Prevalence and clinical predictors of spasticity after intracerebral hemorrhage

Abstract Background Spasticity is a common complication of intracerebral hemorrhage (ICH). However, no consensus exists on the relation between spasticity and initial clinical findings after ICH. Methods This retrospective study enrolled adult patients with a history of ICH between January 2012 and October 2020. The modified Ashworth scale was used to assess spasticity. A trained image analyst traced all ICH lesions. Multivariable logistic regression was used to examine the association between ICH lesion sites and spasticity. Results We finally analyzed 304 patients (mean age 54.86 ± 12.93 years; 72.04% men). The incidence of spasticity in patients with ICH was 30.92%. Higher National Institutes of Health stroke scale (NIHSS) scores were associated with an increased predicted probability for spasticity (odds ratio, OR = 1.153 [95% confidence interval, CI 1.093–1.216], p < .001). Logistic regression analysis revealed that lower age, higher NIHSS scores, and drinking were associated with an increased risk of moderate‐to‐severe spasticity (OR = 0.965 [95% CI 0.939–0.992], p = .013; OR = 1.068 [95% CI 1.008–1.130], p = .025; OR = 4.809 [95% CI 1.671–13.840], p = .004, respectively). However, smoking and ICH in the thalamus were associated with a reduced risk of moderate‐to‐severe spasticity (OR = 0.200 [95% CI 0.071–0.563], p = .002; OR = 0.405 [95% CI 0.140–1.174], p = .046, respectively) compared with ICH in the basal ganglia. Conclusions Our results suggest that ICH lesion locations are at least partly associated with post‐stroke spasticity rather than the latter simply being a physiological reaction to ICH itself. The predictors for spasticity after ICH were age, NIHSS scores, past medical history, and ICH lesion sites.


INTRODUCTION
Spasticity is a common complication of stroke with a pooled incidence of 25.3% (Nam et al., 2019;Zeng et al., 2020). A velocity-dependent increase in resistance during passive stretching is a characteristic of spasticity, which is due to the hyperexcitability of the stretch reflex (Li & Francisco, 2015). Muscle spasticity as a result of cerebral hemisphere injury can lead to long-term disability (Li, 2020;Zeng et al., 2020). The onset of spasticity is highly variable and may occur shortly or more than 1 year after stroke (Ward, 2012). Katoozian et al. (2018) identified a 2.5 times higher risk of spasticity in patients with hemorrhagic stroke than in patients with ischemic stroke during the 3 months post-stroke.
A recent meta-analysis also identified hemorrhagic stroke as a risk factor for post-stroke spasticity (odds ratio, OR = 1.879 [95% confidence interval, CI 1.418−2.490], I2 = 27.3%) (Zeng et al., 2020). However, the role of hemorrhagic stroke in predicting post-stroke spasticity requires further exploration. Moreover, hemorrhagic stroke may not be an independent risk factor for post-stroke spasticity, and damage to the basal ganglia region may lead to greater spasticity in patients with hemorrhagic stroke (Jin & Zhao, 2018;Zeng et al., 2020). It remains uncertain whether the risk relates to the location of the intracerebral hemorrhage (ICH) or indicates a reaction to the ICH lesion itself. Although BoNT-A is an established treatment for focal spasticity (Francisco et al., 2021), advanced well-defined spasticity management strategies have not been conclusively established (Mahmood et al., 2019;Salazar et al., 2019). Patients with spasticity who have low motor recovery potential may benefit from aggressive treatment and care (Li, 2017); therefore, clarifying an association between post-ICH spasticity and ICH lesion location would be beneficial for correctly following up highrisk patients and effectively guiding resources (Zeng et al., 2020).
Herein, we hypothesized that the incidence and degree of post-stroke spasticity would differ among ICH lesion sites, with the highest incidence in patients with lesions in the basal ganglia region. Moreover, despite its high incidence, no consensus exists (Nam et al., 2019) on the number of patients who develop spasticity after ICH and (Zeng et al., 2020) the relation between spasticity and initial clinical findings after ICH. Therefore, this study aimed to investigate the prevalence of spasticity after ICH and identify clinical predictors of subsequent spasticity.

Lesion tracing
Compared with the relative contralateral brain region, lesions were visually identified as having altered fluid-attenuated inversion recovery (FLAIR) and diffusion-weighted imaging (DWI) signal intensities.
We used DWI sequences for MRI within 48 h post-stroke and FLAIR sequences for MRI within 48 h to 7 days post-stroke (Picelli et al., 2014). A trained image analyst traced all lesions, and an experienced clinical neurologist confirmed them. Both were blinded to all clinical data, except for the side of the hemiparesis.

