In Utero Selection Echoes Across a Lifetime: Evidence From Historical Utah Populations

Past work in historical demography finds mixed evidence regarding the role of selective cohort mortality in utero in shaping that cohort's older-age mortality. Here we use, for the first time, high-quality individual-level historical data to test whether adults from birth cohorts with a presumed high level of selection in utero show greater than expected older age survival. We examine a frontier population in Utah, born from 1850 to 1920 and followed to 2020, and use the annual cohort sex ratio (M:F) at birth, a marker of cohort selection in utero against males.

We performed Cox proportional hazards analyses (n = 201 542) to predict the hazard of older-age male mortality (i.e., > 50 years) and used a two-stage correction strategy to control for selective mortality before older age (i.e., before age 50). Analyses also control for strong temporal patterns in cohort survival and individual covariates including month of birth and religious affiliation.

Consistent with our hypothesis, we find that males born to the least selected cohorts-as indicated by a high sex ratio at birth-show greater than expected hazard of older-age (i.e., > 50 years) mortality (p < 0.05). Interestingly, our exploration also shows a greater hazard of older-age mortality among females born to high sex ratio cohorts (p < 0.05).

The sex ratio at birth in 19th and early 20th century cohorts gauges important demographic selection processes that appear detectable well into older age.