Published online before print July 6, 2004, doi: 10.1161/01.HYP.0000130174.70055.ca
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Hypertension. 2004;44:146.)
© 2004 American Heart Association, Inc.
Scientific Contributions |
Trial of Preventing Hypertension
Design and 2-Year Progress Report
From the Division of Cardiovascular Medicine (S.J., N.K.), University of Michigan Medical Center, Ann Arbor; the Emeritus Department of Biostatistics School of Public Health (M.A.S), University of Michigan, Ann Arbor; the Department of Internal Medicine (S.N.), Division of Hypertension, University of Texas Southwestern Medical Center at Dallas; the Division of Clinical Pharmacology (B.E.), Medical University of South Carolina, Charleston; and AstraZeneca, LP (M.G., E.M.), Wilmington, Del.
Correspondence to Stevo Julius, MD, MD (Hon), ScD, Professor of Internal Medicine and Physiology, Fredrick G.L. Huetwell Professor of Hypertension, University of Michigan Medical Center 3918 TC, 1500 E Medical Center Dr, Ann Arbor, MI 48109-0500. E-mail sjulius{at}umich.edu
 |
Abstract |
|---|
 Top Abstract IntroductionStudy DesignProgress Report for Subjects...Proposed Data Analysis: Concepts...DiscussionReferences |
|---|
The TRial Of Preventing HYpertension (TROPHY) study is an investigator-initiated trial to examine whether early pharmacological treatment in subjects with "high-normal" blood pressure (BP) might prevent or delay the development of clinical hypertension. This is a 4-year, multicenter, randomized, double-blind study in untreated subjects aged 30 to 65 years with entry BPs of 130 to 139/
89 or 139/85 to 89. The participants were randomized either to placebo or to a fixed (16 mg once daily) dose of candesartan cilexetil (candesartan). After 2 years, the candesartan group was switched to placebo, and the placebo group continued taking placebo. The main outcome measure was the development of clinical (treatment-requiring) hypertension assessed by an automated (blinded) BP measurement device. We randomized 809 subjects (59% males, average age 49.0±SD 8.1 years) in 71 study centers in the United States. The entry BP was 134±4.3/84.8±3.9 mm Hg. During the first 2 years, 187 subjects (23%) developed clinical hypertension. All have been given antihypertensive treatment, and 170 continue to be followed in study centers. The study dropout rate is 14.8% (120 subjects). The hypertension rates are higher than anticipated, whereas the rates of dropout are within the sample size projections; thus, the study will have sufficient power to evaluate its hypotheses. In this article, we describe baseline characteristics of TROPHY subjects and discuss novel analytical issues and statistical approaches to evaluate the findings in this trial of primary prevention of hypertension.
Key Words: hypertension, detection and control • blood pressure • receptors, angiotensin • angiotensin antagonist
|
Introduction |
|---|
|
TopAbstractIntroductionStudy DesignProgress Report for Subjects...Proposed Data Analysis: Concepts...DiscussionReferences |
|---|
The TRial Of Preventing HYpertension (TROPHY) is an investigator-initiated trial to examine whether early treatment of mild blood pressure (BP) elevations might prevent or delay the development of the treatment-requiring form of hypertension. The rationale of the TROPHY study is rooted in the following facts.
- Literature summarized as early as 1971 suggests that a slight elevation of BP is a precursor of future hypertension.1 More recent observations to the same effect prompted the authors of the recent seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7)2 to call systolic BP of 120 to 139 mm Hg and diastolic of 80 to 89 mm Hg a state of prehypertension and to suggest that this condition requires clinical attention.
- Most of the mortality and morbidity from hypertension occur among subjects with stage 1 hypertension.3 Therefore, if it were possible to prevent the transition from prehypertension to stage 1 hypertension, this would have a major impact on public health.
