Genome-wide meta-analyses identify multiple loci associated with smoking behavior.
Journal: 2010/May - Nature Genetics
ISSN: 1546-1718
Abstract:
Consistent but indirect evidence has implicated genetic factors in smoking behavior. We report meta-analyses of several smoking phenotypes within cohorts of the Tobacco and Genetics Consortium (n = 74,053). We also partnered with the European Network of Genetic and Genomic Epidemiology (ENGAGE) and Oxford-GlaxoSmithKline (Ox-GSK) consortia to follow up the 15 most significant regions (n>> 140,000). We identified three loci associated with number of cigarettes smoked per day. The strongest association was a synonymous 15q25 SNP in the nicotinic receptor gene CHRNA3 (rs1051730[A], beta = 1.03, standard error (s.e.) = 0.053, P = 2.8 x 10(-73)). Two 10q25 SNPs (rs1329650[G], beta = 0.367, s.e. = 0.059, P = 5.7 x 10(-10); and rs1028936[A], beta = 0.446, s.e. = 0.074, P = 1.3 x 10(-9)) and one 9q13 SNP in EGLN2 (rs3733829[G], beta = 0.333, s.e. = 0.058, P = 1.0 x 10(-8)) also exceeded genome-wide significance for cigarettes per day. For smoking initiation, eight SNPs exceeded genome-wide significance, with the strongest association at a nonsynonymous SNP in BDNF on chromosome 11 (rs6265[C], odds ratio (OR) = 1.06, 95% confidence interval (Cl) 1.04-1.08, P = 1.8 x 10(-8)). One SNP located near DBH on chromosome 9 (rs3025343[G], OR = 1.12, 95% Cl 1.08-1.18, P = 3.6 x 10(-8)) was significantly associated with smoking cessation.
Relations:
Content
Citations
(369)
References
(23)
Chemicals
(1)
Genes
(8)
Organisms
(1)
Processes
(5)
Affiliates
(1)
Similar articles
Articles by the same authors
Discussion board
Nat Genet 42(5): 441-447

Genome-wide meta-analyses identify multiple loci associated with smoking behavior

Methods

Methods and any associated references are available in the online version of the paper at http://www.nature.com/naturegenetics/.

Supplementary Material

2

1

2

Click here to view.(655K, pdf)

1

Click here to view.(201K, pdf)

Acknowledgments

This work was funded by the University of North Carolina Lineberger Comprehensive Cancer Center University Cancer Research Fund Award and by US National Cancer Institute K07 CA118412 to H.F. Statistical analyses were carried out on the Genetic Cluster Computer (see URLs), which is supported by the Netherlands Scientific Organization (NWO 480-05-003). Acknowledgments for studies included in TAG are listed in the Supplementary Note.

Abstract

Consistent but indirect evidence has implicated genetic factors in smoking behavior12. We report meta-analyses of several smoking phenotypes within cohorts of the Tobacco and Genetics Consortium (n = 74,053). We also partnered with the European Network of Genetic and Genomic Epidemiology (ENGAGE) and Oxford-GlaxoSmithKline (Ox-GSK) consortia to follow up the 15 most significant regions (n > 140,000). We identified three loci associated with number of cigarettes smoked per day. The strongest association was a synonymous 15q25 SNP in the nicotinic receptor gene CHRNA3 (rs1051730[A], β = 1.03, standard error (s.e.) = 0.053, P = 2.8 × 10). Two 10q25 SNPs (rs1329650[G], β = 0.367, s.e. = 0.059, P = 5.7 × 10; and rs1028936[A], β = 0.446, s.e. = 0.074, P = 1.3 × 10) and one 9q13 SNP in EGLN2 (rs3733829[G], β = 0.333, s.e. = 0.058, P = 1.0 × 10) also exceeded genome-wide significance for cigarettes per day. For smoking initiation, eight SNPs exceeded genome-wide significance, with the strongest association at a nonsynonymous SNP in BDNF on chromosome 11 (rs6265[C], odds ratio (OR) = 1.06, 95% confidence interval (Cl) 1.04–1.08, P = 1.8 × 10). One SNP located near DBH on chromosome 9 (rs3025343[G], OR = 1.12, 95% Cl 1.08–1.18, P = 3.6 × 10) was significantly associated with smoking cessation.

