Brain Shrinkage and Disability in Multiple Sclerosis: What 58 Studies Reveal About the Link

Table of Contents

• Key Points
• Why This Research Matters: MS and Brain Atrophy
• Understanding the EDSS Disability Scale
• How Brain Atrophy Is Measured on MRI Scans
• What This Review Set Out to Do
• Study Methods: How the Research Was Conducted
• Key Findings at a Glance
• Longitudinal Studies: Brain Volume Change Over Time and Disability
• Cross-Sectional Studies: Brain Volume at One Time Point and Current Disability
• Does One Baseline Brain Volume Measurement Predict Future Disability?
• What the "About 1% Per Year" Finding Means
• Clinical Implications: What This Means for People with MS
• Study Limitations: What This Review Could Not Prove
• Recommendations for Patients
• Frequently Asked Questions
• Source Information

Key Points

• 19 of 23 longitudinal studies linked yearly brain shrinkage to worsening disability scores.
• 26 of 29 cross-sectional studies linked smaller brain volume to higher current disability.
• Only 4 of 15 follow-up studies found a single baseline brain-volume scan predicted future disability.
• A yearly loss of about 0.4%–1.3% of total brain volume was associated with disability progression.
• People with MS typically lose 0.5%–1.35% brain volume yearly, 2–10 times the normal aging rate.

Why This Research Matters: MS and Brain Atrophy

Multiple sclerosis (MS) is a widespread neuroinflammatory disease (a condition driven by inflammation in the brain and spinal cord) that also involves progressive neurodegeneration (gradual damage and loss of nerve cells). Typically, MS begins at a young age.

Symptoms of MS can include:

• Altered sensation, such as numbness, tingling, itching, or "pins and needles"
• Impaired coordination and balance
• Weakness
• Muscle spasms or cramps
• Fatigue
• Pain
• Visual disturbances

Brain atrophy (BA) — the medical term for brain shrinkage or loss of brain tissue — was first observed in people with MS (often shortened to PwMS) in the early 1960s. By the mid-1990s, magnetic resonance imaging (MRI) was being used to measure brain atrophy and its connection to physical disability. Those early studies used two-dimensional scans. More recent studies use three-dimensional scans analyzed with specialized software called SIENA (Structural Image Evaluation Using Normalization of Atrophy).

The numbers matter here. Research suggests that people with MS lose about 0.5% to 1.35% of their brain volume (BV) per year. For comparison, an average person without MS loses only about 0.1% to 0.3% of brain volume per year due to normal aging. That means MS-related brain loss is roughly two to ten times faster than the normal aging rate.

Why does this matter for patients? Brain atrophy measurements are thought to represent the net effect of all the degenerative processes happening in MS — not just the visible inflammation or lesions (damaged areas) seen on standard MRI scans. In fact, brain atrophy appears to be a better predictor of clinical outcomes than other MRI measures. For example:

• Brain atrophy seems to predict physical disability better than conventional lesion measures do.
• People with higher levels of brain atrophy are more likely to progress from clinically isolated syndrome (CIS) — a first episode of MS-like symptoms — to full MS.
• Brain atrophy has been linked to increased disability, worsening cognition (thinking and memory), poorer quality of life, more fatigue, and poorer economic outcomes.

Several earlier reviews (published between 2009 and 2016) looked at the link between brain atrophy and cognition, but only three considered physical disability. This matters because an expert consensus group called MAGNIMS (Magnetic Resonance Imaging in Multiple Sclerosis) recently recommended that global brain volume loss (BVL) be used to help define and predict MS severity, and that it be included as a secondary outcome (a secondary measure of success) in therapeutic clinical trials.

Understanding the EDSS Disability Scale

To study disability, researchers need a way to measure it. For the last four decades, the most widely used tool has been the Expanded Disability Status Scale (EDSS). It is an ordinal rating system (a scale where scores are ranked in order) ranging from 0 to 10.

