Tobacco use and dependence : does it originate in utero?

Sammanfattning: Prenatal nicotine exposure (such as from maternal smoking during pregnancy) has been hypothesized to cause a so called programming effect, where epigenetic changes might result in a long-lasting vulnerability to tobacco use and dependence, manifesting during adolescence or young adulthood. However, previous observational studies show mixed results, perhaps because the association could be influenced by a common genetic predisposition to tobacco use as well as by social factors influencing both parental tobacco use during pregnancy and tobacco use in the offspring. The aim of this thesis was to enhance knowledge about the association between prenatal exposure to nicotine and tobacco use and tobacco dependence later in life. Study I assessed the association between parental tobacco use during pregnancy and tobacco use and dependence in adolescent offspring, based on 3,020 youths living in Stockholm County, who were followed from age 11 to 18. Study II investigated the influence of maternal smoking during pregnancy on tobacco use in adult offspring, based on 1,124 young adults, participating in the Stockholm Public Health Survey in 2006 and 2010. The Swedish Sibling Health Cohort, which consists of 1,538 sibling pairs, 19-27 years old and discordant for maternal smoking during pregnancy, constituted the study population for Study III and IV. These studies assessed the influence of prenatal exposure to maternal smoking on tobacco use (Study III) and dependence (Study IV) in young adults, while taking genetic and environmental factors into account. Study IV was based on two subsamples where both siblings were lifetime daily smokers (193 pairs) or snus users (173 pairs). Results from this thesis showed that prenatal exposure to parental tobacco use was linked with a higher risk of heavy tobacco use and dependence in adolescent girls (Study I). Maternal smoking during pregnancy was associated with an increased risk of established and heavy snus use in young adults (Study II). However, there was no association between maternal smoking during pregnancy and tobacco use (Study III) or dependence (Study IV) when exposure-discordant siblings were compared. In this thesis, prenatal nicotine exposure from parental tobacco use was not associated with tobacco use or dependence in young adults. An association with heavy use and dependence in adolescent girls cannot be excluded, but is more likely caused by residual confounding. These findings do not support the hypothesis of an important programming effect caused by prenatal nicotine exposure. If such an effect exists it is most likely weaker than the influence of genetic and early-environmental factors. Instead, this thesis emphasizes the utmost importance of genetic and early-environmental influences on the development of tobacco use and dependence.