Bias in Gender Research

Much transgender healthcare research recruits through specialist clinics.

Selection Bias

Much transgender healthcare research recruits through specialist clinics. As a result, the findings may not be representative of everyone with gender-related concerns. In online surveys (for example, Littman 2018), recruitment through specific communities also plays a role.

Loss to Follow-Up

Long-term studies often suffer from high dropout rates. When participants who drop out differ from those who remain (for example, because detransitioners are less likely to report back to the clinic), the results become distorted.

Lack of Control Groups

Because of the nature of the field, RCTs are rare. Cohort studies often compare treated individuals with the general population, rather than with a comparable group of untreated transgender people — making causal claims difficult.

Publication Bias

Studies with positive outcomes are published more often and more quickly. The systematic reviews by York and SBU explicitly highlighted the challenges surrounding the quality of evidence in this field.

Short Follow-Up Periods

Many studies follow participants for one or two years. That is a short period for treatments with lifelong consequences. Effects on bone density, fertility and sexual functioning may only become apparent later, and regret often does not emerge until years afterward. A study that shows a favorable picture after one year therefore says little about the situation ten years later.

Multiple Treatments at Once

Young people who receive puberty blockers or hormones generally also receive psychological support. If they subsequently do better, it is impossible to determine whether this is due to the medication, the counseling, growing older, or the attention and validation that are part of a treatment pathway. Without a comparison group, these influences cannot be separated.

Self-Reporting and Measurement Tools

Most outcomes are measured using questionnaires completed by the participants themselves. Someone who has high expectations of a treatment and has waited a long time for it may be inclined to rate the outcome favorably. Studies also use a variety of questionnaires, making the results difficult to compare and combine.

What Low-Quality Evidence Does and Does Not Mean

Low-quality evidence does not mean that a treatment does not work. It means that the available studies cannot reliably answer the question: the actual effect may be greater, smaller or absent. That uncertainty is at the heart of the debate. You can read more about how evidence is weighed in the evidence pyramid and the systematic reviews. An overview of the methodological criticism of transgender research, including what the reviews say about it, is available at transgenderidentiteit.nl.