The EDSS has a well-known weakness, though. A frequent criticism is that the distance between points on the scale is not equal. A change from 1 to 2 does not represent the same amount of disability change as a change from 6 to 7. Here is how the scale breaks down:

• Scores 0–4 measure neurological impairments (problems with nerve function)
• Scores 4–6 indicate impacts on walking ability
• Scores 6–7 indicate loss of walking ability beyond a few metres
• Scores 7–7.5 indicate wheelchair dependency

Another limitation: the EDSS correlates only weakly with neuropsychological impairment (cognitive problems) and with patient-reported outcomes (how patients themselves feel and function in daily life). Despite these criticisms, the EDSS remains the standard measure of disability progression (DP) in MS research.

How Brain Atrophy Is Measured on MRI Scans

There are several technical approaches to measuring global (whole-brain) atrophy with MRI, and each has challenges.

The longitudinal approach (tracking over time): This method analyzes two or more MRI scans taken of the same person at different time points. The scans are spatially matched (aligned to each other) and subtracted (compared pixel by pixel). The brain surface is automatically modelled and registered (aligned) to the outer skull so that serial images can be compared. The result is expressed as percentage of brain volume change (PBVC), which reflects changes in the brain surface relative to the skull, normalized to skull size. SIENA is the most frequently used software for this approach.

The cross-sectional approach (single scan): Brain atrophy can also be estimated from a single MRI scan. The image is registered to a standard space (a reference brain template), normalized to the skull surface, and the tissue is segmented (separated) into brain and non-brain regions. The most frequently used software here is SIENAX (Structural Image Evaluation Using Normalization of Atrophy-Cross-Sectional). This produces a normalized brain volume (NBV) measurement, which can then be compared with EDSS scores or compared between groups (for example, people with MS who did or did not show disability progression).

The proportion-based approach: Another method estimates the proportion of the intracranial volume (the space inside the skull) that is occupied by brain tissue. This is called the brain parenchymal fraction or normalized brain parenchymal volume (NBPV).

Throughout this review, researchers measured brain volume loss as PBVL (percentage of brain volume loss), or used NBV or NBPV as their brain-volume measures.

What This Review Set Out to Do

The authors of this review had three main goals. Their primary objective was to compare longitudinal studies measuring the association between change in whole brain volume over time and change in EDSS score or disability progression measured by EDSS.

Their secondary aims were:

1. To compare cross-sectional studies measuring the association between normalized whole brain volume (or fraction) and EDSS score at the same point in time.
2. To compare longitudinal studies measuring whether baseline brain volume (the volume measured at the start of the study) predicts future EDSS score at follow-up.

Why now? Because brain atrophy is increasingly used as an outcome in clinical trials (per the MAGNIMS recommendation), researchers need a clearer understanding of the relationship between brain atrophy and physical disability. An earlier review covered studies published before 2013. This review builds on that work by examining studies published from January 2013 onward.

Study Methods: How the Research Was Conducted

This was a systematic review, meaning the researchers followed a strict, pre-planned protocol to find, screen, and analyze all relevant studies. The protocol was designed according to the PRISMA-P 2015 statement (Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols) and was finalized in January 2022, before the final literature search was performed.

What studies were included? The researchers included studies that met all of these criteria:

• People aged 18 or over, diagnosed with clinically isolated syndrome (CIS) or MS
• Longitudinal observational studies, clinical trials, and modelling studies measuring the link between change in whole brain volume and EDSS score or disability progression (confirmed or non-confirmed)
• Cross-sectional studies measuring the link between normalized brain volume (NBV) or fraction (NBPV) and EDSS score
• Original articles written in English, published in a peer-reviewed journal between 1 January 2013 and 3 February 2022

What was excluded? Studies were excluded if:

• 20% or more of the study population had primary progressive MS (PPMS). This was a pragmatic decision to keep the review focused on CIS and relapsing-remitting MS (RRMS), because brain volume loss proceeds faster in the earlier stages of MS, and disease-modifying therapies (DMTs) — medications that slow MS activity — are more effective at that time.
• The study measured only spinal cord atrophy or regional (localized) atrophy, rather than global (whole-brain) atrophy.
• EDSS was reported only as part of a composite "No Evidence of Disease Progression" outcome. NEDA-3 is defined as no new MRI lesions, no new relapses, and no change in EDSS over 1 year. NEDA-4 adds one more requirement: no brain volume loss of 0.4% or more per year.
• The association was not reported separately for people with MS versus healthy controls.

