Workplace and social support, treatment satisfaction, and their impact on quality of life in Swedish women with multiple sclerosis: a cross-sectional survey study.

STRENGTHS AND LIMITATIONS OF THIS STUDY

The study integrates data from the Swedish MS Registry, which covers 87% of Sweden’s population with multiple sclerosis (MS), with a 71.4% survey response rate from women, offering a comprehensive overview of individuals with MS in Sweden.

Despite a high response rate among women, a minority of women did not complete the survey, potentially introducing non-response bias into the findings.

The cross-sectional design of the study limits our ability to a cause-and-effect relationship, meaning we cannot conclude that perceived support directly impacts health-related quality of life for people with MS.

People with MS may receive support but this does not automatically ensure it is perceived as supportive.

Introduction

Multiple sclerosis (MS) is a chronic and potentially disabling neurological disease, typically diagnosed between the ages of 20 and 40. 1  Women are two to three times more likely to be diagnosed with MS compared with men. 1 MS can manifest with a wide range of cognitive, mental and physical symptoms, with fatigue being one of the most prevalent and burdensome in the patient’s daily life. 2 As follows, women with MS must adapt to and oversee this array of symptoms that can limit engagement across various aspects of life, including work, having a family and social interactions. 3 4 Further, the impact of parenthood on the work capacity of women with MS is more pronounced compared with men, 5 and sustaining employment is positively associated with better health-related quality of life (HRQoL). 6

One of the most common measures to assess HRQoL in MS is the EuroQol five-dimensional questionnaire (EQ-5D). 7 People with MS (PwMS) report lower levels of HRQoL when compared with the general population. 8 9 Specifically, domains like work, social life, relationships, housework, mood, fatigue, cognitive processing speed, walking and balance have been identified as relevant in quality of life in MS. 10 Indeed, both fatigue and cognitive difficulties, prevalent MS symptoms regardless of disability levels, 4 11 12 have demonstrated a major detrimental effect on HRQoL of PwMS. 13 Further, increased disability is known to have a direct impact on HRQoL. 9 13 However, women seem to report less impact of MS on HRQoL as compared with men. 9 Regardless, patients with MS have demonstrated improved results in their HRQoL when receiving support from their workplace, as well as social support. 14 15 Since employment is positively linked to the overall well-being of PwMS, 16 we were particularly interested in exploring workplace support alongside support from family and friends, hereby defined as social support. In addition, recent advancements in MS treatment, which focus on reducing disability progression, have demonstrated a direct positive impact on the quality of life for PwMS. 17 Accordingly, treatment satisfaction has been associated with higher HRQoL or an improvement in HRQoL. 18

To the best of our knowledge, no other study has explored HRQoL with the perceived support from different life dimensions (work, social and treatment) among working women with MS. Using data from the largest cross-sectional survey of patients with MS conducted in Sweden to date, we aim to explore the relationship of sociodemographic and clinical characteristics, and self-reported support from work, social support and treatment with HRQoL among aged-working women. Second, we aim to explore which factors best predict satisfaction with perceived support in any or all three life dimensions.

Methods

A cross-sectional study was conducted using a web-based survey answered by PwMS related to their health, work and perceived support, as well as individual-level linked register data.

Data sources and study population

Individuals listed in the Swedish MS Registry (SMSreg), 19 with ages between 20 and 50 years, and living in Sweden were invited to answer a web questionnaire between May and September 2021. The survey was sent out to 8458 PwMS, with a response rate of 52% (4412 participants), with a higher response rate among women (71.4%; n=3148), the sample included in this study. Further details of the study population are published elsewhere. 4 Statistics Sweden administered the survey and linked the responses to individual-level clinical data from the SMSreg and demographic data from Statistics Sweden. 20 The anonymised data was then delivered to the researchers.

Following, exclusion criteria were applied if: (1) reporting ‘no occupation’ (n=415), (2) not actively working (n=26) or (3) unavailable information on self-reported health (n=10). The final study population consisted of 2967 working women (see online supplemental figure 1 for further details for criteria application).

Register variables

The long integrated database for health insurance and labour market studies (LISA), held by Statistics Sweden, was used to obtain socio-demographic variables (sex, educational level, country of birth, marital status, living with children at home and type of living area). Age was calculated as the difference between the day of birth and the date of survey response. Additionally, the following clinical variables were obtained from the SMSreg:

Type of MS (relapsing-remitting, secondary-progressive, primary-progressive, missing).

