Author + information
- John W. McEvoy, MB BCh BAO, MEHP, MHS∗ (, )
- Yuan Chen, MS and
- Elizabeth Selvin, PhD, MPH
- ↵∗Johns Hopkins Ciccarone Center for the Prevention of Heart Disease, Division of Cardiology, Johns Hopkins University School of Medicine, Welch Center for Prevention Epidemiology and Clinical Research at the Johns Hopkins Bloomberg School of Public Health, 600 North Wolfe Street, Blalock 524C, Baltimore, Maryland 21287
We appreciate the interest in our recent paper in the Journal (1). Dr. Marcus and colleagues express concerns that findings from our observational study should not be extrapolated to clinical care. We agree that our results cannot definitively prove causal effects; as with all observational studies, the possibility of residual confounding cannot be eliminated. We also agree with Dr. Marcus and colleagues that analyses from recent randomized trials (e.g., SPRINT [Systolic Blood Pressure Intervention Trial] and HOPE [Heart Outcomes Prevention Evaluation]-3) can help confirm our observational data and inform how our findings can be translated into clinical practice. We note that the former limitation and the latter suggestion were included in our original paper.
Nonetheless, placed in the context of the evidence to date, we maintain that our observational results are robust and compelling. For example, they meet the following Bradford Hill Criteria: 1) for strength, we found a 50% increased risk of coronary heart disease (CHD) (p value of <0.001) among those with diastolic blood pressure (DBP) <60 mm Hg after rigorous adjustment, results that make residual confounding as the sole explanation unlikely; 2) for consistency, our results are consistent with a wealth of prior data; 3) for temporality, we found an association of low DBP with cross-sectional elevations in high-sensitivity troponin and with temporal change in troponin over the following 6 years; 4) for biological gradient, the lower the DBP category is the stronger our findings are for myocardial damage and CHD; 5) for plausibility, our results are plausible given that we know coronary perfusion depends on diastolic driving pressure; and 6) for coherence, we showed for the first time coherence between epidemiologic findings (e.g., CHD events) and laboratory testing for ischemia (high-sensitivity troponin).
Consistent with this, our conclusions are supported by results from the HOT (Hypertension Optimal Treatment) trial, the only randomized trial evaluating specific DBP targets (2). The HOT trial demonstrated that, among participants with ischemic heart disease, those treated to a DBP target of 80 mm Hg had a higher rate of myocardial infarction (8.3 per 1,000 patient-years) than those treated to a target of 85 mm Hg (6.8 pr 1,000 patient-years; relative risk: 1.22) (3).
Drs. Battista Danzi and Cuspidi suggest that heart rate may be an important factor in the association between low DBP and coronary ischemia. This important suggestion was also raised by Dr. Namasivayam and esteemed colleagues (4). While we do not have complex data such as systolic pressure-time loading, heart rate is available in the ARIC (Atherosclerosis Risk In Communities) study dataset. Thus, we performed additional analyses adding heart rate to our models. As can be seen in Table 1, the addition of heart rate did not appreciably alter our results, suggesting that the association between low DBP and adverse coronary outcomes is not mediated by heart rate.
Finally, the letter by Dr. Spence is a sobering reminder that, for some elderly patients, DBP measured by an arm cuff may actually be higher than central aortic DBP (the latter being the most relevant for coronary perfusion), a phenomenon that further helps to explain our results.
Please note: Dr. Selvin has served on the advisory board for Roche Diagnostics. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2017 American College of Cardiology Foundation