Abstract

Previous genome-wide association studies (GWAS) for smoking behavior (Supplementary Table 1) have identified a chromosome-15 nicotinic acetylcholine receptor gene cluster as being associated with smoking quantity3. The Tobacco and Genetics (TAG) Consortium conducted GWAS meta-analyses across 16 studies originally designed to evaluate other phenotypes (for example, cardiovascular disease and type 2 diabetes). We examined four carefully harmonized smoking phenotypes (see Online Methods)—smoking initiation (ever versus never been a regular smoker), age of smoking initiation, smoking quantity (number of cigarettes smoked per day, CPD) and smoking cessation (former versus current smokers)—among people of European ancestry (Table 1). Smoking cessation contrasted former versus current smokers, where current smokers reported at interview that they presently smoked and former smokers had quit smoking at least 1 year before interview. Smokers who had quit smoking for less than 1 year at interview were excluded from the analysis to minimize misclassification, as relapse after initial smoking cessation occurs in 70% to 80% of former smokers within the first year4.

Table 1

Descriptive characteristics of the 16 studies participating in the TAG Consortium

Studyn (% female)Age, mean (s.d.)Ever smokers (%)CPD, mean (s.d.)bAge of initiation of smoking,
mean (s.d.)b
Former smokers (%)b
Population-based cohort studies
Atherosclerosis Risk in Communities (ARIC)8,330 (52.9)54.3 (5.7)60.421.0 (11.7)18.6 (5.1)57.4
Baltimore Longitudinal Study of Aging (BLSA)856 (46.0)48.1 (17.8)54.0NA19.3 (5.9)NA
Cardiovascular Health Study (CHS)3,236 (60.8)72.3 (5.4)52.317.8 (11.8)19.6 (6.6)77.8
Invecchiare in Chianti (InCHIANTI)1,200 (55.2)68.4 (15.5)43.914.8 (9.4)32.2 (16.7)57.0
Rotterdam Study5,610 (60.3)69.1 (8.9)59.215.8 (11.7)20.4 (8.2)62.6
Framingham Heart Study (FHS)7,257 (53.7)45.4 (10.9)54.215.5 (10.8)17.9 (4.2)61.7
Women’s Genome Health Study (WGHS)22,037 (100)54.7 (7.1)49.216.0 (11.0)NA75.2
Case-control studies
Atherosclerotic Disease Vascular Function and Genetic
Epidemiology (ADVANCE)
585 (58.8)45.8 (5.9)47.713.1 (14.2)17.0 (4.6)65.2
Atherosclerosis, Thrombosis and Vascular Biology Italian
Study Group (ATVB)
3,260 (11.6)39.6 (4.9)68.123.4 (14.7)17.4 (4.0)21.3
Diabetes Genetic Initiative (DGI)2,504 (50.0)61.6 (10.6)37.7NA19.0 (5.5)NA
Finland-United States Investigation of NIDDM Genetics
(FUSION)
1,055 (52.8)64.0 (7.5)46.816.3 (12.4)21.0 (7.0)65.0
International Agency for Research on Cancer (IARC)8,381 (24.7)59.6 (10.1)75.219.3 (12.9)18.7 (5.6)31.4
Myocardial Infarction Genetics Consortium (MIGen)2,647 (38.5)48.8 (8.2)64.3NANA41.1
Nurses’ Health Study (NHS)2,249 (100)70.5 (6.4)53.818.5 (10.5)19.6 (3.6)88.7
Netherlands Twin Registry-Netherlands Study of Depression
and Anxiety (NTR/NESDA)
3,438 (66.9)43.8 (13.4)64.914.5 (9.8)16.4 (4.2)52.6
MGS (GAIN):controls1,390 (54.1)51.1 (17)55.919.3 (16.4)NA62.9
Age in years.
Calculated among ever regular smokers.

NA, not available.

The 16 TAG studies performed their own genotyping, quality control and imputation (see Supplementary Tables 2 and 3 and Online Methods). Studies ranged in size from n = 585 to n = 22,037 and were genotyped on six different platforms. Genotype imputation5 resulted in a common set of ~2.5 million SNPs (Supplementary Table 3). Imputed allele dosages for each SNP (that is, the number of copies of the minor allele) were tested for association with each smoking phenotype, using an additive model.

