Is airway damage during physical exercise related to airway dehydration? Inputs from a computational model

Abstract

In healthy subjects, at low minute ventilation (V̇e) during physical exercise, the water content and temperature of the airways are well regulated. However, with the increase in V̇e, the bronchial mucosa becomes dehydrated and epithelial damage occurs. Our goal was to demonstrate the correspondence between the ventilatory threshold inducing epithelial damage, measured experimentally, and the dehydration threshold, estimated numerically. In 16 healthy adults, we assessed epithelial damage before and following a 30-min continuous cycling exercise at 70% of maximal work rate, by measuring the variation pre- to postexercise of serum club cell protein (cc16/cr). Blood samples were collected at rest, just at the end of the standardized 10-min warm-up, and immediately, 30 min and 60 min postexercise. Mean V̇e during exercise was kept for analysis. Airway water and heat losses were estimated using a computational model adapted to the experimental conditions and were compared with a literature-based threshold of bronchial dehydration. Eleven participants exceeded the threshold for bronchial dehydration during exercise (group A) and five did not (group B). Compared with post warm-up, the increase in cc16/cr postexercise was significant (mean increase ± SE: 0.48 ± 0.08 ng·L^-1 only in group A but not in group B (mean difference ± SE: 0.10 ± 0.04 ng·L^-1). This corresponds to an increase of 101 ± 32% [range: 16%-367%] in group A (mean ± SE). Our findings suggest that the use of a computational model may be helpful to estimate an individual dehydration threshold of the airways that is associated with epithelial damage during physical exercise.NEW & NOTEWORTHY Using a computational model for heat and water transfers in the bronchi, we identified a threshold in ventilation during exercise above which airway dehydration is thought to occur. When this threshold was exceeded, epithelial damage was found. This threshold might therefore represent the ventilation upper limit during exercise in susceptible individuals. Our results might help to prevent maladaptation to chronic exercise such as exercise-induced bronchoconstriction or asthma.

Keywords: airway dehydration threshold; computational modeling; exercise ventilation; healthy participants; serum cc16.

Graphical abstract

Figure 1.

Modeling of the evaporation flux in the bronchial tree of each of the 16 male participants. Each graph represents a participant (n = 16 participants, n = 16 pictures). Participants’ graphs are ranked according to their V̇e, from the smallest at the top left to the largest at the bottom right. Data from the participants in group B are shown in the first five graphs on the top left. The dehydration threshold of 0.32 to 0.35 × 10⁻³ µL·mm⁻²·s⁻¹ (10) is represented by a horizontal gray zone. The first generation represents the trachea. FRC, functional residual capacity; V̇e, minute ventilation during 30-min exercise.

Figure 2.

Mean evaporation flux per surface unit in the bronchial tree (J_evap,i) for functional residual capacity of 3 L (A) and 3.5 L (B). J_evap,i was determined at the level of the trachea (1st generation) up to the last generation of the conducting airways (17th generation). The horizontal shaded area delimits the limit at which the water supply is insufficient to cover losses (between 0.32 and 0.35 × 10⁻³ µL·mm⁻²·s⁻¹) (10). The first generation is the trachea. The values correspond to an ambient temperature of 19°C and a relative humidity of 55%.

Figure 3.

Correlation between the change in cc16/cr (Δcc16/cr) and mean ventilation (V̇e) during exercise in the 16 male subjects. The correlation coefficient is a Pearson coefficient (r = 0.69, P = 0.003, n = 16). The linear regression equation is y = 0.0112x-0.6103. cc16, club cell protein; cr: creatinine.

Figure 4.

Maximum change in cc16/cr ratio according to the ventilation status in 16 healthy male participants, after a 30-min continuous exercise at 70% of their maximal work rate. The median [Q1–Q3] was of 106.9% [42.7%–136.6%] in group A and of 15.8% [0%–25.7%] in group B. The middle line in the box is the median and top and bottom line of the boxes are, respectively, Q3 and Q1. The mean is represented by a cross inside the box. The top and bottom bars represent maximum and minimum values, excluding extreme points. The single point is an extreme value. Groups: Participants exceeding the threshold of 0.35 × 10⁻³ µL·mm⁻²·s⁻¹ for bronchial dehydration (group A “above,” n = 11) and those who did not (group B “below,” n = 5). A Mann–Whitney rank sum test was used to compare delta cc16/cr ratio between groups. cc16, club cell protein; cr, creatinine.

Figure 5.

Individual values of cc16/cr from baseline to postexercise in both groups. Baseline values are post warm-up values. Groups: Participants exceeding the value of 0.35 × 10⁻³ µL·mm⁻²·s⁻¹ for bronchial dehydration (group A “above,” n = 11) and those who did not (group B “below,” n = 5). A two-way ANOVA for repeated measures was done with factors being time (pre- and postexercise) and group (A vs. B) (Interaction Time × Group: F = 12.74, P = 0.003). The Holm–Sidak post hoc test for multiple comparisons was applied to localize any difference (see P values on the graph). cc16, club cell protein; cr, creatinine.