Global prevalence of asthma-COPD overlap (ACO) in the general population: a systematic review and meta-analysis

Background Asthma-COPD overlap (ACO) is a term that encompasses patients with features of both asthma and COPD. To date, the global prevalence of ACO in the general population remains unknown. The objective of this study was to estimate the prevalence of ACO in the general population using a systematic review and meta-analysis. Methods A systematic search of ISI Web of Knowledge, MEDLINE/PubMed, and Scopus was performed up to May 2019 to identify studies reporting the prevalence of ACO. Reference lists from identified studies and relevant review articles were also searched. Eligibility criteria were studies reporting the prevalence of ACO, performed in general population, and published in English language. Pooled prevalence of ACO with 95% confidence interval (CI) was calculated using random effects Meta-analysis. Results A total of 27 studies were included in this meta-analysis. The Cochran Q test and I2 statistics revealed substantial heterogeneity among studies. Based on the random-effects model, the pooled prevalence of ACO was 2.0% (95% CI: 1.4–2.6%) in the general population, 26.5% (95% CI: 19.5–33.6%) among patients with asthma, and 29.6% (95% CI: 19.3–39.9%) among patients with COPD. In addition, for included studies, the global prevalence of asthma-only was 6.2% (95% CI: 5.0–7.4%) and COPD-only was 4.9% (95% CI: 4.3–5.5%). Conclusion We estimated the global prevalence of ACO based on population-based studies and found that 2.0% of the general population is affected. However, the prevalence of ACO depends on its diagnostic criteria. Therefore, there is a vital need to better define the ACO diagnostic criteria, management and treatment. It is worth noting that the limitations of the present study include lack of studies in some region of the world and small number of studies included in the subgroup analyses.


Background
Asthma and chronic obstructive pulmonary disease (COPD) are major public health problem and represent a leading cause of morbidity and mortality worldwide [1,2]. Asthma and COPD are the most common chronic respiratory diseases worldwide each with a unique natural history and pathophysiology [1]. Asthma is usually characterized by chronic airway inflammation whereas COPD is characterized by persistent respiratory symptoms and chronic inflammation of the airways [3,4]. However, patients can sometimes have clinical features of both diseases, and this condition has been termed asthma-COPD overlap (ACO), recommended by joint GINA (Global Initiative for Asthma) and GOLD (Global Initiative for Chronic Obstructive Lung Disease) guideline. According to this guideline, ACO is characterized by "persistent airflow limitation with several features usually associated with asthma and several features usually associated with COPD". Furthermore, many studies use the term asthma-COPD overlap syndrome (ACOS), but based on the recent change recommended by the GINA and GOLD in 2017, we will use 'ACO' as a term for this condition as it is not considered a single entity but a group of phenotypes [4]. Although asthma and COPD have been well defined, there is currently no consensus on the definition of ACO.
Most previous studies revealed that patients with ACO have more respiratory symptoms, more frequent exacerbations, poor quality of life, higher mortality rate, increased health care utilization, and higher prevalence of comorbidities than those with either asthma or COPD only [5][6][7][8][9].
Numerous population-based studies have been carried out to estimate the prevalence of ACO throughout the world, especially in the USA and Europe . However, there is a considerable variation among the studies. The prevalence of ACO has varied widely in these studies from 0.3 to 5.0% in the general population, from 3.2 to 51.4% in patients with asthma, and from 12.6 to 55.7% in patients with COPD. Although there are currently a limited number of population-based studies in this context, it has been growing in recent years.
The global prevalence of ACO remains unknown, and no systematic review and meta-analysis of populationbased studies has yet been conducted. On the other hand, due to the considerable heterogeneity among the reported prevalence of ACO, and its significant public health impact, the exact prevalence of ACO is critical for strategic plan and health policy. We therefore conducted a systematic review and meta-analysis of the published literature to examine this parameter. We examined the prevalence both in the general population and among patients with asthma or COPD to understand better the absolute burden of this condition.

Methods
This systematic review adheres to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist [37].

Search strategy
A systematic literature search was performed in ISI Web of Knowledge, MEDLINE/PubMed, and Scopus databases up to 30 May 2019 to identify studies reporting the prevalence of ACO. The search strategy in ISI Web of Knowledge is outlined in detail in Table 1. The other databases were searched with similar terms. Reference lists from identified studies and relevant review articles were also searched for studies eligible for inclusion.

