Abstract
BACKGROUND: Chronic infection with Pseudomonas aeruginosa in patients with cystic fibrosis (CF) is associated with increased morbidity. Chronic infection can cause limb and respiratory muscle compromise. Respiratory muscle function can be assessed via maximal inspiratory pressure (PImax), maximal expiratory pressure (PEmax), and the pressure-time index of the respiratory muscles (PTImus). We studied the effect of chronic P. aeruginosa infection on respiratory muscle function in patients with CF.
METHODS: This cross-sectional study assessed PImax, PEmax, PTImus, FEV1, FVC, maximum expiratory flow during the middle half of the FVC maneuver, body mass index, and upper arm muscle area in 122 subjects with CF, in 4 subgroups matched for age and sex at different stages of P. aeruginosa infection, according to the Leeds criteria. We compared respiratory muscle function in the subgroups according to P. aeruginosa infection state.
RESULTS: Median PImax was significantly lower in CF subjects with chronic P. aeruginosa infection (PImax = 62 cm H2O), compared to subjects who were never infected (PImax = 86 cm H2O, P = .02), free of infection (PImax = 74 cm H2O, P = .01), or intermittently infected (PImax = 72 cm H2O, P = .02). Median PTImus was significantly increased in CF subjects with chronic P. aeruginosa infection (PTImus = .142), compared to subjects who were free of infection (PTImus = .102, P = .006). Median upper-arm muscle area was significantly lower in CF subjects with chronic P. aeruginosa infection (upper-arm muscle area = 2,219 mm2), compared to subjects who were never infected (2,754 mm2, P = .03), free of infection (2,678 mm2, P = .01), or intermittently infected (2,603 mm2, P = .04). Multivariate logistic regression revealed P. aeruginosa state of infection as a significant determinant of PTImus (P = .03) independently of sex, upper-arm muscle area, and FEV1.
CONCLUSIONS: CF subjects with chronic P. aeruginosa infection exhibited impaired respiratory muscle function and decreased inspiratory muscle strength, and chronic P. aeruginosa infection independently impacts respiratory muscle function in subjects with CF.
Introduction
Pseudomonas aeruginosa infection in patients with cystic fibrosis (CF) is a major determinant of lung disease, and is associated with severe pulmonary disease1 and increased morbidity and mortality.2 P. aeruginosa infection is associated with gradually declining pulmonary status in children and young adults with CF, as assessed by lung function studies.3,4 Furthermore, chronic pulmonary infection with P. aeruginosa is associated with deteriorating nutrition state, to the point of malnourishment.5 Sex differences relating to the natural history of P. aeruginosa infection have been described in the literature: women suffer higher rates of colonization and younger age of conversion to the more aggressive mucoid phenotype, compared to men.6,7 Chronic infection has been linked to compromised diaphragm function in animal models8 and human patients.9
Respiratory failure in CF is caused by parenchyma destruction, recurrent infection, and bronchiectasis, and respiratory muscle dysfunction might also play a critical role. Respiratory muscle compromise in CF can lead to respiratory muscle fatigue and thus contribute to respiratory failure. Respiratory muscle strength can be assessed by measurement of maximal inspiratory pressure (PImax) and maximal expiratory pressure (PEmax).10 Respiratory muscle strength has been assessed in numerous studies, and conflicting evidence has been presented on whether it is decreased, maintained, or decreased in patients with CF, as was recently highlighted.11 While some studies advocate that the chronically increased work load against which the respiratory muscles are forced to operate in CF exerts a conditioning effect,12 other studies support that chronic malnutrition and hyperinflation impact on respiratory muscle strength.13,14 Hyperinflation, airway obstruction, and malnutrition have been recognized as the major determinants of respiratory muscle compromise in CF patients.13–15
To our knowledge, the effect of chronic P. aeruginosa infection on respiratory muscle function has not been previously studied in CF patients. We hypothesized that CF patients chronically infected with P. aeruginosa would have impaired respiratory muscle function, compared to CF patients who are not chronically infected. Our aim was to compare respiratory muscle function by measurement of PImax, PEmax, and the pressure-time index of the respiratory muscles (PTImus) between CF patients at different stages of P. aeruginosa infection in a large cohort of children and young adults.
QUICK LOOK
Current knowledge
Chronic infection with Pseudomonas aeruginosa in patients with cystic fibrosis is associated with increased morbidity, including limb and respiratory muscle dysfunction.
What this paper contributes to our knowledge
Cystic fibrosis patients with chronic P. aeruginosa infection had impaired respiratory muscle function and decreased inspiratory muscle strength. Chronic P. aeruginosa infection independently impacted respiratory muscle function in patients with cystic fibrosis.
Methods
This study was approved by the ethics committee of, and performed at, Aghia Sophia Children's Hospital, Athens, Greece. All subjects, or their parents or legal guardians, gave informed written consent prior to the study.
Subjects
We recruited subjects from patients attending their follow-up appointments in our CF department. CF diagnosis was confirmed by sweat-test and expanded mutation analysis. All the subjects received standard daily chest physical therapy. We excluded patients who were unable to perform reproducible lung function tests; patients who had an exacerbation during the month prior to assessment;16 patients on steroids, anti-inflammatories, or azithromycin; and patients who had positive respiratory cultures for Aspergillus species, Stenotrophomonas species, Scedosporium species, Burkholderia species, methicillin-resistant Staphylococcus aureus, or any pathogen other than S. aureus and Haemophilus influenzae in the 5 years prior.
Sample Size
Power analysis was conducted to determine the sample size required to identify PTImus differences between 4 subgroups: never had P. aeruginosa infection; no P. aeruginosa infection at assessment; intermittent P. aeruginosa infection; and chronic P. aeruginosa infection. The PTImus standard deviation was set at 0.03.15 The power analysis indicated that detecting a PTImus increase of 0.05 at a power of 0.9 and a statistical significance of P < .05 would require 8 subjects in each subgroup.
Measurements
Flow was recorded with a pneumotachograph (Mercury F100L, GM Instruments, Kilwinning, Scotland) connected to a differential pressure transducer (DP45, Validyne Engineering, Northridge, California, range ± 3.5 cm H2O). Airway pressure was measured from a side port on the pneumotachograph, with a differential pressure transducer (DP45, Validyne Engineering, Northridge, California, range ± 225 cm H2O). A carrier amplifier (CD280, Validyne Engineering, Northridge, California) was used to amplify the signals from the differential pressure transducers. The amplified signals were recorded with analog-to-digital sampling at 100 Hz (NI PCI-6036E, National Instruments, Austin, Texas) and analyzed with data analysis software (Labview, National Instruments, Austin, Texas).
Respiratory Pressures
We measured PImax, PEmax, airway-occlusion pressure at 0.1 s after the start of inspiration (P0.1), breathing frequency, tidal volume, minute ventilation, inspiratory time (TI), and total breathing cycle time (Ttot). P0.1 was measured during quiet breathing; at least 5 airway occlusions were performed, and the average P0.1 value was calculated. PImax was measured starting at residual volume, with a maximal inspiratory effort against an occluded airway.17 PEmax was measured starting at total lung capacity, with a maximal expiratory effort against an occluded airway.17 Five maximal reproducible respiratory efforts were performed, and the maximum value was recorded.10 The occlusions were performed with a unidirectional valve connected to the mouthpiece (total dead space 8 mL). Care was taken to eliminate any leak around the mouthpiece; a small leak prevented artificial glottic closure.10 Only PImax and PEmax maneuvers with plateau pressure for at least 1 s were accepted for analysis.10 We also calculated the percent-of-predicted PImax and PEmax values.18
Pressure-Time Index of the Respiratory Muscles
Respiratory muscle function was evaluated as PTImus, calculated as:
where PI is airway pressure during inspiration, calculated as
PTImus is a composite index of respiratory muscle function15,19 that describes the efficiency of the respiratory muscles and the balance between neuromuscular competence and respiratory load. A higher PTImus indicates inefficient respiratory muscle function and is related to increased risk of respiratory muscle fatigue.20
Lung Function Tests
FEV1, maximal expiratory flow during the middle half of the FVC maneuver, and FVC were measured (MasterScreen, Erich Jaeger/CareFusion, San Diego, California) per the European Respiratory Society guidelines, and are expressed as percent-of-predicted values.21,22 The values recorded were those achieved before the use of bronchodilator drugs.
