- Email: [email protected]
The extended disability status scale (EDSS) as a predictor of impairments of functional activities of daily living in multiple sclerosis
132
Journal of the Neurological Sciences, 115 (19931 132-135 Elsevier Science Publishers B.V.
JNS 03953
The extended disability status scale (EDSS) as a predictor of impairments of functional activities of daily living in multiple sclerosis R o n a l d A. C o h e n
a,
H o w a r d R. Kessler b a n d M a r c Fischer c
" Departments of Neurology and Psychiatry University of Massachusetts Medical Center Worcester, MA, USA h Department of Rehabilitation Medicine Maine Medical Center Portland, ME, USA, and c Department of Neurology University of Massachusetts Medical Center and Medical Center of Central Massachusetts Worcester, MA, USA
(Received 28 February, 1992) (Revised, received 11 September, 1992) (Accepted 26 September, 19921
Key words: Multiple sclerosis; Disability; ADL; EDSS
Summary We investigated the relationship between the Kurtzke expanded disability status scale (EDSS) and an activities of daily living (ADL) scale that assesses disability across multiple functional domains based on self report of patients with multiple sclerosis (MS). Forty-three patients with definite MS responded on a 42-item ADL inventory and were also assigned an EDSS rating as part of their regular clinical examination. While a strong correlation was found between the EDSS and total ADL disability level, the ADL domain of 'Mobility' fully accounted for this relationship. The distribution of ADL scores as a function of EDSS level was curvilinear, as ADL variance was inconsistent across EDSS levels. Little ADL disability was evident when EDSS levels were below five. At higher EDSS levels, ADL variability increased substantially, making predictions regarding ADL level less exact.
Introduction The expanded disability status scale (EDSS) (Kurtzke, 1983) is the most commonly used quantitative system for measuring disability status in multiple sclerosis (MS). It was developed because other scales for measuring functional disability of activities of daily living (ADL) were felt by some clinical researchers to be inadequate for characterizing MS disability. While the EDSS has been extensively used in research, it generally does not correlate with neuropsychological disturbance in MS (Marsh 1980; Heaton et al. 1985; Franklin et al. 1989; Baumhfner et al. 1990), or with structural brain abnormalities as measured by magnetic resonance imaging (Rao et al. 1985). The EDSS may also lack reliability, as there is high variability among raters, with as little as 49% agreement in one study (Amato et al. 1988). Reliability appears weakest with respect to sensory and cognitive symptoms. While the EDSS lacks sensitivity to the neuropsychological impairments of MS, there is now evidence
Correspondence to: R.A. Cohen, Ph.D., Department of Neurology, UMMC, 55 Lake Ave, Worcester, MA 01655, USA. Tel.: (508) 856-5664; Fax: (508) 856-6778.
suggesting a substantive relationship between disability and neuropsychological status (Heaton et al. 1985; Kessler et al. 1991). Chronic progressive MS patients tend to perform more poorly than remitting-relapsing patients on neuropsychological tests (Heaton et al. 1985; Beatty et al. 1989). Furthermore, functional disability status as measured by deficits in activities of daily living has been associated with level of cognitive dysfunction (Staples and Lincoln 1979; Franklin et al. 1989; Kessler et al. 1991). We demonstrated that disability level reported on an A D L scale strongly correlated with memory impairments, particularly with those operations requiring effortful processing and focused attention (Kessler et al. 1991). This finding indicates that the observed relationship between MS disability and neuropsychological status may depend on the chosen approach to disability assessment. Since the EDSS continues to be the standard disability rating scale in MS research, it is important to determine how it corresponds with alternative approaches to disability assessment that are more sensitive to cognitive dysfunction. We conducted the present investigation to evaluate the relationship between the EDSS and a A D L inventory whose sensitivity to cognitive dysfunction has been previously established (Kessler et al. 19911.
