An effective cough is an important host defense mechanism to clear the airway.1 The phases of cough are classified as 1) inspiratory, 2) compressive, and 3) expiratory.1-4 During the inspiration phase, normal subjects have pre-cough volumes that are about 85% to 90% of their inspiratory capacity and have a total cough volume of about 2.3 L to obtain optimal cough flows.5 After enough air volume is inhaled, the glottis automatically closes and prevents any outflow of inhaled air.6,7 The contraction of the expiratory muscles while the glottis is open initiates the expulsion phase during which air from the lungs is forcibly expelled.4 An effective cough cannot be obtained if one of these phases fails.

Respiratory muscle weakness and decreased cough capacity are the main causes of pulmonary complications that result in morbidity and mortality in patients with neuromuscular disorders (NMDs) and in tetraplegics. To prevent pulmonary complications in these patients, it is necessary to understand the mechanics of the respiratory system and to accurately evaluate pulmonary function.

In patients with Duchenne muscular dystrophy (DMD), diaphragmatic function is relatively well preserved despite generalized muscle dysfunction.8 However, patients with amyotrophic lateral sclerosis (ALS) experience profound diaphragm weakness, early in the disease process.9-11 Finally, in many patients with spinal cord injury (SCI) if injured at high cervical levels, the functions of intercostal and abdominal muscles are impaired, while diaphragm function is preserved.8-10,12,13

Previous studies have examined respiratory function in patients with NMDs by analyzing the differences between vital capacity measured while patients are in various positions.8-10,13 Several researchers have also studied the correlation between expiratory muscle strength and cough.4,14 However, the relationship between cough capacity and differential respiratory muscle weakness patterns has rarely been discussed. Although expiratory muscle function is crucial for an effective cough, the lung volume attained before expiratory muscle contraction also plays an important role. If insufficient pre-cough volume is obtained due to inspiratory muscle weakness, cough capacity decreases in spite of functional expiratory muscles.15 As a result, a lack of deep breathing leads to atelectasis, where the lung and chest walls lose their elasticity and compliance in the chronic stage.16-18 In addition, a loss of lung compliance may exacerbate cough weakness by restricting dynamic airway compression.15

The correlation between cough capacity and the inspiratory muscle strength necessary to obtain adequate lung volume has rarely been studied.19,20 For this reason, we studied patients with clinically common diseases that accompany restrictive lung disease (RLD) to establish how weakened respiratory muscles affect cough capacity. We evaluated patients with ALS, cervical SCI, and DMD that had different but well-defined patterns of respiratory muscles weakness.

Table 1 lists patient characteristics. All 43 SCI patients showed complete motor paralysis at the cervical level; 32 were classified as having ASIA-A. Twenty-one of these patients were injured at C6, six patients were injured at C7, and five patients were injured at C8. Eleven patients had ASIA-B tetraplegia; seven of these patients had levels of injury at C6, three patients were injured at C7, and one patient was injured at C8.

Table 2 shows that in the cervical SCI group, FVC and %FVC while supine was significantly greater than FVC and %FVC while sitting (p < 0.01). In the ALS group, however, FVC and %FVC while sitting was significantly greater than while supine (p < 0.01). In the DMD group, there was no statistically significant difference between the mean FVCs while sitting or supine. The mean PCF and %FVC is also noted in Table 2.

The SCI group had a relative MIP value much greater than the relative MEP value (p < 0.01) (Table 3). In the ALS group, the relative MEP value was lower than the relative MIP value, but there was no significant difference between the two values (Table 3). In DMD patients, the relative MIP value was significantly higher than the relative MEP value (p < 0.01) (Table 3).

Respiratory muscle strength and PCF were significantly correlated in all three groups. In the SCI group, both MEP (Pearson's coefficient of correlation r = 0.534, p < 0.01) and MIP (r = 0.609, p < 0.01) (Table 4) were significantly correlated with PCF, but MIP (p = 0.007) was more strongly correlated than MEP (p = 0.132) (Table 5) with PCF. The results of the ALS group were somewhat different: again, both MEP (r = 0.528, p < 0.01) and MIP (r = 0.339, p < 0.05) (Table 4) were significantly correlated with PCF, but in this group, MEP (p = 0.003) was more strongly correlated than MIP (p = 0.751) (Table 5). The DMD group showed similar results to the ALS group: MEP (r = 0.742, p < 0.01) and MIP (r = 0.637, p < 0.01) (Table 4) were significantly correlated with PCF, but MEP (p = 0.000) was more strongly correlated than MIP (p = 0.051) (Table 5).

In this study, we compared the FVC of subjects in both sitting and supine positions. In the SCI group, the FVC was larger in the supine position than in the sitting position, which reflects preserved diaphragmatic function but impaired function of the intercostal and abdominal muscles. However, in ALS patients the FVC in the supine position was much smaller than that in the sitting position, suggesting both rapidly progressing muscle weakness with profound diaphragm weakness. In contrast, because in DMD patients the diaphragm retains its function as the primary inspiratory muscle, there is scant difference in vital capacity when the patient's position changes. These results are similar to those of previous studies.8-13,24 We confirmed that measuring FVC in different positions is important to fully understand the weakness patterns of inspiratory and expiratory muscles in patients with RLD.

In this study, the cervical SCI group showed markedly decreased MEP, moderately decreased MIP, and high %MIP/%MEP (2.42) (Table 3), indicating expiratory muscle weakness as the predominant respiratory dysfunction in the SCI group. In the ALS group, both MIP and MEP were markedly decreased, in addition to a lower %MIP/%MEP value (1.12) (Table 3), which suggests that the inspiratory and expiratory muscles had similar and profound levels of weakness. In the DMD group, both MIP and MEP were low, but the ratio of %MIP/%MEP (1.67) (Table 3) was midway between that of the SCI and ALS groups. This indicates that the DMD group exhibited both inspiratory and expiratory muscle weakness, but retained more inspiratory muscle function than the ALS group.

In our correlation study, both MIP and MEP showed significant correlations with PCF in all three groups. However, in the SCI group, multiple regression analysis showed that MIP (p = 0.007) was more highly correlated with PCF than MEP (p = 0.132) (Table 5). This may indicate that inspiratory muscles are more important for cough production in motor complete cervical SCI patients who have lost a large percentage of expiratory muscle power. Conversely, in the ALS group, MEP (p = 0.003) was more highly correlated with PCF than MIP (p = 0.751) (Table 5). Due to the inspiratory and bulbar muscle weakness in patients with ALS, it is difficult to attain sufficient pre-cough volume and to hold the breath just before coughing. Thus, in ALS patients, cough may depend predominantly on expiratory muscle strength. Finally, in the DMD group, MEP (p = 0.000) was more highly correlated with PCF than MIP (p = 0.051) (Table 5). This indicates that although patients with DMD retain more of their inspiratory muscle strength, both inspiratory and expiratory muscle weakness are profound. Because inspiratory muscle strength is insufficient to generate enough pre-cough volume, cough in patients with DMD depends more upon expiratory muscle strength than in patients with SCI.

It is easily understood that coughing is a kind of exhalation process, which leads to the conclusion that expiratory muscle strength is more strongly correlated than inspiratory muscle strength with cough capacity. In light of our results, however, this assumption can be misleading, depending on the disease process and its effects on respiratory muscle weakness. Like the SCI cases in this study, inspiratory muscle strength can correlate more strongly with cough capacity. Thus, we conclude that both expiratory and inspiratory muscle function are important to build an effective cough. Given this assertion, it is difficult to achieve a sufficiently strong cough by assisting only expiratory muscles without assisting inspiratory muscles in RLD patients whose pre-cough volumes decrease.15 Therefore, an adequate cough can be generated by not only expiratory phase assistance, but also by pre-cough volume support.25,26

This study has several limitations. We think it is worthy to note that inspiratory muscle strength may be important for producing an effective cough in patients with predominantly expiratory muscle dysfunction, in light of the fact that inspiratory muscle strength is one of the determinants of pre-cough lung volume. However, we were unable to determine the pre-cough lung volume directly. We believe that the functional residual capacity (FRC) is an indicator of pre-cough lung volume, but our study would be strengthened by obtaining additional lung volume data (total lung capacity, functional residual capacity, residual volume, etc.). With this information, the correlation between FRC and PCF could be analyzed in detail. However, in our hospital, those parameters are not included in routine pulmonary evaluation of patients with severe respiratory compromise.

In conclusion, 1) inspiratory muscle strength is important for producing an effective cough in certain patients with respiratory muscle dysfunction, even though coughing is a type of exhalation process; and 2) pre-cough lung volume may play an important role in establishing an effective cough in patients with respiratory muscle dysfunction whose expiratory muscles are highly impaired.