Background

Multiple sclerosis (MS) is an inflammatory immune-mediated disease of the central nervous system (CNS) [1,2,3]. MS affects about 2.8 million people worldwide [4], and in 2018 its incidence was 40.4/100,000 of the population in the Kingdom of Saudi Arabia (KSA) [5]. It is more common in young adults aged 20–40 years of age [6] and in women than men (ratio 2:1) around the world, including in Saudi Arabia [4].

The most common clinical manifestations of MS are numbness, urinary/bowel symptoms, ataxia, tremor or dysmetria, visual disturbance, cognitive impairment (CI), and motor dysfunction (MD) [3, 7, 8]. Motor and cognitive dysfunction are the most common complications in patients with MS that negatively impact their quality of life (QoL) [9,10,11].

The MD seen in MS can take several forms including mobility problems, muscle weakness, spasticity, motor incoordination, and balance dysfunction [8, 12]. About 41% of patients with MS have walking difficulties that increase the fall risk and affect their daily living [13,14,15].

CI is also common in patients with MS, with a reported prevalence ranging between 43% and 70% [16,17,18]. It varies in severity and presents in all types of MS [19]. The most affected domains of CI are executive dysfunction, lack of attention, loss of memory, and slow information processing [20,21,22]. CI may disturb social interactions, work participation [23], and daily living in patients with MS [15, 24].

Understanding the relationship between motor and cognitive impairments is crucial to designing appropriate treatment plans that consider the patient’s unique symptom profile and that provides holistic care relevant to the success of physical therapy. In general, physical therapists do not give adequate attention to cognitive assessment in clinical practice. However, practitioners must be aware that the clinical assessment of cognitive function is no less important than a physical assessment of patients with MS, because it helps to determine the site of impairment and its relationship with physical abilities [21, 25, 26].

Previous studies have shown that there is a relationship between MD and CI in other neurological diseases, such as Alzheimer’s and Parkinson’s disease [27,28,29]. To date, similar previous studies on multiple sclerosis have focused on examining the relationship between motor and cognitive deficits [10, 21, 25, 30,31,32,33,34] and its negative effect on the overall functioning of MS adults [15]. Moreover, a retrospective study has revealed a significant correlation between cognitive function and a decrease in walking speed and an increase in the frequency of falls in patients with MS. However, these findings about walking speed were documented by nurses in the clinic visit and also the data of the repetition of falls were self-reported by patients or/and their family members based on their memory of the recent falls [34]. In addition, an observational pilot study reported that recurrent fallers have higher CI compared to single-time fallers among elderly people with MS [22]. While most studies focus on cognitive impairment as a predictor of the risk of fall, walking speed [34], gait variability [35], and postural control [32], it remains unclear whether the muscle strength, motor coordination, balance, gait, and/or risk of fall can be considered as predictors of CI in adults with MS.

Therefore, the aims of this study were to: (1) investigate the association between CI and muscle weakness, motor incoordination, poor balance, gait abnormality, and high fall risk in patients with MS; and (2) examine if muscle weakness, motor incoordination, poor balance, gait abnormalities, and/or increased fall risk can be adopted as the best indicator of CI in patients with MS.

Methods

Participants and inclusion and exclusion criteria

Seventy patients with MS were recruited for this cross-sectional study over 6-month period from July 2020 to January 2021. The sample size was estimated based on the G*Power3 analysis program when the effect size = 0.15, α error probability = 0.05, power (1-β error probability) = 0.8, and the number of predictors = 3 [36] which is reported according to STROBE guidelines [37] (Additional file 1). Patients were assessed in the Rehabilitation Department of King Khalid University Hospital, Physical Therapy Clinics of King Saud University and Sultan Bin Abdulaziz Humanitarian City. All measurements were conducted by a trained physical therapist with over 12 years of experience in rehabilitation of MS. Adults aged between 20 and 60, with a diagnosis of MS, able to read and write, and able to walk with or without assistive aids were included. Patients with communication difficulties, severe muscle weakness and spasticity, vision and hearing problems, severe CI, relapse within the last 3 months, neurological diseases other than MS, and/or a history of psychiatric illness were excluded.