133 3-
Subjects and methods We studied a sample of 43 patients (15 male, 28 female) with definite MS (Poser et al. 1983), recruited consecutively from the MS clinic of the University of Massachusetts Medical Center and the Medical Center of Central Massachusetts-Memorial. They had a mean age of 47.1 + 13.1 years (mean + s.d) and averaged 13.6 + 9.4 years since symptom onset. The patients showed four patterns of disease progression: Benign (6), remitting-relapsing (15), progressive-stable (5) and chronic-progressive (19). Subjects averaged 12.6 + 2.8 years of education. Functional disability was measured using the Activities of Daily Living (ADL) Scale, a 42-item inventory, previously used by Staples and Lincoln (1979) to evaluate the relationship between the intellectual and functional impairments of MS. This A D L inventory contained items representing the following functional domains: Mobility (10 items), Communication (4 items), Personal Care (5 items), Domestic Activity (18 items), Education (1 item), Employment (1 item) and Social Activity (3 items). Each test item was scored on a 1-3 scale, with higher scores indicating greater functional impairment. Each subject completed the A D L inventory during their regular neurologic examination. At the same time the attending neurologist determined subjects' EDSS rating. The Education and Employment domains were excluded from subsequent analyses, as subjects had difficulty responding to these items. Raw scores were obtained for each of the other domains by summing contributing items. A total A D L disability score was obtained by summing scores across all items. A Spearman rank-order correlation coefficient was computed to determine the overall relationship between EDSS and total A D L disability scores, as well as between each of the five domains comprising the total A D L score. A stepwise regression was then conducted to determine the relationship of the EDSS rating to the five A D L domains. The A D L scores of patients with high ( > 5) and moderate ( < 5) EDSS ratings were then compared using analysis of variance procedures (ANOVAs).
2 ¸ iii
A
td)
o
_1
¢/1
o I-'o
, °
-1
•
i
i
i
2
4
6
EDSS DISABILITY SCORE
Fig. 1. The curvilinear relationship between the A D L and EDSS disability is depicted for the MS patient sample. A D L scores are standardized, so that A D L = 0 reflects the mean A D L score for t h e sample, and A D L = 1 corresponds to an A D L disability score that is one standard deviation greater than the mean (i.e., greater A D L disability). A D L = - 1 corresponds to an A D L disability score that is one standard deviation less than the mean (i.e., less ADL disability).
whether their EDSS rating was greater or less than 5. Sixteen patients had an EDSS rating < 5, twenty-seven patients had an EDSS rating > 5, and no patient had an EDSS rating = 5. The moderate disability EDSS group had a mean EDSS rating of 2.06 + 1.06, while the high disability EDSS group had mean rating of 6.37 + 0.49 (mean + SD). The moderate disability group had a range of scores of 0-4, whereas the high disability group had a range of scores of 6-7. An analysis of variance indicated that the high EDSS group was significantly more impaired than the moderate disability EDSS group with respect to their total A D L score (F(1, 41) = 17.04, P < 0.0001). Mean rating for MS patients with EDSS < 5 was 44.68 + 5.53, as compared to 64.59 + 18.72 for patients with EDSS > 5. As evident in Fig. 1, the relationship between EDSS and A D L standardized scores was curvilinear, as specified by the polynomial regression equation: Y = - 1.33 + 0.96X + 0.38X 2 + 0.04X 3
Results Overall functional disability as measured by the total score on the A D L disability inventory correlated strongly with EDSS severity (r = 0.82, P < 0.0001). The subjects had a mean EDSS rating of 4.77 + 2.33 (mean + SD) and a mean total score on the A D L inventory of 57.19 + 17.97, indicating an overall moderate level of EDSS and A D L disability. Subjects were next divided into two groups based on
A D L scores increased very little as a function of increasing EDSS when ratings were less than 5, but then increased dramatically when the EDSS rating exceeded 5. This finding indicated that MS patients with EDSS ratings below 5 had little A D L impairment, while patients who had higher EDSS ratings showed greater A D L impairment as a function of their EDSS score. The distribution of A D L scores as a function of EDSS level were then analyzed. A striking difference was found between the distributions of the variances of
134 TABLE 1 CORRELATIONS BETWEEN ADL DOMAINS, DISABILITY AND EDSS DISABILITY LEVEL
Mobility C o m m u n i c a t i o n (C) P e r s o n a l care (PC) D o m e s t i c activity ( D A ) Social activity (SA) Total A D L disability (TD)
TOTAL
ADL
C
PC
DA
SA
TD
EDSS
0.56 -
0.65 0.46
0.79 0.65 0.56
0.76 0.49 0.62 0.78
0.90 0.66 0.62 0.96 0.81
0.78 0.57 0.53 0.77 0.66 0.82
. .
