Blood Pressure Usually Considered Normal Is Associated with an Elevated Risk of Cardiovascular Disease

Transcription

1 The American Journal of Medicine (2006) 119, CLINICAL RESEARCH STUDY Blood Pressure Usually Considered Normal Is Associated with an Elevated Risk of Cardiovascular Disease Abhijit V. Kshirsagar, MD, MPH, a Myra Carpenter, PhD, b,c Heejung Bang, PhD, d Sharon B. Wyatt, RN, CANP, PhD, e Romulo E. Colindres, MD, MSPH a a UNC Kidney Center, School of Medicine, b Department of Epidemiology, c Collaborative Studies Coordinating Center, Department of Biostatistics, School of Public Health, University of North Carolina at Chapel Hill; d Division of Biostatistics and Epidemiology, Department of Public Health, Weill Medical College, Cornell University, New York, NY; and e Schools of Nursing and Medicine and Jackson Heart Study Examination Center, University of Mississippi Medical Center, Jackson. ABSTRACT PURPOSE: Research on the risk of cardiovascular disease among individuals with prehypertension (blood pressure 120/80 to 139/89 mm Hg) is incomplete. Additional information among individuals with a high risk of cardiovascular disease complications may help to focus current and future efforts. SUBJECTS AND METHODS: We performed a prospective cohort analysis among 8960 middle-aged adults in the Atherosclerosis Risk in Communities (ARIC) study. The exposure variables were blood pressure levels: high normal blood pressure, systolic blood pressure mm Hg or diastolic blood pressure mm Hg; and normal blood pressure, systolic blood pressure mm Hg or diastolic blood pressure mm Hg. The outcome was incident cardiovascular disease defined as fatal/nonfatal coronary heart disease, cardiac procedure, silent myocardial infarction, or ischemic stroke. Subgroup analysis was performed among blacks, diabetics, individuals aged years, individuals with renal insufficiency, and among individuals with varying levels of low-density lipoprotein (LDL) cholesterol and body mass index (BMI). RESULTS: Compared with optimal blood pressure (systolic blood pressure 120 mm Hg and diastolic blood pressure 80 mm Hg), the relative risk (RR) of cardiovascular disease for high normal blood pressure was 2.33 (95% confidence interval [CI], ), and RR for normal blood pressure was 1.81 ( ); among blacks: RR for high normal blood pressure was 3.29 (95% CI, ); among diabetics: RR for high normal blood pressure 4.10 (95% CI, ); age years: RR for high normal blood pressure 2.41 (95% CI, ) among individuals with renal insufficiency: RR for high normal blood pressure was 1.90 (95% CI, ); among individuals with BMI 30 kg/m 2 :RRfor high normal blood pressure was 3.56 (95% CI, ); and among individuals with LDL 160 mg/dl, RR for high normal blood pressure was 1.85 (95% CI, ). CONCLUSIONS: Individuals with prehypertensive levels of blood pressure have an increased risk of developing cardiovascular disease relative to those with optimal levels. The association is pronounced among blacks, among individuals with diabetes mellitus, and among those with high BMI Elsevier Inc. All rights reserved. KEYWORDS: Prehypertension; Cardiovascular disease; Cohort study In 2003, the Seventh Joint National Committee (JNC 7) on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure introduced a new and distinct category of blood pressure, prehypertension. 1 The committee recommended close follow-up and possible intervention for individuals with Requests for reprints should be addressed to Abhijit V. Kshirsagar MD MPH, CB Burnett-Womack Hall, Chapel Hill, NC sagar@med.unc.edu systolic blood pressure mm Hg or diastolic blood pressure mm Hg 1,2 because of an elevated risk of incident cardiovascular disease. 3,4 The designation of prehypertension, blood pressure levels usually considered normal, has raised concerns. The new category will swell the ranks of individuals targeted for blood pressure modification by millions in the United States alone. The influx would strain the infrastructure of all but the wealthiest nations. More importantly, the prehyperten /$ -see front matter 2006 Elsevier Inc. All rights reserved. doi: /j.amjmed

2 134 The American Journal of Medicine, Vol 119, No 2, February 2006 CLINICAL SIGNIFICANCE sive group represents a heterogeneous cohort of individuals with varying risk profiles for cardiovascular disease; undoubtedly, some individuals would be treated unnecessarily with pharmacologic agents. These concerns may help to explain the reluctance of clinicians and societies to adopt the recommendations of the JNC 7. 5,6 Quantifying the impact of prehypertension in groups with a high risk for cardiovascular disease complications may help to focus efforts. There is sparse information for individuals with diabetes mellitus, individuals with renal insufficiency, overweight/obese individuals, individuals with hypercholesterolemia, minorities, or for individuals with relatively advanced age. We therefore investigated the association of prehypertension levels of blood pressure with the risk of incident cardiovascular disease in the Atherosclerosis Risk in Communities (ARIC) study. This large, biracial cohort of middle-aged men and women with varied demographic and medical characteristics allowed us to examine the association of prehypertension levels of blood pressure with cardiovascular disease in several subgroups of interest. METHODS Study Participants The ARIC study is a prospective epidemiologic study of the etiology and natural history of atherosclerotic diseases in men and women. 