In a large, national cohort study of U.S. veterans with non–dialysis dependent chronic kidney disease, lower systolic and diastolic blood pressures were associated with lower mortality rates – but only when the diastolic value was higher than about 70 mm Hg.
In addition, mortality rates were significantly increased among those patients with "ideal" blood pressure values (less than 130/80 mm Hg), "because of the inclusion of patients with low SBP and DBP," reported Dr. Csaba P. Kovesdy, chief of nephrology at the Memphis Veterans Affairs Medical Center, and his associates. The study was published in the Annals of Internal Medicine on Aug. 19 (Ann. Intern. Med. 2013;159:233-42).
Dr. Csaba P. Kovesdy
The results indicate that current guidelines for patients with chronic kidney disease (CKD), which recommend a systolic blood pressure (SBP) of 130 mm Hg or lower "at the expense of lowering DBP [diastolic blood pressure] to less than approximately 70 mm Hg," may be harmful, they concluded. However, one of the limitations of the study was that it was an observational study and cannot establish a causal association, so "clinical trials are needed to inform us about the ideal BP target for antihypertensive therapy in patients with CKD," they added.
Using more than 18 million BP readings, the study evaluated the association of SBP and DBP values separately and SBP/DBP combinations on all-cause mortality in almost 652,000 U.S. veterans with CKD, who were not dependent on dialysis, between 2005 and 2012. Their mean age was 74 years, most were male (97%), 88% were white, 9% were black, 43% had coronary artery disease, and 43% had diabetes. The mean SBP values at baseline were 135 mm Hg while the mean DBP was 72 mm Hg; the mean glomerular filtration rate (GFR) was 50.4 mL/min per 1.73 m2. The study looked at 96 different SBP/DBP combinations. During the time period of the study, 238,640 patients died.
They identified a U-shaped curve when analyzing mortality with SBP and DBP separately, "with both lower and higher levels showing a substantial and statistically significant association" with mortality risk. Based on the adjusted hazard ratios for the combinations of SBP and DBP, the lowest mortality rates were associated with blood pressures of 130-139/90-99 mm Hg, and 130-159/70-89 mm Hg, adjusted for factors that included age, sex, race, diabetes, and cardiovascular and cerebrovascular disease, age and medication use).
But combinations of lower SBP and DBP values "were associated with relatively lower mortality rates only if the lower DBP component was greater than approximately 70 mm Hg," they said.
When evaluating risk based on JNC 7 (Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure) categories,they found that that those with stage 1 hypertension (SBP of 140-159 mm Hg or DBP of 90-99 mm Hg) were associated with the lowest mortality rates, while those in the normal category (an SBP lower than 120 and a DBP below 80) had the highest mortality rates," results that were independent of confounding factors and were statistically significant.
The authors described an elevated SBP combined with a low DBP, which is common in CKD patients, as "an especially problematic BP pattern," they said, pointing out that 33% of the patients had an SBP greater than 140 mm Hg and a DBP less than 70 mm Hg at some point during the study period.
The study strengths included the large size and the representation of the U.S. veterans’ population, but the limitations included the mostly male population and the observational design of the study, so more studies are needed, the authors said. "Until such trials become available, low BP should be regarded as potentially deleterious in this patient population, and we suggest caution in lowering BP to less than what has been demonstrated as beneficial in randomized controlled trials," they concluded.
Dr. Kovesdy, professor of medicine at University of Tennessee Health Science Center, Memphis, disclosed having received grants from the National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, and nonfinancial support from the Department of Veterans Affairs while the study was conducted. Four authors had no disclosures, and one author disclosed having received NIH grants during the study. The remaining two authors disclosed having received research grants or personal fees from different pharmaceutical companies.