Birth Parameters and Cognitive Ability in Older Age: A Follow-Up Study of People Born 1921–1926

Abstract

Background: Birth weight throughout the normal range correlates with cognitive ability in childhood. Some studies suggest this may persist into adulthood, but it is unclear whether this continues into old age. Objective: To investigate the relationship between birth parameters (birth weight, birth length, placental weight and social class) and cognitive ability in old age. Methods: 128 community-dwelling volunteers born in one hospital in Edinburgh, UK, between 1921 and 1926 had birth details extracted from the records. All aged 75–81 underwent cognitive tests. The National Adult Reading Test (NART) was used to estimate prior ability. Tests of fluid cognitive ability showed positive intercorrelations, and therefore a general cognitive factor (g) was extracted using principal components analysis. Results: Estimated prior ability (NART) correlated with birth length (r = 0.25, p = 0.02), and non-significantly with birth weight (r = 0.15, p = 0.19). There were no statistically significant correlations between birth parameters and cognitive ability aged around 80 (birth weight r = 0.12, p = 0.27; length r = 0.19, p = 0.08). Placental weight was not associated with cognitive ability. There was a trend towards an association between social class at birth and prior cognitive ability (more deprivation predicting poorer performance) (rho = –0.17, p = 0.06). Social class was not significantly associated with cognitive ability in old age (rho = 0.07, p = 0.47), or with estimated cognitive change (rho = –0.03, p = 0.73). Conclusion: In a small sample of community-dwelling older people, an influence was detected from birth to earlier cognitive ability using a test of crystallized intelligence. However, this effect was not significant into old age. In this sample, social class at birth did not affect later cognitive ability. The effect size of early life influences on cognition is small, with birth parameters explaining about 1% of the variance. Research with larger samples should consider influences operating across the whole life course, and their relative importance over time.