Abstract
Background : A number of questions have been used to assess chronic cough in epidemiological surveys of children with and without wheeze. The comparative validity of these questions is unknown. In a population-based cohort study, we determined the association of different cough questions with current wheeze.
Methods : In a cohort study of unselected children, we assessed prevalence of respiratory symptoms repeatedly by questionnaires, including various questions on chronic cough. Using data collected in 2010, at age 13-17 years, we determined the strength of association of different cough questions with current wheeze using a logistic regression.
Results : Of 2542 participants, 359 reported current wheeze, 1621 had never wheezed and 784 had wheezed earlier but not in the last 12 months. Adjusting for age, sex and ethnicity, the following cough questions were most strongly associated with current wheeze (Odds ratio, OR>5): child coughs more than others (OR 12.1 95% CI [8.8-16.7]), cough lasts >2months (6.04; [3.2; 11.6]), phlegm at any time of day or night (6.5 ; 5.0; 8.4); frequent colds (23.1 [9.2; 57.8]), extended duration of colds (4.36 [1.2; 15.7]), cough triggered by exercise (8.0; [6.1; 10.3]) and by pets (7.7 [4.7; 12.7]). Less strongly associated were: night cough (ISAAC question; 4.9; [3.8; 6.4], cough apart from colds (3.4 [2.6; 4.4]).
Conclusion : Questions on chronic cough commonly used in respiratory questionnaires vary greatly in their association with current wheeze. This demonstrates the need to determine the comparative validity and usefulness of these questions in epidemiological surveys.
Funding :
SNF PDFMP3 137033; SNF 32003B-144068; Asthma UK 07/048.
- © 2013 ERS
