Asthma–chronic obstructive pulmonary disease (COPD) overlap (ACO) is a condition characterized by overlapping clinical features of asthma and COPD. It has recently been recognized as a disease, but debate remains whether it is a single disease entity distinct from asthma and COPD.

Patients with features of both asthma and COPD have been excluded from clinical studies on either disease, resulting in a lack of information about ACO from prevalence to course and outcomes of the disease. Furthermore, as there are as yet no unified diagnostic criteria for ACO, the exact prevalence of ACO is unknown because it varies according to the diagnostic criteria used to define the disease in each study with estimates ranging from between 9 and 55% (1).

Efforts have been made to identify patients with ACO in several chronic airway disease cohorts focused on either asthma or COPD. Several cohort studies have reported relatively consistent data on ACO, indicating that they have more severe respiratory symptoms, poorer quality of life status and lung function, experience more frequent exacerbations, and have higher mortality rates than patients with COPD alone (2–6). This clinical pattern of patients with ACO was similar in a Korean COPD cohort even defined using several different diagnostic criteria (7). Although there have been racial comparisons within a single cohort, no studies have compared racial differences in ACO between distinct large cohorts.

Therefore, this study was performed to evaluate differences in the prevalence of ACO using identical diagnostic criteria according to race and ethnicity in two large-scale nationwide cohorts of COPD and to compare the clinical parameters and the risk for exacerbation between patients with ACO of different races.

The prevalence rates of ACO using identical diagnostic criteria composed of BDR and blood eosinophil count were similar in two cohorts but lower in the AA group when divided into three racial and ethnic groups. Asian patients with ACO were more likely to be male, older, with lower BMI and smoking pack-years, and fewer symptoms (fewer dyspneic and chronic bronchitis symptoms) compared to the NHW and AA patients with ACO. Lung function was lower in Asian patients with ACO, and the proportion of BDR positive subjects was small, and they were prescribed more maintenance inhaler treatment. Patients with ACO of the three racial groups experienced moderate-to-severe exacerbations more frequently compared to patients with non-ACO, and although the effects were not significant within each racial group, the use of ICS in total patients with ACO reduced moderate-to-severe exacerbation risk.

We defined ACO by BDR >15% and 400 ml and/or blood eosinophil count ≥300/μl among patients with COPD classified by age, smoking history, and fixed airflow limitation. As there are no unified diagnostic criteria for ACO, studies have yielded inconsistent results depending on the criteria used. We applied the same criteria consistent with the most recently proposed definition of ACO by the Spanish Guidelines for COPD and GEMA (11, 13). The results indicated similar prevalence rates of ACO among the three racial groups.

Asian patients with ACO are more likely to be older, male, and have less smoking history compared to NHW and AA patients with ACO. The low percentage of females among Asian patients with ACO was due to the low rate of smoking among females in Asia. The results of this study were compatible with previous reports. A previous prospective study indicated that the percentage of females among Asian patients with COPD was 5% (14), and the percentage of females in the IMPACT trial in Asia was 5% (15). In addition, Asian patients with ACO had significantly lower BMI compared to NHW and AA patients, which may have been due to genetic factors and differences in nutrition and dietary habits. The low BMI in Asian patients with COPD was consistent with previous studies. For example, the mean BMI in the Hokkaido (Japanese) cohort was 22 kg/m² compared to 27 kg/m² in the ECLIPSE cohort (16), and the mean BMI of Asian subjects in the FLAME study was 22.4 kg/m² (17).

