Background: Objective validation of smoking status is necessary. Earlier studies have used saliva cotinine concentrations between 14.2 and 30 ng/ml as cut-off values to distinguish pregnant smokers from non-smokers. However, these cut-offs derive from studies including men and non-pregnant women. This constitutes a problem, as recent studies have reported an accelerated metabolism in pregnant smokers. The aim of this study was to determine the optimum cut-off cotinine level distinguishing pregnant smokers from pregnant non-smokers.
Methods: An observational study of 620 pregnant women, 359 self-reported smokers and 261 self-reported non-smokers, with complete data on smoking status and saliva cotinine. The study was conducted at a large university hospital in Copenhagen, Denmark. Saliva was sampled at the first antenatal visit to the midwife and analyzed for cotinine level by gas chromatography. Participants completed a questionnaire immediately after the first visit.
Results: A saliva cotinine cut-off level of 13 ng/ml, corresponding to a sensitivity of 0.99 and specificity 0.98, was found to be the optimum cut-off value separating pregnant smokers from non-smokers. The sum of the sensitivity and specificity was at its maximum, 1.981. A 95% bootstrap confidence interval for the optimum cut-off was (8-14 ng/ml).
Conclusion: From the present data we recommend that in future studies on smoking cessation and pregnancy a cut-off level of 13 ng/ml should be applied to distinguish between pregnant smokers and non-smokers.