A CDC survey analysis shows a sharp drop in lesbian or bisexual identification among younger US teenagers, especially ninth-grade girls. But while the numbers have prompted competing explanations, they don’t reveal why students’ answers have changed, making careful interpretation just as important as the headline figures.

Essential Takeaways

  • The sharpest reported decline: In an analysis of CDC survey data, 17% of ninth-grade girls identified as lesbian or bisexual in 2025, down from 30% in 2021.
  • Bisexual identification also fell: Among ninth-grade girls, the share identifying as bisexual dropped from 26% to 13% over the same period.
  • The trend varies by age: Changes were smaller among older teenagers than among ninth-graders.
  • There’s no established cause: The CDC survey records students’ responses; it doesn’t explain why they may change over time.
  • Different studies measure different groups: A separate report on university students examines identity trends in a different population from the high-school survey.

The biggest change is among younger girls

The striking detail in Jean Twenge’s analysis is how much the figures differ by age. Among ninth-grade girls, 17% identified as lesbian or bisexual in 2025, compared with 30% in 2021. The drop is driven largely by fewer students selecting bisexual: that share went from 26% to 13%.

Twenge, a psychology professor at San Diego State University, analysed data from the CDC’s Youth Risk Behavior Survey. The Christian Today and Washington Post reports both describe a broader fall in reported LGB identification among younger teenagers, with a less pronounced change among older students.

That age difference matters. A single headline figure can make the shift sound uniform, when the reported pattern is more concentrated among the youngest high-school students.

The survey records answers, not explanations

The CDC’s Youth Risk Behavior Surveillance System gathers information from US high-school students on health-related behaviours and experiences. Its surveys include topics such as sexual behaviour, mental health and substance use, with questionnaires and technical material available for readers to examine.

But a survey can show what respondents said at the time; it can’t, by itself, establish why their answers changed. The CDC’s data guide explains how survey responses, categories and weighted estimates should be interpreted, while also underscoring why percentages need to be read in context.

Twenge has said the reason for the decline is unclear. She plans to examine possible roles for the pandemic, politics and social media, but the survey doesn’t prove any one of those factors caused the shift.

Commentators disagree on what’s behind the numbers

Some commentators have pointed to cultural changes. Glenn T. Stanton of Focus on the Family suggested that shifting social influences might help explain the pattern, while acknowledging that the cause is not known. Those comments are interpretations, not findings established by the CDC data.

Terry Schilling, president of the American Principles Project, attributed the trend to federal funding cuts. Twenge noted that the decline began during the Biden administration and questioned why a political explanation would have a greater effect on ninth-graders than on older students.

The Washington Post also reported a caution from Jody Herman of UCLA’s Williams Institute: without confidence intervals, it’s harder to assess whether differences in the analysis are statistically significant, and more evidence over time is needed. In other words, the figures invite questions, but they don’t settle them.

Identity and behaviour are related, but not interchangeable

Twenge’s analysis also looks at reported same-sex partners among sexually active students. Her Generation Tech post says reports among girls in this subgroup peaked in 2021 and later declined, particularly among younger students.

That measure isn’t the same as sexual identity. It covers only sexually active students and asks about reported behaviour, while the identity figures reflect students’ selections from survey response options. Treating the two as interchangeable risks blurring what each statistic actually describes.

For readers comparing figures across survey years, the CDC’s questionnaires and technical documentation are useful places to start. Check who was included, how a category was defined and which students form the denominator before drawing conclusions.

A separate study looks at university students

The story also sits alongside a 2025 report by Eric Kaufmann, published through the Manhattan Institute and examining survey data from US undergraduates. It describes a decline in transgender and queer identification in that student population, but those findings concern university students, not ninth-graders.

The populations and measures are different, so the two sets of figures shouldn’t be treated as direct comparisons. They offer a wider view of changing identity responses among young people, but neither replaces careful attention to what each survey asked and whom it surveyed.

For now, the clearest conclusion is also the most restrained: the reported figures have changed, but the reasons remain open to investigation.

A trend is worth noticing; understanding it takes more than one percentage.

Source Reference Map

Story idea inspired by: [1]

Sources by paragraph: