Reprinted from the Journal of the American Medical Association August 17, 1970 Volume 213 -Copyright 1970, American Medical Association Elects of Treatment on Morbidity in Hypertension II. Results in Patients With Diastolic Blood Pressure Averaging 90 Through 114 rn~m Hg Veterans Administration Gooperative Study Group on Antihypertensive Agents Three hundred and eighty male hypertensive patients with diastolic blood pressures averaging 90 to 114 mm Hg were randomly assigned to either active antihypertensive agents or placebos. The estimated risk of developing a morbid event over a five-year period was reduced from 55% to 18% by treatment. Terminating morbid e-vents occurred in 35 patients of the control group as compared to 9patients in the treated group. Nineteen deaths related to hypertension or at herosclerosis occurred in the control group and 8 in the actively treated group. In addition to morbid events, 20 control patients developed persistent diastolic levels of 125 mm Hg or higher. Treatment was more effective in prevent- ing congestive heart failure and stroke than in preventing the complications of coronary artery disease. The degree of benefit was related to the level of prerandomization blood pressure. I n a previous publication in this journal' the Veterans Admin- istration Cooperative Study For complete list of participants, see page 1152. Reprint requests to 50 Irving St NW, Washington, DC 20422 (Dr. Freis). JAMA, Aug 17, 1970 o ??o 213, No 7 Group on Antihypertensive Agents reported on the beneficial effects of antihypertensive drugs on morbidi- ty in patients with moderately severe hypertension. These were patients with initial diastolic blood pressures averaging 115 through 129 mm Hg who had been randomized into a prospective double-blind trial of active antihypertensive drugs vs placebos. Twenty-seven patients de- veloped assessable events in the control group as compared to two patients in the group receiving ac- tive antihypertensive agents. This striking result favoring treatment I was in agreement, with the results of other prospective trialF in pa- tients with hypertension of similar severity. In hypertension of lesser severity, however, there are little or no con- trolled data available on the value of antihypertensive drug therapy. Resolution of this question is of great importance not only because of the large number of patients with mild hypertension but also because the potential benefits of drug treat- ment have been questioned espe- cially in this group of hypertensive patients.' The present report pre- sents the results of a prospective, controlled trial of drug treatment on morbidity and mortality in a group of 380 patients with mild or moderate hypertension whose initial Morbidity in Hypertension 1143 diastolic blood pressure averaged 90 through 114 mm Hg. Plan of investigation The clinical trial included 523 male veterans who, while not re- ceiving antihypertensive treatment, exhibited diastolic blood pressures averaging 90 through 129 mm Hg. Randomization of patients began in! April 1964. However, in May 1967, the study was terminated in the subgroup of 143 patients whose diastolic blood pressures averaged 115 through 129 mm Hg prior to randomization. Termination of the study of this group as previously re- ported' was necessitated by the high incidence of morbid events in the control as compared to the treated patients, demonstrating at a relatively early date a highly sig- nificant (P < 0.001) effect of treat- ment. Such a significant difference was not evident at the time, how- ever, in the patients whose diastolic blood pressures averaged below 115 mm Hg prior to randomization. These latter patients were con- tinued in the randomized trial until 1969 and are the subject of the present communication. The experimental design has been described in previous reports.`s5 Initially all patients were hospital- ized for diagnosis and evaluation of the severity of their hyperten- sion. Patients whose diastolic blood pressure averaged 90 through 129 mm Hg during the fourth through sixth hospital day were accepted for further follow-up. Patients whose diastolic averages fell below 90 mm Hg or rose above 129 mm Hg dur- ing this period of hospitalization were excluded. Following hospitalization the pa- tients entered a prerandomization observation period of two to four months' duration during which time they received placebos of antihy- pertensive agents. The patients whose diastolic blood pressures dur- ing the last two clinic visits of the observation period averaged 90 through 129 mm Hg were entered into the trial, providing there were no other reasons for exclusioh. Blood pressure was measured by a physician with the patient in a sitting position. Other reasons for excluding pa- tients from the trial, in addition to diastolic blood pressure, are detailed in other reports.