Relationship between motor dysfunction, the respiratory muscles and pulmonary function in stroke patients with hemiplegia: a retrospective study

This was a retrospective study that received approval from ShangHai Xuhui Central Hospital Ethics Committee (No.2022-043). Due to the nature of a retrospective study, the need for informed consent was waived by ShangHai Xuhui Central Hospital Ethics Committee. Data were de-identified prior to analysis.

Ultimately the clinical history data of 134 patients at hospital admission, including gender, age, height, weight, disease type, duration of onset, motor function assessment results and pulmonary capacity results were collected and analyzed. In addition, 60 of 134 patients’ motor function assessment and pulmonary capacity results at discharge were obtained.

Through electronic medical records, the data of stroke patients hospitalized in the Rehabilitation Department of Xuhui District Central Hospital between December 2018 and March 2022 were reviewed. Patients had to meet the following criteria to be included in the study: 1) a diagnosis of cerebral infarction or cerebral hemorrhage; 2) met the diagnostic criteria for stroke; 3) hemiplegia patients with a first episode of stroke; 4) aged ≥ 18 years; 5) sitting balance of ≥ grade 1; and 6) had a report of a pulmonary ventilatory function test. Exclusion criteria were: 1) had consciousness disorder and severe cognitive dysfunction; and 2) had acute diseases of the heart, brain, kidney and other organs. The research process is shown schematically in Fig. 1.

Trunk function has been identified as an important early predictor of functional outcomes after stroke. The TIS is a valid tool for examining person with hemiplegia’ trunk control ability, allowing for both qualitative and quantitative assessment of trunk deficits [20]. TIS consists of 3 subscales: namely, static, dynamic sitting balance and trunk coordination in a sitting position, with a total maximum score of 23 [21]. According to previous studies, non-ambulatory patients had a median TIS score of 8 and ambulatory patients had a median score of 14 (11–18) [22]. Although it was not a direct reference basis, we believe it was a relatively appropriate method for our research. We took 8 points and 18 points as two nodes and divided patients into 3 groups based on their TIS scores and classified thus: a patient score of 0–7 points as a severe level, 8–17 points a moderate level and 18–23 points a slight level.

The Berg Balance Scale (BBS) is a comprehensive balance function examination scale, which evaluates the patient’s ability to shift actively their centre of gravity through multiple functional activities, including dynamic and static balance tasks while sitting and standing. It had 14 item list with each item consisting of a 5-point ordinal scale ranging from 0 to 4. 0 indicated the lowest level of function and 4 was the highest level. Scores of 0 to 20 suggested that patients had poor balance and needed a wheelchair. Scores of 21–40 indicated that patients had a certain balance ability and could walk with assistance. Scores of 41–56 showed that patients can walk independently [23].

The BI was used to measure the degree of assistance required for patients’ activities involved in daily living, which included 10 personal activities with a total score of 100 points. Guidelines for interpreting BI scores are: scores of 0–20 indicate “total” dependency; 21–60 indicate “severe” dependency; 61–90 indicate “moderate” dependency; and 91–99 indicate “slight” dependency [24].

Spirometric data were evaluated through a portable pulmonary function instrument (Xeek X1, China) with the patients in a sitting position. The measured value of the parameter was expressed as a percentage of the predicted value. The main collected parameters were as follows: 1) maximal inspiratory pressure (MIP), male normal value ≥ 75% (of predicted value), female normal value ≥ 50%; 2) maximal expiratory pressure (MEP), male normal value ≥ 100%, female normal value ≥ 80%; 3) forced vital capacity (FVC): the total volume of air exhaled with maximum strength and fastest speed; 4) forced expiratory volume in 1 s(FEV₁), evaluated the volume of air exhaled by the patient using the maximum force and the fastest speed in 1 s, which reflected ventilation dysfunction and was dependent on the respiratory muscles and airway status [25]. It is divided into 5 different levels: mild, FEV₁ over 70%; moderate, FEV₁ 60–70%; moderately severe, FEV₁ 50–60%; severe, FEV₁ 35–50%; very severe, FEV₁ < 35% [26]. 5) peak expiratory flow (PEF) reflecting the highest flow rate during forced expiration, which is an important index reflecting the strength of the expiratory muscles [27]; and 6) maximal mid-expiratory flow (MMEF) representing the mean expiratory flow rate at which 25%-75% of vital capacity is exhaled with force.

The data were analyzed using SPSS 23 (SPSS Inc., 233S. Wacker Dr., IL). Categorical data are presented as percentages and continuous data as means ± standard deviation (SD) and medians (interquartile ranges).

