Investigating distributions of inhaled aerosols in the lungs of post-COVID-19 clusters through a unified imaging and modeling approach

Abstract

Background:

Recent studies, based on clinical data, have identified sex and age as significant factors associated with an increased risk of long COVID. These two factors align with the two post-COVID-19 clusters identified by a deep learning algorithm in computed tomography (CT) lung scans: Cluster 1 (C1), comprising predominantly females with small airway diseases, and Cluster 2 (C2), characterized by older individuals with fibrotic-like patterns. This study aims to assess the distributions of inhaled aerosols in these clusters.

Methods:

140 COVID survivors examined around 112 days post-diagnosis, along with 105 uninfected, non-smoking healthy controls, were studied. Their demographic data and CT scans at full inspiration and expiration were analyzed using a combined imaging and modeling approach. A subject-specific CT-based computational model analysis was utilized to predict airway resistance and particle deposition among C1 and C2 subjects. The cluster-specific structure and function relationships were explored.

Results:

In C1 subjects, distinctive features included airway narrowing, a reduced homothety ratio of daughter over parent branch diameter, and increased airway resistance. Airway resistance was concentrated in the distal region, with a higher fraction of particle deposition in the proximal airways. On the other hand, C2 subjects exhibited airway dilation, an increased homothety ratio, reduced airway resistance, and a shift of resistance concentration towards the proximal region, allowing for deeper particle penetration into the lungs.

Conclusions:

This study revealed unique mechanistic phenotypes of airway resistance and particle deposition in the two post-COVID-19 clusters. The implications of these findings for inhaled drug delivery effectiveness and susceptibility to air pollutants were explored.

1. Introduction

There have been more than a hundred million reported cases of COVID-19 infection in the United States (CDC 2023). Among the survivors, some have exhibited lung injuries, including conditions such as small airway disease and pulmonary fibrosis (Cho et al., Wang et al., 2020). These post-COVID-19 patients, who exhibit a spectrum of enduring physical, cognitive, and mental health symptoms, are diagnosed with long COVID or Post-Acute Sequelae of COVID-19 (PASC) (Groff et al., 2021). The ability to predict the occurrence of long COVID and assess pulmonary complications in COVID survivors is of paramount importance in the realm of medical care and management.

The impact of structural variations in airways on lung functions and drug delivery has been investigated, taking into account factors such as sex (Dominelli et al., 2018, Choi et al., 2015, Kim et al., 2011, Christou et al., 2021), age (Choi et al., 2015, Occhipinti et al., 2017), height (Dominelli et al., 2018, Choi et al., 2015), genetic variation (Smith et al., 2018), and dysanapsis (Smith et al., 2020, Vameghestahbanati et al., 2023). Recently, both sex (female) and age have been recognized as factors associated with an elevated risk of developing long COVID (Perlis et al., 2022, Kessler et al., 2023). For instance, the electronic health records from 2190,579 patients in the National COVID Cohort Collaborative (N3C) were used to train machine learning models for predicting long COVID during the acute infection period (Antony et al., 2023). Age, sex, albuterol usage, obesity, chronic lung disease, and several other factors were identified as key contributors to the model’s predictive accuracy.

Quantitative CT imaging has emerged as a valuable tool for exploring various lung disease manifestations, creating quantifiable biomarkers that can assess a wide range of risk factors and associated lung abnormalities (Hoffman, 2022). Computational fluid and particle dynamics (CFPD) has also emerged as a powerful tool for assessing both the efficiency of drug delivery and the severity of respiratory diseases (Lin et al., 2013, Koullapis et al., 2021, Zhou et al., 2024, Yang et al., 2024, Huang et al., 2021, Kadota et al., 2022). Furthermore, in the era of big data, discovering patient clusters and unique biomarkers specific to these clusters is invaluable (Lin et al., 2021, Choi et al., 2017). These biomarkers serve to monitor disease progression, define disease subtypes, predict health trajectories, and inform product and clinical trial designs (Couper et al., 2014). For instance, a recent study, using deep learning in post-COVID-19 CT lung images, identified two distinct clusters (Li et al., 2022). Cluster 1 (C1) was characterized by a female-dominant group with small airway disease, while Cluster 2 (C2) consisted of older individuals with fibrotic-like lung characteristics. Compared with the healthy control group Cluster 0 (C0), C1 subjects showed an increase in the percentage of voxels with trapped air (AirT%, C0 vs. C1 vs. C2: 2.34±3.26 % vs. 8.15±8.71 % vs. 4.01±4.72 %) and functional small airway disease (fSAD%, C0 vs. C1 vs. C2: 1.08±1.97 % vs. 4.10±5.94 % vs. 1.68±2.59 %). In contrast, C2 subjects displayed a significant increase in the percentage of voxels characterized as ground glass opacity (GGO%, C0 vs. C1 vs. C2: 0.39±0.23 % vs. 3.68±3.80 % vs. 15.85 %±7.20 %), indicating the presence of pulmonary fibrosis.

