Blood pressure and cognition in the Aberdeen 1936 birth cohort

The relationship between blood pressure and cognition in old age remains unclear. Some data indicate that elevated blood pressure causes cognitive deficits whilst others show that lower early life mental ability predicts high blood pressure in adulthood. Longitudinal studies in which mental ability earlier in life is known are needed to clarify the relationship.

To measure the effect of blood pressure on cognition in late adulthood after adjusting for early life mental ability.

The sample comprised survivors of the 1947 Scottish Mental Survey who had validated IQ scores at age 11. Six cognitive tests - Mini-Mental State Examination, Raven's Progressive Matrices, Rey Auditory Verbal Learning Test, Uses of Common Objects Test, Digit Symbol Test and Block Design - were administered at ages 64, 66 and 68 years. Sitting and standing blood pressure was also measured at each wave of follow-up. Mixed general linear models were constructed with each cognitive test treated as a repeated measure of an underlying cognitive trait and with wave of testing also set as a repeated measure. Mental ability at age 11 was entered as a covariate.

504 participants were tested at wave 1 with 368 returning at wave 2 and 300 at wave 3. Age 11 mental ability did not predict any of the blood pressure measures. There were several significant associations between blood pressure variables and cognitive test scores in univariate models. After adjusting for significant effects of wave of testing, type of cognitive test, the interaction between these, age 11 mental ability, age, gender and occupation in a multivariate model, the main effect of BP trait was no longer significant (p = 0.44) nor its effect over time (p = 0.26), though there was a significant interaction between blood pressure trait (BP) and test type with a distinctly negative effect of BP on Auditory Verbal Learning Test (p = 0.007, -0.13 points per mm Hg higher, 95% CI -0.22 to -0.033).

The effects of blood pressure on cognition in old age are finely nuanced. Multivariate repeated measures models reveal a differential effect of blood pressure on verbal recall.