Contribution of rostral fluid shift to intrathoracic airway narrowing in asthma

In supine asthmatic subjects, application of positive pressure to the lower body caused appreciable increases in respiratory system resistance and stiffness. Moreover, these changes in respiratory mechanics correlated positively with increase in thoracic fluid volume. These findings suggest that fluid shifts from the lower body to the thorax may contribute to overnight intrathoracic airway narrowing and worsening of asthma symptoms.

Abstract

In asthma, supine posture and sleep increase intrathoracic airway narrowing. When humans are supine, because of gravity fluid moves out of the legs and accumulates in the thorax. We hypothesized that fluid shifting out of the legs into the thorax contributes to the intrathoracic airway narrowing in asthma. Healthy and asthmatic subjects sat for 30 min and then lay supine for 30 min. To simulate overnight fluid shift, supine subjects were randomized to receive increased fluid shift out of the legs with lower body positive pressure (LBPP, 10-30 min) or none (control) and crossed over. With forced oscillation at 5 Hz, respiratory resistance (R5) and reactance (X5, reflecting respiratory stiffness) and with bioelectrical impedance, leg and thoracic fluid volumes (LFV, TFV) were measured while subjects were seated and supine (0 min, 30 min). In 17 healthy subjects (age: 51.8 ± 10.9 yr, FEV₁/FVC z score: −0.4 ± 1.1), changes in R5 and X5 were similar in both study arms (P > 0.05). In 15 asthmatic subjects (58.5 ± 9.8 yr, −2.1 ± 1.3), R5 and X5 increased in both arms (ΔR5: 0.6 ± 0.9 vs. 1.4 ± 0.8 cmH₂O·l⁻¹·s⁻¹, ΔX5: 0.3 ± 0.7 vs. 1.1 ± 0.9 cmH₂O·l⁻¹·s⁻¹). The increases in R5 and X5 were 2.3 and 3.7 times larger with LBPP than control, however (P = 0.008, P = 0.006). The main predictor of increases in R5 with LBPP was increases in TFV (r = 0.73, P = 0.002). In asthmatic subjects, the magnitude of increases in X5 with LBPP was comparable to that with posture change from sitting to supine (1.1 ± 0.9 vs. 1.4 ± 0.9 cmH₂O·l⁻¹·s⁻¹, P = 0.32). We conclude that in asthmatic subjects fluid shifting from the legs to the thorax while supine contributed to increases in the respiratory resistance and stiffness.

NEW & NOTEWORTHY In supine asthmatic subjects, application of positive pressure to the lower body caused appreciable increases in respiratory system resistance and stiffness. Moreover, these changes in respiratory mechanics correlated positively with increase in thoracic fluid volume. These findings suggest that fluid shifts from the lower body to the thorax may contribute to overnight intrathoracic airway narrowing and worsening of asthma symptoms.

nocturnal worsening of asthma is common. This worsening is characterized by chest tightness, frequent awakening, increased use of asthma medications, increased airway hyperresponsiveness, and nocturnal decline in lung function (33). Circadian variations in airway inflammation contribute in part to nocturnal worsening of asthma (16), and anti-inflammatory treatment and bronchodilators often improve nocturnal asthma symptoms. Nevertheless, nocturnal asthma remains a major concern (1, 23). Up to 66% of asthmatic patients report nocturnal symptoms at least once a week, indicating poor asthma control (3). In one study, it was shown that 53% of fatal asthmatic attacks occurred at night (30). Moreover, nocturnal asthma leads to fragmented sleep, early morning awakening, and daytime fatigue, leading to a poor quality of life (37).¹

In patients with chronic asthma, high-resolution CT demonstrates thickened lower airway walls (26), which may increase airway hyperresponsiveness (4). Postmortem analysis of asthmatic lungs shows that the airway walls are thicker in patients who died from fatal asthma attacks than in those dying of other causes (6). Acute changes in airway wall thickness can result from inflammation (17), enlarged pulmonary capillaries (5), and local edema due to vascular leakage in the wall (35). In anesthetized dogs, high-resolution CT demonstrated that intravenous infusion of physiological saline (0.9% weight per unit volume of NaCl) increased pulmonary blood volume, thickened airway walls by 50%, narrowed the airway lumen (5, 27), and increased airway response to histamine (4). In healthy humans rapid infusion of intravenous saline leads to mild airway obstruction (27) and enhanced airway hyperresponsiveness (27, 31), likely due to acute increases in airway wall edema.