. .
.
. .
.
.
total A D L scores of patients from the two disability groups. The variance of A D L scores for the high EDSS disability group (EDSS > 5: s 2 = 350.71) was much greater than that of the moderate disability group (EDSS < 5: s 2 = 30.62). Analyses of the homogeneity of group variances using the Cochran test, Fmax statistic (F(1, 42) = 3.39, P < 0.001) and the Bartlett test X 2 = 19.39, d f = 1, P < 0.001) confirmed a highly significant difference between the group variances. Therefore, while MS patients with high EDSS ratings ( > 5) were much more likely to have significant A D L impairments, the extent of A D L impairment at higher EDSS levels is less predictable. As shown in Table 1, the five functional domains of the A D L inventory correlated differentially with EDSS severity. A stepwise regression was performed to determine which A D L domains best predicted EDSS disability. Raw scores for each of the five disability domains were entered as independent variables into the stepwise regression model to predict EDSS rating. An alpha level of 0.05 was used as the basis of entering and removing particular variables from the analysis. This analysis yielded a strong multiple correlation between the five A D L domains and EDSS rating which was highly significant (R 2 = 0.41, F = (1, 41) = 28.72, P < 0.0001). However, only the Mobility variable was retained by the stepwise procedure, indicating that the degree of impairment of mobility accounted for most of the variance in the relationship between A D L and EDSS disability.
Discussion
We have previously documented that this A D L inventory correlates strongly with memory dysfunction in MS (Kessler et al. 1991). In contrast, the EDSS has been repeatedly shown to lack sensitivity to the cognitive impairments of MS (Marsh 1980; Heaton et al. 1985; Franklin et al. 1989; Baumhfner et al. 1990). Therefore, while both the EDSS and A D L inventory presume to measure MS disability, they must differ in some fundamental way, to account for this dissociation. The present results suggest that both the EDSS and
ADL inventory may be sensitive to some of the same characteristics of MS disability, since a strong correlation between total scores for the two scales was found. Yet, the strong relationship between EDSS and ADL ratings was accounted for primarily by the contribution of the 'Mobility' items from the A D L inventory. This finding corroborates prior observations that the EDSS is most sensitive to gait and lower extremity dysfunction (Willoughby and Paty, 1988). Mobility is, however, only one component of functional disability status. An ideal disability rating system should account for the full range of A D L functions, since other causes of disability may be associated with cognitive status. Therefore, memory/cognitive dysfunction is likely to be betterpredicted by an ADL rating which reflects the involvement of multiple neural systems. Unlike the EDSS, the A D L inventory reflects the patient's self report and provides a comprehensive assessment of disability across multiple functional domains. While the EDSS and A D L inventories correlate as a function of the mobility domain, the A D L inventory is also sensitive to other functional domains of disability. We have also demonstrated a statistical characteristic of the EDSS that has bearing on its utility as an index of functional disability. At moderate EDSS levels, patients exhibited almost no A D L disability. At high EDSS levels ( > 5), patients not only exhibit greater A D L disability, they also exhibit much greater variance of total A D L scores This effect was not due to a restriction of range of the EDSS ratings for this group, since as we noted previously, the range and variance of the moderate EDSS group (EDSS: 0-4) was actually greater than that of the high EDSS group (EDSS: 6-7). Therefore, a wide range of EDSS scores was associated with a limited range of A D L disability when EDSS < 5. Conversely, a small range of EDSS scores was associated with a wide range of A D L disability when EDSS > 5. Consequently, the ability of the EDSS to predict overall A D L disability becomes less exact at higher EDSS levels. While we can be relatively certain of minimal A D L disability when EDSS levels are below 5, we can be less certain of the extent of A D L disability when EDSS exceed 5. The nonlinear relationship between A D L and EDSS ratings also potentially limits use of the EDSS as a measure of overall A D L disability. A D L disability does not increase significantly until EDSS levels exceed 5. When EDSS scores exceed 5, A D L disability increases dramatically, with a rapid increase in the slope of the underlying regression line. As a result two disability states: seem to be evident when an EDSS cutoff of 5 is used: ADL impaired-not impaired. This suggests that the EDSS has utility as a predictor of overall ADL functional disability, if one is willing to classify patients as functionally impaired or unimpaired based on this EDSS cutoff score.