7 Participants were aged 45 to 64 years at the baseline examination (Visit 1) conducted from This population-based cohort was recruited from Forsyth County, NC; Jackson, Miss; suburbs of Minneapolis, Minn; and, Washington County, Md. The present investigation was restricted to black or white participants without prevalent coronary heart disease based on electrocardiogram (ECG) evidence or self-report of physician diagnosis and without self-reported physician diagnosis of stroke at baseline. Individuals at the initial examination with diastolic blood pressure 90 mm Hg or systolic blood pressure 140 mm Hg, or who reported taking a medication to control hypertension, or missing eligibility data were excluded. Data Collection Three follow-up examinations (Visits 2, 3, and 4) were completed at approximately 3-year intervals following Visit 1. Visit 4 concluded in January Annual telephone contacts that include follow-up for incident cardiovascular events have been conducted since the baseline examination and are continuing. Prehypertension is associated with an elevated risk of cardiovascular disease, especially coronary heart disease. The risk of cardiovascular disease is more pronounced among certain groups with prehypertension diabetics, African Americans, and obese individuals. Intervention trials are urgently needed to quantify the extent of any potential benefit of therapy among prehypertensive individuals. Study Variables Blood pressures were measured utilizing standardized procedures 8 by trained and certified technicians at each of the 4 ARIC examinations. A random-zero mercury manometer was used to measure blood pressures after the participant had been sitting quietly for 5 minutes. Three hierarchal categories of blood pressure were formed according to JNC 6 recommendations. 9 Systolic and diastolic pressures were independently classified as optimal, normal, or high normal. An individual s blood pressure was then classified as the worse of these two categories. For example, a reading of 130/60 mm Hg yields a high normal systolic pressure and an optimal diastolic pressure resulting in an overall classification of high normal. High normal blood pressure: 130 mm Hg systolic blood pressure 140 mm Hg or 85 mm Hg diastolic blood pressure 90 mm Hg. Normal blood pressure: 120 mm Hg systolic blood pressure 130 mm Hg or 80 mm Hg diastolic blood pressure 85 mm Hg. Optimal blood pressure: systolic blood pressure 120 mm Hg and diastolic blood pressure 80 mm Hg. Race identification was reported by the participant or a member of the participant s household using categories (white, black, American Indian/Alaskan Indian, or Asian/ Pacific Islander). Certified technicians adhering to standardized procedures drew and processed blood. 10 Fasting glucose (hexokinase method), fibrinogen (thrombin-time titration method), von Willebrand factor (vwf, ELISA procedure), and highdensity lipoprotein cholesterol (HDL, enzymatic method) were measured by central laboratories. White blood cell count (WBC, by Coulter counters) was measured by laboratories local to the examination sites. Low-density lipoprotein cholesterol (LDL) 11 was calculated. Categories of LDL cholesterol were based on National Cholesterol Education Program (NCEP) guidelines. 12 Serum creatinine concentration was determined by the modified kinetic Jaffe method at a central laboratory. A correction constant of 0.24 mg/dl was subtracted from each measured value of serum creatinine concentration in this analysis to correct for noncreatinine chromagens. 13 Renal function was estimated by the Cockcroft-Gault creatinine clearance (CrCl), 14 normalized to 1.73 m Renal insufficiency was defined as CrCl less than 90 ml/min/1.73 m 2, and further categorized by severity according to the recently published guidelines for classification of chronic kidney disease. 16 Creatinine clearance ml/min/1.73m 2 is based on Stage 2 chronic kidney disease, and CrCl 60 ml/min/1.73m 2 is based on stage 3 or higher chronic kidney disease. Body mass index was computed using weight and height collected by trained and certified staff using standardized pro-