Time since diagnosis, calculated as the difference, in years, between the date of MS diagnosis and the date of survey response.

MS severity was assessed with the most recently available Expanded Disability Status Scale (EDSS) score 12 and occurring within 3 years of the survey response (n=257 set to missing) and participants with secondary-progressive or primary-progressive MS with an EDSS score of 0 were set to missing (n=9). 21 The scores were then categorised (as mild (EDSS=0–2.5), moderate (3–5.5) and severe (6–9.5) or missing).

Cognitive processing speed assessed with the Symbol Digit Modalities Test (SDMT) and occurring within 3 years of the survey response (n=347 set to missing).

Ongoing treatment at the time of survey (yes, no).

Survey measures

Several measures were selected from the questionnaire. Specifically, closed-ended questions related to self-rated health (EQ-5D-3L and EuroQol Visual Analogue Scale (EQ-VAS)), fatigue, depression and anxiety subscales (Neuro-QoL), disclosure and perceived support. In addition, an open-ended question was included related to reporting MS restrictions. Responses from this open-ended question ‘Which MS symptom do you experience as the most limiting?’ informed on the most limiting symptom and were grouped into three categories following the European and American MS guidelines for symptom management, as reported elsewhere 4 21 :

No symptom .

Visible symptoms (movement/ambulation problems, vision problems, bladder, bowel and sexual dysfunction and dysphonia, dysarthria and dysphagia).

Invisible symptoms (fatigue and cognitive difficulties, sensory and pain, mood problems/depression, other—medication-related symptoms, unspecified symptoms or other psychosocial implications of MS symptoms).

Disclosure at work was explored with the following question: ‘Have you told your boss about your MS diagnosis?’ with four possible answers: Yes; No, but I will; No; Not applicable.

Further, perceived support from work or social support was investigated with the following statements: ‘I receive the support I would like from my closest boss’ or ‘I receive the support I would like from my friends and family’. Both questions could be answered with five options: Agree completely; Agree to some extent; Do not agree at all; Not applicable; Have not told.

Finally, participants were asked ‘What treatment/s do you currently have for your MS?’ followed by a closed-end question to explore their satisfaction with the current treatment: ‘Are you satisfied with your treatment?’ with three possible answers: Yes; Partly; No. However, only the latter question was used in the study, as information about ongoing treatment was based on the registered data (‘ongoing treatment’).

The web-based survey has been translated into English and is included as an Annex in the online supplemental material .

Outcome measure

The EQ-VAS was used to determine the self-rated health among the actively working women with MS. It derives from the EQ-5D, which assesses mobility, self-care, usual activities, pain/discomfort and anxiety/depression. The EQ-VAS form, however, is a vertical thermometer-like scale ranging from 0 (worst imaginable health) to 100 (best imaginable health), where respondents rate their overall health status. EQ-VAS scores have shown consistency with the severity of health states from the EQ-5D-3L. 22

Statistical analyses

Descriptive statistics was used to characterise the study population using frequencies and proportions for all categorical variables and mean and SD for continuous variables, after confirming normal distribution. Comparisons between categorical or continuous variables were performed using χ 2 and analysis of variances, respectively, using Benjamini–Hochberg correction for multiple comparisons.

Simple linear regressions were applied to estimate the association between all sociodemographic and clinical variables, as well as other answers from the survey (limiting symptoms and support from work, social and treatment), with the health-related quality of Life (EQ-VAS). All variables were examined to ensure they met the necessary assumptions for multiple linear regression analysis. Certain variables (depression and anxiety subscales) were excluded after introducing multicollinearity issues. Categorical variables with more than two levels were converted into dummy variables, using one of the levels as a reference (constant in the model). Missing data was not included in the analyses. Subsequently, variables which individually showed a significative association with HRQoL were then included in a multiple linear regression with a backward elimination method. Additional analyses (not shown) were performed to confirm that the selection of variables included in the model contained the highest proportion of variance (R 2 ) and adjusted variance (adj. R 2 ) considering the number of predictors in the model.

Finally, the likelihood of reporting having support from none, any or all dimensions (based on ‘completely agree’ or ‘yes’ responses to support from work, family and treatment satisfaction), was estimated using multinomial logistic regression. Results are reported as OR with 95% CI. Furthermore, an additional sensitivity analysis was conducted, which involved participants who had disclosed their condition to their boss and could thus be queried about receiving support from their workplace.