Intriguingly, the two predictors for long COVID, sex and age, which are derived from clinical data, align with the characteristics of the two clusters, derived from CT lung data acquired from COVID survivors 3–4 months post-viral infection. Hence, this study seeks to elucidate the relationships between the structure and function within each cluster and the corresponding mechanistic phenotypes that may explain their responsiveness or vulnerability to inhaled aerosols. We accomplish this through a unified approach that combines quantitative CT imaging (Choi et al., 2015) with CFPD modeling (Zhang et al., 2022).

2. Methods

2.1. Human subject data

The Institutional Review Board at the University of Iowa approved the study, and all subjects provided informed consent. The study involved 245 individuals. Among them, 140 had previously recovered from SARS-CoV-2 and were prospectively recruited for follow-up examinations (Cho et al., 2022). The remaining 105 participants were healthy non-smoking controls who had not contracted SARS-CoV-2. A comprehensive dataset was collected from each participant, including demographic information, pulmonary function test (PFT) results, and CT scans obtained at both total lung capacity (TLC) and residual volume (RV). A comparison of major demographic and PFT data between the two groups (post-COVID subjects vs. healthy controls) is the following: Age (years): 45.88±15.93 vs. 44.59±14.05; BMI (kg/m²): 32.13±7.58 vs. 25.73±3.59; TLC (L): 5.56±1.35 vs. 5.83±1.24; RV (L): 1.57±0.56 vs. 1.66±0.62; RV/TLC (%): 29.28±8.94 vs. 28.39±7.74; Female (%): 66.42 vs. 51.43. The average time interval between the initial COVID-19 diagnosis and the follow-up visit was 112±39 days.

2.2. Image processing and feature analysis

The VIDA Vision software (VIDA Diagnostics, Iowa) was used to perform segmentation and generate a one-dimensional (1D) airway tree from CT images obtained at TLC. Average luminal diameters ($D=\sqrt{4\times luminalarea/\pi }$) for CT-resolved airways at TLC were computed, along with the homothety ratio D/D_parent, where D is the daughter branch diameter and D_parent is the parent branch diameter. A critical D/D_parent value of 0.79, established for an ideal dichotomic bronchial tree (Mauroy et al., 2004), serves as a gage to determine airway health or remodeling (Grainge et al., 2011, Jones et al., 2016). Additionally, hydraulic airway diameters D_h, which consider both luminal area and perimeter, were calculated to assess airflow resistance in noncircular airways.

To mitigate the impact of sex and age on airway dimensions, the normalization schemes based on sex, age, and height (Choi et al., 2015, Choi et al., 2017) were adopted to calculate normalized hydraulic airway diameter ($D_h^{*}$) and normalized airway wall thickness ($W_T^{*}$). Furthermore, the above CT-based 1D airway tree was used to create an anatomically-consistent whole-lung conducting airway tree using a volume-filling algorithm (Tawhai et al., 2004). When comparing regional airway diameters, we used the subscript ‘Segmental’ to refer to the airways at generations 3–7 and the subscript ‘Terminal’ to denote those at generations 8–27.

The Adapted Multiple Feature Method (AMFM) was employed to quantify GGO% and bronchovascular% (the percentage of the bronchovascular region in the entire lung, measuring the extent of airway passages filled with fluids) (Uppaluri et al., 1999). An image registration method that matches CT images at TLC and RV (Haghighi et al., 2018, Yin et al., 2009) was used to compute the Jacobian (a measure of lung expansion, Jac ≥ 1) (Amelon et al., 2011), regional ventilation, AirT%, and fSAD% (Choi et al., 2015, Choi et al., 2013, Haghighi et al., 2018).

