Multiple sclerosis (MS) is a chronic, debilitating neurological condition that affects an estimated two million people worldwide. Despite decades of research, the aetiology of MS and factors that affect disease progression and relapse rate are still debated [
Several large national MS registries collect longitudinal data on patient outcomes and measure the effectiveness of a range of therapies [
We aimed to recruit followers of online media engaging people with MS and examine in detail their health and lifestyle behaviours and determine the relationship of these factors to self-reported disability, disease activity, and quality of life, with a follow-up period of five years. For many people with chronic diseases, the internet is an important tool for self-education, emotional support, and practical advice from people facing similar challenges [
This study provides a snapshot of current lifestyle and risk-modifying behaviours of a large international group of people with MS, as well as an ongoing platform for analysing the association between these variables and disease progression, that have not previously been examined in detail. They will also help to inform future research, and people with MS, of the potential contribution of lifestyle to their health-related quality of life, disease activity, and physical disability. This paper reports in detail the methodology of the study and an overview of the characteristics of participants recruited. Future studies will seek to analyse and report associations between lifestyle variables and disease data in detail.
This survey, collecting the baseline data for the health outcomes and lifestyle interventions in a sample of people with multiple sclerosis (HOLISM) study, was conducted using the online software SurveyMonkey. A webpage was created, inviting people to take part in the study, with a description of the study aims. Online recruitment for the study took place over a 15-week period, using websites, a mailing list, and web 2.0 platforms such as blogs, forums, Facebook, and Twitter. The principle investigator (GJ) had developed an online presence in the MS community over the 12 years prior to the study due to his extensive work advocating for lifestyle modification and integrated MS management, including moderating a website dedicated to this field (
The survey webpage linked individuals to a participant information sheet, which they were asked to read before giving consent. This was required for continued participation in the survey. Anyone formally diagnosed with MS by a medical doctor was encouraged to take part, but participants were excluded if they were under 18 years of age. This was verified in the survey: if the year of birth selected indicated that they were under 18 years of age, skip logic took the participant to the end of the survey. All participants were required to complete their contact details to facilitate followup. Data were stored in a reidentifiable form, and data security measures were undertaken to ensure that only members of the research team had access to participant information. The research team was available to answer participants’ questions by phone and email. Ethics approval was granted by St Vincent’s Hospital Melbourne Human Research Ethics Committee (LRR 055/12).
The survey consisted of a maximum of 163 questions, taking around 40 minutes to complete, although skip logic enabled participants to avoid some questions not relevant to them. Three members of the research team searched and reviewed validated tools for use in the survey. Where possible, a tool was chosen that was psychometrically sound and had been tested in a similar study (Table
Summary of validated tools used.
Outcome variable | Instrument (reference) | Number of items | Authors (reference) |
---|---|---|---|
Disability | Patient determined disease steps (PDDS) | 1 | Hohol et al., 1995 [ |
Comorbidities | Self-administered comorbidity questionnaire (SCQ) | 13 | Sangha et al., 2003 [ |
Health-related quality of life | Multiple sclerosis quality of life-54 (MSQOL-54) | 54 | Vickrey et al., 1995 [ |
Dietary habits | Diet habits questionnaire (DHQ), modified | 20 | McKellar et al., 2008 [ |
Physical activity | International physical activity questionnaire (IPAQ) | 7 | Craig et al., 2003 [ |
Social support | Single item measure of social support (SIMSS) | 1 | Blake and McKay, 1986 [ |
Fatigue | Fatigue severity scale (FSS) | 9 | Krupp et al., 1989 [ |
Depression | Patient health questionnaire short version (PHQ-2) | 2 | Kroenke et al., 2003 [ |
The survey consisted of the following domains.
The statistical package for the social sciences (SPSS) version 20.0 was used to calculate statistics. Univariate analyses were performed and continuous data reported using mean (95% CI) or median (IQR) and categorical data using number and percentage. Due to variation in item completion, analyses were calculated using item response as the denominator. Where possible, summary scores from validated tools were derived according to scoring instructions or as suggested in the literature. The following are explanations of how new variables and summary scores were derived for the purpose of this paper.
