Clinical characteristics of chronic obstructive pulmonary disease patients with superoptimal peak inspiratory flow rate

Characteristics of chronic obstructive pulmonary disease (COPD) patients with superoptimal peak inspiratory flow rates (PIFR) has not been thoroughly investigated. This study aimed to compare the characteristics between COPD patients with superoptimal PIFR and those with optimal and sub-optimal PIFR. PIFR was measured using In-Check DIAL G16 and categorized into sub-optimal (PIFR lower than that required by the patient’s device), optimal, and superoptimal (peak PIFR ≥ 90 L/min). Considering COPD patients with sub-optimal PIFR as the reference group, analyses were performed to identify PIFR-related factors. Subgroup analysis was performed according to the forced expiratory volume in 1 s (FEV1) % of the predicted value (%pred). Among 444 post-bronchodilator-confirmed COPD patients from seven tertiary hospitals in South Korea, 98, 223, and 123 were classified into the sub-optimal, optimal, and superoptimal PIFR groups, respectively. The superoptimal PIFR group were younger, had an increased proportion of males, a higher body mass index, lowest number of comorbidities and less frequent exacerbation in the previous year, as well as the highest forced vital capacity %pred. The adjusted odds ratio for frequent exacerbation in the previous year was lower in the superoptimal PIFR group than in the sub-optimal PIFR group and was more pronounced in patients with an FEV1%pred of < 70%. COPD patients with superoptimal PIFR have clinical characteristics different from those patients with the sub-optimal and optimal PIFR. Having a high inspiratory flow may be a favorable trait in COPD.

to effectively use the device, coordination, and their physical abilities.Among them, the inspiratory flow rate is an essential factor when determining the appropriate inhaler type for an individual 1 .
A sub-optimal peak inspiratory flow rate (PIFR) is a common problem and leads to insufficient drug delivery into the lungs to induce effective bronchodilation or other clinical effect.Studies have largely focused on the difference between optimal and sub-optimal PIFR groups [2][3][4][5] and have shown that patients with sub-optimal PIFR are more likely to have advanced-stage disease, older age, and lower lung function compared to patients with optimal PIFR 3 .Sub-optimal PIFR is also related to a shorter time to exacerbation 5 and readmission 6 .However, the optimal PIFR group in these studies included patients with a PIFR > 90 L/min, which is considered a superoptimal, excessive, or a fast PIFR 7,8 .A recent study in stable COPD patients investigated excessive PIFR at > 90 L/min and showed that the majority of excessive PIFR was observed against low-resistance DPI devices, regardless of age, sex, body mass index (BMI), symptom score, and degree of airflow limitation 8 .This high PIFR group was considered problematic based on a priori premise from the study by Usmani et al. 9 .In that study, fast PIFR resulted in drug deposition, mainly in the upper respiratory tract 9 .However, these data were derived from an aerosol generator and not real patients and devices.Another study revealed that high PIFR in COPD patients who were using DPIs exhibited a more favorable inhalation profile than that associated with low PIFR 10 .Thus, there exists a knowledge gap regarding the association between the severity and degree of PIFR and the clinical characteristics of patients with COPD, especially those with a superoptimal PIFR.
In this context, this multi-center observational study in South Korea aimed to compare the clinical characteristics between COPD patients with superoptimal PIFR and those with optimal and sub-optimal PIFR in real-world clinical setting.

Results
Among the 444 COPD patients using DPI, 98 (22.1%), 223 (50.2%), and 123 (27.7%) were classified into the sub-optimal, optimal, and superoptimal PIFR groups, respectively (Table 1).The superoptimal group consisted of younger patients, higher proportion of males, higher BMI, lower Charlson comorbidity index (CCI) scores, higher forced expiratory volume in one second (FEV 1 ) % of the predicted value (%pred), higher forced vital capacity (FVC) %pred, and higher PIFR values compared to the same parameters associated with the sub-optimal group.The proportion of frequent exacerbations in the previous year was significantly lower in the superoptimal PIFR group than in the optimal and sub-optimal groups (Fig. 1, p for trend = 0.015).
In univariable multinomial logistic regression analysis, patients in the superoptimal PIFR group were more likely to be younger, male sex, higher BMI, and current smokers and to have less comorbidity, better lung  a The value was obtained in post-bronchodilator spirometry.b Frequent exacerbation was defined as ≥ 2 moderate or ≥ 1 severe exacerbation in the previous year.*p < 0.017 versus sub-optimal in post-hoc analysis with Bonferroni correction.† p < 0.017 versus sub-optimal in post-hoc analysis with Bonferroni correction.‡ p < 0.017 versus optimal in post-hoc analysis with Bonferroni correction.2).The degree of association, presented as odds ratio (OR) with 95% confidence interval (CI), was more prominent in the superoptimal PIFR group than in the optimal PIFR group.The factors affecting the PIFR in the multivariate multinomial logistic regression model are presented in Table 3.Among the three PIFR groups, the superoptimal group had the youngest age, highest proportion of male sex, highest BMI, lowest CCI score, fewest frequent exacerbations in the previous year, and highest FVC %pred.However, no significant differences were observed between the optimal and sub-optimal PIFR groups, except for males.
Subgroup analysis revealed prominent associations in terms of age, sex, BMI, frequent exacerbations in the previous year, and FVC %pred in COPD patients with more severe airflow limitation and FEV 1 < 70%pred (Table 4).In addition, the superoptimal PIFR group was less likely to experience frequent exacerbations in the previous year than the sub-optimal PIFR group, but this relationship was only observed in patients with FEV 1 < 70%pred (Fig. 1).Exacerbations in the previous year did not significantly differ between the sub-optimal and optimal PIFR groups.

