Effects of Forceful Breathing Maneuvers of Spirometry on Respiratory Impedance Measured by Oscillometry
Article information
Abstract
Background
Deep inspiratory maneuvers can modify airway diameter, leading to shortterm fluctuations in respiratory impedance (Zrs). However, the impact of the forceful breathing maneuvers used in spirometry on Zrs has not been systematically investigated. This study was designed to assess and compare the effects of spirometry maneuvers on Zrs, as measured by oscillometry, in adult patients with obstructive airway diseases (group I) versus those presenting with respiratory symptoms but without airflow obstruction on spirometry (group II).
Methods
All participants underwent oscillometry assessments both before and immediately following spirometry. Paired t-tests and unpaired t-tests were conducted to compare differences within and between groups, respectively. Bland-Altman plots were utilized to display the percentage change in Zrs parameters against the mean, along with the limits of agreement (LoA).
Results
In this cross-sectional study, 166 patients were enrolled (53% male), with a mean age of 40.4 years. Group I accounted for 62% of cases, while group II composed 38%. Patients in group I demonstrated greater impairment in both spirometry and Zrs parameters. Nearly all patients experienced changes in Zrs after spirometry compared to pre-spirometry values, regardless of group assignment. Except for R5 in group I, statistically significant paired differences in Zrs parameters were not observed between pre- and post-spirometry in either group. The cohort demonstrated mean biases between pre- and post-spirometry as follows: R5 4.6% (LoA: –44.8%, 54%); X5 5.8% (LoA: –69.5%, 81.2%); AX 4.3% (LoA: –93.9%,102.6%); and Fres 0.5% (LoA: –30.8%, 31.8%). The broad and random LoA reflect marked inter-individual variability.
Conclusion
Spirometry maneuvers cause fluctuations in Zrs parameters, especially in R5. Oscillometry performed after spirometry may cause clinically meaningful changes in Zrs parameters.
Introduction
Lung oscillometry is a noninvasive method for measuring respiratory impedance (Zrs), which is composed of respiratory system resistance (Rrs) and respiratory system reactance (Xrs). During oscillometry, pressure oscillations at multiple frequencies are superimposed on tidal breathing, permitting evaluation of Zrs across a spectrum of frequencies. A principal benefit of oscillometry is that it requires less patient cooperation compared to spirometry, as forceful breathing maneuvers are not required. Oscillometry has been recommended as an alternative diagnostic modality to spirometry for assessing obstructive airway diseases [1].
A deep inspiratory maneuver (DI) can influence airway diameter, resulting in short-term variability in Rrs. The impact of DI on Rrs varies between individuals with bronchial asthma and healthy controls [2-4]. In patients with asthma, DI may cause either an increase or a decrease in Rrs, depending on underlying structural changes in the airway walls or functional changes in the airway smooth muscle [3]. How et al. [5] found that DI worsens Rrs in patients with bronchial asthma. As spirometry requires forceful breathing maneuvers, performing oscillometry after spirometry may affect Zrs measurements, particularly in bronchial asthma patients. In recognition of this, the European Respiratory Society (ERS) recommends performing oscillometry prior to any lung function test involving DI [6]. However, DI maneuvers assessed in earlier studies were not identical to the forceful breathing maneuvers used during spirometry. Although spirometry is essential for diagnosing and grading the severity of airflow obstruction, oscillometry is increasingly utilized to assess small airway dysfunction. Implementation of the ERS guidelines regarding oscillometry before spirometry, especially for patients suspected of obstructive airway disease, may be challenging in routine diagnostic workup [6]. Therefore, evaluation is necessary to determine whether the forceful breathing maneuvers of spirometry modify Zrs, and if so, whether these changes have clinical significance.
This study was designed to investigate and compare the effects of forceful spirometric maneuvers on Zrs, as measured by oscillometry, in adults with obstructive airway disease and those presenting respiratory symptoms without spirometric evidence of airflow obstruction.
Materials and Methods
1. Study design
This cross-sectional observational study was conducted between September 2024 and December 2024. The Institutional Ethics Committee approved the study protocol (Approval No. 4547/IEC-AIIMSRPR/2024, dated 15.07.2024). Written informed consent was obtained from every participant.
