Birds of a feather: choosing a partner leaves traces in the genes

It is well known that partners often resemble each other. Psychologist Gretchen R. B. Saunders of the University of Minnesota has now examined whether that similarity is also reflected in DNA and whether it is growing across generations; the study appeared in the journal Behavior Genetics and was reported by PsyPost on 9 October. She analysed UK Biobank data – including 82,398 likely spouse or long-term cohabiting pairs, 16,273 sibling pairs and 333,491 unrelated individuals – as well as 228,875 participants in the US All of Us Research Program. The traits studied were education, height, body mass index and five smoking and drinking measures.

Couples were most alike in educational attainment (correlation 0.45), followed by weekly alcohol consumption (0.42) and quitting smoking (0.33). The genetic similarity was much weaker, by about 80 per cent on average – for education, the correlation of genetic predispositions was 0.16. Yet the signature also appeared among unrelated people, which would not be expected if partners were chosen at random. In the US data (birth years 1900 to 2005), the education signal rose from about 0.04 for people born in the 1930s to about 0.09 for those born around 2000; in Britain (birth years 1936 to 1970) it stayed largely stable. A Lausanne-led study using British and Estonian biobank data, published in Nature Human Behaviour in August, likewise concluded that genetic partner similarity for height and education has increased in more recent generations.

The limitations are considerable: only people of European ancestry were studied, and in the UK Biobank couples were inferred from shared addresses rather than marriage records. The comparison is between people born in different periods – participation, survival and population change can skew the picture. Shared environments, mutual influence and meeting in the same neighbourhoods or workplaces cannot be cleanly separated; the PsyPost report gives no information on funding. Psychologist Norbert Meskó of the University of Pécs, who was not involved, cautioned against reading the results as genes determining whom we fall in love with. “The main takeaway is that patterns of similarity between partners can leave detectable genetic signatures across generations,” he said.