‘The Night-time might be the Right Time’ for Cardiovascular Event Prediction

The ability to ascertain circadian variability of blood pressure (BP) via ambulatory recording devices has led investigators to evaluate the importance of the components of the 24-hour profile, including the overall 24-hour mean as well as the awake and sleep periods. For nearly 30 years, BP monitoring researchers have become fascinated with the shape of the 24-hour curve and have classified patients according to the ‘dip’ in the curve during sleep. Clinical terms such as dipper, non-dipper, and reverse dipper (or ‘riser’) have evolved to include those people who decline by at least 10% during sleep, those who decline less than 10% but still have some reduction compared to their awake BP, and those who actually have a higher BP during sleep and/or supine periods than while awake (1). These defined patterns of circadian BP variation have also been related to target organ involvement in cross-sectional studies (2,3) as well as to cardiovascular outcomes in longitudinal cohort analyses (4,5).

Two patterns of circadian BP variability have evolved as markers of excessive cardiovascular risk: those with higher than expected nocturnal BP (non-dippers and reverse dippers) (4-6) and those with a large early morning surge in BP, particularly in older people (7,8). In isolation, the findings from both patterns make pathophysiologic sense. A high nocturnal BP versus a lower nocturnal BP increases the BP burden in a given 24 hours. A surge in BP in the early morning versus a lack of a morning BP surge could theoretically increase shearing force on the arterial wall, increase myocardial wall tension and energy demand, and coincides with attenuating concentrations of antihypertensive therapy (9). However, there has always been a concern that conceptually these 2 types of BP profiles are at odds with each other – an elevated nocturnal BP with a non-dipper or reverse dipper profile should be associated with a small morning BP surge while a large morning BP surge could be associated with a dipper or extreme dipper profile in which the nocturnal BP might be relatively low.

In this issue of Hypertension, Verdecchia and colleagues have reported on new findings from their longstanding cohort analysis in 3012 patients from Italy in which patterns of circadian BP were evaluated in relation to major cardiovascular outcomes (10). As might be expected, a smaller daytime-nighttime dip was associated with a lesser surge in the morning BP. Furthermore, multivariate regression analyses showed that the nocturnal BP was the key predictor of cardiovascular outcomes and not the morning surge in BP. In addition, there was a graded response (as has been reported previously by this group) among the various circadian BP profiles (extreme dipper to reverse dipper) and cardiovascular event rates over a 15 to 20 year follow-up period. There was also a distinct separation between those individuals who had the lowest quartile of morning BP surge and the three higher quartiles and cardiovascular events, regardless of the method of calculating the morning surge in BP. In other words, using the ‘sleep-trough morning surge’ method, the investigators from Perugia found that the group with the lowest BP surge (≤ 19.5 mmHg) had a substantially worse prognosis than the group whose sleep-trough systolic BP surge was > 19.5 mmHg. In contrast, using the International Database on Ambulatory Blood Pressure in Relation to Cardiovascular Outcomes (IDACO), Li et al (7) showed negligible increases in cardiovascular risk if the morning surge in BP was < 20 mmHg, which also correlated with lower absolute systolic BP values at the peak of the morning surge. Of note, the analysis by Verdecchia et al (10) does not clarify what the absolute BP values were in association with the varying degrees of morning surge.

There are important methodological strengths of the study by Verdecchia and colleagues (10) in contrast to prior analyses of a similar nature. First, the investigators used a 4-component endpoint of major adverse cardiovascular events (MACE) that included cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, and hospitalized heart failure rather than a broad composite of these ‘hard’ endpoints mixed together with softer endpoints such as transient ischemic attack or angina pectoris. The MACE composite used in the present study is verifiable by an adjudication process, has greater clinical importance than a broader, mixed set of cardiovascular endpoints and is not diluted by less severe and more vague endpoints. Secondly, the analysis linking elevated nocturnal BP rather than the morning surge in BP to future MACE was performed with carefully considered clinical covariates to rule out relevant confounding. However, there are also limitations in this analysis – besides the race and ethnicity issue brought up by the authors, there is also the concern that anti-hypertensive interventions and their impact on 24-hour BP and variability over time cannot be ascertained. While the participants of this study were untreated at the time of the baseline ambulatory BP measurement, over the next 15 to 20 years that MACE were being accrued, it is not known whether the circadian BP abnormalities remained fixed or were altered due to aging or BP lowering agents.

It is of note that prior work has led us to believe that both the morning surge in BP (7-9) and a non-dipper profile (1, 4-6) are independent risk factors for the development of cerebrovascular and cardiac events. However, the data for a relation between nocturnal BP elevation and target organ involvement and future cardiovascular events may be more robust (11, 12). The relative importance of nighttime BP to daytime BP in association with cardiovascular events, including stroke in older patients (1, 4, 6, 12) may be due in part to the superior reproducibility of nocturnal BP compared to other time periods within a 24-hour profile, including the early morning surge period. In fact, we have previously shown that the early morning BP is perhaps the least reproducible clinical parameter that is derived from ambulatory BP monitoring while nocturnal BP is among the most reproducible (13). This remains a hypothetical conjecture since in the study by Verdecchia et al (10), it is not possible to assess the reproducibility of the various BP parameters as we only have the index ambulatory BP study to characterize circadian patterns in the cohort.

Progression from cross-sectional and prospective cohort studies to clinical trials using ambulatory BP values as the target of intervention is indeed the appropriate next step in the evolution of development of out-of-office monitoring of the BP for clinical decision-making. While a massive intervention trial does seem practically infeasible, as suggested by the authors of the index article (10), smaller outcome studies in high risk population are quite possible and should be an important part of present and future research.