Pulmonary function with expiratory resistive loading in healthy volunteers

Expiratory flow limitation is a key characteristic in obstructive pulmonary diseases. To study abnormal lung mechanics isolated from heterogeneities of obstructive disease, we measured pulmonary function in healthy adults with expiratory loading. Thirty-seven volunteers (25±5 yr) completed spirometry and body plethysmography under control and threshold expiratory loading of 7, 11 cmH2O, and a subset at 20 cmH2O (n = 11). We analyzed the shape of the flow-volume relationship with rectangular area ratio (RAR; Ma et al., Respir Med 2010). Airway resistance was increased (p<0.0001) with 7 and 11 cmH2O loading vs control (9.20±1.02 and 11.76±1.68 vs. 2.53± 0.80 cmH2O/L/s). RAR was reduced (p = 0.0319) in loading vs control (0.45±0.07 and 0.47±0.09L vs. 0.48±0.08). FEV1 was reduced (p<0.0001) in loading vs control (3.24±0.81 and 3.23±0.80 vs. 4.04±1.05 L). FVC was reduced (p<0.0001) in loading vs control (4.11±1.01 and 4.14±1.03 vs. 5.03±1.34 L). Peak expiratory flow (PEF) was reduced (p<0.0001) in loading vs control (6.03±1.67 and 6.02±1.84 vs. 8.50±2.81 L/s). FEV1/FVC (p<0.0068) was not clinically significant and FRC (p = 0.4) was not different in loading vs control. Supra-physiologic loading at 20 cmH2O did not result in further limitation. Expiratory loading reduced FEV1, FVC, PEF, but there were no clinically meaningful differences in FEV1/FVC, FRC, or RAR. Imposed expiratory loading likely leads to high airway pressures that resist dynamic airway compression. Thus, a concave expiratory flow-volume relationship was consistently absent–a key limitation for model comparison with pulmonary function in COPD. Threshold loading may be a useful strategy to increase work of breathing or induce dynamic hyperinflation.


Introduction
Chronic obstructive pulmonary disease (COPD) and asthma are broadly characterized by airway narrowing, high airway resistance, expiratory flow limitation, and exercise intolerance. Expiratory flow limitation in obstructive disease often leads to abnormal lung mechanics such as static and dynamic hyperinflation-powerful drivers of dyspnea [1][2][3]. However, obstructive diseases are exceedingly heterogeneous. For example, patients with COPD suffer from a variety of abnormalities in skeletal muscle [4][5][6], cardiovascular dysfunction [7], physical inactivity [8,9], and high systemic inflammatory burden [10]. Isolating the effects of expiratory flow limitation on lung mechanics alone is challenging for these and other reasons.
For restrictive diseases such as pulmonary fibrosis and interstitial lung disease, chest wall strapping is a simple and useful model for understanding lung mechanics by reducing lung volumes and increasing chest wall stiffness [11]. Such a simple solution for obstructive diseases is elusive. Expiratory resistive loading is effective for increasing the work of breathing [12], dyspnea, exercise intolerance [13][14][15] and is useful with or without metronome-paced tachypnea for inducing dynamic hyperinflation in healthy volunteers [16]. This is true for both reduced aperture or Starling resistors [16,17]. Reducing the aperture during expiration is especially effective for reproducing the characteristic flow-volume deficits in upper airway obstruction [18,19], however upper airway obstruction is very different in presentation of the flowvolume loop than small airway disease such as COPD.
Imposing adjustable external expiratory resistance with a threshold device [15] may provide a non-invasive method to simulate the effects of obstruction and to serve as a model for studying abnormal lung mechanics at rest and during exercise. This could provide supplemental experimental model options to that of either reduced aperture or Starling resistors for expiratory loading [14,16,17]. A comprehensive pulmonary function assessment is not available in the literature for examining healthy people with imposed threshold pressure expiratory loading. Thus, we aimed to measure the effects of expiratory loading on pulmonary function in young, healthy volunteers. We hypothesized that external expiratory loading would produce dose-response impairment in the expiratory segment of the flow-volume envelope, similar to what is present in those with obstructive diseases.

Study participants
Forty-three healthy adults were recruited to participate (25 ± 5 yrs, 171.8 ± 10.2 cm, 72.2 ± 13.3 kg, 24 men, 19 women). Volunteers were included on the basis of age, 18 years and older, and were screened for cardiopulmonary disease using the physical activity readiness questionnaire (PAR-Q). Exclusion criteria included abnormal spirometry defined as an FEV 1 / FVC < 0.70. Written informed consent was obtained and the study protocol was approved by the Institutional Review Board of San Diego State University.