135 T h e E D S S was d e v e l o p e d to provide specific inform a t i o n a b o u t MS f u n c t i o n a l disability; in p a r t i c u l a r disability associated with m o t o r system dysfunction. W h i l e this i m p o r t a n t d o m a i n of MS disability is well characterized by the EDSS, o t h e r A D L m e a s u r e s m a y provide m o r e d e t a i l e d i n f o r m a t i o n a b o u t o t h e r functional domains, i n c l u d i n g those with g r e a t e r d e p e n d e n c e o n cognitive abilities. T h e use of the E D S S t o g e t h e r with a m o r e c o m p r e h e n s i v e A D L inventory, such as the scale r e p o r t e d in this study, may provide the best m e a n s of fully characterizing the f u n c t i o n a l i m p a i r m e n t s of MS.
4
5
6
7
8 9
References 1 Amato, M.P., L. Fratiglioni, C. Groppi et al. (1988) Interrater reliability in assessing functional systems and disability on the Kurtzke scale in multiple sclerosis. Arch. Neurol., 45: 746-748. 2 Beatty, W.W., D.E. Coodkin, N. Monson, and P.A. Beatty (1989) Cognitive disturbances in patients with relapsing remitting multiple sclerosis. Arch. Neurol., 46: 1113-1119. 3 Baumhfner, R.W., W.W. Tourtellotte, K. Syndulko, V. Waluch, G.W. Ellison et al. (1990) Quantitative multiple sclerosis plaque assessment with magnetic resonance imaging: its correlation with
10
11
12
clinical parameters, evoked potential, and intra-blood-brain barrier IgG synthesis. Arch. Neurol., 47: 19-26. Franklin, G.M., L.M. Nelson, C.M. Filley and R.K. Heaton, Cognitive loss in multiple sclerosis: Case reports and review of the literature. Arch. Neurol., 1989: 46: 162-167. Heaton, R.K., L.M. Nelson, D.S. Thompson, et al. (1985) Neuropsychological findings in relapsing/remitting and chronic/ progressive multiple sclerosis. J. Cons. Clin. Psych., 53: 103-110. Kessler, H.R., R.A. Cohen, K. Lauer and D.F. Kausch (1991) The-relationship between disability and memory dysfunction in multiple sclerosis. Int. J. Neurosci., 60: 124-137. Kurtzke, J.F. (1983) Rating neurological impairment in multiple sclerosis: An expanded disability status scale (EDSS). Neurology, 33: 1444-1452. Marsh, G.G. (1980) Disability and intellectual function in multiple sclerosis patients. J. Nerv. Ment. Dis., 168: 758-762. Poser, C.M., L. Paty Scheinberg et al. (1983) New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann. Neurol., 13: 227-231. Rao, S.M., Glatt, S., Hammeke, T.A. et al. (1985) Chronic/ progressive multiple sclerosis: Relationship between cerebral ventricular size and neuropsychological impairment. Arch. Neurol., 42: 678-682. Staples, D. and N.B. Lincoln (1979) Intellectual impairment in multiple sclerosis and its relation to functional abilities. Rheum. Rehab., 18: 153-160. Willoughby, E.B. and D.W. Paty (1988) Scales for rating impairment in multiple sclerosis: a critique. Neurology, 38: 1793-1798.
