Self-reported smoking status and exhaled carbon monoxide: results from two population-based epidemiologic studies in...

To investigate the validity of self-reported responses to questions on current smoking in two cohorts based in Northern England.

A cross-sectional population-based study (the Newcastle Heart Project [NHP]) and a follow-up of the Newcastle Thousand Families birth cohort established in 1947.

Participants included 1,189 members of the NHP and 410 members of the Newcastle Thousand Families cohort who completed a health and lifestyle questionnaire, including questions on current smoking, and attended a clinical examination, including testing for exhaled carbon monoxide between April 1993 and December 1998.

The number of self-reporting smokers for whom very low (ie, < 6 ppm) exhaled carbon monoxide levels were recorded varied between 9% in the Newcastle Thousand Families Study and 26% among the members of the NHP who were of South Asian origin. Using a cutoff of 8 ppm, 80% of self-reported smokers were identified in both the Newcastle Thousand Families study and in the NHP European population, but only 60% were identified in the NHP South Asian population. In each population, < 7% of nonsmokers had exhaled carbon monoxide measurements of > 6 ppm, with nonsmoking men more likely to have higher levels than nonsmoking women. Among the nonsmokers, the levels of exhaled carbon monoxide did not vary with respect to the smoking status of a partner or socioeconomic status.

Using a cutoff value of 6 ppm would potentially miss a large number of smokers, although this may vary with ethnicity. Epidemiologic studies should continue to use biochemical markers to validate responses to smoking surveys. However, the use of exhaled carbon monoxide measurements as a method of assessing the validity of self-reported smoking status may require additional analyses of whether the cutoff level should vary for different populations.