We performed a fixed-effect meta-analysis for each smoking phenotype by computing pooled inverse variance–weighted β coefficients, s.e. values and z-scores for each SNP6. Fixed-effects analyses are regarded as the most efficient method for discovery in the GWAS setting78. Heterogeneity across studies was investigated using the I statistic9. Random-effects analyses are presented in Supplementary Table 4. We used a significance threshold of P < 5 × 10 (refs. 1011).

In the initial TAG meta-analysis, only one locus contained SNPs that exceeded genome-wide significance for one of the four phenotypes (Fig. 1 and Supplementary Table 4). A total of 130 SNPs in the 15q25.1 nicotinic receptor gene cluster were significantly associated with CPD (n = 38,181, minimum P = 4.2 × 10 at rs12914385 in CHRNA3). One SNP approached significance for smoking cessation (n = 41,278, minimum P = 5.5 × 10 for rs7872903, located ~17 kb 5′ of DBH on chromosome 9). No SNPs were significantly associated with ever versus never regular smokers (n = 74,035, minimum P = 2.2 × 10 at rs16941640 in CDC27) or age of smoking initiation (n = 24,114, minimum P = 1.6 × 10 at rs2806464, located 3′ of DISC1) in the initial TAG meta-analysis.

An external file that holds a picture, illustration, etc.
Object name is nihms222027f1.jpg

Genome-wide association results for the TAG Consortium. Manhattan plots showing significance of association of all SNPs in the TAG Consortium meta-analyses for four smoking phenotypes. (ad) Manhattan plots show SNPs plotted on the x axis according to their position on each chromosome against, on the y axis (shown as −log10P value), the association with CPD (a), former versus current smoking (b), ever versus never smoking (c) and age of smoking initiation (d).

To follow up associations identified in the TAG Consortium analyses, we partnered with the ENGAGE and Oxford-GlaxoSmithKline (Ox-GSK) consortia and conducted a reciprocal exchange of summary results for the 15 most significant genetic regions for three smoking phenotypes1213. Our regions were defined by clusters of P values < 10 (that is, where the correlations (r) were >0.5 and/or the SNPs were located <50 kb apart; Supplementary Table 5). Sample sizes across the three consortia were n = 143,023 for smoking initiation, n = 73,853 for CPD and n = 64,924 for smoking cessation (data on age of smoking initiation were not available in ENGAGE or Ox-GSK).

Results of the most significant SNPs for each smoking phenotype across the three consortia are summarized in Table 2. We identified three loci associated with CPD. The synonymous SNP rs1051730 in CHRNA3 showed the strongest association: each copy of the A allele corresponded to an increase in smoking quantity of 1 CPD (β = 1.03, s.e. = 0.056, P = 2.8 × 10, I = 0.66; Fig. 2) and accounted for 0.5% of the variance in CPD. The SNP rs16969968[G], which has been proposed as a causal variant in this region14, was the second most significant SNP associated with CPD (P = 5.57 × 10; Supplementary Fig. 1). In tests of association for SNPs within the 15q25.1 region conditional on rs1051730, we observed residual associations, with the most significant signals at rs684513[G] (P = 6.3 × 10), in CHRNA5, and rs9788682[G] (P = 1.06 × 10) and rs7163730[G] (P = 1.22 × 10), in LOC123688 (Supplementary Fig. 2 and Supplementary Table 6). Our results suggest that several markers within this region may influence CPD independently. Fine mapping and the use of the 1000 Genomes Project data should help refine these signals. We investigated whether the 15q25.1 region was associated with smoking initiation and smoking cessation as well, but no SNP in that region exceeded genome-wide significance (smoking initiation minimum P = 0.98; smoking cessation minimum P = 1.75 × 10).

An external file that holds a picture, illustration, etc.
Object name is nihms222027f2.jpg

Forest and regional plots of significant associations for CPD from meta-analyses of the TAG, Ox-GSK and ENGAGE consortia. (af) Regional association plots show SNPs plotted by position on chromosome against −log10P value with each smoking phenotype. Estimated recombination rates (from HapMap-CEU) are plotted in light blue to reflect the local LD structure on a secondary y axis. The SNPs surrounding the most significant SNP (red diamond) are color coded to reflect their LD with this SNP (using pairwise r values from HapMap-CEU): blue, r ≥ 0.8–1.0; green, 0.5–0.8, orange, 0.2–0.5; gray, <0.2. The gray bars at the bottom of the plot represent the relative size and location of genes in the region.

