@ERSpublicationsAll exacerbation data, whether reported, self-managed or untreated, is required to recognise health burdens in COPD http://ow.ly/usGyB Chronic obstructive pulmonary disease (COPD) exacerbations are episodes of symptom worsening that have major impacts on a variety of health outcomes, including quality of life, hospital admission and death [1]. Exacerbation reduction is now a key outcome for pharmacological therapies in COPD and, thus, if exacerbation can be prevented how does this affect disease progression and lung function decline?The issue of the importance of lung function decline in COPD first arose when FLETCHER and PETO [2] explored lung function in a cohort of 792 male British postal workers. These subjects were studied over 8 years and it was found that exacerbations were associated with the ratio of forced expiratory volume in 1 s (FEV1) to the cube of the height, although the exacerbation definition was limited at that time. This work was followed in close succession by a study from Utah in which KANNER et al. [3] showed that exacerbation frequency was related to COPD severity, thus implicating exacerbation frequency as a factor in FEV1 decline. However, detecting the relationship proved elusive until the Lung Health Study showed that in smokers and intermittent quitters each additional exacerbation reported to physicians caused a greater annual decline of 7 mL [4]. The following year, the London COPD cohort study reported that in patients with a history of more frequent exacerbations there was an increased rate of decline of 8 mL per year compared with infrequent exacerbators [5]. The relationship between exacerbation frequency and FEV1 decline in COPD has also been seen in the UPLIFT study [6] and another study from Greece [7]. All of these studies were performed on European or North American subjects with COPD.In diseases other than COPD, an occasional relationship has been reported between exacerbation rates and FEV1 decline. In a study of asthmatics over a period of 11 years, a relationship was seen between exacerbation frequency and FEV1 decline [8]. In a Canadian study, WATERS et al. [9] showed that severe cystic fibrosis exacerbations (requiring hospitalisation) were associated with greater FEV1 decline. Although the European Cystic Fibrosis Society study did not report on exacerbations they have reported a link between chronic Pseudomonas aeruginosa infection (a known risk factor for cystic fibrosis exacerbations) and FEV1 decline [10], from which such a relationship may be inferred and this supports the findings of the Canadian study.In this issue of the European Respiratory Journal, the relationship between exacerbations and lung function decline is addressed in a cohort of 279 Japanese subjects followed over a 5-year period in Hokkaido [11]. The detection rate for exacerbations in this study was 0.78 events per year using a symptom-based definition. The study found no significant relation between exacerbation rate and FEV1 decline. The annual decline in FEV1 in the Hokkaido...