Partner smoking and maternal cotinine during pregnancy: implications for negative control methods

Comparison of the associations of maternal and mother's partner smoking with offspring outcomes is, in theory, a useful method for assessing whether there may be an intrauterine effect of tobacco exposure on these outcomes. However, this approach assumes that the effects of passive smoking from exposure to partner smoking during pregnancy are minimal. We evaluated this assumption using a biochemical measure of tobacco exposure in pregnant women.

Cotinine levels taken during the first trimester of pregnancy were measured in a sample of 3928 women from the Avon Longitudinal Study of Parents and Children. Median cotinine values were compared across categories of smoking heaviness (cigarettes per day) of the women during the first trimester and in non-smoking women by the smoking heaviness of their partner.

Cotinine levels were substantially higher in women who smoked compared to non-smokers (range of medians across smoking heaviness categories: 900-5362 ng/ml versus 20 ng/ml, interquartile range (IQR) (0-63) for non-smokers). In contrast, cotinine levels in non-smoking women were only very weakly related to partner smoking status (range of medians in women with smoking partners: 34-69 ng/ml versus 12 ng/ml, IQR (0-48) in women with non-smoking partners).

Levels of tobacco exposure from partner smoking, as assessed by cotinine, were low in non-smoking pregnant women. This suggests that using mother's partner's smoking as a negative control for investigating intrauterine effects is valid.