Blood Pressure Values During the Year Prior to Events is More Meaningful in Assessing the J-Curve Phenomenon Versus Mean On-Treatment Blood Pressure

21st European Meeting on Hypertension and Cardiovascular Prevention

Xuezhong Zhang, MD, Clinical Trials and Research Center, Chinese Hypertension League, Beijing, China, presented a new more sensitive approach to assessing the J-curve phenomenon, which shows that the nadir blood pressure (BP) during the 2 months prior to the cardiovascular (CV) event is most meaningful.

A J-curve effect, in which the risk of CV events increases when BP is below 120/70 mmHg, has been shown in several clinical trials of hypertension reduction. Although the suggestion of this effect is based on a limited number of subjects and outcomes and its appearance may be influenced by the method used to report BP rates (eg, mean, last observation, or all observations prior to events), the possibility of such a phenomenon has contributed to the debate concerning how aggressive clinicians should be in pursuing blood pressure goals.

At the European Society for Hypertension’s 2010 meeting, Prof. Zhang presented data showing that when analyzing the relationship between BP levels and CV events, an approach using moving events per 1000 patient observations (MEPPO) is more sensitive than using mean on-treatment BP. He explained that when using the MEPPO approach each patient has not just one BP observation (average on-treatment) but as many observations as the number of visits in which BP is measured. Event information is collected and event occurrence or absence is referenced to the BP measurement prior to the event. In addition each observation is moved forward in 1 mmHg steps so that it is entered 10 times into 10 different consecutive 10 mmHg ranges, thus dramatically increasing the data analyzed.

At this year’s meeting, using data from the FEVER (Felodipine Event Reduction: NCT01136863) Trial, Prof. Zhang reported on a new version of MEPPO in which only BP observations closest to the event are included. In the FEVER Trial patients (n = 9,711) were stratified by 10 mmHg increments according to their mean on-treatment BP and CV events in each group were measured over 60 months, during which there were a total of 159,844 blood pressure observations. Each observation came from 3 systolic and 3 diastolic readings over 30 minutes. In this new analysis, only the BPs 1, 6, and 12 months before the event are included.

The new approach was more sensitive for J-curve assessment, because it uses an equal number of observations in all patients. The J-curve is evident when using BP observations during the one year prior to the CV events. When analyzed in this way there is a steep decrease in incidence of all outcomes (all CV events, strokes, cardiac events, CV and all deaths) to reach a rather flat nadir of SBP (Table 1).

Table 1. J-curves Nadir in the During the One Year Prior to the Events