- A substantial proportion of previously treated subjects with hypertension remain normotensive for long periods after discontinuation of the treatment. The proportion of such "cures" of hypertension is higher among subjects with lesser levels of initial BP elevation.4–6 Furthermore, in spontaneously hypertensive rats (SHR), an early brief treatment with angiotensin-converting enzyme inhibitors repeatedly attenuated development of hypertension.7–10
- Weight loss,11 salt restriction,12 exercise,13 and use of the Dietary Approaches to Stop Hypertension (DASH) diet14 are proven to lower BP. However, intensive efforts to promote healthy lifestyles in the general population seem to have little lasting effect. The rates of obesity worldwide have increased dramatically,15 and incidence rates of hypertension have not decreased.
Against this background, we proposed to investigate whether treating individuals with high-normal BP values (defined according to JNC 6 criteria) with candesartan (AstraZeneca) would prevent or postpone the development of stage 1 hypertension. We chose candesartan because angiotensin-receptor blocking agents are well tolerated. Moreover, animal experiments suggested that drugs that interfere with the renin-angiotensin system (presumably because of their ability to prevent hypertrophy of resistance arterioles) are particularly effective in ameliorating hypertension in SHR.16
|
Study Design |
|---|
|
TopAbstractIntroductionStudy DesignProgress Report for Subjects...Proposed Data Analysis: Concepts...DiscussionReferences |
|---|
TROPHY is a 4-year, multicenter, randomized, double-blind study in untreated subjects aged 30 to 65 years with entry BPs of 130 to 139/
89 or 139/85 to 89 mm Hg. At entry and during all subsequent clinical visits, BP is measured in the sitting position after 5 minutes of rest by an automated device (Omron 705-CP; Omron Healthcare, Inc.) and by a regular manual clinical device. The readings were taken in weekly intervals during 3 consecutive clinic visits. Patients qualified for the study if their average untreated BP taken by the automated device at 3 separate clinic visits (3 readings at each session) met study entry criteria and if the BP reading during the first clinic visit did not exceed 155/99 mm Hg. If a patient did not qualify initially, the entire 3-week BP measurement protocol could be repeated. BP was assessed with regular and automated devices at each subsequent clinic visit.
The study subjects also measured their BP at home with the automated device daily for 1 week before randomization and before the 12-, 24-, 36-, and 48-month clinic visits.
The study design is shown in Figure 1. During the first 2 years, the subjects were randomized either to placebo or to a fixed (16 mg QD) dose of candesartan. After 2 years of study participation, subjects in the candesartan group began receiving placebo tablets, and the placebo group continued taking placebo tablets.
|
Subjects were seen at months 1 and 3 and every 3 months thereafter until the 24-month visit. At the beginning of the third study year, as in the first study year, the subjects returned to the clinic after 1 month (month 25)and after 2 additional months; they will continue returning to the clinic at 3-month intervals until the end of the study (June 2005).
After a complete baseline physical examination, a 12-lead ECG was obtained, and weight, height, and skinfold thickness were measured. The study participants completed the short-form 36 health status questionnaire. Thereafter, blood samples for baseline measurement of hematocrit, fasting insulin, glucose, cholesterol, high-density lipoprotein (HDL) cholesterol, and triglycerides were drawn, and a urine sample was tested for protein by dipstick. In female participants of childbearing potential, a pregnancy test (ß-human chorionic gonadotropin) was performed to exclude pregnancy. The short-form 36 questionnaire, ECG, and blood and urine analyses are repeated annually or when subjects reach a study end point.
The main outcome in this study is the development of clinical hypertension. According to the study protocol, clinical hypertension occurs if:
- The average clinic BP taken by the automated device was
140/90 mm Hg at 3 different clinic visits during the 4-year study period.
- The average clinic BP taken by the automated device was 160/100 mm Hg at any single visit during the 4-year study period.
- The subject developed signs of target organ damage or had a BP condition that, in the judgment of the clinical investigator, required pharmacological treatment.
- The average clinic BP taken by the automated device was 140/90 mm Hg at the 48-month visit.
