AND DEBATES

Should We Treat Prehypertension in Diabetes? What are the cons?

EHUD GROSSMAN, MD When both risk factors were present in the same patient, the incidence rate in- creased to 48 per 1,000 (5). levated (BP), even uncomplicated hypertension to below Several clinical studies have indicated within the normal range, is associ- 140/90 mmHg and in the diabetic hyper- that diabetes is associated with cardio- E ated with cardiovascular (CV) mor- tensive patient to below 130/80 mmHg myopathy that is independent of ath- bidity and mortality. Therefore, the Joint (1,2). Solid evidence exists showing that erosclerotic coronary artery (6). National Committee (JNC) VII intro- the benefits of BP lowering are far more Congestive is substantially duced the term “prehypertension” in the pronounced in the diabetic than in the increased in diabetic patients irrespective general population, which is defined as nondiabetic hypertensive patient. In light of coronary artery disease and hyperten- BP levels of 120–139 mmHg and 80–89 of the benefits of BP lowering in diabetic sion (7). The Framingham study data re- mmHg for systolic and diastolic BP, re- patients, there is a dilemma as to whether vealed a fourfold greater incidence of spectively (1). Prehypertension includes diabetic patients with prehypertension congestive heart failure in diabetic men two different categories of BP: normal should be medically treated to lower BP. I and an eightfold increase in diabetic (systolic of 120–129 mmHg or diastolic will endeavor to analyze the available data women, compared with nondiabetic sub- BP of 80–84 mmHg) and high-normal to determine what therapeutic approach jects (8). In the DIGAMI (Diabetes Melli- (systolic of 130–139 mmHg or diastolic should be adopted for diabetic patients tus Insulin-Glucose Infusion in Acute BP of 85–89 mmHg). The risk of CV with prehypertension. ) trial, congestive events is increased by two- to fourfold heart failure accounted for up to 66% of with the coexistence of hypertension and RISK OF HYPERTENSION IN mortality during the first year postmyo- type 2 diabetes. Lowering BP is particu- DIABETES — Hypertension is a major cardial infarction in diabetic patients (9). larly effective in patients with type 2 dia- modifiable risk factor for CV morbidity Longstanding hypertension leads to betes. Therefore, guidelines recommend and mortality. Diabetes is associated with the development of cardiomyopathy, lowering BP to below 130/80 mmHg in a high risk of CV disease and is the leading which is associated with impaired cardiac diabetic patients. Thus, the term “prehy- cause of end-stage renal disease, blind- function (10). We showed that in hyper- pertension” is inadequate for patients ness, and nontraumatic amputations in tensive patients, contractility deteriorated with type 2 diabetes. It is clear from western countries (3). Although the ef- as left ventricular mass increased (11). A guidelines that in diabetic patients, the fects of diabetes and hypertension on the progressive decline in ventricular func- high-normal BP category of prehyperten- CV system vary somewhat, and are often tion may lead to congestive heart failure. sion should be pharmacologically treated. distinct, their combined presence in the Data from the Framingham study showed However, there is no evidence that drug same patient is destructive (4). that hypertension was the primary cause treatment is beneficial in the normal BP Coronary artery disease is far more of congestive heart failure in 35% of cases category of prehypertension. Therefore, common in diabetic hypertensive patients and played a role in this condition in an- despite the devastating effect of elevated than in patients suffering from hyperten- other 40% (12). BP in type 2 diabetes, drug treatment is sion or diabetes alone (5). For all 2,681 The coexistence of diabetes and hy- not always recommended for all diabetic men in the PROCAM trial who had none pertension results in more