated syndrome, relapsing-remitting, primary-progressive, and secondary-progressive MS — and added modifiers for disease activity and progression, precisely because the older categorical scheme obscured the continuum between inflammatory relapse and neurodegenerative progression (Lublin et al., 2014).

How Demyelination Disrupts Conduction

In a healthy myelinated axon, the impulse does not travel smoothly along the membrane; it leaps between the nodes of Ranvier — the gaps in the myelin sheath where voltage-gated sodium channels are concentrated — in a process called saltatory conduction. Myelin is a low-capacitance, high-resistance insulator, and by confining current flow to the nodes it lets conduction proceed at up to a hundred metres per second while sparing metabolic cost. Demyelination removes this insulation. Current now leaks across the bare internodal membrane, the local circuit is degraded, and conduction slows dramatically or fails altogether — a conduction *block* (Compston & Coles, 2008).

Demo 1 · Demyelination Slows, Then Blocks, Conduction

A myelinated axon conducts by leaping between nodes of Ranvier. As myelin is stripped away, current leaks across the bare membrane and conduction velocity falls steeply; past a critical loss the impulse fails to propagate at all. Adjust the myelin lost and watch the conduction velocity.

block1000velocity (m/s)myelin lost (%)
Conduction velocity: 55 m/s (from 100 m/s intact). Partial demyelination slows the impulse and spreads out its arrival time, degrading computations that depend on synchronous input.

Partial demyelination slows conduction and, critically, introduces variability in conduction time: impulses that once arrived in tight synchrony now spread out in their arrival, degrading any computation that depends on temporal coincidence. Because so much of neural processing is a matter of signals converging within a narrow time window, even modest slowing across many tracts compounds into a measurable slowing of behaviour. Over time the demyelinated axon is also left metabolically vulnerable, and the accumulation of irreversible axonal transection within lesions — the severing of the nerve fibre itself — converts a reversible conduction problem into permanent disconnection (Trapp et al., 1998). This distinction between reversible demyelination and irreversible axonal loss underlies the clinical difference between a relapse that remits and disability that persists.

Measuring Disability and Disease Course

The severity of MS has been quantified for four decades by the Expanded Disability Status Scale (EDSS), which rates neurological impairment on a 0-to-10 ordinal scale anchored in the examination of functional systems and, at its higher reaches, in walking ability (Kurtzke, 1983). The EDSS is heavily weighted toward ambulation, and one of its recognised limitations is precisely that it is comparatively insensitive to cognitive change — a person can decline substantially in processing speed and memory while their EDSS score, tracking the legs, barely moves. This insensitivity is one reason cognitive assessment became a separate enterprise in MS.

Demo 2 · Disease Course and Disease-Modifying Therapy

Relapsing-remitting MS steps up disability with incomplete recovery from each relapse, then converts to a steadily climbing secondary-progressive course. Therapy acts mainly on the relapsing, inflammatory component. Toggle therapy and compare the twenty-year trajectory on the Expanded Disability Status Scale.

progressive phase100EDSSyears from onset
Untreated, residual disability accumulates from each incompletely remitting relapse and then climbs steadily in the progressive phase, reaching EDSS 5.6 at 20 years.

The clinical course is tracked against this scale over years. In the relapsing-remitting phase, disability steps up with incomplete recovery from relapses; in the secondary-progressive phase, it climbs steadily between or in the absence of relapses. Disease-modifying therapies act chiefly on the inflammatory, relapsing component: high-efficacy agents such as the B-cell-depleting monoclonal antibody ocrelizumab reduce relapse rates and new lesion formation substantially relative to older interferon treatments (Hauser et al., 2017). Their more limited effect on the progressive, neurodegenerative component is a central reason MS remains disabling despite effective relapse control.

Cognitive Impairment in Multiple Sclerosis

Cognitive impairment is now recognised as a core feature of MS rather than an incidental complication, present in an estimated 40 to 65 percent of people with the disease across its course (Chiaravalloti & DeLuca, 2008). The foundational survey establishing its frequency and pattern found that impairment was common, only weakly related to physical disability, and selective rather than global — sparing some functions while reliably degrading others (Rao et al., 1991).

The signature deficit is a slowing of *information-processing speed*: the rate at which mental operations are executed. On timed tasks such as the Symbol Digit Modalities Test, people with MS complete fewer items per interval, and this slowing is both the earliest cognitive change and the strongest single discriminator between impaired and unimpaired patients (Benedict et al., 2020). Because processing speed is upstream of so much else, its slowing masquerades as impairment in other domains: a person given unlimited time may recall or reason normally, but under the ordinary time pressure of daily life appears broadly impaired.