3 Kshirsagar et al Prehypertension and CVD 135 cedures and equipment. Categories of body mass index (BMI) were based on Federal guidelines. 17 The sport index for physical activity was derived from a physical activity interview developed by Baecke and colleagues. 18 The sport index reflects frequency, duration, and intensity of reported sport activities augmented by information on leisure time physical activity and sweating. Cholesterol-lowering and anti-hypertensive medication use reflected medications taken during the 2 weeks prior to the examination. Education level was self-reported. Diabetic status was defined based on a fasting (8 or more hours) glucose of 126 mg/dl, or nonfasting glucose of 200 mg/dl, self-reported physician s diagnosis of diabetes, or use of an antidiabetic medication during the 2 weeks prior to the examination. Cigarette smoking status was classified as either current, former, or never, based on self-report. Outcomes All participant hospitalizations occurring after the baseline examination were reviewed systematically for cardiovascular outcomes. Hospitalizations were identified from two information sources participant self-report and review of discharge lists prepared by area hospitals. Hospital medical records were reviewed and ones meeting specified criteria for possible cardiovascular disease were abstracted in detail. Incident cardiovascular disease was defined as the occurrence of either incident coronary heart disease or incident ischemic stroke. Procedures for ascertainment of incident coronary heart disease and for ascertainment and classification of ischemic stroke 23 are published elsewhere. Statistical Analyses Cox proportional hazards regression was used to estimate the hazards ratio (HR) or relative risk between blood pressure levels and cardiovascular disease event rate. 24 The relative risk between blood pressure levels and incident coronary heart disease or incident stroke was also estimated with Cox proportional hazards regression. Follow-up time was defined as the number of days from Visit 1 to the date of the first occurrence of cardiovascular disease as defined previously, death, last contact, or December 31, We treated all the risk factors, including blood pressure, as fixed at baseline as well as time-varying unless information was available only at baseline. Time-varying covariates can capture the dynamic nature of changes in covariate status. The time-varying covariates included systolic blood pressure, diastolic blood pressure, diabetic status, HDL cholesterol, LDL cholesterol, smoking status, BMI, cholesterol-lowering medication use, and sport index. Although the resulting HRs from the analysis using fixed covariates, as expected, were slightly attenuated, the estimates from both analyses were highly consistent. Thus, we decided to present the results from the analysis allowing covariates to be time-dependent. For constructing time-varying covariates, we carried over the value from the previous visit, if the respective information was missing or not measured at the subsequent visit. A participant s data was censored from analysis at the date of the visit prior to the one at which they reported having started taking an antihypertensive medication. For example, if an individual reported the use of blood pressure medications at Visit 3, he/she was excluded from analysis starting at Visit 2. We expected the exclusion of these individuals, who likely had a heightened risk profile for cardiovascular disease, to bias the results toward the null hypothesis. Models included potential confounding variables sequentially. An initial model adjusted for key demographic variables (sex, age, race, and study site). Traditional risk factors for cardiovascular disease (BMI, smoking status, diabetes mellitus, LDL cholesterol, HDL cholesterol, cholesterol-lowering medication use, physical activity, and education level), in addition to available markers of inflammation (fibrinogen, vwf, and white blood cells) were controlled in the next model. We repeated the analyses for the subgroups. The Kaplan-Meier method was used to estimate the (unadjusted) risk of cardiovascular disease up to about 14 years, stratified by blood pressure status at baseline. 25 All analyses were performed using SAS statistical software (SAS Institute Inc., Cary, NC), and the type I error of 5% with a 2-sided hypothesis was used. RESULTS The study population consisted of 8960 middle-age men and women after the necessary exclusion of individuals with baseline cardiovascular disease and hypertension (Figure 1). The majority of individuals in the study population had blood pressure in the optimal range (n 5622), followed by normal blood pressure (n 2059), and high normal blood pressure (n 1279). The numbers of individuals that comprise each subgroup vary from approximately 600 diabetics to nearly 4000 individuals with renal insufficiency. Of the cohort of 8960 individuals at baseline, 8332 (93%) completed Visit 2, 7671 (86%) completed Visit 3, and 7095 (79%) completed the final evaluation at Visit 4. At baseline there were important trends in demographic and medical characteristics by category of blood pressure (Table 1). Individuals with high normal blood pressure had a greater prevalence of traditional risk factors for cardiovascular disease compared with the optimal blood pressure group, including male sex, nonwhite race, diabetes mellitus, and others. Risk factor measures for individuals in the normal blood pressure group generally fell between values for the optimal blood pressure and high normal blood pressure groups. The proportion of current smokers and renal function did not vary meaningfully across the 3 levels of blood pressure. The incidence of cardiovascular disease increased monotonically across the 3 