Nutrition Assessment
Height and weight were measured, and the corresponding body mass index (BMI) and BMI Z score were calculated.23 Midarm muscle circumference was measured halfway between the acromion and the olecranon to the nearest centimeter, right hand hanging relaxed.24 Triceps skinfold thickness was measured (Harpenden Skinfold Caliper, Baty International, West Sussex, United Kingdom) to the nearest millimeter, halfway over the triceps muscle, skinfold parallel to the upper arm longitudinal axis.24 Upper-arm muscle area was calculated from the midarm muscle circumference and triceps skinfold thickness.25
Classification of P. aeruginosa Infection State
The Leeds criteria were used to classify P. aeruginosa infection state.26 The infection was classified as chronic when > 50% of the months sampled had P. aeruginosa positive culture; intermittent when ≤ 50% of the months sampled had P. aeruginosa positive culture; free of infection when P. aeruginosa culture had been negative over the previous 12months, after having previously been P. aeruginosa culture positive; and never infected when P. aeruginosa had never been cultured from sputum or cough swab. At least 6 airway cultures were acquired in separate months over the year before the assessment.26 Sputum was collected in sterile disposable containers, stored at ambient temperature, and processed within 4 h from collection. Sputum samples were inoculated and incubated aerobically at 37°C for 48 hours, then analyzed for P. aeruginosa and other pathogens. P. aeruginosa positive cultures were identified as either mucoid or non-mucoid phenotype.
All the subjects who were chronically infected with P. aeruginosa were regularly treated with inhaled antibiotics.
Protocol
All the subjects were assessed in the same setting, with the same medical instruments. All the subjects were in stable clinical condition and had received their medications as usual. They were evaluated in the morning hours, and the assessments were in the following order: nutrition assessment, pulmonary function testing, respiratory muscle assessment. All the subjects were evaluated in a sitting position, and a nose clip was used in the pulmonary function and respiratory muscle studies.
Statistical Analysis
Data were checked for normality with the Kolmogorov-Smirnoff and Shapiro-Wilk tests. Differences between groups were assessed for significance with the Kruskal-Wallis rank-sum test and the Pearson chi-square test, as appropriate. If a significant difference was detected, the Mann-Whitney rank-sum test was used for subsequent pairwise comparisons between the subgroups. Multivariate logistic regression was performed to determine which variables contribute to alterations of PTImus. P < .05 was considered a significant difference. Statistical analysis was performed with statistics software (SPSS 17.0, SPSS, Chicago, Illinois).
Results
Subjects
Between October 2009 and June 2010, 122 subjects (68 male) were included in the study. The median age was 13 y (IQR 10–17 y), and 13 subjects were > 19 years old. The median BMI Z score was 0.22 (IQR −0.49 to 0.84). The median percent-of-predicted FEV1 was 99% (IQR 75–119%).
The 4 subgroups were matched for age and sex: group 1 consisted of 11 subjects that had never been infected by P. aeruginosa, group 2 consisted of 33 subjects who were free of P. aeruginosa infection at the time of assessment, group 3 consisted of 39 subjects with intermittent P. aeruginosa infection, and group 4 consisted of 39 subjects with chronic P. aeruginosa infection. As the power analysis showed that 8 subjects in each group were required, the sample size per group was deemed appropriate. In group 4, 17 of 39 subjects (43.6%) were chronically infected with the mucoid P. aeruginosa phenotype. Since the data were not normally distributed, non-parametric tests were applied to compare the subgroups.