Which databases were searched? The team searched six major medical databases: Medline, Embase, the Cochrane Library, the Cochrane Central Register of Controlled Trials (CENTRAL), the Cochrane Database of Systematic Reviews (CDSR), and the Cumulative Index to Nursing and Allied Health Literature (CINAHL).

How was screening done? Duplicates were removed using PROQUEST and manual checks in Excel. Two researchers independently screened titles and abstracts using a standard data collection form. Discrepancies were resolved by discussion. Full-text screening was done the same way, with a third reviewer available if needed (though not required). Data were extracted into a standardized spreadsheet by one researcher and validated by a second.

How were studies grouped? Studies were first grouped by design (longitudinal or cross-sectional). Longitudinal studies were then grouped by how they defined disability increase: as a change in EDSS score, or as disability progression. Studies that used disability progression were further grouped by their specific definition of progression, given the range of baseline EDSS scores reported.

The screening process can be summarized in numbers. A total of 3,114 abstracts were screened, yielding 715 papers and abstracts for more detailed consideration. After limiting to full-text articles from 2013 onward, 320 articles were screened in full. (164 articles published before 2013 and 231 conference abstracts from 2016 onward were set aside.) Of those 320 articles, 197 measured brain atrophy, and 131 of those also focused on disability outcomes. After applying all exclusion criteria, 58 studies were included in the final review.

Key Findings at a Glance

Here is the big picture before we dive into details. The review found:

• 19 of 23 longitudinal studies (about 83%) observed a significant association between global brain volume loss over time and change in EDSS score or disability progression.
• 26 of 29 cross-sectional studies (about 90%) observed an association between baseline brain volume measures and current EDSS score.
• Only 4 of 15 follow-up studies (about 27%) found that baseline brain volume predicted future EDSS score or disability progression.

In plain terms: tracking brain volume changes over time in the same person is strongly linked to disability. A single brain-volume snapshot, however, is a weak predictor of what will happen to that person in the future.

Longitudinal Studies: Brain Volume Change Over Time and Disability

The 23 longitudinal studies followed people with MS over time, measuring both brain volume loss and disability. These studies had a median study size of 180 people with MS (ranging from just 16 to 3,635 participants), a median follow-up time of 4 years (range 1–12 years), and participants with a median age of 37 years (range 29–48).

Key characteristics of the participants in these studies:

• 70% were female at baseline (range 54%–80%)
• 0% had primary progressive MS at baseline (range 0%–17%)
• Mean EDSS score at baseline was 2.0 (range 1.3–6)
• Mean disease duration at baseline was 7.5 years (range 1–14 years)

All of the longitudinal studies measured brain atrophy as PBVL (percentage of brain volume loss), although one study categorized PBVL for its analysis rather than using it as a continuous number.

Brain volume loss and change in EDSS score: Eight studies investigated the association between PBVL and change in EDSS score. Six of those eight (75%) observed a significant association (p < 0.05, meaning there is less than a 5% chance the result was due to random luck). Two small studies were inconclusive: one found a non-significant association in 38 people, and the other observed no change in EDSS over the study period in just 16 people.

Brain volume loss and disability progression: Fifteen studies investigated the association between PBVL and disability progression, defined by changes in EDSS score. Thirteen of those 15 studies (about 87%) reported a significant association between PBVL and disability progression.