Previously, we have shown that upon moving from the upright to the recumbent position, as in lying down to sleep, fluid accumulated in the legs because of gravity moves out of the legs and accumulates in the thorax and neck, which we refer to here as a rostral fluid shift (38, 40). During this process, fluid accumulating in the neck increases the neck circumference, narrows the pharyngeal airway, increases pharyngeal airway resistance and collapsibility, and worsens sleep apnea (12, 29, 44). It has been suggested that excess thoracic fluid can cause airway narrowing by entering the airway lumen, thickening airway walls, and reducing airway diameter (42, 43), but the potential impact of rostral fluid shifts upon intrathoracic airway narrowing in asthma has not been studied. The present study examines the extent to which fluid accumulating in the thorax while supine contributes to intrathoracic airway narrowing in asthma.

METHODS

Subjects.

Inclusion criteria were age between 30 to 75 yr, a body mass index (BMI) < 35 kg/m², and never-smokers or ex-smokers with less than 10 pack-years smoking history. Exclusion criteria for asthmatic subjects were current smokers or past smokers with a smoking history of <10 pack-years, a history of hypertension, any cardiovascular, renal, or neurological condition, or use of any medication for these conditions, or use of medication that might influence fluid retention in the body such as diuretics. Subjects with a clinical diagnosis of asthma were enrolled from the respirology clinics at the University Health Network and St. Michael’s Hospitals, Toronto. Their asthma was managed according to the Canadian guidelines for asthma management (18). Healthy subjects were recruited from the local community. Given that menstruation can affect fluid retention, all female subjects were scheduled to undergo the study protocol outside of the menstrual period. All participants provided informed written consent for participation. The experimental protocol was approved by the ethics board of the University Health Network and St. Michael’s Hospitals.

Lower body positive pressure and leg and thoracic fluid volumes.

As described in previous studies (9), lower body positive pressure (LBPP) of 40 mmHg was applied to shift fluid out of the legs similar to the volume that will be moved during nocturnal rostral fluid shift. Application of LBPP increases central venous pressure without increasing left ventricular pressure (32). Furthermore, LBPP does not alter end-expiratory lung volume in healthy subjects (9). With subjects lying supine, deflated antishock medical trousers were wrapped around both legs from hips to ankles, and LBPP was applied by rapidly inflating the trousers to 40 mmHg for 20 min (LBPP study arm). The trousers remained deflated during the control study arm.

Leg and thoracic fluid volumes (LFV and TFV, respectively) were measured continuously with the bioelectrical impedance technique (Biopac Systems EBI100C) as previously described by Yadollahi et al. (40). Bioelectrical impedance is a noninvasive technique that is based on the principle that tissue fluid volume is inversely related to tissue impedance to electrical current. To measure fluid volume in a body segment, a small-amplitude (400 μA), high-frequency (12.5–100 kHz) current is injected into the body segment with a pair of electrodes and the resulting voltage drop across the segment is measured with another pair of electrodes to estimate electrical resistance (R) of the segment. The fluid volume of the segment is estimated assuming that the segment is a truncated cone as (13, 40):

$$\text{V}=\frac{{\text{ρ}}^{2\hspace{0.17em}/\hspace{0.17em}3}}{3{\left(4\text{π}\right)}^{1\hspace{0.17em}/\hspace{0.17em}3}}{\left(\frac{L}{C_1{C}_{2}R}\right)}^{2\hspace{0.17em}/\hspace{0.17em}3}L\left(C_1+C_2+C_1{C}_2\right)$$ (1)

where C₁ and C₂ are the circumferences of the segment at the level of voltage-measuring electrodes in centimeters, L is the distance between the voltage-measuring electrodes in centimeters, and ρ is resistivity of the extracellular fluid for the body segment (45).

Before electrode placement, the skin was cleaned with an exfoliating gel. The electrodes were imbedded in a conductive gel and secured to the skin with adhesive tape. The voltage-measuring electrodes were placed on the ankle and the upper thigh of the right leg for LFV and on the midline of the posterior aspect of the chest at the superior border of the scapula and at the same level as the xiphoid process for TFV. For each body segment, the current-injecting electrodes were placed 1 in. on either side of the voltage-measuring electrodes. Frequencies used for the leg and the thorax were 25 kHz and 100 kHz, respectively (40).

