Adult sex friend
Among non-Hispanic whites in Add Health, KING estimates of genetic homophily were about two-thirds the magnitude of our previous KING estimates of genetic similarity among spouses in the US Health and Retirement Study (friend similarity = 0.031, CI = 0.022–0.036, compared with spousal similarity = 0.045 from ref. REAP estimates of genetic similarity were somewhat smaller.
Our second analysis tested whether friends were more similar to one another on specific phenotype-related genetic dimensions.
At the most recent interview in 2008, ∼12,000 Add Health participants provided DNA for genotyping and genome-wide single-nucleotide polymorphism (SNP) data were assayed.
We linked these genetic data with social network information from the original school-based surveys along with information about personal characteristics and social environments accumulated across Add Health follow-up waves.
In step 3, we evaluate a potential implication of genetic similarity among friends: social–genetic effects, or the association between the genotypes of one’s social peers and one’s own phenotype (net of own genotype).
We tested whether friends were more genetically similar to one another compared with random pairs of individuals.
We considered the genetics of three phenotypes: height, body mass index (BMI) (a measure of adiposity), and educational attainment.
Recent research has evaluated the possibility that unobserved genotypes may play an important role in the creation of homophilous relationships.To address potential confounding of analysis by ancestry, we focused on unrelated respondents of European ancestry.We also conducted analysis of genetic relatedness of the full Add Health sample using the REAP algorithm (26), which is designed to estimate genetic similarity in the presence of population stratification. Estimates of genetic similarity among friends were positive (Table 1).Our study reported significant findings of a “social genome” that can be quantified and studied to understand human health and behavior.In a national sample of more than 5,000 American adolescents, we found evidence of social forces that act to make friends and schoolmates more genetically similar to one another compared with random pairs of unrelated individuals.The degree to which genetics are implicated in the formation and consequences of social relationships is of growing interest to the new field of sociogenomics (1, 2).Analysis of spousal genotypes suggests that spouses are more genetically similar to one another compared with random pairs of individuals in the population (3–9). In previous analyses, we estimated that genetic homogamy was about one-third the magnitude of educational homogamy (3), even when specifically examining education-associated genotypes (8).This subtle genetic similarity was observed across the entire genome and at sets of genomic locations linked with specific traits—educational attainment and body mass index—a phenomenon we term “social–genetic correlation.” We also find evidence of a “social–genetic effect” such that the genetics of a person’s friends and schoolmates influenced their own education, even after accounting for the person’s own genetics.Humans tend to form social relationships with others who resemble them.Adult friends are, on average, more genetically similar than random pairs from the population (13).Genetic similarity among friendship networks is important for at least two reasons.