British Medical Journal: UNC study finds numerous virgin births reported in US since mid-1990s

Dec. 17, 2013

The prestigious British Medical Journal has published a study by University of North Carolina at Chapel Hill researchers that found numerous pregnancies reported by self-described virgins in The United States since the mid 1990s.

The researchers, led by a team from UNC’s Gillings School of Global Public Health, discovered 45 “virgin pregnancies” while analyzing data for a much larger study on adolescent health.


Dr. Amy Herring

“Actually, we weren’t looking for virgin births at all,” said Amy Herring, PhD, professor of biostatistics at UNC’s Gillings School of Global Public Health and the study’s lead author. “While analyzing data for a separate project that examined correlates of virginity in adulthood, we were surprised to discover that a number of these individuals who stated they were virgins also reported pregnancies.  Once we confirmed these were not programming errors, we became interested in understanding factors related to this type of response pattern.”

Herring adds that at no time were the women asked directly if they had virgin births.

“It is important to note that these women did not report their experiences directly as virgin pregnancies,” she said. “They answered a series of questions on pregnancy history and on history of vaginal intercourse, from which virginity status at the time of pregnancy was derived.”

The present study comprised 7,870 women, a subset of participants in the U.S. National Longitudinal Study of Adolescent Health (Add Health), who have been interviewed confidentially over a 15-year period between adolescence and adulthood (average age, 16 to 29 years).

Add Health has been and continues to be known as a leader in both scientific innovation and rigor, successfully integrating high-quality survey, biomarker and contextual data as it has followed a sample of more than 20,000 adolescents into adulthood. State-of-the-art technology, including computer-assisted self-interview methods, was used to enhance the candor of the respondents.

Add Health has been described by external reviewers as “the closest thing to a ‘national treasure’ available to the NICHD [National Institute of Child Health and Human Development] research community,” said Carolyn Tucker Halpern, PhD, professor of maternal and child health at UNC’s Gillings School.

At each interview, participants reported their history of vaginal intercourse, use of assisted reproductive technology and pregnancy history. These data were used to classify women by their virginity status at the time of reported pregnancy.

Other information, such as age, importance of religion and presence of a chastity pledge, was recorded, and respondents also indicated their knowledge of different birth-control methods.  Parents of many participants answered numerous questions regarding communication with their children about sex and birth control. Women whose reports were consistent with virgin pregnancy were more likely than other virgins to report having signed chastity pledges and were also more likely than other virgins to have parents whose responses indicated lower levels of communication regarding sex and birth control.

“I am not aware of any medically validated virgin pregnancies in humans,” Herring said.

Of 7,870 women, 5,340 reported a pregnancy, and 45 of these (0.8 percent) consistently affirmed their status as virgins and did not use assisted reproductive technology, yet reported a virgin birth.   Roughly three times that many women were classified as “born-again virgins” who also reported pregnancies but provided inconsistent reports of their virginity status over time.

Although the study used carefully designed questions and state-of-the-art self-interview technology, the authors point out that self-reported measures of potentially sensitive topics are subject to some degree of respondent bias.

Some who have examined the study data have suggested that these reports of virgin birth are a result in errors produced by accidental misreporting of data by the study participants, or a reticence on the part of participants to admit to vaginal intercourse.  Herring admitted this possibility and added an interesting clarification.

“Assuming all of these births are due to errors of some sort,” she said, “0.5 percent is a rather low rate for misclassification of dates, respondent bias and other methodological issues that challenge researchers who must rely on self-reported data to study potentially sensitive topics.” 

The full study and reported data can be found online.


Gillings School of Global Public Health contact: David Pesci, director of communications, (919) 962-2600 or