Sexual orientation is associated with 2D:4D finger length ratios in both sexes: An updated and expanded meta-analysis

Ashlyn Swift-Gallant ^1* Toe Aung ² Stephanie Salia ¹ S. Marc Breedlove ³ David Puts ^4*

• ¹ Department of Psychology, Memorial University of Newfoundland and Labrador, St. John’s, NL, Canada
• ² Department of Psychology and Counseling, Immaculata University, Immaculata, Pennsylvania, United States
• ³ Neuroscience Program, Michigan State University, East Lansing, Illinois, United States
• ⁴ Department of Anthropology, The Pennsylvania State University (PSU), University Park, United States

The ratio of the lengths of the 2 nd and 4 th fingers (2D:4D) is a putative marker for prenatal gonadal hormone signaling and has been linked to human sexual orientation. Although 2D:4D is consistently found to be lower in males than females, the association with sexual orientation is variable across studies, with one meta-analysis finding lower (more masculine) digit ratios in lesbians than heterosexual females, but no overall association in males. However, this previous meta-analysis considered neither unpublished datasets nor bisexual individuals separately from homosexual and heterosexual individuals. Moreover, 17 datasets examining relationships between 2D:4D and sexual orientation have been published since that time, and we located an additional 11 unpublished datasets. We therefore conducted an updated and expanded meta-analysis comprising 51 studies, including 44 male and 34 female datasets, totaling 227,648 participants. This meta-analysis also explored whether 2D:4D differed between heterosexual and bisexual and/or non-exclusive individuals in both sexes. Results indicate lower (more male-typical) digit ratios in homosexual women (right hand g = 0.26, left hand g = 0.16; both adjusted following trim-and-fill), and higher (more female-typical) ratios in homosexual men (right hand g = -0.17, left hand g = -0.20; both adjusted) compared to heterosexual same-sex counterparts. Moderator analyses do not support publication bias for females. For males, positive findings were more likely to be published, but robustness tests, including trim-and-fill and leave-one-out, support the findings' robustness. No significant differences were observed in 2D:4D between male or female bisexual and heterosexual individuals. These findings are consistent with evidence that prenatal androgens increase attraction to females and/or that prenatal estrogens increase attraction to males.