Journal of the Neurological Sciences, 115 (19931 132-135 Elsevier Science Publishers B.V.
JNS 03953
The extended disability status scale (EDSS) as a predictor of impairments of functional activities of daily living in multiple sclerosis R o n a l d A. C o h e n
a,
H o w a r d R. Kessler b a n d M a r c Fischer c
" Departments of Neurology and Psychiatry University of Massachusetts Medical Center Worcester, MA, USA h Department of Rehabilitation Medicine Maine Medical Center Portland, ME, USA, and c Department of Neurology University of Massachusetts Medical Center and Medical Center of Central Massachusetts Worcester, MA, USA
(Received 28 February, 1992) (Revised, received 11 September, 1992) (Accepted 26 September, 19921
Key words: Multiple sclerosis; Disability; ADL; EDSS
Summary We investigated the relationship between the Kurtzke expanded disability status scale (EDSS) and an activities of daily living (ADL) scale that assesses disability across multiple functional domains based on self report of patients with multiple sclerosis (MS). Forty-three patients with definite MS responded on a 42-item ADL inventory and were also assigned an EDSS rating as part of their regular clinical examination. While a strong correlation was found between the EDSS and total ADL disability level, the ADL domain of 'Mobility' fully accounted for this relationship. The distribution of ADL scores as a function of EDSS level was curvilinear, as ADL variance was inconsistent across EDSS levels. Little ADL disability was evident when EDSS levels were below five. At higher EDSS levels, ADL variability increased substantially, making predictions regarding ADL level less exact.
Introduction The expanded disability status scale (EDSS) (Kurtzke, 1983) is the most commonly used quantitative system for measuring disability status in multiple sclerosis (MS). It was developed because other scales for measuring functional disability of activities of daily living (ADL) were felt by some clinical researchers to be inadequate for characterizing MS disability. While the EDSS has been extensively used in research, it generally does not correlate with neuropsychological disturbance in MS (Marsh 1980; Heaton et al. 1985; Franklin et al. 1989; Baumhfner et al. 1990), or with structural brain abnormalities as measured by magnetic resonance imaging (Rao et al. 1985). The EDSS may also lack reliability, as there is high variability among raters, with as little as 49% agreement in one study (Amato et al. 1988). Reliability appears weakest with respect to sensory and cognitive symptoms. While the EDSS lacks sensitivity to the neuropsychological impairments of MS, there is now evidence
Correspondence to: R.A. Cohen, Ph.D., Department of Neurology, UMMC, 55 Lake Ave, Worcester, MA 01655, USA. Tel.: (508) 856-5664; Fax: (508) 856-6778.
suggesting a substantive relationship between disability and neuropsychological status (Heaton et al. 1985; Kessler et al. 1991). Chronic progressive MS patients tend to perform more poorly than remitting-relapsing patients on neuropsychological tests (Heaton et al. 1985; Beatty et al. 1989). Furthermore, functional disability status as measured by deficits in activities of daily living has been associated with level of cognitive dysfunction (Staples and Lincoln 1979; Franklin et al. 1989; Kessler et al. 1991). We demonstrated that disability level reported on an A D L scale strongly correlated with memory impairments, particularly with those operations requiring effortful processing and focused attention (Kessler et al. 1991). This finding indicates that the observed relationship between MS disability and neuropsychological status may depend on the chosen approach to disability assessment. Since the EDSS continues to be the standard disability rating scale in MS research, it is important to determine how it corresponds with alternative approaches to disability assessment that are more sensitive to cognitive dysfunction. We conducted the present investigation to evaluate the relationship between the EDSS and a A D L inventory whose sensitivity to cognitive dysfunction has been previously established (Kessler et al. 19911.