For comparability with terminology in other trials and for statistical analyses, the development of clinical hypertension is considered the end point in the TROPHY study. If an end point occurs, the study site continues to follow the patient according to the TROPHY study schedule. Subjects who reach an end point are offered additional therapy with 50 mg of metoprolol succinate (Toprol XL, AstraZeneca) or 12.5 mg hydrochlorothiazide (Microzide, Watson) at no cost, but the physician is free to prescribe other antihypertensive medication (excluding angiotensin receptor-blocking agents).
The primary study objective is to determine whether treatment with 16 mg of candesartan compared with placebo reduces the incidence of hypertension in subjects with high-normal BP evaluated after a 4-year study consisting of 2 years of double-blind treatment followed by 2 years of single-blind placebo treatment.
The secondary objectives are to (1) evaluate the reduction of BP after 2 years of treatment with 16 mg of candesartan versus placebo; (2) evaluate the incidence of hypertension after 2 years of treatment with of 16 mg candesartan versus placebo; and (3) evaluate the quality of life during 2 years of treatment with 16 mg of candesartan versus placebo.
|
Progress Report for Subjects Who Reached 2 Years of Study Participation |
|---|
|
TopAbstractIntroductionStudy DesignProgress Report for Subjects...Proposed Data Analysis: Concepts...DiscussionReferences |
|---|
A total of 809 subjects were randomized at 71 study centers of the TROPHY study. All subjects read and signed an approved informed consent form.
Selected baseline characteristics of TROPHY participants at the time of randomization are given in the Table. The average age was 49 years (minimum 29 and maximum 66). The average body mass index (BMI) of 29.95 kg/m2 is in the overweight range. As defined by the protocol, the average BP taken by the automated device was within the high-normal range. The average BP values obtained in the clinic by either a standard or an automated BP measurement device were similar. The average automated home BP readings were similar to average clinic readings, but the SD of home BP readings was larger than in the clinic. The average heart rate of TROPHY participants, taken either in the clinic or at home, was very similar to average heart rates observed in subjects with borderline hypertension in the Tecumseh study.17 In the Tecumseh study and many other reports,18 the average heart rates of subjects with mild BP were significantly higher than in control subjects.
|
In this article, we consider only the data through the 24-month study visit. In the first 2 years of the TROPHY study, 137 of 809 participants who had baseline measures discontinued participation in the study. None required breaking the study blind. However, 17 of these subjects dropped out of the study after they reached the primary study end point of clinical hypertension. Thus, for the purposes of this study, the true dropout total stands at 120 subjects, 14.8% of the initial study population.
During this observation period, a total of 187 subjects (23% of the 809 subjects at risk) reached the clinical hypertension end point and another 15 had a secondary end point. Of the 15 subjects with a secondary end point, 10 had new onset diabetes mellitus, 2 had angina, 1 had coronary heart disease, 1 had an increase of serum creatinine, and 1 developed arrhythmia. Most of these 15 subjects continue to return for clinic visits, and some of them may still develop a BP end point. The Kaplan-Meyer curve of BP end points during the first 2 years of the study was linear (Figure 2).
|
|
Proposed Data Analysis: Concepts and Methods |
|---|
|
TopAbstractIntroductionStudy DesignProgress Report for Subjects...Proposed Data Analysis: Concepts...DiscussionReferences |
|---|
The TROPHY study is a novel (untested in a randomized trial) approach to the prevention of new-onset hypertension in a previously untreated population of subjects with high-normal BP. As such, it affords a number of interesting analytic issues. It is assumed that candesartan will be an effective antihypertensive agent. Therefore, the primary research question of this study is what happens to subjects with high-normal BP whose BP is maintained in the normal range with drug therapy when this therapy is discontinued? Direct comparison of hypertension incidence rates of this candesartan-treated group with placebo subjects during the final 2 years of study, when all subjects are taking placebo, is fraught with difficulties. Most notable is that a number of subjects in the placebo group are expected to develop hypertension during the first 2 years of treatment. Consequently, the placebo subgroup that did not develop hypertension during the final 2 years of the study might be considered "resistant to developing hypertension." A comparison of this preselected group to the previously treated group just starting placebo in the second half of the study would be neither equitable nor unbiased. Furthermore, a comparison of unadjusted incidence rates at the end of the study would favor the candesartan group because subjects on this regimen had their BP suppressed by antihypertensive treatment in the first 2 years of the study. We will invoke for all analyses the intent-to-treat principle on the basis of group assignment. Because the focus of this trial is to study differential effects of randomized therapy at end points, and in view of the concerns outlined above, we present several conceptual alternative hypotheses. These scenarios will guide the analyses and are presented here to ensure a priori objectivity at the conclusion of the study.
