Blanchard · Archives of sexual behavior 2018 · Meta-analysis of observational studies · n=26 studies (7,140 homosexual and 12,837 heterosexual males)

Fraternal Birth Order, Family Size, and Male Homosexuality: Meta-Analysis of Studies Spanning 25 Years.

Cited 196 times in the scientific literature.

Level 3 - non-randomized controlled study

Meta-analysis of observational case-control and cross-sectional studies.

PubMed 28608293 · doi:10.1007/s10508-017-1007-4 · record verified 2026-08-26

What was done

A meta-analysis was conducted on 26 studies containing 30 homosexual and 30 heterosexual comparison groups, comprising 7,140 homosexual and 12,837 heterosexual males over a 25-year span. The fraternal birth order effect was quantified using the Older Brothers Odds Ratio (OBOR), calculating the ratio of older brothers to other siblings (older sisters, younger brothers, younger sisters) in homosexual versus heterosexual males. Subgroup analyses evaluated effect differences by family size (small vs. large) and gender expression (12 feminine/transgender groups vs. 18 other homosexual groups).

What was found

The pooled Older Brothers Odds Ratio across all samples was 1.47 (p < .00001). Across the 30 individual group comparisons, OBOR was significantly >1.00 in 20 instances, non-significantly >1.00 in 9 instances, and non-significantly <1.00 in 1 instance. Subgroup comparisons revealed a significantly larger effect size in the 12 feminine or transgender homosexual groups compared to the remaining 18 homosexual groups. Family size had no detectable effect on the magnitude of the OBOR.

Why it matters

This study provides comprehensive meta-analytic confirmation that older brothers consistently correlate with increased odds of male homosexuality across diverse samples, ruling out family size as a confounding artifact and highlighting variation across gender identity subgroups.

Limits

The underlying studies were observational and retrospective, relying on self-reported sibling compositions and sexual orientation. Potential sampling biases across individual primary studies could not be eliminated, and causal biological mechanisms cannot be directly established from demographic sibling counts alone.

Cited by