Introduction

Multiple sclerosis (MS) is a neurological disease affecting over 1.8 million people worldwide1, characterized by immune-mediated damage to the brain and spinal cord that can impair cognitive, emotional, motor, sensory, and visual functions. As it is known, MS predominantly affects women, with a female-to-male ratio of 3:1, typically diagnosed in young adulthood, between 20 and 40 years of age. Furthermore, according to the International Advisory Committee on Clinical Trials of MS2, this disease is currently divided into four different types, namely clinically isolated syndrome (CIS), Relapsing-remitting MS (RRMS), Secondary progressive MS (SPMS), Primary progressive MS (PPMS). Moreover, although not considered a type of MS, the Radiologically isolated syndrome (RIS) has been used to classify cases with abnormalities of the brain and/or spinal cord consistent with MS lesions3.

Given its chronic and unpredictable nature, MS profoundly affects the quality of life (QoL) of people, not only through direct physical impairments but also through psychological and social dimensions, including increased levels of anxiety, stress, and depression4. People with multiple sclerosis (PwMS) frequently report elevated levels of anxiety and depressive symptoms, which often exceed those experienced by the general population5,6. Depression and anxiety in PwMS are associated with higher disability levels, increased fatigue, cognitive impairment, and a general decline in QoL, significantly impacting their daily functioning and overall well-being4,6. Additionally, the distress resulting from managing unpredictable symptoms and disease progression further contributes to psychological burdens, creating a cyclical deterioration in mental health and QoL7. Beyond these psychological correlates, PwMS often face stigma associated with chronic illness, which further complicates their psychosocial adaptation and QoL8,9.

Stigma is broadly defined as the devaluation of an individual based on perceived differences10,11. In particular, perceived stigma is characterized by the awareness or expectation that others hold negative stereotypes about one’s condition, leading to feelings such as fear of judgment, and rejection12. On the other hand, enacted stigma is what actually happens, such as discriminatory remarks, exclusion, and unequal treatment from others13. Internalized stigma, instead, is turning the social stereotypes inward, producing self-devaluation, shame, and low self-esteem13. Despite these constructs are related, are also distinct: perceived/enacted are social and external; internalized is intrapsychic and self-directed. Perceived or enacted stigma can, over time, contribute to self-stigma, but this transition is not inevitable. In PwMS stigma has been associated with higher rates of depression, social withdrawal, concealment of diagnosis and lower health-related QoL14,15. Furthermore, the literature shows that stigma can act through multiple pathways, exacerbating disability and limiting access to care and social participation16. In addition to what has been said so far, it is important to underline that there is growing recognition of the critical role of physical activity (PA) in mitigating the negative outcomes and enhancing QoL in PwMS17. It has been demonstrated that regular PA leads to beneficial effects, including reduced fatigue, improved mood, better management of stress and anxiety, and enhanced overall physical function18. However, PwMS often reduce or modify their PA after diagnosis, switching from group to individual training19. This transition reflects not only physical limitations but also stigma. This transition reflects not only physical limitations but also stigma. As hypothesized by Powel and collaborators (2024)16, stigma, which may be enacted by others or internalized, would discourage participation in activities that disclose symptoms. Such avoidance reduces PA and reinforces the cycle of disability progression and diminished perceived self-efficacy. The Health Belief Model helps explain these behaviors by highlighting how perceived susceptibility, severity, benefits, barriers, incentives for action, and self-efficacy influence engagement in PA20[,21. It is important to note that barriers in MS include fatigue, limited access to adapted programs, and fear of negative evaluation (i.e., stigma as a psychosocial barrier), while internalized stigma can undermine self-efficacy, which decreases the likelihood of starting or sticking with exercise10. Furthermore, internalized stigma can further reduce self-efficacy and decrease the likelihood of starting or maintaining exercise. When stigma is considered in this framework, it emerges as an important psychosocial barrier that undermines adherence and limits the health benefits of exercise in MS. Despite the above, it is interesting to note that the relationship between stigma and PA in PwMS has remained largely unexplored.

