Abstract

Introduction:

Multiple sclerosis (MS) affects different cognitive domains, including social cognition. Immersive Virtual Reality (VR) may provide a novel rehabilitative approach to treat motor and cognitive symptoms of MS. This exploratory pilot study evaluated the effects of immersive VR rehabilitation on social cognition in MS patients and explored related cortical neurophysiological signatures.

Methods:

Seven MS patients underwent immersive VR rehabilitation with the CAREN system (3 sessions/week, approximately 45 min of active training per session, about 1 h including preparation, 8 weeks), while seven healthy controls (HC) did not undergo any intervention. Patients were evaluated at baseline (T0) and post-treatment (T1) with standardized measures of cognitive, emotional, and motor functioning. EEG data were acquired from all participants, and, after artifact removal, spectral parameterization decomposed signals into aperiodic (exponent, offset) and periodic oscillatory components (alpha and beta power). Power spectral density was analyzed using group comparisons and Pearson correlations with neuropsychological measures.

Results:

Compared with HC, MS patients showed reduced alpha-band power, mainly over frontal and parieto-occipital regions, whereas aperiodic parameters did not differ between groups. In patients, alpha and beta power correlated with the Positive Emotions Self-Efficacy Scale (alpha: r = 0.92, p = 0.003; beta: r = 0.83, p = 0.020). Alpha power is also correlated with RAO SRT–LTS (r = 0.85, p = 0.016), and beta with EQ-CE (r = 0.82, p = 0.023). Overall, alpha and beta power were correlated with emotional self-efficacy, balance, memory, and empathy, suggesting that oscillatory markers are potential indicators of clinical outcomes.

Discussion:

Rehabilitation via immersive VR has shown promising clinically significant effects in the cognitive, emotional, and motor domains, supported by convergent EEG spectral signatures. Future studies employing predictive modeling approaches will be required to assess their prognostic value.

1 Introduction

Social cognition (SC) refers to a set of neurocognitive processes that enable individuals to perceive, interpret, and respond to the intentions, emotions, and behaviors of others (Adolphs, 2009; Goettfried et al., 2025). It encompasses diverse interconnected domains, including emotion recognition, empathy, theory of mind (ToM), and social decision-making, all of which are essential for adaptive interpersonal functioning (Melloni et al., 2014; Mitchell and Phillips, 2015). Among these domains, empathy plays a pivotal role and is commonly divided into cognitive empathy, which involves understanding the mental states of others, and affective empathy, the ability to share in the emotional experiences of others (Shamay-Tsoory, 2011; Decety et al., 2012). These processes rely on a complex and distributed brain network, including the medial prefrontal cortex, temporoparietal junction, and other limbic-related regions such as the anterior cingulate cortex and insula (Apps et al., 2016; Arioli et al., 2018; Krendl and Betzel, 2022). Alterations in these systems have been documented in several neurological and neurodegenerative conditions, in which impairment to the fronto-limbic and temporo-parietal pathways impairs the ability to process social and emotional signals (Melloni et al., 2014; Vandekerckhove et al., 2020). Among these, multiple sclerosis (MS) is increasingly recognized as a disorder that extends beyond motor and sensory impairment, affecting higher-order cognitive and socioemotional functions (Henry et al., 2016; Benedict et al., 2020). MS represents a particularly relevant clinical model for investigating SC impairments, as disease-related structural and functional alterations frequently involve fronto-limbic and temporoparietal networks supporting emotional processing, empathy, and theory of mind. These alterations may lead to deficits in socio-emotional functioning that negatively affect interpersonal relationships and quality of life, highlighting the importance of developing targeted rehabilitation approaches for this domain. Some authors have highlighted that the demyelinating and neurodegenerative processes of MS could influence the networks involved in emotional processing and social understanding, particularly within the prefrontal-limbic circuits (Krause et al., 2009; Kalin, 2019). Consequently, individuals with MS frequently exhibit deficits in emotion recognition, reduced empathy, and impaired ToM skills (Pöttgen et al., 2013; Raimo et al., 2017), which significantly impact interpersonal relationships, psychosocial adjustment, and overall quality of life (Phillips et al., 2011).