A planned variable, “disease activity,” was derived from data reporting relapse rates (doctor diagnosed) for those with relapsing-remitting MS only. It was categorised as increasing, decreasing, or stable, where relapse rate in the preceding 12 months was higher, lower, or the same, respectively, as the 5-year annualised relapse rate. For the purpose of this paper, conditions listed in the SCQ were summed to determine the proportion of participants that had one or more comorbidity. “Other” free-text responses will be categorised and reported in a future study along with reporting of the treatment and limitations relevant to each condition.
The MSQOL-54 was scored according to the scoring instructions with a set number of items required to be completed in order to give rise to the subscores, which in turn were required for calculation of the composite scores; hence, there was variability in the completion rates.
Cutoff scores were set as an aggregate score ≥ 3 in the PHQ-2 to screen positive for major depression, and a mean score ≥ 4 in the FSS to indicate clinically significant fatigue, as defined in the literature. To derive both of these summary scores, full item completion was required.
Based on recency of licensing and mode of action, the medications were grouped into seven categories (Table
Classification of MS medication types.
First generation disease modifying drugs | Interferons |
Glatiramer acetate | |
| |
Second generation disease modifying drugs | Alemtuzumab |
Cladribine | |
Daclizumab | |
Dimethyl fumarate | |
Fingolimod | |
Laquinimod | |
Rituximab | |
Teriflunomide | |
Natalizumab | |
| |
Chemotherapy or immunosuppressants | Azathioprine |
Cyclophosphamide | |
Methotrexate | |
Mitoxantrone | |
Mycophenolate mofetil | |
| |
IVIG or plasmapheresis | Immunoglobulins IVIG |
Plasmapheresis | |
| |
Generic drugs | Low-dose naltrexone |
Minocycline | |
| |
Steroids | Adrenocorticotropic hormone |
Prednisolone | |
| |
Symptom modifying drugs | Baclofen |
Fampridine |
A total of 3053 participants consented to participate, seven of whom were pilot study participants, whose data were included in analysis. Of those that consented, 2519 had been formally diagnosed with MS and therefore met the criteria for study inclusion. 2518 of these provided contact details for followup (Figure
Participant study inclusion criteria.
The majority of respondents were women, comprising 82.2% of the sample (Table
Characteristics of the study sample.
Characteristic | Number (%) |
---|---|
Gender | |
Male | 413/2318 (17.8) |
Female | 1905/2318 (82.2) |
Age in 2012 (years) | |
18–29 | 124/2443 (5.1) |
30–39 | 621/2443 (25.4) |
40–49 | 794/2443 (32.5) |
50–59 | 656/2443 (26.9) |
60–69 | 231/2443 (9.5) |
>70 | 17/2443 (0.7) |
Country of location | |
USA | 827/2518 (32.9) |
Australia | 649/2518 (25.8) |
UK | 416/2518 (16.5) |
NZ | 216/2518 (8.6) |
Canada | 107/2518 (4.2) |
Other* | 303/2518 (12.2) |
Country of birth | |
USA | 790/2510 (31.5) |
Australia | 518/2510 (20.6) |
UK | 502/2510 (20.0) |
NZ | 174/2510 (6.9) |
Canada | 111/2510 (4.4) |
Other† | 415/2510 (16.5) |
Marital status | |
Married | 1511/2475 (61.1) |
Single | 359/2475 (14.5) |
Cohabitating/partnered | 313/2475 (12.6) |
Separated/divorced | 262/2475 (10.6) |
Widowed | 30/2475 (1.2) |
Family status | |
No children | 767/2477 (31.0) |
One or more biological children | 1644/2466 (66.7) |
One or more step children | 239/2466 (9.7) |
No. of children |
2 (0–2) |
Employment status | |
Employed full time | 807/2508 (32.2) |
Employed part time | 526/2508 (21.0) |
Stay at home parent/carer | 191/2508 (7.6) |
Student full time | 57/2508 (2.3) |
Unemployed‡ | 203/2508 (8.1) |
Retired due to age | 79/2508 (3.1) |
Retired due to medical reasons or disability | 588/2508 (23.4) |
Other | 57/2508 (2.3) |
Education status | |
No formal schooling | 3/2504 (0.1) |
Primary school | 55/2504 (2.2) |
Secondary school | 570/2504 (22.8) |
Vocational training | 405/2504 (16.2) |
Bachelor’s degree | 897/2504 (35.8) |
Postgraduate degree | 574/2504 (22.9) |
†Includes 52 other countries.