With regard to comorbidities, compared to NHW and AA groups, Asian patients with ACO had lower rates of ischemic heart disease, dyslipidemia, cerebrovascular disease, and gastroesophageal reflux disease. Similar patterns were reported in previous COPD studies (18–20). As smoking is well-known to be the leading cause of COPD and atherosclerosis-related vascular disease (21), both environmental and genetic factors may have contributed to these findings. In this study, the prevalence of the previous history of asthma was significantly higher in AA and Asian patients with ACO compared to NHW patients. Compared to NHW, AA patients have a higher prevalence of childhood asthma (22), and this maybe related to the relatively younger age in the AA group in the present study, and Çolak et al. (23) reported that the percentage of patients with asthma was significantly elevated in cases of early COPD. These sociodemographic characteristics may interact with the prevalence of comorbidities as risk factors for COPD and ACO, and it is assumed that they also contributed to the observed racial differences. When the demographic and clinical features of patients with non-ACO COPD in the three racial and ethnic groups were compared, similar to those observed in patients with ACO were identified, although there were some differences (data are not shown).

There is broad agreement that patients with ACO experience frequent exacerbations, have a poor quality of life, more rapid decline in lung function, higher mortality rate, and greater use of healthcare resources compared to patients with asthma or COPD alone (1). We further compared patients with ACO and non-ACO in each cohort. In COPDGene, ACO subjects had a male predominance with fewer current smokers and higher rates of asthma history, favorable lung function, more BDR positivity, and higher blood eosinophil count than patients with non-ACO. Inhaler prescription status did not differ between ACO and non-ACO patients (see Supplementary Table 2). In KOCOSS, patients with ACO were also older, more BDR positivity, with higher blood eosinophil count compared to patients with non-ACO. There were no differences in previous asthma history or inhaler prescription status between patients with ACO and non-ACO (see Supplementary Table 3). These results were consistent with previously reported clinical findings in patients with ACO. During 1 year of follow-up, patients with ACO experience more exacerbations than non-ACO patients, and the difference was significant in NHW and Asian racial groups. Moreover, the rates of severe exacerbations were significantly higher in total and NHW patients with ACO. As the risk for exacerbations tends to be higher in patients with ACO than patients with non-ACO (6, 24, 25), our results were consistent with previous studies despite the low prescription rate of maintenance inhaler treatment regardless of ACO or non-ACO.

Although Asian patients with ACO had fewer past exacerbations than other ethnic group of ACOs, they experienced more exacerbations during 1-year follow-up. This is thought to be related to several risk factors of subsequent exacerbation risk, such as older age, had lower BMI, lower lung function, higher eosinophil count, and male predominance in Asian patients with ACO. Moreover, ~55.9% of patients with COPD in COPDGene and 16.6% of patients with COPD in KOCOSS did not use any long-acting bronchodilator and/or ICS. This implies undertreatment of patients with COPD regardless of ACO in real-world practice and suggests that more active treatment is needed in these patients. Also, Asian patients with ACO were receiving more maintenance inhaler treatment at the time of enrollment, and this suggests that they may already be more severe patients than other ethnic group of ACOs.

There have been inconsistent reports regarding the effects of ICS use in patients with ACO (7, 26–28). Recently, it was reported that ICS has reduced the risk for exacerbation in the KOCOSS cohort, but the protective effect of ICS was found only in ACO by the GINA/GOLD criteria among the five sets of diagnostic criteria for ACO (7). As asthma and COPD are umbrella labels for heterogeneous conditions, each of which includes several different clinical phenotypes with several different underlying mechanisms, ACO is also considered a clinical term including several different clinical phenotypes. Therefore, the use of ICS in ACO may or may not be protective in some cases, and this will depend mainly on the clinical features that determine the diagnosis of ACO. In this study, we found a protective effect of ICS in risk reduction of moderate-to-severe exacerbation in total patients with ACO, but the effect was not significant in any racial group of patients with ACO. On the other hand, the use of ICS in total patients with non-ACO led to increased risk for exacerbation (aIRR, 1.30; 95% CI, 1.11–1.54 for moderate-to-severe exacerbation and aIRR, 1.38; 95% CI, 1.05–1.83 for severe exacerbation), and the effects were also significant in Asian patients with non-ACO COPD. In fact, ICS was prescribed similarly in patients with non-ACO and ACO of all races (33.5 vs. 37.6% in NHW, 38.4 vs. 35.2% in AA, and 34.4 vs. 37.0% in Asian patients with non-ACO vs. ACO, respectively). This may have influenced the effects of ICS on exacerbation in both patients with ACO and non-ACO. There is some debate regarding the protective effect of ICS on exacerbation in ACO, but as it maybe harmful in patients with non-ACO, more attention will be needed to reduce overuse of ICS in COPD, particularly in patients with non-ACO.

This study had some limitations. First, KOCOSS is a single nationwide cohort study. Although Korean patients with COPD had similar overall clinical features to other Asian patients with COPD, such as Japanese, Chinese, and Taiwanese patients, the degree of exposure to cigarette smoke or biomass fuels differs between countries (18), which may limit the generalizability of our findings to patients with COPD in other Asian countries. Second, considering the GOLD recommendation of the “ABCD” assessment since 2011, the differences in the timing of patient enrollment in the two cohorts may have affected therapeutic approaches. Third, we used extremely high BDR to define ACO, which is similar to the previous study (29), but there were no quantitative chest CT data in the KOCOSS cohort, which made it impossible to compare differences in imaging findings, such as the proportion of emphysema, across the ACO groups. Fourth, another criterion for identifying ACO in our study was high blood eosinophil count. There has been a report that the risk of exacerbation increases in patients with COPD whose blood eosinophil counts of 300/μl or greater (30), but identifying high blood eosinophil count in COPD as a marker of eosinophilic COPD or ACO is controversial. Fifth, to apply the same criteria for defining COPD, non-smoking COPD could not be considered by including smoking history, thus, there is a possibility of bias due to the difference in smoking rates according to gender in the two countries. Sixth, only the use of ICS at the baseline was included in the analysis, so the number of cases in which inhaler therapy was changed during the follow-up period was not reflected. Lastly, exacerbations were ascertained differently in the American and Korean COPD cohorts. Korean subjects were interviewed in person every visit, thus, the frequency of exacerbations maybe higher in Korean subjects.

In conclusion, this is the first report of similar prevalence rates of ACO among three different race and ethnicity groups using identical diagnostic criteria in large COPD cohorts. There were several differences in clinical features and inhaler prescription patterns, but the tendency of increased risk for exacerbation in ACO was similar across all racial groups. The use of ICS in ACO reduced moderate-to-severe exacerbation risk during 1 year of follow-up, but the effect was not significant in each racial group of patients with ACO. However, the use of ICS in patients with non-ACO COPD increased the risk for exacerbation. Therefore, further attention and greater caution are needed with regard to the use of ICS in patients with COPD with a low likelihood of ACO.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

All hospitals involved in the KOCOSS and COPDGene cohort studies obtained approval from the relevant Institutional Review Board and patients provided informed consent, including at KONKUK University Medical Center (IRB No. KHH1010338). The patients/participants provided their written informed consent to participate in this study.

YSJ and CKR take responsibility for the data and analysis and designed the study. Members of the Korean Asthma Research Group contributed to the initiation of the ACO project and data acquisition. YSJ performed statistical analysis of data and wrote the first draft of the manuscript. All authors provided critical reviews. All authors contributed to the article and approved the submitted version.

This research was supported by the Research Program funded Korea National Institute of Health (2016ER670100, 2016ER670101, 2016ER670102, 2018ER67100, 2018ER67101, 2018ER67102, and 202104044B8-00) for the KOCOSS cohort, and COPDGene was supported by Award Numbers U01 HL089897 and U01 HL089856 from the National Heart, Lung, and Blood Institute. COPDGene is also supported by the COPD Foundation through contributions made to an Industry Advisory Board that has included AstraZeneca, Bayer Pharmaceuticals, Boehringer-Ingelheim, Genentech, GlaxoSmithKline, Novartis, Pfizer, and Sunovion.

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Heart, Lung, and Blood Institute or the National Institutes of Health.

CR received consulting/lecture fees from MSD, AstraZeneca, GSK, Novartis, Takeda, Mundipharma, Boehringer-Ingelheim, Teva, Sanofi, and Bayer, but the sponsors had no role in the design of the study, the collection and analysis of data, or writing of the manuscript. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.