Anthropometry and Nutrition
There were no significant differences in height or weight between the 4 subgroups (Table 1). There were significant differences in BMI Z score between the 4 subgroups (P = .02). Post hoc analysis revealed that the median BMI Z score was significantly lower in group 4 than in group 1 (P = .009) or group 3 (P = .02). There were significant differences between the 4 groups in midarm muscle circumference (P = .001), triceps skinfold thickness (P = .02), and upper-arm muscle area (P = .03). Post hoc analysis revealed significantly lower midarm muscle circumference in group 4 than in group 1 (P = .003), group 2 (P = .003), or group 3 (P = .002), and significantly lower triceps skinfold thickness in group 4 than in group 2 (P = .05) or group 3 (P = .006).
Anthropometric, Pulmonary, and Respiratory Muscle Function Data According to Different Stages of Pseudomonas aeruginosa Infection
Lung Function, Respiratory Muscle Function, and Breathing Cycle Components
There were no significant differences between the 4 subgroups in breathing frequency, tidal volume, tidal volume per kilogram, minute ventilation, inspiratory flow, TI, Ttot, TI/Ttot, P0.1, mean PI, or PEmax. There were significant differences in PImax (Fig. 1) and percent-of-predicted PImax (P = .043 and .037, respectively), and post hoc analysis revealed significantly lower PImax in group 4 than in group 1 (P = .044), group 2 (P = .01), or group 3 (P = .046), and significantly lower percent-of-predicted PImax in group 4 than in group 2 (P = .009) or group 3 (P = .02). Non-parametric testing revealed significant differences in mean PI/PImax and PTImus (P = .02 and P = .03, respectively). The mean PI/PImax was significantly lower in group 4 than in group 2 (P = .005), and PTImus (Fig. 2) was significantly higher in group 4 than in group 2 (P = .006). In group 4 the subjects who had mucoid P. aeruginosa had a median PImax of 59 cm H2O, whereas the group-4 subjects who had non-mucoid P. aeruginosa had a median PImax of 66 cm H2O (P = .15).
Maximum inspiratory pressure in 4 subgroups of patients with cystic fibrosis. In each data bar the horizontal line represents the median, the bottoms and tops of the bars represent the 25th and 75th percentiles, and the whisker bars represent the 5th and 95th percentiles. P = .043 for the comparison of all the groups.
Pressure-time index of the respiratory muscles in 4 subgroups of patients with cystic fibrosis. In each data bar the horizontal line represents the median, the bottoms and tops of the bars represent the 25th and 75th percentiles, and the whisker bars represent the 5th and 95th percentiles. P = .006 for the comparison of all the groups.
Multivariate logistic regression analysis revealed that P. aeruginosa infection state was significantly related to PTImus, independently of FEV1, upper-arm muscle area, and sex (Table 2).
Multivariate Logistic Regression Analysis With Pressure-Time Index of the Respiratory Muscles as the Outcome Variable
Discussion
Our study found that PTImus was significantly increased and PImax was significantly decreased in CF subjects with chronic P. aeruginosa infection, compared to those who were free, never infected, or intermittently infected. The chronically infected patients also had compromised somatic muscular indices, such as midarm muscle circumference and upper-arm muscle area.
Our results suggest that chronic P. aeruginosa infection impacts respiratory muscle function in the context of normal lung-function parameters, probably identifying chronic P. aeruginosa infection as an independent determinant of respiratory muscle compromise in CF. Our findings emphasize the multifactorial origin of respiratory muscle impairment in the pathophysiology of respiratory failure in CF.