Conceptual Alternative Hypotheses
We envisage 2 unlikely extreme scenarios and 3 "in-between" alternatives.
- Masking (Figure 3, left). Because of its antihypertensive effect, candesartan treatment suppresses (masks) the manifestation of new onset of hypertension but does not affect the underlying evolution of hypertension. Conclusion: The study did not prove its hypothesis of primary prevention.
- Elimination (Figure 3, right). Two years of treatment with candesartan totally prevented the onset of hypertension for 2 years after discontinuation of treatment. Conclusion: Primary prevention is possible; treatment with candesartan altered the underlying pathophysiology of evolving hypertension. The duration of this effect over a longer period of time should be investigated.
- Delaying (Figure 4A). Conclusion: Treatment with candesartan postpones the development of hypertension but does not affect the underlying evolution of hypertension. Nevertheless, such a postponement might be clinically useful. Are periods of "treatment vacations" clinically feasible? In whom? Are they cost effective? How do such regimens affect compliance with treatment? Ultimately, with regard to cardiovascular events, how does early intermittent treatment compare with the present strategy of waiting for stage 1 hypertension to develop and then instituting lifelong treatment?
- Profoundly delaying (Figure 4B). Conclusion: Treatment with candesartan not only postponed the development of hypertension but also considerably affected the pathophysiology of evolving hypertension. This outcome would obviously be clinically attractive. A number of interesting research questions would arise. Would a longer period of treatment or a larger dose have a better effect? Would treatment in younger subjects with prehypertension be more effective than in the middle-aged subjects seen in the TROPHY study?
- Slow unmasking (Figure 4C). Conclusion: Prevention of hypertension is not feasible. However, useful research data might be garnered if further analyses suggest that the outcome is not homogenous (ie, some subjects quickly return to hypertension, whereas others remain protected for a longer period of time).
|
|
Proposed Statistical Analyses
The primary end point is the development of treatable hypertension. To assess differentiable incidence measures for this end point, statistical analyses will use 2 techniques. We shall use methods that consider the presence or absence of the end point, namely methods for rates of events. Fisher exact test is the required first procedure to test the primary study hypothesis: The proportions of subjects developing hypertension between the treatment groups are significantly different at the end of the 4-year study. We shall also use logistic regression, which enables adjustment for covariates. To more completely investigate our alternative scenarios, we shall apply methods that not only assess the existence of the event but also the time of occurrence. These procedures include Kaplan-Meier survival analysis and the Cox model. We will routinely include the covariates of age, sex, body mass index, and baseline BP in all models, in addition to a dichotomous variable indicating randomized therapy.
Further, because as noted in Figures 3 and 4
, we are particularly interested in discerning if or when the slopes of incidence curves for the 2 groups statistically converge, stay parallel, or diverge, we shall develop procedures on the basis of logistic regression, Kaplan-Meier curves, and Cox modeling to test the differential significance of various patterns. Such procedures involve adding covariate indicators to assess "breaks" in incidence rate slopes between the candesartan and placebo groups over time. For example, we shall consider using an indicator declaring when or if a subject became hypertensive (in the first 2 years of study, in the last 2 years of study, or never) or, alternatively, using several indicators to more finely delineate when a change (break) in slope may have occurred, such as use of hypertensive status at each of the 3 monthly clinic visits.
|
Discussion |
|---|
|
TopAbstractIntroductionStudy DesignProgress Report for Subjects...Proposed Data Analysis: Concepts...DiscussionReferences |
|---|
Study Conduct
Sample size for the TROPHY study was estimated on the basis of the incidence rates of hypertension observed in a similar population in the Trial Of dietary Prevention of Hypertension (TOPH).19 We hypothesized that among subjects in the placebo group, 10% per year would develop hypertension (40% at the end of the study). Further, in subjects receiving candesartan, we assumed that there would be a 30% reduction in the incidence rate of hypertension during the 4 years of the study, for a cumulative rate of 28%. To detect this 30% reduction (40% versus 28%) with 95% power using a 2-side
-error of 5%, we calculated that 420 subjects per arm of the study would be needed, or a total of 840 subjects. To compensate for possible loss of follow-up, a goal of 1000 randomized subjects was established. In fact, 809 subjects were randomized. However, in the first 2 years, the incidence rates of hypertension in the entire study cohort were higher than anticipated. We have observed 187 new cases. We expected in the placebo group a rate of 
10% per annum. Thus, we anticipated 80 new hypertensive patients in the placebo group during the first 2 years of the study. We also assumed that 28% of the candesartan group would become hypertensive during the 4 years of the study. Using the null hypothesis effect rate of 10% per annum in years 3 and 4, we expect maximally a 4% per annum rate for the candesartan group during years 1 and 2. This hypothetical rate (4% per annum) translates into an anticipated total of 32 new hypertensives in the candesartan group in years 1 and 2. Thus, we expected 112 (80+32) new cases. On the basis of this appreciably higher-than-expected incidence rate of hypertension (187 cases seen, 112 expected), we are quite optimistic that our final sampling of 809 subjects will be more than adequate to test the primary hypotheses with excellent power. In the progress report section of this article, we discussed questions related to final data analysis. That such analyses are needed is well illustrated by difficulties in interpreting the findings of published reports of the Diabetes Prevention Program (DPP) group. Akin to the TROPHY protocol in which, for the first 2 years, we used a drug (candesartan) that pharmacologically suppresses the variable (BP) on which the diagnosis is made (hypertension), the DPP group22,23 used a drug (metformin) that lowers blood glucose and on which a diagnosis of diabetes depends. They first reported the incidence rates of diabetes during the active treatment period and concluded that treatment with metformin reduced the incidence of diabetes.22 However, these incidence rates of diabetes might alternatively be called rates of failure to control blood sugar with active treatment. In a second report,23 the authors addressed the issue of primary prevention by allowing a 1- to 2-week washout from placebo and the active drug in patients who did not develop diabetes during the trial. The clinical relevance of a short time of drug discontinuation is doubtful, and a comparison of such highly preselected groups is fraught with difficulties. A better way to study the issue of prevention is to stipulate that the washout period after treatment will last as long as the previous treatment period and then compare slopes of development rates of the disease (in our case, hypertension) throughout the entire study.
Clinical Implications
Clinical hypertension evolves from a background of long-standing slightly higher BP readings. The Framingham study reported that more subjects with high-normal BP tend to develop hypertension than people with lower-normal BP values.20,21 Records of periodic exams in the Tecumseh study17 show that subjects in the age range of 20 to 40 years with borderline hypertension (average BP 130.7/ 93.8 mm Hg) had significantly higher childhood (average age 6 years) BP levels than the rest of the normotensive population.
We believe that the time has come to test whether early treatment of prehypertension might ameliorate the natural history of subsequent hypertension. We are particularly encouraged by the recognition in the recent JNC 7 document2 that prehypertension represents a major public health problem. Furthermore, in the United States BP control rates have not increased, and the rates of cardiovascular consequences of hypertension do not continue to decrease. Under these circumstances, a fresh look at treatment practices in hypertension appears to be well warranted. The practice of lifelong treatment has evolved from observations that on discontinuation of treatment, BP increase is likely to return in advanced hypertension. This knowledge has then been applied to all forms of hypertension despite evidence that such a post-treatment BP rebound is less likely to occur in milder forms of hypertension.4–6 We believe, but cannot prove, that fear of receiving a "sentence" to lifelong treatment is a serious deterrent to early diagnosis and compliance with treatment in hypertension. The TROPHY study seeks only the proof of principle that early pharmacological treatment of prehypertension might delay or prevent development of clinical (stage 1) hypertension. If the TROPHY results confirm this hypothesis, further studies would be warranted to investigate how this affects the compliance with treatment and cardiovascular outcomes in hypertension.
|
Acknowledgments |
|---|
This investigator-initiated study is supported by a grant from AstraZeneca LP. S.J. is supported in part by the Fredrick G.L. Huetwell professorship in hypertension. The authors thank Terry Flanagan for his contribution to the planned statistical analysis. We also thank the staff at the study sites and the participants for their important contributions.
Received February 11, 2004; first decision February 25, 2004; accepted April 19, 2004.
|
References |
|---|
|
TopAbstractIntroductionStudy DesignProgress Report for Subjects...Proposed Data Analysis: Concepts...DiscussionReferences |
|---|
1. Julius S, Schork MA. Borderline hypertension—a critical review. J Chronic Dis. 1971; 23: 723–754.[CrossRef][Medline] [Order article via Infotrieve]
2. U.S. Department of Health and Human Services, National Institutes of Health. Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.NIH Publication No. 03–5233. May 2003.
3. Neaton JD, Kuller L, Stamler J, Wentworth DN. Impact of systolic and diastolic blood pressure on cardiovascular mortality. In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis and Management. 2nd ed, vol 1. New York, NY: Raven Press; 1995: 127–144.
4. Medical Research Council Working Party on Mild Hypertension. Course of blood pressure in mild hypertensive safer withdrawal of long term antihypertensive treatment. Br Med J. 1986; 293: 988–992.
5. Langford HG, Blaufox MD, Oberman A, Hawkins CM, Curb JD, Cutter GR, Wassertheil-Smoller S, Pressel S, Babcock C, Abernethy JD, Hotchkiss J, Tyler M. Dietary therapy slows the return of hypertension after stopping prolonged medication. JAMA. 1985; 253: 657–664.
6. Marland LJ, Lutz LJ, Castle CH. Effects of withdrawing diuretic therapy on blood pressure in mild hypertension. Hypertension. 1983; 5: 539–544.
7. Harrap SB, Mirakian C, Datodi SR, Lever AF. Blood Pressure and lifespan following brief ACE inhibitor treatment in young spontaneously hypertensive rats. Clin Exp Pharmacol Physiol. 1994; 21: 125–127.[Medline] [Order article via Infotrieve]
8. Harrap SB, Van Der Merwe WM, Griffin SA, Macpherson F, Lever AF. Brief angiotensin converting enzyme inhibitor treatment in young spontaneously hypertensive rats reduces blood pressure long-term. Hypertension. 1990; 16: 603–614.
9. Lee RM, Berecek KH, Tsoporis J, McKenzie R, Triggle CR. Prevention of hypertension and vascular changes by captopril treatment. Hypertension. 1991; 17: 141–150.
10. Wu JN, Berecek KH. Prevention of genetic hypertension by early treatment of spontaneously hypertensive rats with the angiotensin-converting enzyme inhibitor captopril. Hypertension. 1993; 22: 139–146.
11. Hypertension Prevention Trial Research Group. The hypertension prevention trial: three-year effects of dietary changes on blood pressure. Arch Intern Med. 1990; 150: 153–162.
12. Swales JD. Dietary sodium restriction in hypertension. In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis, and Management. 2nd ed, vol 1. New York, NY: Raven Press; 1995: 283–298.
13. Jennings G, Nelson L, Nestel P, Esler M, Korner P, Burton D, Bazelmans J. The effects of changes in physical activity on major cardiovascular risk factors, hemodynamics, sympathetic function, and glucose utilization in man: a controlled study of four levels of activity. Circulation. 1986; 73: 30–40.
14. Appel LJ, Moore TJ, Obarzanek E, Vollmer M, Svetkey L, Sacks FM, Bray GA, Vogt TM, Cutler JA, Windhauser MM, Lin PH, Karanja N, Simons-Morton D, McCullough M, Swain J, Steele P, Evans M, Miller ER, Harsha DW. A clinical trial of the effects of dietary patterns on blood pressure. The DASH Collaborative Research Group. N Engl J Med. 1997; 336: 1117–1124.