categories of blood pressure level (Figure 2 and Table 2) for the entire cohort. During the follow-up period (average 11.6 years) there were a total of 772 new cardiovascular disease events. The vast majority of these events were related to coronary heart disease. Individuals with diabetes mellitus had the highest proportion of cardiovascular disease events, followed by individuals with LDL 160 mg/dl, and age years. Events related to coronary heart disease events comprised the majority of cardiovascular disease events (data presented in Appendix). Normal blood pressure and high normal blood pressure were each associated with an increased risk for cardiovascular disease compared with optimal blood pressure (Table 3). Among individuals with normal blood pressure, the hazards

4 136 The American Journal of Medicine, Vol 119, No 2, February 2006 Figure 1 Study population. *Not mutually exclusive. NBP normal blood pressure; HNBP high normal blood pressure. ratio for incident cardiovascular disease compared with the optimal group was 1.81 (95% confidence interval [CI], ) after adjustment for demographic factors, and remained significant after further adjustment. In the high normal blood pressure group, the hazards ratio for incident cardiovascular disease compared with optimal blood pressure was 2.49 (95% CI, ) after adjustment for baseline demographic factors, and remained significant after adjustment for traditional risk factors. Both the normal blood pressure and high normal blood pressure groups were associated with incident coronary heart disease. The point estimates for incident stroke for the normal blood pressure and high normal blood pressure groups compared with optimal blood pressure were significant only with the demographic model for normal blood pressure, but the number of incident stroke events was relatively small (n 128). Table 1 Baseline demographic and medical characteristics of the study population by blood pressure Optimal blood pressure Normal blood pressure High normal blood pressure Variable n 5622 n 2059 n 1279 Age (years) 53 (6) 54 (6) 55 (6) Sex (% male) 42% 51% 51% Race (% white) 86% 77% 72% Diabetes mellitus 5% 7% 10% Education ( high school) 16% 21% 23% Current smoker 28% 25% 24% Systolic BP (mm Hg) 106 (8) 123 (4) 132 (5) Diastolic BP (mm Hg) 67 (7) 75 (7) 79 (8) LDL cholesterol (mg/dl) 134 (38) 137 (38) 137 (38) HDL cholesterol (mg/ll) 53 (17) 52 (17) 53 (18) Serum glucose (mg/dl) 101 (28) 105 (31) 108 (33) CrCl ml/min/1.73m 2 95 (19) 95 (20) 96 (21) Fibrinogen mg/dl 291 (60) 299 (62) 299 (63) vwf % 111 (43) 115 (46) 117 (45) WBC count 10 3 cells/mm (1.9) 6.1 (1.8) 6.0 (1.8) BMI 26.0 (4.2) 27.5 (5.0) 28.1 (5.3) Sport index 2.5 (0.8) 2.5 (0.8) 2.4 (0.8) Cholesterol medication usage 1.7% 2.2% 2.2% Mean (standard deviation) for continuous variables. BP blood pressure; CrCl creatinine clearance; vwf von Willebrand factor; WBC white blood cell.

5 Kshirsagar et al Prehypertension and CVD 137 Figure 2 Incidence of CVD by blood pressure level. The associations of prehypertension levels of blood pressure with incident cardiovascular disease among the subgroups of interest are presented in Table 4. Normal and high normal blood pressure were both significantly associated with incident cardiovascular disease among all subgroups. The majority of the cardiovascular disease events were related to coronary heart disease rather than stroke. Analysis of the outcome by coronary heart disease or stroke revealed that the hazards ratios were significant for coronary heart disease alone, and not for stroke (data not shown). The point estimates for cardiovascular disease were especially high among blacks, diabetic individuals, individuals with LDL mg/dl, and BMI 30 kg/m 2. Among blacks, the Hazard Ratio for cardiovascular disease was 3.29 for high normal blood pressure compared with optimal blood pressure after full adjustment. Among individuals with diabetes mellitus, normal blood pressure was associated with over a 2-fold greater risk of incident cardiovascular disease compared with optimal blood pressure after adjustment for demographic and traditional risk factors; furthermore, high normal blood pressure was associated with a 4-fold increase for new cardiovascular disease events compared with the optimal blood pressure group. Among individuals with LDL cholesterol mg/ dl, high normal blood pressure was associated with over a 2-fold increase in cardiovascular disease compared with optimal levels of blood pressure. Finally, among individuals with BMI 30 kg/m 2, normal blood pressure and high normal blood pressure were associated with a more than a 3-fold increase in cardiovascular disease compared with optimal blood pressure. Table 2 Incident CVD, CHD, and stroke events by blood pressure category for the entire cohort (proportion in parentheses) Optimal blood pressure Normal blood pressure High normal blood pressure Total CVD 393/5622 (0.07) 221/2059 (0.11) 158/1279 (0.12) 772/8960 (0.09) CHD 343/5622 (0.06) 192/2059 (0.09) 136/1279 (0.11) 671/8960 (0.07) Stroke 65/5622 (0.01) 38/2059 (0.02) 25/1279 (0.02) 128/8960 (0.01) CVD cardiovascular disease; CHD coronary heart disease.