Table 2

Meta-analytic results from three GWAS smoking consortia

TAG meta-analysisOx-GSK meta analysisENGAGE meta analysisCombined results




SNPAllelesCoded AFnβs.e.P valuenβs.e.P valuenβs.e.P valuenβs.e.P value
CPD: CHRNA3
rs1051730G/A0.6538,181−1.02070.0868.00 × 10−3314,952−1.15930.1398.88 × 10−1720,720−0.96480.0892.15 × 10−2773,853−1.02090.0562.75 × 10−73
rs16969968G/A0.6538,181−1.01500.0854.48 × 10−3314,952−1.11530.1373.72 × 10−1620,720−0.94260.0892.07 × 10−2673,853−1.00290.0565.57 × 10−72
CPD: in LOC100188947
rs1329650T/G0.2838,181−0.43170.0912.33 × 10−614,952−0.25680.1457.61 × 10−220,720−0.34640.0921.73 × 10−473,853−0.36730.0595.67 × 10−10
rs1028936C/A0.1837,284−0.55450.1161.57 × 10−614,952−0.24510.1761.65 × 10−120,720−0.42520.1131.77 × 10−472,956−0.44640.0741.29 × 10−9
CPD: EGLN2, near CYP2A6
rs3733829G/A0.3638,1810.35380.0907.67 × 1014,9520.04770.1457.43 × 10−120,7200.42040.0892.90 × 10−673,8530.33280.0581.04 × 10−8
Smoking initiation (ever versus never smokers): BDNF
rs6265T/C0.2174,035−0.06300.0151.72 × 10−534,226−0.04480.0224.48 × 10−234,762−0.07620.0241.39 × 10−3143,023−0.06140.0111.84 × 10−8
rs1013442T/A0.2674,035−0.05680.0143.39 × 10−534,226−0.03860.0216.36 × 10−234,762−0.06740.0209.60 × 10−4143,023−0.05510.0103.31 × 10−8
rs4923457T/A0.2374,035−0.06000.0142.08 × 10−534,226−0.04210.0225.05 × 10−234,762−0.07520.0241.91 × 10−3143,023−0.05860.0113.33 × 10−8
rs4923460T/G0.2374,035−0.05980.0142.22 × 10−534,226−0.04270.0224.81 × 10−234,762−0.07340.0242.51 × 10−3143,023−0.05830.0114.08 × 10−8
rs4074134T/C0.2374,035−0.06030.0141.90 × 10−534,226−0.04210.0225.08 × 10−234,762−0.07250.0242.81 × 10−3143,023−0.05820.0114.11 × 10−8
rs1304100G/A0.2674,035−0.05570.0144.86 × 10−534,226−0.04600.0212.62 × 10−234,762−0.06510.0222.88 × 10−3143,023−0.05540.0104.44 × 10−8
rs6484320T/A0.2474,035−0.05970.0142.04 × 10−534,226−0.03870.0216.78 × 10−234,762−0.07230.0242.13 × 10−3143,023−0.05710.0104.91 × 10−8
rs879048C/A0.2374,035−0.05980.0142.28 × 10−534,226−0.04090.0225.86 × 10−234,762−0.07280.0242.41 × 10−3143,023−0.05780.0114.94 × 10−8
Smoking cessation (former versus current smokers): near DBH
rs3025343G/A0.8441,2780.11770.0265.68 × 10−623,6460.12950.0411.76 × 10−3NANANANA64,9240.12100.0223.56 × 10−8

All SNPs coded to NCBI Build 36/UCSC hg18 forward strand. Coded allele frequency refers to the allele analyzed as the predictor allele; it is not necessarily the minor allele. For CPD, 174 SNPs followed up across three consortia; 130 exceeded genome-wide significance and the two top SNPs are presented. NA, not available.

CPD was analyzed as a continuous variable representing the number of cigarettes smoked per day. Smoking initiation and smoking cessation were analyzed as dichotomous variables, contrasting ever versus never and former versus current smokers, respectively.

In addition, markers within regions on chromosomes 10q23 and 19q13 were significantly associated with CPD. The SNPs rs1329650[G] (β = 0.367, s.e. = 0.059, P = 5.7 × 10; Fig. 2) and rs1028936[A] (β = 0.446, s.e. = 0.074, P = 1.3 × 10; Supplementary Fig. 1) are located in a noncoding RNA (LOC100188947), where each additional copy of a risk allele corresponded to an increase in smoking quantity of ~0.5 CPD. Linkage disequilibrium (LD) between these SNPs is moderate (r = 0.46), suggesting that they may represent one signal. To our knowledge, this region has not been previously investigated in relation to smoking behavior or other addiction phenotypes.