All statistical analyses were performed using SPSS .27 and a p value of <0.05 was considered statistically significant.

Patient and public involvement

None.

Results

The average age of the study participants was 40.1 ( table 1 ). The majority had a university educational level (68.9%), were born in Sweden (89.2%), were not married/not in partnership (53.5%), had children living at home (60.4%) and lived in cities or suburban areas (82.9%). Relapsing-remitting MS (RRMS) was the most common type of MS (94.4%). Most participants had an EDSS reflecting a low to mild MS severity (84.7%). Further, most women with MS were on treatment (90.8%) at the time of the survey, where 99.6% of these treatments were disease-modifying treatments (DMT). Available results for SDMT and fatigue scores reflected average scores, although the variance was wide-ranging ( table 1 ).

Table 1 Sociodemographic and clinical characteristics of all working women included in the study Sociodemographic characteristics Women with MS(n=2697) n (%) M (SD) Age  20–29 years 257 (9.5) 40.1 (8.1)  30–39 years 897 (33.3)  40–49 years 1384 (51.3)  50–51 years 159 (5.9) Educational level  Non-university 685 (31.1)  University 1518 (68.9) Country of birth  Sweden 2405 (89.2)  Other 292 (10.8) Marital status  Not married nor cohabitant/divorced/widowed 1444 (53.5)  Married/cohabitant 1253 (46.5) Children <18 at home  Yes 1628 (60.4)  No 1069 (39.6) Type of living area  Cities 1207 (44.8)  Suburbs/towns 1028 (38.1)  Rural areas 462 (17.1)
Clinical characteristics (register data)

n (%)

M (SD)
Type of MS  Relapsing remitting (RRMS) 2547 (94.4)  Secondary progressive (SPMS) 96 (3.6)  Primary progressive (PPMS) 31 (1.1)  Missing 23 (0.9) Time since diagnosis  0–4 years 746 (27.7) 8.8 (6.2)  5–9 years 731 (27.1)  10–14 years 554 (20.5)  15+ years 493 (18.3)  Missing 173 (6.4) MS severity (EDSS)  Mild (0–2,5) 1857 (84.7) 1.5 (1.4)  Moderate (3–5,5) 289 (13.1)  Severe (4–9) 49 (2.2) Cognitive processing speed (SDMT)  Available result 1928 (65.0) 58.1 (12.2)  (min.–max.) (11–110) Fatigue (Neuro-QoL)  Available result 2691 (90.7) 50.4 (95.1)  (min.–max.) (29.5–74.1) Ongoing treatment  Yes 2449 (90.8)  Yes (of which DMTs) 2439 (99.6)  No 248 (9.2) Note: some 50-year-old participants turned 51 by the time they conducted the survey. DMTsdisease modifying treatmentsM, meanmax.maximummin.minimumMS, multiple sclerosis; n, sample size

Sociodemographic and clinical characteristics of all working women included in the study

Note: some 50-year-old participants turned 51 by the time they conducted the survey.

DMTsdisease modifying treatmentsM, meanmax.maximummin.minimumMS, multiple sclerosis; n, sample size

Regarding the survey questions ( table 2 ), the majority reported invisible symptoms (63.1%) as the most limiting symptom, had disclosed to their closest boss (77.2%) and were satisfied with support from work (62.2%), social support (62.8%) and treatment (79.3%).

Table 2 Number and proportion of responses to open-ended and closed-ended survey questions All women with MS(n=2697) Most limiting symptom  No symptom 326 (12.1)  Visible symptom 363 (13.5)  Invisible symptom 1703 (63.1)  No response 305 (11.3) Disclosure: ‘Have you told your boss about your MS diagnosis?’  Yes 2082 (77.2)  No, but I will; No; Not applicable; No response 615 (22.8) * Support from work: ‘I receive the support I would like from my closest boss’  Agree completely 1294 (48.0)  Agree to some extent 550 (20.4)  Do not agree at all 214 (7.9)  No response 24 (0.9)  Missing (did not disclose to their boss) 615 (22.8) * Social support: ‘I receive the support I would like from my family and friends’  Agree completely 1694 (62.8)  Agree to some extent 759 (28.1)  Do not agree at all 100 (3.7)  Not applicable/Have not told 126 (4.7)  No response 18 (0.7) Treatment satisfaction: ‘Are you satisfied with your treatment?’  Yes 2140 (79.3)  Partly 456 (16.9)  No 55 (2.0)  No response 47 (1.7) Note: treatment satisfaction relates to the participant’s opinion regarding their treatment and should not be confused with ‘ongoing treatment’, which is data extracted from the clinical register regarding if the participant has a treatment or not. * Participants who responded not having disclosed to their boss or not applicable or missing, where not enquired if receiving support from work (thus, they are set as missing). Hence, from those enquired (n=2082): aAgree completely (62.2%), Agree to some extent (26.4%), Do not agree (10.3%), or nNo response (1.2%). MSmultiple sclerosis

