Role of ultrasonography in assessing respiratory muscle loss: insights from a cross-sectional study on neurological patients with long-term bed rest with and without tracheostomy

Abstract

Objective

Prolonged immobilization is the main cause of muscle loss, particularly in patients with tracheostomy. The parasternal intercostal muscle (ICM) and diaphragm are commonly evaluated using ultrasonography; however, the importance of the fifth ICM has been overlooked. We compared the thickness and echogenicity of the second and fifth ICMs and diaphragm in patients with and without tracheostomy to determine the associations with onset duration and other factors and to characterize the respiratory muscle loss.

Patients and methods

This single-center, cross-sectional study compared the thickness and grayscale of respiratory muscles in patients without tracheostomy admitted to a general rehabilitation ward and in patients with tracheostomy admitted to an intensive care rehabilitation ward and correlated them with onset duration, functional capacity, dyspnea, and other parameters.

Results

In patients without tracheostomy, nutritional parameters were associated with delta values of the diaphragm and fifth ICM thickness between the paretic and healthy sides. In contrast, in patients with tracheostomy, the onset duration was associated with grayscale delta values of the second ICM, especially in those with an onset of ≥ 60 days, where the grayscale of the healthy side was significantly lower than that in patients with an onset of <60 days. In patients with and without tracheostomy, the thickness or grayscale of the second or fifth ICMs correlated with the patients’ functional independence measure and Borg scores, although this correlation was weak for diaphragm thickness.

Conclusions

In comparison to the diaphragm, the ICM provide a more comprehensive understanding of pulmonary function in patients undergoing extended bed rest, irrespective of the presence of a tracheotomy. Notably, the thickness of the second ICM and the grayscale of the fifth ICM are indicative of dyspnea scores, while the grayscale of the second ICM correlates with scores related to activities of daily living. These indices exhibit differential correlations in patients with and without a tracheotomy. It is imperative that these factors be assessed and compared in clinical practice.

Take home message

In the intensive care unit, ultrasound was used as a quick and efficient tool for assessing respiratory function. While the diaphragm and second intercostal muscle are often considered the key muscles of interest, focusing exclusively on these two structures requires further evaluation. Our findings suggest that comparing the characteristics of the respiratory muscles on the unaffected side and assessing the ultrasound features of the crucial fifth intercostal muscle are important, regardless of whether a tracheotomy is present.

Introduction

Most survivors of neurological impairments are bedridden. Endurance training is a critical aspect of preparing such patients to sit or stand. However, there are numerous obstacles, including respiratory-related issues and muscle atrophy issues resulting from long-term bed rest. Multiple central nervous system diseases, including stroke, may impair respiratory function due to prolonged bed rest [1]. Patients who are bedridden for a long time may also have pendant pneumonia, which increases mucus production, and disused muscles such as the diaphragm and abdominal muscles cause diminished ability to cough up secretions. Respiratory muscles are associated with pulmonary function and overall health outcomes [2]. Atrophy and further weakness of the respiratory muscles may occur because of respiratory failure, leading to the need for mechanical ventilation and potential further complications related to the respiratory muscles. This can result in long-term bed rest and further deterioration associated with weakness, thereby creating a vicious circle. Muscle loss may be prominent during the acute and subacute phases of the disease. These conditions can lead to severe pneumonia and atelectasis, ultimately resulting in hypercapnia, hypoxemia, and even death [3].

Prolonged bed rest has been reported to cause muscle atrophy and subsequent respiratory impairment [4, 5], however, it remains unclear whether the decline in respiratory function observed in all patients is solely due to prolonged bed rest. This ambiguity warrants further investigation. Few studies have provided sample data on respiratory muscle-related aspects for patients with stroke or long-term bed rest. Pulmonary function testing is the gold standard method for assessing respiratory capacity. However, the pulmonary function testing cannot be readily performed in patients with frailty, particularly those who have undergone tracheostomy, as they often cannot cooperate with the examination.

In patients with neurological injuries, the main muscles studied are the diaphragm and parasternal intercostal muscle (ICM), which play roles in expanding the thoracic cavity. The diaphragm functions synergistically with the parasternal and external intercostals to expand the rib cage during inspiration [6, 7]. Ultrasonographic examination of muscle thickness and grayscale is a safe and reliable alternative for assessing respiratory function in patients who are unable to cooperate with the examination. The parasternal ICM has an important reserve capacity during respiration, moreover, the thickening of the parasternal ICM fraction is linked to the index of respiratory capacity [8]. However, most research has focused on the second ICMs, with few studies examining the lower border muscles of ribs especially in long-term bedridden patients [6, 9, 10]. As ICM thickness gradually increases with inspiration from the second to sixth ICM, the contribution of these muscles in the two planes differs during inspiration [11].

