Genome-wide association study identifies nine novel loci for 2D:4D finger ratio, a putative retrospective biomarker of testosterone exposure in utero

Nicole M Warrington, Enisa Shevroja, Gibran Hemani, Pirro G Hysi, Yunxuan Jiang, Adam Auton, Cindy G Boer, Massimo Mangino, Carol A Wang, John P Kemp, George McMahon, Carolina Medina-Gomez, Martha Hickey, Katerina Trajanoska, Dieter Wolke, M Arfan Ikram, The 23andMe Research Team, Grant W Montgomery, Janine F Felix, Margaret J Wright, David A Mackey, Vincent W Jaddoe, Nicholas G Martin, Joyce Y Tung, George Davey Smith, Craig E Pennell, Tim D Spector, Joyce van Meurs, Fernando Rivadeneira, Sarah E Medland, David M Evans

The ratio of the length of the index finger to that of the ring finger (2D:4D) is sexually dimorphic and is commonly used as a noninvasive biomarker of prenatal androgen exposure. Most association studies of 2D:4D ratio with a diverse range of sex-specific traits have typically involved small sample sizes and have been difficult to replicate, raising questions around the utility and precise meaning of the measure. In the largest genome-wide association meta-analysis of 2D:4D ratio to date (N = 15,661, with replication N = 75,821), we identified eleven loci (nine novel) explaining 3.8% of the variance in mean 2D:4D ratio. We also found weak evidence for association (β = 0.06; P = 0.02) between 2D:4D ratio and sensitivity to testosterone (length of the CAG microsatellite repeat in the androgen receptor gene) in females only. Furthermore, genetic variants associated with (adult) testosterone levels and/or sex hormone-binding globulin were not associated with 2D:4D ratio in our sample. Although we were unable to find strong evidence from our genetic study to support the hypothesis that 2D:4D ratio is a direct biomarker of prenatal exposure to androgens in healthy individuals, our findings do not explicitly exclude this possibility, and pathways involving testosterone may become apparent as the size of the discovery sample increases further. Our findings provide new insight into the underlying biology shaping 2D:4D variation in the general population.