Outcomes
Our primary outcome of interest was prevalence of ACO in the general population. Secondary outcomes included the prevalence of ACO in patients with asthma and patients with COPD, prevalence of asthma-only, and prevalence of COPD-only. Table S1 presents definitions of asthma, COPD, and ACO in included studies.

Inclusion and exclusion criteria
The following inclusion criteria were used to select studies: (1) studies reporting the prevalence of ACO, (2) population-based studies, (3) studies published in English language. No time restriction for publication dates was used. The following articles were excluded: (1) studies with no usable data, (2) repeated or overlapping studies, and (3) reviews, meta-analysis, and case reports articles.

Data extraction and quality assessment
Data were abstracted independently by two authors (SM and AAH), and discrepancies were resolved by consensus. The following study characteristics were extracted: first author's name, year of publication year, country, region, study name, study period, age, sample size, and prevalence of ACO (point or period prevalence). Study quality was assessed independently by two authors (SM and AAH) utilizing the Joanna Briggs Institute's critical appraisal checklist for studies reporting prevalence data [38]. This tool includes 9 items each of which is rated as either yes, no, not clear, or not applicable [38].

Statistical analysis
All meta-analyses were performed using the Stata version 13.0 (Stata Corp, College Station, TX, USA). The Cochran Q test (P < 0.10 was considered indicative of significant heterogeneity) and the I 2 statistic (values of 25, 50, and 75% were considered representing low, medium and high heterogeneity, respectively) were used to assess the heterogeneity between studies [39]. Due to the substantial heterogeneity between studies, pooled #2 ("Asthma-chronic obstructive pulmonary disease overlap syndrome") OR ("Asthma and chronic obstructive pulmonary disease overlap syndrome") OR ("asthma-COPD overlap syndrome") OR ("asthma-COPD overlap") OR ("asthma-COPD") OR ("ACOS") OR ("mixed asthma-COPD phenotype") OR ("Asthma combined with COPD") OR ("coexistence of asthma and COPD") OR ("COPD with asthmatic features") OR ("overlap of asthma-COPD") 1692 #3 ("Pulmonary Disease, Chronic Obstructive") OR ("Chronic Obstructive Pulmonary Disease") OR ("COPD") OR ("asthma") OR ("asthmatic") OR ("Asthmatic Patient") 237,190 #4 #1 AND #2 AND #3 258 prevalence of ACO was calculated using a randomeffects model. Potential sources of heterogeneity were explored through meta-regression and subgroup analyses with regard to the study region, year of publication, and study quality. Egger's test and a visual inspection of funnel plots were used to assess the presence of publication bias [40,41].

Study selection
A flow diagram of the study selection process is presented in Fig. 1. A total of 787 articles were retrieved from the different sources. After removal of duplicates, 402 records remained. By screening titles and abstracts, we identified 47 articles; after the inclusion and exclusion criteria were applied, 27 articles remained.

Study characteristics
The study characteristics of the included studies are summarized in  [19,22,36], three from the Latin America and 6 low-and middle-income countries and Australia [14,24,31]. The majority of the included studies were published after 2015. As presented in Table S2, the risk of bias was low in the majority of the criteria in the included studies.
Prevalence of ACO in the general population As seen in Fig. 2 Table 3). Furthermore, the shape of funnel plots did not reveal any evidence of obvious asymmetry (publication bias) (Fig. 3). Egger's test also did not reveal any evidence of publication bias (P = 0.576).

Subgroup analysis
Potential sources of heterogeneity were explored through subgroup analyses with regard to the study region, year of publication (before 2015 vs after 2015), and study quality (≤7 vs ≥8). According to our results, prevalence of ACO was not related to study region (P = 0.994), year of publication (P = 0.995), and study quality (P = 0.865) ( Table 4).