In addition, while few studies have examined how stigma differs by MS type or may be modulated by sociodemographic variables, no study to date has systematically investigated the interplay between stigma and a comprehensive constellation of factors in PwMS, including clinical subtypes, years of illness, physical and mental well-being, and sociodemographic factors. The aims of this study were to determine, through a mixed-method approach, the complex interplay between stigma, health perception, psychological well-being, and PA in PwMS. Specifically, we aimed: (i) to quantify, using regression models, the effects of perceived stigma on levels of depression, anxiety, stress, and PA; (ii) to examine whether perceived physical and mental health status are associated with the degree of stigma; (iii) to assess whether these relationships are further shaped by clinical variables such as age, gender, BMI, years of MS and its subtype, and the presence of symptoms; (iv) to explore PwMS perspectives on QoL and PA through the analysis of open-ended responses. We hypothesize that perceived stigma would be associated with poorer psychological well-being and lower engagement in PA, and that perceived health status would be related to perceived stigma.

Materials and methods

Participants and data collection

A total of 127 participants were recruited, all of whom completed an online questionnaire comprising several standardized psychometric tools. The mean age of the sample was 45.09 years (SD = 11.57), with age ranging from 19 to 69 years and an average BMI of 24.57 (SD = 5.23). The sample consisted of 90% women (n = 114) and 10% men (n = 13). Inclusion criteria were age ≥ 18 years, clinically confirmed diagnosis of any subtype of multiple sclerosis, ability to independently provide informed consent, and adequate cognitive status to complete the questionnaire. Exclusion criteria included the presence of comorbid neurological or psychiatric disorders, the presence of an acute relapse phase, and incomplete or inconsistent responses to the questionnaire. To ensure self-report validity, we asked the neurologist to exclude participants with one or more of the following current or recent (within the past 12 months) psychiatric disorders, which are typically associated with marked cognitive/behavioral deterioration or acute clinical risk: (i) schizophrenia spectrum disorders and other psychotic disorders (including schizoaffective disorder); (ii) bipolar disorder in a current manic or hypomanic episode (or with psychotic features); (iii) major depressive episode with psychotic features or with active suicidal ideation/intent requiring urgent treatment; iv) moderate to heavy substance use (alcohol or drugs) in the past 12 months; v) mental disability or severe neurocognitive disorder precluding informed consent or reliable completion of the questionnaire; vi) current psychiatric hospitalization or involuntary treatment. On the other hand, frequent ambulatory depressive or anxiety symptoms/disorders (without acute risk or psychosis) were not excluded, as they are common in MS and integral to the psychosocial context we wished to characterize. The participants, after the routine neurological examination, noted by the physician that the inclusion criteria were present and the exclusion criteria were absent, were asked to fill out the online questionnaires by providing them with a QR code. There was no financial compensation or incentives for participation. This study was approved by the Ethics Committee of Split University, Croatia, with the number Class: 003–08/25 − 04/001 and Official number: 2181-205-02-05-25-045. Written informed consent was obtained from all participants. All procedures performed in the study were in accordance with the 1964 Declaration of Helsinki and its later amendments. The study was conducted as part of a PRIN 2022 project (code: 2022CTPP38) funded by the European Union – Next Generation EU, Mission M4 Component C2.

Questionnaires

Short Form Health Survey – 12 items (SF-12). The SF-1222 is a self-report instrument consisting of 12 items designed to assess perceived health status. It is a short version of the SF-36 and maintains strong correlations with its full form. The 12 items evaluate eight health domains: physical functioning, role limitations due to physical health, bodily pain, general health, vitality, social functioning, role limitations due to emotional problems, and mental health. Items are rated using various Likert scales (e.g., 1 = ‘Excellent’ to 5 = ‘Poor’ for general health; 1 = ‘Not at all’ to 5 = ‘Extremely’ for functional limitations). Scoring requires a weighted algorithm that generates two component summary scores: Physical Component Summary (PCS), Mental Component Summary (MCS). These scores are normalized to a standard scale based on population norms.