6 138 The American Journal of Medicine, Vol 119, No 2, February 2006 Table 3 Hazards ratio* of CVD relative to the optimal blood pressure group by level of blood pressure (95% CI) Model Optimal blood pressure Normal blood pressure High normal blood pressure Demographic CVD ( ) 2.49 ( ) CHD ( ) 2.64 ( ) Stroke ( ) 1.43 ( ) Traditional CVD ( ) 2.33 ( ) CHD ( ) 2.44 ( ) Stroke ( ) 1.31 ( ) Demographic model: center, age, race, sex. Traditional model: demographic model variables and body mass index, diabetes mellitus, smoking status, low-density lipoproteins, high-density lipoproteins, education level, sport index, cholesterol lowering medication use, fibrinogen, von Willebrand factor, white blood cell count (WBC). *Time-dependent covariates were used. CVD cardiovascular disease; CHD coronary heart disease. DISCUSSION Previous classification schemes had emphasized a threshold for intervention to reduce cardiovascular disease complications at a blood pressure 140/90 mm Hg for most individuals. Clinical practice patterns have generally reflected these recommendations. The results of our analysis lend additional support for the new JNC 7 recommendation: individuals with prehypertension require close follow-up at the very least, and probably, lifestyle modification. Furthermore, our findings suggest that focusing efforts on certain groups of individuals with a high risk of cardiovascular complications may be warranted. After adjustment for demographic and traditional risk factors, normal blood pressure and high normal blood pressure were associated with a 69% and 133% greater risk of incident cardiovascular disease, respectively, compared with optimal blood pressure. Importantly, coronary heart disease events comprised the majority of the cardiovascular disease outcomes. The overall number of stroke events was small, and prevent us from drawing any meaningful inferences. Furthermore, subgroup analysis revealed important heterogeneity of the relationship of prehypertension and incident cardiovascular disease events. The point-estimates were especially high among blacks, individuals with diabetes mellitus, high BMI, Table 4 Hazards ratio* of CVD relative to optimal blood pressure for groups with high baseline risk (95% CI) Model Normal blood pressure High normal blood pressure Demographic Blacks 2.37 ( ) 3.61 ( ) Diabetics 2.91 ( ) 3.90 ( ) CrCl ml/min/1.73m ( ) 1.99 ( ) Age years 1.79 ( ) 2.51 ( ) LDL mg/dl 1.80 ( ) 3.24 ( ) mg/dl 1.54 ( ) 2.37 ( ) 160 mg/dl 1.60 ( ) 1.83 ( ) BMI kg/m ( ) 2.10 ( ) kg/m ( ) 2.36 ( ) 30 kg/m ( ) 3.38 ( ) Traditional Blacks 2.09 ( ) 3.29 ( ) Diabetics 2.36 ( ) 4.10 ( ) CrCl ml/min/1.73m ( ) 1.90 ( ) Age years 1.68 ( ) 2.41 ( ) LDL mg/dl 1.59 ( ) 2.66 ( ) mg/dl 1.54 ( ) 2.42 ( ) 160 mg/dl 1.55 ( ) 1.85 ( ) BMI kg/m ( ) 2.09 ( ) kg/m ( ) 2.14 ( ) 30 kg/m ( ) 3.56 ( ) Demographic model: center, age, sex, and race when appropriate. Traditiona model: demographic model variables and body mass index, smoking status, low-density lipoproteins, high-density lipoproteins, education level, sports index, cholesterol lowering medication use and diabetes mellitus, fibrinogen, von Willebrand factor, white blood cell count (WBC). *Time-dependent covariates were used. LDL low-density lipoprotein; BMI body mass index; CrCL creatinine clearance.

7 Kshirsagar et al Prehypertension and CVD 139 and relatively low LDL-cholesterol values. Individuals with mild renal insufficiency, with higher age, or with moderately high levels of LDL-cholesterol demonstrated a 60-90% greater risk of cardiovascular disease for prehypertension levels of blood pressure compared with the optimal blood pressure group. The potential impact of prehypertension is large. The population attributable risk (PAR) represents the proportion of the disease in the whole population that would be prevented if the risk factor were eliminated. For the entire cohort, this would amount to almost 30% of new cardiovascular events 13.7 % from the normal group, and 16.0 % from the high normal group. Among blacks, the elimination of normal and high normal blood pressure would prevent 24.1% and 33.4% of new cardiovascular events, respectively. For persons with LDL cholesterol mg/dl, the corresponding PAR is estimated as 11.5% for the normal group, and 19.2% for high normal. Nearly 26% and 40% of new cardiovascular events might be prevented by eliminating normal blood pressure and high normal blood pressure, respectively, among diabetic individuals. Among the group of individuals with BMI 30 kg/m 2, 38% and 36% of new cardiovascular events could be prevented with the elimination of normal and high normal blood pressure, respectively. These estimates may be understating the true magnitude of prehypertension given the rising prevalence of obesity. The results of the present analysis also reiterate the importance of increased blood pressure as a risk factor for cardiovascular disease Many causative/predictive factors influence the clinical occurrence of cardiovascular disease, among which are the interplay of dyslipidemia, diabetes mellitus, hypertension, smoking, and other (some yet-to-be determined) host and environmental factors. In our analysis, adjustment for most of these risk factors did not greatly attenuate the relationship of blood pressure and cardiovascular disease. Recently, there has been debate about whether traditional risk factors for cardiovascular disease hold for individuals with renal insufficiency. 