Calculation of the Minimal Clinically Important Difference in Upper and Lower Limb Motor Assessment in Spinal Muscular Atrophy

ABSTRACT

Objectives:

Physical function assessments in patients with spinal muscular atrophy (SMA) are important indicators for assessing the effectiveness of treatment and changes over time in rehabilitation therapy. However, few reports exist on this indicator. This study calculated the minimal clinically important difference (MCID) for assessing motor function in the upper and lower limbs of individuals with SMA to estimate the degree of change within a functional score that is considered clinically meaningful.

Methods:

This cohort study relied on individual participant measurements. A distribution-based approach was used to calculate the MCID values, incorporating data from 26 patients with SMA for the 6-Minute Walk Test (6MWT), Hammersmith Functional Motor Scale Expanded (HFMSE), Revised Upper Limb Module (RULM), and grip and pinch strength.

Results:

The standard errors of measurement for all patients were: 58.38 m for 6MWT; 4.71 points for HFMSE; 3.25 points for RULM; 10.93 N and 9.86 N for right and left grip strength, respectively; 5.42 N and 4.73 N for right and left Palmar pinch; and 11.96 N and 8.66 N for right and left Key pinch. Significant correlations were observed between the physical function assessments.

Conclusions:

We calculated MCID values for physical function evaluations of SMA and, as a sub-analysis, determined the SMA type and ambulatory status. These findings are expected to contribute to future SMA treatment and rehabilitation and promote the selection of appropriate physical function assessments.

INTRODUCTION

Spinal muscular atrophy (SMA) is a rare neuromuscular disorder characterized by progressive symmetrical muscle weakness caused by anterior horn cell degeneration. SMA leads to various impairments attributed to muscle weakness.^1,2) Muscle weakness can result in gait impairment, upper limb dysfunction, difficulty in activities of daily living, and, in advanced cases, severe respiratory complications.¹⁾ The incidence of SMA is estimated to be 1 in 10,000 people worldwide, with 7.8 to 10 cases per 100,000 people reported globally. Differences between regions may be attributed to the population size, the source of data, and the availability of cross-border testing. The prevalence of SMA is estimated to be 1–2 cases per 100,000 people.³⁾ Traditionally, the clinical classification of this disease has been based on the achievement of motor milestones during childhood. SMA spans a wide range of phenotypes, including children who cannot sit (SMA type 1), those who can sit but cannot walk (SMA type 2), and those who can walk with difficulty (SMA type 3).⁴⁾ The lack of fundamental treatment for this disease has been addressed since 2016; indeed, gene therapy targeting the causative gene for 5q-related SMA has significantly changed the treatment landscape for SMA.³⁾ Consequently, rehabilitation approaches for SMA have shifted from focusing on maintaining function, preventing complications, and acquiring compensatory strategies for functional impairment to extending functional maintenance and improving abilities. Although rehabilitation recommendations for SMA based on limited evidence have been reported, new evidence since the advent of recent therapeutic agents is lacking. Moreover, owing to the rarity of the disease, there is currently insufficient evidence regarding SMA rehabilitation.^5,6) Limited evidence exists about the extent to which improvements in physical function following rehabilitation or therapeutic interventions can be considered clinically significant. Specifically, there is no clear interpretation of objective evaluation values for tracking changes in physical function over time. Functional scores, such as the Hammersmith Functional Motor Scale Expanded (HFMSE), Revised Upper Limb Module (RULM), and 6-Minute Walk Test (6MWT), are used to assess disease severity, clinical progression, and treatment effects in SMA.^7,8,9) However, although these tests are primary assessments of functional ability, the primary impairment in SMA is muscle weakness, and the clinical significance of evaluations related to muscle strength has not been thoroughly investigated. Some studies have reported changes in scores from natural history studies for the HFMSE, RULM, and 6MWT, but only one study calculated the minimal clinically important difference (MCID) for these assessments.^10,11,12,13) Stolte et al.¹³⁾ calculated MCID values for 51 patients with SMA types 2 (n=36) and 3 (n=15) with a mean age of 35.8 ± 12.7 years (age range: 18–71 years) and reported MCID values of 4.3 and 2.9 points for the HFMSE and RULM, respectively. The MCID, the smallest change that indicates a clinically meaningful improvement for patients, is crucial in medical care and clinical research to determine whether a treatment or intervention has had a significant impact. The MCID may also aid in evaluating treatment efficacy. Limited evidence exists regarding the extent to which improvements in physical function following rehabilitation or therapeutic interventions can be considered clinically significant. Therefore, there remains urgent need to consider clinical applications of drug treatments, establish effective rehabilitation strategies, and generate supporting data. This study calculated the MCID values for upper and lower limb motor function evaluations of SMA, including muscle strength, with a view to assist the development of future rehabilitation treatments.