Stigma Scale for Chronic Illness – 8 items (SSCI-8). The SSCI-823 is a brief version of the original SSCI-2424 developed to assess perceived internal (i.e., internalized or self-stigma) and external stigma (i.e., enacted) in individuals with chronic illnesses. It includes 8 items, each rated on a 5-point Likert scale ranging from 1 (Never) to 5 (Always). Due to the right-skewed distribution of these variables, we decided to categorize it into quartiles. After calculating the quartile values, for the variable internalized stigma, all values between 1 and 3 were assigned to the first quartile, all values between 4 and 5 to the second quartile, all values between 6 and 9 to the third quartile and all values above 10 to the last quartile. For the variable enacted stigma, the first quartile comprises values between 1 and 4, the second quartile between 5 and 6, the third quartile between 7 and 9 and the fourth quartile all values greater than 10.

Depression Anxiety Stress Scales – 21 items (DASS-21). The DASS-2125 is an abbreviated version of the original DASS-4226 designed to measure symptoms across three domains of negative affectivity: depression, anxiety, and stress. The instrument consists of 21 items, with 7 items per subscale: Depression (e.g., dysphoria, anhedonia), Anxiety (e.g., autonomic arousal, situational anxiety), Stress (e.g., tension, irritability). Each item is rated on a 4-point Likert scale ranging from 0 (‘Did not apply to me at all’) to 3 (‘Applied to me very much or most of the time’). Scores for each subscale are calculated by summing the item scores and multiplying by 2 to match the original DASS-42 scaling. Subscale scores range from 0 to 42, with cut-offs available to classify symptom severity (further information in Table 1).

Table 1 Levels severity of each subgroup of the DASS-21 questionnaire.
Full size table

Oxford Happiness Questionnaire – Short Form (OHQ-8). The OHQ-827 is a short form of a OHQ-2928 developed to measure subjective happiness and psychological well-being. Each item is rated on a 6-point Likert scale ranging from 1 (Strongly disagree) to 6 (Strongly agree). Three of the eight items are reverse-scored. Scoring involves computing the mean of all item scores, resulting in a total score ranging from 1 to 6. Higher scores indicate greater levels of happiness and well-being. This variable was further divided into quartiles: the first quartile ranges from 1 to 24, the second quartile from 25 to 30, the third quartile from 31 to 35 and the fourth quartile above 36.

International Physical Activity Questionnaire - Single Item. The Single-Item IPAQ29, is a validated self-report measure designed for large-scale population surveillance. The item asks participants: “In the past week, on how many days have you done a total of 30 minutes or more of physical activity, which was enough to raise your breathing rate? This may include sport, traditional games, exercise and brisk walking or cycling for recreation or to get to and from places, but should not include housework or physical activity that may be part of your job”. An open-response scale was used, with valid responses ranging from 0 to 7 days for the past-week, where each point corresponds to the number of days with ≥ 30 min of moderate-to-vigorous PA. As reported by authors, the single-item days measure classified 29.9% meeting the threshold of 5 days per week, but this increased to 59.7% when the single-item days measure used the cut point of 3 or more days per week. Therefore, we consistently chose to divide the participants into those engaging in PA less than or equal to three days per week and those exceeding three days per week.

Qualitative analysis

Participants responded to the following single open-ended question included in the aforementioned online questionnaire: “What do you think could improve your quality of life as a person living with multiple sclerosis?”. We chosen a solution-focused wording to minimize burden and ambiguity, and to elicit participant-prioritized facilitators with direct translational relevance, given that negatively valenced constructs (e.g., […] what could affect your quality of life) were already assessed in other questionnaires. The open-ended responses were analyzed using a qualitative content analysis approach with an inductive framework30,31. Initial coding was conducted manually by two independent researchers with backgrounds in health psychology and qualitative research. Codes were then grouped into broader themes through constant comparison and iterative refinement. Discrepancies were resolved through discussion until consensus was reached. Saturation was achieved when no new themes emerged during analysis. Data were managed using Microsoft Excel and MAXQDA 2022 to facilitate coding and theme organization. Frequencies were calculated for each theme to highlight prevalence, though qualitative richness was preserved in interpretative reporting. A reflective attitude was maintained throughout the process to ensure transparency and mitigate researcher bias.

