ABSTRACT
In this chapter we focus on the design features and results of a series of three investigations of relationships between blood pressure and cognitive functioning employing the Framingham Heart Study population (Elias, Wolf, D’Agostino, Cobb, & White, 1993; Farmer et al., 1990; Farmer, White, Abbott, et al. 1987). The results of the Elias et al. (1993) study indicated that blood pressure level and chronicity of hypertension were inversely related to measures of neuropsychological test performance, particularly those measures sensitive to memory processes and learning dependent on memory. This was true even though blood pressure and chronicity of hypertension were assessed 12 to 14 years prior to neuropsychological testing. By reviewing the relevant literature, by comparing the Elias et al. (1993) study with two previous studies of the same population (Farmer et al., 1990; Farmer, White, Kittner, et al., 1987), and by performing additional data analysis for the third, we illustrate design features that may have influenced the outcome of these studies. Such features include using multiple versus single or few measures of blood pressure on one or several occasions, time interval between blood pressure and neuropsychological assessments, medication status of the study participants, and analytic techniques that make use of the full range of test scores (linear regression models) rather than discrete categories (logistic regression models). Finally we discuss the limitations of clinical neuropsychological tests with regard to discriminating among cognitive processes such as attention, memory, and learning, and we discuss several explanatory models relating levels of blood pressure to levels of cognitive functioning.
