Abstract
Objective: To compare the Expanded Disability Status Scale (EDSS) and self-rated quality of life scores (SF-36 Health Survey) as measures of disease impact in a representative sample of MS patients.
Background: The EDSS is the most common outcome measure of impairment/disability for MS patients but is heavily weighted toward mobility. Sensitive outcome measures are needed that also capture other aspects of the effects of MS.
Methods: The authors performed a cross-sectional study of the cohort of all individuals with onset of MS between 1976 and 1986 who were diagnosed before 1995 in Hordaland County, Norway. A total of 194 patients (94%) participated.
Results: The patients had lower mean scores for all eight SF-36 health dimensions compared with sex- and age-adjusted scores in a general population. EDSS scores correlated highly with physical functioning (r = −0.86, R2 = 0.73), and explained some of the variation in social functioning (r = −0.48, R2 = 0.23) and general health (r = −0.46, R2 = 0.21) but little for the other dimensions.
Conclusions: The SF-36 captures the broad effects of MS, and the results showed that patients also are bothered frequently with health problems such as bodily pain and low vitality. These problems, which are not reflected in the Expanded Disability Status Scale, should be given more attention in the treatment of MS and when evaluating interventions.
MS is a chronic neurologic disease with onset primarily between the ages of 20 and 45 years. MS produces various symptoms and signs and may lead to severe disability.1,2 The Expanded Disability Status Scale (EDSS)3 is the most common measure of impairment/disability for MS patients and of outcome in clinical trials. Nevertheless, the EDSS has been criticized for overemphasizing mobility and for not capturing all the elements involved in the global impact of MS.4-6 It is also criticized for poor reliability and nonlinearity.7-12
Quality of life measures are alternative indicators of the impact of the disease. Assessing quality of life focuses more attention on MS patients as a whole, in addition to focusing on physical problems.13,14 Health care workers often underestimate the difficulty people have with activities of daily living and basic social routines.15,16
The primary goal of this study was to describe the self-assessed burden of MS using a structured quality of life instrument in a representative sample and to compare these scores with those in a general population. The second goal was to estimate the extent to which variation in EDSS scores and functional systems (FS) scores can explain the variation in quality of life scores. Two previous, small, hospital-based studies17,18 have shown no significant correlation between EDSS scores and quality of life scores except for physical functioning. We wanted to estimate these correlations in a population-based sample.
Methods.
Design and patients.
The study was cross-sectional, including the cohort of all patients with onset of MS between 1976 and 1986, and diagnosed before 1995 in Hordaland County in western Norway. Several prevalence and incidence studies have been performed in this county.19-21 All potential MS patients in the county are referred to the Department of Neurology, University Hospital of Bergen, and the cohort therefore probably includes almost all patients with onset of MS during this time period in the county. The cohort comprised 218 patients; 207 were still alive in 1995, and 194 (94%) agreed to participate in a neurologic examination and a questionnaire including a quality of life measure. All patients were examined in 1995 and had had MS for 9 to 19 years. The patients were scored using Kurtzke’s EDSS—a neurologist-rated scale that grades impairment/disability caused by MS on a continuum of 0 (normal neurologic examination) to 10 (death due to MS) in 20 steps.3 EDSS scoring is based on the FS, consisting of mental, visual, brainstem, sensory, pyramidal, cerebellar, and bowel/bladder functions. Table 1 shows the demographic and clinical characteristics of the study population.
Clinical and demographic characteristics in 1995 of 194 MS patients with disease onset between 1976 and 1986 in Hordaland, Norway
Quality of life assessment.
The patients assessed their quality of life using the SF-36 Health Survey23—one of the most commonly used measures of quality of life. The questionnaire was developed empirically by including items that have been evaluated by patients to include important aspects of their quality of life. It assesses health concepts that represent basic human values that are relevant to everyone’s functional status and well-being, also called health-related quality of life.23 It measures primarily the functional effect of an illness as perceived by the patient. SF-36 has been validated and tested for reliability in several studies.23-25 It has been shown to be sensitive to early changes in MS5 and to adverse events in a treatment trial on interferon-α.26 SF-36 includes eight multi-item scales of functioning and well-being that represent physical and mental health status. The physical functioning scale (10 items) assesses limitations in activities ranging from self-care to vigorous activities; the general health scale (five items) evaluates global health issues such as feeling well or ill. The mental health scale (five items) assesses general mood or affect, including depressive symptoms, anxiety, and positive well-being. The bodily pain scale (two items) determines the frequency of pain and how it interferes with daily activities. The role limitation scales measure the extent to which poor physical health (four items) or emotional problems (three items) interfere with regular daily activities. The social functioning scale (two items) assesses the extent to which health interferes with social activities with family and friends, and the vitality scale (four items) assesses perceived energy levels. The standard version (SF-36 1.0) was used, and the focus was the situation during the past 4 weeks. All items scores were coded, summed, and transformed into a scale of 0 (poor health) to 100 (optimal health). Missing substitution was performed to calculate the score for dimensions when less than 50% of the questions were unanswered.23 This was performed for 2.6% of the dimension scores.
