Background
The diverse symptoms associated with multiple sclerosis (MS) adversely impact health-related quality of life (HRQL) which, in turn, is manifested in extensive physical, psychosocial and economic burden [
1‐
3]. Although the assessment of HRQL in MS is well recognized as an important clinical assessment tool [
4], burden of morbidity of persons with MS in comparison with the general population is largely unknown.
The Expanded Disability Status Scale (EDSS) is the primary disease specific health measure for MS [
5], but it is heavily weighted toward ambulation and is unable to provide a broader comparison of HRQL attributes among different conditions and the general population. The use of a generic health measure to complement the disease specific health measure is typically advocated for the appraisal of the overall impact of MS.
The evaluation of HRQL of persons with MS has been primarily in clinical or patient study groups. Relying solely on these cohorts limits the external validity of these findings and generates possible selection bias [
6]. Few investigations have compared the burden of illness in MS to a reference group or to the general population [
7‐
10]. Subsequently, only some studies have made comparisons by statistically adjusting for differences between persons with MS and a reference population. Limited evidence indicates that physical attributes such as ambulation are lower in persons with MS than the general population; however, it is unclear whether other attributes such as pain and emotion are relatively lower than the general population. Using the SF-36, lower scores were reported for only physical dimensions in persons with MS as compared to the US population; however, mental health scores were comparable to the general population [
8]. Alternatively, both physical and mental health components were lower for patients with MS than the Norwegian general population [
7]. Others have reported problems with balance, cognition, visual disturbance, bowel and bladder difficulties, spasticity, depression, anxiety, bipolar disorders, speech problems and fatigue for persons with MS who reside within the community [
9,
11,
12].
The comparison of HRQL in the general population to a sample of persons with MS provides quantitative baseline estimates of the impact of MS which, in turn, can be used for therapeutic intervention, program and healthcare evaluations. The primary aim of this study was to compare the HRQL of persons who have MS to those persons without MS, using a cross-sectional representative sample of the general population. A secondary aim was to identify the attributes associated with the burden of MS relative to the general population adjusting for various socio-demographic factors. To assist in identifying the effect of MS independent of other comorbidities, we also compared the HRQL of persons with MS alone to the health status of respondents without any chronic medical conditions.
Results
In this community-based population, respondents with MS were approximately four years older on average than the general population (48.7 versus 44.8 years of age, p < 0.05) (Table
1). A larger proportion of respondents with MS than respondents without MS was female (68.3% versus 50.9%, p < 0.05). Unadjusted overall HUI3 scores were considerably lower for respondents with MS (0.57 versus 0.88); this difference was more than 10 times that would be considered clinically important (Table
1). Clinically important differences were also seen with the single attributes; MS respondents had lowers scores for ambulation, dexterity, cognition and pain (Table
1).
Table 1
Demographic characteristics
Age – mean (95% CI) | 48.7 (46.6 to 50.8)* | 44.8 (44.7 to 44.8) |
Sex (% Female) | 68.3* | 50.9 |
Education – % | | |
Less than high school | 17.9 | 22.2 |
High School | 19.0 | 20.4 |
Some post-secondary/college/trade school | 42.9 | 36.5 |
University degree | 20.1 | 20.9 |
Marital Status (% Married) | 71.0 | 64.3 |
Social Assistance (% Receiving Social Assistance) | 11.1* | 5.1 |
Number of other Medical Conditions mean (95% CI) | 2.6 (2.3 to 3.0)* | 1.6 (1.6 to 1.6) |