severe cardio- patients with prehypertension. of the three risk factors (i.e., hyperten- myopathy than would be expected with Hypertension is perhaps best defined sion, diabetes, or hyperlipidemia), the either hypertension or diabetes alone by the BP level that has a negative impact coronary artery disease incidence was (10). Clinical studies with echocardiogra- on the CV system. Thus, numerical defi- 6/1,000 over 4 years. In contrast, the in- phy also showed an increased left ventric- nitions, although hotly debated by nu- cidence of coronary artery disease in par- ular mass in diabetic hypertensive merous guideline committees, are not ticipants who were suffering from patients (13) increased septal and poste- helpful to practicing physicians. Recent hypertension or diabetes was 14 and 15 rior wall thickness in patients with hyper- guidelines set the target level of BP for per 1,000 over 4 years, respectively. tension and diabetes, compared with ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● nondiabetic hypertensive patients (13). From the Internal Medicine Department D and the Hypertension Unit, The Chaim Sheba Medical Center, Tel Prevalence of left ventricular hypertrophy Hashomer, and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel. Corresponding author: Ehud Grossman, [email protected]. was 72% in diabetic hypertensive patients The publication of this supplement was made possible in part by unrestricted educational grants from Eli and only 32% in the nondiabetic hyper- Lilly, Ethicon Endo-Surgery, Generex Biotechnology, Hoffmann-La Roche, Johnson & Johnson, LifeScan, tensive patients who had a similar degree Medtronic, MSD, Novo Nordisk, Pfizer, sanofi-aventis, and WorldWIDE. of hypertension. Because left ventricular DOI: 10.2337/dc09-S324 © 2009 by the American Diabetes Association. Readers may use this article as long as the work is properly hypertrophy is known to predispose pa- cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons. tients with hypertension to CV morbid org/licenses/by-nc-nd/3.0/ for details. and fatal events, the finding of a high

S280 DIABETES CARE, VOLUME 32, SUPPLEMENT 2, NOVEMBER 2009 care.diabetesjournals.org Grossman prevalence of left ventricular hypertrophy tality (20). The relationships between BP to initiate antihypertensive treatment in in diabetic hypertensive patients may par- and mortality exist over a wide BP range, diabetic patients with diabetic prehy- tially explain their increased morbidity starting from 115/75 mmHg. On this ba- pertension. and mortality. Cardiomyopathy of diabe- sis, the JNC VII introduced a new cate- tes and hypertension is associated with gory of “prehypertension.” This category TREATMENT OF impaired ventricular function and a high is defined as a systolic BP level of 120– PREHYPERTENSION — In the prevalence of congestive heart failure 139 mmHg and/or diastolic BP level of general population, there are no outcome (10). 80–89 mmHg. Several studies showed studies showing any benefit of drug treat- Diabetes is one of the leading causes that “prehypertension” is common and is ment in prehypertension. Only two stud- for end-stage renal disease (ESRD) (14). associated with the metabolic syndrome ies evaluated the efficacy of drug Hypertension is a well-defined risk factor and other CV risk factors (25,26), such as treatment in prehypertension (30,31). for end-stage renal disease and accounts obesity, elevated triglycerides, elevated The Trial of Preventing Hypertension for 27% of all end-stage renal disease LDL cholesterol, and low levels of HDL (TROPHY) study investigated whether cases in the U.S. and 33.4% of end-stage cholesterol. Furthermore, during follow- pharmacologic treatment of prehyperten- renal disease cases among African Ameri- up, subjects with prehypertension are sion prevents or postpones stage 1 hyper- cans (14). more susceptible to developing true hy- tension (30). A total of 809 subjects with When hypertension is superimposed