New VCU Statistical Meta-Analysis Challenges Longstanding Claims of High Transgender Desistance Rates Among Youth

New VCU meta-analysis reveals that claims of 90% transgender desistance are statistically unsupported. Explore why modern science challenges these figures.

By: AXL Media

Published: Apr 28, 2026, 6:02 AM EDT

Source: Information for this report was sourced from EurekAlert

The Statistical Flaws in Gender Identity Renouncement Claims

A comprehensive meta-analysis led by Virginia Commonwealth University has challenged the validity of frequently cited statistics regarding transgender youth. For years, proponents of restrictions on gender-affirming care have argued that a vast majority of children eventually renounce their transgender identities, citing figures as high as 90%. However, according to lead author Catherine Wall, an assistant professor at VCU, these numbers are not supported by rigorous statistical scrutiny. The research, published in APA PsycNet, indicates that these estimates fluctuate wildly depending on how data is interpreted, rendering the current legislative reliance on them scientifically questionable.

Historical Context and the Evolution of Pathological Frameworks

The study examined 11 original papers from a prominent 2016 blog post alongside five more recent publications to understand the origin of these figures. A significant portion of the data stems from research conducted before 1990, a period when the medical community viewed transgender identities through a lens of pathological mental illness. During this era, clinical goals often focused on encouraging children to conform to their assigned gender at birth, a practice now associated with conversion therapy. Catherine Wall noted that these early frameworks were designed to actively discourage gender expression, which inherently biased the reported outcomes of those studies.

Methodological Errors and the Problem of Participant Nonresponse

Researchers identified critical flaws in how early studies categorized "desistance," or the act of a child returning to their birth-assigned gender identity. Many older studies included small sample sizes and, crucially, counted any participant who dropped out or failed to respond to follow-ups as having desisted. According to the VCU analysis, assuming that a loss of contact equals an identity shift is a statistically unsound practice. When the team ran simulations accounting for different reasons for study withdrawal, they found that desistance and persistence rates could both be estimated anywhere from zero to 100%, highlighting a complete lack of data stability.