First author [ref.] | Source of research participants | Variables | Cluster method | Summary of clusters |
Burgel 8 | Respiratory clinic patients with COPD, excluding asthma | Age, BMI, dyspnoea score, exacerbation rate, FEV1 % pred, HAD scale, pack-yrs, SGRQ score | Principal components analysis on variables, Euclidean distance, Ward's method on derived variables | Young, severe disease, underweightYoung, moderate diseaseOld, mild disease, overweightOld, moderate disease, overweight |
Wardlaw 4 | Not stated, well-characterised patients with asthma or COPD | Age, bronchodilator reversibility, FEV1/FVC ratio, FEV1 % pred, pack-yrs, serum IgE, sex, % sputum eosinophils | Z-score distance measurement, Ward's method | Classical COPDAsthma/COPD overlapAsthma with high sputum eosinophils and high IgEAsthma with low sputum eosinophils and low IgE |
Weatherall 6 | Community-based random sample with FEV1/FVC <70% or a history of current wheeze | Bronchodilator reversibility, DL,CO/VA % pred, FeNO, FEV1/FVC ratio, FEV1 % pred, FRC % pred, pack-yrs, serum IgE, sputum production | Gower's distance, average distance method | EmphysemaAsthma/COPD overlapAtopic asthma with increased FeNOAirflow obstruction and sputum production without increased FeNOAirflow obstruction without other features |
FEV1: forced expiratory volume in 1 s; FVC: forced vital capacity; BMI: body mass index; % pred: % predicted; HAD: hospital anxiety and depression; SGRQ: St George's Respiratory Questionnaire; Ig: immunoglobulin; DL,CO: diffusing capacity of the lung for carbon monoxide; VA: alveolar volume; FeNO: fraction expired nitric oxide; FRC: functional residual capacity.