Cognitive development and cortisol patterns in mid-life: findings from a British birth cohort

Life-course associations among cortisol, cognitive development and educational attainment in the general population are not well understood. Using the 1958 British birth cohort, our aim was to establish whether cortisol patterns at age 45y are associated with childhood cognition and qualification level by adulthood. We measured salivary cortisol in 6527 individuals, 45min after waking (t(1)) and 3h later (t(2)). To identify lack of morning cortisol peak and diurnal rhythm, we defined groups with: (a) t(1) cortisol in the bottom 5% of the distribution, or (b) 'flat't(1)-t(2) cortisol. Data on cognitive tests at ages 7, 11 and 16y and educational level were used. All childhood cognitive tests (maths, reading, verbal and non-verbal ability) were inversely associated, although not always significantly, with low t(1) and flat t(1)-t(2) cortisol. For example, at age 11 for males, a standard deviation (SD) increase in maths score was associated with a 28% decreased odds for lowest t(1) cortisol, and with a 13% decreased odds of flat t(1)-t(2) cortisol. Associations for lowest t(1) and flat t(1)-t(2) cortisol were attenuated after adjustment for qualification level at 33y among males, although adjustment for childhood socio-economic position had little effect. Weaker associations for lowest t(1) cortisol among females were either unchanged or strengthened after adjustment for qualification level. Our results for males, but less so for females, are compatible with a causal relationship in either direction, namely from cortisol to cognitive ability or vice versa.