Spinal muscular atrophy (SMA) is an autosomal-recessive degenerative neuromuscular disease affecting lower motor neurons due to the absence of the survival of motor neurons (SMN1) gene (1). This results in progressive denervation and muscle weakness, predominately affecting the proximal muscles (2). SMA occurs in 1 out of 10,000 to 20,000 live births (3, 4), with age of onset and clinical severity often correlated with the number of survival motor neurons 2 (SMN2) copies (5).

Those individuals with one or two SMN2 copies are typically non-ambulatory, whereas individuals with three to four SMN2 copies can walk either household or community distances (6). Although this trend is noted, recent systematic review suggests that SMN2 copy number alone does not predict walking ability (7). Studies have also found that sex, along with SMN2 copy number, is a predictor of disease severity related to ambulatory function, with males more severely affected than females (5, 8). Although it is not clear how sex-based differences are affected by ambulatory function, specifically.

As the disease progresses, individuals with SMA may lose their ability to ambulate due to weakness, fatigue, or injuries resulting from falls. So far, clinicians have not been able to accurately predict loss of ambulation in SMA. A history of falls is often reported prior to complete loss of ambulation (9, 10). Limited research assessed the association between ambulation and falls in SMA, including a multiple-case study design of five adults and two children showing that strength, gait velocity, and variability in gait pattern were not predictors of falls but stride-length variability was a significant risk factor (9). In contrast, gait speed is a validated predictor of fall risk in other neurological diseases such as Huntington’s disease (11, 12), Parkinson’s disease (13, 14), and stroke (15).

Within neuromuscular disease, there is limited data on gait speed in adults. One study assessed children with Charcot–Marie Tooth Disease finding a gait speed of 0.94 m/s, significantly slower when compared to healthy controls (16). Another study found that gait speed was slower and stride length was shorter in children with neuromuscular disease (Duchenne Muscular Dystrophy, Becker Muscular Dystrophy, and Charcot–Marie-Tooth) compared to typical developing peers (17). In healthy adults between 30 and 39 years old, gait speeds tend to range from 1.32 to 1.55 m/s in men and 1.19 to 1.48 m/s in women (18). Clinicians use the 10 Meter Walk Test or 6 Minute Walk Test (6MWT) to assess gait speed and stratify an individual as a community or household ambulatory. In healthy adults, those with a gait speed of <0.8 m/s were found to be at an increased risk of falling and were classified as household ambulators (19). In an upper and lower motor neuron disease such as amyotrophic lateral sclerosis (ALS), changes in gait including shorter stride length, increased stride duration, slower walking speed, and increased gait variability were noted in comparison to healthy adults (20–23). These spatial–temporal changes in gait were associated with lack of independence, lower quality of life, and increased risk for falls (24). A study investigating the changes in lower extremity function across clinical phenotypes in ALS also found that a history of falls serves as an early indicator of disease progression and severity (25). Assessing fall risk in the neuromuscular population is important as it can signal a progressive loss in ambulation (25).

Fatigue, defined as an overwhelming sense of tiredness, lack of energy, feeling of physical or mental exhaustion or both (26–28), is reported in 81–100% of patients with SMA and found to be a major barrier for participation in physical activities, recreational activities, and overall mobility (29–31). The term fatigue can be broken down into two domains: physical and mental fatigue (26). Physical fatigue can be captured as fatigability which can be measured by the performance on an outcome measure (26). Studies have used the 6MWT to capture rate of speed decline across the minutes of the test as a means of capturing fatigability in ambulant patients. One study found a 17% decline in speed between the first minute and the last minute of the 6MWT in individuals with SMA, capturing fatigability during ambulation (32). In a healthy cohort, a study found a typical percent decline in speed from the first minute to the last minute during the 6MWT was 1.4% (33).

However, to date, no study has been able to associate fall risk with a quantifiable value in SMA. The purpose of this study was to investigate falls in adults with genetically confirmed 5q-SMA and the predicting factors associated with falls. We hypothesized that individuals with reported falls would express greater fatigability and slower gait speed when compared to individuals who are not falling.

Fourteen ambulatory patients (9 men, 5 women; median age 33.50; range 17–48) with SMA met enrollment criteria and were classified as fallers (F_all, n = 10) and non-fallers (NF_all, n = 4) based on subjective report during clinic visits. Patients had clinic visits every 6 months for a minimum of 2 years. The average follow-up time was 36 months (range 12–66 months). All participants were receiving Nusinersen except for one in the fall cohort who was receiving Risdiplam consistently throughout the entire study period. None of the participants used assistive devices during the 6MWT. However, in the community for longer distances, 60% of F_all cohort used wheelchairs compared to 25% in the NF_all cohort. Age, sex, initial age at symptom onset, and BMI were similar between the falling and non-falling cohorts as shown in Table 1. All patients with 3 SMN2 copies were in the fall cohort, whereas people with 4 or more copies were distributed between the two cohorts.