MATERIALS AND METHODS

Study Design and Participants

This cohort study relied on individual participant measurements and targeted patients with SMA types 2 and 3 who visited the Department of Rehabilitation Medicine at the National Center of Neurology and Psychiatry (NCNP) between December 2017 and August 2024. The following inclusion criteria were used: (i) SMA was diagnosed through genetic testing; (ii) the copy number of survival motor neuron 2 (SMN2) was known; and (iii) physical function evaluation was performed by the Department of Rehabilitation Medicine. Patients from whom consent could not be obtained after opting out were excluded from the study. Eligible patients were required to undergo physical function assessment during inpatient or outpatient rehabilitation. Regarding clinical care, outpatients received routine rehabilitation treatment at the Department of Rehabilitation Medicine on the same day as outpatient consultations by a neurologist or pediatric neurologist. All patients were evaluated every 3–6 months. However, inpatients were admitted for regular evaluation of their condition. At this time, routine rehabilitation was provided at the Department of Rehabilitation Medicine. For cases in which a treatment drug had already been administered, only data collected before drug administration were used, because it was suggested that such drugs could potentially influence physical function. For cases in which multiple assessments were performed, data collected immediately before drug administration were selected. Regarding sample size, no setting was made because the method used to calculate MCID in this study is independent of sample size. The extracted data included age, sex, SMA type, date of diagnosis, genetic data (including SMN2 gene copy number), and walking ability. Physical evaluations included the 6MWT, HFMSE, grip strength, and pinch strength (Palmar and Key pinch). The assessments were performed by three therapists (each with more than 5 years of experience in SMA rehabilitation) under the supervision of a rehabilitation physician. After the evaluation, the results were confirmed by a doctor, and the data were registered by the doctor who acted as the data manager to ensure accuracy. Data were extracted from the hospital databases. Three physicians from the Department of Physical Rehabilitation analyzed and verified the data. Final confirmation of the findings was performed by the first author (TH).

Ethics

This study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the National Center of Neurology and Psychiatry (A2024-024). This study was registered with the University Hospital Medical Information Network Clinical Trials Registry (UMIN-CTR) (UMIN000054591). Information was released on an opt-out method and informed consent was obtained in writing. Patients could withdraw from the study at any time. Information was provided on the National Center of Neurology and Psychiatry website and potential participants could withdraw from the study.

Study Procedures

To evaluate physical function, the tests were performed in the following order: 6MWT, HFMSE, RULM, grip strength, and pinch strength. All physical tests were performed on the same day. For patients with ambulatory difficulties, the 6MWT was omitted and HFMSE was performed first.

6MWT

The 6MWT is a walking test designed to evaluate exercise tolerance and overall functional capacity. In this study, a flat, straight course was prepared, and the patients were instructed to wear comfortable walking shoes and to rest prior to the test. At the start of the test, patients were instructed to “walk as far as possible in 6 min.” They were allowed to stop and rest if they felt fatigued but were encouraged to resume walking as long as time remained. The distance covered by the patient after 6 min was measured. The 6MWT is a valid and reliable test in outpatients with SMA.^14,15)

HFMSE

The HFMSE was originally developed as the Hammersmith Functional Motor Scale (HFMS) but was later expanded into its current form. HFMSE is used to quantitatively evaluate motor function in patients with SMA, helping monitor disease progression and treatment effects. It is particularly useful for evaluating the motor capabilities and limitations in patients with different types of SMA (types 1–3). The HFMSE comprises 33 motor tasks, such as rolling over, rising from a seated position, standing, ascending and descending stairs, and walking. It provides a detailed assessment of motor function. Each task was scored from 0 to 2: 0 points for an inability to perform the task, 1 point for partial ability, and 2 points for full ability, with a maximum possible score of 66 points. This scale has demonstrated high intra-rater reliability (Intraclass Correlation Coefficient: ICC =0.959–0.99).^16,17)

