Risks of different types (natural, social, moral, recreational etc.) have been a driving force for both human evolution and development. Working with international collaborator, Prof. Wang Xiaotian from the University of South Dakota, Li Shu research group from the Key Laboratory of Behavioral Sciences of the Institute of Psychology explored the genetic and environmental influences on human risk taking in different task domains. The researchers used the twin research paradigm and meta analysis method to support the field of risk-taking propensity, namely, whether it is in the behavior of expression, or in the genetic mechanism, the correlation between the risk-taking propensity of various fields are low.

In order to develop a valid tool for measuring individual differences in risk-taking propensity involving both evolutionarily typical and modern risks, the researchers integrated several domain-specific risk-taking and developed a synthetic scale based on the results of factor analyses and validly tests. The end product is a Domain-Specific Risk-Taking Scale across Seven Domains (DOSPERT-7): cooperation/competition, safety, reproduction, natural/physical risk, moral risk, financial risk, and gambling.

Then there was a twin study with a total of 240 same-sex twin pairs (108 female pairs and 132 male pairs) sampled from the Beijing Twin Study (BeTwiSt) registry. Using the DOSPERT-7 the researchers estimated genetic and environmental influences on individual differences in risk-taking propensity over the seven domains. The effects on risk propensity were partitioned into four components: additive genetic (A), dominant genetic (D), shared environmental (C), and non-shared environmental (E) effects.  AE (additive genetic plus non-shared environmental effects) models had the best fit for most of the domains, except for gambling and safety domains where CE models had the best fit, suggesting strong shared and non-shared environmental influences. Supporting the notion of risk-domain specificity, both the behavioral and genetic correlations among the 7 domains were generally low. Among the relatively few correlations between pairs of risk domains, the analysis revealed a common genetic factor that regulates moral, financial, and natural/physical risk taking.

After a series of meta-analyses of extant twin studies across the 7 risk domains, the results showed that individual differences in risk-taking propensity and its consistency across domains were mainly regulated by additive genetic influences and individually unique environmental experiences. The heritability estimates from the meta-analyses ranged from 29% in financial risk taking to 55% in safety.