2. Patients
The inclusion criteria consisted of adult patients presenting with respiratory symptoms (e.g., breathlessness, chest tightness, cough, or wheezing), no evidence of structural lung abnormalities on chest radiography, and who had undergone lung function testing as part of their diagnostic assessment. Asthma was diagnosed, in accordance with the Global Initiative for Asthma guidelines, by a history of respiratory symptoms that varied in time and intensity, along with expiratory airflow limitation [7]. The clinical diagnosis of chronic obstructive pulmonary disease (COPD) was established based on the patient's clinical history, exposure to risk factors, and the presence of airflow obstruction on spirometry [8]. At the time of enrolment, patients with bronchial asthma or COPD were not receiving oral or inhaled corticosteroids or bronchodilators. Oscillometry was performed before spirometry and repeated immediately after the spirometry. Bronchodilator responsiveness (BDR) was assessed with both oscillometry and spirometry whenever deemed appropriate by the treating physician. The enrolled cohort was divided into two groups. Group I included patients diagnosed with bronchial asthma or COPD. Group II included patients who had respiratory symptoms but no evidence of airflow obstruction on spirometry.
3. Parameters
Anthropometric data (age, sex, height, and body weight) as well as clinical diagnoses were recorded. The Resmon Pro Full device (RestechSrl, Milan, Italy) was utilized in accordance with ERS guidelines [6]. Participants underwent oscillometry in the seated position, wearing a nose clip and supporting their cheeks throughout the procedure. A minimum of three trials, with each trial continuing until ten acceptable breaths were recorded. The mean value from three trials was used in the analysis, provided the coefficient of variation (CoV) of respiratory system resistance at 5 Hz (R5) was <10%. Oscillometry parameters evaluated in this study included total R5, the difference in resistance between 5 and 19 Hz (R5-19), total respiratory system reactance at 5 Hz (X5), resonant frequency (Fres), and the area under the reactance curve from X5 to resonant frequency (AX). R5 was considered abnormal if the z-score was >1.64, and X5 was considered abnormal if the z-score was <–1.64, following the ERS technical standards [6]. The z-scores were calculated using normative equations specific to the local population [9]. Both oscillometry and spirometry were repeated 15–20 minutes after administration of 400 μg salbutamol through a metered-dose inhaler with a spacer to evaluate BDR. For BDR assessment, oscillometry was conducted first, followed by spirometry. BDR in oscillometry was defined as a ≥40% reduction in R5, a ≥50% improvement in X5, and a ≥80% reduction in AX compared to pre-bronchodilator values, according to the ERS technical standards [6].
Spirometry was conducted using the PowerCube Body+ (GANSHORN Medizin Electronic, Niederlauer, Germany) following established recommendations [10]. Airflow obstruction on spirometry was defined by a forced expiratory volume in 1 second (FEV1) to forced vital capacity (FVC) ratio of <0.70. The Global Lung Function Initiative reference equations were utilized to determine FEV1% predicted [11]. Airflow obstruction severity was classified according to FEV1% predicted, with FEV1% predicted ≥70% considered mild, and <70% categorized as moderate-to-severe. The BDR in spirometry was identified as an increase of >10% in predicted FEV1 or FVC, in line with recent ERS guidelines [12].
4. Statistical analysis
Statistical analysis was performed using MedCalc statistical software for Windows, version 23.0.8 (Med-Calc Software, Ostend, Belgium). Data are presented as mean±standard deviation and as percentages. The chi-squared test was applied to compare categorical proportions between groups. Pre- and post-spirometry Zrs parameters within each group were compared using the paired t-test. Intergroup differences in Zrs parameters were evaluated with an unpaired t-test. Bland-Altman plots for pre- and post-spirometry Zrs parameters within each group were generated, with the percentage difference plotted against the mean value. The upper and lower limits of agreement (LoA) were defined by the 95% confidence interval of the differences (mean difference ±1.96 times the standard deviation of the differences). Statistical significance was defined as a p-value <0.5.
