Why was the cohort set up?Chronic obstructive pulmonary disease (COPD) affects 5-10% of people worldwide (1), is rising in prevalence (2) and is the third most common cause of death (3). The annual burden of COPD regarding healthcare (mainly exacerbations resulting in emergency admissions) and societal (predominantly lost productivity) costs was estimated to be around $49.9billion in the USA (2010 prices (4)) and €48.4billion in the EU (2011 prices (5)). A substantial proportion of those with COPD are of working age, but there is some evidence that they have poorer employment history (6), higher rate of sickness absence (7) and poorer work performance (presenteeism) (8) compared to the general population.There remains much uncertainty about the natural history of COPD (9, 10) and which interventions are effective in altering the course of early disease. Furthermore, up to 85% of cases (11-13) are undiagnosed; representing many with potentially unmet need. Partly in response to reports (14)(15)(16) highlighting the burden of COPD, extent of underdiagnosis and uncertainty about prognosis of early disease, expert reviews have highlighted a need for further longitudinal data (9, 10). However, established cohorts usually represent secondary care patients with more advanced disease, with short duration of follow up and, generally small samples (17)(18)(19). While large population cohorts have sometimes addressed questions relevant to COPD (20-28), limitations in outcome measures and quality of lung function testing provide insufficient data to inform the COPD arena. Importantly, there are no primary care COPD cohorts with case-found patients and few with patients representing the full range of disease severity, particularly those with mild to moderate disease, and diverse socioeconomic mix.In recent years, several studies have also focused on patients reporting respiratory symptoms but who have normal lung function (former GOLD severity stage 0 (29)). The evidence on progression to COPD is limited and contradictory (23, 30, 31) and methods for assessing symptoms are inconsistent (23, 32). Thus there is also a paucity of evidence on the clinical relevance and natural history for this patient group.Better understanding of natural history and prognostic factors is needed to facilitate consultations, and to inform management decisions and health service planning. Existing COPD prognostic indices (PI) mainly focus on predicting mortality risk (17,(33)(34)(35)(36), though others were developed to predict additional outcomes such as exacerbations (37, 38), COPD-related hospitalisation (39), respiratory hospital attendance/admission (40), exacerbation or hospitalisation (41, 42). Only three indices (38,41,42) were derived in primary care populations despite this being where most COPD patients are managed, and most included patients with more severe established disease. No indices were developed in populations that included case-found patients. The methods and basis for selecting prognostic variables are rarely described, and...