Number and proportion of responses to open-ended and closed-ended survey questions

Note: treatment satisfaction relates to the participant’s opinion regarding their treatment and should not be confused with ‘ongoing treatment’, which is data extracted from the clinical register regarding if the participant has a treatment or not.

Participants who responded not having disclosed to their boss or not applicable or missing, where not enquired if receiving support from work (thus, they are set as missing). Hence, from those enquired (n=2082): aAgree completely (62.2%), Agree to some extent (26.4%), Do not agree (10.3%), or nNo response (1.2%).

MSmultiple sclerosis

Simple linear regressions between HRQoL and sociodemographic variables showed that only non-university level and living in cities were associated with lower and higher HRQoL, respectively ( online supplemental table 1 ). In addition, a worse clinical profile (progressive types of MS, reporting visible and invisible symptoms, higher EDSS, lower cognitive processing speed and higher fatigue) was associated with worse HRQoL. However, when all key factors were considered, the previous sociodemographic associations with HRQoL disappeared. The remaining results were lower MS severity and fatigue along with satisfaction with support from work had a positive association with HRQoL and were associated with higher HRQoL (see table 3 ). Whereas partly perceived social support had a negative association with HRQoL. This multiple regression model explained 66% of the variance in HRQoL, with an adjusted R 2 of 43%.

Table 3 Multiple regression association of health-related quality of life (HRQoL) with sociodemographic and clinical characteristics as well as self-reported health and perceived support responses from the survey (n=1309) Explanatory variable R Adj. R 2 Unstandardised coefficient Standardised coefficient T P value B SE Beta (Constant) 118.964 2.790 42.642
<0.001
University education 0.015 0.687 0.493 Living in cities 0.017 0.837 0.403 MS severity (EDSS) −2.359 0.253 −0.214 −9.318
<0.001
Cognitive processing speed (SDMT) 0.049 0.028 0.038 1.763 0.078 Fatigue (Neuro-QoL) −0.895 0.040 −0.516 −22.190
<0.001
Limiting symptoms  Visible symptom −1.891 1.026 −0.041 −1.843 0.066  Invisible symptom −0.013 −0.380 0.704 Support from work 0.437 0.434  Agree completely 3.878 0.804 0.115 4.824
<0.001
 Agree to some extent 0.030 0.271 0.787  Do not agree at all 2.371 1.248 0.044 1.900 0.058 Social support  Agree completely −0.002 −0.039 0.969  Agree to some extent −1.888 0.771 −0.053 −2.448
0.014
 Do not agree at all −0.016 −0.735 0.462 Treatment satisfaction  Partly −0.017 −0.785 0.433  No 0.006 0.288 0.774 Note: Predictive variables included in the model were selected according to a significant result from the simple regression models (Supp. Material, online supplemental table 1 Table 1). Backwards elimination method was performed in the multiple regression model with the results corresponding to the 9thninth model. The model was significant, F (7,196) = 144.029, pp<0.001, indicating more than one predictor significantly affects HRQoL. To avoid collinearity, the type of MS and time since diagnosis were excluded from this model owing to its association with EDSS (r=−0.308, pp<0.001; r=0.228, pp<0.001 and after posterior collinearity diagnostics >0.95). All levels of the categorical variables were dummied (0=no, 1=yes). References for categorical variables: limiting symptoms=No symptoms; Ssupport from work and Ssupport from family and friends=Not applicable or have not told; and treatment satisfaction variables=Yes. Bold values denote statistical significance at the p <0.05 level. EDSSExpanded Disability Status ScaleHRQoLhealth-related quality of LifeMSmultiple sclerosisSDMTSymbol Digit Modality Test