Results
1. Patient characteristics
A total of 166 patients were consecutively enrolled from the outpatient department, with 53% being male. The mean age of the study cohort was 40.4±15.5 years. Group I included 103 patients (58.3% male), and group II included 63 patients (44.4% male). In group I, 90% of patients were diagnosed with bronchial asthma, and the remaining 10% had COPD. The mean CoV of R5 for the entire cohort before spirometry was 5.23%, and after spirometry was 5.05%. Table 1 compares the demographic data, spirometry parameters, and pre-spirometry Zrs parameters of the two groups. Group I patients were slightly older, showed lower spirometry values, and demonstrated higher Zrs parameters than those in group II. In group I, 25.2% of patients had mild airflow obstruction, while 74.8% had moderate-to-severe obstruction on spirometry.
Comparison of demographic characteristics, pre-spirometry respiratory impedance, and spirometry parameters between group I and group II
BDR in oscillometry was evaluated in 107 patients, with 79.4% belonging to group I. The proportion of BDR detected in R5 was not significantly different between the groups (30.6% vs. 13.6%, p=0.089), while BDR in X5 was significantly more frequent in group I (29.4% vs. 4.8%, p=0.013). Of the 115 patients assessed for BDR by spirometry, 81.7% were from group I and 18.3% from group II. BDR in spirometry was observed in 51.1% of group I patients and in none from group II.
2. Comparison of pre- and post-spirometry impedance parameters of the cohort
Following spirometry, 46.4% of the cohort demonstrated a reduction in R5 compared to pre-spirometry values. The percentages of patients experiencing a post-spirometry decrease in R5 were comparable between group I and group II (43.7% vs. 50.8%, p=0.233). Impairment in R5 prior to spirometry was observed in 39.2% of the cohort, with a significantly greater proportion in group I (55.3%) compared to group II (12.7%; p<0.001). The proportion of impaired R5 significantly decreased after spirometry (36.1%; p<0.001). Similarly, 46.4% of the cohort exhibited a decrease in R5-19 post-spirometry, with almost identical proportions between group I (46.6%) and group II (46.8%).
The proportion of patients in whom post-spirometry X5 became more negative was not significantly different between group I (54.4%) and group II (56.7%; p=0.5), resulting in an overall proportion of 54.2%. Before spirometry, impairment in X5 was observed in 27.7% of the cohort, with a higher proportion in group I (41.7%) than in group II (4.8%), and remained largely unchanged after spirometry (39.8% in group I and 7.9% in group II). After spirometry, decreases in AX and Fres were found in 48.5% and 49.7% of the cohort, respectively, and similar proportions occurred in both groups (AX: 49.0% vs. 53.2%, p=0.359; Fres: 51.0% vs. 49.2%, p=0.476). Only a small number of patients showed no alterations in Zrs parameters following spirometry.
3. Effects of spirometry on the respiratory impedance of group I
Figure 1 demonstrates the paired differences in Zrs between pre- and post-spirometry measurements among patients in group I. After spirometry, R5 was significantly reduced, with a mean paired difference of 0.3 cm-H₂O/L/sec (p=0.013). In contrast, mean paired changes in R5-19 were minimal and not statistically significant (0.05 cmH₂O/L/sec, p=0.408). X5 also exhibited no significant change, as indicated by a mean paired difference of –0.09 cmH₂O/L/sec (p=0.39). Although reductions in Fres and AX were noted post-spirometry, these changes were minor and did not reach statistical significance, with mean paired differences of 0.29 Hz (p=0.424) and 0.97 cmH₂O/L/sec (p=0.418), respectively.
Paired t-test comparison of respiratory impedance parameters measured before and after spirometry in group I. Each set of connected points corresponds to a single patient. (A) Resistance of the respiratory system at 5 Hz (R5), (B) reactance at 5 Hz (X5), (C) resonant frequency (Fres), and (D) reactance area (AX).
