*3.3. HGS*

The mean HGS of both hands improved significantly at postoperative 3 months and 1 year, compared with the preoperative measures, in all patients (*p* = 0.000 for both; ANOVA) (Figure 2 and Table 4). A significant difference was found between the male and female groups for every measure (*p* = 0.000; independent *t*-test).

**Figure 2.** The mean HGS of both hands improved significantly at postoperative 3 months and 1 year compared with the preoperative measures in all patient groups (*p* = 0.000 for all; ANOVA). \* indicates the statistical difference between measures.


**Table 4.** Hand grip strength measurements.

Statistical analyses were performed by independent *t*-test comparing the male and female groups.

#### *3.4. Multiple Regression Analyses of Parameters Associated with Falls and Fall-Related Mobility Tests*

In male patients, preoperative falls were correlated positively with symptom duration (beta ± standard error = 0.003 ± 0.001, *p* = 0.000) and mFi (beta ± standard error = 0.362 ± 0.043, *p* = 0.000) and negatively with EQ-VAS (−0.002 ± 0.001) and HGS (beta ± standard error = −0.004 ± 0.001, *p* = 0.000). Falls at postoperative 3 months and 1 year were not correlated with any parameter.

In female patients, preoperative falls were correlated negatively with mJOA score (beta ± standard error = −0.057 ± 0.015, *p* = 0.000) and HGS (beta ± standard error = −0.035 ± 0.005, *p* = 0.000). At postoperative 3 months, number of falls was positively correlated with mFi (beta ± standard error = 0.246 ± 0.017, *p* = 0.000) and NDI (beta ± standard error = 0.050 ± 0.006, *p* = 0.000). Fall measures at 12 months were positively correlated with NDI (beta ± standard error = 0.066 ± 0.000, *p* = 0.000) and WC (beta ± standard error = 0.046 ± 0.000, *p* = 0.000) and negatively with HGS (beta ± standard error = −0.049 ± 0.000, *p* = 0.000). Other correlations with functional mobility tests are listed in Table 5: Correlations between parameters not presented in Table 5 lacked

statistical significance. Additionally, univariate linear regression analyses were presented in the Table S1.

**Table 5.** Multiple regression analyses of fall-related functional mobility tests.


\* modified JOA grade (16~17 = Grade 0, 12~15 = Grade 1, 8~11 = Grad 2, 0~7 = Grade 3, higher grades reflect worse functional status); Neck Disability Index (NDI) (=higher scores indicate worse functional status), Euro-QoL Visual Analog Scale (VAS) (=higher scores represent better QoL status), modified Japanese orthopedic association (JOA) score (=higher scores reflect better functional status).

#### **4. Discussion**

Surgical treatment for DCM is associated with improvements in functional, disabilityrelated, and QoL outcomes and reduced incidences of both falls and fall-related deterioration of subjective symptoms [5,32,33]. Compared with lumbar stenosis, the lack of

data on DCM patients and the related risk of falls therein makes it difficult to predict surgical outcomes and postoperative rates of improvement in preoperative neurologic deficits. Additionally, prior studies that have characterized grip strength in association with myelopathic symptoms have presented mixed evidence with postoperative improvement, or no difference [34–36].

Compared with a recently published lumbar stenosis study, the present study confirmed differences in correlations between male and female sex and the postoperative risk of falling [19]. The previous study excluded cervical stenosis patients with upper-extremity motor deficits to focus on the sarcopenic conditions of the patients [37]. The present study focused on cervical myelopathy-related HGS weakness and postoperative functional changes according to sex. As expected, differences between the male and female groups were observable. Meanwhile, different from other available studies, all of the enrolled patients developed cervical myelopathic symptoms, and more than half also showed spinal cord signal changes (65.5%; 133/203). We confirmed that the spinal cord signal changes were not necessarily correlated with actual falls and other outcomes, such as functional mobility tests and QoL (Table 5), and the direction of correlations varied from positive to negative depending on the measured time and the sex, a finding that is consistent with the literature [38]. Healing of the spinal cord after surgical decompression is based on the intrinsic ability of the spinal cord to heal itself. Thus, the pre-operative health of the cord is paramount to post-operative improvement [39]. For the enrolled male and female patients in the present study, preoperative status, including the general condition and duration of symptoms (Table 1), could differ, and these could affect the observed variations in correlations with fall and fall-related parameters. Although there was a negative correlation between postoperative fall-related functional tests and HGS in female patients, it was smaller than that in the male patients in this study.

