Association of childhood intelligence with risk of coronary heart disease and stroke: findings from the Aberdeen Children of the 1950s cohort study

Objectives Associations of cognitive function assessed in adulthood with coronary heart disease (CHD) and stroke might reflect a causal effect or could be explained by residual confounding or a common underlying pathology (atherosclerosis) that links both declines in cognitive function and increased risk of cardiovascular disease (i.e. reverse causality). Our objective was to examine the association of childhood intelligence (assessed at an age when generalised atherosclerosis would be extremely unlikely) with risk of CHD and stroke in later life in a cohort of females and males on whom information on a wide range of potential confounding factors is available. Methods Cohort study of 11,125 individuals born in Aberdeen, Scotland, between 1950 and 1956, who had childhood intelligence measured at ages 7, 9, 11 and who have been followed up by linkage to hospital admissions and mortality data. Results The cohort contributed 264,672 person years of follow-up and over this time 93 females experienced CHD and 56 experienced a stroke; 264 males experienced CHD and 67 a stroke. There were inverse associations of childhood intelligence measured at all 3 ages (7, 9, 11 years) with both CHD and stroke, with some evidence that the association with intelligence assessed at age 11 was stronger than at younger ages. The magnitude of associations were similar for CHD and stroke. Adjustment for a range of potential confounding factors did not markedly attenuate the associations and there was evidence that the association was stronger in females than in males. For example, the age adjusted hazard ratio of a combined outcome of CHD and stroke per 1 standard deviation (SD; 1SD = 15 points) difference in intelligence score at age 11 in females was 0.52 (95% CI: 0.42, 0.64) and in males was 0.78 (95% CI: 0.68, 0.90), with adjustment for all potential confounding factors these became 0.60 (95% CI: 0.48, 0.76) and 0.84 (95% CI: 0.72, 0.98) respectively; P-value for interaction with gender in both models =0.002. Adjustment for educational attainment attenuated both associations to the null and removed evidence of a gender difference. Conclusions Our results suggest an association of intelligence in childhood with future CHD and stroke that is unlikely to be explained by reverse causality and was robust to adjustment for a wide range of confounding factors. This association appeared to be mediated by educational attainment. The gender difference seen in our study requires replication in other studies.