2.3. Cluster-specific clinical and imaging features

The cluster-specific characteristics, as detailed in (Li et al., 2022), are outlined below (see also Table S1) to aid in establishing correlations with CFPD-based variables. C1 had a female-dominant composition (68.37 % females), while C0 and C2 had a more balanced distribution of males and females. C2 subjects (mean age: 62.04±11.85 years) were notably older than those in C0 (44.15±13.93 years) and C1 (43.59±14.76 years). Additionally, C2 subjects displayed impaired lung function, showing reduced FEV₁% and FVC% compared with others. Both C1 and C2 clusters showed an increase in BMI. C1 subjects exhibited the highest percentages of AirT% and fSAD% among all clusters (p<0.05). In contrast, C2 subjects had the highest GGO%, Bronchovalscular%, Tissue%, and RV/TLC%.

2.4. CFPD modeling

A CT-based subject-specific CFPD model of the entire lung was utilized to determine the flowrate fraction within each airway segment (Choi et al., 2019, Yoon et al., 2020) and to predict the airway resistance and particle deposition in both individual branches and the whole lungs (Zhang et al., 2022). The deposition probability in each airway segment was calculated based on a combination of diffusion, sedimentation, and impaction mechanisms. A sinusoidal waveform with a 5-second period was used to mimic tidal breathing, and tidal volume was estimated based on each subject’s ideal body weight (Fuller et al., 2017). The particle diameter ranged from 0.01 to 10.0 μm.

The CFPD model was applied to all three cluster subjects: C0, the control group; C1, a female-dominant subgroup with small airway diseases; and C2, relatively elderly individuals with fibrotic-like lungs. To assess deposition effectiveness, we computed the deposition fraction (DF%), which is defined as the ratio of particles deposited in a specific region to the total particles released. To evaluate ventilation heterogeneity, we computed the coefficient of variation (CV = σ/μ), where μ represents the mean of concentration of both deposited and suspended particles at the end of a breathing cycle, and σ is the standard deviation.

2.5. Statistical analysis

An ANOVA test was performed to determine whether the results across all three clusters were statistically different, with a significance level of p < 0.05 set as the threshold. If an ANOVA test showed significant difference among all clusters, Tukey’s Honest Significant Difference test was utilized as the post-hoc test with a significance level of p < 0.05. For categorical results, logistic regression was used to determine whether one cluster is significantly different from others. Pearson’s correlation coefficient r was applied to evaluate the strength of the relationship between two variables, with |r| > 0.6 considered as strong.

3. Results

3.1. Airway structure and lung deformation

The average percentage of small airways (D < 2 mm) increased in C1, while it decreased in C2 compared with C0 subjects (Fig. S1, C0 vs. C1 vs. C2: 10.5 % vs. 14.5 % vs. 3.1 %, p<0.05). Fig. 1(a) exhibits the $D_h^{*}$. A 3 % decrease in $D_h^{*}$ was observed in C1 subjects compared with C0 subjects. In contrast, C2 subjects exhibited a 2 % increase in $D_h^{*}$. As for $W_T^{*}$, Fig. 1(b) shows that C1 subjects had an 8 % increase compared with C0, while C2 subjects had a 12 % increase. Fig. 2(a) shows the D /D_parent of the three clusters. The ratio decreased in C1, while it increased in C2 compared with C0 (C0 vs. C1 vs. C2: 0.80 vs. 0.77 vs. 0.82, p<0.05). In Fig. 2(b), a notable reduction in Jac was observed in C1 and C2 compared with C0, indicating decreased lung deformation in post-COVID-19 subjects, with the lowest Jac value in C2.

Fig. 1.

Airway structure analysis of post-COVID-19 subjects in each cluster. (a) Normalized hydraulic diameter of CT-resolved airways. (b) Normalized airway wall thickness of CT-resolved airways. Post-hoc testing was done with a Tukey Honest Significant Difference test. *: p<0.05; **: p<0.01; ***: p<0.001.

Fig. 2.

(a) Homothety ratio and (b) Jacobian of post-COVID-19 subjects in each cluster. Post-hoc testing was done with a Tukey Honest Significant Difference test. *: p<0.05; **: p<0.01; ***: p<0.001.