pertension and coronary atherosclerosis high-normal BP (high prehypertension) on diabetes, it accelerates the decrease in (25,27). Prehypertension includes two were randomly assigned to receive 2 years renal function. Blood pressure control different categories of BP that are used by of either candesartan (409 subjects) or can slow the progression of renal disease the European Society of Hypertension: placebo (400 subjects), followed by 2 in diabetic patients (15). normal BP (systolic 120–129 mmHg, or years of placebo for everyone. All subjects Diabetes adversely affects cerebrovas- diastolic 80–84 mmHg) and high- were instructed to change their lifestyles cular arterial circulation. The risk of normal BP (systolic 130–139 mmHg or to reduce BP. During the first 2 years, can- is increased by 150–400% for pa- diastolic 85–89 mmHg) (2). Grotto et al. desartan reduced the risk of incident hy- tients with diabetes (16). In the Multiple (26) showed that subjects with high pre- pertension by 66.3% (P Ͻ 0.001); Risk Factor Intervention Trial, subjects hypertension, which is equivalent to hypertension had developed in 154 sub- taking medications for diabetes were high-normal BP, have elevated levels of jects in the placebo group and 53 of those three times as likely to suffer a stroke (17). glucose, total cholesterol, triglycerides, in the candesartan group. After 4 years, In particular, diabetes increases the risk of and BMI and lower levels of HDL choles- candesartan reduced the risk of incident stroke among younger patients. The prev- terol than those with low prehypertension hypertension by 15.6% (P Ͻ 0.007); hy- alence of diabetes increases the risk of equivalent to normal BP. Vasan et al. (28) pertension had developed in 240 subjects stroke-related dementia more than three- showed that the risk for CV disease is 2.5- in the placebo group and 208 of those in fold (18), doubles the risk of recurrence, and 1.6-fold higher among women and the candesartan group. In the recent Pre- and increases total and stroke-related men, respectively, with high-normal BP vention of Hypertension study, using the mortality (19). than in those with optimal BP (Ͻ120/80 ACE inhibitor ramipril in patients with Hypertension, mainly systolic, is mmHg). Thus, prehypertension is associ- high-normal BP (PHARAO), a total of strongly and directly related to stroke in ated with other metabolic abnormalities 1,008 subjects with high-normal office BP all age-groups (20), and lowering BP re- and increased CV risk. Within the prehy- were randomized to treatment with either duces the rate of stroke remarkably (21). pertension group, there is further stratifi- ramipril (n ϭ 505) or placebo (n ϭ 503) The occurrence of diabetes more than cation into two risk categories: normal and were followed up for 3 years (31). doubles the risk of stroke in hypertensive and high-normal BP. Treatment with ramipril reduced the risk patients (22), and lowering BP in these of progression to hypertension by 34.4% patients reduces the risk of stroke by 44% DIABETIC (155 subjects with ramipril vs. 216 sub- (23). PREHYPERTENSION — In pa- jects with placebo). Despite the reduction Diabetes may cause diabetic reti- tients with type 2 diabetes, elevated BP is in progression to hypertension, ramipril nopathy that is characterized by neovas- more harmful than in nondiabetic sub- failed to reduce CV events and death. cularization and formation of jects. There is clear evidence that lowering Both studies succeeded in showing that microaneurysms. Hypertension acceler- BP is more beneficial in diabetic than in blockers of the renin angiotensin system ates the development of diabetic retinop- nondiabetic patients. Aggressive lowering reduce progression to hypertension, but athy. Knowler et al. (24) found that in of BP is beneficial in type 2 diabetes, even they did not show a reduction of CV diabetic subjects not taking insulin, the in patients without hypertension. High- events. Further long-term studies with incidence of exudates in those with sys- normal BP or high prehypertension is additional antihypertensive agents are re- tolic BP of Ͼ145 mmHg was more than considered hypertension in type 2 diabe- quired to evaluate whether pharmacologi- twice that of those with pressures of tes and requires antihypertensive treat- cal treatment can improve clinical Ͻ125 mmHg. The combination of hyper- ment. Regarding this issue, elsewhere we outcomes in patients with prehypertension. tensive and diabetic retinopathy is often suggested that diabetic prehypertension devastating and remains one of the lead- should be defined as systolic BP of 110– TREATMENT OF DIABETIC ing causes of blindness. 129 mmHg and/or diastolic BP of 70–79 PREHYPERTENSION — Several mmHg (29). studies evaluated the effect of pharmaco- PREHYPERTENSION — A recent The question, therefore, should not logical treatment in diabetic patients with meta-analysis showed that casual BP is be whether to treat prehypertension in normal BP (32,33). In the normotensive strongly associated with age-specific mor- patients with type 2 diabetes, but whether Appropriate Blood Pressure Control in care.diabetesjournals.org DIABETES CARE, VOLUME 32, SUPPLEMENT 2, NOVEMBER 2009 S281 Prehypertension in diabetes

Diabetes (ABCD) study (33), 480 type 2 mmHg is marginal. Thus, it seems that disease: an update. Hypertension 2001;4: diabetic patients with baseline normal BP with the present evidence, it would be un- 1053–1059 (Ͻ140/90 mmHg) were randomized to justified to recommend drug treatment in 5. Assmann G, Schulte H: The Prospective intensive (10 mmHg below the baseline diabetic prehypertension. Cardiovascular Munster (PROCAM) diastolic BP) or moderate (80–89 study: prevalence of hyperlipidemia in mmHg) diastolic BP control. Over a persons with hypertension and/or diabe- CONCLUSIONS tes mellitus and the relationship to coro- 5-year follow-up period, intensive BP — The definition of prehypertension in type 2 diabetes is dif- nary heart disease. Am Heart J 1988;116: control (average of 128/75 mmHg) was 1713–1724 associated with less progression to incip- ferent from that in the general population. High-normal BP (high prehypertension) 6. Blendea MC, McFarlane SI, Isenovic ER, ient or overt diabetic nephropathy, less Gick G, Sowers JR: Heart disease in dia- progression to diabetic retinopathy, and is considered hypertension in type 2 dia- betes and requires drug treatment. How- betic patients. Curr Diab Rep 2003;3: less incidence of stroke than moderate 223–229 (137/81 mmHg) BP control. In the recent ever, diabetic-prehypertension requires 7. Gustafsson I, Hildebrandt P. Early failure Action in Diabetes and Vascular disease lifestyle modification and not pharmaco- of the diabetic heart. Diabetes Care 2001; preterAx and diamicorN MR Controlled logical treatment. Long-term studies with 24:3–4 Evaluation (ADVANCE) trial, 11,140 antihypertensive treatment in diabetic 8. Kannel WB, Hjortland M, Castelli WP. patients with type 2 diabetes were ran- prehypertension will be beneficial in Role of diabetes in congestive heart fail- domized to treatment with a fixed com- teaching us whether or not to modify our ure: the Framingham study. Am J Cardiol current approach. 1974;34:29–34 bination of perindopril and indapamide 9. Malmberg K, Ryden L, Efendic S, Herlitz or matching placebo (32). After a mean of J, Nicol P, Waldenstrom A, Wedel H, We- 4.3 years of follow-up, active treatment Acknowledgments— No potential conflicts lin L. Randomized trial of insulin-glucose (BP 136/73 mmHg) reduced the relative of interest relevant to this article were infusion followed by subcutaneous insu- risk of a major macrovascular or micro- reported. lin treatment in diabetic patients with vascular event by 9%, compared with the acute myocardial infarction (DIGAMI placebo treatment (BP 140/73 mmHg). study): effects on mortality at 1 year. J Am The authors stated that the study treat- Coll Cardiol 1995;26:57–65 