HUI3 Scores – mean (95% CI)† | | |
Overall HUI3 score | 0.57 (0.52 to 0.63)* | 0.88 (0.88 to 0.88) |
Vision | 0.93 (0.91 to 0.95)* | 0.97 (0.97 to 0.97) |
Hearing | 0.99 (0.98 to 0.99) | 0.99 (0.99 to 0.99) |
Speech | 0.99 (0.99 to 0.99) | 1.00 (0.99 to 1.00) |
Ambulation | 0.71 (0.65 to 0.77)* | 0.98 (0.98 to 0.98) |
Dexterity | 0.93 (0.90 to 0.96)* | 1.00 (1.00 to 1.00) |
Emotion | 0.93 (0.91 to 0.95)* | 0.97 (0.97 to 0.97) |
Cognition | 0.89 (0.86 to 0.92)* | 0.96 (0.96 to 0.96) |
Pain | 0.75 (0.69 to 0.80)* | 0.93 (0.93 to 0.93) |
After adjusting for the model covariates (age, sex, education, marital status, and social assistance), clinically important differences in overall HUI3 scores and single attribute scores persisted. The mean difference in overall HUI3 scores between respondents with and without MS was reduced from 0.31 to 0.25 after adjustments (95% confidence interval (CI): 0.20 to 0.31, p < 0.05) (Table
2). This difference, however, was still more than eight times what would be considered clinically important. On the single attributes, a difference of 0.05 is considered clinically meaningful. Particularly large differences in scores were observed for ambulation and pain, with differences of 0.26 (95% CI: 0.20 to 0.32, p < 0.05) and 0.14 (95% CI: 0.09 to 0.19, p < 0.05), being observed, respectively. Clinically important differences on the dexterity and cognition attributes were also observed, although differences were not as large as those observed for ambulation and pain (Table
2). No clinical differences were seen with sensory and emotion attributes.
Table 2
Adjusted† mean scores and differences in overall and single attribute utility scores for respondents with and without MS
Overall HUI3 score | 0.58 | 0.84 | 0.25 (0.20 – 0.31)* |
Vision | 0.94 | 0.96 | 0.03 (0.01 – 0.05)* |
Hearing | 0.99 | 0.99 | 0.00 (-0.01 – 0.004) |
Speech | 0.99 | 1.00 | 0.01 (-0.003 – 0.006) |
Ambulation | 0.72 | 0.98 | 0.26 (0.20 – 0.32)* |
Dexterity | 0.93 | 1.00 | 0.06 (0.03 – 0.10)* |
Emotion | 0.92 | 0.95 | 0.03 (0.01 – 0.05)* |
Cognition | 0.89 | 0.94 | 0.05 (0.02 – 0.09)* |
Pain | 0.77 | 0.91 | 0.14 (0.09 – 0.19)* |
When persons with MS alone were compared to persons without MS or any other chronic conditions, differences in overall and single attribute utility scores were similar to those observed in the entire sample (Table
3). The difference in overall HUI3 scores between the subgroups without any chronic conditions was 0.29 (95% CI: 0.18 to 0.41, p < 0.05) (Table
3). Again, the largest differences on the single attributes were observed for ambulation and pain (Table
3). Clinically important differences were also observed on the dexterity and cognition attributes.
Table 3
Adjusted† mean scores and differences in overall and single attribute utility scores for respondents with and without MS, but no other chronic conditions
Overall HUI3 score | 0.64 | 0.93 | 0.29 (0.18 to 0.41)* |
Vision | 0.95 | 0.98 | 0.03 (-0.01 to 0.07)* |
Hearing | 1.00 | 0.99 | 0.00 (0.00 to -0.01) |
Speech | 0.99 | 1.00 | 0.00 (-0.01 to 0.02) |
Ambulation | 0.69 | 1.00 | 0.31 (0.17 to 0.44)* |
Dexterity | 0.95 | 1.00 | 0.05 (-0.01 to 0.11) |
Emotion | 0.93 | 0.97 | 0.04 (-0.01 to 0.09)* |
Cognition | 0.88 | 0.97 | 0.09 (0.02 to 0.16)* |
Pain | 0.88 | 1.00 | 0.11 (0.03 to 0.20)* |
Discussion
Within the context of a national population health survey, the burden of illness for persons with MS was quantified using a generic health measure, HUI3 in the community dwelling population. We found that the MS population experienced large deficits in overall HRQL relative to the general population without MS. When the effect of other chronic conditions was removed, persistent large deficits of HRQL existed for persons with MS. Given the underlying neuropathologic changes that occur to the central nervous system and the diverse clinical features, it is not surprising that HRQL would be affected by the disease. In particular, our findings quantify significant difficulties with pain, ambulation, dexterity, and cognition in persons with MS.
