Impulse oscillometry (IOS) is a non-invasive technique used to assess respiratory mechanics by measuring airway resistance and reactance during tidal breathing [1-3]. Unlike spirometry, which requires forced expiratory maneuvers and active patient cooperation, IOS provides information about both central and peripheral airway function by applying pressure oscillations at different frequencies. Low-frequency signals (5 Hz) penetrate the peripheral airways, while high-frequency signals (20 Hz) primarily reflect central airway function [4-6]. In recent years, IOS has emerged as a valuable tool for assessing respiratory function in asthma patients, especially for identifying small airway dysfunction and monitoring disease control [7-9]. The technique provides several key parameters: total respiratory system resistance at 5 Hz (R5), resistance at 20 Hz (R20), reactance at 5 Hz (X5), resonant frequency (Fres), area under the reactance curve (AX), and frequency dependence of resistance between R5 and R20 (R5-R20). These parameters have demonstrated significant associations with spirometric indices and may be more sensitive in detecting airway obstruction compared to traditional spirometry [10-12]. In this context, IOS offers a distinct advantage over spirometry by providing an effort-independent assessment and enhanced sensitivity to peripheral airway abnormalities. This allows IOS to capture clinically relevant airway dysfunction that may not be reflected by conventional spirometric measures alone [3,6]. Asthma exacerbations are a crucial aspect of disease management, often linked to increased airway inflammation and obstruction. Studies have shown that IOS parameters can identify uncontrolled asthma more effectively than conventional spirometry, especially in patients with normal spirometric values [13-15]. The %-predicted R5-R20 has emerged as a particularly useful parameter, with studies demonstrating its superior diagnostic ability in detecting uncontrolled asthma compared to forced expiratory volume in 1 second (FEV₁) [16].

However, standardized cutoff values for IOS parameters have not yet been established, and further research is needed to assess their clinical applicability. To address this gap, this study examined the relationship between IOS parameters and exacerbation events in asthma patients, emphasizing their potential for risk stratification and management. Additionally, it evaluated the diagnostic performance of individual IOS indices in predicting exacerbations.

This study reveals significant associations between IOS parameters and exacerbation events in asthma patients, highlighting their clinical relevance in modern asthma management. Patients experiencing exacerbations displayed distinct IOS profiles, marked by increased airway resistance, elevated Fres, and greater frequency dependence of resistance. These changes reflect pathophysiological alterations in the small airways that often precede noticeable clinical deterioration. The findings support the idea that IOS can serve as a sensitive indicator of small airway involvement, providing insights that traditional spirometry may overlook.

From a pathophysiological standpoint, small airway dysfunction is characterized by airflow heterogeneity, dynamic airway closure, and impaired elastic recoil, which can persist even in clinically stable asthma. Abnormalities derived from IOS, especially increased R5-R20 and more negative X5 or elevated AX, may represent ongoing peripheral airway impairment that spirometry does not adequately capture. Such residual small airway issues can exacerbate airway instability and ventilation inhomogeneity during inflammatory events, thus lowering the threshold for acute exacerbations. In this context, the observed links between IOS metrics and exacerbation events are clinically plausible and reflect significant airway pathology rather than mere statistical correlations.

Among the parameters analyzed, AX emerged as the most robust marker, exhibiting the strongest correlations with spirometric indices (r=-0.51 with FEV₁, r=-0.42 with FEF_25-75%) and the largest effect size (d=0.91). In this study, most patients showed preserved FEV₁/ FVC ratios, indicating relatively normal spirometric profiles. In this context, AX may offer superior discriminatory ability compared to R5-R20, as it not only reflects the frequency dependence of resistance but also integrates information on lung compliance, airway closure, and the heterogeneity of peripheral airway function. This likely explains why AX demonstrated a higher AUC than R5-R20 in our cohort, suggesting that AX is particularly sensitive to subtle impairments in small airway function that conventional spirometry may miss.

Consistent with the findings of Kim et al. [21], who reported a significant association between AX and asthma in patients with preserved lung function, our results further emphasize the diagnostic value of AX in detecting small airway dysfunction, even when spirometric parameters appear normal. Fres (d=0.74) and R5-R20 (d=0.66) demonstrated moderate-to-large effect sizes, reinforcing their importance as indicators of exacerbation risk. Although detailed smoking history (packyears) was not available for all participants and could not be fully incorporated into multivariable modeling in this study, subclinical smoking-related airway changes may partially influence IOS, particularly AX, which integrates resistance and reactance. Therefore, residual confounding by smoking/chronic obstructive pulmonary disease overlap cannot be excluded. In contrast, R20 showed a negligible effect, while X5 exhibited a moderately large negative effect, underscoring the differential diagnostic value of various IOS indices. Effect size analysis further highlights that these differences are not only statistically significant but also clinically meaningful, extending the interpretation beyond p-values alone.