RULM

The RULM is an assessment scale developed to evaluate upper limb function in patients with SMA, particularly in those who can maintain a sitting position but have difficulty walking or who experience progressive muscle weakness in their upper limbs. The RULM quantitatively assesses upper limb function in daily life activities, including lifting arms, gripping and lifting objects, raising hands above the head, bringing hands to the mouth, and moving objects across a table. The task comprises19 items, 18 of which were scored using a 0–2 point scoring system and 1 item was scored using a 0–1 point scoring system, with a maximum possible score of 37.¹⁸⁾ An analysis of 126 patients with SMA using RULM demonstrated high reliability, with an ICC of 0.948.¹⁶⁾

Grip Strength and Pinch Strengths

Grip strength was measured with the participants seated on a chair with their feet flat on the floor and their elbows bent at a right angle with the forearm in a neutral position. The wrist was slightly extended, and the participant gripped a dynamometer placed on a table using the thumb and the all four fingers. The grip width of the dynamometer was adjusted to fit the participant’s hand size, ensuring that the proximal interphalangeal joint of the index finger was nearly at a right angle. The grip dynamometer was positioned such that the carpal bones were aligned with the handles. Pinch strength was measured in a seated position using the Palmar pinch (between the thumb and index finger) and the Key pinch (between the thumb and side of the index finger). Each measurement was performed twice on each hand, and the average value for each side was recorded. Grip and pinch strengths were measured in newtons (N), and the values were normalized by body weight. The ICC for grip strength was 0.91 and 0.85 for the right and left hands, respectively.^19,20) For pinch strength, the ICC ranged from 0.752 to 0.903 for the Palmar pinch and from 0.712 to 0.881 for the Key pinch.²¹⁾

Statistical Analyses

The MCID values and correlations were calculated using the assessed 6MWT, HFMSE, RULM, grip strength, and pinch strength. The MCID values were calculated based on a previous report.¹³⁾ Previous studies have evaluated the following three approaches to determining the MCID value: (1) anchor-based methods, which compare the change in score with other databases or clinical (e.g., physiological or laboratory) measures²²⁾; (2) distribution-based methods, such as half a standard deviation (1/2 SD), one-third of SD (1/3 SD), or standard error of measurement (SEm), which estimate the MCID value based on the distribution of measurements within a sample^22,23,24,25); and (3) the Delphi method, which seeks to reach a consensus by repeatedly presenting a questionnaire containing each of the previous results to experts.²⁶⁾

The number of patients with SMA in Japan is small, and there are often only a few patients at each institution. Indeed, no study in Japan has tracked changes in physical function in SMA patients over time. Furthermore, there are no reports of physical function assessments that could serve as anchor values for SMA. In addition, many patients had already started drug treatment when the study plan was formulated, making it difficult to calculate the MCID in a prospective study. For this reason, the Delphi method was not used in this study. Therefore, we calculated the MCID values using the 1/2 SD, 1/3 SD, and SEm methods.

The following formula was used to calculate SEm:

$$SEm=SD\sqrt{1-R\text{xx}}$$

The test–retest reliability (Rxx) of each evaluation was calculated using the ICC, based on previous reports. For the 6MWT, literature values of ICC ranged from 0.85 to 0.992.^14,15) To ensure a conservative calculation, we adopted 0.85 for our analysis. For the HFMSE, previous studies reported ICC values between 0.959 and 0.99.^16,17) We selected the lower value of 0.959. For the RULM, we used an ICC of 0.948.¹²⁾ The ICC values for grip strength were 0.91 and 0.85 for the right and left hands, respectively, based on previous studies.^19,20) Low values of ICC were also selected for pinch strength (Palmar, 0.752; Key, 0.712).²¹⁾ Finally, MCID values were calculated for subgroups based on SMA types (2 and 3) and ambulatory status (ambulatory and non-ambulatory). The categorization of patients into the ambulatory and non-ambulatory groups was based on progressive muscle weakness that begins with the onset of symptoms. The peak of muscle strength and functional impairment occurs at symptom onset, and there is not necessarily a rapid decline with age. In SMA types 2 and 3, the onset period may differ, and subanalysis based on other criteria, such as age, has poor clinical validity.

The differences in each item between the study groups were assessed by first determining the homogeneity of variance. If homogeneity of variance was confirmed, a t-test was conducted; otherwise, Welch’s test was performed. To determine correlations between each assessment, normality was examined using the Shapiro–Wilk test. Because not all items showed normality, Spearman’s rank correlation coefficient was used to assess the correlations. Bonferroni correction was applied to account for multiple comparisons.

Statistical significance was set at P <0.05. All statistical analyses were performed using SPSS Statistics for Windows, version 27 (IBM, Armonk, NY, USA).

