Meaningful differences in maternal smoking behaviour during pregnancy: Implications for infant behavioural vulnerability

Background: Smoking during pregnancy has been consistently associated with risk of problem behaviour in offspring. There is debate about whether this association reflects a teratological effect or is a marker for problematic maternal characteristics. We test these "competing" hypotheses by examining whether (1) exposure is associated with an early risk pathway by testing its association with infant temperamental difficultness, and (2) whether pregnancy quitting is associated with an early protective pathway, testing its association with easy infant temperament. Methods: We used the 9-month-old sweep of the Millennium Cohort Study, a cohort of over 18 000 infants born in 2000-2. Mothers were classified as pregnancy non-smokers, quitters and light or heavy smokers. Temperamental positive mood, receptivity to novelty and regularity were assessed with the Carey Infant Temperament Scale. Results: Pregnancy quitters had infants with the highest scores of easy temperament and heavy smokers had infants with the lowest scores (F = 28.51, p<0.001). Pregnancy smoking also predicted difficult temperament: heavy smoking was associated with increased risk of low positive mood (OR = 1.17, p = 0.09). In contrast, pregnancy quitting exerted a protective effect with decreased risk of distress to novelty (OR = 0.79, p<0.01) and irregularity (OR = 0.89, p = 0.02) in these infants. Conclusions: Pathways from pregnancy smoking to offspring behaviour are complex and multi-determined. These findings suggest that both exposure and maternal characteristics associated with pregnancy smoking status contribute to offspring behavioural patterns. Research that characterises differences between quitters and persistent smokers and examines the role of these differences in prediction of early vulnerabilities and problems in adaptation over time will be important for elucidating these pathways.