To compare mortality associated with various blood pressure components in middle-aged men during up to 32 years of follow-up.Design
A prospective cohort study.Setting
We studied 3267 initially healthy men, aged 30–45 years, who participated in health check-ups from 1964 onwards.Main outcome measures
Cox regression was used to relate baseline blood pressure components to all-cause (n = 701) and cardiovascular disease (CVD) mortality (n = 325).Results
Systolic (SBP) and diastolic (DBP) blood pressures, pulse pressure and mean arterial pressure singly predicted CVD mortality. With SBP 160 mmHg and DBP 90 mmHg as cut-off values, four blood pressure subgroups were identified: normotension (n = 1919), isolated systolic hypertension (ISH, n = 17), isolated diastolic hypertension (IDH, n = 1013), and combined systolic and diastolic hypertension (SDH, n = 318). IDH was subdivided into IDH-1 with SBP 140–159 mmHg (n = 667) and IDH-2 with SBP less than 140 mmHg (n = 346). With normotension as reference, only SDH and IDH-1 predicted CVD mortality [relative risk (RR) 2.71, 95% confidence interval (CI) 2.00 to 3.66, and RR 1.39, 95% CI 1.04 to 1.87, respectively]. Risk with IDH-2 (RR 1.14, 95% CI 0.77 to 1.69) was not statistically significant. SDH and IDH-1, but not IDH-2, were also associated with increased all-cause mortality risk. Use of antihypertensive medication did not explain the results.Conclusion
These results demonstrate the often neglected role of SBP in predicting long-term CVD risk in middle-aged men. When SBP is less than 140 mmHg, IDH is not associated with significantly increased risk of mortality. Administrative guidelines, which affect population health, should also take due note of SBP.