Associations between parenthood and dementia in men and women: biology or confounding?

Abstract

Background: High parity and extremes of age at first birth have been linked with increased dementia risk in women, with exposure to pregnancy-associated physiological changes proposed as an explanation. However, confounding by socioeconomic and lifestyle factors could also produce such associations, whereby men would share similar patterns of association. We investigated whether these associations hold for both sexes. Methods: In a cohort study including all women (N = 2,222,638) and men (N = 2,141,002) ≥ 40 years of age in 1994–2017 in Denmark, we used Cox regression to evaluate associations between number of children, age at first birth, and dementia risk separately for women and men. Results: During follow-up, 81,413 women and 53,568 men (median age at diagnosis, 83.3 and 80.3 years, respectively) developed dementia. Compared with having one child, having two or more children was associated with modest decreases in overall dementia risk in both sexes (hazard ratio [HR] range 0.82–0.91, Pdifference men vs. women = 0.07). Although the associations between childlessness and overall dementia risk differed statistically for men and women, the association magnitudes differed only slightly (HRmen 1.04, 95% confidence interval [CI] 1.01–1.06; HRwomen 0.99, 95% CI 0.97–1.01; P = 0.002). Associations between age at becoming a parent and overall dementia were also similar for women and men, with the exception of older (≥ 40 years) first-time parents (HRmen 1.00, 95% CI 0.96–1.05; HRwomen 0.92, 95% CI 0.86–0.98; P = 0.01). With few exceptions, sub-analyses by dementia subtype and timing of onset also revealed similar patterns and effect magnitudes for women and men. Conclusions: Associations between number of children, age at becoming a parent, and dementia risk were similar for both sexes. Lifestyle and socioeconomic factors are more likely to explain the observed associations than normal pregnancy-related physiological changes.