Instruments and procedures

Montreal Cognitive Assessment (MoCA)

The MoCA is a brief test that is used to assess cognitive function with a universal cutoff point of 26 to detect CI [38, 39]. The MoCA assesses several cognitive domains: (i) executive function, which involves an alternating trial-making task, a verbal fluency task, and a verbal abstracting task; (ii) the short-term memory domain, consisting of two trials of five words, learning and recalling them after about 5 min; (iii) the visuospatial domain, which includes drawing a clock and copying a cube; (iv) the language domain, covering naming, repetition of two sentences, and verbal fluency; (v) the attention assessment, which involves reading a list of digits forwards and backwards and serial subtraction; and the (vi) vigilance and (vii) abstraction domains as well as (viii) orientation to time and place [38]. In this study, the Arabic version of the MoCA was used by an examiner trained and certified in the MoCA to assess cognitive function [40]. The final score was the total of all correct points, and an additional point was given if the participant’s educational level was equal to or less than 12 years [38].

Muscle strength: handheld dynamometers

Isometric muscle strength was measured using “Jamar hydraulic hand dynamometer (Lafayette Instruments, Lafayette, Indiana, USA)” [41] for hand grip and “hand-held dynamometer (Commander Muscle Tester, JTech, USA)” [42] for knee extensor muscles. Three readings were recorded, and the highest value was selected, as previously [43]. For handgrip strength, the participant was in a sitting position and was asked to squeeze the dynamometer with the shoulder adducted and the elbow flexed 90°. For knee extensors, the participant was in a sitting position with the knee in 90° flexion. The examiner held the handheld dynamometer on the anterior lower third of the participant’s leg during isometric knee extension.

Lower extremity coordination test: timed alternate heel-to-knee test

In an alternative heel-to-knee test, the participant was in a supine position and was asked to bend the knee and drag the heel of the tested leg to reach the level of the contralateral knee, then extend the tested knee completely. The examiner used a stopwatch to time in seconds how long it took to finish ten repetitions for each side as quickly and accurately as possible [44, 45]. Two trials were performed for each side, and the faster one was selected [46].

Upper extremity coordination test: timed rapid alternating movement

The participant was in a sitting position and was asked to perform the test for each side separately after verbal instructions and a visual demonstration performed by the examiner. The participant was asked to perform alternate supination–pronation movements ten times as quickly and accurately as possible. The examiner used a stopwatch and counted the time in seconds from the beginning until the end of ten repetitions. The participant was asked to perform two trials, and the faster one was selected [46].

Tinetti Performance Oriented Mobility Assessment (POMA)

The POMA is a performance-based test to assess both balance (POMA-B) and gait (POMA-G) separately, and the total score indicates the fall risk [47, 48]. The POMA consists of nine tasks for balance assessment and seven tasks for gait assessment, with each task given a score of 0, 1, or 2, where 0 implies a low independence level and 2 a normal independence level [48]. The fall risk is considered high when the total score is ≤ 18, moderate with a score between 19 and 23, and low when it is ≥ 24 [48]. For balance assessment, the subject sat on a hard, armless chair and followed the examiner’s instructions. The assessment included sitting balance, rising, attempts to rise, immediate standing balance within the first five seconds, standing balance, nudged, eyes closed while standing, turning 360°, and sitting down [48]. For gait assessment, the subject had to walk 15 feet at their usual speed and back at safe rapid speed; participants could use an assistive device in this task. The assessment included gait initiation, step length and height, step symmetry, step continuity, path, trunk sway, and walking distance [47].

Statistical analysis

All statistical analyses were performed using SPSS (v23, IBM Statistics, Armonk, NY). Participant characteristics were presented by descriptive analysis. A normality test was conducted on all variables to select the appropriate (parametric or non-parametric) statistical test. Pearson’s correlation coefficient analysis was used to assess the strength and direction of the association between cognition, muscle strength of knee extension, and motor coordination of upper extremity variables. The Spearman correlation coefficient was computed to assess associations between cognition, muscle strength of the hand grip, motor coordination of lower extremity balance, gait, and risk of fall. In addition, the Spearman correlation coefficient was used to assess the correlation between the MoCA domains and all motor variables. Motor variables significantly correlated with the MoCA were used in a multiple linear regression to determine the predictor variables [33, 49]. All data were expressed as mean ± standard deviation (SD), mode, median, or frequency and percentage, as required for each data type. The significance level was a p value ˂0.05 with a 95% confidence interval (CI).