RESULTS

The data of 28 patients were extracted for this study. One of these patients received initial treatment at our hospital at the sixth administration of their medication. Therefore, this patient was excluded because no data was available before drug administration. Another case of an infant with type 1 SMA was also excluded from this study. Based on the participant characteristics extracted from the data, 26 patients were enrolled. All patients underwent the assessments targeted for evaluation in this study (except for 6MWT, which was not performed for non-ambulatory patients). No adverse events or missing data were observed in the assessments. Patient background information is given in Table 1. Twenty-six participants of an equal sex distribution aged 2–50 years were enrolled. Nine participants (34.6%) were ambulatory, median weight was 29.85 kg [interquartile range (IQR) 25.95 kg], most were diagnosed 8 years before evaluation, and their median age at assessment was 12.5 years. Fifteen participants (57.6%) had SMA type 3, and nine of these (60%) could walk. Of the 11 participants (42.3%) with SMA type 2, none could walk. Participant characteristics did not differ between study groups, except for SMN copy number variation, which represents pathology discrimination.

Table 1.  Patient characteristics.

Characteristic	All patients (n = 26)	SMA type 3 (n=15)	SMA type 2 (n=11)	P value
Age at assessment, years	12.5 [23.75]	13 [26.5]	11 [21.0]	>0.05
Age range, years	2–50	3–50	6–31	>0.05
Male/female, n (%)	13 (50.0)/13 (50.0)	7 (46.7)/8 (53.3)	6 (54.5)/5 (45.5)	>0.05
Ambulatory, n (%)	9 (34.6)	9 (60.0)	0 (0)	N/A
Body weight, kg	29.85 [25.95]	36.0 [22.8]	26.2 [27.1]	>0.05
SMN2 copy number, n (2/3/4)	1/19/6	1/9/5	0/10/1	>0.05
Years between diagnosis and evaluation	8 [9]	7 [9]	9 [13.5]	>0.05

Data given as median [IQR] or number (percentage).

The physical function assessment results of each patient are presented in Table 2. The median 6MWT score for all patients was 0 [91] m, HFMSE was 15 [48.0] points, RULM was 11.5 [26] points, grip strength was 7.8 [16.2] N for the right hand and 10.8 [20.6] N for the left hand, Palmar pinch was 2.9 [5.9] N for the right hand and 3.4 [5.35] N for the left hand, and Key pinch was 5.8 [11.3] N for the right hand and 4.9 [7.9] N for the left hand. All patients were right-handed. The 6MWT, HFMSE, and RUML scores for participants with SMA type 3 were higher than those of participants with SMA type 2. The grip strength of participants with SMA type 3 was higher than that of participants with SMA type 2. The grip strength in the right hand was lower than that in the left hand for participants with SMA type 3 but did not differ by limb for participants with SMA type 2. The grip strength by weight for participants with SMA type 3 was higher than that of participants with SMA type 2. The grip strength by weight in the right hand was higher than that in the left hand for participants with SMA type 3, but it was lower than that in the left hand for those with SMA type 2. The distributions of 6MWT, HFMSE, and RULM in all patients with SMA, SMA type 3, and SMA type 2 are illustrated in Fig. 1

Table 2.  Physical function evaluation data.

Physical function	All patients (n=26)	SMA type 3 (n=15)	SMA type 2 (n=11)	P value
6MWT, m	0 [91]	74.5 [307.8]	0 [0]	0.009
HFMSE	15 [48.0]	46 [37.5]	3 [3.5]	0.01
RULM	11.5 [26]	30 [19.5]	4 [2]	0.01
Right grip, N	7.8 [16.2]	18.6 [67.8]	3.9 [4.9]	0.012
Left grip, N	10.8 [20.6]	22.5 [43.6]	3.9 [8.8]	0.011
Right Palmar pinch, N	2.9 [5.9]	6.8 [18]	1.9 [2.0]	0.015
Left Palmar pinch, N	3.4 [5.35]	6.3 [14.2]	1.9 [1.5]	0.01
Right Key pinch, N	5.8 [11.3]	13.2 [43.3]	1.9 [3.7]	0.01
Left Key pinch, N	4.9 [7.9]	7.7 [30.1]	1.9 [2.5]	0.011
Normalized right grip, N/kg	0.34 [0.81]	0.93 [1.50]	0.20 [0.23]	0.006
Normalized left grip, N/kg	0.32 [0.62]	0.79 [1.52]	0.22 [0.20]	0.01
Normalized right Palmar pinch, N/kg	0.15 [0.31]	0.27 [0.36]	0.07 [0.09]	0.05
Normalized left Palmar pinch, N/kg	0.13 [0.28]	0.32 [0.31]	0.07 [0.08]	0.011
Normalized right Key pinch, N/kg	0.22 [0.77]	0.62 [0.88]	0.12 [0.15]	0.011
Normalized left Key pinch, N/kg	0.16 [0.42]	0.41 [0.46]	0.07 [0.09]	0.016

