Is it useful to distinguish between asthma and chronic obstructive pulmonary disease in respiratory epidemiology.

Disagreement exists on how to define asthma in population studies. Features characteristic of asthma also have been demonstrated in nonasthmatic subjects. For example, although in a random sample of 339 subjects investigated in the Netherlands, subjects with a history of asthmatic attacks, wheeze and/or dyspnea, or chronic dry cough on average were the most responsive (mean PD20 values for histamine of 11.5, 17.0, and 15.0 mg/ml) and asymptomatic subjects were the least responsive (mean PD20, 45.4 mg/ml), the distributions of PD20 values in all groups overlapped considerably. The best way to investigate the association between exposure and disease is in a cohort study. It is argued that follow-up studies in patients may have limited meaning because of a high risk for selection bias. Furthermore, in general, diagnostic disease entities will not represent homogeneous exposure groups. It is concluded that end points of disease should be defined as objectively as possible, for instance, as the level of FEV1, or as simple respiratory symptoms such as dyspnea, wheeze, or chronic cough. Terms such as "asthma," "COPD," or "CNSLD" will not be helpful in defining end points because there is a risk that they could introduce bias, misclassification, and hence confusion.