The third locus identified for CPD lies in the first intron of EGLN2 on chromosome 19q13, 40 kb from the 3′ end of CYP2A6. One SNP, rs3733829, exceeded genome-wide significance, and each copy of the G allele corresponded to an increase in smoking quantity of <0.5 CPD (β = 0.333, s.e. = 0.058, P = 1.0 × 10; Fig. 2). CYP2A6 is an established candidate gene for smoking, as it encodes for an enzyme involved in the metabolic inactivation of nicotine to cotinine15. Many allelic variants of CYP2A6 result in slower metabolism of nicotine16 and are associated with lower prevalence of smoking and lower amounts of cigarette use1617. We interpret this finding with caution, as only one SNP upstream of CYP2A6 was observed and the strength of its association was moderate. However, the 19q13 region merits continued investigation given its biological plausibility as involved in nicotine metabolism and because several markers within this region were identified in the ENGAGE Consortium12. The SNP identified in our study (rs3733829) lies directly between, and shows moderate LD with, the two most significant markers identified in ENGAGE.

Eight SNPs around BDNF exceeded genome-wide significance for smoking initiation analyses across the three consortia (Fig. 3). The minimum P value was at the missense variant rs6265 (P = 1.8 × 10) located in the first exon of BDNF on chromosome 11. Each copy of rs6265[C] conferred a 6% increase in the relative risk of regular smoking (OR = 1.06, 95% c.i. 1.04–1.08); rs6265 accounted for 0.03% of the variance. BDNF belongs to a family of neurotrophins that regulate synaptic plasticity and survival of cholinergic and dopaminergic neurons18. The eight SNPs overlap an antisense transcript (BDNFos). BDNF is expressed at high levels in the prefrontal cortex and hippocampus, which are brain regions implicated in the cognitive-enhancing effects of nicotine19. Although the molecular mechanisms underlying this association have yet to be elucidated, it is plausible that genetic variation at BDNF could alter the rewarding effects of nicotine through modulation of dopamine reward circuits and could contribute to the salience of nicotine’s effects by altering formation of drug-related memories that promote continued use after initial exposure. The SNP rs6265 has been found to be associated with substance-related disorders, eating disorders and schizophrenia20. Most recently, it was identified in a GWAS for body mass index21; the allele associated with a greater body mass index was the same allele associated with regular smoking in our study.

An external file that holds a picture, illustration, etc.
Object name is nihms222027f3.jpg

Forest and regional plots of significant associations for smoking behavior. (ad) Shown are plots for smoking initiation (a,b) and smoking cessation (c,d) from meta-analyses of the TAG, Ox-GSK and ENGAGE consortia. Regional association plots show SNPs plotted by position on the chromosome against −log10P value with each smoking phenotype. Estimated recombination rates (from HapMap-CEU) are plotted in light blue to reflect the local LD structure on a secondary y axis. The SNPs surrounding the most significant SNP (red diamond) are color coded to reflect their LD with this SNP (using pairwise r values from HapMap CEU): blue, r ≥ 0.8–1.0; green, 0.5–0.8; orange, 0.2–0.5; gray, <0.2. The gray bars at the bottom of the plot represent the relative size and location of genes in the region.

For smoking cessation, one SNP, located 23 kb 5′ of DBH on chromosome 9, achieved genome-wide significance: rs3025343[G] was associated with former smoking status (OR = 1.12, 95% c.i. 1.08–1.18, P = 3.6 × 10; Fig. 3) and accounted for 0.19% of the variance in smoking cessation. Because DBH catalyzes conversion of dopamine to norepinephrine, there has been interest in DBH as a candidate gene for various psychiatric phenotypes, including smoking behavior22. Although the SNP identified in this study does not cause amino acid residue changes in DBH, gene expression may be modified either directly or through other variant(s) in strong LD. This view is supported by evidence that a genetic variant (C1021T or rs1611115), located upstream of the DBH translational start site, accounts for 51% of the variation in plasma-DBH activity in European-Americans22. Alternatively, the SNP identified in our study or a variant in LD may influence expression of other genes nearby (ADAMTSL2, FAM163B or SARDH), which would introduce new pathways to our current understanding of addiction biology.

