Race/ethnicity, completed parity, and women's later-life health in the Health and Retirement Study.

OBJECTIVES: This study examines associations between completed parity, including childlessness (zero parity), and a range of medical conditions (individually and cumulatively) and self-rated health, including differences associated with race/ethnicity-related factors.

METHODS: We conduct correlational and multivariable regression analyses using a historical HRS/AHEAD sample of ever-married, non-Hispanic Black and White women born between 1920 and 1941. Multivariable models adjust for early-life-course social determinants of health, including contextual birthplace exposure to Jim Crow-era structural racial inequalities. We focus on childlessness/zero parity by using zero-inflated count regression methods that statistically identify women with zero probability of having had a live birth (a latent "always-zero" group).

RESULTS: Black women had three times the odds of "always-zero" group membership and, if not in that group, an expected parity 11% higher than White women. In correlational analyses, the predicted probability of "always-zero" group membership (Predicted Infecundity Probability or PIP)-rather than reported childlessness status-and high parity were associated with high-fatality medical conditions and self-rated fair/poor health. In multivariable models, PIP and high parity were positively associated with number of conditions, while PIP and number of conditions were positively associated with self-rated fair/poor health. Net of parity, consistent with weathering, younger Black women had a higher-than-expected number of medical conditions relative to all women.

DISCUSSION: In the context of historical structural racial inequality, race/ethnicity-related differences in childlessness and high parity were strongly associated with later-life, post-reproductive health disparities. Women's number of births was associated with the accumulation of medical conditions in a U-shaped pattern.