133 3-
Subjects and methods We studied a sample of 43 patients (15 male, 28 female) with definite MS (Poser et al. 1983), recruited consecutively from the MS clinic of the University of Massachusetts Medical Center and the Medical Center of Central Massachusetts-Memorial. They had a mean age of 47.1 + 13.1 years (mean + s.d) and averaged 13.6 + 9.4 years since symptom onset. The patients showed four patterns of disease progression: Benign (6), remitting-relapsing (15), progressive-stable (5) and chronic-progressive (19). Subjects averaged 12.6 + 2.8 years of education. Functional disability was measured using the Activities of Daily Living (ADL) Scale, a 42-item inventory, previously used by Staples and Lincoln (1979) to evaluate the relationship between the intellectual and functional impairments of MS. This A D L inventory contained items representing the following functional domains: Mobility (10 items), Communication (4 items), Personal Care (5 items), Domestic Activity (18 items), Education (1 item), Employment (1 item) and Social Activity (3 items). Each test item was scored on a 1-3 scale, with higher scores indicating greater functional impairment. Each subject completed the A D L inventory during their regular neurologic examination. At the same time the attending neurologist determined subjects' EDSS rating. The Education and Employment domains were excluded from subsequent analyses, as subjects had difficulty responding to these items. Raw scores were obtained for each of the other domains by summing contributing items. A total A D L disability score was obtained by summing scores across all items. A Spearman rank-order correlation coefficient was computed to determine the overall relationship between EDSS and total A D L disability scores, as well as between each of the five domains comprising the total A D L score. A stepwise regression was then conducted to determine the relationship of the EDSS rating to the five A D L domains. The A D L scores of patients with high ( > 5) and moderate ( < 5) EDSS ratings were then compared using analysis of variance procedures (ANOVAs).
2 ¸ iii
A
td)
o
_1
¢/1
o I-'o
, °
-1
•
i
i
i
2
4
6
EDSS DISABILITY SCORE
Fig. 1. The curvilinear relationship between the A D L and EDSS disability is depicted for the MS patient sample. A D L scores are standardized, so that A D L = 0 reflects the mean A D L score for t h e sample, and A D L = 1 corresponds to an A D L disability score that is one standard deviation greater than the mean (i.e., greater A D L disability). A D L = - 1 corresponds to an A D L disability score that is one standard deviation less than the mean (i.e., less ADL disability).
whether their EDSS rating was greater or less than 5. Sixteen patients had an EDSS rating < 5, twenty-seven patients had an EDSS rating > 5, and no patient had an EDSS rating = 5. The moderate disability EDSS group had a mean EDSS rating of 2.06 + 1.06, while the high disability EDSS group had mean rating of 6.37 + 0.49 (mean + SD). The moderate disability group had a range of scores of 0-4, whereas the high disability group had a range of scores of 6-7. An analysis of variance indicated that the high EDSS group was significantly more impaired than the moderate disability EDSS group with respect to their total A D L score (F(1, 41) = 17.04, P < 0.0001). Mean rating for MS patients with EDSS < 5 was 44.68 + 5.53, as compared to 64.59 + 18.72 for patients with EDSS > 5. As evident in Fig. 1, the relationship between EDSS and A D L standardized scores was curvilinear, as specified by the polynomial regression equation: Y = - 1.33 + 0.96X + 0.38X 2 + 0.04X 3
Results Overall functional disability as measured by the total score on the A D L disability inventory correlated strongly with EDSS severity (r = 0.82, P < 0.0001). The subjects had a mean EDSS rating of 4.77 + 2.33 (mean + SD) and a mean total score on the A D L inventory of 57.19 + 17.97, indicating an overall moderate level of EDSS and A D L disability. Subjects were next divided into two groups based on
A D L scores increased very little as a function of increasing EDSS when ratings were less than 5, but then increased dramatically when the EDSS rating exceeded 5. This finding indicated that MS patients with EDSS ratings below 5 had little A D L impairment, while patients who had higher EDSS ratings showed greater A D L impairment as a function of their EDSS score. The distribution of A D L scores as a function of EDSS level were then analyzed. A striking difference was found between the distributions of the variances of
134 TABLE 1 CORRELATIONS BETWEEN ADL DOMAINS, DISABILITY AND EDSS DISABILITY LEVEL
Mobility C o m m u n i c a t i o n (C) P e r s o n a l care (PC) D o m e s t i c activity ( D A ) Social activity (SA) Total A D L disability (TD)
TOTAL
ADL
C
PC
DA
SA
TD
EDSS
0.56 -
0.65 0.46
0.79 0.65 0.56
0.76 0.49 0.62 0.78
0.90 0.66 0.62 0.96 0.81
0.78 0.57 0.53 0.77 0.66 0.82
. .