‡Collapsed from unemployed: seeking work/not seeking work.
A median of 13 years (IQR 7–21) had passed since participants first experienced symptoms of MS, whilst participants had formally been diagnosed with MS for a median of six years (IQR 3–12), with 45.0% having being diagnosed within the previous five years (Table
Diagnostic characteristics and level of disability.
Number (%) | |
---|---|
Number of years since diagnosis | |
<1–5 | 1107/2459 (45.0) |
6–10 | 577/2459 (23.5) |
11–15 | 393/2459 (16.0) |
16–20 | 186/2459 (7.6) |
21–25 | 98/2459 (4.0) |
26–30 | 48/2459 (2.0) |
>30 | 50/2459 (2.0) |
Disability (PDDS) | |
Normal | 734/2314 (31.7) |
Mild disability | 352/2314 (15.2) |
Moderate disability | 170/2314 (7.3) |
Gait disability | 373/2314 (16.1) |
Early cane | 265/2314 (11.5) |
Late cane | 175/2314 (7.6) |
Bilateral support | 139/2314 (6.0) |
Wheelchair/scooter | 103/2314 (4.5) |
Bedridden | 3/2314 (0.1) |
Currently experiencing symptoms due to a recent relapse* | |
Yes | 643/2429 (26.5) |
No | 1455/2429 (59.9) |
Unsure | 331/2429 (13.6) |
Diagnosed subtype of MS.
Benign | Relapsing-remitting | Primary progressive | Secondary progressive | Progressive relapsing | Unsure/other | |
---|---|---|---|---|---|---|
Type of MS first diagnosed with number (%) | 99/2455 (4.0) | 1880/2455 (76.6) | 157/2455 (6.4) | 47/2455 (1.9) | 14/2455 (0.6) | 258/2455 (10.5) |
Type of MS currently diagnosed with number (%) | 101/2460 (4.1) | 1500/2460 (61.0) | 184/2460 (7.5) | 287/2460 (11.7) | 47/2460 (1.9) | 341/2460 (13.9) |
The level of disability, as measured with the PDDS, was across all disability spectrums, but 31.7% reported having no symptoms or mild symptoms that return to normal after an attack: “normal” (Table
Over the previous 12 months, relapsing-remitting participants self-reported an average of 1.09 relapses, and over the last five years, they self-reported an average of 0.97 relapses per year (Table
Relapse rate for relapsing-remitting participants.
Self-diagnosed relapse rate | Doctor-diagnosed relapse rate | |||||
---|---|---|---|---|---|---|
|
Mean (95% CI) | Median (IQR) |
|
Mean (95% CI) | Median (IQR) | |
Number of relapses over last 12 months* | 1475/1500 | 1.09 (1.01–1.17) | 1 (0.0–2.0) | 1456/1500 | 0.73 (0.67–0.78) | 0 (0.0–0.1) |
Number of relapses over last 5 years, annualised* | 1409/1500 | 0.97 (0.92–1.03) | 0.67 (0.4–1.2) | 1399/1500 | 0.66 (0.62–0.70) | 0.5 (0.2–1.0) |
Median summary scores from the MSQOL-54 were 68.4 (IQR 55.0–81.65) for the overall quality of life subscore, 59.2 (IQR 42.3–77.2) for the physical health composite, and 72.0 (IQR 51.2–84.1) for the mental health composite (Table
MSQOL summary scores.
Summary score |
|
Distribution (Kolmogorov) | Mean (95% CI) | Median (IQR) |
---|---|---|---|---|
Overall quality of life subscore | 2275 (244) | 0.00 | 66.9 (66.1–67.7) | 68.4 (55.0–81.7) |
Physical health composite | 1944 (575) | 0.00 | 59.1 (58.1–60.0) | 59.2 (42.3–77.2) |
Mental health composite | 2222 (297) | 0.00 | 66.7 (65.8–67.6) | 72.0 (51.2–84.1) |
Depression and fatigue screening.