Chronic pulmonary infection with P. aeruginosa affects the majority of CF subjects by adulthood,27 and increases mortality and morbidity in CF patients,2 and harms pulmonary status in children with CF.3 P. aeruginosa plays a central role in the vicious cycle of pulmonary infection, pulmonary inflammation, lung tissue damage, and consequent respiratory failure.28
Although systemic inflammation is not a major component of CF disease, pulmonary inflammation has been suspected to cause limb and respiratory muscle wasting and weakness via “spill-over” of inflammatory mediators. Pulmonary inflammation and consequent injury is linked to systemic inflammation in patients with COPD,29 and increased circulating inflammatory markers have been detected in CF subjects.30,31 Infection induces respiratory muscle weakness in animal models,32 and respiratory muscle weakness is associated with upper-respiratory-tract infections in humans.33 Induced chronic bronchopulmonary infection by P. aeruginosa significantly decreased diaphragm and limb strength in infected mice,28 and tumor necrosis factor alpha depressed the diaphragmatic tetanic force in murine diaphragm and limb muscle preparations.8 Furthermore, in mice, endotoxin caused diaphragm weakness and contractile dysfunction.34
Respiratory muscle function is compromised in CF, according to some studies, which reported decreased maximal respiratory pressures in hyperinflated, malnourished CF patients with airway obstruction.13–15,35 PTImus was abnormal in CF patients, indicating respiratory muscle impairment related to nutritional compromise, decreased somatic muscle mass, hyperinflation, and airway obstruction.13–15
In our study, PTImus was selected to describe respiratory muscle function because it is measured noninvasively and it is a global respiratory-muscles index that incorporates properties of time, respiratory load, and neuromuscular competence.20 As a limitation of PTImus in CF we should mention that measurement of P0.1 might be affected by the increased time-constant of the CF lung and the ensuing delayed transmission of the pressure changes from the alveoli to the mouth, as exhibited in COPD subjects.36 Sniff nasal inspiratory pressure has been proposed as an alternative noninvasive test of respiratory muscle function in CF, but nasal inspiratory pressure underestimates esophageal pressure, probably because of nasal obstruction and dampening of the pressure changes secondary to the increased time-constant of the CF lung.37
BMI Z score, midarm muscle circumference, and upper-arm muscle area were significantly decreased in our CF subjects with chronic P. aeruginosa infection. Muscular indices and lean body mass correlate well with respiratory muscle indices.15 It has been suggested that decreased muscular synthesis and impaired muscle regeneration occur in the presence of abundant pro-inflammatory cytokines in the chronically infected lung, which spill over and cause chronic systemic inflammation.38
Recent work addressed sex differences in the course of P. aeruginosa infection in patients with CF, and highlighted that estrogen induces mucoid conversion of P. aeruginosa in women with CF and is associated with more frequent exacerbations.6 On the other hand, male CF patients with impaired skeletal muscle strength have normal testosterone levels.39 Mucoid conversion of P. aeruginosa in CF is associated with resistance to antibiotics and increased morbidity and mortality.40 In the present study, which was not designed to address those questions, sex and the mucoid state of P. aeruginosa were not significant determinants of respiratory muscle function in CF subjects. The lack of statistical difference in PImax between subjects infected with mucoid versus non-mucoid P. aeruginosa raises the question of whether muscle weakness is the result of the chronic disease itself rather than the P. aeruginosa infection state. Of note, PImax in CF subjects chronically infected with the non-mucoid strain was nonsignificantly higher than in CF subjects chronically infected with the mucoid strain. This might represent the course of disease progression in CF, reflecting mucoid-strain conversion and increasing disease severity with age.
Compared to previous studies, our study reports similar maximal respiratory pressure and PTImus values. PEmax was not significantly different between the 4 subgroups, which might indicate a preferential action of chronic P. aeruginosa infection on the diaphragm rather than on the expiratory muscles, which might be explained by way of proximity. A preferential-to-the-diaphragm weakness was demonstrated in induced sustained P. aeruginosa infection in animal models.28
Our subjects had good lung function, probably due to idiosyncratic reasons. This was not deliberately done, although enrollment was limited to stable out-patients. Furthermore, our study included younger subjects than have previous studies,14,15,38 who probably (due to limited disease progression) had milder lung disease, implying that further investigation in older CF patients with lower baseline lung function may demonstrate an even more pronounced effect.