ment was not affected by the initial BP References 10. Grossman E, Messerli FH. Diabetic and 1. Chobanian AV, Bakris GL, Black HR, hypertensive heart disease. Ann Intern levels. However, the mean initial BP of the Cushman WC, Green LA, Izzo JL, Jr, studied population was 145/81 mmHg, Med 1996;125:304–310 Jones DW, Materson BJ, Oparil S, Wright 11. Grossman E, Oren S, Messerli FH. Left which is clearly hypertension in type 2 JT Jr, Roccella EJ.The Seventh Report of diabetes, and 7,655 (68.5%) patients had ventricular mass and cardiac function in the Joint National Committee on Preven- patients with . J a history of current antihypertensive tion, Detection, Evaluation, and Treat- Hum Hypertens 1994;8:417–421 treatment. Moreover, analysis of sub- ment of High Blood Pressure: the JNC 7 12. Kannel WB, Castelli WP, McNamara PM, report. JAMA 2003;289:2560–2572 groups revealed that in patients with no McKee PA, Feinleib M. Role of blood pres- 2. Mancia G, De Backer G, Dominiczak A, history of hypertension, active treatment sure in the development of congestive Cifkova R, Fagard R, Germano G, Grassi did not reduce CV events. heart failure: the Framingham study. G, Heagerty AM, Kjeldsen SE, Laurent S, N Engl J Med 1972;287:781–787 Narkiewicz K, Ruilope L, Rynkiewicz A, 13. Grossman E, Shemesh J, Shamiss A, Tha- SHOULD WE USE Schmieder RE, Boudier HA, Zanchetti A, ler M, Carroll J, Rosenthal T. Left ventric- PHARMACOLOGICAL Vahanian A, Camm J, De Caterina R, Dean TREATMENT IN DIABETIC V, Dickstein K, Filippatos G, Funck-Bren- ular mass in diabetes-hypertension. Arch PREHYPERTENSION? — There is tano C, Hellemans I, Kristensen SD, Intern Med 1992;152:1001–1004 no evidence that antihypertensive treat- McGregor K, Sechtem U, Silber S, Ten- 14. Lea JP, Nicholas SB. Diabetes mellitus and ment is beneficial in patients with diabetic dera M, Widimsky P, Zamorano JL, Er- hypertension: key risk factors for kidney dine S, Kiowski W, Agabiti-Rosei E, disease. J Natl Med Assoc 2002;94 prehypertension. Data from the U.K. Pro- (Suppl. 8):7S–15S spective Diabetes Study (UKPDS)-23 Ambrosioni E, Lindholm LH, Viigimaa M, Adamopoulos S, Agabiti-Rosei E, Ambro- 15. American Diabetes Association. Stan- showed that a 10-mmHg decrease in sys- dards of medical care in diabetes: 2008. tolic BP was associated with a 12% reduc- sioni E, Bertomeu V, Clement D, Erdine S, Farsang C, Gaita D, Lip G, Mallion JM, Diabetes Care 2008;31 (Suppl. 1):S12– tion in risk of any complication related to Manolis AJ, Nilsson PM, O’Brien E, Poni- S54 type 2 diabetes. The lower the systolic BP, kowski P, Redon J, Ruschitzka F, Tama- 16. Folsom AR, Rasmussen ML, Chambless the lower the risk of complications, and rgo J, van Zwieten P, Waeber B, Williams LE, Howard G, Cooper LS, Schmidt MI, no threshold of systolic BP was observed B, 2007 Guidelines for the Management Heiss G. Prospective associations of fast- for a substantive change in risk for any of of Arterial Hypertension. The Task Force ing insulin, body fat distribution, and di- the outcomes examined. These findings for the Management of Arterial Hyperten- abetes with risk of ischemic stroke: the do not necessarily mean that we need to sion of the European Society of Hyperten- Atherosclerosis Risk in Communities force BP treatment to a goal Ͻ130/80 sion (ESH) and of the European Society of (ARIC) Study Investigators. Diabetes Care 1999;22:1077–1083 mmHg. When considering lowering the (ESC). J Hypertens 2007;25: 17. Jorgensen H, Nakayama H, Raaschou HO, BP goal in type 2 diabetes, and medically 1105–1187 3. Sowers JR. Recommendations for special Olsen TS. Stroke in patients with diabe- treating patients with diabetic prehyper- populations: diabetes mellitus and the tes: the Copenhagen Stroke Study. Stroke tension, the cost and benefit of drug treat- metabolic syndrome. Am J Hypertens 1994;25:1977–1984 ment should be taken into account. It 2003;11:41S–45S 18. Luchsinger JA, Tang MX, Stern Y, Shea S, seems that the benefit, if any, of lowering 4. Sowers JR, Epstein M, Frohlich ED. Dia- Mayeux R. Diabetes mellitus and risk of BP in diabetic patients below 130/80 betes, hypertension, and cardiovascular Alzheimer’s disease and dementia with