Demo 3 · Timed Speed Falls While Untimed Accuracy Holds

The Symbol Digit Modalities Test measures how many symbol-to-digit substitutions can be made in 90 seconds — a pure test of processing speed. Untimed recognition memory taps stored knowledge instead. Raise the white-matter lesion load and compare how the two measures respond.

Timed SDMT (speed)52Untimed recognition15/15SDMT z-score vs age-matched norm: -0.3
Timed score 52 (z = -0.3); untimed recognition 15/15. As lesion load rises, the timed measure falls long before untimed accuracy does — why processing speed is the core screening measure.

Beyond speed, the reliably affected domains are episodic memory, complex attention and working memory, and aspects of executive function; language and general intellect are relatively spared (Sumowski et al., 2018). The memory deficit itself has a revealing structure: much of it reflects impaired *acquisition* — getting information into memory efficiently — rather than accelerated forgetting, so that once material is learned to criterion, retention and recognition are comparatively preserved. This is why recognition testing so often outperforms free recall in MS, and why the profile points to a retrieval-and-encoding-efficiency problem rather than a storage lesion of the kind seen in amnesic syndromes.

Worked Example

Consider a hypothetical assessment illustrating why processing speed dominates the MS cognitive profile. A 38-year-old with relapsing-remitting MS and a low EDSS score of 2.0 — minimal physical disability — is referred for cognitive complaints. On the Symbol Digit Modalities Test, a task requiring the examinee to pair abstract symbols with digits as quickly as possible for 90 seconds, they complete 38 correct substitutions. The normative mean for their age and education is 55, with a standard deviation of 10.

The z-score is (38 − 55) / 10 = −1.7, placing performance at roughly the 4th percentile — clearly impaired. Now the same person is given an untimed list-learning task and, across repeated trials, eventually learns all the words; on a later recognition test they correctly identify 14 of 15 items. Their *acquisition* was slow, requiring more trials than a healthy peer, but their *retention* and *recognition* are near-ceiling.

The dissociation is the point. A clinician who screened only with the EDSS would record almost no disability. A clinician who tested only untimed recognition memory would also miss the deficit. It is the *timed* measure that exposes the impairment, because the underlying pathology — slowed and desynchronised conduction across demyelinated tracts — taxes the speed of every mental operation while leaving the knowledge stores themselves comparatively intact. The worked numbers show why processing-speed measures are the recommended core of MS cognitive screening (Sumowski et al., 2018).

Discussion

The cognitive profile of MS makes a general point about the architecture of cognition. The deficits do not respect the folk taxonomy of separate faculties — memory here, attention there — but instead track a single physical variable: the integrity of the white-matter tracts that connect processing regions. When those tracts are demyelinated, the cost is paid in *timing*. Signals arrive late and desynchronised, the coincidence-detection on which cortical computation depends is degraded, and the behavioural expression is a global slowing that only looks domain-specific because the most time-pressured tasks reveal it first (Chiaravalloti & DeLuca, 2008).

This disconnection account also explains the weak correlation between physical and cognitive disability that has been observed since the earliest systematic studies (Rao et al., 1991). Motor disability reflects lesion burden in specific motor and cerebellar pathways; cognitive slowing reflects diffuse tract damage and cortical atrophy, a partly independent process. Two people with identical EDSS scores can have entirely different cognitive status, which is why the field has increasingly treated cognition as a distinct outcome requiring its own measurement rather than a correlate of the neurological examination (Benedict et al., 2020).

Cognitive Implications

Multiple sclerosis earns a place in cognitive psychology because it is an unusually clean natural experiment in the cognitive cost of *disconnection*. Where classical neuropsychology built its models on focal lesions that removed a circumscribed function — a stroke that abolished naming, a lesion that erased new learning — MS instead degrades the wiring between intact modules, and the resulting profile tests theories that treat cognition as a set of communicating processors rather than a set of storage boxes. The disproportionate slowing of processing speed, and its status as the deficit upstream of apparent impairment elsewhere, is direct behavioural evidence for the claim that conduction timing is a limiting resource for higher cognition (Chiaravalloti & DeLuca, 2008; Sumowski et al., 2018).

