Abstract:

For roughly 100 years, there has been a widely recognized relationship between high blood pressure (hypertension) and cardiovascular disease. Currently, the prevailing wisdom concerning blood pressure is "the lower, the better." The Joint National Commitee on Hypertension sets blood pressure cut-points independent of age and gender. An analysis of the Framingham Study data suggests both of these approaches are wrong. A multiple logistic spline regression model, which incorporates age and gender and does not assume a simple linear relationship between blood pressure and risk, yields results that may help physicians treat hypertensive patients more intelligently.