Along with HGS, the present study demonstrated sex differences in the recovery of QoL reflected in the outcomes and related functional mobility results. For male patients, because baseline HGS and muscle strength are much greater than those in female patients, a higher increase in HGS was expected postoperatively. Although a lesser amount of recovery of HGS and related function was observed in the female group by postoperative 3 months, the larger delayed recovery between postoperative 3 months and 1 year (Table 4) could lead the patients and medical team to encourage functional rehabilitation to improve muscle strength and lower the risk of falling up to postoperative 1 year. [40].

In a study by Kalsi-Ryan et al., [14,15] a more specific hand assessment study was suggested. Unfortunately, in this study, the patients were enrolled from March 2017 to August 2019, and therefore the specific test was not ye<sup>t</sup> available. The authors believe that the hand assessment study would be better to describe upper extremity function in DCM patients in future studies.

The surgical effect of decompression in patients with DCM could differ in relation to a variety of factors. Since HGS improved after surgical decompression, the recovery of HGS was not only related to preoperative HGS but also to the overall functional outcome originating from compressive myelopathy-related pyramidal tract dysfunction [36]. Improved concordant motor function and muscle coordination with the resolution of myelopathy symptoms postoperatively elicited better functional mobility tests related to the risk of falling and actual falls [5,32].

The key findings of the present study are the following: postoperative HGS may be correlated with postoperative falling and functional outcomes differently in male and female patients. Meanwhile, surgical intervention for DCM significantly reduced the incidence rate of falls to less than 40% of the preoperative rate. The incidence of falls decreased significantly from 17.2% (35/203) to 6.8% (14/203) after surgery. Frequent falling is one of the most common symptoms in patients with DCM, and our analysis revealed that the incidence of both actual falls and multiple falls decreased significantly during postoperative follow up (Table 2). The decrease in actual falls during follow-up, however, made multiple regression analyses thereof in relation to other parameters impossible.

In another study, the incidence of postoperative falls peaked at 5 to 6 months after surgery, likely because many patients may have increased their daily walking activity during this period, leading to a transiently increased fall rate [5]. However, only a limited number of patients fell during follow up and no aggravation of symptoms and related fractures were reported in the present study. This finding could be explained by the peri- and postoperative fall prevention education program provided by our institution to emphasize the risk and caution of postoperative falls to patients and caregivers during admission and at every outpatient clinic follow up, based on previous publications [2–4,19].

Another possible reason for the decreased number of falls during follow-up could be the low BMI (mean: 24.30 ± 3.82 kg/m2) of the enrolled patients. A higher BMI is an independent risk factor for falls, and an association between increasing BMI (ranging from 25.0 to 29.9 kg/m<sup>2</sup> and 30.0 kg/m<sup>2</sup> and higher) and the risk of falls has been reported [41]. However, no significant association was found between increasing BMI and fall-related injury in the present study: correlations between functional mobility tests and BMI are presented in Table 5.

Our study had several strong points compared with previous studies. We evaluated a comprehensive range of risk factors, including the duration of symptoms and comorbidity. As the general condition of the patients is related to the preoperative and postoperative recovery of function, the overall condition of the patients is an important factor [42,43]. Additionally, we included more severe spondylotic myeloradiculopathic cases that had undergone combined anterior–posterior surgery [44–46], and as such the rate of combined anterior–posterior surgeries was much higher than that in another study [5]. Moreover, we report not only actual falls but also the objective measures of functional mobility tests and HGS, which all affect patient subjective symptoms.

A limitation of the present study was that the radiologic factors for the risk of falling were not reported at the same time. However, regarding the functional evaluation in the present study, all parameters, including mFi and HGS, would help clarify the postoperative recovery patterns of DCM patients. The results concerning radiologic evaluation and analyses are now being prepared for a future study. Despite these limitations, this is the first study to analyze correlations between HGS and the risk of falls in relation to functional tests and actual falls, as well as QoL, in DCM.