No falls or injuries occurred during the testing. The distances walked and corresponding gait speeds at 6 min between the falling and non-falling groups were not associated with predicting fall risk as shown in Table 1.

In the F_all cohort, 7 of the 10 patients (70%) had multiple falls over the course of the study period, with two (20%) of those individuals having falls sustaining serious lower extremity fractures resulting in hospitalizations and loss of ambulation. Per physical therapy and physician documentation, no other serious injuries or hospitalizations occurred in the remaining patients.

The gait speeds across minutes 1, 2, and 6 in the 6MWT are illustrated in Figures 1, 2 for both cohorts. We compared the median pre-fall gait speed in the F_all cohort to the slow and fast median gait speeds of the NF_all cohort to decern if there was a range of speeds where patients are at a higher risk for falls. Despite considerable overlap between pre-fall gait speeds for fallers and median speeds for non-fallers, a distinct trend emerged. In the F_all cohort, the median speeds at 1, 2, and 6 min, measured at the visit preceding the fall, were 0.92 m/s, 0.89 m/s, and 0.77 m/s, respectively, with one individual reaching 1.32 m/s at 6 min. For non-fallers, the median slowest speeds were 1.18 m/s, 1.11 m/s, and 1.09 m/s, and the median fastest speeds were 1.39 m/s, 1.37 m/s, and 1.31 m/s at minutes 1, 2, and 6, respectively. Therefore, we found that the slowest median speeds at 1, 2, and 6 min for non-fallers were nominally faster than the speed recorded prior to a fall. To further explore the distribution of gait speeds at 1, 2, and 6 min between fallers and non-fallers, the median gait speeds for each participant were calculated across study period and compared between the two groups. In the F_all cohort, the median speeds at 1, 2, and 6 min were 0.92 m/s, 0.92 m/s, and 0.85 m/s, respectively. For the NF_all cohort, the median gait speeds across the study period were 1.33 m/s, 1.28 m/s, and 1.25 m/s at minutes 1, 2, and 6, respectively. We therefore noted a trend where individuals who reported falls tended to have gait speeds below 1.0 m/s while those without reported falls tended to have gait speeds above 1.0 m/s shown in Table 1 and Figure 2C.

Similarly, a trend was observed where fallers walk with a slower gait speed than non-fallers. The slowest median speeds at 1, 2, and 6 min were 0.80 m/s, 0.84 m/s, and 0.55 m/s for fallers, compared to 1.18 m/s, 1.11 m/s, and 1.09 m/s for non-fallers as shown in Figure 2A. Whereas the fastest median gait speeds at 1, 2, and 6 min were 0.93 m/s, 1.0 m/s, and 0.95 m/s for fallers, compared to 1.39 m/s, 1.37 m/s, and 1.31 m/s for non-fallers shown in Figure 2B.

Additionally, we captured the percent difference in gait speeds across minutes 1, 2, and 6 during the 6MWT in both groups. Patients who fell showed a nominally greater decline in gait speeds between minutes 1 and 6, as well as 2 and 6, compared to non-fallers during the same intervals, though this difference was not statistically significant. One patient who reported falls demonstrated a 66.37% decline in gait speed from minute 1 to minute 6 in the 6MWT, which is substantially greater than the greatest percent decline in the NF_all cohort of 26.45%.

Our retrospective study assessed the association between reported falls and both gait speed and fatigability in adults with SMA. We cannot draw firm conclusions due to the small sample size of 14 individuals. However, our findings highlight important trends found in the SMA population in regard to describing falls. There was not a statistically significant association between falls and gait speed in this small pilot study; however, we observed a tendency for NF_all cohort to have faster gait speeds compared to F_all cohort. In addition, we found that the F_all cohort had a greater relative decline in gait speed between the first and the last minute during the 6MWT compared to the NF_all cohort. In our study population, the majority of the F_all cohort experienced multiple falls, with some serious injuries resulting in loss of ambulation. Individuals who reported falls were more likely to own and use a wheelchair for community distances than individuals who did not report falls. This may indicate an increased reliance on assistive devices for mobility in those with a history of falls to reduce risk of injury. Further investigation should focus on analyzing gait and fatigue to assess for changes in ambulation that may require assistive devices or intervention to ensure safe ambulation.

Our study assessed fall risk in adults with SMA by examining gait speed and fatigability. A trend emerged indicating that lower gait speeds might increase an individual’s risk of falls. The data also suggested a more pronounced decline in gait speed over the course of the 6MWT among fallers, highlighting fatigability as a possible determinant of fall risk that requires further evaluation. Overall, the findings emphasize the importance of continuing to assess gait speed and fatigability in this population to better predict fall risk. Future research should not only continue to assess gait speed but also expand to include other gait parameters in real time using wearable devices.