Lung function and respiratory impedance.

Spirometry was performed in the seated position in accordance with ATS/ERS guidelines (20) with a portable spirometer (Microloop, CareFusion Respiratory Care). Test results were expressed using the Global Lung Function Initiative reference equations (28).

Assessment of airway narrowing by spirometry requires deep inspirations that may change airway resistance (25). An alternative is to use impulse oscillometry system (IOS) techniques to assess respiratory system impedance (34). Impulse oscillometry is a passive and minimally invasive technique that only requires tidal breathing to characterize the mechanical properties of the respiratory system. We used IOS to measure respiratory system impedance over 5–20 Hz for 30 s (MasterScreen IOS, CareFusion, Germany). Briefly, small-amplitude flow impulses were superimposed on the subject’s tidal breathing at a rate of 5 impulses/s, and the resulting pressure fluctuations were measured at the airway opening. Respiratory system impedance at a forcing frequency was determined from the frequency domain relationship between the pressure and flow at that frequency and separated into the resistance and reactance of the respiratory system. Resistance is a measure of airway obstruction and includes resistance of the large and small airways, lung tissue, and chest wall. Reactance reflects elastic properties of the respiratory system that are dominant at low frequencies and inertive properties that become important at high frequencies. Here we present resistance and reactance values at 5 Hz (R5 and X5, respectively) and the frequency dependence of resistance over 5–20 Hz (R5-R20). R5 reflects the resistance of the entire respiratory system and increases with airway obstruction. X5 is a negative value, and a more negative X5 reflects increased stiffness of the respiratory system (24). When an airway narrows beyond a certain degree, the elastic parenchymal tissue downstream is no longer accessible to the IOS, resulting in a reduction in X5. X5 becomes more negative with increasing airway obstruction (8), lower airway narrowing (10, 15), and closure (36) and becomes less negative with use of a bronchodilator (7).

During the IOS measurements, the subjects wore a nose clip and supported both cheeks firmly with their palms to reduce the effect of upper airway shunt (24). They were asked to slightly extend their head to minimize the effects of pharyngeal airway narrowing on the IOS measurements. The IOS flow sensor was calibrated with a 3-liter syringe before each subject measurement, and the device calibration was validated with a standard test load of 2 cmH₂O·l⁻¹·s⁻¹ (CareFusion Respiratory Care). The IOS measurements were repeated if the subjects coughed, swallowed, or talked during the recording.

Study design.

Subjects were enrolled in a randomized, double crossover design. Both study arms were performed on a single day. The subjects were instrumented for fluid volume measurements. Before each study arm, the subjects sat upright in a chair for 30 min and two repeat IOS measurements followed by spirometry were performed. Then the subjects lay supine on a bed with no pillow for 30 min. They received either LBPP of 40 mmHg (from 10 min to 30 min) or 0 mmHg (control arm) and were crossed over to the other study arm. In both study arms, while subjects were supine, two repeat IOS measurements were performed at 0 min and 30 min. Respiratory system impedance was derived as the average of the two repeat IOS measurements.

The asthmatic subjects withheld short-acting β₂ agonists for 6 h, long-acting β₂ agonists for 24 h, and long-acting anticholinergic agents for 72 h before the study measurements.

Statistical analyses.

Data are reported as means ± standard deviation (SD) for normally distributed data or median [interquartile range] otherwise. Normality of data was tested with the Shapiro-Wilk test, and for nonnormal data nonparametric tests were performed. Differences in baseline measures between the healthy and asthmatic subjects were assessed with an unpaired t-test or Mann-Whitney rank sum test. Within-group changes in measures such as LFV, TFV, R5, and X5 were compared with a paired t-test or Wilcoxon signed-rank test. Changes in measures from the seated position to supine 0 min between the two study arms were compared by two-way repeated-measures analysis of variance (ANOVA) with posture change and study arms as factors. Changes in measures during supine from 0 min to 30 min between the two study arms were compared by two-way repeated-measures ANOVA with time and study arms as factors. Between-group changes in the measures were compared with two-way ANOVA with study arms and groups as factors. Correlation was assessed by Pearson correlation analysis. Factors independently associated with the changes in R5 and X5 with LBPP (from 0 to 30 min) were identified with stepwise linear regression. Independent factors were sex, age, weight, baseline FEV₁, FEV₁-to-FVC ratio, baseline R5 in the seated position, and absolute and relative changes in TFV during LBPP. Statistical significance was accepted at P < 0.05, and all analyses were performed in SAS 5.1 (SAS Institute, Cary, NC).