The two that did not find an association included one study of 62 people with a non-significant result and one study of 82 people that found no difference in normalized brain volume. One important caveat: across those 15 studies, researchers used six different definitions of disability progression. Some definitions required a confirmed progression (sustained over a period of months), while others did not. This variation makes direct comparisons between studies difficult.

How were the studies designed? Most longitudinal studies (20 of 23) were retrospective analyses, meaning they used data that had already been collected, often from existing clinical trial cohorts and clinical studies. Three studies were prospective (designed in advance to follow participants forward in time). Of those three prospective studies, two observed a significant association between PBVL and disability progression. The third found a significant association between PBVL and EDSS when using Icobrain software, but not when using SIENA software — a reminder that the choice of analysis software can affect results.

What software was used? All but one of the 23 longitudinal studies used SIENA software to measure brain volume change; one used FreeSurfer. Two studies used additional software alongside SIENA (Icobrain and SPM12). Eighteen studies used regression or survival models (statistical methods that examine relationships while accounting for other factors) adjusted for combinations of sex, age, education, disease phenotype, disease duration, study cohort, lesion volume, MRI field strength, and other baseline variables.

What about the study populations? The majority of participants were treated with disease-modifying therapies (DMTs). Just one study included mostly untreated patients. Many studies also reported regional (localized) brain atrophy measures — 14 of 23 studies (61%) — and for many of those, regional atrophy was actually the main focus. A few studies (4 of 23, about 17%) reported additional physical disability measures, specifically the 9-Hole Peg Test (9HPT), which measures hand dexterity, and the Timed 25-Foot Walk (T25FW), which measures walking speed.

How fast was the brain volume loss? Most studies did not report an annualized (per-year) PBVL, instead reporting the total PBVL over the whole observation period. But for the 5 studies that did report annualized PBVL, the average yearly loss ranged from −0.48% ± 0.93 to −0.9% ± 1.0 (minus sign indicating loss), with observation periods ranging from 2 to 7.5 years.

Cross-Sectional Studies: Brain Volume at One Time Point and Current Disability

Cross-sectional studies take a "snapshot" — they measure brain volume and disability at the same moment and ask whether people with smaller brains tend to have worse disability. These studies compared baseline brain volume measures (NBV, NBPV, or whole brain volume) with baseline EDSS scores.

The 29 cross-sectional studies had a median size of 80 people with MS (range 20–3,635), a median participant age of 41 years (range 31–61), and the following participant profile:

• 68% female at baseline (range 57%–80%)
• 0% with primary progressive MS at baseline (range 0%–17%)
• Mean EDSS score of 2.5 (range 1.2–6.5)
• Mean disease duration of 9.5 years (range 0.5–27 years)

The result was clear-cut: 26 of the 29 studies (about 90%) observed a significant association between brain volume measured at baseline and EDSS score measured at the same time. In other words, people with MS who had more brain shrinkage also tended to have more physical disability. Note that a cross-sectional association cannot by itself prove that brain atrophy causes disability — it only shows the two are linked at a single point in time.

Does One Baseline Brain Volume Measurement Predict Future Disability?

This is a critical question for patients. If a single MRI scan early in the disease could predict who will become disabled later, doctors could target treatments more aggressively. The evidence from this review says: no, not reliably.

Fifteen studies asked this specific question — whether baseline brain volume measures predict follow-up EDSS score or future disability progression. These studies had a median size of 102 people with MS (range 23–1,214), a median follow-up time of 5 years (range 1–30 years), and participants with a median age of 40 (range 31–61).

The participant profile for these follow-up studies was similar to the cross-sectional group:

• 68% female at baseline (range 59%–80%)
• 0% with PPMS at baseline (range 0%–14%)
• Mean EDSS score of 2.5 (range 1.2–5)
• Mean disease duration of 8.5 years (range 4–30 years)

Only 4 of these 15 studies (about 27%) observed that baseline brain volume predicted future EDSS. The other 11 studies did not. This makes biological sense: because normal brain volumes vary widely from person to person, a single measurement cannot easily separate the effect of disease-related shrinkage from a person's natural starting brain size. Measuring change within the same person over time removes that problem, which is exactly why the longitudinal results were much stronger.