Cutoff | Positive, number (%, 95% CI) | Negative, number (%, 95% CI) | |
---|---|---|---|
Depression screen (PHQ-2) | Negative < 3; positive ≥ 3 | 431/2231 (19.3, 17.7–21.0) | 1800/2231 (80.7, 79.0–82.3) |
Fatigue (FSS) mean score | Negative < 4; positive ≥ 4 | 1408/2143 (65.7, 63.7–67.7) | 735/2143 (34.3, 32.3–36.3) |
The following variables are a snapshot of participant responses and not comprehensive of all variables examined; these will be examined in greater detail in subsequent papers. 68.4% of participants reported having one or more of the listed comorbidities in the SCQ. At the time of the survey, 51.4% of respondents were taking a first or second generation DMD. According to the single item measure of social support, the majority (59.5%) had 2–5 people in their life that they could count on in times of difficulty, but 5.3% of the sample had no one. 38.0% of respondents did not consume dairy products, 26.7% did not consume meat products, and 21.5% consumed neither meat nor dairy. 11.7% were current smokers. Over two-thirds (67.0%) intentionally exposed themselves to the sun to try to raise their vitamin D levels, and 82.3% took vitamin D supplements. Nearly two-thirds of respondents (64.3%) took omega-3 supplements. Meditation practice was undertaken at least once per week by 676 respondents (30.0%).
This is the first cross-sectional study examining health and lifestyle behaviours in a large international sample of people with MS using web 2.0 platforms. Our results suggest that the online MS community is a unique sample to study. The proportion of women that participated is an overrepresentation from the estimated incidence of 1.8 female cases for every male [
In general, participants of this study had been diagnosed recently and had a low level of disability. This concurs with results from a self-enrolling online registry—NARCOMS—that reported one-third of their participants enrolled within two years of diagnosis and also had a median disability score of 3 (1–5) on the PDDS [
Comparison of data with NARCOMS and MSBase registries.
HOLISM | NARCOMS [ |
MSBase [ | |
---|---|---|---|
Female (%) | 82.2 | 74.9 | 71.5 |
Age at diagnosis (years) | 37.0 (median) | 36.9 (mean) | 32.2 (mean) |
Age at symptom onset (years) | 31.0 (median) | 32.4 (mean) | Data not provided |
Age at enrolment (years) | 46.0 (median) | 42.8 (mean) | 42.7 (mean) |
Disease duration at enrolment (years) | 6.0 (median) | Data not provided | 10.4 (mean) |
Country of residence.
Country |
|
---|---|
Armenia | 1 |
Australia | 649 |
Austria | 2 |
Belgium | 5 |
Brazil | 7 |
Bulgaria | 1 |
Canada | 107 |
China | 1 |
Croatia | 7 |
Cyprus | 1 |
Czech Republic | 2 |
Denmark | 10 |
Estonia | 2 |
Finland | 4 |
France | 11 |
Gibraltar | 2 |
Guam | 1 |
Germany | 29 |
Greece | 12 |
Iceland | 2 |
India | 3 |
Indonesia | 1 |
Iran | 1 |
Ireland | 36 |
Israel | 2 |
Italy | 6 |
Kuwait | 1 |
Lebanon | 1 |
Luxembourg | 1 |
Malta | 2 |
Mexico | 3 |
Namibia | 1 |
Netherlands | 28 |
New Zealand | 216 |
Norway | 13 |
Philippines | 2 |
Poland | 2 |
Portugal | 6 |
Puerto Rico | 3 |
Qatar | 2 |
Romania | 4 |
Russian Federation | 2 |
Saudi Arabia | 2 |
Serbia | 2 |
Singapore | 2 |
Slovakia | 4 |
Slovenia | 2 |
South Africa | 30 |
Spain | 9 |
Sweden | 17 |
Switzerland | 8 |
Syria | 1 |
Trinidad and Tobago | 1 |
Turkey | 1 |
United Arab Emirates | 4 |
United Kingdom | 416 |
United States | 827 |
It has been noted that clinical outcome measures of relapse rate and disability are insufficient alone to measure the impact of MS as they do not reflect patients’ experiences of the disease [
Wide-ranging health and lifestyle behaviours have not been studied extensively among people with MS [
The MS sample described appear to be a highly engaged and proactive group of patients. This is evidenced by the method of recruitment which was self-selecting from patients utilising online resources. The preliminary data suggests that a significant number in this sample have adopted lifestyle changes, much of which they would have learned about through self-directed learning beyond the clinical setting, and which demands a great deal of self-efficacy and commitment. A surprisingly large proportion of people continued to the end of the survey despite its considerable length. This was likely aided by the fact that participants could exit and return to the survey at their leisure. The large sample size and high item completion rate might indicate that the online MS community is motivated to contribute to novel research. Much of the feedback the research team received through the survey, emails, phone calls, and online comments was very positive. Participants expressed a desire for more research exploring lifestyle factors. Personal empowerment is key to successful adoption of healthful practices; people with MS who show increased levels of activation or self-efficacy also demonstrate positive changes in their self-management behaviours [
