Is hypertension linked to age-related cognitive decline in executive function? Comparison of different statistics analyses

Abstract

Background: To examine the validity of Bucur and Madden (2010) proposal that BP-related cognitive decline is more pronounced for executive function than for elementary perceptual speed tasks, 6 years longitudinal data of BP and cognitive performances was examined by three types of statistical analyses. Methods: Developmental change ratio for 6 years in executive function task was calculated individually for participants (over 60 yrs) and classified into Maintain (N = 36; mean regression coefficient was -3.4) and Decline (N = 62; mean regression coefficient was -12.8) groups. In study 1, performances for both groups in SBP and DBP compared by t-test. In study 2, number of participants of normal BP and high BP in Decline and Maintain groups was compared by Chi2 test. In Study 3, data of executive function tasks was analyzed by covariance analysis with age and gender as covariates and multiple regression analysis. Results: Statistical analysis in Study1 showed a slight group difference in DBP but not in SBD. In Study 2, number of participants of normal BP was significantly larger than that of high BP. In Study 3, statistical results showed significant effect only in age but not in BP group. Conclusions: The validity of proposal remains opaque because it depends on the type of used statistics. It became clear that age contributes strongly in the relation between BP and cognitive performance in elderly. However, several limitations remain in BP measurement control (e.g., measurement time zone and number of times).