To our knowledge, the sample sizes for the TAG Consortium alone and combined with the ENGAGE and Ox-GSK consortia are among the largest genetic meta-analyses yet conducted23. Notably, most of the loci identified in this study reside in or near known candidate genes involved in the neurobiology of smoking, which differs from the results of previous GWAS, in which variants identified have generally not been in regions previously suspected. The lack of findings for smoking initiation and cessation is noteworthy in light of considerable genetic epidemiological data suggesting a role for genetic factors in different aspects of smoking behavior (for example, heritability estimates are often >0.50)1, and we note that the loci identified do not of themselves account for more than small fractions of the phenotypic heritability. Additional smoking behavior loci may be identified with improved genomic coverage and analysis of gene-gene and gene-environment interaction, copy number variation or epigenetic effects. We acknowledge that imprecision in phenotypic assessment and differences across studies could have added noise sufficient to blur all but the most prominent genetic signals. Smoking behavior obtained by questionnaires may be subject to phenotypic misclassification. Recent work24 has shown that genetic variation at 15q25.1 influences cotinine (the main and long-lived metabolite of nicotine) measurements more strongly than it influences CPD values obtained by means of a questionnaire. Future smoking GWAS that use biomarkers or longitudinal assessments that refine phenotypic assessments by incorporating time to quitting or relapsing to smoking may be required. In addition, inclusion of multiple ethnic groups will enhance the investigation of the genetics of smoking.

Notably, the five significant loci identified in these meta-analyses were each associated with only one specific smoking phenotype. Our findings suggests that separate genetic loci contribute modestly to phenotypic variability in each aspect of smoking behavior, which, in turn, may have implications for the way in which smoking cessation therapies and tobacco control efforts are designed and targeted.

Footnotes

Note: Supplementary information is available on the Nature Genetics website.

AUTHOR CONTRIBUTIONS

TAG: study conception, design, management: H.F., P.F.S., Y.K., J. Dackor; TAG Statistical Working Group: D.-Y.L., P.K., J.P.A.I., D.P., H.F., Y.K., J. Dackor, S.P.F., N.F., E.H.L., J.D.M., J.M.V., D.I.B., D.L., B.M.E., E.L.T., B. McKnight, P.F.S., D. Absher; TAG Phenotype Working Group: C. Lerman, J.K., H.H.M., L.M.T., J.A.-M., E.H.L., J.E.R., M.D.L., J.M.V., H.F., Y.K., J. Dackor, S.P.F., P.F.S., E.L.T.; data analysis: Y.K., D.M.A., F.G., E.H.L., J.D.M., J.M.V., A.U.J., L. Bernardinelli, S.R.P., S.-J.H., B.M.E., C. Ladenvall, J.R.B.P., T.T., E.L.T., J.C.B., G.L., S.W.; TAG Manuscript Writing Group: H.F., Y.K., J. Dackor, P.F.S., C. Lerman, M.D.L., J.K., J.A.-M., P.K. All authors reviewed and approved the final version of the manuscript. The corresponding authors had access to the full data set of summary results contributed by each study.

ARIC: study conception, design, management: E.B.; phenotype collection, data management: N.F.; sample processing and genotyping: N.F.; data analysis: Y.K., N.F.

Atherosclerosis Thrombosis and Vascular Biology Italian Study Group: study conception, design, management: L. Bernardinelli, P.M.M., P.A.M., D. Ardissino; phenotype collection, data management: F.M., L. Bernandinelli; data analysis: L. Bernandinelli.

ADVANCE: study conception, design, management: S.P.F., D. Absher, T.Q., C.I., T.L.A., J.W.K.; phenotype collection, data management: S.P.F., T.Q., C.I., T.L.A., J.W.K.; sample processing and genotyping: D. Absher, T.Q.; data analysis: S.P.F., D. Absher, T.L.A., J.W.K.

Baltimore Longitudinal Study of Aging: study conception, design, management: L. Ferrucci; phenotype collection, data management: L. Ferrucci; data analysis: T.T.

CHS: study conception, design, management: B.M.P., J.C.B., C.D.F.; phenotype collection, data management: B.M.P.; sample processing and genotyping: T.H., K.D.T.; data analysis: B.M.P., E.L.T., J.C.B., B. McKnight.