. .
.
. .
.
.
total A D L scores of patients from the two disability groups. The variance of A D L scores for the high EDSS disability group (EDSS > 5: s 2 = 350.71) was much greater than that of the moderate disability group (EDSS < 5: s 2 = 30.62). Analyses of the homogeneity of group variances using the Cochran test, Fmax statistic (F(1, 42) = 3.39, P < 0.001) and the Bartlett test X 2 = 19.39, d f = 1, P < 0.001) confirmed a highly significant difference between the group variances. Therefore, while MS patients with high EDSS ratings ( > 5) were much more likely to have significant A D L impairments, the extent of A D L impairment at higher EDSS levels is less predictable. As shown in Table 1, the five functional domains of the A D L inventory correlated differentially with EDSS severity. A stepwise regression was performed to determine which A D L domains best predicted EDSS disability. Raw scores for each of the five disability domains were entered as independent variables into the stepwise regression model to predict EDSS rating. An alpha level of 0.05 was used as the basis of entering and removing particular variables from the analysis. This analysis yielded a strong multiple correlation between the five A D L domains and EDSS rating which was highly significant (R 2 = 0.41, F = (1, 41) = 28.72, P < 0.0001). However, only the Mobility variable was retained by the stepwise procedure, indicating that the degree of impairment of mobility accounted for most of the variance in the relationship between A D L and EDSS disability.
Discussion
We have previously documented that this A D L inventory correlates strongly with memory dysfunction in MS (Kessler et al. 1991). In contrast, the EDSS has been repeatedly shown to lack sensitivity to the cognitive impairments of MS (Marsh 1980; Heaton et al. 1985; Franklin et al. 1989; Baumhfner et al. 1990). Therefore, while both the EDSS and A D L inventory presume to measure MS disability, they must differ in some fundamental way, to account for this dissociation. The present results suggest that both the EDSS and
ADL inventory may be sensitive to some of the same characteristics of MS disability, since a strong correlation between total scores for the two scales was found. Yet, the strong relationship between EDSS and ADL ratings was accounted for primarily by the contribution of the 'Mobility' items from the A D L inventory. This finding corroborates prior observations that the EDSS is most sensitive to gait and lower extremity dysfunction (Willoughby and Paty, 1988). Mobility is, however, only one component of functional disability status. An ideal disability rating system should account for the full range of A D L functions, since other causes of disability may be associated with cognitive status. Therefore, memory/cognitive dysfunction is likely to be betterpredicted by an ADL rating which reflects the involvement of multiple neural systems. Unlike the EDSS, the A D L inventory reflects the patient's self report and provides a comprehensive assessment of disability across multiple functional domains. While the EDSS and A D L inventories correlate as a function of the mobility domain, the A D L inventory is also sensitive to other functional domains of disability. We have also demonstrated a statistical characteristic of the EDSS that has bearing on its utility as an index of functional disability. At moderate EDSS levels, patients exhibited almost no A D L disability. At high EDSS levels ( > 5), patients not only exhibit greater A D L disability, they also exhibit much greater variance of total A D L scores This effect was not due to a restriction of range of the EDSS ratings for this group, since as we noted previously, the range and variance of the moderate EDSS group (EDSS: 0-4) was actually greater than that of the high EDSS group (EDSS: 6-7). Therefore, a wide range of EDSS scores was associated with a limited range of A D L disability when EDSS < 5. Conversely, a small range of EDSS scores was associated with a wide range of A D L disability when EDSS > 5. Consequently, the ability of the EDSS to predict overall A D L disability becomes less exact at higher EDSS levels. While we can be relatively certain of minimal A D L disability when EDSS levels are below 5, we can be less certain of the extent of A D L disability when EDSS exceed 5. The nonlinear relationship between A D L and EDSS ratings also potentially limits use of the EDSS as a measure of overall A D L disability. A D L disability does not increase significantly until EDSS levels exceed 5. When EDSS scores exceed 5, A D L disability increases dramatically, with a rapid increase in the slope of the underlying regression line. As a result two disability states: seem to be evident when an EDSS cutoff of 5 is used: ADL impaired-not impaired. This suggests that the EDSS has utility as a predictor of overall ADL functional disability, if one is willing to classify patients as functionally impaired or unimpaired based on this EDSS cutoff score.