Discussion
ACO is a newly identified condition with apparently high prevalence, and evidently variable features, which is associated with a high burden of disease. Patients with ACO may have worse outcomes than those with either asthmaor COPD-only. Despite its high prevalence, few population-based studies have examined the prevalence of ACO, and as a result, the epidemiology of this condition remains poorly defined. Furthermore, until 2010 only one study has examined the prevalence of ACO in the population-based studies. Here we present a comprehensive overview of the prevalence of ACO in the general population as well as in patients with asthma-only and COPD-only. The overall prevalence of ACO in the general population was 2.0%. Heterogeneity was observed between included studies. This heterogeneity may be explained by differences in the definitions, diagnostic criteria, disease ascertainment methods, geographic region, population characteristics (e.g., age and smoking), study design, and inherent biases associated with observational studies. Subgroup analysis was performed to assess ACO prevalence in different geographic regions. Most of the studies originated from either North America or Europe, but others came from Latin America, China, Korea, and six low-and middle-income countries. When stratification was performed by geographic region, no differences in ACO prevalence were found. Data were lacking for regions of Africa and Middle East, and Eastern Europe, so we could not examine the prevalence of ACO in these areas. Because data were lacking for regions of Africa and Middle East, and Eastern Europe, ACO estimates in this meta-analysis may overestimate or underestimate  the global health burden. Thus, future population-based studies are required to estimate and compare the prevalence of ACO in the different geographic regions. Ideally, we would have investigated whether the year of study contributed significantly to the heterogeneity observed between included studies; however, the considerable overlap of the time periods prevented this analysis. Because of this limitation, we alternatively replace year of study with year of publication and found that ACO prevalence were not changed by year of publication. Of included studies, only one study reported temporal trend of ACO prevalence, which indicated increasing prevalence in time [26]; however, additional population-based studies are necessary to confirm this finding. This upward trend may be explained by increasing prevalence of both asthma and COPD, and improvement in monitoring of patients. To date, only three studies from Canada, Denmark, and Taiwan examined the incidence of ACO in the general population, emphasizing that such data are missing in the literature. In these studies, the incidence per 1000 person-years was as follows: Canada, 2.1 [26]; Denmark, 0.64 [25]; and Taiwan, 0.70 [42].
The prevalence of ACO among patients with COPD ranged from 12.6 to 55.7% with pooled prevalence of 29.6%. This finding is consistent with a recent metaanalysis, in which the ACO prevalence among patients with COPD in population-based studies was estimated to be 27.0% [8]. In our study, the prevalence of ACO among patients with asthma ranged from 3.2 to 51.4% with pooled prevalence of 26.5%. In previous studies, the prevalence of ACO among patients with asthma also has varied widely from 11.8 to 27.4% [43][44][45][46][47].
For included studies in this review, the global prevalence of asthma-only was 6.2% and COPD-only was 4.9%. In a recent meta-analysis, the global prevalence of COPD based on the GOLD definition was found to be 12.2% [48]. This difference may be due to the different diagnostic criteria and/or including high risk population in the analysis. According to Global Burden of Disease Study 2015, asthma and COPD were the most prevalent chronic respiratory disease, affecting 358 and 174 million Fig. 4 Forest plot showing prevalence of ACO in patients with asthma. Note. Squares represent study-specific prevalence estimates (size of the square denotes the study-specific statistical weight); Horizontal lines represent 95% confidence intervals (CIs); Diamond represents summary estimate of Prevalence with corresponding 95% CI.   [49]. As a large proportion of patients with asthma and COPD have ACO, as well as a worse condition of ACO than asthma-and COPD-only, it is crucial to increase the awareness of ACO and provide clear management and treatment to reduce the disease burden. This study has several limitations that should be taken into account when interpreting the results. First, there were no data on ACO prevalence for large regions such as Africa and Middle East, and Eastern Europe, thus the generalizability of the results may be limited. Further population-based studies, particularly in Asia and Africa, are required to better estimate the global prevalence and incidence of ACO. Second, regarding the issue that values of ACO prevalence depends on diagnostic methods, we could not perform subgroup analysis for this source of between-study heterogeneity. Third, the quality of the included studies was not optimal. Fourth, the number of studies included in the subgroup analyses was relatively small, so the prevalence of ACO in these subgroups may not be accurately represented. Fifth, although our search was comprehensive, we did not search some other database such as Embase, CINAHL and DOAJ, and include non-English language publications and non-published studies.

Conclusion
This study showed that 2.0% of the general population is affected by ACO. This finding indicates the considerable clinical impact and major burden posed by ACO. Given that there is no consensus on ACO definition in the literature, and the dependence of its prevalence on diagnostic method, it is important to provide clear diagnostic criteria, management and treatment for this condition.
Additional file 1: Table S1. Definitions of asthma, COPD, and ACO in included studies. Table S2. Quality assessment of included studies in the meta-analysis using Joanna Briggs Institute's .Critical appraisal checklist for studies reporting prevalence data.
Abbreviations ACO: Asthma-COPD overlap; ACOS: Asthma-COPD overlap syndrome; CI: Confidence Interval; COPD: Chronic obstructive pulmonary disease; GINA: Global Initiative for Asthma; GOLD: Global Initiative for Chronic Obstructive Lung Disease; PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses