This study was based on analyses of the 2007–2012 National Health and Nutrition Examination Survey (NHANES) data, which were the years that spirometry testing was conducted on the survey. NHANES is a cross-sectional survey of the civilian, non-institutionalized U.S. population conducted by the National Center for Health Statistics (NHANES - National Health and Nutrition Examination Survey Homepage). Data were collected via household interviews and standardized physical examinations in specially equipped mobile examination centers (MEC). The NHANES samples are selected through a complex, multistage, probability design. The 2007–2012 NHANES cycles oversampled major U.S. demographic subgroups including Hispanic and Black persons, low income white persons, and persons aged 80 years and older. Additionally, Asian persons were oversampled in 2011–2012. The procedures to select the sample and conduct the interview and examination have been specified elsewhere (Zipf et al.). Informed consent was obtained from all participants and the NCHS Research Ethics Review Board approved the protocol (NHANES - NCHS Research Ethics Review Board Approval).

Analyses were conducted on data collected from adults aged 40 to 79 years old. The lower boundary of the analyzed age cohort (40 years) was based on guidelines for diagnosis which recommend considering spirometry in symptomatic patients over the age of 40 years (GOLD 2017 Global Strategy for the Diagnosis, Management and Prevention of COPD). The upper boundary of the analyzed age cohort (79 years) was selected as this was the maximum eligible age for the NHANES spirometry examination.

During 2007–2012, 14,825 persons aged 40–79 years were eligible for the survey, 10,531 (71%) were interviewed and 10,219 (69%) were examined in the NHANES mobile examination center (MEC).

During the in-home interview participants were asked detailed questions on specific medical conditions. A participant was classified as having healthcare provider-diagnosed COPD if they responded affirmatively to either of the following questions on emphysema or chronic bronchitis: “Has a doctor ever told you that you have emphysema?” or “Do you still have chronic bronchitis?” (Mannino, Gan, et al.; Halldin et al.).

Demographic data were collected during the household interview. Age was categorized as 40–59 and 60–79 years. Self-reported race and Hispanic origin were categorized as non-Hispanic white, non-Hispanic black, and Hispanic. Due to small sample size and heterogeneity of group, participants who reported other race or Hispanic groups (including multiple races) were not reported separately but were included in total estimates. Self-reported education was categorized as less than 12^th grade, high school graduate or equivalent, and some college to include anyone who had attended college for any length of time.

General health status was categorized as very good or excellent, good, and fair or poor. The number of times in the past year a person reported seeing a doctor or healthcare professional was categorized as collected: none, 1, 2 or 3, 4 to 9, and 10 or more with the exception of the two highest categories (10–12 and 13 or more) which were combined for a more balanced data distribution. Presence of co-morbid conditions (coronary heart disease, stroke, cancer/malignancy, diabetes and hypertension) were identified based on affirmative responses to being told by a doctor or other health professional that he/she had the particular condition. Number of co-morbid medical conditions was derived based on the presence of these five conditions.

A current smoker was a respondent who reported smoking at least 100 cigarettes in their lifetime and at the time of the interview reported smoking every day or some days OR by a serum cotinine value greater than 10ng/mL (Hukkanen et al.). A former smoker was a respondent who reported smoking at least 100 cigarettes during their lifetime but currently did not smoke. A never smoker was a participant who reported smoking fewer than 100 cigarettes during their lifetime.

Presence of chronic cough was defined as affirmative responses to “Do you usually cough on most days for 3 consecutive months or more during the year?” and the follow-up question that the coughing had been for at least 2 years. Presence of chronic phlegm was defined as affirmative responses to “Do you bring up phlegm on most days for 3 consecutive months or more during the year?” and the follow-up question that bringing up of phlegm had been for at least 2 years. Presence of wheezing was defined by the question, “In the past 12 months, how many attacks of wheezing or whistling have you had?” Number of attacks was categorized as 0, 1–3, and 4 or more. Presence of dyspnea was based on an affirmative response to “Have you had shortness of breath either when hurrying on the level or walking up a slight hill?” Number of respiratory symptoms was derived based on the presence of these four symptoms.

Participants were defined as using a respiratory medication if they reported past 30-day use of any of the following drug classes based on the Global Initiative for Chronic Obstructive Lung Disease (GOLD) treatment guidelines: long-acting bronchodilator, short-acting bronchodilator, inhaled corticosteroid, systemic corticosteroid, methylxanthine, and leukotriene modifier.

