With genetic causal inference methods, researchers showed that testosterone does not affect socioeconomic status or health in men, countering prior observational studies.
Testosterone and socioeconomic status
Testosterone is best known as a sex hormone, but there is also much interest in its influence on human behaviour. In the laboratory, testosterone has been shown to promote aggressive behaviour and risk tolerance. Both traits are believed to bring financial rewards in careers, for instance, by increasing willingness to engage in wage bargaining.
These findings have seeded a widespread belief that testosterone levels contribute to socioeconomic success. In fact, male executives with higher circulating testosterone levels were observed to have a higher number of subordinates. Some have suggested that the reverse is true – that socioeconomic position influences testosterone levels. Factors associated with low socioeconomic status, such as smoking and obesity, have been observed to lower testosterone levels in men.
However, as the saying goes, correlation is not causation. Most previous work has been observational, thereby failing to establish whether there is a causal relationship between testosterone and socioeconomic status. Furthermore, most studies have only involved specific occupational samples and focused almost exclusively on men.
UK BioBank data
A recent paper, published in Science Advances, refuted previous work that had reported associations between testosterone and socioeconomic status. Here, researchers at the MRC Integrative Epidemiology Unit of the University of Bristol analysed 306,248 participants of the UK BioBank.
These participants were more representative of the general population, involving both men and women across different occupations. The comprehensive dataset also allowed the researchers to examine numerous health exposures and socioeconomic outcomes with great precision.
With a sex-stratified genome-wide association analysis (GWAS), the researchers identified single nucleotide polymorphisms associated with bioavailable testosterone in men and women.
Debunking the testosterone myth
Consistent with previous observational studies, multivariable-adjusted regression analysis indicated that testosterone was positively associated with advantaged socioeconomic status and health in men. Interestingly, the effect of testosterone was the opposite for women.
To determine whether the relationships were causal, the researchers performed Mendelian randomisation analysis – a genetic causal inference method. Analogous to a randomised controlled trial, Mendelian randomisation is a powerful tool to evaluate whether a risk factor exerts a genuine causal effect on an observed outcome. Since genotypes are assigned randomly, genetic variants act as a proxy for the randomised assignment of clinical interventions. Researchers assessed the impact of testosterone-associated variants on socioeconomic outcomes, such as income, employment status, neighbourhood-level deprivation, educational qualifications, as well as overall health and self-reported risk-taking behaviour.
The analysis showed that testosterone levels had no causal effect on socioeconomic position, health or risk-taking in either sex. However, the estimates for women were less precise as GWAS revealed fewer testosterone-associated polymorphisms in women than in men. Using larger female cohorts in the future may remedy this.
Nevertheless, the male results disproved prior observational work, which have probably suffered from multiple confounding variables or reverse causation. The researchers further suggested that previously reported associations more likely reflect the influence of socioeconomic status on testosterone. This provides evidence for the fact that testosterone levels do not directly affect our chance of success and reaffirms the power of genetic analysis for de-bunking common biological misconceptions!
Image credit: freepik – Freepik