Results

Participants and demographics

Seventy subjects were enrolled, 55 females and 15 males, with a mean age of 36.60 ± 9.24 years. The participants’ demographic characteristics are presented in Table 1.

Table 1 Participant demographics
Full size table

Cognitive and motor variables

All cognitive function and motor variable data are summarized in Table 2. Memory was the most affected domain (in 84.3% of cases), followed by executive functions and visuospatial skills at 74% and 71.4%, respectively. Language was only affected in 26% of cases. The worst score from both sides was selected for all motor variables examined bilaterally.

Table 2 Descriptive analysis of the cognitive and motor variables
Full size table

Correlation between cognitive performance (MoCA) and motor function

As highlighted in Table 3, the correlation analysis was computed between Motor variables including muscle strength of handgrip and knee extension, motor coordination of upper and lower extremities, balance, gait, and the risk of fall.

Table 3 Correlations analysis between cognitive and motor scores
Full size table

Motor predictor variables

Motor variables significantly correlated with the MoCA were used in a multiple linear regression analysis to determine predictor variables [50, 51]. Balance (POMA-B) and gait (POMA-G) were excluded from the linear regression, because they were highly correlated with each other (ρ = 0.822, n = 70, p ˂ 0.001), balance (POMA-B) with fall risk (POMA total score) (ρ = 0.931, n = 70, p ˂ 0.001), and gait (POMA-G) with fall risk (POMA total score) (ρ = 0.961, n = 70, p ˂ 0.001). Strength of hand grip and knee extension were also excluded, because they were not significantly correlated with the MoCA. Using the total POMA score (fall risk), upper and lower limb coordination variables were entered into the regression analysis as independent variables and the MoCA as the dependent variable. Using the stepwise method, a significant model 1 emerged (F (1, 68) = 14.150, p < 0.001) with an R2 of 0.172 (Table 4). Therefore, 17.2% of the variance in the MoCA was predictable from the overall POMA score (fall risk). In model 2 (F (2, 67) = 9.814, p < 0.001), the R2 increased from 0.171 to 0.227 after adding the supination/pronation test (Tables 4 and 5). Thus, 22.7% of the variance in the MoCA was predictable from the overall POMA score (fall risk) and supination/pronation test (Fig. 1). The equation describing the best fit of the regression is

Table 4 Summary of multiple linear regression analysis: dependent variable–MoCA total score
Full size table
Table 5 Summary of stepwise multiple linear regression analysis for predicting cognitive impairment (MoCA score)
Full size table
Fig. 1
figure 1

Scatter plot of the MoCA on the overall POMA score (fall risk) and upper extremity coordination score with regression line and individual 95% CIs

Full size image
$$MoCA \, = \, 24.325 +(0.147 \times \left(Fall Risk\right)+(-0.539 \times \left(Supination\setminus Pronation\right))$$

A one-unit increase in the total POMA score increased the MoCA score by 0.147, which was significant (t (67) = 2.216, p < 0.001) (Table 5). In addition, a one-unit increase in the supination/pronation score decreased the MoCA score by 0.539, which was also significant (t (67) =−2.169, p < 0.001).

Discussion

The results of this study showed that CI is significantly associated with motor incoordination of the upper and lower extremities, balance deficits, gait abnormalities, and fall risk. However, there was no significant association between cognition and muscle strength, as estimated by hand grip and knee extension strength. Moreover, reduced upper and lower extremity coordination scores and lower fall risk were associated with higher CI scores. Specifically, the fall risk and upper extremity incoordination scores were predictive of CI in MS patients.