15. World Health Organization. Obesity: Preventing and Managing the Global Epidemic: Report of a WHO Consultation on Obesity. Geneva, Switzerland: World Health Organization; 1998.
16. Giudicelli JF, Freslon JL, Glasson S, Richer C. Captopril and hypertension development in the SHR. Clin Exp Hypertens. 1980; 2: 1083–1096.[CrossRef][Medline] [Order article via Infotrieve]
17. Julius S, Jamerson K, Meija A, Krause L, Schork N, Jones K. The association of borderline hypertension with target organ changes and higher coronary risk: Tecumseh blood pressure study. JAMA. 1990; 264: 354–358.
18. Palatini P, Julius S. Heart rate and the cardiovascular risk. J Hypertens. 1997; 15: 1–15.[Medline] [Order article via Infotrieve]
19. Trials of Hypertension Prevention Collaborative Research Group. Effects of weight loss and sodium reduction intervention on blood pressure and hypertension incidence in overweight people with high-normal blood pressure. Arch Intern Med. 1997; 157: 657–667.
20. Leitschuh M, Cupples LA, Kannel W, Gagnon D, Chobanian A. High-normal blood pressure progression to hypertension in the Framingham heart study. Hypertension. 1991; 17: 22–27.
21. Vasan RS, Larson MG, Leip EP, Evans JC, O’Donnell CJ, Kannel WB, Levy D. Impact of high-normal blood pressure on the risk of cardiovascular diseases. N Engl J Med. 2001; 345: 1291–1297.
22. Diabetes Prevention Program Research Group. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002; 346: 393–403.
23. Diabetes Prevention Program Research Group. Effects of withdrawal from metformin on the development of diabetes in the diabetes prevention program. Diabetes Care. 2003; 26: 977–980.
This article has been cited by other articles:
 |
|
 |
|
  J. A. Cutler, P. D. Sorlie, M. Wolz, T. Thom, L. E. Fields, and E. J. Roccella Trends in Hypertension Prevalence, Awareness, Treatment, and Control Rates in United States Adults Between 1988-1994 and 1999-2004 Hypertension, November 1, 2008; 52(5): 818 - 827. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. M. Egan, S. D. Nesbitt, and S. Julius Review: Prehypertension: should we be treating with pharmacologic therapy? Therapeutic Advances in Cardiovascular Disease, August 1, 2008; 2(4): 305 - 314. [Abstract] [PDF]
|
|
|
|
 |
|
 C. Y. Shim, J.-W. Ha, S. Park, E.-Y. Choi, D. Choi, S.-J. Rim, and N. Chung Exaggerated Blood Pressure Response to Exercise Is Associated With Augmented Rise of Angiotensin II During Exercise J. Am. Coll. Cardiol., July 22, 2008; 52(4): 287 - 292. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Julius, S. D. Nesbitt, B. M. Egan, M. A. Weber, E. L. Michelson, N. Kaciroti, H. R. Black, R. H. Grimm Jr., F. H. Messerli, S. Oparil, et al. Feasibility of Treating Prehypertension with an Angiotensin-Receptor Blocker N. Engl. J. Med., April 20, 2006; 354(16): 1685 - 1697. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. M. Kaplan Treatment of Hypertension: Remaining Issues After the Anglo-Scandinavian Cardiac Outcomes Trial Hypertension, January 1, 2006; 47(1): 10 - 13. [Full Text] [PDF]
|
|
|
|
|
|
N. M. Kaplan An Updated Meta-Analysis With a Few Surprises Hypertension, August 1, 2005; 46(2): 257 - 258. [Full Text] [PDF]
|
|
|
|
|
|
N. M. Kaplan Trial of Preventing Hypertension (TROPHY): Design and 2-Year Progress Report Hypertension, August 1, 2004; 44(2): 125 - 126. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||