Statistical analysis

Descriptive statistics were performed for the categorical covariates such as the three subgroups of the DASS-21 questionnaire (depression, anxiety, stress) and their severity levels, the type of multiple sclerosis, the presence or absence of symptoms, internalized and enacted stigma. Ordinal logistic regression analysis was performed to analyze the relationship between the DASS-21 subgroups as dependent variables and eight independent variables: Age, BMI, gender (F = 0; M = 1), years of MS, type of MS, internalized and enacted stigma. The same model was used with the dependent variable replaced by the happiness variable. To examine associations with perceived stigma (dependent variable), we included PCS and MCS as independent variables. In each of these analyses, the proportional odds assumption was tested, and the results of the Pearson test and the deviance test were considered to assess the goodness of fit and the Pseudo-R2 value (Nagelkerke method). Binary logistic regression was performed to examine the effects of self and enacted stigma on PA levels. In this model, the Box-Tidwell procedure was performed to test the linearity of the logit. In addition, the Hosmer and Lemeshow test was performed to assess the goodness of fit of the model. The α was set to 0.05. Data were analyzed using IBM SPSS Statistics, version 29.0.1.0.

Results

Quantitative results

Descriptive statistics

The percentage distribution of participants’ responses to the DASS-21 questionnaire is summarized in Table 2.

Table 2 Descriptive statistics for the categorical covariates.
Full size table

DASS-21 questionnaire

The analysis addressing depression revealed statistically significant model (χ²(8) = 41.04, p < 0.001), with Pseudo-R2 of 0.34, and Pearson (χ²(370) = 351.53, p = 0.747) and Deviance (χ²(370) = 171.08, p = 1.000) test indicating a good fit of the model to the data with internalized-stigma as significant predictor (B = 0.452, OR = 1.57 [CI95% = 1.27–1.94], p < 0.001). For anxiety, the model was statistically significant (χ²(8) = 22.82, p = 0.004), with Pseudo-R2 of 0.19, and Pearson (χ²(370) = 346.93, p = 0.80) and Deviance (χ²(370) = 222.40, p = 1.000) test indicating a good fit of the model to the data with internalized stigma as significant predictor (B = 0.215, OR = 1.24 [CI95% = 1.03–1.49], p = 0.021). Lastly, for stress the model was statistically significant (χ²(8) = 19.60, p = 0.012), with Pseudo-R2 of 0.22, and Pearson (χ²(244) = 280.38, p = 0.055) and Deviance (χ²(244) = 111.29, p = 1.000) test indicating a good fit of the model to the data with internalized stigma as significant predictor (B = 0.382, OR = 1.47 [CI95% = 1.15–1.87], p = 0.002; Fig. 1).

Fig. 1
Fig. 1The alt text for this image may have been generated using AI.
Full size image

Odds ratio estimates of the internalized stigma predictor of the three domains of negative affectivity of DASS-21 questionnaire: depression, anxiety, and stress. Error bar: 95% Confidence Interval; Red line: null value of the association.

OHQ- 8 questionnaire

The ordinal logistic regression analysis addressing happiness revealed statistically significant model (χ²(8) = 46.78, p < 0.001), with Pseudo-R2 of 0.35, and Pearson (χ²(244) = 359.86, p = 0.232) and Deviance (χ²(244) = 220.03, p = 0.863) test indicating a good fit of the model to the data with internalized stigma and gender as significant predictors (B = −1.02, OR = 0.36 [CI95% = 0.24–0.56], p < 0.001; B = 1.69, OR = 5.42 [CI95% = 1.48–19.87], p = 0.011; Fig. 2).

