The heritability of smoking initiation (SI) and number of cigarettes smoked (NC) was determined in 3657 Dutch twin pairs. For SI a heritability of 36% was found and for NC of 51%. Both SI and NC were also significantly influenced by environmental factors shared by family members. The etiological factors that influence these traits partly overlap. Linkage analyses were performed on data of 536 DZ twins and siblings from 192 families, forming 592 sibling pairs. Results suggested QTLs on chromosome 6 (LOD ¼ 3.05) and chromosome 14 (LOD ¼ 1.66) for SI and on chromosome 3 (LOD ¼ 1.98) for NC. Strikingly, on chromosome 10 a peak was found in the same region for both SI (LOD ¼
INTRODUCTIONLarge-scale population-based twin and family studies have shown that genetic factors contribute to individual differences in smoking behavior. 1-4 Several different, possibly correlated, dimensions of smoking behavior can be distinguished: smoking initiation (SI), number of cigarettes smoked per day and nicotine dependence (ND). 5 Koopmans et al 6 investigated the heritability of SI and the number of cigarettes smoked per day (NC) in adolescent Dutch twins by considering a single liability model, an independent liability model and a combined model. The combined model best described the data and showed that 39% of the variance in SI and 86% of the variance in NC was explained by genetic influences. Kendler et al 7 found that liabilities to SI and ND were substantially correlated but not identical, and that heritable factors played an important role in both SI and in ND.The next step after obtaining evidence for significant heritability is to identify chromosomal regions involved in smoking behavior, either by linkage or association approaches. 8 Both human and animal studies have explored candidate genes for smoking behavior. Association studies point to dopamine receptor genes, dopamine transporter genes, cytochrome P450 and serotonergic genes. 9,10 Association studies have relatively high statistical power, and can detect quantitative trait loci (QTLs) with only small effects. A possible disadvantage of the candidate gene approach is that the focus is on known pathways, which may lead us to overlook genes that are etiologically important, because of our ignorance of other biological systems involved. In contrast, linkage analysis will identify chromosomal regions that harbor known and