Background: The underreporting of environmental tobacco smoke (ETS) exposure by parents of study children may depend on the instrument used and population studied, underlining the need for questionnaire validation in specific study settings. This study explores the validity of parent-reported ETS exposure in a French multicenter study on asthma.
Methods: The study population was composed of 313 children ages 4 to 14 years. Exposure to ETS was evaluated both by questionnaires on recent ETS exposure and by assessment of urinary cotinine by an enzyme immunoassay.
Results: According to parents' reports, about one-third of children were exposed to ETS within the past 2 days before cotinine measurement, and on average 14.9 +/- 15.4 cigarette-equivalent were smoked in their homes. The mean urinary cotinine was 435 +/- 530 nmol/mol creatinine and increased with the reported number of cigarette-equivalents smoked at home but it did not differ between children registered as being exposed to 1-10 cigarettes and children registered as unexposed. Agreement between questionnaire and urinary cotinine was moderate to poor according to our correlation coefficient (0.22) and kappa coefficient (0.09).
Conclusion: These results show that our questionnaire is not discriminating enough to distinguish between nonexposure and mild exposure, but reveals gradients of higher exposure.