What the "About 1% Per Year" Finding Means

The review's headline conclusion: a decrease in global brain volume of about 1% per year (range 0.4%–1.3%) was associated with disability progression in people with MS.

To put that in perspective, remember that normal aging causes only about 0.1%–0.3% brain volume loss per year in people without MS. So when a person with MS loses brain volume at roughly 1% per year — several times the normal aging rate — that pace of loss was consistently linked, across many studies, to measurable worsening of disability on the EDSS scale.

But the researchers add an important caution: direct comparisons between studies should be made carefully. The six different definitions of disability progression used across the studies mean that one study's "progression" may not match another's. Confirmed progression (sustained disability worsening over time) and non-confirmed progression are not equivalent outcomes.

Clinical Implications: What This Means for People with MS

These findings have several practical implications for how MS is monitored and how new treatments are tested.

Brain volume loss is a valid marker of disease worsening. The strong and consistent association between brain volume loss over time and disability progression supports the MAGNIMS recommendation to include global brain volume loss as a secondary outcome in therapeutic clinical trials. When a treatment slows brain volume loss, that is meaningful evidence that it may also slow disability.

Change over time matters more than any single scan. The striking contrast — 83% of longitudinal studies finding an association versus only 27% of single-scan prediction studies — carries a clear message. For individual patients, a single brain volume measurement on one MRI is not a reliable crystal ball. What matters is the trajectory: how much brain volume is lost between one scan and the next, measured using the same method.

The roughly 1%-per-year figure gives doctors and researchers a working benchmark. When brain volume loss reaches about 0.4%–1.3% per year, that range was associated with disability progression in

Frequently Asked Questions

What is brain atrophy in multiple sclerosis?

Brain atrophy means brain shrinkage or loss of brain tissue. It was first observed in people with MS in the early 1960s. It is thought to represent the net effect of all degenerative processes in MS, not just visible inflammation or lesions, and faster brain-volume loss has been linked to greater disability.

How fast does brain volume shrink in MS compared with normal aging?

Research suggests people with MS lose about 0.5% to 1.35% of brain volume per year. For comparison, a person without MS loses only about 0.1% to 0.3% per year from normal aging. That means MS-related brain loss is roughly two to ten times faster than the normal aging rate.

Can a single MRI brain-volume measurement predict future disability?

No, not reliably. Only 4 of 15 follow-up studies (about 27%) found that baseline brain volume predicted later disability. A single snapshot cannot easily separate disease-related shrinkage from a person's natural starting brain size. Measuring change over time in the same person is much stronger and was linked to disability in about 83% of longitudinal studies.

What is the EDSS disability scale used in MS?

The Expanded Disability Status Scale, or EDSS, ranges from 0 to 10. Scores 0–4 measure neurological impairments, 4–6 reflect walking ability, 6–7 indicate loss of walking beyond a few metres, and 7–7.5 mean wheelchair dependency. It remains the standard measure of disability progression in MS research, although points on the scale are not equal distances apart.

How is brain atrophy measured on MRI scans?

There are two main methods. The longitudinal method compares two or more scans of the same person over time, often using software called SIENA, and reports percentage brain volume change. The cross-sectional method estimates brain volume from a single scan using software like SIENAX, producing a normalized brain volume that can be compared with disability scores at that time.

When should I seek a second opinion about brain shrinkage on my MS MRI and disability progression?

A single MRI brain-volume snapshot does not reliably predict future disability; only about 27% of studies found baseline brain volume predicted later EDSS, whereas 83% of longitudinal studies linked yearly volume loss to worsening disability. If your doctor recommends a treatment change or prognosis based on one scan, or if you are unsure whether the roughly 1% per year loss (range 0.4%–1.3%) applies to your trajectory, an independent review of serial MRIs and disability scores can help. Diagnostic Detectives Network provides independent expert second opinions.