Spirometry and plethysmography
Participants completed spirometry and whole-body plethysmography under control conditions and with imposed expiratory threshold loading at 7 and 11 cmH 2 O (n = 43). A subgroup performed an additional spirometry condition with a loading at 20 cmH 2 O (n = 15). The expiratory load was produced with a threshold expiratory training device (Threshold PEP, Respironics, Pittsburgh, PA) installed between the pulmonary filter and pneumotachometer. The threshold device relies on a flow-independent one-way valve that provides constant pressure that is adjustable based on the valve spring tension. The manufacturer's accuracy and reproducibility are ± 1.0 and ± 0.5 cmH 2 O, respectively. Both the spirometer and the plethysmograph were calibrated with a 3L syringe according to manufacturer's instructions. Spirometry (TrueOne, Parvo Medics, Sandy, UT) and plethysmography were measured using commercial PFT systems (Vmax, CareFusion, Yorba Linda, CA) according to ATS/ERS standards. Pulmonary function was evaluated by forced expiratory volume in the first second (FEV 1 ), forced vital capacity (FVC), ratio of the forced expiratory volume in the first second to forced vital capacity (FEV 1 /FVC), and peak expiratory flow rate (PEFR) produced by spirometry and airway resistance (R aw )measured by plethysmography.
During spirometry testing, participants were seated upright, holding the pneumotachometer to their mouth and wearing a nose clip during all trials. The order of conditions was counterbalanced to ensure no order effects were present. They were then instructed to maximally inhale to total lung capacity (TLC), followed by a forced expiration lasting at least 6 seconds and ending when zero flow was reached. This process was performed for each control and resistance condition until 3 trials with FEV 1 and FVC values were within 150mL for reproducibility, with the best trial for each condition used in the analysis.
During plethysmography testing, participants were seated upright in a sealed box while breathing through a spirometer. Participants were instructed to breathe quietly until achieving a reproducible functional residual capacity (FRC). Volunteers were then instructed to pant at a queued pace (1 Hz) with the spirometer shutter open (R aw ) and closed (lung volumes) for approximately 5 breaths each. This process was performed for each condition until 3 trials with R aw values were within 10% for reproducibility. R aw was averaged for each experimental condition. Assumptions and calculations for R aw are as follows: Where P ATM is atmospheric pressure, P A is alveolar pressure, and R is resistance. Resistance is defined as where η is the viscosity of the gas, l is the length of the airway, and r is the radius of the airway. Thus, _ V is more completely defined as where ΔP is the pressure gradient. Finally, R aw can be described and measured through body plethysmography as

Rectangular Area Ratio
Rectangular Area Ratio (RAR) is a geometric analysis that allows the characterization of the shape of the expiratory limb of the flow-volume envelope [20][21][22]. To measure RAR, a rectangle is drawn bounded by the peak expiratory flow and zero flow at residual volume. A ratio is taken of the area under the expiratory segment of the flow-volume loop to the total area of the rectangle. When the ratio value is greater than 0.5, this indicates the shape of the limb is convex. When the value is below 0.5, this indicates the shape is concave. Each RAR value was obtained using custom-designed software using MATLAB (MathWorks, Natick, MA) to identify the PEF, zero flow, and the area under the expiratory limb of the flow-volume relationship. RAR was calculated as follows [20]: where V@ _ V EE and V@ _ V max are volumes at end-expiratory flow and peak expiratory flow, and V T is tidal volume.

Statistical analysis
Differences in pulmonary function variables were compared using a one-way repeated measures ANOVA. All spirometry variables and R aw were examined for correlation with RAR using the Pearson Product-Moment Correlation method. In the case of a significant omnibus test, a Bonferroni post hoc test was used for follow-up analysis (Prism, GraphPad, San Diego, CA).

Results
Six participants with an FEV 1 /FVC < 0.70 were excluded from analysis on the basis of abnormal spirometry, and this included four from the smaller subset. We examined four primary spirometry variables of FVC, FEV 1 Fig 3).

Discussion
We aimed to measure the effect of imposed expiratory loading on pulmonary function (FVC, FEV 1 , FEV 1 /FVC, PEF, FRC, and R aw ) in healthy adults. Imposed expiratory loading resulted in increased R aw and reduced FEV 1 , FVC, and PEF in healthy adults with no clinically significant change in the ratio of FEV 1 /FVC or FRC. Additionally, RAR did not show a clinically meaningful reduction to suggest alteration in the shape of the flow-volume envelope. There was no perceptible dose-response of imposed loading on pulmonary function, even up to a supra-physiologic loading of 20 cmH 2 O threshold pressure.

