Some Families May Be Flipping a Weighted Coin on Baby Sex

Three daughters and one more try. The odds on a fourth girl may run higher than a fair coin would allow.

Researchers at Harvard T.H. Chan School of Public Health combed through data on 58,007 U.S. women who had at least two live births, covering about 146,000 pregnancies spanning 1956 through 2015. It was a longitudinal analysis paired with a genome-wide scan, drawn from Nurses' Health Study II and Nurses' Health Study 3, and it appeared in Science Advances in July 2025.

Some families skewed toward one sex more than chance alone would predict. Mothers with at least three kids were likelier to end up with a house of only sons or only daughters.

Age seemed to matter a little too. Mothers whose first child arrived over age 28 carried a slight edge toward those single-sex groups. The authors floated age-related body changes as a possible reason, though that remains speculation.

Variants in two maternal genes also turned up. TSHZ1 tracked with families of boys only, and NSUN6 with families of girls only. The senior author, Jorge Chavarro, said the team cannot say why those genes would matter, which raises fresh questions.

Caution applies. The cohort was about 95 percent white, included data from mothers only and could not account for fathers, so the results may not carry to other populations. The findings show an association and need replication.

Two genes, TSHZ1 and NSUN6, sit on file for all-son and all-daughter families. The researchers cannot say what either is doing in the nursery.

Read the full story at Harvard T.H. Chan School of Public Health, July 22, 2025

Hot Take: Every family trying for one more has been running a trial with no protocol. This study suggests the randomization may have been compromised before enrollment.