PROGNOSTIC VALUE OF AMBULATORY OR HOME BLOOD PRESSURE MEASUREMENT COMPARED WITH OFFICE BLOOD PRESSURE MEASUREMENT IN A GENERAL ITALIAN POPULATION
Measurement of office blood pressure (OBP) continues to be used to make therapeutic decisions in patients with hypertension despite the fact that recent studies have suggested that ambulatory blood pressure (ABP) monitoring is superior to OBP measurement in predicting end‐organ damage and future cardiovascular (CV) events. Furthermore, two recent prospective studies, one from Japan and one from France, have shown that home blood pressure (HBP) monitoring is superior to OBP measurement in predicting fatal CV events. Information on the general population, however, including both hypertensive and nonhypertensive individuals, is currently lacking. The Pressioni Arteriose Monitorate e Loro Associazioni (PAMELA) Study sought to determine whether ABP or HBP measurement was superior to conventional OBP measurement in predicting future CV or all‐cause mortality in a representative cross‐sectional population of a small northern Italian town.
Between 1990 and 1993, 3200 citizens between 25 and 74 years of age were randomly selected from the general population of Monza, Italy, a town outside of Milan. With 64% of those interviewed agreeing to participate, participants were similar to nonparticipants with respect to demographics and baseline CV risk factors. OBP, HPB and ABP measurements were obtained on every participant. OBP consisted of three sphygmomanometric measurements performed with the subject in the seated position 10 minutes after the start of the initial study visit. ABP was obtained with a commercially available 24‐hour oscillometric ABP monitor (Model 90207, Spacelabs Medical, Issaquah, WA) set to take readings 20 minutes apart through the entire 24‐hour period. HBP was obtained using a commercially available semiautomatic device (Model HP 5331, Royal Philips Electronics, Amsterdam, The Netherlands); subjects were instructed to take two home readings, at 7 a.m. and 7 p.m., on the same day as the ABP device was worn. All subjects underwent all three methods of blood pressure (BP) measurement. Additional information was obtained concerning the baseline presence of other CV risk factors (smoking, overweight, cholesterol, or diabetes) or previous CV events.
The three OBP readings and two HBP readings were averaged separately for each subject. ABP readings were edited for artifacts according to prespecified criteria and averaged for the 24‐hour period, as well as for daytime (7 a.m.–11 p.m.) and nighttime (11 p.m.–7 a.m.). Cox proportional hazard models were fitted to explore the relationship between BP obtained by each method and the relative risk of CV or all‐cause death. For each model, the likelihood ratio test was used to determine whether the method of BP monitoring explained a significant portion of the risk of death. Letters were sent out throughout the study requesting information on the subjects' vital status, with death certificates required to verify the cause of death.
During the average follow‐up of 131 months there were 186 deaths (9.1% of all subjects), with only 56 deaths (2.1% of all subjects) due to CV causes. Subjects who died of CV disease were older, more likely to be male or overweight, or to have a history of hypertension, hypercholesterolemia, diabetes, or previous CV disease. In the cohort as a whole, average OBP was the highest, and ABP the lowest, with HBP intermediate. BP measured by all three methods was higher in subjects who died compared with those who were alive at the end of the study. The difference between average systolic OBP and 24‐hour mean systolic ABP was approximately 11 mm Hg in subjects who remained alive and about 25 mm Hg in subjects who died. The difference between systolic OBP and systolic HBP was approximately 8 mm Hg in subjects who remained alive and about 11 mm Hg in subjects who died. Differences in diastolic BP between methods were also significant, although of lesser magnitude. Night ABP was considerably lower than day or 24‐hour ABP.
Regardless of the method of measurement used, the relationship between BP and risk of death was less for diastolic than systolic BP. Although not highlighted by the authors, goodness of fit between BP and risk of CV or all‐cause death was highest with systolic HBP. The risk of CV and all‐cause mortality increased more steeply from OBP to HBP, day ABP, 24‐hour mean ABP, and night ABP. In other words, for each 10 mm Hg increase in OBP, there was a correspondingly greater attributable risk of death than with a 10 mm Hg increase in BP by other methods. The risk of death showed no correlation with heart rate measured by any method. In a multivariate model, the addition of HBP or night ABP to OBP improved the goodness of fit significantly, but the addition of 24‐hour or day ABP to OBP did not improve the model's predictive ability.
The authors conclude that over the 11 years of follow‐up, OBP, HBP, and ABP were all higher in subjects who died of CV or other causes than in people who survived, but that there was no prognostic superiority of either ABP or HBP over OBP measurement. The addition of HBP or night ABP to OBP did add some predictive value.—Sega R, Facchetti R, Bombelli M, et al. Prognostic value of ambulatory and home blood pressures compared with office blood pressure in the general population: follow‐up results from the Pressioni Arteriose Monitorate e Loro Associazioni (PAMELA) study. Circulation. 2005;111:1777–1783.