Discussion
Using real-world clinical data of spirometry-confirmed COPD patients across seven tertiary hospitals in South Korea, we have demonstrated that COPD patients with superoptimal PIFR have different characteristics from those with optimal and sub-optimal PIFR.Among the three PIFR groups, COPD patients with superoptimal PIFR  had the youngest age, highest proportion of male sex, highest BMI, lowest CCI score, least frequent exacerbations in the previous year, highest FVC %pred, and highest FEV 1 %pred.This association was more prominent in COPD patients with FEV 1 < 70%pred than in those with ≥ 70%pred.In a real-world clinical setting, clinicians may gain additional insights into PIFR, considering various clinical characteristics, through the measurement of this value using devices such as the In-Check Dial.Notably, frequent exacerbations in the previous year were fewest in the superoptimal PIFR group, whereas no differences were observed between the sub-optimal and optimal PIFR groups.This association was more prominent in COPD patients with FEV 1 < 70%pred than it was in those with ≥ 70%pred.This result suggests that superoptimal PIFR is a distinguishing phenotype in COPD patients using DPI, with a lower probability of exacerbation.This is an extension of the observation that a superoptimal PIFR is closely associated with young age, male sex, higher BMI, lower CCI score, and higher FVC %pred.In line with our findings, previous studies have shown that frequent exacerbation in COPD is related to older age 11 , female sex 12 , lower BMI 13 , and higher comorbidities 14 , all of which imply a bundle of characteristics against the superoptimal PIFR.www.nature.com/scientificreports/Other factors associated with the superoptimal PIFR group were younger age, male sex, higher BMI, lower comorbidity burden, and higher FVC %pred.These factors were more closely related to superoptimal PIFR in COPD patients with FEV 1 < 70%pred than in those with FEV 1 ≥ 70%pred.Our result confirms prior findings that male sex and younger age were more likely to be associated with higher PIFR than female sex and older age were 1,3,15 , Given that inspiratory muscle strength depends on sex, age, and anthropometric indices, the observed finding in our study may be not surprising 16,17 , Respiratory muscle power, which was assessed using the maximum inspiratory pressure, was higher in obese individuals than in eutrophic individuals 16 .In terms of FVC %pred, a weak but significant correlation (r = 0.37, p < 0.001) with PIFR has been reported 3 , which is similar to our results (r = 0.316, p < 0.001) obtained from a Pearson's correlation analysis.In COPD patients with FEV 1 ≥ 70%pred, the airway obstruction may not be sufficiently severe to create a significant effect on PIFR or cause notable differences related to clinical factors, such as BMI or sex.Consequently, the observed difference according to the FEV 1 %pred suggests that superoptimal PIFR in COPD patients with severe airflow limitation can be a favorable trait.In contrast, a sub-optimal PIFR may represent a treatable trait.Notably, inspiratory muscle training increases PIFR in patients with severe COPD 18 .This finding may have clinical implications, suggesting that patients with severe COPD who are unable to achieve an optimal PIFR against DPI may significantly benefit from inspiratory muscle training and that this may represent a treatable trait 19 .
Our study suggests that superoptimal PIFR may be considered as another phenotype of COPD patients who are using DPI, although further longitudinal studies are necessary.Superoptimal, excessive PIFR, is often regarded as inappropriate for optimal drug delivery to the lung [7][8][9] .A previous study showed that a faster inspiratory flow (> 60 L/min) decreased particle deposition in the lungs and increased oropharyngeal deposition 9 .Another study in children with asthma suggested an optimal PIFR range, showing similar clinical outcomes within a range between 30 L/min and 60 L/min or 90 L/min of PIFR for Turbohaler and Diskus, respectively 20 .The concept that there is a maximal value of proper PIFR is based on the observation that more oropharyngeal deposition is related to faster PIFR 21 .However, the actual mean values of the PIFRs for Turbohaler and Diskus were 82.8 L/min and 105.6 L/min, respectively 21 .Similarly, another study showed that the mean PIFR against the R1 device was approximately 80 L/min 4 , and a high proportion of PIFR > 90 L/min was observed in the low-resistance device, which is consistent with our findings.Therefore, considering the heterogeneity within the PIFR group, formerly uniformly categorized as the optimal PIFR group [2][3][4][5]21 , and the difference in clinical characteristics among the sub-optimal, optimal, and superoptimal PIFR groups, further studies are warranted to elucidate the longitudinal effects of superoptimal PIFR in COPD patients. Inthe additional subgroup analysis performed for patients with a superoptimal PIFR according to the FEV 1 % pred and PIFR, the residual volume/ www.nature.com/scientificreports/total lung capacity (RV/TLC%) was lower in those with a PIFR ≥ 100 L/min than in those with a PIFR < 100 L/ min, irrespective of the FEV 1 %pred (Table S1).Additional studies may provide insights into the physiological factors underlying the negative correlation between the PIFR and RV/TLC%.For instance, it would be helpful to measure the total lung capacity across more patients, analyse small and large airway abnormalities using other techniques such as computed tomography or oscillometry, and verify conditions linked to the inspiratory strength (such as muscle strength) 22,23 .It may also be the case that superoptimal PIFR reflects the individuals underlying fitness and thus the effects of delivery of inhaled medication to the lungs becomes less relevant.However, we feel that this would lead to the potential for both effects to cancel each other out: less efficient delivery of drug and better underlying health status.
Our study had some limitations.First, as this was a cross-sectional study, the results should be interpreted with caution.There was a lack of temporality, and causal relationship was not explained.For example, it is inappropriate to conclude that a superoptimal PIFR is beneficial for the future risk of exacerbation.Further longitudinal studies are required to differentiate the clinical course of COPD patients with superoptimal PIFR.Second, there were no data on eosinophil counts or use of inhaled corticosteroids.Given the close relationship among blood eosinophil count, maintenance device therapy, and exacerbation 24 , the application of these factors could alter the observed findings.Third, only the In-Check Dial was used to assess PIFR and categorize the patients into PIFR groups.Although using this device is a popular way to evaluate patients' ability to generalize inspiratory flows, considering other parameters, such as pressure drop, would provide a more relevant way to optimize the DPI device 25 .Also, the assessment of PIFR does not consider inspiratory duration which also has to be adequate to enable effective deposition of treatment into the lungs from a DPI.Finally, although the In-Check Dial has a red-colorized boundary indicating the upper optimal value as 90 L/min, and we utilized the cut-off value of 90 L/min in accordance with previous reports 7,8 , it is important to acknowledge that this is arbitrary and will be affected by the intrinsic resistance of the inhaler device and should be validated in future studies.
In conclusion, superoptimal PIFR can be another phenotype with characteristics different from those of the optimal and sub-optimal PIFR groups.In particular, patients in the superoptimal PIFR group are more likely to be younger and men and have higher BMI, lower comorbidities, fewer frequent exacerbations in the previous year, and higher FVC %pred.This is more pronounced in COPD patients with FEV 1 < 70%pred than in those with predicted FEV 1 ≥ 70%pred, suggesting that superoptimal PIFR may be a favorable trait in severe COPD and encouraging patients in the sub-optimal PIFR group to receive inspiratory muscle training to improve their PIFR.In a real clinical practice, by measuring the PIFR using devices such as the In-Check Dial, clinicians may gain additional insights into PIFR, considering various clinical characteristics.Further longitudinal studies are necessary to identify the clinical course of COPD patients with superoptimal PIFR.