RESULTS

Demographics.

Nineteen healthy subjects (6 women) and 21 asthmatic subjects (13 women) consented to participate in the study. In the healthy group, data from two men were excluded from the analysis: one who declined to continue with the study after being instrumented and another who had an upper respiratory infection on the day of measurements. In the asthma group, one woman was excluded from the analysis since she had cough on the day of measurements. We also excluded data from one man and four women because of high variation in their seated respiratory impedance values between the two study arms. The remaining 17 healthy subjects (6 women) and 15 asthmatic subjects (8 women) showed no differences in age, height, BMI, and lengths of the thorax and right leg (Table 1). Asthma medications are detailed in Table 2.

Table 1.

Subject demographics and lung function

	Healthy (n = 17)	Asthma (n = 15)	P Value
Sex (male/female)	11/6	7/8
Age, yr	51.8 ± 10.9	58.5 ± 9.8	0.08
Height, cm	174.5 ± 10.1	170.1 ± 11.7	0.26
Weight, kg	75.7 [61.4, 90.5]	80.7 [75.5, 81.7]	0.68
BMI, kg/m2	25.2 ± 3.9	27.6 ± 3.5	0.08
Right leg length, cm	65.0 [61.1, 67.1]	64.2 [60.1, 69.8]	0.80
Thoracic length, cm	24.8 [21.8, 26.0]	24.7 [22.9, 27.2]	0.90
Lung function, seated
FEV1, % pred	101.3 ± 12.6	67.6 ± 17.4	<0.0001*
FEV1 z score	0.07 ± 0.9	−2.12 ± 1.1	<0.001*
FVC, % pred	105.5 ± 13	85.4 ± 10.4	<0.0001*
FVC z score	0.35 ± 0.9	−1.0 ± 0.7	<0.001*
FEV1/FVC, % pred	95.5 [89.0, 100.2]	86.1 [70.5, 89.9]	0.0004†
FEV1/FVC z score	−0.4 ± 1.1	−2.1 ± 1.3	0.0004*
Fluid volume, supine 0 min
Leg, ml	8,884.7 [6,975.5, 9,189.1]	8,554.5 [7,561.3, 10,281.7]	0.79
Thorax, ml	6,311.6 [5,453.8, 8,716.4]	6,038.3 [4,145.2, 8,511.9]	0.35
Respiratory impedance, seated
R5, cmH2O·l−1·s−1	3.2 ± 0.95	6.8 ± 2.7	0.0002*
X5, cmH2O·l−1·s−1	−0.87 [−1.2, −0.71]	−1.7 [−2.65, −1.32]	0.0002†
Respiratory impedance, supine 0 min
R5, cmH2O·l−1·s−1	4.4 ± 1.4	8.5 ± 3.3	0.0004*
X5, cmH2O·l−1·s−1	−1.4 [−1.8, −1.1]	−3.4 [−4.8, −2.3]	0.0001†

Values are mean ± SD or median [25th percentile, 75th percentile] for n subjects. Leg fluid volumes indicate fluid volume from both legs. Spirometry data are expressed as % of predicted (% pred) and as z scores using the Global Lung Initiative reference equations (30). X5 values are negative. To avoid confusion, in the text we discuss the changes in the magnitude of X5. BMI, body mass index; FEV₁, forced expiratory volume in 1 s as % of predicted; FVC, force vital capacity as % of predicted; R5, respiratory system resistance at 5 Hz; X5, respiratory system reactance at 5 Hz.

^*Unpaired t-test,

^†Mann-Whitney rank sum test between healthy and asthmatic subjects.

Table 2.

Asthma subjects’ medication details

Medications	No. of Subjects
Short-acting β2 agonist (SABA)	1
SABA and corticosteroid (CS)	2
SABA and a combination of CS and long-acting β2 agonist (LABA)	3
SABA, CS, and a combination of CS and LABA	1
SABA, CS, and a combination of CS and LABA, leukotriene receptor antagonist (LTRA)	1
CS and a combination of CS and LABA	1
A combination of CS and LABA	1
Steroid (nasal spray), a combination of CS and LABA, LABA and antihistamine	1
Prednisone, SABA, LABA, theophylline, anticholinergic	1
Prednisone, SABA, and a combination of CS and LABA	1
Anticholinergic and a combination of CS and LABA	1
SABA, a combination of CS and LABA, and a monoclonal anti-IgE antibody	1

Fluid volumes and changes in fluid volumes: supine position.

Within either the healthy or the asthma group, LFV and TFV measured at the beginning of the supine position were similar between the two study arms (P > 0.05 for both). Also, LFV and TFV were similar between healthy and asthmatic subjects in both study arms (Table 1).

After 30 min in the supine position, in both the control and LBPP study arms and in both groups, LFV decreased and TFV increased significantly (P < 0.001 for all). In both groups, the changes in LFV and TFV during LBPP were larger than in the control study arm (P < 0.001 for all; Fig. 1). Furthermore, the decreases in LFV during both study arms were similar between the healthy and asthmatic subjects. Similarly, the increases in TFV in the LBPP arm were similar between the two groups. However, the increases in TFV during the control arm were larger in the healthy subjects than in the asthmatic subjects (Fig. 1; 78.3 ± 31.3 ml vs. 53.2 ± 23.9 ml, P = 0.01).

Fig. 1.

Changes in the leg and thoracic fluid volumes (ΔLFV and ΔTFV, respectively) from 0 to 30 min supine during the control and lower body positive pressure (LBPP) study arms. Data are represented as means ± SD. Black and gray bars represent data from the healthy and asthmatic subjects, respectively. In both groups, LFV decreased and TFV increased in both study arms. However, the changes were significantly larger in the LBPP arm than in the control arm. Except for the increases in TFV in the control arm (P = 0.01), the changes in the fluid volumes were similar between the healthy and asthmatic subjects (NS, not significant).

Lung function and respiratory impedance: seated position.

While seated, asthmatic subjects had smaller FEV₁, FVC, and FEV₁/FVC % predicted values and z score values than the healthy subjects (Table 1). R5 and X5 were larger in the asthmatic subjects than in the healthy subjects in both seated and supine positions (0 min) during the control (Table 1) and LBPP study arms. Compared with the healthy subjects, the frequency dependence of resistance (R5–R20) was significantly larger in the asthmatic subjects in the seated position and at beginning of the supine position in both study arms.

Changes in respiratory impedance.

In both groups and in both study arms, upon moving from the seated to the supine position R5 and X5 increased significantly (P < 0.05 for all). In healthy subjects, R5 increased after 30 min in the supine position in both study arms, with a similar degree of magnitude (Fig. 2, left). In the asthmatic subjects, while R5 increased after 30 min in both study arms, the increases in R5 were significantly larger in the LBPP arm than in the control arm (Fig. 2, right). Also, the increases in R5 with LBPP were larger in the asthmatic subjects than in healthy subjects (P = 0.0003). In the healthy subjects, X5 did not change after 30 min in the supine position in either study arm (Fig. 3, left). However, in the asthmatic subjects, X5 decreased (became more negative) after 30 min in both study arms and the reduction in X5 was larger during LBPP than during the control arm (Fig. 3, right). The reduction in X5 was significantly larger in asthmatic subjects than the healthy subjects (P = 0.0001).

Fig. 2.

Changes in the respiratory system resistance at 5 Hz (R5) in healthy subjects (black) and asthmatic subjects (gray) from 0 to 30 min in the control and lower body positive pressure (LBPP) study arms. Each line represents an individual subject. Bars on either side of data lines represent group mean value with SE.

Fig. 3.

Changes in the respiratory system reactance at 5 Hz (X5) in healthy subjects (black) and asthmatic subjects (gray) from 0 to 30 min in the control and lower body positive pressure (LBPP) study arms. Each line represents an individual subject. Bars on either side of data lines represent group mean value with SE. Note that X5 is negative, but to avoid confusion, in the text we discuss the changes in the magnitude of X5.

In the healthy subjects, the changes in R5 and X5 from seated to supine posture (0 min) were larger than the changes in R5 and X5 during supine posture (from 0 min to 30 min) in either the control or LBPP study arm (P < 0.05 for all). Similarly in the asthmatic subjects, the increases in R5 from seated to supine posture were larger than those from 0 to 30 min supine in either study arm (P < 0.05 for all). In the asthmatic subjects, the decreases in X5 from seated to supine were larger than those while supine in the control arm. However, the decreases in X5 from seated to supine were similar to those due to fluid shift with LBPP (Fig. 4).

Fig. 4.

Average changes in the respiratory system reactance at 5 Hz (ΔX5) with posture change (moving from seated to supine position) and rostral fluid shift in lower body positive pressure (LBPP) study arm in the asthmatic subjects. Error bars indicate SD.

Upon moving from the seated to the supine position, R5–R20 increased in both groups in both study arms (P < 0.05 for all). In the healthy subjects after 30 min in the supine position, the changes in R5–R20 were similar in the LBPP and control study arms (P = 0.1). In the asthmatic subjects after 30 min supine, R5–R20 increased in both the control (Δ = 0.4 ± 0.5 cmH₂O·l⁻¹·s⁻¹, P = 0.009) and LBPP (Δ = 0.9 ± 0.6 cmH₂O·l⁻¹·s⁻¹, P = 0.0001) arms. However, the increases in R5–R20 during the LBPP arm were larger than those during the control arm (P < 0.0001) and larger than in the healthy subjects (P = 0.0001).

Correlations.

In the asthmatic subjects, the increases in R5 (normalized to R5 values at supine 0 min) in the LBPP arm were strongly and positively correlated with the increases in TFV (Fig. 5; r = 0.73, P = 0.002). Indeed, increase in TFV was the only independent correlate of increase in R5 and accounted for 54% of its variability among all asthmatic subjects (stepwise regression). There were no associations between the seated spirometry measures (FEV₁ and FEV₁/FVC) and the increases in R5 with LBPP (P > 0.05) in the asthmatic subjects. In the healthy subjects, no association was observed between the increases in R5 and TFV with LBPP (ρ = 0.2, r = 0.45).

Fig. 5.

Correlation between increases in the respiratory system resistance at 5 Hz, R5 (normalized to R5 values at supine 0 min) and thoracic fluid volume (TFV) in the asthmatic subjects (n = 14) during the lower body positive pressure (LBPP) study arm (r = 0.73, P = 0.002).

DISCUSSION

The principal findings of this study are as follows: 1) While supine, fluid shifting from the legs to the thorax by application of LBPP increased airway narrowing in all subjects. The airway narrowing was larger in asthmatic than healthy subjects as indicated by larger increases in R5 and reduction in X5, thus indicative of increased respiratory system resistance and stiffness. 2) Consistent with the hypothesized link between fluid shift to the thorax and airway narrowing, the magnitude of airway narrowing assessed by R5 was associated with the amount of increase in TFV. 3) In the asthmatic subjects, the effects of fluid shift on lung mechanics as assessed by X5 were similar to those due to the posture change from seated to supine. Together our findings suggest that in subjects with asthma, when supine, fluid displaced from the legs and accumulated in the thorax, causing an appreciable increase in the intrathoracic airway narrowing.

The application of LBPP caused fluid shift from the legs that was similar to the magnitude of spontaneous fluid shift overnight in nonobese patients with obstructive sleep apnea (29). Thus it is possible that overnight fluid shift from the legs and subsequent thoracic fluid accumulation in the supine position while sleeping contribute to nocturnal airway narrowing in asthma. An implication of our findings is that treatments designed to reduce daytime fluid retention in the leg, as simple as exercising or wearing compression stockings during the day, may help reduce nighttime fluid shift to the thorax and airway narrowing, and thus improve asthma control.

It is well established that in asthma there is excess airway fluid due to edema and inflammation (42, 43). The excess fluid can thicken the airway wall and reduce airway diameter, thereby decreasing luminal cross-sectional area available for airflow. Moreover, if the fluid leaks into the lumen, it can further amplify airway narrowing by reducing luminal area. These mechanisms may explain the increases in R5 and decreases in X5 in our asthmatic subjects. More importantly, in the asthmatic subjects the magnitude of airway narrowing attributable to LBPP-induced fluid shift was comparable to that induced by a bronchial agonist (2). The increase in R5 with LBPP in the asthmatic subjects (1.1 ± 0.8 cmH₂O·l⁻¹·s⁻¹) was similar to that seen with methacholine-induced fall in FEV₁ of >20% (1.3 ± 1.7 cmH₂O·l⁻¹·s⁻¹, n = 10) (2). This suggests that in the asthmatic subjects the response to fluid shift was clinically significant.

We recognize the effects of body posture on airway mechanics. Upon moving from the seated to the supine position, respiratory resistance increases over all frequencies and respiratory reactance decreases at low frequencies (21). Respiratory resistance has been considered as a measure of airway obstruction and includes resistance of intrathoracic and extrathoracic airways, lung tissue, and chest wall. Reactance at 5 Hz reflects stiffness of the respiratory system and decreases (becomes more negative) with increasing airway obstruction in asthma (8). Increases in resistance and decreases in reactance with posture change have largely been attributed to reduction in end-expiratory lung volume (21). A novelty of our study is that we controlled the effects of posture and compared the effects of lying supine with minimal fluid shift (control study arm) vs. lying supine with increased fluid shift (LBPP study arm) on airway narrowing. Thus the increases in R5 and X5 with LBPP were likely associated with the increased amount of fluid displaced to the thorax.

During the supine posture, there is an increase in the intrapulmonary blood volume with augmented venous return (14), which may also cause a reduction in the end-expiratory lung volume. While we did not measure end-expiratory lung volumes, previous experiments from our group have shown that in healthy subjects, LBPP of 40 mmHg has an insignificant effect on end-expiratory lung volume (9). While it may be the case that there is a change in end-expiratory lung volume in asthmatics, the more likely explanation for the increases in R5, X5, and R5–R20 with LBPP is increased airway narrowing and respiratory system stiffness due to an increased amount of fluid in the thorax.

We also compared the effects of increased fluid shift with LBPP and the effects of posture change from seated to supine on lung mechanics. In both asthmatic and healthy subjects, R5 increased more with the posture change (Table 2) than during LBPP (Fig. 2). Since the pharyngeal airway is significantly narrower in the supine position compared with the seated position, pharyngeal airway narrowing may explain larger increases in R5 with the posture change compared with the increased fluid shift with LBPP. In healthy subjects, X5 decreased more with the posture change (Table 2) than during LBPP (Fig. 3, left). However, it is important to note that in the asthmatic subjects, the decreases in X5 with the posture change (Table 2) were similar to those during LBPP (Fig. 3, right). These findings suggest that the magnitude of increase in the respiratory system stiffness due to fluid shift might be comparable with that due to the posture change, underscoring the importance of fluid shift in airway narrowing in asthma while supine.

Modeling studies suggest that in asthma excess airway fluid can amplify the degree of airway narrowing with airway smooth muscle contraction (41, 43). With no fluid in the airway lumen, a plateau in the resistance dose-response curve can be achieved during a bronchial agonist challenge (41). However, when there is fluid in the airway lumen, with an increase in the airway smooth muscle contraction airway resistance may increase without any limit. Supporting evidence has been presented by Pellegrino et al., who found that in healthy humans intravenous infusion of saline for 20 min enhanced airway responsiveness to methacholine likely due to airway wall edema encroaching the bronchial lumen (27). Also, in patients with mitral valve disease, lying supine for 2 h and the consequent increase in TFV were shown to raise airway hyperresponsiveness (22). Furthermore, animal studies have shown that fluid overloading by intravenous infusion of saline enhanced airway response to histamine (4). These findings suggest that in asthmatics, while supine, excess fluid in the thorax may contribute to increased airway hyperresponsiveness. Future studies are warranted to test the potential role of fluid shift on airway hyperresponsiveness in asthma.

This study has some important clinical implications. If overnight fluid shift contributes to nocturnal airway narrowing in asthma, treatments that help reduce daytime fluid retention in the legs, such as exercise (19) or wearing compression stockings during daytime (39), may be used to reduce airway narrowing due to fluid shift. In fact, in patients with sleep apnea, wearing below-the-knee compression stockings reduces leg fluid retention and nocturnal fluid shift and improves sleep apnea severity (39). Similarly, in patients with coronary artery disease, four weeks of walking for 30 min/day reduces daytime fluid retention in the legs and subsequent overnight fluid shift out of the legs, inducing significant reduction in sleep apnea severity (19). Previous studies in asthmatic subjects have shown that exercise improves asthma control (11). Although the underlying mechanisms of improved asthma control with exercise have not been fully investigated, reduced fluid retention in the legs could be a contributing factor.

Both healthy and asthmatic subjects showed appreciable increases in R5 with fluid shift in both study arms. These increases could be in part due to an increase in the pharyngeal airway resistance, which was not separately measured in this study. Indeed, previous studies have shown that while subjects are supine fluid shifted to the neck narrows the pharynx, increases pharyngeal airway resistance (9, 32), and worsens sleep apnea (12, 29, 44). Future studies could examine the effects of fluid shift on pharyngeal airway narrowing in asthmatic subjects with sleep apnea.

This study has raised immensely interesting questions and impetus for future work. In the supine position, the fluid displacement from the legs to the thorax will result in the accumulation of fluid in both intrapulmonary and extrapulmonary blood vessels. Determining the site of fluid accumulation in the thorax is an important research question. Future investigations could measure intrapulmonary capillary blood volume to estimate the amount of fluid accumulated in the lungs. Second, it would be interesting to explore the effects of fluid shift on lung mechanics in fluid-retaining patients such as pregnant women and patients with heart or renal failure.

A limitation of our study is that we did not determine whether the effects of fluid shift on airway narrowing relate to asthma control. Future studies should investigate the relationship between fluid shift and asthma control in order to classify subpopulations among asthmatics that are at increased risk of the effects of fluid shift.

Conclusions.

Our results suggest that fluid shift from the legs to the thorax while supine may exacerbate airway narrowing and worsen asthma. Future studies are required to investigate the effects of overnight fluid shifts on airway narrowing during sleep for elucidation of the mechanisms contributing to nocturnal asthma.

GRANTS

This work was supported by grants from the Ontario Lung Association, Sleep and Biological Rhythms Toronto, a Canadian Institutes of Health Research (CIHR)-funded Research and Training Program, and the Canadian Research Respiratory Network (CRRN). The CRRN is supported by grants from the CIHR-Institute of Circulatory and Respiratory Health; Canadian Lung Association (CLA)/Canadian Thoracic Society (CTS); British Columbia Lung Association; and Industry Partners Boehringer-Ingelheim Canada Ltd, AstraZeneca Canada Inc., and Novartis Canada Ltd. Funding for training of graduate students and new investigators within the network was supported by the above funding sponsors as well as by GlaxoSmithKline Inc. The funders had no role in the study design, data collection and analysis, or preparation of the manuscript.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

S.A.B. and A.Y. performed experiments; S.A.B. analyzed data; S.A.B., M.D.I., J.J.F., S.M.T., O.D.L., G.K., and A.Y. interpreted results of experiments; S.A.B. prepared figures; S.A.B. drafted manuscript; S.A.B., M.D.I., J.J.F., S.M.T., O.D.L., and A.Y. edited and revised manuscript; S.A.B., M.D.I., J.J.F., S.M.T., O.D.L., G.K., and A.Y. approved final version of manuscript; A.Y. conceived and designed research.

Glossary

BMI

Body mass index (kg/m²)

C₁

Top circumference of a body segment (cm)

C₂

Bottom circumference of a body segment (cm)

FEV₁

Forced expiratory volume in 1 s (liters)

FVC

Forced vital capacity (liters)

L

Distance between the voltage measuring electrodes (cm)

LBPP

Lower body positive pressure (cmH₂O/l)

LFV

Leg fluid volume (ml)

R

Resistance of a body segment to electrical current (Ω)

R5

Respiratory system resistance (cmH₂O·l⁻¹·s⁻¹)

R5–R20

Respiratory resistance at 5 Hz minus respiratory resistance at 20 Hz (cmH₂O·l⁻¹·s⁻¹)

TFV

Thoracic fluid volume (ml)

V

Fluid volume of a body segment (ml)

X5

Respiratory system reactance (cmH₂O·l⁻¹·s⁻¹)

ρ

Resistivity of the extracelluar fluid for a body segment (Ω·cm)