Binary logistic regression models predicted group membership with notable accuracy: 81.5% for sex, 77.7% for male sexual orientation, and 69.5% for androphilic male gender identity/presentation. A separate discriminant function analysis (DFA) successfully classified 69.8% of participants on the basis of their facial metrics alone. These accuracies are statistical associations measured within this specific sample; the study does not establish that facial morphology causes orientation or gender presentation, nor that the models would generalize to individuals outside these groups.

The DFA revealed two independent patterns. The first function traced a masculinity–femininity continuum, ordering the groups from most to least masculinized facial morphology: gynephilic men scored highest, followed by muxe nguiiu, while muxe gunaa and androphilic women clustered together at the feminine end. This ordering aligns with the possibility that feminine-presenting androphilic males share facial characteristics with women, and that masculine-presenting androphilic males occupy an intermediate position.

The second function was less expected. It separated both muxe categories from gynephilic men, with muxe gunaa scoring above muxe nguiiu, indicating a hyper-masculinization continuum in which muxes displayed certain facial metrics exceeding those of typical gynephilic men. Univariate partial correlations, which tested each facial metric individually while controlling for other variables, showed significant associations between facial metrics and sex in 29 of the 62 metrics, between metrics and male sexual orientation in 26 of 62, and between metrics and androphilic male gender identity/presentation in 12 of 62.

The sex-related findings partially replicated those of Skorska et al. (2015), but the male sexual orientation findings did not, underscoring that individual facial measurements are weaker and less consistent predictors than the multivariate models that combine them. Integrating the univariate differences across all 62 metrics improved predictive accuracy for distinguishing among the four groups by sex, sexual orientation, and gender identity/presentation, confirming that facial morphology encodes these traits as a coordinated pattern rather than as isolated features. Because the sample was drawn from a non-Western cultural context, the findings provide cross-cultural evidence that associations between facial morphology, sex, and sexual orientation are not artifacts of Western sampling alone.

The prediction accuracies describe group-level classification within this sample rather than individual diagnosis, and the models were built from photographs of these 400 participants. Nonetheless, by including muxe nguiiu and muxe gunaa as distinct groups, the researchers could statistically separate sexual orientation from gender identity/presentation within the same population. The results provide further evidence suggesting that associations between facial morphology, sex, and sexual orientation are present cross-culturally, while leaving open the biological mechanisms the authors discuss for how such coordinated variation arises during development.

Subject of Research: Facial Morphology Predicts Sexual Orientation and Gender Identity/Presentation in Androphilic Males Among the Istmo Zapotec

Article Title: Facial Morphology Predicts Sexual Orientation and Gender Identity/Presentation in Androphilic Males Among the Istmo Zapotec

Article References: García-Acosta, S., Sadr-Bazzaz, M., Jiménez, F. R. G., & Vasey, P. L. (2026). Facial Morphology Predicts Sexual Orientation and Gender Identity/Presentation in Androphilic Males Among the Istmo Zapotec. Archives of Sexual Behavior. https://doi.org/10.1007/s10508-026-03550-x

Image Credits: AI Generated

DOI: 10.1007/s10508-026-03550-x

Keywords: Facial, Morphology, Predicts, Sexual, Orientation, Gender, Identity, Presentation, Androphilic, Males, Istmo, Zapotec