`,' Such reasons included a history of a severe hy- pertensive complication such as a cerebral or subarachnoid hemor- rhage, hypertensive neuroretinop- athy, dissecting aneurysm, or renal failure, but did not include athero- sclerotic complications such as cor- onary artery disease or cerebrovas- cular thrombosis. Also excluded were (1) patients with surgically curable hypertension, (2) with un- related fatal diseases such as malig- nant tumors, (3) those unwilling or unable to return to clinic, and (4) poorly motivated or otherwise un- cooperative or unreliable patients. The outpatient prerandomization observation period provided a fur- ther opportunity to check on the reliability of the patients. Ribo- flavin, which produces bright yellow fluorescence of the urine, was incor- porated in the placebos. At each clinic visit a urine specimen was examined under ultraviolet ?ight. In addition, pill counts were made at each clinic visit. No patient was accepted into the randomized trial unless the urine exhibited fluores- cence and the pill counts were with- in a stipulated range, at each of two successive visits during the preran- domization observation period. Accepted patients were then ran- domly assigned double-blind to either active drugs or placebos. Active drugs consisted of two types of tablets, one being a combination tablet containing 50 mg hydrochlo- rothiazide and 0.1 mg reserpine which was given twice daily. The other was 25 mg of hydralazine hydrochloride given three times 1144 JAMA, Aug 17, 1970 . Vol 213, No 7 daily. The latter medication was raised to 50 mg three times daily if the diastolic blood pressure re- mained at 90 mm Hg or higher. Obviously, practically all of the patients in the placebo group had their "doses" raised to this level. Provision was made for reduction of doses if hypotensive reactions or other disturbing side effects oc- curred. Patients in the control group received placebos identical in taste and appearance to the active drugs. Indicated symptomatic treatment, including drugs other than antihy- pertensive agents, was permitted in all patients. Postrandomization clinic visits were at monthly intervals for the first two months and at bimonthly intervals thereafter. Annual exami- nations included taking a history and a physical examination, roent- genogram of the chest, electro- cardiogram, pertinent chemical an- alyses of the blood, and renal func- tion tests. Additional interim visits could be scheduled when indicated. Characteristics of Patients Three hundred and eighty pa- tients with diastolic blood pressure;: averaging 90 through 114 mm Hg were randomized into the trial. Of this number, 186 received active drugs while 194 were given place- bos. Tables 1 and 2 indicate that the two groups were comparable according to the indicated variables. The median ages were 49.2 and 48.1 years and the average ages were 52.0 and 50.5 years in the con- trol and treatment groups, respec- tively. Negro patients comprised 42% of the control group and 417" of the treated group. Blood pressure as measured in the clinic during the posthospitalization observation period prior to randomization aver- aged 165.1/104.7 mm Hg in the control group and 162.1/103.8 mm Hg in the treated patients. There were no significant differences be- tween the control and treated pa- Morbidity in Hypertension tients with regard to fmdings from renal function tests, fasting blood sugar value, serum cholesterol value, uric acid level, and left ventricular enlargement as assessed by x-ray films and electrocardiography. By all factors measured the two groups were comparable. Table l.-Background of Randomized Patients: Numeration Data Characteristic Total randomized Negro Other" Heart size by roenJgenogram Ungerleider enlarged Electrocardiogram Left ventricular hypertrophy Control Group Treatment Group -No. n n Total ,- I- 194 186 380 ~- Si 42 76 41 157 . 114 58 109 59 223 42 22 53 29 95 32 16 30 16 62 Duration of Observation *In addition to whites, this group includes four patients of Asiatic extraction, two in the con- trol group and two in the treated group. Patients were entered into the trial from April 1964 to September 1968, and the study was terminated in October 1969. Thus, the earliest entrants were observed for 5.5 years and the latest entrants for a mini- mum of 1 year. The average poten- tial duration of observation, disre- garding losses and terminations, was 3.9 years for the control group and 3.7 years for the treated pa- tients. However, because of the losses and terminations due to ele- vated diastolic blood pressure de- scribed below, the actual duration of postrandomization observation was 3.3 years for the control group and 3.2 years for the treated pa- tients. Table 2.