Data given as median [IQR]. Maximum possible scores: HFMSE, 66; RULM, 37. Normalized data are relative to body weight.

Fig. 1.

 Box-plot diagrams showing the variability in each physical function evaluation. a) HFMSE, b) 6MWT, c) RULM, d) grip strength, e) Palmar pinch, and f) Key pinch. RT, right; Lt, left.

The calculated MCID values are presented in Table 3. For all participants, the SEm value was 58.38 m for the 6MWT, 4.71 points for the HFMSE, 3.25 points for the RULM, 10.93 N on the right and 9.86 N on the left for grip strength, 5.42 N on the right and 4.73 N on the left for Palmar pinch, and 11.96 N on the right and 8.66 N on the left for Key pinch. Figure 2 shows the distribution of the MCID values for the HFMSE, 6MWT, and RULM. For the HFMSE, all MCID values were higher in SMA type 3 group than in the SMA type 2 group. In contrast, SEm was higher in the non-ambulatory group than in the ambulatory group. For the 6MWT, the MCID value for the SEm in the ambulatory group was slightly higher than that in the entire SMA group. For the RULM, the trend was similar to that for the HFMSE. Figure 3 shows the distribution of MCID values for grip strength, Palmar pinch, and Key pinch on the left and right sides and also shows the values normalized for body weight. Generally, SMA type 3 had higher MCID values for all items than SMA type 2, and the ambulatory group had higher MCID values for all items than the non-ambulatory group. For correlations between each evaluation, 6MWT, HFMSE, RULM, grip strength, Palmar pinch, and Key pinch showed significant correlations with all other evaluation items (P <0.01) (Table 4).

Table 3.  Minimal clinically important differences (MCIDs) in physical assessment of SMA.

	Total (n=26)			SMA type 3 (n=15)			SMA type 2 (n=11)			Non-ambulatory (n=17)			Ambulatory (n=9)
	SEm	1/2 SD	1/3 SD	SEm	1/2 SD	1/3 SD	SEm	1/2 SD	1/3 SD	SEm	1/2 SD	1/3 SD	SEm	1/2 SD	1/3 SD
6MWT, m	58.38	75.37	50.25	58.38	75.37	50.25	N/A	N/A	N/A	N/A	N/A	N/A	60.15	77.66	51.77
HFMSE	4.71	11.63	7.75	4.66	11.50	7.67	2.12	5.24	5.40	2.19	5.41	3.60	1.22	3.01	2.01
RULM	3.25	7.13	4.75	3.16	6.92	4.62	0.63	1.38	0.92	1.60	3.52	2.35	0.69	1.51	1.00
Rt grip, N	10.93	18.21	12.14	12.88	21.47	14.31	1.21	2.02	1.35	3.87	6.46	4.31	14.11	23.51	15.68
Lt grip, N	9.86	12.73	8.48	11.56	14.92	9.95	1.92	2.48	1.65	3.13	4.04	2.70	11.86	15.31	10.21
Rt Palmar pinch, N	5.42	5.44	3.63	6.46	6.49	4.32	0.90	0.90	0.60	1.91	1.92	1.28	6.85	6.88	4.58
Lt Palmar pinch, N	4.73	4.75	3.17	5.52	5.54	3.70	0.65	0.65	0.43	1.75	1.76	1.17	5.76	5.78	3.86
Rt Key pinch, N	11.96	11.14	7.43	13.93	12.98	8.65	2.17	2.02	1.35	6.38	5.94	3.96	13.83	12.88	8.59
Lt Key pinch, N	8.66	8.07	5.38	10.15	9.45	6.30	2.17	1.43	0.96	4.48	4.17	2.78	10.45	9.73	6.49
Rt grip, N/kg	0.22	0.37	0.25	0.28	0.46	0.31	0.05	0.08	0.05	0.08	0.13	0.09	0.25	0.41	0.27
Lt grip, N/kg	0.26	0.33	0.22	0.32	0.41	0.27	0.07	0.10	0.06	0.09	0.12	0.08	0.26	0.34	0.23
Rt Palmar pinch, N/kg	0.13	0.13	0.09	0.16	0.16	0.11	0.08	0.08	0.06	0.07	0.07	0.05	0.15	0.15	0.10
Lt Palmar pinch, N/kg	0.11	0.11	0.07	0.13	0.13	0.08	0.05	0.05	0.03	0.04	0.04	0.03	0.10	0.10	0.07
Rt Key pinch, N/kg	0.26	0.25	0.16	0.29	0.27	0.18	0.16	0.15	0.10	0.14	0.13	0.09	0.20	0.18	0.12
Lt Key pinch, N/kg	0.20	0.18	0.12	0.29	0.27	0.14	0.11	0.10	0.07	0.10	0.09	0.06	0.20	0.18	0.12