COMMENT
It is increasingly recognized that OBP measurements correlate poorly with BP measurements in other settings. This has led to an increased interest in the importance of BP measurements in out‐of‐office settings. Potential benefits of ABP monitoring include identification of white coat hypertension, absence of observer bias, and the ability to predict end‐organ damage and future CV events more accurately than OBP measurement. HBP monitoring may provide similar benefits at lower cost and with the ability to monitor the response to antihypertensive therapy over time. However, HBP measurement is associated with an increased risk of operator error and manipulation of data, and its predictive value is not as well supported in the literature. The current PAMELA study does not support the conclusion that ABP or HBP monitoring is prognostically superior to OBP.
Previous studies demonstrating the superiority of ABP were performed in hypertensive subjects; in contrast, PAMELA was performed in the general population, the majority of whom did not have hypertension. Perhaps ABP is prognostically superior to OBP in patients with hypertension, but not in those with normal BP. In addition, there are important limitations to this study that may also help explain these data. First, a major finding of this study is that the risk of mortality increased more steeply for OBP; however, ABP and HBP averages were distributed over a much narrower range than OBP, which calls into question the validity of this conclusion. Second, the number of CV deaths was small (n=56), and the study may have been underpowered to demonstrate a significant difference in attributable risk between BP methods. Third, the use of only two HBP measurements may have decreased the predictive value of this modality; perhaps a model using a significantly greater number of HBP readings would have been more predictive.
These limitations aside, there are a number of important findings from PAMELA. First, all three methods of BP measurement—OBP, HBP, and ABP—were predictive of future mortality. Use of any of the three is probably reasonable to make the diagnosis of hypertension, currently based on in‐office measurement. Second, average measured HBP and ABP were significantly lower than the corresponding OBP measurement, a finding often seen in hypertensive individuals. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7) recommends that the diagnosis of hypertension be considered in patients with average ABP exceeding 135/85 mm Hg, whereas these data suggest that ABP or HBP corresponding to an OBP of 140/90 mm Hg may be considerably lower. Third, the fact that night ABP was more predictive of mortality than day or 24‐hour ABP is interesting and supports the previous finding that subjects who do not experience a “nocturnal dip” on ABP monitoring have a worse prognosis. Fourth, regardless of the method of BP measurement utilized, systolic BP was more predictive of death than diastolic BP. Finally, the addition of HBP to OBP did increase the predictive value of the model somewhat over using OBP alone.
These findings suggest a potential strategy for making the diagnosis of hypertension. OBP should continue to be used as the primary method in screening for hypertension. While HBP measurement can be used as an alternative method of screening or to confirm a diagnosis of hypertension made by OBP, the use of ABP should be reserved for subjects with high OBP readings, especially when white coat hypertension is suspected. Measuring ABP in the general population is not cost effective. More data are needed to determine the appropriate BP levels for the diagnosis of hypertension and prehypertension when using HBP instead of OBP measurement. It remains unclear if either of these modalities can be used for monitoring patients with hypertension longitudinally.
ANGIOTENSIN‐CONVERTING ENZYME INHIBITOR USE IS ASSOCIATED WITH ANEMIA IN PATIENTS WITH HEART FAILURE
Heart failure (HF) remains a significant public health problem; almost five million Americans are affected. Antagonism of the renin‐angiotensin system (RAS) with either angiotensin‐converting enzyme (ACE) inhibitor or angiotensin receptor blocker (ARB) therapy has reduced the morbidity and mortality associated with HF. The presence of anemia in patients with HF increases mortality. This is often under‐appreciated, and not often addressed in clinical guidelines. For example, in the Studies of Left Ventricular Dysfunction (SOLVD) trial, each 1% reduction in prerandomization hematocrit value was associated with a 6% increased risk of death.
Studies have demonstrated that both ACE inhibitors and ARBs are effective agents for reducing hemoglobin concentrations in people with polycythemia. In addition, small studies or case series have found that ACE inhibitors reduce hemoglobin concentrations in healthy individuals, as well as in patients on dialysis and with high‐altitude polycythemia. Little information exists on the effects of ACE inhibitors on the hemoglobin concentrations in patients with HF. Investigators evaluated the effect of the ACE inhibitor enalapril in subjects with symptomatic and asymptomatic systolic dysfunction in the SOLVD trial.
Using the SOLVD database, investigators compared the odds of developing anemia at 1 year in patients who were not anemic on entry and were randomized either to enalapril or placebo. They were all receiving other usual care for heart failure. Defining anemia as a hematocrit less than 40% in men and less than 37% in women, they determined whether the development of anemia, irrespective of etiology, was associated with an increased risk of all‐cause mortality, and whether enalapril retained its beneficial effects in anemic patients with HF compared with those who did not develop anemia.
Pooling the 2569 individuals in the prevention arm and 4228 individuals in the treatment arm from the original study, investigators excluded 361 participants from the analysis who had incomplete data. As another 1187 were anemic on entry into the trial, they were excluded from the development of new anemia calculation but were included in the overall analysis to determine if enalapril benefited people with anemia on entry or those who developed anemia at the end of 1 year. Of the 5249 individuals with no anemia at baseline, 1075 were excluded for incomplete data at 1 year, allowing complete data to be evaluated in 4174 of the 6797 total subjects.
Compared with placebo, enalapril therapy was associated with a decrease in hematocrit that began as early as 6 weeks after randomization. At 1 year, 11.3% of those randomized to enalapril developed anemia compared with 7.9% of patients assigned to placebo (unadjusted odds ratio [OR], 1.48; 95% confidence interval [CI], 1.20–1.82). In a logistic model adjusting for confounders, including changes in serum creatinine and weight gain, use of enalapril continued to be associated with an increased risk of developing new anemia at 1 year (adjusted OR, 1.56; 95% CI, 1.26–1.93). In multivariate analysis, patients who were anemic on entry had a 44% increase in all‐cause mortality (hazard ratio, 1.44; 95% CI, 1.31–1.66) whereas development of anemia at 1 year was associated with a 108% mortality increase (hazard ratio, 2.08; 95% CI, 1.82–2.38). After adjusting for confounders, each percentage point increase in hematocrit was associated with a 3% reduction in mortality in the placebo group and a 2% reduction in individuals in the enalapril group. The use of enalapril was associated with a survival benefit after adjusting for the presence of anemia on entry or the development of anemia during the trial.
Enalapril was associated with increased odds of developing anemia at 1 year. As the enalapril dose was determined by clinical factors individualized to each subject, no evaluation occurred of the relationship between the development of anemia and the dose of enalapril utilized. People with anemia at baseline had an increased risk of death of about half greater than those who developed anemia during the trial. Despite the increased risk for the development of anemia on enalapril, its use continued to be associated with a survival benefit.—Ishani AI, Weinhandl MS, Zhao Z, et al. Angiotensin‐converting enzyme inhibitor as a risk factor for the development of anemia, and the impact of incident anemia on mortality in patients with left ventricular dysfunction. J Am Coll Cardiol. 2005; 45:391–399.
COMMENT
ACE inhibitors continue to be widely used in the treatment of hypertension and are the only class of antihypertensive agents that received a recommendation for use in all six compelling indications in the Seventh Report of the Joint National Committee on the Prevention, Evaluation, Detection, and Treatment of High Blood Pressure (JNC 7). Multiple clinical outcome trials have confirmed their benefit in patients with HF with reduced ejection fractions, where they are often used as initial therapy or add‐on therapy to block the effects of the RAS.
Recently, ACE inhibitor use has been associated with reduced hemoglobin concentrations in healthy individuals as well as in patients on dialysis, and patients with post‐transplant erythrocytosis or high‐altitude polycythemia. Most of these studies have either been small case series or crossover studies. The changes have usually been small and clinically insignificant, and clinicians may still be unaware of this association. The present analysis is the first to observe this finding in a large randomized clinical trial; the use of enalapril for 1 year in individuals with HF with reduced left ventricular function was associated with an increased incidence of anemia.
The mechanism by which ACE inhibitors cause anemia remains unclear. Much interest has focused on the RAS and the role of angiotensin II to stimulate erythroid precursors through activation of the ATI receptor on erythroid progenitor blast‐forming units. In addition, the RAS is intricately linked with the production of endogenous erythropoietin in the peritubular fibroblasts of the kidney. While activation of this system will enhance erythropoietin production, ACE inhibitors decrease circulating angiotensin II levels, with subsequent inhibition of erythroid precursors. Other studies have demonstrated that ACE inhibition lowers insulin‐like growth factor‐1, which has also been associated with erythroid stimulation. In addition, ACE inhibitors increase plasma levels of N‐acetylseryl‐aspartyl‐lysyl‐proline, a natural stem cell regulator that inhibits the formation of pluripotent hematopoietic stem cells while it also inhibits the production of interleukin‐12, a cytokine known to stimulate erythropoiesis.
Anemia occurring at 1 year in patients with HF associated with reduced left ventricular function is common and increases the risk of death. Treatment with an ACE inhibitor increases the odds of developing anemia at 1 year. While outcome is best in individuals without anemia either at baseline or during treatment, overall mortality is further improved on ACE inhibitor therapy, even in those who develop anemia.