-Measurement Data Prior to Randomization Characteristic Atie (Y r) Age (median, yr) :; * Height, cm (ft, in) I-4) Weight, kg (lb) ,I Duration known hypertension (yr) (m&100 cc) Average hospital diastolic pressure (mm Hg) Average hospital systolic pressure (mm tig) A..^_^__ I:-:_ L:_-A-I:- --_--..-- I--- .I-\ nvarc_g;r Lllrlll; alabl"11c pressure Irnr" ng, Average clinic systolic pressure (mm Hi) Total severity SCOI Renal score (O-4 Cardiac score (( CNSt score (O-4 Serum creatinine I"4.1 52.0 I"3.6 165.1 50.5 162.1 49.2 6.7 0.2 48.1 6.8 0.2 175.3 (5,9) 0.8 l72.7 (5.8) 0.9 82.0 (180.9) 0.3 79.8 (176.1) 0.3 4.4 4.6 1.26 101.3 1.24 156 100.2 167 157.5 154.0 .,.* . .^^ ^ Changes in Blood Pressure *Detailed criteria for grades 0 through 4 given in reference 6. tCNS signifies central nervous system. $PSP signifies phenolsulfonphthalein. Systolic and diastolic blood pres- sure fell promptly and significantly in the treated patients and re- mained at reduced levels through- out the trial. The changes in blood pressure at the fourth month of observation in the treated and con- trol patients are depicted in Fig 1. The mean change in systolic blood pressure was an increase of 4.2 mm Hg in the control group and a fall of 27.2 mm Hg in the treated pa- tients from the levels recorded dur- ing the prerandomization observa- tion period. The mean change in diastolic blood pressure was a rise of 1.2 mm Hg in the control pa- tients and a fall of 17.4 mm Hg in the treated group during this same interval. The distribution of the changes in blood pressure as shown in Fig 1 indicates a marked shift to the left into the "decrease" zone for the treated patients as compared to the control group. Also apparent is the wide variation in individual responses particularly with regard to systolic blood pressure. laxis. Postmortem examination was carried out in both of these patients. Losses Other Than Assessable Events Deaths Due to Unrelated Condi- tiom-Four patients died of dis- orders unrelated to hypertension. Two of the patients were in the control group. One died of general- ized carcinomatosis demonstrated at autopsy and the other of ure-mia secondary to carcinoma of the uri- nary bladder. One patient in the treated group died of a subdural hematoma following a skull fracture and another of penicillin anaphy- Losses Due to Drug Toxicity.- Two patients in the treatment group developed reactions thought to be due to drug toxicity. The first pa- tient developed orbital edema with fever and malaise. Roentgenogrnm of the chest revealed infiltrates in the lungs. There was no dermatitis or arthritis. Lupus cells were not found in the blood although the antinuclear antibody test was pos- itive. Protocol drugs were discon- tinued because of the possibility of lupus syndrome associated with hydralazine. The second patient de- veloped purpura one month after beginning active drug treatment. Findings from examination in the hospital, including biopsy, were con- JAMA, Aug 17, 1970 . Vol 213, No 7 Morbidity in Hypertension 1145 PLACEBOS `ji DECREASE ACTIVE DRUGS -32 -16 0 +I6 +32 +a0 DECREASE -48 -32 -16 0 +16 +32 +48 ACTIVE DRUGS Ti i 1. Changes in systolic (left) and diastolic blood pressure (right) after four months of treatment in patients given placebos (top) and in patients treated with active drugs (bottom). Mean of changes (x). sistent with anaphylactoid purpura. The purpuric lesions cleared two weeks after protocol treatment was discontinued and reappeared within three days after administration of active drugs began again. Protocol treatment was, therefore, discon- tinued. Drop-Outs.