Spasticity measure
Spasticity assessments of the upper/lower extremities and trunk were performed. For MAS, higher grades equated to more severe spasticity. The MAS is considered to provide reliable and reproducible results (Bohannon & Smith, 1987). The assessments were done by a trained physiotherapist at a 1-year follow-up. Patients with MAS scores ≥1 in any one joint were considered to have spasticity.

Statistical analysis
All variables in the study were calculated using descriptive statistics.

RESULTS
In total, 304 patients with ICH (mean age 54.86 ± 12.93 years; 72.04% men) were eligible for the study. The patient characteristics are shown in S1. Among the 304 patients, 94 experienced spasticity (31%). The results indicated that patients with spasticity were younger than those without spasticity (p = .014; Table S1). NIHSS scores were higher in ICH patients with spasticity (p < .001; Table S1).
The spasticity characteristics at different lesion locations are shown in

DISCUSSION
This analysis of patients with ICH found that the occurrence of poststroke spasticity was associated with NIHSS scores, and the degree of spasticity was related to age, NIHSS scores, smoking, drinking, and lesion location. Our findings are consistent with those of previous studies showing that NIHSS scores are significant clinical predictors of post-stroke spasticity (Glaess-Leistner et al., 2021;Schinwelski et al., 2019). Moreover, predictors for the development of spasticity after ICH were age, NIHSS scores, past medical history, and ICH lesion sites.
The relationship between age and spasticity has never been specifically explored in stroke, although it has been identified as an important predictor of stroke severity, risk factors, and complications (Lundstrom et al., 2008;Shin et al., 2018;Tedesco Triccas et al., 2019). Our analysis showed that moderate-to-severe post-stroke spasticity is more likely to occur in the younger population, which is consistent with the results of Lundström et al. (2008). However, several studies have reported an association between increasing age and risk for increasing muscle tone (Persson et al., 2020;Shin et al., 2018;Tedesco Triccas et al., 2019).
Further research should be conducted to investigate the influence and mechanism of age on spasticity.
In our study, patients with more severe spasticity had higher NIHSS scores. These results reflect those of Shin et al. (2018), who also found that the NIHSS had the most significant correlation with the aggravation of spasticity between 3 and 12 months after stroke. Therefore, as mentioned by Opheim et al. (2015), the assessment of NIHSS scores at admission may provide a good indication of the probability of a patient developing spasticity.
Studies have shown that smoking and drinking were associated with stroke (O'Donnell et al., 2016;Yang et al., 2021), while their association with post-stroke spasticity was unclear. Our research identified smoking history as a predictor for more severe disability from spasticity. This finding was also reported by Leathley et al. (2004). Moreover, our results revealed that a history of drinking was associated with severe spasticity. However, we could further examine the relationship between smoking/drinking and spasticity because the study data did not include a detailed smoking/drinking history. Future studies examining post-stroke spasticity should consider smoking/drinking history.
Spasticity is a symptom of an upper motor neuron syndrome (Tedesco Triccas et al., 2019), which occurs due to a change in the net balance of excitatory and inhibitory inputs on spinal-level circuitry secondary to injury in higher-order centers or their descending pathways (i.e., pyramidal or parapyramidal fibers) (Li & Francisco, 2015 (Li & Francisco, 2015). Moreover, the dorsal RST provides a powerful inhibitory effect on the spinal stretch reflex (Li et al., 2019), and abnormalities in RST outflow are considered to play a major role in the genesis of spasticity in humans (Li & Francisco, 2015).
Therefore, damage to the brainstem can result in the highest incidence of spasticity. Another important finding was that the risk of moderateto-severe spasticity after thalamic hemorrhage was reduced compared to that in the basal ganglia. These results may be due to imbalanced descending regulations (Li & Francisco, 2015). The corticospinal tract (CST) and corticoreticular tract are adjacent to each other in the corona radiata and internal capsule (Jang & Lee, 2019;Yeo et al., 2020); thus, damage often occurs in both the CST and corticoreticular tract, resulting in the loss of cortical facilitatory input to the medullary inhibitory center (Mukherjee & Chakravarty, 2010). Moreover, the CST is one of the most important descending pathways of the human motor system. Therefore, spastic hemiplegia is frequently observed in patients with basal ganglia ICH.

LIMITATIONS
The present study was a cross-sectional analysis and not a longitudinal analysis; thus, it had some limitations. First, it was a single-center study though the study has an appropriate sample size; therefore, the results may not be generalizable to all patients. Moreover, some patients with ICH may present with multiple lesions. Such cases might have attenuated the differences in spasticity observed across ICH lesion locations. Further, limited clinical information was available (e.g., detailed smoking/drinking history), so we could not further examine the relationship between smoking/drinking and spasticity. Finally, the predictors identified in this study only appear to explain some of the variability of the incidence/severity of post-stroke spasticity, indicating other contributing factors that we did not consider.