Article Text


Diagnosing COPD in non-smokers: splitting not lumping
  1. Edwin R Chilvers,
  2. David A Lomas
  1. Respiratory Medicine Division, Department of Medicine, University of Cambridge School of Clinical Medicine, Addenbrooke's Hospital, CUHNHSFT and Papworth Hospital NHS Foundation Trust, Cambridge, UK
  1. Correspondence to Professor Edwin R Chilvers, Respiratory Medicine Division, Department of Medicine, University of Cambridge School of Clinical Medicine, Box 157, Level 5 Addenbrooke's Hospital, Hills Road, Cambridge CB2 0QQ, UK; erc24{at}

Statistics from

Chronic obstructive pulmonary disease (COPD) is the diagnostic label given to patients with chronic airflow obstruction that is poorly reversible.1 As such, it is a non-specific disease entity and, like stroke or atrial fibrillation, can be the end result of multiple and often very differing conditions. A diagnosis of COPD based on spirometry alone should therefore trigger further investigation to identify the underlying cause and the application, if possible, of a more appropriate and specific diagnostic label. Hence while sarcoidosis, chronic asthma, organ-specific autoimmune disease, cystic fibrosis and multiple other pathologies can all cause COPD (and in a technical sense allow all such patients to be labelled as having COPD) most would recognise this as diagnostic duplicity.

Recent studies have highlighted the apparently high proportion of patients with COPD who have never smoked. Collectively, these studies suggest that even in developed countries cigarette smoking causes COPD in only 50–70% of patients. Indeed, the Swedish OLIN and US NHANES III studies reported that the population-attributable risk of COPD from smoking in these countries was 45% and 44%, respectively.2 3 More recent studies in the UK suggest that 15% of adults above the age of 65 who have never smoked also have COPD.4 These data, however, do not gel with most UK physicians' practice, at least at a secondary care level. While there is little doubt that exposure to biomass smoke in developing countries and certain occupational insults in other settings can lead to COPD,5 6 clinical practice would suggest that the proportion of patients with genuine COPD in the UK who have never smoked is extremely small. Suggesting that COPD is common in never smokers living in developed countries (and hence by implication could be termed “idiopathic”) runs the significant risk of leaving such patients underdiagnosed and inappropriately treated.

So what has led to the view that obstructive lung disease is common in non-smokers? This matter has been reviewed recently by Salvi and Barnes7 who present mainly epidemiological studies, where a diagnosis of COPD has been made most commonly by questionnaire and spirometry alone. The problems associated with such studies are obvious and include poor case validation, failure to exclude patients with other common airway diseases such as asthma, bronchiectasis, tuberculosis and prior viral bronchiolitis, and insufficient scrutiny of patients' true exposure to tobacco smoke. These studies are further hampered by age-related changes in physiology; surely it is inappropriate to label an elderly patient with cough or breathlessness who has never smoked but who has mildly “obstructive” spirometry as having COPD? This matter was recently highlighted in a study by Hardie and colleagues who found that the Global Initiative for Obstructive Lung Disease (GOLD) guidelines for diagnosing and treating COPD falsely classify a substantial number of healthy, asymptomatic, never-smoking older adults as having COPD. Indeed, in their study, undertaken in residents of Bergen, Norway, >50% of such individuals aged over 80 years would have been classified as having COPD.8 This conclusion is supported by a number of other studies conducted in primary care that highlight the pitfalls of using the current recommended spirometric criteria for COPD in the elderly.9 Hence we must avoid using spirometric data in such an absolute way as, say, a diabetologist might when using a blood sugar level to define diabetes and avoid making a diagnosis of COPD in all those with chronic airflow obstruction. This view does not, of course, undermine the importance of certain events such as poor socioeconomic status, low birth weight, concurrent asthma, previous tuberculosis, etc. to operate as important sensitisers to the effects of cigarette smoke7; however this does not equate with independent causation. Moreover, there seems to be little evidence to support the view that outdoor air pollution causes COPD.

Hence, while accepting that different phenotypes exist under the diagnostic label of “smoking-related COPD”, this term would appear to be an entirely appropriate label for patients with a significant (eg, >5 pack year) smoking history, obstructive spirometry (postbronchodilator forced expiratory volume in 1 s (FEV1)/forced vital capacity (FVC) ratio <70%) and evidence of chronic bronchitis or emphysema; we would suggest that other causes of airflow obstruction must be sought in individuals in the UK who do not fulfil these criteria. This will avoid COPD becoming a “catch all” diagnostic label applied to multiple unrelated respiratory pathologies with clearly diverse aetiology. COPD would run the risk of becoming degraded into the “lung failure” equivalent of “heart failure” with little thought as to its true aetiology. The call to include non-smoking patients with “COPD” in trials designed to assess novel treatments7 is also likely to confound and confuse rather than add clarity. Cigarette smoking may be a less common cause of COPD in developing countries, but in the UK smoking should be considered a diagnostic prerequisite and its absence should trigger more detailed investigation. It is our view that the diagnostic process should start and not stop at the point of a spirometric diagnosis of airflow obstruction.


The work in the author's laboratories is supported by the MRC, Wellcome Trust, EPSRC, Asthma-UK, BBSRC, GlaxoSmithKline and NIHR Cambridge Biomedical Research Centre


View Abstract


  • Competing interests None.

  • Provenance and peer review Commissioned; not externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.