NHANES spirometry data were used to identify persons with lung obstruction and assess severity levels. Spirometry testing was performed on NHANES participants aged 6–79 years during 2007–2012. It was conducted in accordance with recommendations of the American Thoracic Society (Miller et al.) using Ohio 822/827 dry-rolling seal volume spirometers. The full exam protocol is detailed elsewhere (NHANES - Respiratory Health - Bronchodilator Procedures Manual). Using pre-bronchodilator spirometry values, participants were defined as having evidence of COPD when the ratio of the forced expiratory volume in one second (FEV1) to the forced vital capacity (FVC) is less than the lower limit of normal (LLN) representing the lower 5^th percentile based on person’s age, sex, height, and race/ethnicity (Pellegrino et al.). LLN values were determined using normative reference equations developed from NHANES III data by Hankinson et al. (Hankinson et al.).

Adults showing no spirometric evidence of COPD were categorized as having no measured lung obstruction. Those showing spirometric evidence of COPD were further categorized as having mild (FEV1>70% predicted) or moderate or worse (FEV1≤70% predicted) lung obstruction (Pellegrino et al.). Percent of predicted FEV1 was defined as the observed FEV1 value divided by the predicted FEV1 value estimated for a person of the same age, gender, race/ethnicity, and height using race-specific reference equations (Hankinson et al.). Reference values are available for non-Hispanic white, non-Hispanic black, and Mexican Americans. For the Hispanic group, the reference values provided for Mexican Americans were used.

Prior to the spirometry examination, participants were asked if they currently had a breathing problem that required the use of supplemental oxygen during the daytime. Persons who reported a healthcare provider diagnosis of COPD and were excluded from the spirometry exam due to daily supplemental oxygen use, which is suggestive of advanced disease, were assigned a lung obstruction severity of moderate or worse (n=84).

Statistical analyses were performed using STATA™ version 13.1 (StataCorp, College Station, TX). Examination sample weights were used to account for differential probabilities of selection, nonresponse, and non-coverage. Taylor series linearization was used to calculate standard errors to account for the complex sampling design. Differences in the prevalence of diagnosed COPD between subgroups was evaluated using T-tests for categorical variables and linear trend tests using linear regression and orthogonal contrasts for continuous variables. Chi-square tests were performed to examine the association between covariates and the presence of a COPD diagnosis. Multivariable logistic regression was used to assess predictors of diagnosis along with their associated odds ratios and statistical significance. All reported estimates have a relative standard error ≤30% (NHANES - Analytic Guidelines, 1999–2010).

The percentage of missing data at the individual variable level ranged from 0% (multiple variables) to approximately 20% for the spirometry measures with the overall fraction of missing information for the aggregated NHANES data being roughly 4%. To assess the impact of the missing data, sensitivity analyses were conducted on the proportions with diagnosed COPD and the characteristics of those with and without diagnosed COPD using 40 multiply imputed datasets. As estimates were similar between those based on the unmodified (complete case) data and those based on the multiply imputed data, all estimates presented here are based on the unmodified data for ease of reproducibility.

During 2007–2012, 5.2% of adults aged 40–79 years reported being diagnosed with COPD, which includes ever emphysema and current chronic bronchitis. Prevalence was similar by sex, higher among non-Hispanic white (5.9%) and non-Hispanic black (4.6%) adults than Hispanic (2.0%) adults, and higher among adults aged 60–79 (8.3%) than 40–59 (3.6%) years (Table 1). Prevalence of diagnosed COPD significantly decreased with increasing education going from 9.2% among adults with less than a high school education to 3.7% among those with any college education. Prevalence was 14.3% among adults who reported fair or poor health status compared to 5.0% who reported good health and 1.7% who reported very good/excellent health. Prevalence of diagnosed COPD increased as the number of health care visits in the past year increased going from 1.8% among those with 1 visit to 12.6% among those with 10 or more visits.

Prevalence of COPD diagnosis was greater among adults who were current smokers (10.2%) than past (6.6%) or never (2.1%) smokers. A third (33.7%) of adults with moderate or worse lung obstruction had been diagnosed with COPD compared to 4.1% of those with mild lung obstruction and 3.1% with no lung obstruction. As the number of respiratory symptoms increased the prevalence of diagnosed COPD also increased. Specifically, prevalence was less than 1% (0.9%) when no respiratory symptoms were reported, 5.8% with 1 symptom, 14.9% with 2, and 36.7% with 3 or more symptoms.

