The ability of diastolic, systolic, mean arterial, and pulse pressures to predict cardiovascular disease (CVD) morality has not been assessed for persons with prehypertension (diastolic pressure 80 to 89 mm Hg or systolic pressure 120 to 139 mm Hg).Methods:
Cox's regression analyses were conducted using 4849 subjects aged 33 to 87 years from the First National Health and Nutrition Examination Survey Epidemiologic Follow-up Study. A correction was made for the regression-dilution bias.Results:
There were 327 cardiovascular disease and 258 coronary heart-disease deaths during an average follow-up of 8.6 years. For nonelderly prehypertensives, systolic blood pressure (BP) was a stronger predictor than diastolic BP. The multivariate single predictor hazard ratios (95% confidence interval) for CVD mortality were 1.43 (0.65-3.14) and 2.11 (1.28-3.49), for a 10 mm Hg increment diastolic and systolic BP, respectively. For elderly prehypertensives, it was reversed. The equivalent results were 1.53 (1.10-2.13) and 1.25 (0.89-1.60), respectively. For nonelderly hypertensives, diastolic BP was a stronger predictor than systolic BP, and for elderly hypertensives it was reversed. Diastolic and systolic BP provided as much as or more predictive information than pulse and mean arterial pressure in all analyses.Conclusions:
For nonelderly prehypertensives, systolic BP was the strongest predictor, whereas for hypertensives the strongest predictor was diastolic BP. For elderly subjects this pattern was reversed.