Supplementary Material for: Genetic and environmental factors underlying parallel changes in body mass index and alcohol consumption: a 36-year longitudinal study of adult twins

Introduction: While the genetic and environmental underpinnings of body weight and alcohol use are fairly well-known, determinants of simultaneous changes in these traits are still poorly known. We sought to quantify the environmental and genetic components underlying parallel changes in weight and alcohol consumption, and to investigate potential covariation between them. Methods: The analysis comprised 4461 adult participants (58% women) from the Finnish Twin Cohort with four measures of alcohol consumption and body mass index (BMI) over a 36-year follow-up. Trajectories of each trait were described by growth factors, defined as intercepts (i.e., baseline) and slopes (i.e., change over follow-up), using Latent Growth Curve Modeling. Growth values were used for male (190 monozygotic pairs, 293 dizygotic pairs) and female (316 monozygotic pairs, 487 dizygotic pairs) same-sex complete twin pairs in multivariate twin modeling. The variances and covariances of growth factors were then decomposed into genetic and environmental components. Results: The baseline heritabilities were similar in men (BMI: h2=79% [95% Confidence Interval: 74,83]; alcohol consumption: h2=49% [32,67]) and women (h2=77% [73,81]; h2=45% [29,61]). Heritabilities of BMI change were similar in men (h2=52% [42,61]) and women (h2=57% [50,63]), but the heritability of change in alcohol consumption was significantly higher (p=0.03) in men (h2=45% [34,54]) than in women (h2=31% [22,38]). Significant additive genetic correlations between BMI at baseline and change in alcohol consumption were observed in both men (rA =-0.17 [-0.29,-0.04]) and women (rA=-0.18 [-0.31,-0.06]). Non-shared environmental factors affecting changes in alcohol consumption and BMI were correlated in men (rE=0.18 [0.06,0.30]). Among women, non-shared environmental factors affecting baseline alcohol consumption and the change in BMI were inversely correlated (rE=-0.11 [-0.20, -0.01]). Conclusions: Based on genetic correlations, genetic variation underlying BMI may affect change in alcohol consumption. Independent of genetic effects, change in BMI correlates with change in alcohol consumption in men, suggesting direct effects between them.引言：尽管体重与饮酒行为的遗传及环境基础已得到较为充分的阐明，但这两种性状同步变化的影响因素仍未得到充分揭示。本研究旨在量化体重与饮酒量同步变化背后的遗传与环境组分，并探究二者间潜在的共变关系。 方法：本研究的分析对象来自芬兰双生子队列（Finnish Twin Cohort），共纳入4461名成年参与者（女性占比58%），在长达36年的随访期间对其饮酒量与身体质量指数（body mass index, BMI）进行了4次测量。采用潜增长曲线模型（Latent Growth Curve Modeling），以截距（即基线水平）和斜率（即随访期间的变化量）作为增长因子，对两种性状的变化轨迹进行刻画。针对男性同卵双生子对（190对）、男性异卵双生子对（293对）以及女性同卵双生子对（316对）、女性异卵双生子对（487对）的完整同性别双生子对，将其增长因子值用于多变量双生子模型分析。随后将增长因子的方差与协方差分解为遗传组分与环境组分。 结果：基线水平的遗传力在男性与女性中相似：男性BMI的遗传力h²=79%[95%置信区间：74,83]，饮酒量的遗传力h²=49%[32,67]；女性BMI的遗传力h²=77%[73,81]，饮酒量的遗传力h²=45%[29,61]。BMI变化的遗传力在男性（h²=52%[42,61]）与女性（h²=57%[50,63]）中相近，但男性饮酒量变化的遗传力（h²=45%[34,54]）显著高于女性（h²=31%[22,38]，p=0.03）。基线BMI与饮酒量变化之间存在显著的加性遗传相关：男性的r_A=-0.17[-0.29,-0.04]，女性的r_A=-0.18[-0.31,-0.06]。影响饮酒量与BMI变化的非共享环境因素在男性中存在相关（r_E=0.18[0.06,0.30]）。在女性中，影响基线饮酒量与BMI变化的非共享环境因素呈负相关（r_E=-0.11[-0.20,-0.01]）。 结论：基于遗传相关性结果，BMI背后的遗传变异可能会影响饮酒量的变化。在男性中，不依赖于遗传效应的BMI变化与饮酒量变化存在相关，提示二者间存在直接作用。