Chronic obstructive pulmonary disease (COPD) remains one of the leading causes of disability and death in the developed world, and is emerging as increasingly important in the developing world. Despite its importance, COPD is not well recognised by the general public and frequently goes undiagnosed in people who have evidence of it. This underdiagnosis of people with evidence of obstruction on spirometry (generally adults with an FEV₁/FVC ratio <70%) has been previously documented in the United States¹ and Korea.²

The paper by Shahab and colleagues in this issue of Thorax shows that underdiagnosis and, in all likelihood, misdiagnosis, is a factor in England also.³ Their key finding was that 13.3% of the population aged 35 and older had evidence of COPD that would, in general, correspond to GOLD stage 1 or more severe disease.⁴ Bronchodilator response was not evaluated, so this would not meet strict GOLD criteria and, if this population is similar to the Norwegian adult population,⁵ one might expect the “post-bronchodilator” prevalence of COPD to be 20–25% lower. Whether this shifting classification is important either clinically or epidemiologically is unclear. While some research has suggested that bronchodilator responsiveness may help to distinguish asthma and COPD or improve the prediction of outcomes,^6,7 pre-bronchodilator classification of lung function using GOLD criteria has been shown to predict mortality and other adverse outcomes in several different populations.^8,9,10

A second important finding was that, among survey participants with evidence of COPD, only 18.8% had a current diagnosis of any lung disease.³ When looking at participants in the most severe category, correlating to GOLD stage 3 or more severe disease, only 46.8% had a diagnosis of lung disease and, of those with a diagnosis, that diagnosis was asthma in 47 of 74 (63.5%). The degree of underdiagnosis of disease in this population is very similar to that previously reported in the United States.¹¹ In the absence of a diagnosis, effective interventions are unlikely to occur. The question of a misdiagnosis by diagnosing COPD as asthma is complex. COPD and asthma, generally thought to represent different pathophysiological processes, share some important similarities in the adult population with regard to disease presentation and treatment,¹² and there may be overlap between COPD and asthma in a significant proportion of the population.¹ The question of whether treating COPD like asthma—with inhaled corticosteroids or leukotriene antagonists—provides benefits with regard to morbidity and mortality is central to several recently completed and ongoing investigations.^13–15

A third important finding in this study is the relation between current smoking and COPD. The authors reported that 34.9% of the population with COPD were current smokers compared with 22.4% of the population without COPD.³ They also reported that the prevalence of current smoking increased as COPD severity increased, from 29.5% in the mildest category to 40.5% in the most severe category. The link between cigarette smoking and COPD is well established and undeniable.^16–18 Similarly, smoking cessation is the most important intervention in COPD, with demonstrated effects on both morbidity and mortality.¹⁹ The authors suggest that smokers with COPD were more “cigarette dependent” than smokers without COPD.³ The 382 smokers with COPD did, in fact, smoke significantly more cigarettes per day than the 1596 smokers without COPD (16.3 v 14.8), had higher cotinine levels (322 v 278 ng/ml) and a higher Heavy Smoking Index dependency rating (3.9 v 3.6). When restricted to participants with COPD, however, there was no evidence of a dose-response relation for any of these measures as severity of COPD increased. In fact, among subjects in the highest severity category for COPD, both the cigarettes smoked per day and the dependency ratings were similar to those seen in smokers without COPD. Are smokers with COPD more cigarette dependent than smokers without COPD? While this remains a possibility, other explanations may be responsible for the results seen in this study. For example, the acute and subacute changes in lung function seen in active smokers may have been enough, in a population study, to classify “normal” subjects who are close to the 70% threshold as those with COPD.^20,21

Another aspect of the relation between smoking and COPD important in this study was the burden of disease among never and former smokers. Overall, 711 of 1093 (65.1%) of those classified as having COPD were not current smokers.³ Among people with clinically relevant COPD corresponding to GOLD stage 2 or higher, 390 of 638 (61.1%) were not current smokers. While some never or former smokers misclassified themselves (based on the cotinine levels), this finding points to the reality that COPD can develop and progress in people who have never smoked or in people who have stopped smoking. In this study, among ever smokers aged 65 and older, over 65% had stopped smoking. In all likelihood these people will suffer from limitations, morbidity, and mortality related to COPD. What factors contribute to the development and progression of COPD in the never smoker? This analysis points to some non-smoking factors previously shown to be risks for COPD such as social class and working in manual labour (as a surrogate for a dusty occupation).^22,23

In conclusion, this survey from England corroborates findings from other studies in other countries with different health delivery systems.^1,2 It also presents some interesting paradoxes: (1) COPD is a common disease among the adult population yet remains a hidden disease that is frequently undiagnosed, even when causing severe impairment; (2) when diagnosis of a “disease” in people with airflow limitation does occur, it is frequently asthma which may represent a misdiagnosis; and (3) smoking cessation remains the most effective intervention, yet the majority of disease in this population could be found in former and never smokers. This study, as those in other countries, shows that objective measures of pulmonary function need to be a more routine part of the assessment of the adult patient.