Multiple regression association of health-related quality of life (HRQoL) with sociodemographic and clinical characteristics as well as self-reported health and perceived support responses from the survey (n=1309)

Note: Predictive variables included in the model were selected according to a significant result from the simple regression models (Supp. Material, online supplemental table 1 Table 1). Backwards elimination method was performed in the multiple regression model with the results corresponding to the 9thninth model. The model was significant, F (7,196) = 144.029, pp<0.001, indicating more than one predictor significantly affects HRQoL. To avoid collinearity, the type of MS and time since diagnosis were excluded from this model owing to its association with EDSS (r=−0.308, pp<0.001; r=0.228, pp<0.001 and after posterior collinearity diagnostics >0.95). All levels of the categorical variables were dummied (0=no, 1=yes). References for categorical variables: limiting symptoms=No symptoms; Ssupport from work and Ssupport from family and friends=Not applicable or have not told; and treatment satisfaction variables=Yes. Bold values denote statistical significance at the p <0.05 level.

EDSSExpanded Disability Status ScaleHRQoLhealth-related quality of LifeMSmultiple sclerosisSDMTSymbol Digit Modality Test

Finally, a higher likelihood to report satisfaction with perceived support from work, social or treatment was associated with younger age, lower EDSS, higher cognitive processing speed and fewer MS restrictions (self-reported invisible and visible symptoms) ( table 4 ). However, when adjusted by age, educational level and all clinical variables, only lower fatigue scores and being on treatment were associated with satisfaction with support from one life dimension, compared with those who reported no support. Similarly, satisfaction with support from two and all three dimensions was associated with lower fatigue, being on treatment and non-university educational level. Importantly, higher HRQoL was associated with higher odds of responding by having support from all three dimensions (work, social and treatment) compared with those who had no support. Descriptive characteristics for multinomial regressions can be found in online supplemental table 2 . Additional sensitivity analysis repeating the adjusted multinomial logistic regression was performed for the subsample who had disclosed their MS at work with no changes in relation to previous results ( online supplemental table 3 ).

Table 4 Unadjusted and adjusted multinomial logistic regression analysis for satisfied support (‘completely agree’ or ‘yes’) from work, social and/or treatment dimensions Unadjusted OR (95% CI) Adjusted OR (95% CI) Support in one dimension Support in two dimensions Support in all three dimensions Support in one dimension Support in two dimensions Support in all three dimensions Age group  20–29 1.295 (0.650–2.580) 2.132(1.110–4.094 ) 2.131(1.097–4.138 ) 1.587 (0.527–4.780) 2.619 (0.896–7.659) 2.492 (0.839–7.404)  30–39 1.253 (0.876–1.792) 1.117 (0.789–1.583) 1.342 (0.942–1.913) 1.311 (0.775–2.217) 1.176 (0.699–1.979) 1.343 (0.789–2.288)  50+ 0.799 (0.408–1.564) 0.870 (0.459–1.649) 1.202 (0.634–2.280) 1.190 (0.416–3.403) 1.234 (0.441–3.455) 1.448 (0.501–4.186) Educational level  Non-university 1.289 (0.895–1.857) 1.418 (0.997–2.018 ) 1.549(1.083–2.216 ) 1.392 (0.803–2.414) 1.724(1.003–2.965 ) 2.062(1.185–3.588 ) Country of birth  Not Sweden 1.142 (0.695–1.876) 0.953 (0.587–1.550) 0.704 (0.423–1.171) Children <18 at home  No 0.729 (0.526–1.010) 0.882 (0.645–1.207) 0.876 (0.636–1.205) Marital status  Not married nor cohabitant 0.859 (0.622–1.187) 0.977 (0.715–1.334) 0.867 (0.631–1.191) Type of living area  Suburbs/towns 0.892 (0.626–1.271) 1.126 (0.801–1.583) 1.286 (0.908–1.820)  Rural areas 0.1098 (0.692–1.743) 1.084 (0.691–1.701) 1.327 (0.841–2.096)  MS severity (EDSS) 0. 851 (0.759–0.955 ) 0. 816 (0.730–0.913 ) 0766 (0.681–0.860 ) 0.935 (0.789–1.106) 1.045 (0.886–1.231) 1.026 (0.864–1.219)  Cognitive processing speed (SDMT) 1.010 (0.992–1.028) 1.023(1.005–1.041 ) 1.029(1.011–1.048 ) 0.999 (0.979–1.018) 1.012 (0.993–1.031) 1.013 (0.993–1.033)  Fatigue (Neuro-QoL) 0 . 952 (0.934–0.970 )
0.920(0.903–0.937)
0 . 900 (0.883–0.917 ) 0 . 961 (0.928–0.996 ) 0 . 930 (0.898–0.963 ) 0 . 919 (0.886–0.952 )  HRQoL (EQ-VAS) 1.023(1.014–1.032 ) 1.038(1.030–1.047 ) 1.056(1.046–1.066 ) 1.007 (0.990–1.024) 1.015 (0.999–1.032) 1.033(1.015–1.052 ) Ongoing treatment  No
0.483(0.310–0.752)

0.450(0.294–0.687)

0.305(0.192–0.485)

0.368(0.173–0.785)

0.356(0.170–0.743)

0.201(0.088–0.461)
Most limiting symptom  Visible symptom 0.399 (0.168–0.948) 0 . 344 (0.151–0.787 ) 0 . 274 (0.119–0.629 ) 0.506 (0.129–1.996) 0.515 (0.135–1.968) 0.513 (0.133–1.981)  Invisible symptom 0.507 (0.236–1.091) 0 . 286 (0.137–0.597 ) 0 . 233 (0.111–0.489 ) 0.786 (0.219–2.822) 0.710 (0.203–2.488) 0.769 (0.218–2.709) References (predictive variables): Aage (40–49 years), Eeducational level (University), Ccountry of birth (Sweden), Cchildren at home (yes); marital status (married or cohabitant), type of living area (cities), ongoing treatment (yes) and most limiting symptom (no symptom). Reference (dependent variable): Nno Ssupport. To note: sample sizes may variatey depending on available data, specifically for EDSS, SDMT and type of MS). Adjusted model samples size n=1689. Bold values denote statistical significance at the p <0.05 level. EDSSExpanded Disability Status ScaleEQ-VASEQ Visual Analogue ScaleMSmultiple sclerosisQoLquality of LifeSDMTSymbol Digit Modality Test

Unadjusted and adjusted multinomial logistic regression analysis for satisfied support (‘completely agree’ or ‘yes’) from work, social and/or treatment dimensions

References (predictive variables): Aage (40–49 years), Eeducational level (University), Ccountry of birth (Sweden), Cchildren at home (yes); marital status (married or cohabitant), type of living area (cities), ongoing treatment (yes) and most limiting symptom (no symptom).

Reference (dependent variable): Nno Ssupport.

To note: sample sizes may variatey depending on available data, specifically for EDSS, SDMT and type of MS). Adjusted model samples size n=1689. Bold values denote statistical significance at the p <0.05 level.

EDSSExpanded Disability Status ScaleEQ-VASEQ Visual Analogue ScaleMSmultiple sclerosisQoLquality of LifeSDMTSymbol Digit Modality Test

Discussion

In this study, we explore the association of several sociodemographic and clinical characteristics, self-reported restricting MS symptoms and perceived support from work, from social and treatment satisfaction with HRQoL among working women with MS. Further, we explore the likelihood of satisfaction with the perceived support in one or more life dimensions (work, social and treatment) among PwMS. Overall, our findings suggest that HRQoL among working women with MS can be individually determined by demographic and disease-related characteristics, self-reported restricting MS symptoms, as well as the perceived support from work, social support or treatment satisfaction. Nevertheless, when all factors are considered, clinical factors such as lower MS severity and fatigue scores, along with reported satisfaction with support from work, emerge as some of the most relevant factors explaining a higher HRQoL among working women with MS. Furthermore, a higher likelihood of reporting satisfaction with perceived support in all three dimensions (work, social or treatment) was associated with lower educational attainment, lower fatigue, having ongoing treatment and higher HRQoL.

Higher HRQoL is generally associated with better overall well-being and functioning. Consequently, individuals with a chronic and disabling condition such as MS, experience reduced HRQoL. 8 9 Numerous and overlapping factors influence HRQoL in PwMS. In our survey of PwMS, we explored important clinical variables and MS symptoms, especially fatigue, MS severity (focused on physical disability), cognitive processing speed and self-reported MS restrictions, as well as some domains of life affected by MS: support from work, social support, as well as treatment satisfaction. Additionally, factors such as having children at home or civil status were considered, as family composition has shown to have a greater impact on the work capacity of women with MS as compared with men. 5 Surprisingly, neither of them exerted a significant impact on HRQoL, contradicting earlier research indicating higher HRQoL among single women compared with married/cohabitant, separated or divorced. 23 However, it is somewhat in line with the observation of no differences in QoL found between mothers with MS and childless women with MS. 24 Differences in the study population and measures of HRQoL could explain these divergent findings. Moreover, it is important to note that further research is merited given the relatively scarce literature of family composition among women with MS and HRQoL.

Other important sociodemographic factors like education and type of living area exerted a notable impact on HRQoL but lost significance after accounting for all clinical factors and perceived support. For example, consistent with our findings, PwMS living in cities have higher HRQoL, possibly due to better accessibility of essential and clinical resources. 25 Furthermore, individuals with higher educational levels had higher HRQoL, which could potentially explain greater awareness of healthy behaviours or preventive measures which directly impact their HRQoL, but also fewer physical jobs, greater economic resources and more opportunities for activities that promote health. 26 Moreover, cognitive difficulties, commonly affected in PwMS, 27 are also a crucial component of overall well-being, influencing the individuals’ perceived HRQoL. Our findings indicate that higher cognitive processing speed levels are also associated with higher HRQoL, as previously reported. 28

Additionally, and consistent with other studies from our group, participants who reported having ‘no limiting symptoms’ had a higher HRQoL 4 and lower MS severity, 21 compared with those with visible or invisible symptoms. It is highly probable that this finding in our study is due to the fact that most participants were in the early stages of disease progression. While it is known that MS relapses contribute to the accumulation of disability, particularly at the early stages of the disease, the slow progression of disability independent of relapse activity has also shown to become a primary factor behind MS severity. 29 30 Whether this ‘smouldering’ (asymptomatic) or the relapse activity associated with disease worsening have different impacts on HRQoL is yet to be defined.

Moreover, our findings suggest that, ultimately, lower MS severity (EDSS) and fatigue, satisfaction with perceived support from work as well as certain satisfaction with social support, are among the factors that better predict higher levels of HRQoL. The association between fatigue and/or MS severity with HRQoL has been repeatedly reported, indicating that lower levels of both measures are related to higher HRQoL or vice versa. 2 12 13 31 Similarly, the impact of support from work on HRQoL has been described among chronic patients, where insufficient support from either supervisors or co-workers was associated with lower HRQoL. 32 We also observed that having no or unsatisfactory support either from work or social was associated with lower levels of HRQoL, consistent with the literature. 14 15 Such findings underscore the importance of a supportive work environment and significant others (family, friends or acquaintances) in promoting a higher quality of life for individuals within a wide grade of restrictions.

Importantly, most of the women with MS included in the study had a low to mild level of MS and reported invisible symptoms as the most important MS limitation. The majority also still responded being satisfied or partly satisfied when asked if they receive the support they would like from their closest boss (at least among those that had disclosed). While receiving support from work does not necessarily imply that support is required or sufficient, the notion of perceived support itself still appears to carry significance when coping with MS disease at the workplace. 33 Altogether, this could be interpreted as perceived support from work being a relevant factor among working women with MS, regardless of the severity of their MS and the ‘evidence’ of the disease’s restrictions. In addition, the impact of the ‘invisible limitations’ of MS, mainly related to fatigue symptoms, cannot be disregarded, as it has previously been shown to affect various aspects of work and private aspects of life (family, leisure activities and contact with friends/acquaintances). 4 This also accounts for its impact on HRQoL. 6 Our findings indicated how fatigue, presence of invisible symptoms and cognitive processing speed were associated with HRQoL. Furthermore, the strong and negative association between fatigue and HRQoL prevailed when all other clinical factors were considered. This finding reinforces the idea that fatigue converges with the overall sense of well-being among PwMS, 31 as a prevalent symptom in PwMS 2 and work capacity. 6

