Thoracic gas compression during forced expiration in patients with emphysema, interstitial lung disease and obesity

During forced expiration alveolar gas is compressed due to muscular force and the combined effect of elastic pressures of the thorax and lungs acting against airway resistances [1]. The compression-free flow-volume curve can be measured by a flow plethysmograph (transmural plethysmograph), employed in a constant pressure, pressure-corrected mode, by integrating air flow through a pneumotachograph in the wall of a body plethysmograph. A flow plethysmograph measures the volume displacement of the chest movement simultaneously with the spirometric measurement of integrated flow at the mouth. This method permits accurate measurements of changes in thoracic gas volume (TGV) during forced expiratory manoeuvres [2, 3].

Several reports exist on increased thoracic gas compression in asthma, the compression increasing with increasing airways resistance [4‐6]. Alveolar gas compression during forced expiration causes a considerable artifact on flow-volume curves derived from mouth flow [7‐9]. This divergence is also seen in flow-volume loops during bronchodilation testing, especially in the mid-range of forced expiratory flow [5, 6]. Elevated gas compression at the mid-fraction of the forced expiratory flow-volume loop has been reported for smoking subjects [5] and in chronic obstructive pulmonary disease (COPD) [10, 11]. However, as far as we know, no data are available on gas compression in interstitial lung disease (ILD) or obesity.

Gas compression can modify the results of conventional forced spirometry. Therefore, we were interested in studying gas compression in ILD and obesity relative to healthy controls and patients with emphysema.

Thoracic gas compression was investigated in four groups of subjects with different pulmonary mechanics: healthy non-smoking subjects (controls), healthy non-smoking obese subjects (obese), patients with interstitial lung disease (ILD) and patients with emphysema (emphysema).

Spirometry was measured with a Medikro SpiroStar spirometer (Medikro Kuopio, Finland) according to recent international guidelines [17]. Bronchodilation testing was performed using 0.4 mg of salbutamol aerosol (Ventoline®, Glaxo UK) given with a spacer. Post-bronchodilator spirometry was recorded 15 minutes after the salbutamol dose.

Diffusing capacity measurements were performed according to international guidelines [18]. Vital capacity (VC) was measured first, allowing standardization of the inspiratory volume to 90% of individual maximal VC [14]. The mean value of the parameters of two diffusing capacity measurements was presented as the result. Measurements were corrected for haemoglobin concentration.

The body plethysmographic measurements were performed first in conventional constant-volume mode (MasterScreen Body Version 4.3, Würzburg, Germany). The atmospheric pressure was measured using a Vaisala device (Vaisala, Finland). Flow-volume calibration according to the three-flow protocol was performed with a Jaeger 3 L calibration syringe. If the deviation of the registered volume was within ±3.5% for all flows, the calibration was regarded as successful. Box verification/calibration included two procedures. The time constant of the box was verified to lie within the range of 4–7 seconds. Box shift volume was calibrated following the recommendations of the manufacturer.

For flow plethysmography, corresponding calibration methods were applied. To allow flow-volume calibration of the spirometer connected to the box chamber, the box was set to the flow plethysmographic (transmural) mode by removing the closure on the box wall hole and closing the door of the box. The calibration pump strokes were directed through the box chamber pneumotachograph. A volume calibration was accepted if the stroke volumes did not exceed the ±3.5% range.

All investigations were performed in conventional constant volume mode first. The measurements of specific resistance breathing loops and functional residual capacity (FRCpleth) were performed with a panting frequency of 0.5 Hz [14]. The 4 second shutter manoeuvre (FRCpleth) was linked to a maximal spirometric vital capacity manoeuvre for determination of residual volume (RV) and total lung capacity (TLC). The system automatically derived total specific conductance (sGtot) from the breathing loops and determined total respiratory resistance (Rtot).

In flow plethysmographic mode, the uncompressed flow-volume curve at the chest wall and the compressed flow-volume loops at the mouth were measured simultaneously. The differences between compression-free or thoracic flows and corresponding compressed mouth flows were regarded as rough estimates of the degree of gas compression at levels of peak flow (PEF) and maximal expiratory flows at 75%, 50% and 25% of the remaining forced vital capacity (MEF75, MEF50 and MEF25). The volume reference was the same in both thoracic and mouth flow measurements, i.e. the forced vital capacity.

The results of spirometry, diffusing capacity and constant volume body plethysmography for the patient groups are summarized in Table 2. In ILD, restrictive ventilatory impairments and reduction of diffusing capacity were found. Reduced FEV1/FVC ratio, increased Rtot and reduced sGtot were noted as signs of obstruction, and limited DLCOc and DLCOc/VA (KCOc) as signs of emphysema. In obesity, low ERV and FRCpleth were observed with a normal diffusing capacity.

A schematic presentation of gas flows during forced expiration at the thorax and at the mouth is provided in Figure 1, enabling the differences in gas compression for each of the four groups to be visualized. The quantitative results are given in Table 3. In all groups, a significant difference between thoracic and mouth flows was found at MEF50 (Table 3). In controls, a significant difference was measured also at MEF75, in emphysema at PEF and MEF75 and in obese subjects at MEF75 and as the only group additionally at MEF25.

At the level of PEF, the difference between thoracic and mouth flows was significantly larger in patients with emphysema than in controls or patients with ILD (Figure 2).

When the patient groups were compared with each other, ILD showed the lowest difference between thoracic and mouth flows at MEF75 (Figure 3).

At MEF50 and MEF25, the difference between thoracic and mouth flows was significantly larger in obese subjects than in patients with emphysema (Figures 4 and 5).

Significant differences between thoracic and mouth flows at forced expiration were observed in all patient groups in this study, at least at the MEF50 level. Results clearly indicate distortion of forced expiratory flow-volume curves measured by conventional spirometric methods. Some inter-individual variation and overlapping between study groups occurred. However, significant differences between thoracic and mouth flows can be explained by differencies in pulmonary mechanics, e.g. lung stiffness, obesity and decreased elastic recoil in emphysema which cause different airways’ resistances. Differences in flows were reduced in ILD at MEF75 and increased in emphysema at PEF and MEF75. In obesity, significant deviances were seen also at the level of MEF25.

Mead [19] developed the volume displacement body plethysmograph, and Jaeger and Otis [20] analysed volume displacement at the thorax and at the mouth and found the latter to be lower. This difference in volumes increased with increasing airway resistance, respiratory rate and lung volume. Later on, a flow plethysmograph was developed, easily changeable from constant volume mode to constant pressure mode (3).

In normal subjects, an increase of alveolar pressure in forced expiration parallels a compression of intrabronchial gas, upstream of the equal pressure point [1]. Forced inspiration, on the contrary, induces a decompression of alveolar gas. In bronchial obstruction, the intrabronchial gas compression in forced expiration has been reported to be increased [4, 5, 9], explained by the elevated airways resistance. Since in asthma the airways obstruction may vary, only patients with stable pulmonary diseases, without a significant bronchodilator response measured with spirometry, were included in the study.

The differences of thoracic and mouth flows in emphysema patients at the level of PEF were larger than in controls. PEF and MEF75 represent mostly characteristics of proximal airways, and MEF50 of more peripheral airways [12]. In emphysema, the support of bronchi is limited because of the emphysematic process increasing peripheral inhomogeneity and furthermore increasing airway resistance, which may explain the findings. Our results confirm earlier investigations in smokers [5] and COPD [11]. In emphysematic COPD, the peripheral obstructions typically affect the MEF50 and MEF25 values. In this study, MEF25 in patients with emphysema was very low because of elevated airway resistance. The differences of maximal expiratory flows measured by the thoracic and mouth pneumotachographs at this flow level were also elevated. In addition, in emphysematic COPD probably lowered muscle force in forced expiration may influence the expiratory flow values, although this component was not investigated here.

In obesity, the differences between thoracic and mouth flows also occurred at MEF25 level. Even though the difference at MEF25 was significant only in comparison with emphysema, this finding is remarkable. Decreased lung volume causes decreased airways calibers increasing airways resistance, which might be the source of this finding. The lungs have to work against increased intra-abdominal pressure (not measured here) [21‐25]. Trapped gas and low expiratory reserve volume also contribute to unfavourable respiratory mechanics in obesity. For most of the obese subjects the specific conductance was in normal limits ≥46% of predicted value [14], which indicates that no bronchial obstruction was present in them.

The development of interstitial fibrosis decreases pulmonary compliance and is associated with inhomogeneous gas exchange. This study revealed that patients with ILD had unexpectedly low differences between thoracic and mouth flows at the beginning of maximal expiration. In ILD, elastic recoil forces are increased and airways stiffened, which decreases airways resistance. In a consequence dynamic airways compression in the large bronchi is decreased or abolished.

Especially in several patients with ILD, we observed an inverse behaviour between thoracic and mouth flows. It is difficult to explain the reasons to that finding. It could be possible that during forced expiration the upwards movement of the diaphragm might be attenuated due to stiff pulmonary tissue in ILD. Therefore, dynamic reduction of thoraco-abdominal volume during forced expiration may be decreased thus counteracting the pressure and volume displacements within the body box thereby decreasing the flow measured through the wall of body box (thoracic flow).

Another factor for the negative flow difference in ILD might be an instrumental error. Assuming that no phase shift occurs between measurements of mouth flow and thoracic flow and no instrumental errors exist, mouth flow exceeding thoracic flow would imply dilatation of some gas compartments in the lungs or elsewhere in the body during forced expiration.

As presented in Table 2 the total specific conductance (sGtot) was the highest in ILD compared with the other groups suggesting the least resistance. Based on the high specific conductance the changes in intrathoracic pressure, lung volume and air flow during forced expiration are, probably, more abrupt in ILD patients than in other patient groups. This might augment the effect of a possible phase shift between the measurements of mouth flow and thoracic flow. As presented by Goldman et al. [2] the volume change recorded by integrated flow across the plethysmograph wall is slower than that measured at mouth because of a temporary loss of volume during the initial decrease in plethysmographic air pressure. For comparison, in obstruction the time constant of the mouth flow could be high enough to fit with the compensation of the time constant of the box. In ILD, the time constant of the mouth flow is low and an incomplete correction of the box flow probably leads to negative differences.

The patient groups are well defined, and typical functional findings were made within each group. However, finding suitable patients for the study was difficult. The group of emphysema patients was on average older than other groups. Patients with pulmonary fibrosis were mostly on peroral steroids, therefore having tendency for overweight. Locating lean patients with ILD was thus challenging. In addition, it was challenging to find healthy elderly obese non-smoking men. In consequence, the BMI of patients with emphysema was lower than that of other patient groups and healthy subjects. The group of obese subjects mostly comprised women. A small, non-significant difference in age was also present between groups. To overcome these differences, comparisons between all participants were adjusted for these confounders, as indicated in the Methods section. However, the subgroups were rather homogenous and adjustments were not needed when comparisons within these groups were performed. Most of the variables were normally distributed. Because the available statistical program did not allow adjustment for non-parametric tests, parametric tests were used for all comparisons. The results in non-parametric and parametric testing did not differ significantly; therefore, no bias was likely caused by this restriction.

We have used a panting frequency of 0.5 Hz because the Finnish reference values for body plethysmography [14] have been determined for this panting frequency. Compared with medium or high panting frequencies, there might be slight differences in the functional residual capacity (FRC) values [26]. Originally, Jaeger and Otis (1964) [20] used spontaneous breathing and recommended avoidance of panting in body plethysmographic measurements.

In conclusion, patient groups with distinct respiratory mechanics show different profiles of gas compression. All groups had elevated differences between thoracic and mouth flows in the mid-fraction of forced expiratory manoeuvre. The remarkable difference in thoracic and mouth flows found in emphysema is probably associated with inhomogeneity of the lung and decreased elastic forces increasing peripheral obstruction. The low difference between thoracic and mouth flows in ILD, can be explained by increased elastic forces decreasing airway resistance in the proximal airways.

Our results revealed remarkable levels of gas compression also in obesity in the middle and last parts of the maximal expiratory manoeuvre. These findings may help in the interpretation of conventional spirometric results, especially in obese subjects. A decrease in forced expiratory flow in non-smoking healthy obese subjects is more likely a result of gas compression than of real bronchial obstruction.

PP, Chief Physician at the Unit of Clinical Physiology, Helsinki University Central Hospital.

UH, Senior Consultant at the Pulmonary Clinic, Helsinki University Central Hospital.

H-J S, engineer with a methodological background in lung function testing and physiological expertise.

AS, Professor of Clinical Physiology, Helsinki University.