Thus, we propose the following hypothesis that the respiratory muscles—particularly the second and fifth ICMs, along with the diaphragm—may offer distinct insights into pulmonary function. Variations in muscle thickness and grayscale measurements may correlate with disease duration or other clinical conditions, such as tracheostomy status.

Patients and methods

Study design

This single-center, cross-sectional, observational study was conducted among patients with tracheostomy in the intensive rehabilitation area and among patients without tracheostomy in the general rehabilitation area of the rehabilitation department between January and April 2024. Before participation, all individuals and their family members were provided comprehensive information regarding the objectives and procedures of the study, and written informed consent was obtained from each patient. The study was conducted in accordance with the Declaration of Helsinki.

Participants

The inclusion criteria for this study were age between 18 and 80 years and a confirmed diagnosis of ischemic or hemorrhagic stroke or traumatic brain injury. Muscle weakness was defined as prolonged duration of bed rest, and participants were required to have a manual muscle test (MMT) score of ≤ 4 in one or both lower limbs (quadriplegia was defined as an MMT score of ≤ 4 in both the lower limbs), with an onset duration of < 120 days. Various factors can contribute to extended periods of bed rest; however, patients exhibiting insufficient muscle strength may experience prolonged bed rest, potentially exacerbating further muscle strength deterioration [12, 13]. Therefore, in this study, we categorize patients with inadequate muscle strength (Manual Muscle Testing score ≤ 4) as those subjected to prolonged bed rest. For patients with quadriplegia, the affected side was defined as the lower limb with the worse MMT score. Patients with tracheostomy were required to have minimal respiratory muscle function (spontaneous breathing) for at least 1 week. Respiratory rhythm depends on the ventral medulla oblongata [14]; therefore, patients with brain injuries on the ventral side of the medulla oblongata were excluded. Moreover, to exclude potential errors in muscle thickness measurements caused by primary diseases, patients with chronic inflammatory conditions (e.g., diabetes, chronic kidney disease, chronic heart failure, and chronic obstructive pulmonary disease), sarcopenia, hereditary neuromuscular diseases, cachexia secondary to cancer, or a history of rheumatic diseases were excluded [15]. Patients who attempted to occlude the tracheal cannula in further 3 days or occluded in approximately 3 days were also excluded from our study to avoid influencing the results. Additionally, during the grayscale analysis, image data of poor quality that could potentially lead to inaccurate results were excluded from the analysis.

Measurements of diaphragm and ICM thickness

Ultrasound data were collected by a physician with 5 years of specialized experience in musculoskeletal ultrasound and subsequently reviewed by another physician with two decades of expertise in the same field to ensure reliability of data. Participants were instructed not to take analgesics or muscle-relaxing medications for 1 week before the study. During the recordings, participants were asked to lie supine, with arms at their sides, avoiding movement of the head and neck or engaging in conversation during the recording process. A 6–15 MHz linear-array transducer with M-mode ultrasonography (Sonimage HIS; Konica Minolta, Tokyo, Japan) was used to measure diaphragm thickness. All acquisitions were performed at the end of inspiration. For ICM measurements, the ultrasound probe was positioned perpendicular to the ribs at the midclavicular line between the 2nd and 3rd intercostal ribs (second ICM) and between the 5th and 6th intercostal ribs (fifth ICM). The ultrasound probe was placed parallel to the ribs between the midclavicular and anterior axillary lines [9, 11, 16]. On the right side, diaphragm thickness was measured at the level of the upper border of the liver, and on the left side, it was measured at the level of the upper border of the spleen, excluding the pleura and peritoneum during measurements of muscle thickness and grayscale [17]. To minimize errors, the measurement process was conducted without engaging in joint activities, thus preventing any potential impact on the muscle thickness assessment.

Processing of images

Ultrasound images were saved in JPEG format, and the analysis was performed using ImageJ software (https://imagej.nih.gov/ij/index.html; NIH, Bethesda, MD, USA). Detailed measurement methods was adopted from studies conducted by Wang et al. [18] and Formenti et al. [17]. Images were reviewed by a second reader who was not directly involved in image acquisition.

Outcome

The primary outcomes were the thickness and grayscale of the diaphragm and second and fifth ICMs on the affected and unaffected sides. The importance of muscle thickness is well-documented in the literature [19]. Nonetheless, measurements of grayscale values in respiratory muscle ultrasound extend beyond mere assessment of muscle quality; they also provide an evaluation of muscle function, encompassing the detection of intramuscular fat and fibrous tissue [20]. This parameter is progressively being adopted in clinical practice as a biomarker for muscle quality. Increased muscle thickness may be indicative of enhanced functional capacity, whereas elevated grayscale values might suggest diminished muscle quality as a result of fatty infiltration [21]。.