Rt, right; Lt, left.

Fig. 2.

 Illustration of MCIDs in a) HFMSE, b) 6MWT, and c) RULM. Three distribution-based approaches for MCID were calculated for each subgroup: SEm (triangle), 1/2 SD (diamond), and 1/3 SD (circle).

Fig. 3.

 Illustration of MCIDs in a) right grip, b) left grip, c) right Palmar pinch, d) left Palmar pinch, e) right Key pinch, and f) left Key pinch. Three distribution-based approaches for MCID were calculated for each subgroup: SEm (triangle), 1/2 SD (diamond), and 1/3 SD (circle). Plots marked with asterisk indicate results for data normalized according to body weight. Rt, right; Lt, left.

Table 4.  Spearman correlation coefficients for each physical function assessment.

	6MWT		HFMSE		RULM		Rt grip		Lt grip		Rt Palmar pinch		Lt Palmar pinch		Rt Key pinch
	ρ	P value	ρ	P value	ρ	P value	ρ	P value	ρ	P value	ρ	P value	ρ	P value	ρ	P value
6MWT	-	-
HFMSE	0.799	0.000003	-	-
RULM	0.820	0.000031	0.862	0.000009	-	-
Rt grip	0.685	0.000438	0.825	0.000002	0.882	0.000003	-	-
Lt grip	0.659	0.001000	0.792	0.000011	0.801	0.000113	0.964	0.000000	-	-
Rt Palmar pinch	0.742	0.000076	0.882	0.000000	0.876	0.000004	0.829	0.000011	0.774	0.000015	-	-
Lt Palmar pinch	0.758	0.000044	0.892	0.000000	0.863	0.000008	0.869	0.000000	0.839	0.000000	0.954	0.000000	-	-
Rt Key pinch	0.715	0.000186	0.889	0.000000	0.827	0.000043	0.835	0.000001	0.832	0.000010	0.904	0.000000	0.947	0.000000	-	-
Lt Key pinch	0.695	0.000332	0.854	0.000000	0.816	0.000064	0.849	0.000000	0.858	0.000000	0.898	0.000000	0.943	0.000000	0.974	0.000000

Rt, right; Lt, left.

DISCUSSION

In this study, we calculated MCID values for various functional assessments related to physical disabilities in patients with SMA and performed a subgroup analysis based on SMA types (2 and 3) and ambulatory status (ambulatory and non-ambulatory). We identified correlations between different physical function evaluations. To the best of our knowledge, only one prior study has reported MCID values for SMA, and no such reports exist for Japan.¹³⁾ This previous report only calculated the MCID values for 6MWT, HFMSE, and RULM.¹³⁾ In addition, reports on the MCID values for grip and pinch strengths are lacking. This study is the first to report MCID values related to grip strength, which makes the findings significant.

Based on previous reports, the MCID was calculated using SEm, 1/2 SD, and 1/3 SD, which are the most used distribution-based approaches.^22,23,24,25) Stolte et al.¹³⁾ reported MCID values of 4.3 and 2.9 points for the HFMSE and RULM, respectively. Although our study involved a younger patient cohort, these results are consistent with our MCID values of 4.71 and 3.25 points for the HFMSE and RULM, respectively. This suggests that the reliability of the MCID values derived in this study, including those for grip and pinch strengths, is robust. The calculation of MCID values is crucial for the establishment of future rehabilitation treatments for SMA. MCID values may also serve as important milestones in evaluating the progression of physical function over time and assessing treatment efficacy when new therapies are developed. This study reaffirmed that the HFMSE and RULM are indispensable for evaluating physical function in clinical trials and rehabilitation.