External loading and dynamic airway compression
Imposed resistance at the mouth generates lung function abnormalities that are substantially different to that of patients with resistance within the airways. While expiratory flow limitation results from dynamic airway compression in relatively distal airways in patients with COPD, resistance applied at the mouth can maintain airway pressures, the equal pressure point, and prevent airway compression. Naturally, mitigation of dynamic airway compression is one of the mechanisms by which pursed lip breathing is effective [23,24]. We suspect this is the mechanism for why the flow-volume loop failed to show a scooping, or concave profile, similar to that commonly found in COPD where FEV 1 /FVC is reduced. Further evidence to this point is provided by experiments in which negative pressures are imposed at the mouth in patients with COPD. Applying negative expiratory pressures during tidal breathing in people without flow limitation can improve flow as long as the mouth pressures are modest, such as -5 cmH 2 O [25]. In individuals with severe flow limitation, imposed negative expiratory pressure leads to decreases in flow [25] due to exacerbation of dynamic airway compression [2]. At greater unloading (-10 cmH 2 O/L/s), dynamic airway compression is accentuated [2]. This also increases the sensation of dyspnea and provides an alternative method that may be more closely associated with dyspnea scores than spirometry measurements [26].

Absence of dose-response for external threshold loading
There was no apparent spirometry dose-response caused by imposed threshold expiratory resistance. This is despite a clear dose-response in the airway resistance itself. The majority of the decline in lung function appeared to occur at moderate flow limitation. While surprising, it may be that expiratory loading of 11 and 20 cmH 2 O only serve to raise airway pressures, such that the equal pressure point is maintained similarly to that of when 7 cmH 2 O is imposed on the participant. This again highlights a key limitation in our experiment where resistance outside of the airways deviates substantially from resistance within an airway.
It is not clear what external loading would be necessary to produce a dose-response for spirometry. Again, reduced aperture loading appears to accomplish this for modelling upper airway obstruction [18,19], however we were unable to reproduce the behavior using a threshold device. Part of the explanation may be that healthy people are capable of very high maximal expiratory pressures (MEP) sufficient to overcome just about any commercial threshold device. In healthy individuals, MEP is upwards of 125 cm H 2 O [27] and even in COPD MEP may be~75 cmH 2 O [28]. Providing threshold loading with an opening pressure closer to the MEP would certainly further increase the work of breathing and provide more resistance. This amount of resistance placed outside the airways, however, is magnitudes larger than what is present in obstructive disease. There may be no easy solution to impose obstruction outside of the airways.

Indicators of obstruction and restriction
FEV 1 /FVC was reduced with imposed loading, however only by 3-4%. Furthermore, mean FEV 1 /FVC did not fall below the obstructive threshold. These modest changes are due to the proportional reduction in FEV 1 and FVC. The reduction in FVC results from mouth pressure at low lung volumes failing to maintain threshold pressure. Interestingly, the FVC did not differ across the conditions, however it may be a signal:noise limitation for such small change in residual volume. Reductions in FVC are also common in restrictive lung diseases, with minimal changes, or even increases in FEV 1 /FVC [29]. In this manner, external loading has some limited application for restrictive disease, however the absence of reduced TLC or increase in lung stiffness makes this largely inappropriate as a model.

Graphical analysis of the flow-volume relationship
Chronic obstructive diseases are often progressive with no cure. Therefore, lung function can be expected to worsen over time, and it is no surprise that spirometry is used to monitor progression. However, monitoring of symptoms, exercise tolerance, comorbidity, and smoking habits are also highly recommended to guide treatment, and assess management options [30,31]. While spirometry alone cannot give a complete picture of the progression of COPD, graphical analysis of the flow-volume relationship [20,32] may provide an important supplement. With this in mind, we examined the relationship of RAR with spirometry variables that are currently used to diagnose and monitor restrictive and obstructive diseases. We found no relationship between RAR and these variables (Figs 6 and 7). The implication is that RAR may be providing additional information about the flow-volume envelope not captured by traditional variables. To test this hypothesis, longitudinal data are needed to track progression of lung function. Ideally, this may constitute monitoring healthy people and patients with COPD not just with lung function tests, but other clinical outcomes such as imaging, quality of life, and, of course, tracking hospitalization and mortality.

Conclusions
We examined the effect of external expiratory loading on airway resistance and the flowvolume relationship. Expiratory loading resulted in a dose-response increase in the total airway resistance and a reduction in FEV 1 and PEF. With external loading, FVC decreased similarly to FEV 1 and, therefore, FEV 1 /FVC did not change. There was no dose-response in spirometry variables with 7, 11, or even 20 cmH 2 O threshold loading. Graphical analysis of the expiratory segment of the flow-volume relationship did not reveal a clinically meaningful change in the shape of the envelope. The rectangular area ratio was not related to any pulmonary function variables, thus there was no perceptible concavity, or scooping effect, seen in the expiratory limb of the flow-volume envelope. This may be attributed to the mitigation of dynamic airway compression caused by the additional load maintaining airway patency. The absence of concavity in the flow-volume relationship renders expiratory loading inappropriate to mimic obstructive pulmonary disease. The intervention approach may, however, be a useful strategy to increase work of breathing. Additionally, threshold loading may also be useful to induce dynamic hyperinflation in otherwise healthy volunteers. This has already been demonstrated using varied methods of expiratory loading and provides a model to study abnormal lung mechanics without the systemic complexities in patients with obstructive diseases.