Spearman’s correlation coefficient was employed to analyze the relationship between values in motor function scales and respiratory indicators. The Wilcoxon signed-rank test was utilized to analyze any differences in spirometric data among individuals with varying degrees of motor dysfunction, as well as the differences in motor function across different spirometric indexes. Additionally, we used the Minimal Clinically Important Difference (MCID) of TIS, BBS and BI to determine if genuine changes in a patient’s function had occurred [28]. We also analyzed the changes in lung function indexes after 3 weeks of physical therapy intervention. Finally, the study investigated the differences in the change values of spirometric indexes among individuals with different levels of motor dysfunction following the 3-week physical therapy intervention. A significance level (p) of 0.05 was determined for the statistical comparisons.

The results of Spearman’s correlation coefficient analysis are presented in Fig. 2. TIS was shown to have slight positive correlations with MIP, MEP, FVC, FEV₁ and PEF.

BBS showed moderate positive correlations with MIP (r = 0.37, P < 0.001, PEF (r = 0.40, P < 0.001), and slight positive correlations with MEP, FVC, FEV₁ and MMEF. BI indicated moderate positive correlations with MIP (r = 0.33, P < 0.001), MEP (r = 0.32, P < 0.001), FVC (r = 0.3, P < 0.001), and slight positive correlations with MEP, FVC, FEV₁ and MMEF. FEV₁ and PEF and MMEF.

TIS determined a patient’s trunk control ability at 3 levels, namely severe, moderate and slight. The MIP values were 29.0 (20.3–44.0), 33.7 (22.4–45.3) and 38.4 (28.1–59.4) for the 3 levels, respectively with statistical significance (P = 0.029). The MEP was 29.0 (19.4–38.0), 33.3 (23.1–46.8) and 45.8 (29.4–80.5) for the 3 levels, respectively with significant differences (P = 0.002). Significant effects were also found with FVC (P = 0.014) and FEV₁ (P = 0.011) for the 3 levels of TIS (Table 2, Fig. 3a). The findings indicated that different trunk control abilities showed differences in these respiratory indicators. Thus, increasing trunk control ability will clearly help in improving these respiratory indicators.

BBS evaluates a patient’s balance ability at 3 levels, namely severe, moderate, and slight. Table 2 and Fig. 3b illustrate the spirometric data at 3 different balance ability levels. MIP values were 29.1 (18.5–38.0), 36.2 (28.4–47.3) and 43.9 (33.2–64.6), respectively for the 3 levels with statistical significance reached (P < 0.001). The MEP was 28.4 (19.2–39.2), 37.2 (28.9–59.6) and 54.6 (36.1–71.9), respectively for the 3 levels all with significant differences (P < 0.001). Significant differences were also found in FVC (P = 0.002), FEV₁ (P < 0.001), PEF (P < 0.001), and MMEF (P = 0.01) at the 3 BBS levels. The results indicated that different balance abilities reflect differences in these respiratory indicators, thus increasing balance ability should help to improve respiratory indicators.

BI assessed the patient’s dependency on daily living activities, which included 5 levels. This study included 3 levels, namely total, severe and moderate dependency due to the highest BI scores being ≤ 90 points. The results indicated that patients’ respiratory muscle strength decreased with increasing dependency on living activities. MIP values were 20.5 (17.2–28.5), 32.2 (22.3–44.7) and 46.8 (32.8–66.7) for 3 levels with significant differences (P < 0.001). MEP values were 26 (22.6–29), 31.8 (23.6–43.8) and 56.3 (34.3–71.9), respectively for 3 levels with significant differences (P = 0.008). Significant differences were associated with FVC (P = 0.005) and PEF (P = 0.04) on three BI levels (Table 2, Fig. 3c). The results indicated that various daily living abilities exhibited differences in these respiratory indicators, thus increasing daily living abilities should help to facilitate these respiratory indicators.

Between normal and abnormal MIP levels, TIS was 13 (10–15) and 15 (13–16) points, with no significant differences (P = 0.099), whereas BBS were 17 (3–32.5) and 33 (29–42), with significant differences (P = 0.02), and BI were 50 (35–55) and 62.5 (50–70), also with significant differences (p = 0.022) (Table 3).

Between normal and abnormal MEP levels, TIS were 13 (10–15) and 13 (7.5–15) points, with no significant differences (P = 0.944), BBS were 20 (3–35) and 31 (17.5–3.5), with no significant differences (P = 0.597), and BI were 50 (35–60) and 55 (42.5–57.5), again with no significant differences (P = 0.876) (Table 3).