S282 DIABETES CARE, VOLUME 32, SUPPLEMENT 2, NOVEMBER 2009 care.diabetesjournals.org Grossman

stroke in a multiethnic cohort. Am J Epi- six-year follow-up study in Pima Indians. blocker. N Engl J Med 2006;354:1685– demiol 2001;154:635–641 N Engl J Med 1980;302:645–650 1697 19. Tuomilehto J, Rastenyte D, Jousilahti P, 25. Grossman A, Grossman C, Barenboim E, 31. Luders S, Schrader J, Berger J, Unger T, Sarti C, Vartiainen E. Diabetes mellitus as Azaria B, Goldstein L, Grossman E. Pre- Zidek W, Bohm M, Middeke M, Motz W, a risk factor for death from stroke: pro- hypertension as a predictor of hyperten- Lubcke C, Gansz A, Brokamp L, spective study of the middle-aged Finnish sion in military aviators: a longitudinal Schmieder RE, Trenkwalder P, Haller H, population. Stroke 1996;27:210–215 study of 367 men. Aviat Space Environ Dominiak P. The PHARAO study: preven- 20. Lewington S, Clarke R, Qizilbash N, Peto Med 2006;77:1162–1165 tion of hypertension with the angiotensin- R, Collins R. Age-specific relevance of 26. Grotto I, Grossman E, Huerta M, Sharabi converting enzyme inhibitor ramipril in usual blood pressure to vascular mortal- Y. Prevalence of prehypertension and as- patients with high-normal blood pres- ity: a meta-analysis of individual data for sociated cardiovascular risk profiles sure: a prospective, randomized, con- one million adults in 61 prospective stud- among young Israeli adults. Hypertension trolled prevention trial of the German ies. Lancet 2002;360;1903–1913 2006;48:254–259 Hypertension League. J Hypertens 2008; 21. Staessen JA, Gasowski J, Wang JG, Thijs 27. Pletcher MJ, Bibbins-Domingo K, Lewis 26:1487–1496 L, Den Hond E, Boissel JP, Coope J, Ek- CE, Wei GS, Sidney S, Carr JJ, Vittinghoff 32. Patel A, MacMahon S, Chalmers J, Neal B, bom T, Gueyffier F, Liu L, Kerlikowske K, E, McCulloch CE, Hulley SB. Prehyper- Woodward M, Billot L, Harrap S, Poulter Pocock S, Fagard RH. Risks of untreated tension during young adulthood and cor- N, Marre M, Cooper M, Glasziou P, and treated isolated systolic hypertension onary calcium later in life. Ann Intern in the elderly: meta-analysis of outcome Med 2008;149:91–99 Grobbee DE, Hamet P, Heller S, Liu LS, trials. Lancet 2000;355:865–872 28. Vasan RS, Larson MG, Leip EP, Evans JC, Mancia G, Mogensen CE, Pan CY, Rodg- 22. Grossman E, Messerli FH, Goldbourt U. O’Donnell CJ, Kannel WB, Levy D. Im- ers A, Williams B. Effects of a fixed com- High blood pressure and diabetes melli- pact of high-normal blood pressure on the bination of perindopril and indapamide tus: are all antihypertensive drugs created risk of . N Engl on macrovascular and microvascular out- equal? Arch Intern Med 2000;160:2447– J Med 2001;345:1291–1297 comes in patients with type 2 diabetes 2452 29. Leibowitz A, Grossman E. How to define mellitus (the ADVANCE trial): a random- 23. U.K. Prospective Diabetes Study Group. prehypertension in diabetes/metabolic syn- ised controlled trial. Lancet 2007;370: Tight blood pressure control and risk of drome. Diabetes Care 2009;32 (Suppl. 2): 829–840 macrovascular and microvascular com- S275–S279 33. Schrier RW, Estacio RO, Esler A, Mehler plications in type 2 diabetes: UKPDS 38. 30. Julius S, Nesbitt SD, Egan BM, Weber MA, P.Effects of aggressive blood pressure BMJ 1998;317:703–713 Michelson EL, Kaciroti N, Black HR, control in normotensive type 2 diabetic 24. Knowler WC, Bennett PH, Ballintine Grimm RH Jr, Messerli FH, Oparil S, patients on albuminuria, retinopathy EJ. Increased incidence of retinopathy in Schork MA. Feasibility of treating prehy- and . Kidney Int 2002;61:1086– diabetics with elevated blood pressure: a pertension with an angiotensin-receptor 1097

care.diabetesjournals.org DIABETES CARE, VOLUME 32, SUPPLEMENT 2, NOVEMBER 2009 S283