Table 2 provides the characteristics of those with diagnosed COPD and those without diagnosed COPD, which includes those with undiagnosed COPD as well as those without the condition. Adults aged 40–79 years with diagnosed COPD were significantly different from those without diagnosed COPD in all sociodemographic and health characteristics examined, except gender (Table 2). Specifically, among those with diagnosed COPD, a higher proportion were non-Hispanic white (81.2% vs 71.4%), aged 60–79 (54.5% vs. 33.4%), and had less higher education (40.9% vs 58.4%).

Over one half (50.8%) of adults with diagnosed COPD reported fair or poor health while only 13.8% reported very good or excellent health compared to 17.3% reporting fair/poor health and 44.2% reporting very good or excellent health among those without diagnosed COPD. Slightly more than three-quarters of adults (75.4%) with diagnosed COPD reported one or more of the following co-morbid conditions: hypertension, diabetes, cancer, heart disease, or stroke compared to 50.3% reporting one or more of these conditions for those without diagnosed COPD. 72.6% of adults with COPD reported having 4 or more healthcare visits in the prior year compared to only 40.7% of adults without COPD.

Nearly half (47.1%) of adults with diagnosed COPD currently smoked cigarettes and 33.4% were former smokers. These proportions were significantly different from adults without diagnosed COPD where more than one-half (51.3%) were never cigarette smokers and 22.7% currently smoked.

Among adults with diagnosed COPD, 39.5% reported having been diagnosed with emphysema alone, 45.3% reported current chronic bronchitis alone and 15.2% reported both diseases. Nearly half (49.1%) reported using a prescribed respiratory drug. Of those with diagnosed COPD, 55.7% showed no spirometric evidence of lung obstruction while 7.9% presented with mild lung obstruction and 36.4% had moderate or worse lung obstruction. Among adults without diagnosed COPD, 87% had no measured lung obstruction. However, 9.3% had mild lung obstruction and 3.6% had moderate or worse lung obstruction.

Respiratory symptoms were prevalent among adults with diagnosed COPD with 76.5% reporting dyspnea or shortness of breath, 41.4% having chronic cough, 30.9% chronic phlegm, and 59.3% at least one wheezing episode in the past year. Nearly 90% (89.4%) reported having at least one of these four respiratory symptoms compared to 35.9% among those without diagnosed COPD.

Table 3 details the adjusted odds ratios, confidence intervals, and associated p-values from multivariable logistic regression modeling used to examine the association of various covariates with a COPD diagnosis outcome. After controlling for all the other variables in the model, non-Hispanic blacks were 49% less likely (AOR=0.51, 95% CI = 0.35 to 0.74) and Hispanics were 68% less likely (AOR=0.32, 95% CI = 0.19 to 0.55) than non-Hispanic whites to be diagnosed with COPD. Persons aged 60–79 years were 65% more likely (AOR=1.65, 95% CI = 1.07 to 2.53) than persons aged 40–59 years to be diagnosed with COPD. Those reporting fair or poor health were nearly three times as likely to be diagnosed with COPD (AOR=2.91, 95% CI = 1.55 to 5.46) than those reporting being in very good or excellent health. Persons reporting 4 to 9 healthcare visits in the past year were 3.49 times as likely (AOR=3.49, 95% CI = 1.30 to 9.39) to be diagnosed with COPD than those with no healthcare visits in the prior year and those with 10 or more healthcare visits in the past year were over 5 times as likely (AOR=5.06, 95% CI = 1.62 to 15.77) to be diagnosed than those with no healthcare visits in the previous year.

Former smokers and current smokers were 75% and 70% more likely (AOR=1.75, 95% CI = 1.23 to 2.48 and AOR=1.70, 95% CI = 1.17 to 2.48, respectively) than never smokers to be diagnosed with COPD.

In terms of lung obstruction severity, adults with moderate or worse lung obstruction were nearly five times as likely (AOR=4.90, 95% CI = 3.28 to 7.32) than those with no measured lung obstruction to be diagnosed with COPD. Having mild disease was not significantly associated with being diagnosed with COPD.

Finally, the likelihood of being diagnosed with COPD significantly increased as the number of reported respiratory symptoms increased. Specifically, adults reporting 1 respiratory symptom were nearly 4 times as likely (AOR=3.91, 95% CI = 2.11 to 7.24), those reporting 2 respiratory symptoms were nearly 10 times as likely (AOR=9.57, 95% CI = 5.43 to 16.88) and those reporting 3 or more respiratory symptoms were about 22 times as likely (AOR=22.07, 95% CI = 12.03 to 40.49) than those with no respiratory symptoms to be diagnosed with COPD.