The results confirmed that the mean total MoCA score indicated mild cognitive impairment and in about two thirds of patients with CI, consistent with previous studies [52, 53]. Moreover, several studies have shown that the most affected cognitive domains in MS patients were memory, abstract/conceptual reasoning, information processing, attention, and visuospatial skills [26, 54]. Similarly, we found that memory, visuospatial skills, executive function, attention, and language were affected in patients with CI based on the MoCA scale.

This study found no association between isometric knee extensor strength and CI. Conversely, a recent study reported a significant correlation between CI and knee extension strength in MS patients [33]. Sandroff et al. [55] reported that the peak torque of knee extension was associated with cognitive processing speed but not verbal and visuospatial learning and memory in 62 people with MS. These contradictory results might be due smaller sample sizes or the type of dynamometer used. Similar to other studies, there was no correlation between hand grip strength and CI as measured by the total MoCA score [33], but there was a significant correlation between hand grip strength and the language and attention domains.

The current study revealed significant correlations between motor incoordination of the upper and lower extremities and CI in patients with MS. Poor performance on rapid alternating movement and heel-to-knee tests is usually due to cerebellar dysfunction [56]. Furthermore, several studies have shown that the cerebellum plays an important role in both the cognitive and motor functions of MS patients [56,57,58]. This might be because the cerebello-cerebral network, consisting of the forward cortico-ponto-cerebellar pathway and the backward cerebello-thalamo-cortical pathway, also modulates cognition [59, 60].

A balance component of POMA was significantly moderately correlated with cognitive performance in MS patients. This is the first study examining the association between cognition and balance in patients with MS using the POMA-B scale. Alfonso et al. [61] similarly concluded that cognitive impairment in adults with MS including processing speed deficit was associated with difficulty selecting appropriate corrective movements to compensate for balance disturbances. This speculation is consistent with a previous study that found postural control becomes more difficult in MS individuals due to somatosensory and visual input deficits. Consequently, additional compensatory mechanisms are needed to overcome this postural instability which also requires the support of several cognitive domains [62]. Moreover, other studies have examined postural–cognitive interference by incorporating the dual-task paradigms in healthy people and those with MS [63, 64]. A recent review concluded that patients with MS showed impaired balance when they simultaneously performed a cognitive and postural task, which increased their fall risk in most daily activities [63].

The gait component of the POMA was moderately positively correlated with cognition, consistent with previous studies demonstrating a significant association between gait abnormalities and cognitive decline in people with MS [31, 35]. A previous study reported that step length and step time variabilities were associated with cognitive processing speed in MS patients [31]. It is interesting to note that no study used POMA-G for the assessment of gait and its correlation with cognition in MS individuals. The current study is, therefore, unique in its use of a simple, short, inexpensive scale that does not need specialist equipment in clinical practice.

Fall risk was significantly correlated with cognitive impairment and predicted CI in MS individuals, consistent with previous studies [34, 65]. Moreover, a higher POMA score was associated with a decrease in fall risk, which was associated with higher MoCA scores. This is particularly important when investigating the fall risk during motor examination in patients with MS and provides a clue about the presence of cognitive impairment. Furthermore, fall frequency has been shown to be significantly correlated with general intelligence, speed of cognitive processing [22], and executive functioning, while verbal memory was found to be a significant predictor of falls in patients with MS [34].

This study has several limitations. The sample size was small, and further studies are required with a larger sample size to improve the robustness of regression. Most of the participants had relapsing–remitting MS which limiting the generalizability of the results on the whole MS population and future studies should involve all MS subtypes. The POMA-G used in this study is based on the examiner’s observations and is less sensitive for detecting gait abnormalities than other instruments. Therefore, future studies using advanced gait analysis systems are now needed to better detect temporal and visuospatial parameters.

Conclusions

CI is significantly associated with motor incoordination, poor balance, gait abnormalities, and increased fall risk. The fall risk and upper extremity incoordination were the best indicators of CI in patients with MS. Thus, motor assessment can provide physical therapists with clues about the presence of CI in patients with MS. In addition, incorporating coordination and balance training into the rehabilitation program may enhance cognitive functions in patients with MS, although this requires empirical testing.