Study design and patients
This multi-center cross-sectional study was conducted in seven tertiary hospitals in South Korea.COPD patients were recruited between June 2021 and November 2021 to evaluate their PIFR who met the following inclusion criteria: (1) aged ≥ 40 years, (2) diagnosis of COPD by post-bronchodilator ratio of FEV 1 /FVC < 0.7 26 , (3) treatment with DPI > 3 months before the recruitment, and (4) regular outpatient visit.During the recruitment process, COPD patients with the following conditions were excluded: (1) patients with a history of asthma or asthma-COPD overlap, (2) patients receiving home oxygen therapy, (3) patients with significant morphological underlying lung diseases such as tuberculosis-destroyed lung or bronchiectasis, and (4) patients with a recent history of severe cardiovascular disease or end-stage cancer.Ultimately, 444 COPD patients were identified.
The study protocol was approved by the Institutional Review Board of Ulsan University Hospital (no.2019-07-038).This study was conducted following the Declaration of Helsinki.All procedures were performed in accordance with relevant guidelines and regulations.

Groups according to PIFR
The PIFR (L/min) generated in the presence of different inhalational resistances was measured using an In-Check Dial G16 (Clement Clarke, UK).The In-Check Dial G16 can be set to the intrinsic resistance of the inhaler that the patient uses.The patients were instructed to fully exhale and then inhale as hard and as fast as possible.The maximum PIFR was obtained during three attempts.The maximum PIFR for each device was recorded separately.

Variables
The most recent values of pulmonary function test measured within 3 months of recruitment were collected.Both pre-and post-bronchodilator results were collected.Data on FEV 1 (L, % pred), FVC (L, % pred), and FEV 1 / FVC (%) were collected.Data on DLCO and residual volume and total lung capacity were available for 301 and 172 patients, respectively.
Exacerbation history in the year prior to recruitment was also collected.A moderate exacerbation was defined as an outpatient visit with a prescription of antibiotics or systemic glucocorticoids.Severe exacerbations were defined as patient visits to the emergency room or requirement of hospitalization because of exacerbation.We

Figure 1 .
Figure 1.Proportion and adjusted aOR of exacerbations in the previous year (any ≥ 2 moderate of ≥ 1 severe) by PIFR and FEV 1 %pred.*OR was adjusted for age, sex, BMI, smoking status, CAT score, CCI score, and FVC %pred.aOR adjusted odds ratio, BD bronchodilator, BMI body mass index, CAT COPD assessment test, CCI Charlson comorbidity index, COPD chronic obstructive pulmonary disease, FEV 1 forced expiratory volume in 1 s, FVC forced vital capacity, %pred % of the predicted value.

Figure 2 .
Figure 2. Distribution of PIFR groups among different DPI resistances.PIFR peak inspiratory flow rate, DPI dry powder inhaler.

Table 1 .
Characteristics of COPD patients according to PIFR.COPD chronic obstructive pulmonary disease, PIFR peak inspiratory flow rate, BMI body mass index, CAT COPD assessment test, mMRC modified medical research council, CCI Charlson comorbidity index, FEV 1 forced expiratory volume in 1 s, FVC forced vital capacity, BD bronchodilator, DLCO diffusion capacity, RV residual volume, TLC total lung capacity, % pred % of the predicted value.Continuous and categorical variables are presented as means with standard deviations and numbers with percentages, respectively.

Table 2 .
Factors affecting PIFR using univariable multinomial logistic regression analysis.PIFR peak inspiratory flow rate, BMI body mass index, CAT COPD assessment test, mMRC modified medical research council, CCI Charlson's comorbidity index, FEV 1 forced expiratory volume in 1 s, FVC forced vital capacity, BD bronchodilator, DLCO diffusion capacity, RV residual volume, TLC total lung capacity, LR likelihood ratio, OR odds ratio, CI confidence interval.%pred % of the predicted value.a The value was obtained in postbronchodilator spirometry.b Frequent exacerbation was defined as ≥ 2 moderate or ≥ 1 severe exacerbation in the previous year.

Table 3 .
Factors affecting PIFR using multinomial multivariable logistic regression analyses.PIFR peak inspiratory flow rate, PFT pulmonary function testing, BMI body mass index, CAT COPD assessment test, CCI Charlson comorbidity index, FEV 1 forced expiratory volume in 1 s, FVC forced vital capacity, LR likelihood ratio, OR odds ratio, CI confidence interval.%pred % of the predicted value.a The value was obtained in postbronchodilator spirometry.b Frequent exacerbation was defined as ≥ 2 moderate or ≥ 1 severe exacerbation in the previous year.

Table 4 .
Subgroup analysis in by post-bronchodilator FEV 1 %pred of factors affecting PIFR using multinomial multivariable logistic regression analyses.PIFR peak inspiratory flow rate, PFT pulmonary function testing, BMI body mass index, CAT COPD assessment test, CCI Charlson comorbidity index, FEV 1 forced expiratory volume in 1 s, FVC forced vital capacity, LR likelihood ratio, NA not available.%pred % of the predicted value.a The value was obtained in post-bronchodilator spirometry.b Frequent exacerbation was defined as ≥ 2 moderate or ≥ 1 severe exacerbation in the previous year.