Fig. 2
Fig. 2The alt text for this image may have been generated using AI.
Full size image

Odds ratio estimates of the internalized stigma predictor of the OHQ and IPAQ questionnaire, and Gender predictor of the ordinal logistic model which used happiness as the dependent variable. Error bar: 95% Confidence Interval; Red line: null value of the association.

IPAQ questionnaire

The logistic regression for PA did not reach the prespecified α = 0.05 (χ²(8) = 14.40, p = 0.072; Pseudo-R²=0.156). The Hosmer and Lemeshow test suggested a good fit to the data (χ²(8) = 4.8, p = 0.760). While the internalized-stigma coefficient was statistically significant (B = − 0.69, OR = 0.50 [CI95% = 0.31–0.83], p = 0.007), inference at the predictor level should be interpreted cautiously (Fig. 2). We therefore treat this as a hypothesis-generating indication that requires replication with a priori model specification and more granular PA measures.

Perceived health status (PCS, MCS) versus perceived stigma

The ordinal logistic regression analysis conducted for the dependent variable internalized stigma revealed statistically significant model (χ²(8) = 69.60, p < 0.001), with Pseudo-R2 of 0.45, and Pearson (χ²(370) = 398.02, p = 0.152) and Deviance (χ²(370) = 271.95, p = 1.000) test indicating a good fit of the model to the data with both PCS and MCS as significant predictors (B = −0.109, OR = 0.90 [CI95% = 0.86–0.94], p < 0.001; B = −0.116, OR = 0.89 [CI95% = 0.86–0.93], p < 0.001). For enacted stigma, the ordinal logistic regression analysis revealed statistically significant model (χ²(8) = 31.91, p < 0.001), with Pseudo-R2 of 0.24, and Pearson (χ²(370) = 371.02, p = 0.475) and Deviance (χ²(370) = 304.86, p = 0.994) test indicating a good fit of the model to the data with both PCS and MCS as significant predictors (B = −0.066, OR = 0.94 [CI95% = 0.90–0.97], p < 0.001; B = −0.055, OR = 0.95 [CI95% = 0.92–0.98], p = 0.002; Fig. 3).

Fig. 3
Fig. 3The alt text for this image may have been generated using AI.
Full size image

Odds ratio estimates of the PCS and MSC versus perceived stigma. Error bar: 95% Confidence Interval; Red line: null value of the association; PCS: Physical Component Summary; MSC: Mental Component Summary.

Qualitative results

Emergent themes

Five core themes were identified from participants’ responses, reflecting their perceived needs to improve their QoL:

Psychological support and emotional care

The most prominent theme was the need for continuous psychological support, mentioned in over half of the responses. PwMS highlighted a desire for accessible psychological support, particularly from professionals trained to understand the specific psychological burden of MS:

“I need a psychological path tailored to someone with MS, not generic counseling”.

“This disease affects me more mentally than physically. I feel judged when I talk about it”.

Some PwMS emphasized the importance of psychological help as an integral part of care, especially in the early stages post-diagnosis:

“Psychological support should be automatically included after diagnosis, not optional”.

Need for physical activity assisted by experts

Another widely cited theme concerned the importance of PA guided by professionals. Responses highlighted the need for personalized exercise programs more than physiotherapy ones:

“Physical activity supported by experts”.

“Expert-assisted physical activity is essential, it helps my mind and body alike”.

Furthermore, some PwMS referred to regional disparities and the lack of financial or logistical access to such services.

Healthcare system efficiency and organization

Participants reported structural and systemic difficulties in accessing services, managing appointments, and receiving consistent care:

“There should be a hospital care pathway for exams and appointments related to multiple sclerosis, automatic and centralized”.

“If the national healthcare system worked as it should, our lives would already be better”.

Many noted delays, lack of coordination between specialties, and a sense of abandonment, especially regarding follow-up care and communication with neurologists.

Socioeconomic and occupational support

The impact of MS on job and financial independence emerged as a salient concern. Some participants suggested part-time employment solutions and better disability-related benefits:

“People with MS should be allowed to work part-time without penalties”.

“Economic serenity is crucial for living with this disease”.

