Asthma and chronic obstructive pulmonary disease (COPD) are common respiratory health conditions. Recent estimates in the United States indicate that in 2010, 11% of children 5–14 years of age, 9% of individuals 15–34 years of age, and 8% of adults aged 35 and older experienced current asthma [1]. The same year, nearly 6% of adults aged 25 years and older reported physician-diagnosed COPD, with the percentages increasing to 10% among adults 65 years and older [2]. The co-occurrence of asthma and COPD diagnoses, often referred to as ‘overlap syndrome,’ is well described, especially at older ages [3–8].

Despite the substantial overlap, the extent to which the two diagnoses are manifestations of the same disease or distinct entities remains unresolved [4, 6, 7, 9–15]. Asthma affects individuals at younger ages, whereas COPD occurs at older ages [1, 2]. Asthma is associated with eosinophilic inflammation, whereas COPD is associated with neutrophilic inflammation [15]. Common risk factors for both asthma and COPD include outdoor air pollution and occupational exposures [16–19], though cigarette smoking continues to be the most well-described, modifiable risk factor for COPD [20]. A history of asthma itself has been associated with the increased risk of COPD [21] and accelerated lung function decline [22, 23]. In a 20-year follow-up study by Silva et al. [21], individuals with active asthma were found to be at increased risk of developing symptoms of chronic bronchitis, being diagnosed with emphysema, and fulfilling criteria consistent with COPD than were individuals with inactive asthma or no asthma history, even after adjustment for age, sex, and smoking history. The fact that individuals with active asthma are at increased risk of developing COPD relative to individuals with inactive asthma suggests that asthma control may affect the occurrence of COPD.

Improving our understanding of the burden of COPD among adults diagnosed with asthma may provide insight into the role of asthma control in the observed overlap between asthma and COPD. We conducted these analyses to examine the prevalence of physician-diagnosed COPD among adults with a history of asthma using data collected routinely for surveillance of asthma in the United States. We report age-specific prevalences of COPD among adult respondents in the Behavioral Risk Factor Surveillance System (BRFSS) Asthma Call-back Survey with and without active asthma and across categories of cigarette smoking history.

Table 1 shows characteristics of the 71,639 Asthma Call-back Survey respondents, estimated to represent nearly 26.4 million adults with a history of asthma, including active and inactive asthma. In this population, an estimated 72% were identified as having active asthma and an estimated 29% were identified as having physician-diagnosed COPD. Among those with active asthma, the estimated prevalence of physician-diagnosed COPD was 34.6% (95% CI: 33.7, 35.6); among the remaining 28% with inactive asthma, the estimated prevalence of physician-diagnosed COPD was 14.7% (95% CI: 13.4, 15.9). Among respondents with active and inactive asthma, unadjusted percentage of respondents with physician-diagnosed COPD increased across categories of age; age-specific percentages were consistently highest among current smokers and lowest among lifetime nonsmokers (Fig. 2).

Adjusted associations between respondent characteristics, including active compared to inactive asthma status, and the prevalence of self-reported, physician-diagnosed COPD are shown in Table 2. Adjusted for age, educational attainment, race/ethnicity, and sex, active asthma was associated with a 26.7% (95% CI: 22.5, 30.9) higher prevalence of physician-diagnosed COPD among current smokers, a 20.6% (95% CI: 18.0, 23.3) higher prevalence among former smokers, and an 8.3% (95% CI: 6.1, 10.5) higher prevalence among lifetime nonsmokers. In this adjusted model, the prevalence of physician-diagnosed COPD was 4.6% (95% CI: 3.0, 6.2) higher among female respondents than among male respondents and PDs decreased monotonically with increasing educational attainment (Table 2).

When we restricted our categorization of active asthma to include only those respondents who reported experiencing any symptoms of asthma during the past 12 months, we excluded 5,769 respondents who were previously categorized as having active asthma, but who only reported talking to a doctor or other health care professional about his/her asthma or taking asthma medication during the past 12 months. Following this restriction, relative to inactive asthma, active asthma was associated with a 27.6% (95% CI: 23.3, 31.9) higher prevalence of physician-diagnosed asthma among current smokers, a 21.9% (95% CI: 19.2, 24.7) higher prevalence among former smokers, and an 8.8% (95% CI: 6.7, 11.0) higher prevalence among lifetime nonsmokers (not shown).

We analyzed data from the 2006–2010 Asthma Call-back Surveys and found higher prevalences of physician-diagnosed COPD among adults with active asthma than among adults with inactive asthma. These higher prevalences were observed among current and former smokers as well as lifetime nonsmokers, suggesting an important role for active asthma in the epidemiology of COPD in each category of smoking status.