The mean paired differences in R5 between pre- and post-spirometry were greater in patients with moderate-to-severe airflow obstruction than in those with mild obstruction (0.36±1.31 cmH₂O/L/sec vs. 0.13±0.81 cmH₂O/L/sec, p=0.416), although this difference did not achieve statistical significance. Thus, the bronchodilator effect of the spirometry maneuver on R5 may be more pronounced as the severity of airflow obstruction increases. The mean paired differences in X5 among patients with mild airflow obstruction were minimal and not statistically different compared to those with moderate- to-severe airflow obstruction (–0.11±0.73 cm- H₂O/L/sec vs. –0.08±1.18 cmH₂O/L/sec, p=0.92). No significant differences in mean paired Zrs parameters were found between pre- and post-spirometry when stratified by spirometry-defined BDR (Table 2). However, patients exhibiting BDR in R5 showed a significantly larger mean paired change in R5 only compared to those without BDR. Patients with BDR in X5 exhibited significantly greater mean paired changes in multiple Zrs parameters compared to those without BDR.
Post-spirometry changes in respiratory impedance in group I patients, stratified by BDR assessed by spirometry and oscillometry
The bias between pre- and post-spirometry measurements was small for R5 (6.2%; 95% LoA: –40.8%, 53.3%), R5-19 (23%; 95% LoA: –192.7%, 238.8%), X5 (7.5%; 95% LoA: –75%, 89.9%), AX (7.4%; 95% LoA: –97.7%, 112.5%), and Fres (1.7%; 95% LoA: –32.8%, 36.3%). Nevertheless, the LoA were broad and largely random, and greater variability was seen among patients with more severe baseline impairments.
4. Effects of spirometry on the respiratory impedance parameters of group II
Figure 2 presents the paired differences between pre- and post-spirometry Zrs parameters in patients from group II. The mean paired differences between pre- and post-spirometry were extremely small and did not reach statistical significance for R5 (0.03 cmH₂O/L/sec, p=0.824), R5-19 (0.04 cmH₂O/L/sec, p=0.397), X5 (0.02 cmH₂O/L/sec, p=0.751), Fres (–0.31 Hz; p=0.228), and AX (–0.7 cmH₂O/L/sec, p=0.159). These results suggest that spirometry maneuvers did not induce meaningful changes in Zrs parameters for patients in group II.
Paired t-test comparison of respiratory impedance parameters measured before and after spirometry in group II. Each set of connected points signifies one patient. (A) Resistance of the respiratory system at 5 Hz (R5), (B) reactance at 5 Hz (X5), (C) resonant frequency (Fres), and (D) reactance area (AX).
The bias between pre- and post-spirometry was minimal for R5 (2.1%; 95% LoA: –50.9%, 54.9%), R5-19 (–8.4%; 95% LoA: –154.9%, 138.8%), X5 (3.2%; 95% LoA: –59.1%, 65.5%), AX (–0.65%; 95% LoA: –97.7%, 112.5%), and Fres (–1.6%; 95% LoA: –26.4%, 23.2%).
In the overall cohort, the bias between pre- and post-spirometry measurements remained small for R5 (4.6%; 95% LoA: –44.8%, 54%), R5-19 (11.1%; 95% LoA: –183.2%, 205.3%), X5 (5.8%; 95% LoA: –69.5%, 81.2%), AX (4.3%; 95% LoA: –93.9%, 102.6%), and Fres (0.5%; 95% LoA: –30.8%, 31.8%), as illustrated in Figure 3. However, the wide LoA for each parameter indicates substantial variability at the individual level.
Bland-Altman plots illustrating the 95% limits of agreement for respiratory impedance assessed before and after spirometry. (A) Resistance of the respiratory system at 5 Hz (R5), (B) reactance at 5 Hz (X5), (C) resonant frequency (Fres), and (D) reactance area (AX). The solid lines denote the mean difference, while dash-dotted lines correspond to the 95% limits of agreement. Open circles indicate group I, and solid squares denote group II. SD: standard deviation.
Discussion
This study evaluated and compared the impact of forceful breathing maneuvers during spirometry on Zrs parameters in patients with obstructive airway disease and in individuals presenting with respiratory symptoms despite normal spirometry results. Oscillometry conducted after spirometry revealed that Zrs parameters increased or decreased in nearly all patients, regardless of the underlying condition. Following spirometry, R5 showed a significant reduction in patients with obstructive airway disease, indicating that spirometric maneuvers induce a transient decrease in airway resistance. Notably, these alterations were more evident in patients who demonstrated a BDR on oscillometry. Changes in reactance indices (X5, AX, and Fres) before and after spirometry were not significant in either cohort, suggesting that these parameters are generally stable despite spirometric maneuvers. While the mean bias in Zrs parameters was minimal for both groups, the broad and inconsistent LoA suggest marked variability between individual measurements. The observed mean bias and LoA for Zrs parameters did not differ significantly between groups.
In this study, a few individuals from group II exhibited impairment in Zrs parameters. Interestingly, impairment in Zrs parameters has previously been reported even in individuals with normal spirometry but having respiratory symptoms, as demonstrated in a population-based cohort [13]. Among Zrs parameters, R5 is particularly known for its inherent variability both within and between measurement sessions [14,15]. The ERS technical standards suggested that a within-session CoV for R5 should remain under 10% in adults to maintain measurement reproducibility [6]. Both pre- and post-spirometry oscillometry measurements met this criterion in the current study, thereby supporting the reliability and repeatability of the Zrs assessments.
DI exerts a bronchodilatory effect, leading to airway dilation in both healthy and asthmatic individuals [4]. However, in patients with mild bronchial asthma, abnormal excitation-contraction coupling in airway smooth muscle mitigates the bronchodilatory effect of DI [4]. The bronchodilatory response observed with DI in asthma is often linked to inflammatory changes in airway smooth muscle and the bronchial wall, resulting in altered airway mechanics [16]. Gobbi et al. [2] described variability in Rrs after DI in both asthmatic and healthy individuals, which is in line with the current study’s observations. Schweitzer et al. [3] found that the impact of DI on Zrs in pediatric asthma patients varies according to airflow obstruction severity. They reported that inspiratory Rrs was more markedly reduced in children with mild asthma compared to those with more severe obstruction. Conversely, in the present study, a greater paired difference in Zrs parameters was noted in patients with moderate-to-severe airflow obstruction than in those with mild disease, although this distinction did not reach statistical significance. In alignment with the present findings, Schweitzer et al. [3] also observed no significant alterations in Xrs following DI.
The findings of this study contribute important additional information. Bland-Altman analysis revealed that mean biases between pre- and post-spirometry measurements remained minimal across all Zrs parameters, independent of underlying disease. LoA is a crucial tool for assessing the interchangeability of two measurements. The broader LoA for Zrs parameters suggests that, even in the absence of bronchodilator administration, Zrs can display marked variability when oscillometry is repeated following a spirometry maneuver. Thus, applying a BDR cutoff lower than the ERS guideline threshold may result in overestimation, as alterations in Zrs parameters may not reflect true BDR [6].
The limitations of this study include the inability to distinguish whether changes in Zrs resulted from the spirometry maneuvers themselves or from inherent between-session variability. Furthermore, the limited number of COPD patients restricted the depth of subgroup analyses in group I. Imbalanced group sizes resulted from the consecutive enrolment. As there were no preceding studies to provide estimates for effect size, a formal a priori sample size calculation was not performed.
In summary, the results indicate that spirometry maneuvers can transiently alter airway geometry, leading to changes in Zrs parameters regardless of disease status. A notable reduction in R5 was detected among individuals with obstructive airway disease, especially those with a BDR in Zrs. The variability in Zrs parameters after spirometry, even without bronchodilator usage, may impede individual-level interpretation. The study findings highlight the necessity of maintaining a consistent sequence for oscillometry to prevent misinterpretation. Further research should assess whether implementing a rest period after spirometry can mitigate variability in Zrs parameters. Larger-scale studies are necessary to validate these results and enhance their applicability.
Notes
Conflicts of Interest
No potential conflict of interest relevant to this article was reported.
Acknowledgments
The author acknowledges Ms. Priyanka Nag and Mr. Bheem Prasad Jaiswal for performing oscillometry and collecting data.
Funding
No funding to declare.