Considering the calculated MCID values, SEm for the HFMSE was 4.71, consistent with previous studies.¹³⁾ A change of 3 points in SMA types 2 and 3 is considered clinically significant.^27,28) In addition, a natural history study of 268 patients with SMA types 2 and 3 (aged 2.5 to 55.5 years) found that 15.29% experienced a decrease of 2 points or more, whereas 7.83% experienced an increase of 2 points or more.¹⁰⁾ Therefore, a change of 2 points was considered clinically important. In this study, the calculated SEm for the HFMSE was greater than 2 points, except for the walking group, for which the SEm was 1.22. Low SEm in the ambulatory group was likely related to a ceiling effect. A study evaluating changes in the RULM over a 12-month period in 114 patients with SMA types 2 and 3 (aged 2.7–49.7 years) reported that 80% of patients overall showed a change of 2 or more points.¹¹⁾ In this study, the SEm was above 3 points in the entire SMA group and the SMA type 3 subgroup. However, in other categories, the SEm was lower, indicating the limitations of upper limb function evaluation in patients with advanced SMA, as discussed earlier. Overall, the MCID values were higher in SMA type 3 across all measures than in SMA type 2. This trend was also observed in the ambulatory and non-ambulatory groups. For both the HFMSE and RULM, SMA type 2 scores were generally lower, and the same trend was observed for grip and pinch strength. However, the difference between the scores for SMA types 2 and 3 tended to decrease when adjusted for body weight. Because muscle mass correlates with muscle strength, adjusting for body weight eliminates errors in measuring grip and pinch strength, and MCID values reflect the true muscle strength of patients with SMA more accurately.²⁹⁾

By calculating the MCID values of not only the HFMSE and RULM, which indicate SMA functional impairment, but also those for grip and pinch strengths, we constructed an index that could evaluate changes in the “impairment” of SMA. The MCID value of each evaluation was calculated to determine the meaning of the evaluation point; that is, the level of change that has meaning. We demonstrated the ability to simultaneously evaluate “disability” (through HFMSE and RULM) and impairment through grip and pinch strength.

Although the HFMSE and RULM are established measures of disability, adding grip and pinch strengths as part of the evaluation is equally important. In addition, the MCID values for the 6MWT obtained from subgroup analysis can be useful for clinical trials or rehabilitation programs tailored to ambulatory patients. Historically, HFMSE and RULM have been the only available measures for non-ambulatory patients; however, grip and pinch strengths can also be utilized effectively.

All assessments in this study demonstrated significant correlations, which adds value to the findings. Farrar et al.³⁰⁾ indicated a significant correlation between HFMSE and the age of onset for both SMA types 2 and 3, whereas Otto et al.³¹⁾ reported a negative correlation between HFMSE and magnetic resonance imaging-derived fat fraction and fractional anisotropy in SMA types 2 and 3. However, the only study that explored significant correlations between different physical function assessments was a natural history study by Wijngaarde et al.,³²⁾ which focused on the relationship between the HFMSE and the Medical Research Council score. No prior studies have reported correlations for the RULM or 6MWT, highlighting the novelty of our findings. Discovering these correlations between various assessments was a valuable outcome of this study. In clinical settings, the evaluation time is usually a critical factor. Conducting multiple assessments can increase the time required for evaluation; however, fatigue should be considered in patients with neuromuscular disorders, such as SMA, and conducting numerous consecutive assessments can be challenging.³³⁾ Indeed, in our results, the grip strength of the right hand was slightly reduced in all cases, who were right-handed. This may have been related to chronic fatigue or prolonged use. With our results showing significant correlations between the major functional evaluations of SMA types 2 and 3, clinicians can select appropriate assessments based on their understanding of the characteristics of each evaluation. If sufficient time for evaluation is not available in a clinical setting, the HFMSE alone will be useful because it encompasses lower limb function, upper limb function, and muscle strength. However, for non-ambulatory patients, adding assessments, such as the RULM, grip strength, and pinch strength, would be beneficial. Conversely, for detailed upper limb evaluations, considering the floor effect observed in RULM, it may be more advantageous to complement the assessment with grip and pinch strengths. The results of this study provide a rationale for selecting various assessments depending on the clinical context.

