Does parity and risk of dementia follow a U‐shaped relationship? Findings from the Cache County Study of Memory and Aging

Abstract

AbstractBackgroundRecent evidence from the UK Biobank of over 273,000 women indicates a U‐shaped relationship between parity and risk of dementia, with nulliparity and grand multiparity showing increased risk of dementia compared to women having 1‐3 births after adjustment for important sociodemographic and health history factors, but not Apolipoprotein E (ApoE) polymorphism. ApoE4 genotype, a known risk factor for Alzheimer’s disease and Alzheimer’s disease related dementia (AD/ADRD), has been shown to be related to higher potential female fertility due to increased concentrations of progesterone. Our objective was to determine whether a similar U‐shaped relationship is present in the population based Cache County Study of Memory in Aging (CCSMA) cohort, after controlling for important confounding factors including ApoE genotype.MethodsCCSMA was a longitudinal study that investigated risk factors contributing to AD/ADRD and other dementias. CCSMA enrolled 90% (N = 5,092, n = 2928 women) of the Cache County, Utah permanent resident population aged 65 and older as of January 1, 1995. Participants were followed for four triennial waves for AD/ADRD ascertainment. Prior number of live births for CCSMA women was assessed via self‐report and validated via linked Utah Population Database records (99% successful linkage). Adjusted risk ratios using Poisson regression with robust variance were calculated to assess whether parity was associated with later dementia risk, adjusting for important sociodemographic and health history factors (Table 1).ResultsAmong the total sample, 5.6% were nulliparous, 18.4% had 1‐2 births, 54% had 3‐5 births, and 22% had 6+ births. Over the course of the study, 12% were diagnosed with AD and 7% with ADRD. Compared to women with 3‐5 births, women with zero, 1‐2, and ≥6 live births had an aRR of 1.16 (95% CI: 0.80, 1.69), 1.00 (95% CI: 0.76, 1.32), and 1.21 (95% CI: 0.96, 1.53) (Figure 1), respectively. Results were similar for ADRD.ConclusionOur results support a U‐shaped relationship between parity and dementia risk. Whether the relationship between parity and dementia risk is being driven by social and behavioral factors involved in parenting or biological factors associated with childbearing (e.g., total endogenous estrogen exposure) warrants further investigation.