The Science of Gender Incongruence: Evidence, Uncertainty, and a Multilevel Developmental Framework

Gender incongruence is a documented human phenomenon characterized by a persistent incongruence between experienced gender and assigned sex, but its developmental origins remain incompletely understood. Research has implicated genetic variation, prenatal endocrine signaling, neurodevelopment, body-self representation, epigenetic regulation, neurodevelopmental traits, and social-developmental processes. No single mechanism, however, has been demonstrated to be either necessary or sufficient, and apparent conflicts within the literature are compounded by inconsistent phenotype definitions, small samples, cross-sectional designs, treatment-related confounding, and the frequent conflation of gender identity, gendered behavior, gender incongruence, transgender identification, and gender dysphoria. This integrative evidence review and theoretical framework evaluates major biological and developmental hypotheses concerning gender incongruence while explicitly distinguishing established findings, replicated associations, preliminary evidence, mechanistic hypotheses, and speculation. Twin research supports contributions from both genetic and environmental variation, although estimates vary substantially across studies. Candidate-gene and whole-exome studies identify possible sex-steroid signaling and neurodevelopmental pathways but remain exploratory. Natural experiments involving congenital adrenal hyperplasia and other differences of sex development demonstrate that prenatal androgen signaling affects several sex-differentiated behaviors while exerting considerably less deterministic effects on gender identity. Neuroimaging studies report group-level structural and functional differences but do not establish a unitary or diagnostic “transgender brain.” A smaller body of research implicates neural systems involved in own-body representation and self-processing. Autism and gender diversity show a reproducible population-level association, although its causal explanation remains unresolved. Preliminary epigenome-wide findings warrant further investigation but do not establish prenatal or transgenerational epigenetic causation. To organize these findings, this paper proposes the Multilevel Developmental Incongruence Model (MDIM). MDIM treats persistent gender incongruence as a potentially heterogeneous developmental phenotype arising through multiple interacting pathways rather than a single etiological mechanism. Three methodological principles are introduced: the Phenotype Resolution Principle, which requires separation of related but non-equivalent gender-related phenotypes; the Cause–Correlate Separation Principle, which distinguishes developmental causes from biomarkers and consequences; and Identity–Embodiment Separation, which distinguishes gendered self-model, experienced body congruence, and gender-related distress. The framework generates falsifiable predictions and a research program involving adequately powered genomics, longitudinal endocrine and epigenomic studies, multimodal neuroimaging, experimental body-ownership paradigms, developmental cohorts, causal modeling, and independent replication. The available evidence is most consistent with etiological heterogeneity and probabilistic developmental influence rather than genetic, hormonal, neural, psychological, or sociocultural determinism. MDIM is presented not as an established causal theory but as an evidence-weighted framework intended to make competing explanations explicit and experimentally distinguishable.