32,33 Unique physiologic changes occur with the onset of renal insufficiency; thus kidneyspecific risk factors such as osteodystrophy, anemia, uremic toxins, and others have been proposed to account for the high burden of cardiovascular disease. 34,35 Furthermore, among individuals with advanced kidney disease, neither systolic blood pressure nor diastolic blood pressure elevation was associated with cardiovascular disease. 36 The current analysis suggests that even modestly elevated blood pressure remains an important risk factor for cardiovascular disease among individuals with renal insufficiency. There was a stepwise increase in risk across the 3 categories of blood pressure, although the point estimates were smaller among individuals with renal insufficiency than the point estimates for the entire cohort. The magnitude of effect was surprisingly large in a group of individuals with relatively low levels of LDL cholesterol. This finding was not an effect of small sample size or a small number of cardiovascular disease events (see Appendix). Speculatively, at lower levels of LDL cholesterol, the relative effect of other risk factors for cardiovascular disease (such as blood pressure) may predominate. Additional study is necessary to confirm this observation. The current analysis adds to existing information. Previous studies have also demonstrated a similar relationship of prehypertension levels of blood pressure and cardiovascular disease. In one analysis from the Framingham cohort, men and women with systolic blood pressure mm Hg or diastolic blood pressure mm Hg have a 1.6- and 2.5-fold increased risk of fatal and nonfatal cardiovascular disease, respectively, compared with those with blood pressure at or below 120/80 mm Hg. 4 Yet, the Framingham cohort includes few minorities, and the analysis did not investigate subgroups included in the present study. Another study demonstrated that each increment of 20 mm Hg systolic blood pressure or 10 mm Hg diastolic blood pressure doubles the risk of cardiovascular disease death across the entire range of blood pressure, including blood pressure in the prehypertension range. 3 Yet this large meta-analysis may have underestimated the magnitude of clinically relevant cardiovascular disease by using only fatal cardiovascular disease as the outcome of measure. This study should be evaluated in the context of the following limitations. Despite the strong and consistent association of prehypertension with elevated risk of cardiovascular disease, it is not possible to determine whether the elevated risk was solely attributable to the blood pressure. A modest elevation of blood pressure above optimal is closely associated with multiple rheologic, hemodynamic, and metabolic abnormalities; many of these abnormalities independently increase the risk of cardiovascular disease. 37 Two, the age range of the ARIC study was years at baseline, or middle-age. With the aging population, information on the relationship of prehypertension and cardiovascular disease among elderly individuals would be pertinent. In summary, prehypertension levels of blood pressure are clearly associated with a significant increase in incident cardiovascular disease. The effect of prehypertension is especially pronounced among blacks, individuals with diabetes mellitus, elevated BMI, and relatively low LDL cholesterol levels suggesting that clinicians may consider lowering the threshold for action among these groups. These findings may also offer compelling evidence for screening and early detection in vulnerable groups. Randomized intervention studies are needed to quantify the extent of any potential benefit of therapy among individuals with blood pressure levels usually considered normal. ACKNOWLEDGMENTS The Atherosclerosis Risk in Communities Study (ARIC) is carried out as a collaborative study supported by National Heart, Lung, and Blood Institute contracts N01-HC-55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC , N01-HC-55021, and N01-HC Dr. Carpenter s effort is supported by the Jackson Heart Study, National Heart, Lung, and Blood Institute and National Center for Minority Health and Health Disparities contract N01-HC Dr. Wyatt s effort is supported by the Jackson Heart Study, National Heart, Lung, and Blood Institute and the National Center for Minority Health and Health Disparities contract N01-HC The authors thank the staff and participants of the ARIC study for their important contributions.