Our findings complement recent estimates of COPD prevalence (e.g., age-adjusted prevalence: 6.5%; unadjusted prevalence: 6.8%) [2] and extend evidence from prospective, observational studies suggesting that knowledge of asthma history may provide information about the risk of COPD. Specifically, greater annual declines in FEV₁ have been observed among individuals with self-reported asthma than among those without self-reported asthma, including among both smokers and nonsmokers [22]; greater age-related declines in FEV₁ have been reported among participants with asthma and among cigarette smokers [23]; and elevated risks of COPD have been reported among adults with active asthma and among current smokers [21]. Our findings build on this evidence by using public health surveillance data that are collected annually to examine whether the relationship between asthma and physician-diagnosed COPD can be observed in a large, cross-sectional population-based sample. The use of public health surveillance data has been identified as a priority for advancing the prevention and control of COPD [35]. Analysis of such data sources may provide useful information about modifiable factors (e.g., level of asthma control and smoking status), non-modifiable factors (e.g., age and sex), and other characteristics (e.g., educational attainment) associated with the prevalence of physician-diagnosed COPD.

In these analyses, we identified respondents as having active asthma based on positive responses about having talked to a doctor or other health professional about [his/her] asthma (estimated prevalence: 51%), taken asthma medication (estimated prevalence: 58%), and experienced any symptoms of asthma (estimated prevalence: 64%) during the past 12 months. When we restricted our categorization of active asthma to include only those respondents who reported experiencing any symptoms of asthma during the past 12 months, the associations between active asthma and the prevalence of COPD were approximately 1% higher than those observed in our main analysis, where the definition of active asthma included talking to a doctor or other health care professional about asthma or taking asthma medication during the past 12 months. The small change following this restriction suggests that the associations observed in all three categories of smoking status are only modestly attenuated by the outcomes of discussions with doctors or other health care providers or asthma medication use during the past 12 months. In these data, just 4% were categorized as having active asthma by only having talked to a doctor or other health professional about asthma during the past 12 months and just 1% were categorized by only having taken asthma medication during the past 12 months. Using these public health surveillance data, we were unable to identify whether symptoms reported during the past 12 months improved following discussions with doctors or health professionals or use of medications for asthma, among individuals who reported these factors. Additional analyses conducted among populations responding to similar questions at intervals longer than 12 months apart may provide useful data about the contributing roles of health care visits and asthma medications in defining active asthma. Similarly, additional analyses conducted in the Asthma Call-back Survey or other surveillance data may further improve our understanding of how changes in medications and treatment practices, advances in patient education and self-management methods, and trends in second-hand smoke exposures and cigarette smoking may affect the relationship between asthma and COPD.

We analyzed cross-sectional data and do not have in-depth information about factors that may be involved in the development of asthma, COPD, or both. Indeed, because we analyzed surveillance data, we cannot draw conclusions about or contribute to the discussion of whether there exists a causal relationship between asthma and COPD. A major limitation of these data is the lack of information about the extent to which asthma or COPD may have been incorrectly diagnosed or reported; neither the BRFSS nor the Asthma Call-back Survey include means by which the responses may be validated. A comparison of self-reported asthma to asthma classification based on physiologic measures during a clinical exam 10 years earlier has shown that the validity of self-reported asthma varies with asthma severity [36, 37]. In the absence of clinical measures of asthma, participants in survey research are often asked whether they have been told by a doctor, nurse, or other health care provider that they have asthma. Data collected via self-administered questionnaire can be a reliable source of information about asthma [38]; the extent to which validity and reliability of asthma history information collected via interviewer-administered telephone survey are different from validity and reliability of the same information collected via self-administered questionnaire is unknown. Comparable data evaluating the validity and reliability of physician-reported COPD or the magnitude of recall bias affecting self-report of physician-diagnosed COPD are also unavailable. We reiterate that in these data, self-reported, physician-diagnosed COPD status is considered a proxy for actual disease status. Similarly, self-reported smoking status is considered a proxy for actual smoking status and additional description of current and former smoking habits, such as pack-years of cigarettes smoked, is not available in these data. Information about self-reported smoking status cannot be used to make inferences about the extent to which the prevalence of physician-diagnosed COPD varies within groups, such as among occasional smokers, smokeless-tobacco users, individuals exposed to secondhand tobacco smoke, and those who quit smoking recently versus those who quit long ago, or across categories of pack-years of smoking history.

The large size of the pooled annual survey samples allowed us to evaluate the prevalence of physician-diagnosed COPD within strata of smoking status. We examined an alternate statistical method to adjust sampling weights; revising our analyses with alternate weights, computed by multiplying the proportion of respondents in each survey year by the corresponding survey year’s weight, as in previous analyses of Asthma Call-back Survey data [39, 40], did not markedly affect our results or conclusions (data not shown). Pooling data collected over five years provided a sufficiently large sample to enable us to evaluate the prevalence of physician-diagnosed COPD in nonsmoking and young adults, where the prevalences of COPD are relatively low, and in the sparsely populated category of current smokers with inactive asthma. Improving our understanding of the influence of active manifestations of asthma, including recent asthma symptoms, on long-term respiratory health may improve our growing understanding of the epidemiology of COPD. Indeed, cigarette smoking continues to be the most well-described, modifiable risk factor for COPD [20], and smoking prevention efforts continue to be identified as the highest priorities to reduce the onset of new COPD and exacerbation of existing COPD [41]. These findings highlight the importance of appropriate management of asthma among smokers and nonsmokers alike.