The secondary outcomes included the relationship of the delta value of the muscle thickness (delta value = unaffected side muscle thickness – affected side muscle thickness) and grayscale (delta value = unaffected side grayscale – affected side grayscale) with patient characteristics. The time from the measured ultrasonographic examination to the onset of lower limb weakness in patients was defined as the duration of the onset period. Other measurement data, such as age, sex, albumin value, tracheostomy status, Borg scale score (fatigue and breathlessness score during daily life), and functional independence measure (FIM) score (total score, cognition score, and motor score), were also acquired on the day of the ultrasound scan. Higher scores on the Borg scale indicate more severe dyspnea [22], while higher FIM scores denote better functional ability [23].

Grayscale measurements were conducted using the purchased ImageJ software, adhering to the methods outlined by Wang et al. [18] and Formenti et al. [17] The Borg scale score and FIM scores are widely employed tools and are typically regarded as public domain, therefore no specific licenses were acquired.

Statistical analysis

Few studies have compared the diaphragm and ICMs between patients with or without tracheotomy or between those with hemiplegia and quadriplegia. Based on previously published research on diaphragm and ICM atrophy, a feasible sample size of 100 patients (accounting for a 10–15% dropout rate) was planned to ensure an effective sample size of > 80. This was determined using averages from two studies on diaphragm atrophy (54 and 107 patients) and two studies on ICMs (56 and 64 patients) [8, 17, 24, 25]. As most of the data were not normally distributed, a non-parametric test was used. Mann–Whitney U tests were performed for between-group analysis, and correlations were evaluated using Spearman correlation analysis. All statistical tests were two-tailed, and statistical significance was set at P < 0.05.

Results

Of the 100 recruited patients, data from five patients were excluded due to poor image quality. This resulted in a final analysis of 95 patients, comprising 75 with stroke and 20 with traumatic brain injury. Among the included patients, 63 were in the general rehabilitation ward without tracheostomy, while 32 were in the intensive care rehabilitation ward with tracheostomy (Table 1).

Table 1 Demographic characteristics of study patients

Tracheotomy is typically utilized as a therapeutic intervention for severe airway obstruction or respiratory failure [26]. Consequently, the presence of a tracheotomy may be indicative of compromised pulmonary function [27] and an extended duration of hospitalization [26], which was analyzed in subgroup analyses. However, no significant difference in muscle thickness was observed between patients with tracheostomy and those without. In contrast, muscle grayscale was lower in patients with tracheostomy than in patients without tracheostomy, with a significant reduction in the grayscale noted for the second ICM on the unaffected side (p = 0.032) (Table 2).

Table 2 Demographic characteristics of patients with or without tracheotomy and their ultrasonographic data

Due to the longer duration of bed rest and increased complications associated with lower body function in patients with quadriplegia, we also compared patients with and without quadriplegia. We found that only age, FIM total scale score, and FIM motor scale score (p = 0.029, 0.008, and 0.000, respectively) differed between quadriplegia and hemiplegia in patients without tracheostomy group, whereas ultrasound-related parameters for the 2nd, 5th ICM and diaphragmatic muscles showed no significant differences (Table 3).

Table 3 Difference of quadriplegia or hemiplegia in patients with and without tracheostomy

To assess differences in ultrasonographic measurements over time, we compared patients with a disease duration of ≥ 30 days or ≥ 60 days in both the non-tracheostomy and tracheostomy groups. We found that only the grayscale of the second ICM on the healthy side showed significant differences in patients with tracheostomy between those with a disease duration of < 60 days and those with ≥ 60 days (p = 0.019) (Table 1; Fig. 1). No other significant differences were observed between the groups.

Fig. 1

Different between duration and characteristics of ultrasound

Correlation analyses were performed separately for each group of patients with and without tracheostomy. In patients without tracheostomy, age was associated with diaphragm thickness on the healthy side (r = 0.270, p = 0.033), whereas protein parameters were negatively correlated with diaphragmatic muscle thickness (r=-0.259, p = 0.040) and delta values of diaphragmatic muscle thickness (r=-0.261, p = 0.039) but positively correlated with delta values of the fifth ICM thickness (r = 0.264, p = 0.037). The grayscale of the second ICM on the paretic side was associated with the FIM total scale (r = 0.331, p = 0.043) and FIM motor scale scores (r = 0.323, p = 0.045), whereas the grayscale of the fifth ICM on the paretic side was associated with the Borg fatigue (r = 0.315, p = 0.048) and Borg breathlessness scale score (r = 0.475, p = 0.002). In patients with tracheostomy, the duration of the disease was negatively correlated with the grayscale of the second ICM on the healthy side (r =-0.432, p = 0.013) and its delta values (r =-0.367, p = 0.039); FIM motor scale score was negatively correlated with the thickness of the diaphragm on the paretic side (r =-0.606, p = 0.005), grayscale of the fifth ICM on the paretic side (r =-0.460, p = 0.041), and delta values of the second ICM thickness (r =-0.679, p = 0.001). The Borg fatigue score was negatively correlated with muscle thickness of the second ICM on the paretic side (r =-0.480, p = 0.032) (Fig. 2).

Fig. 2

Correlation analysis between function and characteristics of ultrasound (A) Patients with non- tracheostomy; (B) Patients with tracheostomy. △value determine the value of unaffected side mines affected side. FIM, functional independent rating scale; Borg, Borg rating of perceived exertion

Discussions

We performed a comparative analysis of respiratory-related muscle characteristics in patients with tracheostomy in the intensive rehabilitation ward and patients without tracheostomy in the general rehabilitation ward. Our study revealed notable differences in respiratory muscle characteristics between these two groups. Specifically, patients without a tracheostomy demonstrated a stronger association with the fifth ICM, particularly concerning nutritional status. In contrast, patients with a tracheostomy exhibited more pronounced findings related to the second ICM, especially regarding the duration of onset. However, diaphragm thickness did not yield significant insights in our analysis. Nonetheless, the echogenicity and thickness of the second and fifth ICMs appeared to provide valuable information regarding functional independence.

Daily living ability and muscle strength serve as prognostic indicators in patients with central nervous system injuries. Certain correlation exists between muscle thickness and muscle strength [28, 29]. Although some studies suggest that this relationship is evident only in older populations [30], others have proposed that muscle thickness can serve as a reference indicator for predicting muscle strength [31]. Our conclusions differ from those of these studies because we found that the respiratory muscle thickness of patients with tracheostomy was better than that of patients without tracheostomy; however, the FIM scores of patients with tracheostomy were significantly lower than those of patients without tracheostomy. Future studies should include muscles that are not involved in excessive coughing, such as the lower limbs, to clarify the effects of other large muscles during the disuse process caused by long-term bed stay. Previous studies have suggested that changes in thigh muscles on the paretic and unaffected sides are obvious after stroke [32].

Malnutrition can lead to muscle wasting, weakness, and impaired respiratory function [6]. The course of the disease, particularly prolonged bed rest can further contribute to increased patient frailty [33]. In the present study, we found that, in patients without tracheostomy, nutritional status was closely related to muscle performance as assessed by ultrasonography, particularly in the diaphragm and fifth ICM. In contrast, in patients who underwent tracheostomy, ultrasound-related measurements were closely related to the duration of onset, independent function, and breathing function, especially for the second ICM. This finding supports our hypothesis that the function of ICMs varies from the second to the fifth ICM during respiration. The lower lung typically receives a larger volume of air and experiences more substantial expansion [34], as observed in the differences in ultrasonographic appearance of lung sliding between the left and right lungs [35]. Additionally, this may explain the muscle thickening in different lung regions, which relates to varying levels of lung function.

Diaphragm thickness has been reported to be positively correlated with respiratory function [36], with some research suggesting that preserving diaphragm function is a key strategy for weaning patients from mechanical ventilation in the intensive care unit [37]. We observed variations in diaphragmatic thickness; however, these alterations did not demonstrate a definitive correlation with other functional significances. Furthermore, due to the absence of an analysis of diaphragm mobility, we were unable to ascertain whether the diaphragm had experienced dehydration and atrophy in patients with tracheostomy. The relationship between the thickness of the second ICM and respiratory function has been reported [6, 38], whereas the fifth ICM seems to be rarely mentioned. Our findings indicated that both the thickness and grayscale of the fifth ICM were significantly associated with patient function, regardless of tracheostomy status. It may be necessary to compare the combined data of the second and fifth ICMs. Nonetheless, the thickness and grayscale of both the second and fifth ICMs can accurately represent functional status, whereas the diaphragm may require a mobility analysis in addition to thickness or grayscale assessment.

Previous research has reported that decreased diaphragmatic and parasternal ICM thicknesses are associated with prolonged mechanical ventilation and length of intensive care unit stay [16]. However, this conclusion should be considered alongside tracheostomy status and the compensatory function of the unaffected side. While previous studies have suggested that patients with short-term injuries in the intensive care unit tend to recover rapidly [17]. However, our study found that patients with a longer disease duration exhibited better ultrasonographic performance. We found that the delta value of the second ICM grayscale in long-term bedridden patients varied between those with a 60-day clinical course and those without tracheostomy. This result contrasts with previous studies that suggested muscle atrophy can occur within 1–2 weeks of the clinical course, likely because those studies focused on conditions such as spinal cord injury, which directly lead to rapid muscle atrophy. In contrast, other types of central nervous system damage may involve disuse and underuse, resulting in a gradual loss of muscle mass. It is worth noting that in patients with central nervous system injuries, inspiratory strength is typically only 50% [39]. Considering that some studies have used ultrasonographic performance as an indicator of respiratory function or as one criterion for extubation [40], it may be necessary to include additional reference indicators, such as disease duration.

Patients who are bedridden for extended periods inherently experience reduced activity levels. This reduction in activity ultimately leads to muscle weakness, atrophy, and contracture, further exacerbating limitations in mobility [41]. Due to their limited physical activity, these patients’ insufficient respiratory function may be masked. However, respiratory function is crucial in preparing these patients for standing and walking rehabilitation training and should be prioritized [42]. Muscle thickness or size changes with age and requires standardization for height and weight, whereas muscle grayscale does not. Additionally, muscle grayscale is associated with metabolic diseases; therefore, several relevant conditions were excluded from the present study. We found that muscle grayscale may be a better indicator of respiratory function and the duration of immobilization than muscle thickness. Muscles with a high grayscale value may indicate atrophy or soft tissue pathology, as muscle tissue comprises both contractile and non-contractile components, such as fat; a higher grayscale value suggests a more concentrated state [20]. In our study, we observed that the respiratory muscles of patients without tracheostomy were thinner and had higher grayscale values than those of patients with tracheostomy. Furthermore, in patients with tracheostomy, the grayscale of ICMs on the unaffected side was lower in those with a longer disease duration, suggesting that muscle function on the unaffected side may be preserved [40]. This finding is challenging to explain, as the correlation analysis indicated that thicker second ICMs in patients with tracheostomy were associated with less dyspnea, suggesting that respiratory muscle thickness not only reflects respiratory function but also relates to compensatory mechanisms. Patients with tracheostomy often rely on the unaffected side to compensate for the disuse of respiratory muscles on the paretic side during forceful breathing or coughing, which indirectly activates muscle function on the unaffected side [43, 44]. We found that the Borg score was significantly higher in the tracheostomy group than in the non-tracheostomy group. Additionally, respiratory muscle thickness was greater, and muscle grayscale was lower in the tracheostomy group than in the non-tracheostomy group, particularly for the second ICM, whose grayscale value has been reported to be closely related to respiratory function [8].

Limitation

We used MMT to define muscle strength insufficiency in patients with long-term bed rest; however, future research is needed to study the actual duration of bed rest and its relationship with muscles to clarify the impact of time on changes in muscle thickness. For differences in thickness and grayscale between the second and fifth ICMs with respect to time and respiratory function, additional samples and experimental methods may be required to support this result. Moreover, we did not select other muscle groups such as the scalene and sternocleidomastoid muscles. The scalene and sternocleidomastoid muscles exhibit similar respiratory activities during breathing. For example, stretching the scalene muscles can improve vital capacity. However, in the present study, we did not attempt to determine all muscles around the lungs; nevertheless, these large muscles may be related to long-term bed rest. Future studies should match the body mass index between the two groups. However, it is difficult to accurately measure the weights of bedridden patients. In addition, future research using novel ultrasound systems, such as elastography or M-mode, is required to assess soft tissue stiffness and mobility.

Conclusions

To some extent, the respiratory function of bedridden patients can be assessed using ultrasonography. However, additional indicators, particularly the fifth ICM, bedridden time, muscle grayscale, and information on the unaffected side, need to be considered because evaluating respiratory function solely based on the thickness of the diaphragm and second ICM is not sufficient to assess different types of patients. The ICM seems to offer more comprehensive insights into pulmonary function in patients undergoing prolonged bed rest, irrespective of the presence with or without a tracheotomy, when compared to the diaphragm. Notably, the thickness of the second ICM is more strongly correlated with scores related to dyspnea. Grayscale analysis provides supplementary insights, including the correlation of the second ICM with activities of daily living and the duration of bed rest, as well as the association of the fifth ICM grayscale with the severity of dyspnea. Furthermore, the disparities between the affected and unaffected sides hold clinical significance and may inform clinical practice.