The disease also foregrounds the interaction of cognition with affect and effort. Depression is markedly more common in MS than in the general population and than in comparably disabling conditions, and it independently depresses performance on the very speed and memory measures used to assess MS cognition, so that a full account must separate the contribution of demyelination from that of mood and fatigue (Feinstein et al., 2014). This entanglement — a physical disease of white matter, a mood disorder, and a fatigue state all bearing on the same cognitive measures — is itself instructive for cognitive psychology, because it shows how a single test score is jointly determined by structural, motivational, and emotional causes that the science must learn to dissociate.

Current Directions

Contemporary MS research is reorienting cognition from a late-recognised complication to a primary target of assessment and therapy. A central priority identified for the field is the standardisation and routine clinical deployment of brief cognitive screening — processing-speed measures foremost — so that cognitive decline is detected as early and monitored as systematically as physical relapse (Sumowski et al., 2018). A parallel effort seeks the neural substrate of cognitive change in quantitative imaging: grey-matter and thalamic atrophy, lesion location within strategic tracts, and disruption of functional networks are being tested as biomarkers that predict and track cognitive impairment more sensitively than lesion count alone (Benedict et al., 2020).

On the therapeutic side, the arrival of high-efficacy disease-modifying treatments has sharpened a question that older, weaker therapies could not address: whether suppressing inflammatory activity early enough also protects cognition over the long term (Hauser et al., 2017). Alongside pharmacology, cognitive rehabilitation and the concept of cognitive reserve — the finding that greater premorbid intellectual enrichment buffers the cognitive expression of a given lesion load — are active lines of work aimed at modifying the cognitive course rather than only the physical one (Benedict et al., 2020).

Common Misconceptions

Multiple sclerosis inevitably leads to a wheelchair.
The disabling trajectory that shaped the disease's reputation reflects its natural history before effective treatment. Modern high-efficacy disease-modifying therapies substantially reduce relapse rates and new lesion formation, and long-term outcomes have improved accordingly; most people with MS do not become wheelchair-dependent (Hauser et al., 2017). The belief persists because the pre-therapeutic prognosis was genuinely severe.
Cognitive problems in MS mean the person is developing dementia.
The cognitive impairment of MS is selective — chiefly slowed processing speed and inefficient memory acquisition — and it dissociates from the global, progressive loss of a dementia. General intellect and language are relatively spared, and much of the memory difficulty reflects slowed learning rather than accelerated forgetting (Sumowski et al., 2018). The misreading arises because any memory complaint is popularly equated with dementia.
Physical disability and cognitive impairment go together in MS.
The two are only weakly correlated: a person with minimal physical disability can have marked cognitive slowing, and vice versa, because motor and cognitive signs reflect lesion burden in partly independent pathways (Rao et al., 1991). This is precisely why the walking-weighted EDSS misses cognitive decline and why cognition must be assessed on its own.

Glossary

Axonal transection.
The physical severing of a nerve fibre within an MS lesion, an irreversible change that converts reversible conduction slowing into permanent disconnection and drives progressive disability.
Cerebral cortex.
The grey-matter mantle of the brain; its atrophy in MS, alongside white-matter lesions, is a strong correlate of cognitive impairment.
Cognitive reserve.
The capacity of premorbid intellectual enrichment to buffer the cognitive expression of a given amount of brain damage, so that equivalent lesion loads produce unequal impairment.
Conduction block.
The complete failure of an impulse to propagate past a demyelinated axon segment, as opposed to mere slowing; a basis of relapse symptoms.
Demyelination.
The loss of the myelin sheath from an axon, the defining pathological process of MS, which degrades or blocks saltatory conduction.
Dissemination in space and time.
The diagnostic hallmark of MS: lesions in more than one central location and arising on more than one occasion, formalised in the McDonald criteria.
Expanded Disability Status Scale (EDSS).
A 0-to-10 ordinal rating of neurological impairment in MS, heavily weighted toward walking ability and comparatively insensitive to cognitive change.
Information-processing speed.
The rate at which elementary mental operations are executed; its slowing is the earliest and most characteristic cognitive deficit in MS.
Myelin.
The lipid-rich insulating sheath, formed by oligodendrocytes in the central nervous system, that enables rapid saltatory conduction and is the target of immune attack in MS.
Node of Ranvier.
A periodic gap in the myelin sheath, rich in voltage-gated sodium channels, at which the action potential is regenerated during saltatory conduction.
Optic neuritis.
Inflammatory demyelination of the optic nerve causing visual loss and pain on eye movement, a common presenting feature of MS.
Plaque.
A focal, sharply demarcated area of demyelination and gliosis in the central nervous system; the lesion from which multiple sclerosis takes its name.
Relapsing-remitting course.
The most common MS phenotype, in which discrete relapses of symptoms are followed by partial or complete recovery and clinical stability between attacks.
Saltatory conduction.
The rapid, energy-efficient propagation of a nerve impulse by leaping between nodes of Ranvier along a myelinated axon.
Symbol Digit Modalities Test (SDMT).
A brief timed task requiring rapid substitution of digits for abstract symbols; the recommended core measure of processing speed in MS cognitive screening.