Freelance workers, in particular, reported the absence of support and safeguards available to salaried employees.

Holistic and person-centered approach to care

Several responses emphasized the need for an integrated, empathetic approach that goes beyond treating physical symptoms. Patients called for personalized care that values their individuality and life context:

“I want to be seen as a person, not just as a set of symptoms”.

“More understanding, more listening. Sometimes doctors minimize what we feel”.

Subthemes within this category included social connectedness, lifestyle changes (e.g., diet, meditation, nature), and family/community support.

Table 3 Emerging themes with relative descriptions and percentages.
Full size table

Discussion

In the present study it has been shown that internalized stigma is a strong predictor of psychological distress, manifested as increased depression, anxiety, and stress, and that it is also associated with lower levels of happiness and lower likelihood of participation in PA among PwMS. Among the different models tested, internalized stigma emerged as a consistent and robust predictor of negative outcomes across all domains of psychological well-being and showed a strong inverse relationship with happiness. These findings are consistent with existing literature that identifies internalized stigma as a central factor for psychological distress in people with chronic illnesses, including PwMS6,14. Our findings also showed that both PCS and MCS significantly predict internalized stigma, with lower scores on the SF-12 associated with greater levels of self-stigmatization. Therefore, perceived health and psychological status maintain a mutual relationship. Furthermore, lower levels of perceived PCS and MCS may foster internalized stigma by amplifying feelings of inadequacy and self-devaluation in PwMS, thereby reinforcing negative self-perceptions7,16. In addition to this strong predictive model for psychological variables, enacted stigma also reached statistical significance, albeit to a lesser extent than internalized stigma. This distinction may suggest that, rather than others’ judgments per se, their internalization is most strongly associated with mental health. This reflects previous work highlighting the detrimental effects of self-stigma as the proximate determinant of emotional well-being10. In addition, recent evidence has shown that higher alexithymic traits are associated with more negative perceptions of the disease in people with MS and their partners32. Although we did not examine alexithymia, this construct could plausibly influence how symptoms and social feedback are evaluated, and thus intersect with perceived and internalized stigma.On the other hand, one of the most interesting findings of our study concerns the relationship between internalized stigma and PA. Although the overall regression model narrowly missed statistical significance, internalized stigma emerged as a significant negative predictor of weekly PA engagement, with confidence intervals sufficiently distant from unity to suggest a stable effect. This result should be interpreted with relative caution because may indicate that some of the included predictors were not optimal for explaining PA behavior in PwMS. This is consistent with previous meta-analytic research showing that psychological barriers, including perceived incompetence and fear of judgment, hinder PA engagement in PwMS18. Our findings complement this data by suggesting that internalized stigma in particular may act as an implicit inhibitor, contributing to avoidance behaviors and reinforcing a sedentary lifestyle. In addition, recent findings suggest that anticipatory concerns related to fatigue may be a key psychological barrier to PA in PwMS22. Anticipation of fatigue is associated with lower participation in exercise-based interventions, even among individuals with adequate physical capacity. It is likely that anticipatory avoidance interacts with internalized stigma, reinforcing self-limiting beliefs and fostering a cycle of withdrawal and diminished self-efficacy. Such mechanisms support the need for integrated interventions that address both psychological and behavioral health factors. This is particularly concerning given the benefits of PA33,34, from reducing fatigue to improving psychological state and mobility in PwMS17,35. Addressing stigma, especially the internalized form, can therefore not only improve emotional outcomes, but also promote more active and health-promoting behaviors. The importance of this finding is also reflected in the qualitative data. It is important to emphasize that the question of “quality of life” refers to the priority areas related to this topic for PwMS (Table 3) and not directly to the perceived QoL. Consequently, our conclusions regarding QoL are exploratory rather than confirmatory in nature. That being said, more than half of participants cited the need for expert-supported PA as a key factor related to their QoL. This was often described not only in terms of physical rehabilitation, but also as a source of psychological support and reassurance. These findings are consistent with research suggesting that structured, supervised PA – particularly in socially supportive contexts – can mitigate the isolating effects of stigma and promote self-determination18. Beyond the area of PA, the qualitative responses reveal a broader constellation of unmet needs. The most common theme was a desire for psychological support, particularly interventions tailored to the experience of living with MS. This reflects recent findings that PwMS often struggle silently with emotional distress and feel misunderstood or pathologized by mainstream mental health services5. Many participants also highlighted the need for system reform, particularly in relation to fragmented health services and inadequate follow-up care, which exacerbate feelings of abandonment and contribute to perceived stigma. These accounts reflect the growing call for integrated care models that include psychological support as a standard component of MS care9. In addition, participants reported an urgent need for occupational and economic support, as the disease often affects financial autonomy and job security. Although this aspect was not statistically analyzed in the quantitative section, it reflects the literature that emphasizes the intersection of stigma and structural disadvantage in dealing with chronic illness15. The emphasis on holistic and person-centered care suggests that interventions must go beyond symptom control and instead assess the person’s broader psychosocial reality. What emerges in the qualitative analysis is in line with the literature on different types of diseases, which highlights the patients’ need to be recognized as a “person” and not by their disease36,37,38,39. Therefore, it is important to strongly reiterate that medicine must be able to put the patient back at the center of their own treatment pathway40,41. The present study, then, contributes to a growing body of literature pointing out the far-reaching consequences of stigma in PwMS. Regarding PA, we observed an exploratory association between higher internalized stigma and lower weekly PA. Because the omnibus model did not meet significance, this result should not be interpreted as confirmatory evidence but rather as a hypothesis-generating signal that aligns with prior work on psychological barriers to PA in PwMS. Future studies with prospective designs, preregistered models, and richer PA assessments are needed to confirm or refute this association. However, it can be suggested – also in light of qualitative surveys – that the promotion of PA among PwMS should take into account the psychological barriers created by stigma. It is therefore essential to develop interventions tailored for PwMS and that address not only functional impairment, but also the psychological well-being and PA engagement. Similarly, improving QoL requires a model of care that is psychologically informed, socially attuned, and person-centered.

