Estimating the association between parity and mid- to later-life cognitive function in rural South Africa: Evidence from “Health and Aging in Africa: A Longitudinal Study of an INDEPTH Community in South Africa” (HAALSI)

Abstract

BackgroundThe burden of cognitive decline is projected to rise over the next decades, particularly in low- and middle-income countries, contributing to economic hardship, morbidity, and mortality. Reproductive life-history, including parity, may be associated with later-life cognitive ability. Using data from the population-representative “Health and Aging in Africa: A Longitudinal Study of an INDEPTH Community in South Africa” (HAALSI) cohort, we estimated the association between parity and mid- to later-life global cognitive function.MethodData were from in-person interviews with 2242 men and 2604 women aged ≥40 in the rural Agincourt sub-district, Mpumalanga province, South Africa in 2014/15. Cognitive function was assessed as time orientation and immediate and delayed recall of a 10-word list. Number of living children was self-reported. We used multivariable linear regression with parity (number of children: 0, 1-2, 3-4, 5-9, 10+) as the categorical predictor and z-standardized cognitive function as the outcome. Analyses were stratified by sex/gender, controlling for age, education, and literacy.ResultAmong men, there was a U-shaped relationship between parity and cognitive function (p<0.001), whereby men with 3-4 children had the highest mean z-standardized cognitive scores (mean: 0.244, SD: 0.984, Table 1, Figure 1). Women also had a U-shaped relationship between parity and cognitive function (p = 0.0254), whereby women with 5-9 children had the highest mean z-standardized cognitive scores (mean: -0.017, SD: 0.935, Table 2, Figure 1). Among men, adjusted regression models showed that parity was associated with higher cognitive function, across all categories compared to no children (Estimates [95% CI]: 1-2: 0.32 [0.17-0.48], 3-4: 0.50 [0.35-0.65], 5-9: 0.50 [0.36-0.74], 10+: 0.55 [0.36-0.74]; Table 3). Among women, only the 5-9 children group (Estimate [95% CI]: 0.16 [0.00-0.31]) had cognitive scores that were significantly different from having no children (Table 3).ConclusionData from this representative sample in rural South Africa indicate that the relationship between parity and later life cognition varies by sex/gender, with having children being associated with greater cognitive performance in men, compared to women. This sex/gender difference may be due to sex and gender differences in biological and social roles involved in childbearing and rearing.