The results of the subgroup analysis revealed that in comparing the physical function evaluations between SMA types 2 and 3, all scores were lower in SMA type 2. Furthermore, differences were observed between the distributions of the two groups for HFMSE and RULM, with the distribution of SMA type 2 tending to be narrower than that of SMA type 3. This likely reflects the known ceiling and floor effects of the HFMSE and RULM.¹³⁾ Furthermore, the significant score disparity between SMA types 2 and 3 is a challenge in the current evaluation method. In SMA, as the disease progresses, the HFMSE score drastically declines when walking ability is lost. In the RULM, certain items require elevation movements, and when such movements become difficult because of proximal muscle weakness, the score tends to decrease. This highlights the importance of developing standardized and uniform assessments regarding score changes across neurodegenerative diseases (not only SMA). In contrast, although grip and pinch strengths showed large variance within the SMA type 3 group, the differences between SMA types 2 and 3 were not as pronounced as those observed in HFMSE and RULM. Wijngaarde et al.³²⁾ stated that the floor effect in HFMSE across SMA types 1–3 makes it unsuitable for long-term follow-up. This suggests that the evaluation of muscle strength could serve as an alternative measure, underscoring the usefulness of grip and pinch strength assessments in physical function evaluation.

The present study emphasizes the following points. First, only one previous study reported an MCID value for SMA, and that study only calculated MCID values for 6MWT, HFMSE, and RULM.¹³⁾ This study is the first to calculate MCID values related to grip and pinch strengths, and these findings are important. Second, we reported correlations between each assessment of physical function in SMA; however, correlations for RULM or 6MWT have not been reported previously, highlighting the novelty of our findings. Third, although several methods exist for calculating MCID values, this study adopted a distribution-based approach. One advantage of using SEm is that it is independent of sample size. Given that it is difficult to secure a sufficiently large sample size for rare diseases like SMA, SEm is a useful method for calculating MCID values. Fourth, evaluations in this study were conducted by therapists with over 5 years of experience in SMA rehabilitation. Therefore, although the data reflect the assessments of highly skilled therapists, maintaining consistency in the assessments is crucial in multicenter studies.

This study had some limitations. First, this was a retrospective study that used data from patients who provided informed consent. Many patients with SMA were already receiving new treatments, making it difficult to use physical function assessment data when the study decision was made. However, in situations where new treatments are available, conducting studies to calculate MCID values that exclude the effect of therapeutic drugs may be challenging when considering the ethical implications for patients’ well-being. Second, the time from disease onset to assessment was variable because the ages of the participants ranged from 2 to 50 years. The presence of new treatments influenced the development of this study but calculating the MCID values for rare diseases, such as SMA, is challenging because of the difficulty in securing a sufficient sample size. Therefore, our results may not reflect the true MCID of other populations nor be generalizable to other populations. It is also difficult to determine whether it is appropriate to combine data from younger patients recently diagnosed with SMA with data from patients diagnosed over a decade ago. Third, the data were collected from a limited number of facilities. Fourth, there are limitations pertaining to the role of the RULM as a functional assessment tool. Although it correlated with the HFMSE and MCID values, some sub-analyses revealed the influence of a floor effect. If the MCID values for grip strength prove valuable, as previously suggested, the relevance of using RULM as an assessment may diminish.³²⁾ Drawing definitive conclusions based solely on these results is premature; therefore, further research into long-term changes in RULM remains necessary. Fifth, because the calculation of SEm is based on Rxx, the calculation can be influenced by the reliability value. Therefore, when calculating MCID values using SEm, having reports on test–retest reliability and carefully selecting past reports on which to base the calculation are essential. For example, for the 6MWT, the test–retest reliability ranges from 0.85 to 0.992, and using a higher Rxx tends to result in a smaller SEm value.^14,15) Therefore, in this study, we used the lower Rxx value for a more rigorous calculation. However, it is impossible to calculate SEm without test–retest reliability data. In rare diseases with limited sample sizes, calculating test–retest reliability presents a significant challenge. In the absence of test–retest reliability data, the MCID should be calculated using another method, such as the Delphi technique or through calculations using Likert scales, but these approaches can be difficult in rare diseases.^26,34)

CONCLUSIONS

We calculated the MCID value for the physical function assessment of SMA and, as a sub-analysis, determined the SMA type and ambulatory status. We found a correlation between each physical function assessment. These results will aid in facilitating treatment and rehabilitation of SMA and selecting physical function assessments. In preparation for new therapies in the future, it is important to not only accumulate physical function assessments but also to evaluate simple scales that can capture changes alongside longitudinal physical function assessments.