High tobacco consumption lowers body weight: a Mendelian randomization study of the Copenhagen General Population Study

TY - JOUR

T1 - High tobacco consumption lowers body weight

T2 - a Mendelian randomization study of the Copenhagen General Population Study

AU - Winsløw, Ulrik C

AU - Rode, Line

AU - Nordestgaard, Børge G

N1 - © The Author 2015; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association.

PY - 2015/5/27

Y1 - 2015/5/27

N2 - Background: Conflicting evidence has been found regarding the association between high tobacco consumption and body weight among smokers. We tested the hypothesis that high tobacco consumption is causally associated with low body weight. Methods: We conducted a Mendelian randomization study with a genetic variant in CHRNA3 (rs1051730) as proxy for high tobacco consumption. The cohort consisted of 80 342 participants from the Copenhagen General Population Study, with details on body weight, smoking habits and CHRNA3 genotype, including 15 220 current smokers. Results: In observational analyses, high tobacco consumption was associated with high body weight, body mass index, waist circumference, hip circumference and waist-hip ratio. In multivariable adjusted models a 1-cigarette/day higher tobacco consumption was associated with 0.05 kg (95% confidence interval 0.02; 0.08) higher body weight, 0.02 kg/m2 (0.01; 0.02) higher body mass index, 0.10 cm (0.07; 0.12) higher waist circumference and 0.02 cm (0.001; 0.03) higher hip circumference. In contrast, the per-allele increase in CHRNA3 rs1051730 associated with high tobacco consumption was associated with 0.59 kg (0.96; 0.22) lower body weight, 0.23 kg/m2 (0.33; 0.13) lower body mass index, 0.32 cm (0.74; 0.003) lower waist circumference and 0.45 cm (0.66; 0.24) lower hip circumference. No association was found between tobacco consumption and waist-hip ratio in genetic analysis, or among CHRNA3 genotype and any of the outcome variables in former or never smokers. Conclusions: High tobacco consumption causes lower body weight among current smokers. However, smoking does not seem to affect body shape or fat distribution causally. The lack of association between CHRNA3 genotype and body weight among former smokers and never smokers favours smoking as the causal factor for the observed associations.

AB - Background: Conflicting evidence has been found regarding the association between high tobacco consumption and body weight among smokers. We tested the hypothesis that high tobacco consumption is causally associated with low body weight. Methods: We conducted a Mendelian randomization study with a genetic variant in CHRNA3 (rs1051730) as proxy for high tobacco consumption. The cohort consisted of 80 342 participants from the Copenhagen General Population Study, with details on body weight, smoking habits and CHRNA3 genotype, including 15 220 current smokers. Results: In observational analyses, high tobacco consumption was associated with high body weight, body mass index, waist circumference, hip circumference and waist-hip ratio. In multivariable adjusted models a 1-cigarette/day higher tobacco consumption was associated with 0.05 kg (95% confidence interval 0.02; 0.08) higher body weight, 0.02 kg/m2 (0.01; 0.02) higher body mass index, 0.10 cm (0.07; 0.12) higher waist circumference and 0.02 cm (0.001; 0.03) higher hip circumference. In contrast, the per-allele increase in CHRNA3 rs1051730 associated with high tobacco consumption was associated with 0.59 kg (0.96; 0.22) lower body weight, 0.23 kg/m2 (0.33; 0.13) lower body mass index, 0.32 cm (0.74; 0.003) lower waist circumference and 0.45 cm (0.66; 0.24) lower hip circumference. No association was found between tobacco consumption and waist-hip ratio in genetic analysis, or among CHRNA3 genotype and any of the outcome variables in former or never smokers. Conclusions: High tobacco consumption causes lower body weight among current smokers. However, smoking does not seem to affect body shape or fat distribution causally. The lack of association between CHRNA3 genotype and body weight among former smokers and never smokers favours smoking as the causal factor for the observed associations.

KW - Adult

KW - Aged

KW - Aged, 80 and over

KW - Body Mass Index

KW - Denmark

KW - Female

KW - Heterozygote

KW - Homozygote

KW - Humans

KW - Male

KW - Mendelian Randomization Analysis

KW - Middle Aged

KW - Phenotype

KW - Receptors, Nicotinic

KW - Smoking

KW - Tobacco Use

KW - Waist-Hip Ratio

KW - Weight Gain

KW - Weight Loss

KW - Young Adult

U2 - 10.1093/ije/dyu276

DO - 10.1093/ije/dyu276

M3 - Journal article

C2 - 25777141

SN - 0300-5771

VL - 44

SP - 540

EP - 550

JO - International Journal of Epidemiology

JF - International Journal of Epidemiology

IS - 2

ER -