The FEV₁ was divided into five different levels, including mild: FEV₁ over 70%; moderate: FEV₁ 60–70%; moderately severe: FEV₁ 50–60%; severe: FEV₁ 35–50%; very severe: FEV₁ < 35%. Table 3 shows no significant difference in TIS (P = 0.33), however, there were significant differences in BBS (P = 0.004) and BI (P = 0.044) in 5 different ventilatory dysfunction levels. The patients with higher FEV₁ values tended to have better balance ability and independence in daily activities.

Regular rehabilitation training used the Bobath technique to inhibit abnormal posture and movement patterns, to induced postural reflex and balance reaction, and to promote the formation of normal movement patterns. It included joint movement training, supine turnover and bridge training, transfer of the torso while sitting and standing, and ball activities to build muscle strength and endurance, improve static and dynamic balance, and enhance walking ability. All patients did not receive specific breathing training.

The correlation coefficients between the two assessment results of TIS, BBS and BI were 0.840, 0.947 and 0.956. MCID were 1.5, 3.4 and 4.3 points respectively (Table 4). MCID determined whether the change in two assessment results was real or caused by random testing errors. It was considered that changed scores were real when the change value was greater than MCID [29]. The changed scores (T₀-T₁) of TIS, BBS and BI were 3.5, 4 and 10 points, respectively which all exceeded the MCID values, indicating that patients’ scores had really changed after 3-weeks treatment.

Table 5 shows the improvements in respiratory muscle strength and pulmonary volume. MIP at T₁ (38.1 ± 2.0) was significantly higher than at T₀ (31.0 ± 1.8) (P < 0.001). MEP at T₁ (35.4 ± 1.9) was significantly greater than at T₀ (27.2 ± 1.6) (P < 0.001). FVC significantly increased from 65.9 ± 2.5 (T₀) to 71.9 ± 2.5 (T₁) (P < 0.001). FEV₁ significantly promoted from 59.7 ± 2.7 (T₀) to 68.7 ± 2.6 (T₁) (P < 0.001). PEF significantly improved from 29.2 ± 1.8 (T₀) to 36.8 ± 2.0 (T₁) (P < 0.001). MMEF were 46.5 ± 2.9 (T₀) and 57.5 ± 3.5 (T₁), significantly difference(P < 0.001).

The results of Spearman’s correlation coefficient analysis are in Fig. 4. △ was used to represent the variation of the parameters. △TIS had significant positive correlations with △MEP (r = 0.49, P < 0.001), △FVC (r = 0.27, P = 0.034), △FEV₁ (r = 0.49, P < 0.001), △PEF (r = 0.45, P < 0.001) and MMEF (r = 0.42, P = 0.001). △BI showed significant positive correlations with △FEV₁ (r = 0.31, P = 0.016).

The results demonstrated that △MIP were 2.6(0–4.7), 6.4(4.5–11.1) and 8.3, with significant differences (P = 0.026), at severe, moderate, and slight TIS levels, respectively (Table 6, Fig. 5a). Less improvement of inspiratory pressure was found in the severe level of a trunk control disorder. The △MIP were 5 (2.5–8.5), 9 (4.8–12.6) and 9.7 (7.3–14.6), with a significant difference (P = 0.044), for severe, moderate, and mild BBS levels, respectively (Table 6, Fig. 5b). Respiratory parameters showed no differences in 3 different BI levels (Table 6, Fig. 5c).

TIS was used to evaluate the capacity of trunk muscles to maintain an upright posture and execute targeted movements during both static and dynamic postural modifications [30]. BBS assessed a patient’s ability actively to shift the center of gravity. BI was used to measure performance in activities of daily living. MIP and MEP revealed the strength of the maximum inspiratory and expiratory muscles. FVC and FEV₁ were indicators that directly reflect the respiratory capacity, which depends on muscular strength [31, 32]. PEF and MMEF refer to the maximum expiratory flow rate and the average respiratory flow rate in the mid-section during forced exhalation, respectively. These parameters were primarily influenced by lung volume and expiratory muscle strength [33, 34].

It was found that TIS, BBS, and BI were positively correlated with MIP, MEP, FVC, FEV₁, PEF, and MMEF in the present study. The results showed consistency with Fuglmyer’s study, indicating that respiratory dysfunction is prevalent in stroke patients with hemiplegia [2]. BBS and BI exhibited a moderate correlation with MIP in our study. The results also showed that FVC, FEV₁, PEF and MMEF were correlated with respiratory muscle strength, especially the expiratory muscles. Further analysis of the variations in respiratory function among different levels of motor impairment were undertaken.

The study found notable differences in MIP, MEP and pulmonary function indicators at three levels of TIS, as well as BBS and BI. In contrast, there were significant differences in BBS and BI at two levels of MIP and 5 levels of FEV₁. These results are consistent with the previous findings that BBS and BI were moderately correlated with MIP.

The decrease in TIS score after a stroke was primarily due to a reduction in selective trunk activities, which included the ability to perform unconstrained active movements of the trunk such as flexion, dorsiflexion, lateral flexion, and rotation [35]. Patients may exhibit abnormal trunk movement patterns, which require more energy for trunk activities [36] and can lead to a solidification pattern of the trunk, back extensor spasms, limited thoracic movement and diaphragm activity. These factors further inhibit the activity of the deep and superficial core muscles. TIS reflects the ability to move the torso flexibly and does not directly result in changes in respiratory muscle strength or lung volume [37]. Restricted trunk mobility indirectly resulted in a restrictive respiratory syndrome [38].

The lower the BBS and BI scores, then lower pulmonary function parameter values were recorded. A pervious study had also indicated that balance was independently associated with individual activities and participation [39]. This study found a strong correlation between BBS and BI. The analysis of the relationship between BI and respiratory function yielded a similar relationship between balance and respiratory function.

The inspiratory and expiratory muscles are the deep core muscles that directly affected a patient’s the ability to maintain trunk stability, rather than the selective control ability of the trunk. Therefore, there was no difference in MIP or MEP whether TIS was normal or not. However, patients with normal inspiratory muscles had higher BBS scores than those with abnormal activity, but the BBS scores did not differ between those with normal and abnormal MEP as well as for BI. The results indicated that inspiratory muscles were more involved in balance functions or daily activities than expiratory muscles, implying that the inspiratory muscles were more important for maintaining body balance and daily activities [40].

In brief, various motor functions had direct or indirect effects on respiratory function. Conversely, the inspiratory muscles had a direct impact on a patient’s balance and daily activities.

Bobath selective trunk postural control was an essential component of routine rehabilitation training, including turnover and bridge exercise training in the supine position, shifting the torso in the sitting and standing positions, and activating trunk muscles with ball activities. Our study found that after 3 weeks of routine rehabilitation without specific respiratory exercises, increased motor function also improved the strength of the respiratory muscles and pulmonary volume. Changes in TIS were moderately positively correlated with changes in MEP, FEV1, PEF, and FVC.

The results suggest that routine rehabilitation focusing on facilitating active and flexible trunk movement will significantly improve TIS, and will reflect enhanced control of trunk. Deep core muscles, such as the abdominal muscles, which are also the main exhalation muscles, were activated [41]. This activation explains the results, which found that changes in TIS were moderately correlated with changes in MEP. Of a correlation between TIS and MIP was not found, as also reported in the study by Jandt [42]. Zheng’s study demonstrated that routine rehabilitation improved diaphragm mobility but not its thickness [43], implying that routine training did not improve inspiratory muscle strength. Therefore, based on routine rehabilitation, it was necessary to improve the inspiratory muscle strength of patients with hemiplegia through multiple means, including increasing the thoracic range of motion, diaphragmatic mobilization, and muscle strength training.

Moreover, we also found that MIP improved in various ways in patients with different trunk control abilities. Patients with severe trunk control disorders showed less improvement. Possible reasons for this findings included: (1) in routine rehabilitation, patients received passive training supplemented by low-dose active training, which cannot effectively activate the diaphragm; (2) inspiratory training was often neglected due to poor patient cooperation; and (3) the therapist may neglect the active and passive training of the diaphragm in clinical treatment. Therefore, specific inspiratory muscle training should be included in the clinical rehabilitation process, especially for patients with severe trunk control disorders.

After 3 weeks treatment, we observed differences in the improvement of MIP among individuals with varying levels of balance disorders. Patients with higher balance abilities achieved a better degree of improvement in their core muscle groups, which was beneficial for the original role of the respiratory muscles [3].

The present study had some limitations, such as the small number of patients studied with certain impairment levels, which made it difficult to explain certain issues. Additionally, the 3-week treatment period was relatively short for stroke patients due to hospitalization requirements. Future studies will increase the treatment time to observe its impact on the respiratory muscles and pulmonary function.