Our findings were similar to other population samples. A Norwegian community-based cohort reported lower health status in persons with MS compared to the general population [
7]. Findings from a cross sectional survey also reported lower physical functioning, vitality, general health and psychological domains in MS patients than controls [
25,
26]. An association between MS and mental health has been also reported within other population-based samples [
12,
27]. Although SF-36 physical component scores in persons with MS were lower in comparison to the US population, the mental component scores were similar to the general population [
8]. This divergence from other studies may be attributable, in part, to psychometric properties of the components scores for the SF-36. Orthogonal factor rotation is used in the determination of the SF-36 component scores, that is, mental and physical components scores are treated as independent. Subsequently, the algorithm has been shown to significantly under-estimate mental health of patients with MS as compared to the component scores based on the RAND-36 Health Status Inventory [
28].
While the prevalence of pain in MS is well recognized, the severity relative to the general population has been examined by few investigators [
29]. We reported large differences in pain for both unadjusted and adjusted analysis. The difference between the MS and general population while adjusting for other covariates was almost five times that would be considered clinically important. Others have recognized acute and chronic pain in this patient population as a substantial clinical problem [
29‐
31], while others have not [
7,
8]. Congruent with our findings, Svendsen and colleagues reported that the severity of pain is greater in persons with MS than the general population [
29]. The ramifications of pain are also far reaching given the associations with depression, fatigue, and poorer health status [
30,
32]. The assessment and treatment of pain warrants further consideration in this patient population which may directly improve HRQL.
The HRQL of persons with MS reported in this study illustrates that the HRQL is worse than HRQL reported with many other diseases. Maddigan and colleagues reported the overall HUI3 score for persons with diabetes, heart disease, arthritis or stroke ranged from 0.74 to 0.89 [
22]. Moreover, the overall HUI3 scores of various combinations using three of these four conditions still ranged from 0.62 to 0.66 [
22]. In relative terms, one may conclude that the HRQL of burden associated with MS is substantially higher than any one of these four other chronic conditions or in any combination of three of these four conditions. Others have also reported that patients with MS are among one of the most severely impaired in comparison with other chronic conditions such as cardiovascular conditions, cancer, endocrinologic conditions, and chronic respiratory diseases [
6,
33]. This illustrates the severe impairment that is associated with MS, even among community dwelling individuals with the disease.
Although secondary analysis makes use of valuable data, the limitations of these findings are noteworthy. First, ascertainment of MS was via self-report. Although questions regarding the presence of medical conditions specified that the condition was diagnosed by a health professional, there remained potential for individuals to over- or under-report any medical condition, including MS. Likewise, no disease specific health measure for MS or indicator for disease course was included in the CCHS 1.1.
Another limitation of this study concerned the number of respondents who were missing data on covariates and were excluded from the analysis. While this was less than 10.0% of MS respondents, generalizability of these results to the respondents with missing data may be limited. Despite over 98% of the Canadian community dwelling population being represented in the survey, the generalizability of the results to the entire Canadian population with MS is limited by the fact that the sampling frame would not capture those individuals who reside in institutions or on reserve lands. That being said, the true HRQL burden of the entire Canadian population with MS would be under-estimated by these results given that individuals with MS who resided in institutions were more likely to have greater impairment than those residing in the community. Although the impact of MS appears to be more severe in First Nations People, the prevalence rates of MS are relatively low [
34] and would likely have a small impact on the overall HRQL of this sample population.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Drs Jones and Pohar were responsible for the conception of the study. Dr. Pohar analyzed the data. Dr. Jones drafted the article. All authors contributed to the interpretation of the results and revising the article for important intellectual content. All authors read and approved the final manuscript.