DGI: study conception, design, management: L.G.; phenotype collection, data management: P.A.; data analysis: P.A., C. Ladenvall.

FUSION: study conception, design, management: K.L.M., M.B.; phenotype collection, data management: H.M.S., J.T.; data analysis: H.M.S., A.U.J.

Framingham Heart Study: study conception, design, management: R.S.V., E.J.B., D.L.; phenotype collection, data management: S.R.P., R.S.V., S.-J.H., E.J.B., D.L.; data analysis: S.R.P., S.-J.H.

GAIN: study conception, design, management: D.F.L., P.V.G.; phenotype collection, data management: A.R.S., D.F.L., J. Duan, J.S., P.V.G.; sample processing and genotyping: J. Duan, P.V.G.; data analysis: A.R.S., D.F.L., J. Duan, J.S., P.V.G. IARC/ARCAGE/Central European GWAS: phenotype collection, data management: D.Z., N.S.-D., J.L., P.R., E.F., D.M., V.B., L. Foretova, V.J., S. Benhamou, P.L., I.H., L.R., K.K., A.A., X.C., T.V.M., L. Barzan, C.C., R.L., D.I. Conway, A.Z., C.M.H., P.B.; sample processing and genotyping: J.D.M., M.L., P.B.; data analysis: E.H.L., J.D.M.

InCHIANTI: study conception, design, management: T.M.F., J.M.G., S. Bandinelli; phenotype collection, data management: Y.M.; data analysis: J.R.B.P.

MIGEN: study conception, design, management: R.E., V.S., O.M., C.J.O., D. Altshuler; phenotype collection, data management: G.L., S.M.S., R.E., V.S., B.F.V., O.M., S.K., C.J.O.; sample processing and genotyping: S.K., D. Altshuler; data analysis: G.L., B.F.V., D. Altshuler

NESDA: study conception, design, management: B.W.P., J.H.S.; phenotype collection, data management: B.W.P., J.H.S., N.V.; sample processing and genotyping: B.W.P., J.H.S.; data analysis: N.V.

NTR: study conception, design, management: D.I.B., G.W., E.J.C.d.G.; phenotype collection, data management: D.I.B., G.W., E.J.C.d.G., J.M.V.; sample processing and genotyping: D.I.B., G.W., E.J.C.d.G.; data analysis: J.M.V.

NHS: phenotype collection, data management: S.E.H., D.J.H., P.K., F.G.; sample processing and genotyping: S.J.C., S.E.H., D.J.H., P.K.; data analysis: S.J.C., F.G., P.K.

Rotterdam: study conception, design, management: A.H.; phenotype collection, data management: H.T., A.G.U.; sample processing and genotyping: H.T., A.G.U.; data analysis: H.T., A.G.U., S.W., C.M.v.D.

WGHS: study conception, design, management: B.M.E., G.P., D.I. Chasman, P.M.R.; phenotype collection, data management: B.M.E., G.P., D.I. Chasman, P.M.R.; sample processing and genotyping: G.P., D.I. Chasman; data analysis: B.M.E., G.P., D.I. Chasman.

COMPETING FINANCIAL INTERESTS

The authors declare no competing financial interests.

Reprints and permissions information is available online at http://npg.nature.com/reprintsandpermissions/.

Footnotes

References

  • 1. Rose RJ, Broms U, Korhonen T, Dick DM, Kaprio J. Genetics of Smoking Behavior. In: Kim Y-K, editor. Handbook of Behavior Genetics. New York: Springer; 2009. pp. 411–432. [PubMed]
  • 2. Li MDIdentifying susceptibility loci for nicotine dependence: 2008 update based on recent genome-wide linkage analyses. Hum. Genet. 2008;123:119–131.[PubMed][Google Scholar]
  • 3. Thorgeirsson TE, et al A variant associated with nicotine dependence, lung cancer and peripheral arterial disease. Nature. 2008;452:638–642.[Google Scholar]
  • 4. Fiore MC, Smith SS, Jorenby DE, Baker TB. The effectiveness of the nicotine patch for smoking cessation. A meta-analysis. J. Am. Med. Assoc. 1994;271:1940–1947.[PubMed]
  • 5. Li Y, Willer C, Sanna S, Abecasis GGenotype imputation. Annu. Rev. Genomics Hum. Genet. 2009;10:387–406.[Google Scholar]
  • 6. de Bakker PI, et al Practical aspects of imputation-driven meta-analysis of genome-wide association studies. Hum. Mol. Genet. 2008;17:R122–R128.[Google Scholar]
  • 7. Kraft P, Zeggini E, Ioannidis JP. A. Replication in genome-wide association studies. Stat. Sci. 2010 published online, doi:10.1214/09-STS290.
  • 8. Pereira TV, Patsopoulos NA, Salanti G, Ioannidis JPDiscovery properties of genome-wide association signals from cumulatively combined data sets. Am. J. Epidemiol. 2009;170:1197–1206.[Google Scholar]
  • 9. Ioannidis JP, Patsopoulos NA, Evangelou EHeterogeneity in meta-analyses of genome-wide association investigations. PLoS One. 2007;2:e841.[Google Scholar]
  • 10. Pe’er I, et al Evaluating and improving power in whole-genome association studies using fixed marker sets. Nat. Genet. 2006;38:663–667.[PubMed][Google Scholar]
  • 11. Pe’er I, Yelensky R, Altshuler D, Daly MJEstimation of the multiple testing burden for genomewide association studies of nearly all common variants. Genet. Epidemiol. 2008;32:381–385.[PubMed][Google Scholar]
  • 12. Thorgeirsson T, et al Sequence variants at CHRNB3-CHRNA6 and CYP2A6 affect smoking behavior. Nat. Genet. 2010;42:448–453.[Google Scholar]
  • 13. Liu J, et al Meta-analysis and imputation refines the association of 15q25 with smoking quantity. Nat. Genet. 2010;42:436–440.[Google Scholar]
  • 14. Saccone NL, et al Multiple distinct risk loci for nicotine dependence identified by dense coverage of the complete family of nicotinic receptor subunit (CHRN) genes. Am. J. Med. Genet. B. Neuropsychiatr. Genet. 2009;150B:453–466.[Google Scholar]
  • 15. Nakajima M, et al Role of human cytochrome P4502A6 in C-oxidation of nicotine. Drug Metab. Dispos. 1996;24:1212–1217.[PubMed][Google Scholar]
  • 16. Mwenifumbo JC, Tyndale RFMolecular genetics of nicotine metabolism. Handb. Exp. Pharmacol. 2009;192:235–259.[PubMed][Google Scholar]
  • 17. Ray R, Tyndale RF, Lerman CNicotine dependence pharmacogenetics: role of genetic variation in nicotine-metabolizing enzymes. J. Neurogenet. 2009;23:252–261.[Google Scholar]
  • 18. Zhang LI, Poo MMElectrical activity and development of neural circuits. Nat. Neurosci. 2001;4 Suppl:1207–1214.[PubMed][Google Scholar]
  • 19. Levin ED, McClernon FJ, Rezvani AHNicotinic effects on cognitive function: behavioral characterization, pharmacological specification, and anatomic localization. Psychopharmacology (Berl.) 2006;184:523–539.[PubMed][Google Scholar]
  • 20. Gratacòs M, et al Brain-derived neurotrophic factor Val66Met and psychiatric disorders: meta-analysis of case-control studies confirm association to substance-related disorders, eating disorders, and schizophrenia. Biol. Psychiatry. 2007;61:911–922.[PubMed][Google Scholar]
  • 21. Thorleifsson G, et al Genome-wide association yields new sequence variants at seven loci that associate with measures of obesity. Nat. Genet. 2009;41:18–24.[PubMed][Google Scholar]
  • 22. Zabetian CP, et al A quantitative-trait analysis of human plasma-dopamine beta-hydroxylase activity: evidence for a major functional polymorphism at the DBH locus. Am. J. Hum. Genet. 2001;68:515–522.[Google Scholar]
  • 23. Hindorff LA, et al Potential etiologic and functional implications of genome-wide association loci for human diseases and traits. Proc. Natl. Acad. Sci. USA. 2009;106:9362–9367.[Google Scholar]
  • 24. Keskitalo K, et al Association of serum cotinine level with a cluster of three nicotinic acetylcholine receptor genes (CHRNA3/CHRNA5/CHRNB4) on chromosome 15. Hum. Mol. Genet. 2009;18:4007–4012.[Google Scholar]
Collaboration tool especially designed for Life Science professionals.Drag-and-drop any entity to your messages.