–Meta-analysis supports a systolic blood pressure target of 120 mm Hg
A new systematic review and network meta-analysis lends support to more aggressive blood pressure targets.
In recent years US guidelines have recommended more relaxed systolic blood pressure targets, ranging from 140 mm Hg to 150 mm Hg, for people with hypertension. But the results of the recent SPRINT trial led to renewed calls for a more aggressive approach with a target for most people of 120 mm Hg.
Yet questions about the interpretation of SPRINT led some hypertension experts to question how its results should be applied to clinical practice.
Against that background, this new paper, published in JAMA Cardiology, supports the lower systolic blood pressure target of 120 mmHg. Jiang He, MD, of Tulane University, and colleagues analyzed data from 144,220 patients participating in 42 randomized trials. They found “significant and linear associations between average achieved systolic BP and the risk of cardiovascular disease and all-cause death.” People who achieved a systolic BP level of 120 to 124 mm Hg had the lowest risk for cardiovascular disease and all-cause mortality.
When compared to people who achieved a systolic blood pressure of 130 to 134 mm Hg, people who reached the 120 to 124 mm Hg level had a 29% reduction in cardiovascular disease and a 27% reduction in all-cause mortality. The relative risk reduction were even larger in comparisons with groups with higher achieved systolic blood pressure levels.
The authors wrote that their results were “generalizable to populations at large with hypertension” and that they “do not support the existence of a J-shaped association between achieved SBP and the risk of CVD and all-cause mortality.” But they also acknowledged that the benefits may come at the cost of an increased risk of adverse events, including hypotension, electrolyte abnormalities, and kidney injury. They also warned that it is unclear whether their results will apply to patients with diabetes. “The outcomes of intensive SBP reduction for patients with diabetes warrant further exploration.”
In an accompanying editorial Clyde Yancy, MD, and Robert Bonow, MD, both of Northwestern University, wrote that the paper provides “provocative evidence that lower is better and likely so in all cohorts with hypertension.” They noted that it is important to remain aware of the potential hazards of aggressive treatment, “particularly in the elderly,” but they note that the adverse effects “have not yet emerged to be more important than the benefits.”
Asked to comment on the paper, Yale’s Harlan Krumholz, MD, pointed out that the trials included in the analysis did not actually test the different systolic BP targets analyzed in the analysis. “It seems to me that they might be comparing responders to non-responders since no trial randomized people to these different levels – and so the results of the different blood pressure levels needs to be interpreted with caution.”
Sverre Kjeldsen, MD, PhD, of Oslo, Norway said the paper was “a strong and very elegant meta-analysis.” He said that along with previous analyses, the new meta-analysis confirms that it is safe to lower systolic BP down to 120 mm Hg. As a practical measure he said that the systolic BP target “should be between 120 and 140 mm Hg. If we aim at 130 mm Hg, we should almost always be able to hit between 120 and 140 mm Hg,” though the target should probably be higher in elderly people.
Kjeldsen has been a leading critic of the SPRINT trial, calling attention in particular to the questions and uncertainties about the precise methods used to measure blood pressure in the trial. The meta-analysis did not change his view of SPRINT, he said. Given the problems with blood pressure measurement in SPRINT, he said the data from SPRINT may have been used incorrectly in the meta-analysis.
Sanjay Kaul (Cedars-Sinai) sent the following comment:
“Network meta-analyses and indirect comparisons can only be useful if the trial or patient characteristics are similar (assumption of similarity), the observed effects are sufficiently homogeneous (assumption of homogeneity), and there are no major discrepancy between the direct and indirect evidence (assumption of consistency). In general these assumptions of ‘exchangeability’ are difficult to fully verify on the basis of available data. The findings of indirect comparisons and network meta-analyses therefore usually allow for less certainty in conclusions than the findings of appropriate pairwise meta-analyses of head-to-head trials. So, while the results of the current report might ‘sharpen our clarity’, I don’t think they rise to the level of evidence required to formulate public policy or guide clinical practice. A case in point: the favorable results in support of cardiovascular safety of naproxen derived from a network meta-analysis of NSAIDs were not replicated in a dedicated cardiovascular outcome trial designed specifically to address this (PRECISION trial)!”
It’s good that this analysis chose research that had standardized BP measurement technique – more-or-less.
One concern I have is that the technique looks only at BP while seated and relaxed after five to fifteen minutes. What happens during the rest of each subject’s actual daily life?
Another concern is the likelihood of bias – innocent and unconscious or not – influencing criteria and therefore outcome in any meta-analysis.
“I told you so!”
As regards the ideal blood pressure control, this author had commented in the past about the confusing and contradictory clinical trial results and meta-analysis conclusions.
It is worth reminding the academicians and the practicing physicians one comment made on November 10, 2015 in NEJM on the SPRINT trial:
“There is a wide spectrum of genetically determined variability in patients in the expression of B.P., blood sugar and lipids numbers and their response to individual drugs and benefit to their treatments. Intraday, day-to-day, and visit-to-visit B.P. variability, which is considered to have higher risk for complications than sustained hypertension, has been recognized in many patients thanks to home blood pressure monitoring in clinical practice.
Marked variability in patient phenotype due to heterogeneity in its pathophysiology, different stages of the disease process and multiple other patient factors including co-morbidities, will affect the outcome analysis of chronic diseases.
After 40 years of experience following the same patients for decades and listening to different experts and guidelines, it has been my advice to control the B.P. in the range of 110-130/75-85 mm Hg, especially for the younger population, without causing adverse side effects, to reduce the risk of cardiovascular events and renal failure.
De intensification of B.P. control is advised for older patients, as we have done for HbA1c and LDL cholesterol now.”