3.2. CFPD-predicted lobar resistance

Fig. 3(a) shows the lobar resistance for the three clusters at peak inspiration. The results indicate that C1 had higher resistance in all five lobes (p<0.05, except between C0 and C1 in LLL), whereas C2 had lower resistance than C0. Moreover, the reduction in resistance among C2 subjects was more significant in the lower lobes than the upper lobes (p<0.05). To examine the regional contribution to the whole lung resistance, Fig. 3(b) displays the relative weights of the resistance in proximal (generation 0–5), central (generation 6–11), and distal (generation 12+) regions, following the demarcation proposed by (Bokov et al., 2014, Bokov et al., 2010). The presence of small airways (D < 2 mm) becomes noticeable roughly from the point where the proximal airways end at generation 5 (Fig. S1). The results showed that the regions with the highest resistance weight differ for each cluster (C0: central, C1: distal, C2: proximal).

Fig. 3.

Airway resistance analysis of post-COVID-19 subjects in each cluster. (a) Airway resistance in left upper lobe (LUL), left lower lobe (LLL), right upper lobe (RUL), right middle lobe (RML), and right lower lobe (RLL). (b) Airway resistance weight in proximal, central and distal regions. Post-hoc testing was done with a Tukey Honest Significant Difference test. *: p<0.05; **: p<0.01; ***: p<0.001.

3.3. CFPD-predicted particle deposition

The total DF% in C2 subjects was significantly reduced in comparison to C0 and C1 subjects, regardless of particle size (p<0.05; Fig. S2). Moreover, the difference in total DF% between C0 and C1 subjects was statistically insignificant. To further reveal changes in regional deposition among clusters, Fig. 4(a)–(c) shows the deposition weights (ratio of regional deposited particles to the total deposited particles) in the proximal, central, and distal regions for each cluster. As particle size increased, the primary deposition area shifted from the distal to the proximal region in all clusters. Additionally, C0 and C1 subjects had higher deposition weights in the proximal region than C2, while C2 had a higher weight in the distal region than C0 and C1, regardless of particle size. While the total DF% values in C0 and C1 remained about the same across particle sizes, the deposition weights for 1.0 and 10.0 μm particles in the proximal and central regions of C1 were approximately 5 % higher than those of C0 (p<0.05). Moreover, the deposition weight in the distal region of C2 exceeded that of C0 by 57 % for 1.0 μm particles, and this difference further increased to 130 % for 10.0 μm particles (p<0.05).

Fig. 4.

Particle deposition analysis of post-COVID-19 subjects in each cluster. Deposition weight of (a) 0.01 μm, (b) 1.0 μm, and (c) 10.0 μm particles in proximal, central and distal regions. Post-hoc testing was done with a Tukey Honest Significant Difference test. **: p<0.01; ***: p<0.001.

Fig. 5 compares the CVs for 0.5 μm particles in airway segments of all cluster subjects. While C0 and C2 subjects did not differ significantly in CV, C1 subjects had higher CV values than others (p<0.05). The selection of 0.5 μm particles enabled a comparison with our previous study involving COPD subjects (Zhang et al., 2022), which used Technetium-99 m sulfur colloid (< 1.0 μm) for ventilation SPECT imaging.

Fig. 5.

Coefficient of variation of 0.5μm particle in each cluster. Post-hoc testing was done with a Tukey Honest Significant Difference test. ***: p<0.001.

4. Discussion

This study aimed to explore the characteristics linked to the two key predictors of developing long COVID: being female and older age. The investigation was prompted by the observation that the two post-COVID-19 clusters identified from CT lung data were distinctively associated with female gender and older age, respectively. We utilized an integrated approach that combines quantitative CT imaging and CFPD modeling to explore the interplay between structure and function in two post-COVID-19 clusters. Model-predicted airway resistance and particle deposition were used to evaluate each cluster’s susceptibility to inhaled aerosols and to identify mechanistic phenotypes that may exhibit traits or features increasing their responsiveness or vulnerability to inhaled aerosols, linking post-COVID lung state and aerosol inhalation.

The homothety ratio arises from the notion of fractal trees that can fill space efficiently while minimizing dissipation in the design of an optimal human bronchial tree, thus serving as a sensitive biomarker in characterizing airway remodeling and decline in lung function (Mauroy et al., 2004). The homothety ratio is about 0.85 (Mauroy et al., 2004) or in the range of 0.80–0.85 (Bokov et al., 2010) for normal human lungs, while it is reduced to about 0.77 in COPD patients (Bokov et al., 2010). The homothety ratio in C0 subjects was approximately 0.81±0.02, agreeing with the reported range for normal. The ratio in C1 subjects (0.79±0.02) was smaller than normal. In contrast, the ratio in C2 subjects (0.83±0.03) was larger than that of C0, but within the reported range for normal. Although the differences in homothety ratio between clusters were minor, they could influence respiratory performance by increasing and decreasing resistance in C1 and C2 subjects, similar to the findings reported regarding the role of homothety ratio in COPD patients (Bokov et al., 2014, Bokov et al., 2010).

Besides the homothety ratio, other variables might contribute to an increase in resistance in C1 subjects, such as a higher ratio of small airways with decreased diameter and increased wall thickness, and an increase in small airway disease. These features might be manifestations of constrictive bronchiolitis that involves airway inflammation and narrowing in the long-term sequelae of COVID-19 pneumonia (Elicker, 2022). As resistance increases, C1 subjects may necessitate greater energy expenditure for breathing. This could offer insight into the observed dyspnea among post-COVID-19 subjects (Wirth and Scheibenbogen, 2022). Reports suggest that female subjects generally exhibit smaller airway dimensions than males (Dominelli et al., 2018, Choi et al., 2015, Kim et al., 2011, Christou et al., 2021). However, the relationship between this discrepancy and the prevalence of small airway disease in C1 subjects, particularly in the context of female dominance, remains unexplored.

Conversely, the elevation of the homothety ratio among C2 subjects correlates with an expanded airway diameter, which may explain the decrease in resistance compared with C0 subjects. Furthermore, an elevation in GGO% among C2 subjects may be attributed to remodeling due to traction bronchiectasis. These observations are consistent with a previous study that fibrosis can reduce airway resistance (Campbell and Sapra, 2023). A greater decrease in resistance was observed in the lower lobes compared to the upper lobes in C2, suggesting more extensive damage in the lower lobes. This reversal of resistance towards the upper lobes aligns well with the study from Liao et al, which reported a more significant increase in CT score in the lower lobes (Liao et al., 2022). Furthermore, the shift of resistance weight to the proximal airways in C2 subjects could lead to the accumulation of mucus within these airways, contributing to an escalation in airway wall thickness (Khan et al., 2021, Miles Campbell, 2022).

C1 and C2 subjects demonstrated a significant reduction in lung deformation when contrasted with C0 subjects. This feature was evident in the lung size at TLC (Table S1), being in agreement with Zhang et al. (2022). The size reduction at TLC signifying a decline in respiratory muscle strength within both clusters (McNarry et al., 2022). Additionally, the decrease in RV/TLC% observed in C1 and C2 subjects might lower blood oxygen levels due to decreased oxygen intake. This, in turn, might cause vascular enlargement, as evidenced by increased bronchovascular markings (Lv et al., 2020).

Fig. 6(a) presents a correlation map to explore comprehensively the interactions between modeling variables (resistance, DF%, CV, and weights), imaging variables, and clinical data. Several key observations are summarized below:

• D/D_parent exhibits a positive correlation with segmental airway diameter and a negative correlation with resistance. In other words, airway narrowing decreases the homothety ratio, resulting in increased resistance as the resistance weight shifts from the proximal to the distal region.

• Within the variables examined, fSAD%, GGO%, and RV/TLC exhibit the most pronounced negative correlation with the lung expansion measure, Jac. Conversely, FEV₁% demonstrates the most robust positive correlation with Jac.

• DF% shows a negative correlation with both segmental and terminal airway diameters.

• CV exhibits the most robust positive correlation with fSAD%.

Fig. 6.

Pearson correlation maps between (a) modeling, imaging, and clinical variables and (b) resistance weight (RW) and deposition weights (DW).

Fig 6(b) illustrates a correlation map that explores the interactions between resistance weights and deposition weights for particles with diameters of 0.01, 1.0, and 10.0 μm in the proximal, central, and distal regions. Notably, the distal resistance weight positively correlates with the proximal DF weights for 1.0 and 10.0 μm particles. Furthermore, although the correlation coefficients between the distal resistance weight and the deposition weights for 10 μm particles in the central and distal regions (−0.56 and −0.5) do not reach the absolute value threshold of 0.6, they remain close. These observations reveal a consistent pattern: as the distal resistance weight increases, particles with diameters around 1.0 and 10.0 μm are more likely to deposit in the proximal region.

The structural changes in the post-COVID-19 lungs may have implications for the deposition of a range of inhaled aerosols in the respiratory system, including virus-laden aerosols (0.01–0.20 μm), pharmaceutical aerosols (≤ 5.0 μm), and air pollutants (2.5–15.0 μm). While the total deposition fraction during respiration in C1 subjects does not significantly differ from C0 subjects across particle sizes, C1 subjects exhibit airway constriction, leading to a redistribution of small particle deposition around 1.0 μm toward the proximal region. In contrast, C2 subjects, with expanded airways, demonstrate reduced overall deposition compared with C0 and C1 subjects while directing the deposition weight towards the distal region. These observations suggest that pharmaceutical aerosols may have a higher tendency to deposit in the proximal airways of C1 subjects compared with C0, while penetrating deeper into the lungs of C2 subjects. Furthermore, these findings imply that air pollutants might potentially reach deeper into the lungs of C2 subjects, potentially rendering them more susceptible to environmental risk factors. Recent research has shown promising outcomes in the treatment of post-COVID-19 subjects with fibrotic lung using orally administered corticosteroids (Myall et al., 2021). Additionally, prolonged exposure to PM_2.5 over an extended period may heighten the susceptibility to post-COVID symptoms (Yu et al., 2023, Zhang et al., 2023). Improving our understanding of cluster-specific mechanistic phenotypes can inform the design of studies aimed at investigating delivery of and susceptibility to inhaled aerosols.

Fig. 7 illustrates the distribution and heterogeneity of inhaled particles for representative cluster subjects. These subjects were selected based on significant CT and CFPD metrics, such as $D_h^{*}$, $W_T^{*}$, resistance and deposition fraction, being closest to the mean values in their respective clusters. Among them, C1 subjects consistently displayed higher CV values, regardless of particle size, than C0 and C2 subjects. This analysis confirms that it is predominantly small airway disease that plays a major role in the elevated CV values and ventilation heterogeneity observed in C1 subjects. In contrast, C2 subjects consistently demonstrated the lowest CV values across all clusters, irrespective of particle size. This suggests more uniform ventilation attributed to expanded airway lumen areas. Furthermore, the color-coded deposition fraction prominently shows green for large 10-μm particles in C2 subjects, confirming deeper penetration of these larger particles. As compared with COPD patients (Zhang et al., 2022), the CV values in severe COPD patients exceeded those in C1 and C2 subjects, attributed to the rising levels of both small airway disease and emphysema in COPD patients.

Fig. 7.

Spatial distribution and heterogeneity of inhaled particles with diameters of (a) 0.01 μm, (b) 1.0 μm, and (c) 10.0 μm in representative cluster subjects.

5. Limitation

This study has several limitations. First, the sample size is restricted by the cost of CT scanning, resulting in a relatively small cohort. Consequently, there may exist subclusters among C1 and C2 subjects. Second, all subjects were simulated under an estimated tidal breathing condition, introducing uncertainty regarding the impact of tidal breathing volume. Third, the long-term effects of the particle deposition characteristics identified for each cluster remain to be explored through follow-up human subjects studies. Furthermore, the relationships between long COVID symptoms and CFPD variables are yet to be established.

6. Conclusion

Recent studies have identified sex and age as significant predictors of long COVID development. To comprehend the underlying mechanisms linked to these attributes, we examined the intricate relationships between structure and function in two post-COVID-19 clusters characterized respectively by female (C1) and older age (C2), comparing them to healthy controls (C0).

In C1 subjects, characterized by small airway disease, we observed reduced luminal areas and a diminished homothety ratio, leading to increased airway resistance and a redistribution of resistance weight toward the distal region. This, in turn, increased particle deposition in the proximal airways, indicating a potential obstacle to the effective delivery of inhaled pharmaceutical drugs to respiratory regions in C1 subjects.

Conversely, C2 subjects exhibited more pronounced traction fibrosis and expanded lumen areas within the airways. This expansion of the lumen area raised the homothety ratio, reducing airway resistance and shifting resistance weight toward the proximal airways. Although particle deposition decreased in C2 subjects compared with C0 subjects, smaller particles, like pharmaceutical aerosols, showed a propensity to penetrate deeper into the lungs. Similarly, larger particles tended to settle in respiratory regions in C2 subjects, potentially increasing susceptibility to environmental risks.

Data availability

Data will be made available on request.