8 140 The American Journal of Medicine, Vol 119, No 2, February 2006 Appendix Incident cardiovascular disease, coronary heart disease, and stroke events by blood pressure category for the subgroups of interest (risk in parentheses) Optimal blood pressure Normal blood pressure High normal blood pressure Total Age years n 3610 n 1087 n 630 n 5327 CVD 197 (0.055) 99 (0.091) 67 (0.106) 363 (0.068) CHD 183 (0.051) 88 (0.081) 59 (0.094) 330 (0.062) Stroke 18 (0.005) 14 (0.013) 9 (0.014) 41 (0.008) Age years n 2012 n 972 n 649 n 3633 CVD 196 (0.097) 122 (0.126) 91 (0.140) 409 (0.113) CHD 160 (0.080) 104 (0.107) 77 (0.119) 341 (0.094) Stroke 47 (0.023) 24 (0.025) 16 (0.025) 87 (0.024) Blacks n 809 n 482 n 363 n 1654 CVD 36 (0.044) 29 (0.060) 37 (0.102) 102 (0.062) CHD 25 (0.031) 24 (0.050) 29 (0.080) 78 (0.047) Stroke 14 (0.017) 7 (0.015) 9 (0.025) 30 (0.018) Whites n 4813 n 1577 n 916 n 7306 CVD 357 (0.074) 192 (0.122) 121 (0.132) 670 (0.092) CHD 318 (0.066) 168 (0.107) 107 (0.117) 593 (0.081) Stroke 51 (0.011) 31 (0.020) 16 (0.017) 98 (0.013) Diabetics n 303 n 152 n 131 n 586 CVD 61 (0.201) 45 (0.296) 36 (0.275) 142 (0.242) CHD 53 (0.175) 41 (0.270) 33 (0.252) 127 (0.217) Stroke 14 (0.046) 7 (0.046) 4 (0.031) 25 (0.043) Nondiabetic n 5291 n 1897 n 1133 n 8321 CVD 331 (0.063) 176 (0.093) 121 (0.107) 628 (0.075) CHD 289 (0.055) 151 (0.080) 102 (0.090) 542 (0.065) Stroke 51 (0.010) 31 (0.016) 21 (0.019) 103 (0.012) CrCl ml/min/1.73m 2 n 2381 n 893 n 548 n 3822 CVD 177 (0.074) 96 (0.108) 75 (0.137) 348 (0.091) CHD 152 (0.064) 81 (0.091) 61 (0.111) 294 (0.077) Stroke 33 (0.014) 17 (0.019) 15 (0.027) 65 (0.017) CrCl 90 ml/min/1.73m 2 n 3155 n 1123 n 705 n 4983 CVD 206 (0.065) 121 (0.108) 81 (0.115) 408 (0.082) CHD 183 (0.058) 108 (0.096) 74 (0.105) 365 (0.073) Stroke 29 (0.009) 20 (0.018) 9 (0.013) 58 (0.012) LDL 99 n 960 n 318 n 184 n 1462 CVD 37 (0.039) 18 (0.057) 9 (0.049) 64 (0.044) CHD 30 (0.031) 13 (0.041) 4 (0.022) 47 (0.032) Stroke 11 (0.011) 5 (0.016) 5 (0.027) 21 (0.014) LDL n 1693 n 587 n 380 n 2660 CVD 75 (0.044) 47 (0.080) 52 (0.137) 174 (0.065) CHD 61 (0.036) 39 (0.066) 45 (0.118) 145 (0.055) Stroke 18 (0.011) 11 (0.019) 8 (0.021) 37 (0.014) LDL n 1598 n 606 n 357 n 2561 CVD 126 (0.079) 75 (0.124) 47 (0.132) 248 (0.097) CHD 114 (0.071) 66 (0.109) 41 (0.115) 221 (0.086) Stroke 15 (0.009) 12 (0.020) 7 (0.020) 34 (0.013) LDL 160 n 1287 n 504 n 323 n 2114 CVD 149 (0.116) 74 (0.147) 47 (0.146) 270 (0.128) CHD 132 (0.103) 67 (0.133) 43 (0.133) 242 (0.114) Stroke 21 (0.016) 8 (0.016) 4 (0.012) 33 (0.016) BMI 18.5 n 52 n 18 n 13 n 83 CVD 3 (0.058) 2 (0.111) 4 (0.308) 9 (0.108) CHD 3 (0.058) 1 (0.056) 4 (0.308) 8 (0.096) Stroke 1 (0.019) 1 (0.056) 0 (0.000) 2 (0.024) BMI n 2514 n 622 n 381 n 3517 CVD 131 (0.052) 52 (0.084) 43 (0.113) 226 (0.064) CHD 116 (0.046) 47 (0.076) 36 (0.094) 199 (0.057) Stroke 19 (0.008) 6 (0.010) 7 (0.018) 32 (0.009) BMI n 2178 n 911 n 493 n 3582 CVD 197 (0.090) 114 (0.125) 63 (0.128) 374 (0.104) CHD 175 (0.080) 99 (0.109) 51 (0.103) 325 (0.091) Stroke 29 (0.013) 20 (0.022) 12 (0.024) 61 (0.017) BMI 30 n 877 n 508 n 390 n 1775 CVD 62 (0.071) 53 (0.104) 48 (0.123) 163 (0.092) CHD 49 (0.056) 45 (0.089) 45 (0.115) 139 (0.078) Stroke 16 (0.018) 11 (0.022) 6 (0.015) 33 (0.019)

9 Kshirsagar et al Prehypertension and CVD 141 References 1. Chobanian AV, Bakris GL, Black HR, et al. Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension. 2003;42: Chobanian AV, Bakris GL, Black HR, et al. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA. 2003; 289: Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet. 2002;360: Vasan RS, Larson MG, Leip EP, et al. Impact of high-normal blood pressure on the risk of cardiovascular disease. N Engl J Med. 2001; 345: European Society of Hypertension-European Society of Cardiology guidelines for the management of arterial hypertension. J Hypertens. 2003;21: Weber MA. Yet more hypertension guidelines: what do they add? J Hypertens. 2003;21: ARIC Investigators. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. Am J Epidemiol. 