Limitations of the study

Despite the interesting results of this study, several limitations should be taken into account. First, the generalizability to men is limited. Namely, our sample was predominantly female (~ 90%) and was drawn from random clinical samples. To both obtain gender-balanced samples and test gender-specific models, future studies should use quotas/targeted recruitment and a priori stratification (with possible post-stratification weighting). Second, we did not collect other measures, such as the Expanded Disability Status Scale (EDSS) or standardized measures of fatigue to examine confounding/mediation and to stratify estimates by disability level. Third, to reduce the number of questions, we chose to provide the IPAQ-SI for PA and explore perceptions of quality of life through an open-ended question, with the risk of individuals not completing the questionnaire. These decisions favor breadth over the depth and precision that broader (IPAQ) and quantitative (e.g., Quality of Life Scale – QOLS) scales can provide. In particular, the IPAQ-SI, while convenient, provides only an approximate estimate of engagement and does not account for intensity, type, or context (e.g., group activities versus home-based activities). Finally, the cross-sectional design and self-report limit inferences over time and can lead to bias. Nonetheless, the convergence between quanti- and qualitative findings strengthens the validity of the study, suggesting directions for more person-centered interventions in clinical and public health practice.

Conclusions

Our findings show that internalized stigma is more strongly associated with psychological well-being and engagement in PA in PwMS. In particular, we observed an exploratory association between internalized stigma and lower PA, which requires replication and should be interpreted with caution given the non-significant omnibus model and the PA measurement used. While targeting stigma remains a plausible avenue to improve well-being, this result should be interpreted with caution, as it may be due to limitations in the selection of predictors included to account for PA behaviors in PwMS. In addition to efforts to improve physical conditions or functional limitations, it is important to include measures to reduce stigma and strengthen identity in routine care. In addition, expert-led PA promotion, individualized psychological interventions for PwMS, and integrated health care pathways can not only alleviate distress but also actively improve engagement in healthy behaviors and overall life satisfaction.