ROC analysis confirmed the diagnostic potential of IOS. AX demonstrated the highest discriminatory performance for exacerbation events (AUC=0.673), followed closely by R5-R20 (0.658) and Fres (0.644). Although the discriminative ability of these parameters was moderate, they provide valuable complementary information when integrated into a broader clinical evaluation. Importantly, the optimal Fres threshold identified in this study (23.237 Hz) was higher than in previous reports, which may reflect bias introduced by a small number of patients with markedly elevated Fres values. This underscores the need for larger validation studies to establish reliable and clinically applicable cutoffs.

The results of this study are in line with prior studies showing that IOS, especially R5-R20, can identify uncontrolled asthma more effectively than spirometry [22-24]. Liwsrisakun et al. [22] reported that a %-predicted R5-R20 ≥200 was superior to FEV₁ in predicting uncontrolled asthma, with an AUC of 0.939 compared to 0.712. These results, along with the data, emphasize the enhanced sensitivity of IOS in detecting small airway dysfunction, even in patients with preserved FEV₁/FVC ratios where conventional spirometry may seem normal. This is especially important for asthma phenotypes that exhibit disproportionate small airway involvement. In exploratory analyses, FeNO levels and peripheral blood eosinophil counts did not show significant associations with IOS parameters and had limited ability to predict exacerbation risk when analyzed independently. Although there were nonsignificant trends indicating higher values with increased exacerbation severity, the inclusion of AX significantly improved model performance, and AX remained independently linked to exacerbation risk. This disconnect may be attributed to several factors. First, FeNO and blood eosinophils primarily indicate current type 2 inflammatory activity, while IOS assesses mechanical small airway dysfunction that can persist despite varying inflammatory signals. Second, clinical diversity, including mixed asthma phenotypes such as cough-variant asthma, may obscure direct correlations between inflammatory biomarkers and IOS indices. Third, relying on single time-point biomarker measurements may restrict the ability to capture the cumulative or longitudinal inflammatory burden that is relevant to exacerbation risk. Overall, these findings suggest that IOS effectively captures mechanical aspects of small airway dysfunction that complement, but are not fully represented by, conventional type 2 inflammatory biomarkers.

The differences in IOS parameters between patients with and without exacerbations highlight small airway dysfunction as a ‘treatable trait’ in asthma management. The elevated values of R5, Fres, AX, and R5-R20 in the exacerbation group indicate increased peripheral airway obstruction and reduced lung compliance, which align with ongoing airway inflammation and remodeling [25]. This aligns with current precision medicine strategies that emphasize phenotype-directed therapy based on specific biomarkers and functional measurements.

The findings from IOS have practical therapeutic implications. Extra-fine particle formulations of inhaled corticosteroids and bronchodilators have demonstrated increased efficacy in patients with IOS-defined small airway dysfunction. Furthermore, identifying such dysfunction may help guide the use of biologic therapies in patients with type 2-high asthma phenotypes. Pediatric studies also support the value of IOS, with Hu et al. [13] showing that IOS combined with FeNO achieved excellent diagnostic performance (AUC=0.95) for detecting uncontrolled asthma in children. This suggests that multimodal approaches integrating IOS with other biomarkers can enhance phenotyping and risk stratification across age groups.

Looking ahead, the integration of IOS with emerging biomarkers and advanced computational methods shows promise for personalized asthma care. Advances in artificial intelligence and machine learning applied to oscillometry data may enable the development of predictive algorithms that can identify patients at high risk of exacerbations, forecast treatment responses, and guide individualized therapy [9].

This exploratory analysis, while not the primary focus of the study, revealed a slight improvement in predictive performance when combining AX, R5-R20, and Fres compared to single-parameter models. This suggests that multi-parameter approaches could enhance risk prediction. However, these findings need to be validated in larger, prospective datasets.

Several limitations must be acknowledged. The retrospective, single-center design restricts generalizability, and the relatively small sample size, especially in the severe obstruction group, may have limited statistical power. Exacerbation events were identified from medical records, which may underestimate mild episodes managed in the community. Additionally, the cross-sectional design prevents conclusions about causality or the predictive value of baseline IOS for future exacerbations. Future research should prioritize prospective, multi-center validation using standardized definitions for exacerbations, larger sample sizes, and longitudinal follow-up to determine whether IOS parameters can reliably predict future outcomes.

In summary, this study shows that IOS provides clinically relevant insights into respiratory mechanics in asthma and complements spirometry by detecting small airway dysfunction. Parameters like AX, R5-R20, and Fres demonstrated significant differences between patients with and without exacerbation events, exhibiting moderate but consistent diagnostic performance. Effect size analyses further underscore the clinical significance of these findings. While IOS parameters should not be used in isolation, their integration into a comprehensive asthma assessment could enhance risk stratification, inform treatment decisions, and support precision medicine approaches. With further validation and standardization, IOS has the potential to become a vital component of asthma management.