Following, we investigated the factors that influence reporting satisfaction in the perceived support of work-life dimensions and treatment. Our findings suggested that satisfaction in all three dimensions was attained in the presence of lower fatigue, lower educational level, ongoing treatment and higher HRQoL. Findings regarding treatment satisfaction are most likely related to the use of DMTs, given that the majority of women in this study had an ongoing DMT prescription during the study period, even at the very early stages of the disease. Hence, early treatment with DMT may potentially uphold a higher HRQoL for PwMS, as it has been shown to delay disability accumulation by years, with the most significant benefit seen in individuals of younger age and relapsing-remitting MS. 29 However, treatment efficacy and side effects, both unexplored factors in this study, have also shown a significant impact on HRQoL. 34

Additionally, we unexpectedly found that the likelihood of experiencing satisfactory support in two or three dimensions was higher among those with lower educational levels. In this context, we hypothesise that this association could be linked to lower-skilled and potentially more physically demanding occupations, where the limitations of MS might be more noticeable and result in a greater need for work adjustments and support. 21 Another possibility is that women with higher educational levels are also more cognizant of the array of resources and support they should ideally receive, thereby being less satisfied with what is currently provided. Initially, these results could reflect a younger and ‘healthier’ MS profile as shown by the unadjusted results. However, after adjusting for age and educational level our findings still sustained, highlighting the relevance of overall well-being and perceived support. While the direction of this association cannot be unravelled, this finding underscores the need for exploring the patient’s work and life circumstances when assessing their HRQoL.

This study has several strengths. It includes a large population-based sample of working-aged women with MS, and their survey responses were subsequently linked with high-quality register data. The web-based survey had a response rate of 71.4% among women, which could be attributed to its convenient mode of administration, enabling participants to complete it at their own pace and in their preferred environment and possessing significant representativeness. However, there are limitations to consider. The findings in this study are linked to the evolving circumstances of the pandemic. Initially, the Swedish National Board of Health and Welfare classified patients with MS as high-risk for COVID-19, but this was later revised to include only those with severe motor impairments affecting respiratory function. These changes may have affected vaccination order for PwMS as well as their exposure to others (at work, healthcare providers, family and friends), due to distancing measures. Further, it is possible that a bias among survey respondents is present, as non-respondents were likely more unwell and may have been unable to cope with answering the survey. Additionally, the large proportion of participants with mild disability (EDSS<3), the age cut-off of 50 years, and the focus on actively working individuals could lead to a representation of a ‘healthier’ MS population, underrepresenting those with more severe illness. Consequently, the findings related to factors influencing perceived HRQoL may be more relevant to individuals with mild MS and may not fully capture the experiences of those with more advanced disease. Nevertheless, it is important to investigate outcomes in early MS that may help shared decision making between patients and physicians (eg, treatment possibilities and alternative approaches, pros and cons of disclosure at the workplace, etc.), which ultimately may lead to enhanced HRQoL. Moreover, the majority of women with MS in this study were currently on medication, while a smaller proportion had no treatment, which was indicative of progressive types of MS. In fewer instances, this included women who were either pregnant or planning to be (data not presented). While the inclusion of ongoing treatment in the multinomial logistic regression could be argued (as considered a dimension of support and thus, implying circularity), we opted to include this factor as the presence of treatment does not necessarily guarantee adherence, 35 let alone treatment satisfaction. Other limitations to consider involve the cross-sectional design, which reduces our ability to establish causality between all variables and HRQoL, as well as limiting the insights into temporal relationships. Nevertheless, the exploratory nature of this study suggests the potential for further research in HRQoL of PwMS using a more comprehensive approach that goes beyond the clinical perspective.

Conclusion

HRQoL is a multidimensional concept that can be associated with disease-related characteristics and perceived support from both work and personal life. The interplay between fatigue levels, MS severity and satisfaction with support from the workplace and significant others, can have a considerable impact on the overall well-being of PwMS. However, despite facing severe symptoms, women with MS have shown resilience, maintaining a high level of HRQoL. Implications of these findings could be useful for targeting individuals at risk of a decline in HRQoL: for patients with MS to seek help when fatigue or other invisible MS symptoms appear, and to cultivate a supportive social network and healthy work environment; for healthcare practitioners to appraise beyond the evident MS signs and to inform their patients about the importance of achieving a healthy work-life balance; and for employers to create environments that foster support and well-being among their employees. Furthermore, support for individuals with chronic illnesses could be strengthened by raising community awareness about invisible symptoms and lesser-known aspects of the disease, as well as its effects both in and outside the workplace.

supplementary material