135 T h e E D S S was d e v e l o p e d to provide specific inform a t i o n a b o u t MS f u n c t i o n a l disability; in p a r t i c u l a r disability associated with m o t o r system dysfunction. W h i l e this i m p o r t a n t d o m a i n of MS disability is well characterized by the EDSS, o t h e r A D L m e a s u r e s m a y provide m o r e d e t a i l e d i n f o r m a t i o n a b o u t o t h e r functional domains, i n c l u d i n g those with g r e a t e r d e p e n d e n c e o n cognitive abilities. T h e use of the E D S S t o g e t h e r with a m o r e c o m p r e h e n s i v e A D L inventory, such as the scale r e p o r t e d in this study, may provide the best m e a n s of fully characterizing the f u n c t i o n a l i m p a i r m e n t s of MS.
4
5
6
7
8 9
References 1 Amato, M.P., L. Fratiglioni, C. Groppi et al. (1988) Interrater reliability in assessing functional systems and disability on the Kurtzke scale in multiple sclerosis. Arch. Neurol., 45: 746-748. 2 Beatty, W.W., D.E. Coodkin, N. Monson, and P.A. Beatty (1989) Cognitive disturbances in patients with relapsing remitting multiple sclerosis. Arch. Neurol., 46: 1113-1119. 3 Baumhfner, R.W., W.W. Tourtellotte, K. Syndulko, V. Waluch, G.W. Ellison et al. (1990) Quantitative multiple sclerosis plaque assessment with magnetic resonance imaging: its correlation with
10
11
12
clinical parameters, evoked potential, and intra-blood-brain barrier IgG synthesis. Arch. Neurol., 47: 19-26. Franklin, G.M., L.M. Nelson, C.M. Filley and R.K. Heaton, Cognitive loss in multiple sclerosis: Case reports and review of the literature. Arch. Neurol., 1989: 46: 162-167. Heaton, R.K., L.M. Nelson, D.S. Thompson, et al. (1985) Neuropsychological findings in relapsing/remitting and chronic/ progressive multiple sclerosis. J. Cons. Clin. Psych., 53: 103-110. Kessler, H.R., R.A. Cohen, K. Lauer and D.F. Kausch (1991) The-relationship between disability and memory dysfunction in multiple sclerosis. Int. J. Neurosci., 60: 124-137. Kurtzke, J.F. (1983) Rating neurological impairment in multiple sclerosis: An expanded disability status scale (EDSS). Neurology, 33: 1444-1452. Marsh, G.G. (1980) Disability and intellectual function in multiple sclerosis patients. J. Nerv. Ment. Dis., 168: 758-762. Poser, C.M., L. Paty Scheinberg et al. (1983) New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann. Neurol., 13: 227-231. Rao, S.M., Glatt, S., Hammeke, T.A. et al. (1985) Chronic/ progressive multiple sclerosis: Relationship between cerebral ventricular size and neuropsychological impairment. Arch. Neurol., 42: 678-682. Staples, D. and N.B. Lincoln (1979) Intellectual impairment in multiple sclerosis and its relation to functional abilities. Rheum. Rehab., 18: 153-160. Willoughby, E.B. and D.W. Paty (1988) Scales for rating impairment in multiple sclerosis: a critique. Neurology, 38: 1793-1798.