Fig. 2
From: Genetic evidence for assortative mating on alcohol consumption in the UK Biobank
Disentangling mechanisms underlying spousal alcohol use similarities. a Phenotypic concordance. Spousal concordance for alcohol use during their relationship (\(A_{\mathrm{I}}^{\mathrm{R}}\), \(A_{\mathrm{P}}^{\mathrm{R}}\)), as measured in UK Biobank, could be explained by several different possibilities. Assortative mating: Alcohol consumption prior to relationship formation (\(A_{\mathrm{I}}^{\mathrm{O}}\), \(A_{\mathrm{P}}^{\mathrm{O}}\)) influences mate choice C. Comparing assorted pairs induces spousal correlations for Ao and AR. Partner interaction effects: Spouses may influence each other’s alcohol behaviour over time while in a relationship. We represent this stochastic process by the arrows between alcohol use at relationship formation (\(A_{\mathrm{I}}^{\mathrm{O}}\), \(A_{\mathrm{P}}^{\mathrm{O}}\)) and alcohol use at study entry (\(A_{\mathrm{I}}^{\mathrm{R}}\), \(A_{\mathrm{P}}^{\mathrm{R}}\)). Note that effects likely relate to relationship length. Relationship dissolution: Spousal alcohol behaviour during the relationship (\(A_{\mathrm{I}}^{\mathrm{R}}\) and \(A_{\mathrm{P}}^{\mathrm{R}}\)) influences the duration of the relationship D. Comparing non-dissolved pairs induces spousal correlations for AR in the remaining couples. Confounding factors: Unmeasured confounders U influence both C and Ao leading to spousal correlation for AR independent of an effect of Ao on C. b Mendelian randomization framework. An association between an individual’s alcohol influencing genotype ZI and their spouse’s alcohol use \(A_{\mathrm{P}}^{\mathrm{R}}\) suggests that the spousal concordance is explained by assortative mating, partner interaction effects or relationship dissolution. Genetic variants are unlikely to be associated with socio-economic confounders suggesting that the confounding factors possibility is unlikely. c Genotypic concordance. Genotypic concordance for alcohol related genetic variants (ZI,ZP) suggests that some degree of the spousal concordance is explained by comparing assorted or non-dissolved pairs (assortative mating/ relationship dissolution). Partner interaction effects cannot lead to genotypic concordance because genotypes are fixed from birth