1989;129: ARIC Investigators. Atherosclerosis Risk in Communities Study Protocol Manual 11: Sitting Blood Pressure. Chapel Hill, NC: Collaborative Studies Coordinating Center; The sixth report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. Arch Intern Med. 1997;157: ARIC Investigators. Atherosclerosis Risk in Communities Study Protocol Manual 7: Blood Collection and Processing, Version 1.1. Chapel Hill, NC: Collaborative Studies Coordinating Center; Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18: Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 2002;106: Coresh J, Astor BC, McQuillan G, et al. Calibration and random variation of the serum creatinine assay as critical elements of using equations to estimate glomerular filtration rate. Am J Kidney Dis. 2002;39: Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron. 1976;16: Du Bois D, Du Bois EF. A formula to estimate the approximate surface area if height and weight be known. Arch Intern Med. 1916;17: Work Group and Evidence Review Team: Part 4. Definition and classification of stages of chronic kidney disease. Am J Kidney Dis. 2002;39(suppl 1):S46-S Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults The Evidence Report. National Institutes of Health. Obes Res. 1998;6(suppl 2):51S-209S. 18. Baecke JAH, Burema J, Frijters JER. A short questionnaire for the measurement of habitual physical activity in epidemiologic studies. Am J Clin Nutr. 1982;36: ARIC Investigators. Atherosclerosis Risk in Communities Study Protocol Manual 3. Surveillance Component Procedures. Chapel Hill, NC: Collaborative Studies Coordinating Center; Rosamond WD, Chambless LE, Folsom AR, et al. Trends in the incidence of myocardial infarction and in mortality due to coronary heart disease, 1987 to N Engl J Med. 1998;339: Rosamond WD Folsom AR Chambless LE Wang CH Coronary heart disease trends in four United States communities The Atherosclerosis Risk in Communities (ARIC) study Int J Epidemiol. 30 suppl S17-S White AD, Folsom AR, Chambless LE, et al. Community surveillance of coronary heart disease in the Atherosclerosis Risk in Communities (ARIC) Study: methods and initial two years experience. J Clin Epidemiol. 1996;49: Rosamond WD, Folsom AR, Chambless LE, et al. Stroke incidence and survival among middle-aged adults: 9-year follow-up of the Atherosclerosis Risk in Communities (ARIC) cohort. Stroke 1999;30: Therneau TM, Gambsch P. Modeling Survival Data: Extending the Cox Model. New York: Springer; Kaplan EL, Meier P. Non-parametic estimation from incomplete observations. J Am Stat Assoc. 1958: Eastern Stroke and Coronary Heart Disease Collaborative Research Group. Blood pressure, cholesterol, and stroke in eastern Asia. Eastern Stroke and Coronary Heart Disease Collaborative Research Group. Lancet. 1998;352: Kannel WB. Blood pressure as a cardiovascular risk factor: prevention and treatment. JAMA. 1996;275: MacMahon S, Peto R, Cutler J, et al. Blood pressure, stroke, and coronary heart disease. Part 1, Prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution bias. Lancet. 1990;335: Neaton JD, Wentworth D. Serum cholesterol, blood pressure, cigarette smoking, and death from coronary heart disease. Overall findings and differences by age for white men. Multiple Risk Factor Intervention Trial Research Group. Arch Intern Med. 1992;152: Stamler J, Stamler R, Neaton JD. Blood pressure, systolic and diastolic, and cardiovascular risks. US population data. Arch Intern Med. 1993;153: van den Hoogen PC, Feskens EJ, Nagelkerke NJ, Menotti A, Nissinen A, Kromhout D. The relation between blood pressure and mortality due to coronary heart disease among men in different parts of the world. Seven Countries Study Research Group. N Engl J Med. 2000; 342: Garg AX, Clark WF, Haynes RB, House AA. Moderate renal insufficiency and the risk of cardiovascular mortality: results from the NHANES I. Kidney Int. 2002;61: Liu Y, Coresh J, Eustace JA, et al. Association between cholesterol level and mortality in dialysis patients: role of inflammation and malnutrition. JAMA. 2004;291: Jaradat MI, Molitoris BA. Cardiovascular disease in patients with chronic kidney disease. Semin Nephrol. 2002;22: Paparello J, Kshirsagar A, Batlle D. Comorbidity and cardiovascular risk factors in patients with chronic kidney disease. Semin Nephrol. 2002;22: Cheung AK, Sarnak MJ, Yan G, et al. Atherosclerotic cardiovascular disease risks in chronic hemodialysis patients. Kidney Int. 2000;58: Julius S. Borderline hypertension. Clin Exp Hypertens. 1999;21: