Associations of Upper Limb Disability Measures on Different Levels of the International Classification of Functioning, Disability and Health in People With Multiple Sclerosis

Relationships Between Impairments on the Body Functions and Structures Level and Measures on the Activity and Participation Levels

In the total group and both dexterity subgroups, general upper limb strength (MI) was the only impairment that was most strongly associated with upper limb capacity measures (NHPT and ARAT) and perceived performance (MAM-36) on the activity level. In almost all (5/6) regression models, general upper limb strength was one of the retained variables. However, the strength of the associations was higher in the low-dexterity subgroup than in the high-dexterity subgroup, perhaps because participants with MS in the low-dexterity subgroup had significantly more muscle weakness than those in the high-dexterity subgroup. Most participants with MS in the low-dexterity subgroup had an MI score below 90 of 100, whereas 14 participants with MS in the high-dexterity subgroup had the maximum score on the MI. In addition, our findings are in line with those of earlier studies performed in people with MS and different dexterity levels.^7–9 Guclu-Gunduz et al⁷ found a significant moderate correlation (r=−.44) between elbow flexion strength measured with a handheld dynamometer and the NHPT in 22 people with MS and less disability (median EDSS score of 2). However, shoulder abduction and flexion strength were not significantly related to the NHPT.⁷ In contrast, moderate to high correlations (r=.69–.79) between muscle strength (MI) and perceived performance (Motor Activity Log [MAL]) were found in people with MS with more disability (median EDSS score of >7.0).^8,9 On the basis of these results, general upper limb strength is considered to be an important outcome measure explaining upper limb capacity and perceived performance in people with MS, especially in people with MS and a low dexterity level.

Handgrip strength showed low to moderate correlations with capacity and perceived performance measures. Similarly, in other studies, low to moderate correlations were found between handgrip strength and the NHPT,^6,7 MAL,⁹ or MAM-36.⁵ On the basis of these results, one may assume that other motor impairments, such as maximum strength in the fingers, force control, or muscle fatigue, may be more related to upper limb capacity or performance in ADL than maximum handgrip strength. Further research regarding the assessment of these motor impairments and their relationships with the activity and participation levels is needed.

Although the difference between the mean scores of both manual dexterity subgroups was relatively small, participants with MS in the low-dexterity subgroup had significantly less AROM of wrist extension than did participants with MS in the high-dexterity subgroup. Only in the total group and the low-dexterity subgroup was active wrist extension significantly correlated with the activity level measures. Moreover, it was one of the variables retained in the regression model predicting the variance in the ARAT for these groups. These findings confirm, as expected from a clinical point of view, that the performance of ARAT test items requires a minimum AROM of wrist extension (eg, placing blocks on a shelf). However, decreased AROM of wrist extension seems to have no role in explaining the variance in the NHPT and the perceived ease or difficulty of performing ADL tasks (MAM-36). The fact that the median score in the high-dexterity subgroup was close to the normative values (60°–75°)³⁷ may explain the lack of association between the AROM of wrist extension and the activity level in the high-dexterity subgroup.

Impaired tactile sensitivity in the thumb was present in both manual dexterity subgroups but was significantly correlated with activity level measures only in the total group and the high-dexterity subgroup. This result was confirmed by regression analyses in which tactile sensitivity was one of the retained variables in the models predicting variances in the NHPT and the ARAT for the total group. Remarkably, tactile sensitivity was not a retained variable in the regression models for either of the manual dexterity subgroups. In addition to these results, a moderate correlation between tactile sensitivity of the thumb and the NHPT was reported by Guclu-Gunduz et al.⁷ On the basis of these results, one may postulate that impaired tactile sensitivity has a substantial impact on upper limb capacity (NHPT and ARAT) only when motor functions are adequate for task execution. However, it appears to have no impact on the perceived ease or difficulty of performing ADL tasks. These findings suggest that people with MS may have developed compensation strategies for performing their ADL (eg, visual compensation). Furthermore, we acknowledge that other sensory modalities, such as kinesthetic modalities, proprioception, and vibration, may affect the ability to perform ADL.

In conclusion, general upper limb strength was found to be the most important impairment affecting the capacity to perform ADL and perceived upper limb performance in daily life in people with MS. Besides muscle strength, the AROM of wrist extension and tactile sensitivity of the thumb were also important contributors. These findings may help clinicians to select appropriate rehabilitation interventions (eg, strength training or sensory training in people with MS and a high dexterity level) to improve disability on the activity level in people with MS. Furthermore, the results of the present study suggest that impairments on the body functions and structures level and disability on the activity level are not highly related to participation within the community. This finding was expected because other factors, such as walking ability and balance, likely are important factors influencing participation.

Relationships Between Capacity and Perceived Performance Measures on the Activity Level

The capacity measures were able to predict 44% (NHPT) to 42% (ARAT) of the variance in the perceived performance measure in the total group. These results are consistent with those of earlier research performed with a smaller study sample; in that research, regression analyses revealed that the NHPT was able to predict 55% of the variance in the MAL in the nondominant and more impaired arm.⁹ In the latter study, a high correlation was found between the ARAT and the MAL. Despite the fact that the MAL and the MAM-36 have different scoring methods, the associations between perceived performance and capacity measures appeared to be similar, especially in people with MS and a low dexterity level.

The associations between capacity and perceived performance measures in the low-dexterity subgroup were similar to those in the total group. In contrast, only low to nonsignificant correlations were found between perceived performance and capacity measures in the high-dexterity subgroup. The low correlations between the ARAT and the MAM-36 in the high-dexterity subgroup can be explained by the lack of variation in ARAT scores, as most participants with MS in this group had an ARAT score above 53, close to the maximum score (57). In contrast, NHPT scores were more varied (Figure). Most participants with MS in the high-dexterity subgroup performed the NHPT at 0.33 peg per second (median score for the high-dexterity subgroup). Despite the relatively high, almost normal scores on these capacity measures, the MAM-36 scores mostly ranged between 60 and 80, suggesting that these participants with MS performed daily life tasks with different levels of ease or difficulty.

In conclusion, both the NHPT and the ARAT were valid outcome measures for assessing upper limb capacity in people with MS and a low dexterity level, whereas the ARAT had a ceiling effect in people with MS and rather high manual dexterity, that is, those who could perform the NHPT faster than 0.27 peg per second. These results underscore the importance of the consideration of a person's disability in the selection of tailored outcome measures in order to avoid ceiling effects. Furthermore, most people with a high dexterity level and good ARAT scores still reported some degree of difficulty in performing ADL, as measured with the MAM-36, suggesting that perceived performance measures and capacity measures assess different constructs. Therefore, we believe that it is important to include both performance and capacity measures in the evaluation of upper limb function in people with MS (as in patients with stroke³⁸) in order to provide a complete understanding of the health care services needed by people with MS.³⁹

Methodological Considerations

To differentiate people with MS and different dexterity levels, we decided to use a median split for the NHPT scores. This approach resulted in a cutoff value of 0.27 peg per second (corresponding to 33.3 seconds) on the NHPT. The differentiation of low- and high-dexterity levels on the basis of this cutoff value resulted in significant differences between the subgroups on all outcome measures except the visual analog scale for pain. Our proposed cutoff value differs from that proposed by Kierkegaard et al,³ who suggested the use of a score of 0.5 peg per second (18 seconds) on the NHPT as an indicator of functioning and to identify people with MS and at risk of activity limitations and participation restrictions. The cutoff value of Kierkegaard et al³ was calculated on the basis of Katz personal ADL, Katz instrumental ADL, and the Frenchay Activities Index. These outcome measures assess global disability on the activity and participation levels and are strongly influenced by a person's walking ability, the ability to perform a transfer, and cognitive function. The cutoff value of Kierkegaard et al³ can be used to differentiate people with MS and no impaired upper limb dysfunction from people with MS and minimally impaired upper limb dysfunction; our proposed cutoff value differentiates people with MS and different levels of upper limb dysfunction. Further research regarding the clinical use of cutoff values to differentiate people with MS and different upper limb disability levels, as is done for people with stroke,¹⁸ is required.

Limitations

The following limitations need to be addressed. First, the design of the study itself may have some limitations. One of the limitations of a cross-sectional study design is that it is not possible to investigate causality. A cohort study design would provide more information. Second, the results for the high-dexterity subgroup should be interpreted with caution because less variance in scores on the outcome measures was found for this subgroup (Tab. 2). Moreover, some outcome measures with ordinal rating scales showed ceiling effects in people with MS and a high dexterity level. These effects may influence statistical analyses investigating associations of outcome measures in this subgroup. Finally, despite an already comprehensive assessment of upper limb function, we did not measure all impairments that may have had an impact on ADL, such as muscle fatigue, force control, or impaired proprioception or vibration. Further research investigating the specific effects of these impairments on the activity and participation levels is needed.