Traditional rehabilitation approaches have focused primarily on general motor and cognitive outcomes, often neglecting the socio-cognitive and emotional domains, which are crucial determinants of functional recovery and social participation. In this context, immersive Virtual Reality (VR) has emerged as a promising rehabilitation tool capable of combining scenarios mimicking reality with precise experimental control (Tieri et al., 2018; Maggio et al., 2019; Riva et al., 2019; Maggio et al., 2024a). Immersive VR systems, such as the Computer Assisted Rehabilitation Environment (CAREN), can simulate realistic social scenarios, enhance motivation, embodied experience, and multisensory integration, while promote neural plasticity (Cellini et al., 2022; Impellizzeri et al., 2024; Maggio et al., 2024c,2024b; Wankhede et al., 2025). However, despite these advantages, VR-based rehabilitation may also present some limitations, particularly in neurological populations such as individuals with MS. Factors such as fatigue, sensory overload, or cybersickness may influence patients’ tolerance to immersive environments, highlighting the need for individualized training protocols and careful monitoring of exposure time to ensure safety and treatment effectiveness (Feng et al., 2025). This technology offers a unique opportunity to stimulate SC, engaging the same neural networks implicated in empathy and emotion regulation (Maresca et al., 2020; Krendl and Betzel, 2022; Sievers and Thornton, 2024). To better understand how these immersive experiences impact brain function, the integration of neurophysiological measures such as electroencephalography (EEG) is crucial, as highlighted by previous studies (Mishra et al., 2021; Kim, 2025). EEG offers a non-invasive approach to understanding neurophysiological mechanisms underlying such interventions. Oscillatory brain activity, particularly within the alpha and beta frequency bands, has been associated with attentional control, emotional processing, and empathic engagement (Riels et al., 2022; Codispoti et al., 2023). Understanding the relationship between EEG oscillatory dynamics and behavioral outcomes can therefore contribute to the identification of recovery biomarkers and the optimization of individualized neurorehabilitation programs (Babiloni et al., 2020; Esteves et al., 2025). Although EEG has been increasingly used to characterize neurophysiological alterations in MS, findings related to EEG power spectral density (PSD) remain underexplored and heterogeneous (Facchetti et al., 1994; Leocani et al., 2000b; Cogliati Dezza et al., 2015; Babiloni et al., 2016a). Moreover, only one study has investigated resting-state EEG modifications following a motor-based rehabilitation program in MS (Tramonti et al., 2019). In that study, no correlations were found between changes in spectral power (delta, theta, alpha, beta) and improvements in motor performance, as measured by the Timed Up and Go test.

Overall, these findings suggest that immersive VR rehabilitation, combined with EEG monitoring, could represent an innovative and effective approach to improving socio-cognitive functioning in individuals with MS.

This exploratory pilot study aimed to investigate the potential effects of immersive VR training on social cognition in patients with MS and to explore its neurophysiological correlates through EEG spectral analysis. Specifically, the study pursued two main objectives:

(1)

To assess longitudinal changes in cognitive, emotional, and neurophysiological measures (EEG spectral parameters) in MS patients, comparing evaluations conducted before (T0) and after (T1) an immersive VR rehabilitation program;

(2)

To compare the spectral features of resting-state EEG in MS patients at baseline with those of healthy controls, to characterize disease-related neurophysiological differences.

Thus, this study aims to explore the potential of immersive VR as a tool for enhancing social cognition and to identify oscillatory biomarkers associated with emotional and cognitive improvement in MS.

2 Materials and methods2.1 Study population

This pilot study was conducted between November 2023 and May 2024 at the IRCCS Centro Neurolesi “Bonino-Pulejo” in Messina, Italy. The study involved a total of 14 subjects, including 7 patients with MS and 7 HC, matched for age, sex, and education level. Neurological disability in the MS group was assessed using the Expanded Disability Status Scale (EDSS), a standardized clinical measure widely used to quantify disability in multiple sclerosis. In the present cohort, participants with MS showed a moderate level of disability (mean EDSS = 4.36 ± 0.90). Information on disease-modifying therapies was collected for all MS patients. At the time of the study, patients were receiving stable disease-modifying treatments (ocrelizumab, fingolimod, or natalizumab), as reported in Table 1.

VariableMS (n = 7)HC (n = 7)p-valueAge (mean ± SD)41.57 ± 10.20 years42.29 ± 11.18 years0.903Sex (F/M)3 / 43 / 41Disease duration (mean ± SD)15.43 ± 10.21 years–NAPharmacological treatmentOcrelizumab (n = 3)–NAFingolimod (n = 2)Natalizumab (n = 2)EDSS (T0) (mean ± SD)4.36 ± 0.90––EDSS (T1) (mean ± SD)4.36 ± 0.90––BBS (T0) (mean ± SD)47.29 ± 7.23––BBS (T1) (mean ± SD)50.00 ± 7.09––

Demographic and clinical characteristics of MS patients and HC.

EDSS-T0, Expanded Disability Status Scale, baseline; EDSS-T1, Expanded Disability Status Scale, post-intervention; BBS-T0, Berg Balance Scale, baseline; BBS-T1, Berg Balance Scale, post-intervention.

Inclusion criteria for the MS group were: (i) a confirmed diagnosis of relapsing-remitting MS according to the revised 2017 McDonald criteria (Thompson et al., 2018); (ii) stable clinical condition without relapses or corticosteroid therapy in the previous three months, and (iii) the ability to participate in the virtual reality training sessions.

Exclusion criteria included severe cognitive impairment (Mini-Mental State Examination < 10), as individuals with severe cognitive deficits may have difficulty reliably cooperating with neuropsychological assessments and EEG recordings (Tombaugh and McIntyre, 1992), severe psychiatric or neurological comorbidities, visual or hearing impairments that could interfere with VR participation, and contraindications to EEG recording.

Healthy control group participants were recruited from the local community and screened for neurological, psychiatric, or systemic conditions that could impair cognitive performance.

All participants were right-handed and had normal or corrected-to-normal vision.

2.2 Ethics

The study was conducted in accordance with the principles of the Declaration of Helsinki. All participants provided written informed consent prior to enrollment. The research forms part of a larger multi-year project coordinated by the IRCCS Centro Neurolesi “Bonino-Pulejo” and registered on ClinicalTrials.gov (ID: NCT07066137). Ethical approval was obtained from the local Ethics Committee (Protocol code: IRCCSME 47/2023).

2.3 Outcome measure

MS patients completed a comprehensive battery of clinical, cognitive, and socioemotional assessments at T0 and T1 (Table 2). The outcome measures are covered four main domains: SC and emotional processing, alexithymia, emotional self-efficacy, and cognitive-motor functioning.

DomainAssessment toolMain subscales/componentsCut-off/interpretationReferencesSocial cognitionEmpathy Quotient (EQ)Cognitive Empathy (EQ-CE); Emotional Reactivity (EQ-ER); Social Skills (EQ-SS); Total scoreEQ Total < 30 = low empathy (clinical range); 30–50 = below averageBaron-Cohen and Wheelwright, 2004Emotional processing/awarenessToronto Alexithymia Scale (TAS-20)Difficulty Identifying Feelings (TAS-IF); Difficulty Describing Feelings (TAS-DF); Externally Oriented Thinking (TAS-EO); Total score (TAS-TOT)TAS-20 Total ≥ 61 = alexithymia; 52–60 = possible alexithymiaBagby et al., 2014Emotional self-efficacyEmotional Self-Efficacy Scale (ESES)Regulation of Positive Emotions; Regulation of Negative EmotionsNo fixed clinical cut-off; higher scores = greater perceived self-efficacyCaprara et al., 2008Cognitive functioningRao’s Brief Repeatable Battery of Neuropsychological Tests (BRB-N)SRT (LTS, CLTR, D); SPART (Immediate, Delayed); SDMT; PASAT (2, 3 s); WLGSRT-LTS (cut-off: 23.3); SRT-CLTR (15.5); SRT-D (4.9); SPART (12.7); SPART-D (3.6); SDMT (37.9); PASAT 3 (28.4); PASAT 2 (17.1); WLG (17.0 men/women). Scores below cut-off indicate impaired performanceTedone et al., 2024Motor/balance functionBerg Balance Scale (BBS)Total score assessing postural control and fall riskBBS ≤ 45 = increased fall riskBerg et al., 1992

Neuropsychological and clinical assessment tools and reference cut-offs used for MS patients.

SC was evaluated using the Empathy Quotient (EQ), which measures three core components of empathic functioning: Cognitive Empathy (EQ-CE), which reflects the ability to understand the mental states of others; Emotional Reactivity (EQ-ER), which indicates affective reactivity to the emotions of others; and Social Skills (EQ-SS), which measures the ability to manage social interactions (Baron-Cohen and Wheelwright, 2004).

Emotional awareness and regulation were assessed using the Toronto Alexithymia Scale (TAS-20), a standardized self-report questionnaire that measures three aspects of alexithymia: difficulty identifying emotions (TAS-IF), difficulty describing emotions (TAS-DF), and externally oriented thinking (TAS-EO), along with a total score (TAS-TOT) (Bagby et al., 2014).

Participants’ perceived ability to manage emotions was measured using the Emotional Self-Efficacy Scale, which separately assesses effectiveness in regulating positive and negative emotions (Emotional Self-Efficacy—Positive/Negative). Higher scores indicate greater confidence in one’s emotional regulation abilities (Caprara et al., 2008).

Cognitive performance was assessed using the Rao Brief Repeatable Battery of Neuropsychological Tests (BRB-N), widely used in MS research (Tedone et al., 2024). The battery includes: Selective Reminding Test (SRT-LTS, SRT-CLTR, SRT-D) for learning and verbal memory; Spatial Recall Test (SPART, SPART-D) for visuospatial memory; Symbol Digit Modalities Test (SDMT) for processing speed; Pacing Auditory Serial Addition Test (PASAT 2 and PASAT 3) for attention and working memory; and Word List Generation (WLG) for verbal fluency and executive functions.

Functional balance was assessed using the Berg Balance Scale (BBS), a validated clinical measure of postural control, and fall risk (Berg et al., 1992).

All measures were collected at two time points (T0, T1) under standardized conditions. The comprehensive assessment was designed to capture multidimensional changes in cognitive, emotional, and motor domains following immersive VR training.

Healthy subjects were assessed by EEG only.

2.4 EEG data acquisition and processing

EEG activity was recorded from all participants (7 MS patients at T0 and T1 and 7 HC) during resting state with eyes closed. EEG signals were collected using a Brain-Quick System (Micromed) with 19 electrodes positioned according to the international 10–20 system (Fp1, Fp2, F7, F8, F3, F4, Fz, T3, T4, C3, C4, Cz, P3, P4, Pz, T5, T6, O1, O2). Although different electrode configurations or reduced electrode sets can be adopted depending on the study design (Ding et al., 2025), the full 19-channel 10–20 montage represents the conventional configuration used in clinical EEG practice (Seeck et al., 2017). Signals were acquired in a differential montage, referenced to linked earlobes. Electrode impedances were kept below 5 kΩ. Data were acquired at a sampling rate of 1024 Hz. EEG preprocessing was carried out using EEGLAB (v2025.1.0) and custom MATLAB (R2023b) scripts. A notch filter was applied at 50 Hz to reduce line noise and improve the signal-to-noise ratio. Data were then band-pass filtered between 0.5 and 40 Hz and downsampled to 256 Hz. Bad channels and bad segments were automatically removed before running the ICA. The removed channels were then interpolated, and the data were finally average-referenced. All preprocessing steps were performed offline. PSD was computed using Welch’s method, with 2-second Hamming windows and 50% overlap, to obtain a stable estimate of frequency content for each electrode. Spectral parameterization was then performed using the FOOOF algorithm (Donoghue et al., 2020). Traditional EEG spectral analyses typically quantify power within predefined frequency bands, but they do not distinguish between true oscillatory activity and broadband aperiodic components of the neural power spectrum. Consequently, variations in the aperiodic background activity may lead to misinterpretations of oscillatory power changes. The FOOOF approach addresses this limitation by modeling the neural power spectrum as the sum of two distinct components: an aperiodic component, which represents the “1/f” background activity, and periodic components, which correspond to oscillatory peaks that rise above this background. The aperiodic components derive from non-rhythmic EEG activity. Specifically, the exponent reflects the general decay of power with increasing frequency and is thought to estimate overall excitatory-inhibitory balance in cortical circuits, while the offset reflects a uniform shift of power across frequencies (Gao et al., 2017; Donoghue et al., 2020). Importantly, recent clinical research suggests that features of the aperiodic component may represent potential electrophysiological biomarkers across several neurological and psychiatric conditions, highlighting the relevance of separating periodic and aperiodic activity when interpreting EEG spectral changes (Pani et al., 2022).

The functional form is:

where b is the offset (overall vertical translation), k is the “knee” parameter (optional, representing a bend in the spectrum), and χ is the exponent reflecting the slope of the aperiodic component in log–log space. In contrast, the periodic components capture rhythmic oscillations, such as alpha or beta activity, linked to network synchronization (Buzsáki and Draguhn, 2004). Each peak n is modeled by a Gaussian function:

where a is the peak height (power above the aperiodic background), c is the center frequency of the peak, w is the width (bandwidth) of the Gaussian, and F is the array of frequency values. The full model of the neural power spectra (NPS) therefore becomes:

For each participant and electrode, the FOOOF fitting returned two aperiodic parameters (exponent and offset) and periodic peaks for alpha (8–13 Hz) and beta (13–30 Hz) bands. The FOOOF model was initialized with a maximum of six peaks, a minimum peak threshold of 0.5, and peak width limits between 0.5 and 12 Hz. These parameters were chosen to allow the identification of multiple oscillatory components, while minimizing the risk of detecting spurious peaks.

Average band power values were computed across electrodes for group-level analyses.

2.5 Procedures

After providing written informed consent, participants were submitted to baseline assessments (T0). These included cognitive, motor, and socio-emotional assessments, as well as resting-state EEG recordings. Only the MS group participated in the immersive VR rehabilitation program, while healthy controls underwent only baseline EEG acquisition.

The rehabilitation program consisted of 20 sessions, delivered three times per week for approximately 8 weeks, with each session lasting about 45 min of active training (approximately 1 h including preparation and breaks). Training was delivered using the CAREN immersive system, a multisensory platform based on motion capture that integrates a dynamic platform, motion tracking, and multiple synchronized projection screens.

The immersive environment enabled real-time interaction with virtual activities designed to simultaneously stimulate cognitive, motor, and social areas. Four different virtual scenarios were used, focusing primarily on dual-task and socio-emotional training. Dual-task exercises combined cognitive and motor demands to improve attentional flexibility and executive control, while socio-emotional scenarios included social stimuli, such as interactive environments (e.g., a virtual city with a pizza cart), aimed at improving emotional processing and perspective-taking. Task difficulty and sensory feedback were progressively adapted to each patient’s performance to maintain engagement and ensure optimal challenge.

During each session, task complexity and environmental feedback were progressively adapted based on individual performance to maintain engagement and ensure adaptive challenge. Visual and auditory feedback were provided in real time to facilitate self-monitoring and reinforce correct execution.

At the end of the rehabilitation program, MS patients underwent post-intervention assessments (T1) identical to those administered at baseline, which included cognitive, emotional, and motor assessments, as well as EEG recordings. All sessions were supervised by qualified neuropsychologists and physiotherapists with experience in neurorehabilitation. Therapists responsible for the intervention were not involved in outcome evaluations to minimize bias.

2.6 Statistical analysis

Between-group comparisons (MS at T0 vs. HC) were performed using independent-samples two-tailed t-tests, whereas within-patient longitudinal comparisons (T0 vs. T1) were conducted using paired-samples two-tailed t-tests. All t-tests were computed pairwise for homologous electrodes. P-values were corrected for multiple comparisons using the false discovery rate (FDR, p < 0.05). Within the MS group, Pearson correlation coefficients were computed between EEG band power and the neuropsychological measures collected at baseline and post-intervention. All statistical analyses were conducted using MATLAB (R2023b).

3 Results3.1 Demographic and clinical characteristics

Demographic and clinical characteristics of the MS patients and HC are summarized in Table 1. The two groups were comparable in terms of age and sex distribution. In the MS group, disability levels remained stable across the intervention period, as reflected by identical EDSS scores at T0 and T1. No pharmacological treatment was present in the healthy control group.

3.2 MS (T0) vs. MS (T1)3.2.1 Neuropsychological assessment

Neuropsychological performance at T0 and T1 in the MS group is reported in Table 3. Paired-sample t-tests (two-tailed) were conducted to assess longitudinal changes. Since the sample size was limited (n = 7), p-values were not corrected for multiple comparisons, consistent with exploratory pilot-study designs. Effect sizes (Cohen’s d) and their corresponding 95% confidence intervals (CI) were also computed to quantify the magnitude of the observed changes. A statistically significant improvement was observed in the SRT-LTS score (p = 0.011), reflecting enhanced long-term storage verbal memory following the VR-based rehabilitation program. All other cognitive, emotional, and motor measures showed no significant longitudinal variations.

TestT0 (mean ± SD)T1 (mean ± SD)p-valueCohen’s d95% CIEQ-CE4.57 ± 1.815.71 ± 2.630.1030.44[−0.15, 1.27]EQ-ER5.86 ± 2.856.00 ± 1.730.8900.05[−0.96, 1.09]EQ-SS6.14 ± 1.215.86 ± 2.540.715−0.12[−1.08, 0.76]TAS-DF12.71 ± 4.9610.57 ± 4.650.200−0.39[−1.29, 0.30]TAS-IF14.29 ± 4.3113.57 ± 4.390.786−0.14[−1.60, 1.23]TAS-EO17.57 ± 6.8320.86 ± 6.040.3250.44[−0.63, 1.77]TAS-TOT44.57 ± 6.8545.00 ± 11.600.9280.04[−1.11, 1.21]Emotional self-efficacy (positive emotions)29.14 ± 6.0728.29 ± 6.870.356−0.11[−0.48, 0.19]Emotional self-efficacy (negative emotions)24.00 ± 5.3923.57 ± 5.860.802−0.07[−0.80, 0.63]SRT-LTS34.95 ± 14.0951.38 ± 9.100.011*1.20[0.29, 2.75]SRT-CLTR29.83 ± 14.5736.41 ± 19.040.2930.34[−0.42, 1.28]SPART15.08 ± 5.9918.65 ± 7.780.1600.45[−0.27, 1.41]SDMT38.17 ± 11.2637.59 ± 13.590.712−0.04[−0.35, 0.24]PASAT 335.95 ± 15.0740.24 ± 13.610.2620.26[−0.28, 0.95]PASAT 229.21 ± 9.5335.21 ± 14.460.0710.43[−0.07, 1.15]SRT-D7.27 ± 1.778.27 ± 1.930.2180.47[−0.41, 1.61]SPART-D6.49 ± 3.136.63 ± 2.780.8840.04[−0.72, 0.83]WLG25.73 ± 8.2427.16 ± 6.900.5870.16[−0.60, 1.02]

Neuropsychological test scores for MS patients at baseline (T0) and post-intervention (T1).

EQ-CE, Empathic Concern; EQ-ER, Emotional Reactivity; EQ-SS, Social Skills; TAS-DF, Difficulty Describing Feelings; TAS-IF, Difficulty Identifying Feelings; TAS-EO, Externally-Oriented Thinking; TAS-TOT, Toronto Alexithymia Scale Total; Emotional Self-Efficacy—Positive Emotions; Emotional Self-Efficacy—Negative Emotions; SRT-LTS, Selective Reminding Test—Long-Term Storage; SRT-CLTR, Selective Reminding Test—Consistent Long-Term Retrieval; SPART, Spatial Recall Test; SDMT, Symbol Digit Modalities Test; PASAT-3, Paced Auditory Serial Addition Test-3 seconds; PASAT-2, Paced Auditory Serial Addition Test-2 s; SRT-D, Selective Reminding Test—Delayed; SPART-D, Spatial Recall Test—Delayed; WLG, Word List Generation.

*Indicates statistically significant differences (p < 0.05).

3.2.2 Clinical significance of cognitive, emotional, and social-cognitive changes

Although most p-values did not reach statistical significance due to the limited sample size (n = 7), several outcomes demonstrated changes that approached or exceeded established minimal clinically important differences (MCID) (Draak et al., 2019) thereby supporting the potential clinical relevance of the observed effects (Table 4).

DomainOutcome measureMCID/Clinically meaningful thresholdObserved change (T1–T0)Clinical interpretationKey referencesVerbal learning and memorySRT-LTSMCID ≈ +10–15% improvement in verbal learning+16.43 points (≈ +47%)Exceeds MCID; large improvement in long-term verbal storageBoringa et al., 2001Working memory and attentionPASAT-2Clinically meaningful change = +10–15%+6 points (≈ +20.5%)Clinically meaningful despite p > 0.05Working memory and attentionPASAT-3Clinically meaningful = +10–15%+4.29 points (≈ +12%)Meets MCID rangeProcessing speedSDMTMCID = +4 points−0.58 pointsNo clinically meaningful changeBenedict et al., 2020Visuospatial memorySPARTNo fixed MCID; ≥ 10–15% considered relevant+3.57 points (≈ +24%)Clinically coherent improvementBoringa et al., 2001Executive function/verbal fluencyWLGNo formal MCID; 10–20% improvement considered relevant+1.43 points (≈ +5.6%)Small but coherent improvementCognitive empathyEQ-CENo MCID; small (5–10%) changes can be meaningful+1.14 points (≈ +25%)Notable improvement in cognitive empathyPöttgen et al., 2013Emotional reactivityEQ-ERAs above+0.14 points (≈ +2.3%)Social skillsEQ-SSAs above−0.28 points (≈ 4.6%)No improvementRaimo et al., 2017Difficulty identifying feelingsTAS-IFDecreases of 3–5 points considered meaningful−0.72 points (≈ 5%)Below meaningful thresholdBagby et al., 2014Difficulty describing feelingsTAS-DFDecreases of 3–5 points meaningful−2.14 points (≈ 16.8%)Trend toward improvementExternally oriented thinkingTAS-EOSame threshold+3.29 points (≈ 18.7%)No improvementAlexithymia total scoreTAS-TOTDecrease ≥ 5 points considered meaningful+0.43 points (≈ +1%)No improvementPositive emotional self-efficacyESES-positiveNo MCID; small absolute changes meaningful−0.85 points (≈ +2.9%)Slight reductionCaprara et al., 2008Negative emotional self-efficacyESES-negativeSame−0.43 points (≈ 1.8%)No improvement

Clinically meaningful change (MCID) thresholds and observed changes in MS patients.

EQ-CE, Empathy Quotient-Empathic Concern; EQ-ER, Empathy Quotient-Emotional Reactivity; EQ-SS, Empathy Quotient-Social Skills; TAS-DF, Toronto Alexitimia Test-Difficulty Describing Feelings; TAS-IF, Toronto Alexitimia Test-Difficulty Identifying Feelings; TAS-EO, Toronto Alexitimia Test-Externally-Oriented Thinking; TAS-TOT, Toronto Alexithymia Scale Total; Emotional Self-Efficacy—Positive Emotions; Emotional Self