U-shaped association between mortality and blood pressure in a thirteen-year prospective study

The present study was performed to investigate the predictive value of casual blood pressure recordings for mortality in a geographically defined population (n = 1917), for which the primary health care services are responsible. Isotonic regression of blood pressure on age was used to define groups with low, medium and high blood pressure. In the age group 40 to 69 years mortality turned out to be a U-shaped function of blood pressure. In those aged 70 years and over blood pressure appeared to be of less importance as a risk indicator of death. The study has shown that a limited number of subjects in a defined population can be utilized in epidemiological studies of mortality as a function of blood pressure if an appropriate statistical method is used in the analyses, thereby justifying our model for studies in primary health care.