A new analysis published in The Lancet medical journal reports that the Trump administration quietly altered hundreds of U.S. public health datasets by replacing references to “gender” with “sex.” The changes occurred without formal disclosure and affected data managed by agencies including the Centers for Disease Control and Prevention (CDC).
Researchers reviewed 232 datasets released between late January and late March of the administration’s first year. The datasets covered topics such as veterans’ health, global tobacco use, stroke mortality, nutrition, exercise, and obesity. The study excluded datasets that are updated on a rolling basis.
According to the analysis, nearly half of the reviewed datasets were substantially modified during that period, but only 15 noted that any changes had occurred. In most of the modified datasets—106 in total—the term “gender” was replaced with “sex.” As of early May, sections of the datasets where changes are typically documented remained blank.
The study was authored by Dr. Aaron Kesselheim, a professor at Harvard Medical School, and Janet Freilich, a law professor at Boston University. They stated that these changes were not publicly disclosed and may have been made in response to an executive order issued on Donald Trump’s first day in office. The order directed federal agencies to recognise only two sexes, male and female, and to remove content perceived as promoting “gender ideology.”
That executive order was part of a broader set of actions that affected LGTBIQ+ protections. Some gender-related data had previously been removed from federal websites but were restored following a court order.
The researchers stated that it is unclear whether the underlying data values were changed or if the modifications were limited to terminology. They noted that the distinction between “gender” and “sex” is significant for data collection, as individuals may respond differently depending on how questions are phrased. This could affect the accuracy and reliability of the data.
The authors described the lack of transparency surrounding the data changes as a problem for researchers who rely on consistent and clearly documented datasets. They recommended that governments and institutions worldwide consider using additional sources of health data in light of these changes.