Statistical analysis.
The means and 95% CIs for each SF-36 health dimension were compared with the age- and sex-adjusted scores in a general population.27 These adjusted scores were calculated by finding the score for each patient in the general population of the same sex and age group in 5-year age groups.
The SD differs substantially between the health dimensions of SF-36,23 and thus the difference between the mean scores of the MS population and the age- and sex-adjusted scores in the general population for each of the dimensions can not be compared directly. The difference is illustrated by calculating the difference as a factor of the SD in the general population. For example, if the general population mean is 80 points with an SD of 10, a patient score of 65 would be considered to fall 1.5 SDs below the mean of the general population.
The association between EDSS score and the SF-36 health dimensions was investigated by comparing the quality of life scores in three severity groups based on their EDSS score (≤2.5, 3.0 to 6.0, and ≥6.5), and differences were tested by t-tests. The association was also estimated using Pearson’s correlation coefficient. The association between the various FS and SF-36 health dimensions was investigated by multiple regression analysis.
The study was approved by the research ethics committee of health region 3, Norway.
Results.
The patients showed markedly and significantly lower mean scores for all health dimensions of the SF-36 compared with the age- and sex-adjusted scores in a general population of 2,323 Norwegians (figure 1). 27 This difference was especially high for physical functioning; general health; role limitation, physical; vitality, and social functioning, all of which showed mean scores among the MS population equivalent to more than 1 SD below that of the general population (table 2).
Figure 1. Distribution of the eight SF-36 dimensions among the MS patients (hatched bars) compared with age- and sex-adjusted scores in a general population (gray bars) in Norway. Mean and 95% CIs are shown. Phys.func. = physical functioning; Role-phys. = physical role; B.pain = bodily pain; Soc.func. = social functioning; Role-emot. = emotional role; Mental h. = mental health.
Mean score and difference from a general population expressed as a factor of the SD in the general population for the eight SF-36 health dimensions in three Expanded Disability Status Scale (EDSS) severity groups
The EDSS scores, ranging from 0 to 9.5, showed a bimodal distribution also found in other population-based studies.12 The mean score was 4.1, and 64% of the patients scored 4 or lower. The patients were categorized in three EDSS severity groups. The patients with low EDSS scores (≤2.5) had a lower mean score than the general population in all health dimensions except for mental health (see table 2). These patients scored significantly better than the two groups with higher EDSS scores in all dimensions (all, p < 0.005; t-test), whereas there was less difference between the groups with EDSS scores between 3.0 and 6.0 and ≥6.0. The patients with EDSS scores of 3.0 to 6.0 had markedly higher quality of life scores than did the patients with the highest EDSS scores (≥6.5) for physical functioning (p < 0.0001) and social functioning (p = 0.002), whereas these EDSS groups did not differ significantly for the other dimensions.
The EDSS score correlated significantly and negatively with all SF-36 health dimensions except bodily pain. The highest coefficients were for physical functioning, social functioning, and general health (table 3). The EDSS could explain 73% of the variation in physical functioning, 23% in social functioning, 21% in general health, and little for the other dimensions.
Correlations between EDSS scores and eight dimensions of the SF-36
Figure 2 illustrates the strong correlation between physical functioning and EDSS score. There is a nearly linear relationship between the two variables except for EDSS values of 8 and higher, when the physical functioning score is 0 for nearly all patients (floor effect).
Figure 2. Mean score and 95% CIs for the SF-36 dimensions of physical functioning accordinto the Expanded Disability Status Scale (EDSS) score.
The relationship between the various FS and the SF-36 health dimensions was studied with regression analysis with the health dimensions as dependent variables. Because the FS scores were highly correlated (r between 0.90 and 0.52; data not shown), multiple regression analysis was used to find the FS scores with the greatest influence on health dimensions. FS pyramidal showed high partial correlation with physical functioning and social functioning; FS sensory correlated significantly with general health, bodily pain, and vitality (table 4). FS mental also correlated significantly with all health dimensions except general health and bodily pain.
Partial correlation coefficients from multiple regression analyses between the SF-36 dimensions (dependent variables) and the FS scales (independent variables)
Discussion.
A total of 64% of the MS patients had an EDSS score of 4 or less after at least 9 years of duration of MS. This finding contrasts with a recent study of a more selected patient population: 66% needed aids for walking.28 Nevertheless, our MS patients had markedly lower mean scores on all quality of life dimensions than the general population (figure 1). All dimensions except mental health showed a decreased score corresponding to at least a 0.5 SD in the general population. The patients with low EDSS scores had no decreased scores in mental health, but the patients with higher EDSS scores had markedly reduced scores for mental health.
