UCLA

Aims

Whether isolated diastolic hypertension (IDH), as defined by the 2017 American College of Cardiology (ACC)/American Heart Association (AHA) guideline, is associated with cardiovascular disease (CVD) has been disputed. We aimed to further study the associations of IDH with (i) subclinical CVD in the form of coronary artery calcium (CAC), (ii) incident systolic hypertension, and (iii) CVD events.

Methods and results

We used multivariable-adjusted logistic and Cox regression to test whether IDH by 2017 ACC/AHA criteria (i.e. systolic blood pressure <130 mmHg and diastolic blood pressure ≥80 mmHg) was associated with the above outcomes in the Multi-Ethnic Study of Atherosclerosis (MESA). In a random-effects meta-analysis of the association between IDH and CVD events, we combined the MESA results with those from seven other previously published cohort studies. Among the 5104 MESA participants studied, 7.5% had IDH by the 2017 ACC/AHA criteria. There was no association between IDH and CAC [e.g. adjusted prevalence odds ratio for CAC >0 of 0.88 (95% CI 0.66, 1.17)]. Similarly, while IDH was associated with incident systolic hypertension, there was no statistically significant associations with incident CVD [hazard ratio 1.19 (95% CI 0.77, 1.84)] or death [hazard ratio 0.94 (95% CI 0.65, 1.36)] over 13 years in MESA. In a stratified meta-analysis of eight cohort studies (10 037 843 participants), the 2017 IDH definition was also not consistently associated with CVD and the relative magnitude of any potential association was noted to be numerically small [e.g. depending on inclusion criteria applied in the stratification, the adjusted hazard ratios ranged from 1.04 (95% CI 0.98, 1.10) to 1.09 (95% 1.03, 1.15)].

Conclusion

The lack of consistent excess in CAC or CVD suggests that emphasis on healthy lifestyle rather than drug therapy is sufficient among the millions of middle-aged or older adults who now meet the 2017 ACC/AHA criteria for IDH, though they require follow-up for incident systolic hypertension. These findings may not extrapolate to adults younger than 40 years, motivating further study in this age group.