Key Researchers

Ralph H. B. Benedict

Professor at the University at Buffalo, State University of New York; developed the Brief International Cognitive Assessment for MS (BICAMS) and led the 2020 review consolidating cognitive impairment as a core clinical target.
ORCID - Faculty Page

Jean-Martin Charcot

Neurologist at the Salpêtrière Hospital, Paris; gave the first systematic clinico-pathological account of multiple sclerosis in the 1860s, defining the demyelinating plaque and its clinical triad.
Wikipedia - Wikidata

Nancy D. Chiaravalloti

Director of Neuropsychology and Neuroscience Research at the Kessler Foundation; lead author of the 2008 review of cognitive impairment in MS and developer of memory-rehabilitation methods for the disease.
ORCID - Faculty Page

Alastair Compston

Emeritus Professor of Neurology at the University of Cambridge; lead author of the 2008 Lancet seminar on MS and an authority on its genetics and immunotherapy.
Wikipedia - Wikidata - Faculty Page

John DeLuca

Senior Vice President for Research at the Kessler Foundation; a leading authority on processing-speed and memory deficits in MS and the disconnection account of its cognitive profile.
ORCID - Faculty Page

Massimo Filippi

Director of Neurology and the Neuroimaging Research Unit at Vita-Salute San Raffaele University, Milan; lead author of the 2018 Nature Reviews Disease Primers primer and an authority on MS MRI biomarkers.
Faculty Page

Stephen L. Hauser

Director of the UCSF Weill Institute for Neurosciences; led the OPERA trials that established B-cell depletion as an MS therapy and reshaped understanding of the disease's immunopathology.
ORCID - Wikipedia - Wikidata

W. Ian McDonald

Neurologist at the Institute of Neurology, Queen Square, London; chaired the panel that produced the McDonald diagnostic criteria integrating MRI evidence of dissemination in space and time.
Wikipedia

Alan J. Thompson

Professor of Neurology at University College London; chaired the panel behind the 2017 revisions of the McDonald criteria and the 2013 redefinition of MS clinical courses.
ORCID - Wikipedia - Wikidata

Frequently Asked Questions

What causes multiple sclerosis?

Multiple sclerosis is an immune-mediated disease in which the immune system attacks the myelin of central axons; its cause is understood as an interaction of genetic susceptibility with environmental factors, and its lesions are disseminated in space and time throughout the central nervous system (Reich et al., 2018).

How does multiple sclerosis affect cognition?

Roughly 40 to 65 percent of people with MS develop cognitive impairment, most characteristically a slowing of information-processing speed together with deficits in memory acquisition, complex attention, and executive function, while language and general intellect are relatively spared (Chiaravalloti & DeLuca, 2008).

Why is processing speed the main cognitive deficit in MS?

Because MS lesions fall in the white-matter tracts that connect processing regions, they slow and desynchronise conduction, and this degrades the timing on which mental operations depend; the result is a global slowing that appears first on timed tasks (Sumowski et al., 2018).

Does physical disability predict cognitive impairment in MS?

Only weakly: motor disability and cognitive slowing reflect lesion burden in partly independent pathways, so a person with minimal physical disability can have marked cognitive impairment, which is why cognition is assessed separately (Rao et al., 1991).

How is multiple sclerosis diagnosed?

Diagnosis rests on demonstrating dissemination in space and time (lesions in more than one central location and arising on more than one occasion) using clinical and MRI evidence codified in the McDonald criteria and their 2017 revision (Thompson et al., 2018).

Can multiple sclerosis be treated?

Disease-modifying therapies reduce the frequency of relapses and the formation of new lesions; high-efficacy agents such as ocrelizumab produce substantially lower relapse rates than older interferon treatments, though their effect on progressive neurodegeneration is more limited (Hauser et al., 2017).

Is the memory problem in MS a storage problem?

Largely no: much of the memory difficulty reflects slowed and inefficient acquisition rather than accelerated forgetting, so once material is learned, retention and recognition are comparatively preserved, a pattern unlike the storage loss of amnesic syndromes (Benedict et al., 2020).

How common is depression in multiple sclerosis, and does it affect cognition?

Depression is markedly more common in MS than in the general population and independently lowers performance on the speed and memory measures used to assess MS cognition, so mood and fatigue must be distinguished from demyelination when interpreting a cognitive score (Feinstein et al., 2014).

Support Organizations

Organizations that provide information, assessment guidance, and advocacy for people with multiple sclerosis and related conditions.

National Multiple Sclerosis Society — Information, research funding, and support services for people affected by MS. (United States)

MS Society UK — Advocacy, funding, and local support for people living with MS across the United Kingdom. (United Kingdom)

Multiple Sclerosis International Federation — Global federation of national MS organizations; publisher of the Atlas of MS. (International)

References

Benedict, R. H. B., Amato, M. P., DeLuca, J., & Geurts, J. J. G. (2020). Cognitive impairment in multiple sclerosis: Clinical management, MRI, and therapeutic avenues. The Lancet Neurology, 19(10), 860-871. https://doi.org/10.1016/S1474-4422(20)30277-5

Chiaravalloti, N. D., & DeLuca, J. (2008). Cognitive impairment in multiple sclerosis. The Lancet Neurology, 7(12), 1139-1151. https://doi.org/10.1016/S1474-4422(08)70259-X

Compston, A., & Coles, A. (2008). Multiple sclerosis. The Lancet, 372(9648), 1502-1517. https://doi.org/10.1016/S0140-6736(08)61620-7

Feinstein, A., Magalhaes, S., Richard, J. F., Audet, B., & Moore, C. (2014). The link between multiple sclerosis and depression. Nature Reviews Neurology, 10(9), 507-517. https://doi.org/10.1038/nrneurol.2014.139

Filippi, M., Bar-Or, A., Piehl, F., Preziosa, P., Solari, A., Vukusic, S., & Rocca, M. A. (2018). Multiple sclerosis. Nature Reviews Disease Primers, 4, 43. https://doi.org/10.1038/s41572-018-0041-4

Hauser, S. L., Bar-Or, A., Comi, G., Giovannoni, G., Hartung, H. P., Hemmer, B., ... & Kappos, L. (2017). Ocrelizumab versus interferon beta-1a in relapsing multiple sclerosis. New England Journal of Medicine, 376(3), 221-234. https://doi.org/10.1056/NEJMoa1601277

Kurtzke, J. F. (1983). Rating neurologic impairment in multiple sclerosis: An expanded disability status scale (EDSS). Neurology, 33(11), 1444-1452. https://doi.org/10.1212/WNL.33.11.1444

Lublin, F. D., Reingold, S. C., Cohen, J. A., Cutter, G. R., Sorensen, P. S., Thompson, A. J., ... & Polman, C. H. (2014). Defining the clinical course of multiple sclerosis: The 2013 revisions. Neurology, 83(3), 278-286. https://doi.org/10.1212/WNL.0000000000000560

Rao, S. M., Leo, G. J., Bernardin, L., & Unverzagt, F. (1991). Cognitive dysfunction in multiple sclerosis. I. Frequency, patterns, and prediction. Neurology, 41(5), 685-691. https://doi.org/10.1212/wnl.41.5.685

Reich, D. S., Lucchinetti, C. F., & Calabresi, P. A. (2018). Multiple sclerosis. New England Journal of Medicine, 378(2), 169-180. https://doi.org/10.1056/NEJMra1401483

Sumowski, J. F., Benedict, R., Enzinger, C., Filippi, M., Geurts, J. J., Hamalainen, P., ... & Rao, S. (2018). Cognition in multiple sclerosis: State of the field and priorities for the future. Neurology, 90(6), 278-288. https://doi.org/10.1212/WNL.0000000000004977

Thompson, A. J., Banwell, B. L., Barkhof, F., Carroll, W. M., Coetzee, T., Comi, G., ... & Cohen, J. A. (2018). Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria. The Lancet Neurology, 17(2), 162-173. https://doi.org/10.1016/S1474-4422(17)30470-2

Trapp, B. D., Peterson, J., Ransohoff, R. M., Rudick, R., Mörk, S., & Bö, L. (1998). Axonal transection in the lesions of multiple sclerosis. New England Journal of Medicine, 338(5), 278-285. https://doi.org/10.1056/NEJM199801293380502

Walton, C., King, R., Rechtman, L., Kaye, W., Leray, E., Marrie, R. A., ... & Baneke, P. (2020). Rising prevalence of multiple sclerosis worldwide: Insights from the Atlas of MS, third edition. Multiple Sclerosis Journal, 26(14), 1816-1821. https://doi.org/10.1177/1352458520970841