A previous study of quality of life among 198 patients from MS clinics in Canada using SF-36 concluded that MS affects patients’ quality of life dramatically, and this was especially notable among patients with low EDSS scores or when impairment of ambulation was not yet a major symptom.5 Our results support the finding that the quality of life scores are reduced compared with the general population, also for the patients with low EDSS scores (≤2.5), although the scores were higher in the current study. Nevertheless, in contrast to the Canadian study,5 our patients with low EDSS scores scored significantly better in all dimensions than those with higher EDSS scores. The range of duration of MS among the patients with low EDSS scores was 9 to 19 years (mean, 13.8 years), and they therefore satisfy very closely previously defined criteria for a benign type of MS (EDSS score ≤ 3.0) and duration of MS more than 10 years.29 The mean duration of MS in the same group in the Canadian study was 9.2 years,30 and this material was hospital based and therefore more likely included higher EDSS scores and patients in an acute phase even within the range of 0 to 2.5. This might explain the difference in the scores between the two studies.
Two previous studies using SF-36 have found that impairment/disability quantified by the EDSS correlated only with physical functioning.17,18 Although EDSS scores correlated significantly with all dimensions except bodily pain, the EDSS correlation with physical functioning was also highest in our study. This reflects the fact that EDSS weights mobility heavily. The EDSS scores explained 73% of all the variation in physical functioning (see table 3), and the nearly linear relationship between these two measures (see figure 2) indicates that physical disability measured by EDSS might be self-assessed when examination is not feasible, as has been suggested previously.10,17 However, the floor effect for EDSS scores higher than 8 suggests that SF-36 physical functioning might not be the optimal measure for patients with severe disability.
The patients had major role limitations caused by physical problems. However, the correlation between EDSS and physical role limitation was relatively low (r = −0.33; see table 3). This scale measures how physical health problems interfere with regular daily activities, and the relatively low correlation probably means that the EDSS does not reflect suitably the impact of reduced physical functioning. The patients with low EDSS scores early experience role limitations caused by physical problems but seem to stabilize at a low level such that the mean scores do not differ between the two groups with the highest EDSS scores (see table 2). Such a difference might have been expected based on the marked difference in physical functioning between these two groups. The questions in physical role limitation are weighted toward changes in the impact of reduced physical functioning during the past 4 weeks. The patients with relatively stable but poor physical functioning, such as many of the chronic progressive patients, might have redefined their roles and therefore score relatively highly on this dimension compared with their physical functioning.
It has been claimed that the EDSS provides no information on how MS affects the patient’s quality of life.5 The EDSS explained almost all the variation in physical functioning (73%) and much of the variation in social functioning (23%) and general health (21%). This shows that reduced physical health as measured by the EDSS influences considerably self-assessed general health and restricts significantly the patients’ social functioning.
A previous study showed that physicians’ assessment of the relative importance of the domains of health-related quality of life differ from those of patients.17 Most physicians were more concerned than the patients about the physical manifestations of disease, whereas the patients evaluated vitality, role limitations caused by emotional problems, and mental health as important determinants of their overall quality of life. Our patients had markedly reduced scores compared with a general population within these dimensions, especially vitality. The EDSS could not explain much of the variation within these three dimensions. Fatigue is a frequent symptom of MS and is not reflected in the EDSS. MS patients are also bothered frequently by pain, and EDSS scores were not related to bodily pain (see table 3).
The multiple regression analysis including FS showed that the pyramidal, sensory, and mental scales seemed to influence the health dimensions the most. Reduced mental function as measured by FS mental seems to be related to important elements in self-assessed quality of life. FS mental measures primarily the decrease in mentation (mood alteration only gives a score of 1 on the scale from 0 to 5), whereas SF-36 mental health measures emotional well-being. The relatively low, but still significant, correlation between these variables reflects this difference in content but also shows that reduced mentation leads to reduced self-assessed quality of life.
This study investigated the relation between the EDSS and the SF-36. These are incomplete measures of impairment/disability and quality of life, and the interpretation of the results as an estimate of the relation between impairment/disability and quality of life is therefore limited. Nevertheless, the SF-36, in addition to reflecting the patients’ perspective, gives a broader measure of the disease impact in MS because the EDSS is related primarily to physical functioning and because all eight SF-36 health dimensions were reduced compared with the general population.
Acknowledgments
Supported by grants from the Norwegian Society of Multiple Sclerosis, Odd Fellow, Kjell Almes Legacy, and the Norwegian Research Council of Norway.
Acknowledgment
The authors thank Dr. Marit Grφnning and Dr. Christian A. Vedeler for participating in the clinical examination of the patients.
- Received January 14, 1999.
- Accepted April 24, 1999.
References
Disputes & Debates: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.