Is forced expiratory volume in six seconds a valid alternative to forced vital capacity?

We have read with interest the article by HANSEN et al. [1] in a recent issue of the European Respiratory Journal, stating that the use of forced expiratory volume in six seconds (FEV6) reduces the sensitivity of obstructive lung disease detection. Their study population of 3,515 current smokers, with a 17.5% prevalence of obstruction, found a sensitivity of only 76.8% and a specificity of 98.2% when comparing FEV1/FEV6 with FEV1/forced vital capacity (FVC) below their respective 95% lower limits of normal (LLN). Based on this low sensitivity, HANSEN et al. [1] concluded that FEV1/FEV6 is less reliable than FEV1/FVC as a screening parameter for the detection of obstruction. The lower sensitivity obtained by HANSEN et al. [1] with respect to previous studies, in which sensitivity typically ranged between 92–95% [2–5], was attributed to differences in study population. HANSEN et al. [1] contended that previous studies comparing the performance of FEV6 and FVC in detecting obstruction [3–5] had too few mildly obstructive and normal subjects. However, HANSEN et al. [1] cite another study, by AKPINAR-ELCI et al. [2], which showed 92% sensitivity and 98% specificity for the detection of obstruction; this is despite a prevalence of obstruction of only 14.5% in 1,139 nonpatients in a workplace setting. In general, these participants are healthy, and obviously median spirometry values are closer to the LLN than for groups of symptomatic patients referred to a pulmonary function laboratory. In the study by HANSEN et al. [1], the ''minimal obstructed'' group represents the largest group (59%) of all subjects considered obstructive. Minimal obstruction was defined as FEV1/FVC and/or FEV3/FVC below the 95% LLN, combined with FEV1 between 80–120% predicted. However, the rationale for using this criterion of possible morbidity only holds in populations with respiratory diseases; for healthy subjects, the meaning of a low FEV1/FVC with a normal FEV1 is unclear [6]. The proportion of subjects with respiratory diseases studied by HANSEN et al. [1] is not explicitly mentioned, but is likely to be small. Therefore, a substantial proportion of subjects labelled with ''minimal obstruction'' could be normal. Figure 3 in the paper by HANSEN et al. [1] clearly shows that the number of false negatives in the ''minimal obstructed'' group represents more than half of all false negatives; this could introduce a serious bias to the results. In previous studies by us and others [2–5], most of the discordant cases had FEV1/FVC and FEV1/FEV6 …