Adult CF patients with severe lung disease have increased diaphragm thickness and inspiratory strength,38 probably as a result of the training effect that the chronically increased respiratory load imposes on the respiratory muscles. In these patients inflammation was not identified as a predictor of respiratory muscle compromise, whereas fat-free mass and airway resistance were.38 In our cohort of CF subjects with mild lung disease, diaphragm thickening might have not yet occurred, and the respiratory muscles might have been exposed to the deleterious effects of chronic infection and malnutrition in the absence of a counterbalancing diaphragm thickening mechanism.
Limitations
At present, no definition of chronic P. aeruginosa infection has been universally accepted. In keeping with defining chronic infection as an infectious process that persists despite appropriate therapy, chronic infection could be more accurately described by persistent pathological and immune marker abnormalities.26 Due to technical limitations we were unable to collect data on functional residual capacity, precluding the possibility of incorporating hyperinflation as another factor that probably contributes to respiratory muscle impairment in CF, as suggested in previous studies.14,15 Furthermore, it is possible that the lack of lung-function and breathing-cycle differences between the groups was largely because all our subjects were in good pulmonary condition, and if more debilitated subjects had been included more differences might have emerged. We acknowledge this as a probable bias, since CF subjects with more severe lung disease could have respiratory muscle impairment secondary to marked airway obstruction and hyperinflation, on top of the impairment caused by chronic P. aeruginosa infection. Thus, our selection of a mildly lung-function-impaired cohort might have affected the results and their applicability in the CF community. Finally, although subjects with various other pathogens were excluded, for logistical reasons we did not exclude subjects infected with S. aureus or H. influenzae, because such a cohort could not be empowered for statistical analysis, nor did we apply polymerase-chain-reaction testing in the assessment of sputum samples. Therefore the contribution of S. aureus and H. influenzae to our results cannot be safely refuted; however, in practical terms, obtaining enough patients for such a cohort would be very challenging for any CF center. Although it is possible that co-infected CF subjects have worse outcomes, both S. aureus and H. influenzae are frequently found early on in CF patients, and their contribution to CF lung-disease progression and respiratory muscle impairment has not been clearly delineated.
Clinical Applicability
Identifying chronic P. aeruginosa infection as an independent predictor of respiratory muscle compromise implies that aggressive eradication of P. aeruginosa might help delay respiratory muscle fatigue and respiratory failure in patients with CF. Patients chronically infected with P. aeruginosa could be targeted for respiratory muscle function testing, especially if they have other risk factors, such as malnutrition, hyperinflation, or airway obstruction. Treatments that alleviate respiratory load and decrease the work of breathing, such as noninvasive ventilation,41 inspiratory muscle training, and aerobic exercise, could then be initiated on the basis of the respiratory muscle function condition.
Conclusions
CF patients with chronic P. aeruginosa infection had impaired respiratory muscle function. Maximal inspiratory pressure was significantly lower and PTImus was significantly higher in CF subjects with chronic P. aeruginosa infection than in those without chronic P. aeruginosa infection. Midarm muscle circumference and upper-arm muscle area were significantly lower in subjects with chronic P. aeruginosa infection than in subjects who had never been infected, were currently free of infection, or had intermittent infection with P. aeruginosa. Chronic P. aeruginosa infection might be an independent determinant of respiratory muscle compromise in patients with CF.
Acknowledgments
We thank Richard Parker PhD, Centre for Applied Medical Statistics, University of Cambridge, United Kingdom, for assistance with the statistical analysis.
Footnotes
- Correspondence: Theodore G Dassios MD, Addenbrooke's Hospital, Cambridge University Hospitals National Health Service Trust, Hills Road, Cambridge, CB2 0SW, United Kingdom. E-mail: tdassios{at}upatras.gr.
The authors have disclosed no conflicts of interest.
- Copyright © 2014 by Daedalus Enterprises