Recruiting participants through web 2.0 platforms results in rapid access to a heterogeneous sample in a highly cost-effective way and may invite contribution from previously hidden populations [
Although participants in 57 countries took part in the survey, the majority were residing in western countries where English is the primary language. This is likely a direct reflection of the sources of recruitment used by the researchers, which were mostly limited to English language, while the survey itself demanded a high level of literacy. Severe physical disability may have prevented some people with MS from taking part as they may not have been able to complete the survey without assistance. The fact that some participants were directly recruited through a website and associated forums promoting lifestyle modification may have resulted in the participation of individuals with an interest in, and more inclined to undertake, holistic disease management. As such, the findings from this study may not be generalisable to the global online MS community.
When making a decision about which items to include in the survey, serious consideration was given to the time burden placed on participants, given that the survey intended to cover many different domains. Although validated tools were used where possible, and researcher-devised items were carefully constructed, with the survey piloted with a small group, there are a number of limitations to the survey. Firstly, the validated tools have not all been tested for validity and reliability across diverse MS samples. Different cultural understandings of health mean that the use of western terminology and concepts of health with a culturally diverse sample may have allowed bias. It is also unclear whether these validated tools have previously been tested in an online survey format. It is also possible that the study was more likely to select participants with a higher level of socioeconomic status due to the method of recruiting through web 2.0 platforms. Socioeconomic status of participants was not measured in the questionnaire. This is because of the international nature of the survey and difficulty correlating data across regions.
A significant proportion of participants reported experiencing symptoms related to a recent relapse. It may be difficult for patients to distinguish between enduring relapse symptoms and a permanent progression of disability. This has the potential to impact other outcome variables, and future analyses will need account for this. All data were self-reported, which can lead to potential sources of bias due to over- or underreporting. This needs to be considered particularly in light of historic reporting of events which can be limited by poor recall, such as the number of relapses in the preceding five years, or poor knowledge, such as the number of lesions or vitamin D level. We were not able to verify the accuracy of these responses. Validation through medical records or physician’s report would significantly increase the reliability of self-reported data. Consideration will be given to the limitations of data reliant on self-report that may be used in future analyses. Despite these limitations, some self-reported data in MS study participants have previously been found to be a reliable measure of outcomes [
Increasingly, web 2.0 platforms are being embraced by patients and researchers alike, with the shared goal of improving patient-centred health outcomes. The characteristics described above suggest this online MS community recruited to our study is a heterogeneous and unique group who appear actively engaged and keen to contribute to research on health and lifestyle behaviour. Although there are several limitations to conducting research with online communities, this method provides efficient and rapid access to a large sample of people with MS. This study provides a sound platform to undertake exploratory analyses of lifestyle variables that have not previously been examined in such detail. The dataset will provide an important opportunity for novel research into lifestyle and health behaviours and their potential impact on MS outcomes over time.
The authors thank all the participants for taking part in the survey and the Bloom Foundation for funding the project.