-Fifty-six or 15% of the 330 randomized patients were classified as drop-outs during the course of the trial. Of this number 27 had been randomized to receive placebos and 29 to receive active drugs. The average period of follow- up prior to dropping out was 17.6 months with a range from less than 1 month to 49 months. Six patients moved away from the area of the clinic. Two were lost from follow-up because of closure of one participat- ing clinic. Four returned to the care of their private physicians. Fifteen complained of side effects prior to dropping out. Nine of these patients had been receiving drugs, and six were taking placebos. Five patients had psychiatric or alcoholic prob- lems of such severity as to make continued protocol treatment im- practical. In the remaining patients the reason for drop-out could not be determined. It should be noted that three of the patients taking pla- cebos sustained nonterminating morbid events prior to their drop- ping out. Assessable Morbid Events electrocardiographic signs of left ventricular hypertrophy or of roent- genographic evidence of cardiac en- largement, which will be reported in a subsequent communication. All available data pertaining to each organic complication, except the type of protocol treatment and the level of blood pressure, were pre- sented to the reviewers and their decisions regarding the occurrence and classification of an event ac- cording to the definitions given in the protocol (see list of assessable events at the end of the communi- cation) were accepted as final. The records of the patients re- Table 3 summarizes the assessa- ported as having assessable morbid ble events by major categories. Such events were reviewed by two con- events occurred in 98 of the 380 sulting physicians who had not randomized patients, 76 in the con- participated in the trial. All assessa- trol group and 22 in the treated ble events were reviewed except patients. Of this number 20 control those related to the development of patients developed an increase in 1146 JAMA, Aug 17, 1970 . Vol 213, No 7 , Morbidity in Hypertension diastolic blood pressure to levels exceeding 124 mm Hg on three separate clinic visits and persisting for 3 weeks or longer. Since these patients were removed from the trial only because of persistent blood pressure elevations and not for an organic complication, they will not be included in the subse- quent assessment of effectiveness of treatment in preventing morbid events. The remaining 78 patients had organic complications subdivided as follows: 56 of 194 or 28.9% of the control group and 22 of 186 or 11.8% of the treated patients. The most striking evidence of benefit of treatment was manifested in the count of class A events (hyperten- sive complications defmed in the protocol which required removal of the patient from the study.' There were none among the treated pa- tients but 14 among the controls. These included five class A deaths (Table 4) plus nine other class A events (Table 5). When other car- diovascular (class B) deaths and treatment failures were added, the comparisons were still impressive, 35 of 194 patients or lB.Ooj, amongst the controls and only 9 of 186 or 4.8% in the treated group (Table 3). The effectiveness of treatment (difference in percent incidence of complications between control and treated groups divided by the per- cent incidence in the control group) in preventing terminating organic complications was 73% (Table 3). The decision to discontinue the trial was based on this favorable evidence supplemented by the lifetable an- alyses described below which sug- gested that the benefit of treatment was continuing through time and was not solely concentrated in the first year or two of treatment. Terminating Events. - DEATHS RELATED TO CARDIOVASCULAR Drs- EAsE.-Twenty-seven patients died of hypertensive or atherosclerotic complications, 19 occurring in the Table 3.-Summary of Assessable Events Control Group Treated Group t NO. % , No Effectkerr* Terminating morbid eventot 35 18.0 9. 4.6 73 Nonterminating B events 21 13 Total morbid events 56 28.9 22 ii.8 59 Terminated on account of elevated blood pressure 20 0 Total assessable events 76 39.2 22 ii.8 70 No. patients randomized 194 100.0 186 100.0 *See text. tlncludes cardiovascular deaths. class A events, and treatment failures except those due to diastolic levels >124 mm Hg. Table 4.-Causes of Death Cause Treated Group Deaths due to class A events Cerebrovascular hemorrhage Subarachnoid hemorrhage 3 0 1 0 Dissecting aneurysm 1 0 Deaths due to class B events Myocardial infarction 3 2 Sudden death 8 4 Cerebrovascularthrombosis 3 1 _ Ruptured atherosclerotic aneurysm 0 1 Total related deaths* 19 8 *Does not include four unrelated deaths, two in the control group and two in the treated group (see text). Table 5.-Terminating Morbid Events Other Than Death Type of Event Class A events Uncontrolled cardiac failure Dissecting aortic aneurysm Subarachnoid hemorrhage Fundi, striate hemorrhages Acute hyoertensive enceohalooathv 5 0 1 0 1 0 1 0 1 0 _. . . _ Subtotal 9 0 Treatment failures Cerebrovascular thrombosis, severe 4 0 Progressive azotemia 1 0 Fundi. one striate hemorrhage and ? early papilledema Fundi, one striate hemorrhage and ? encephalopathy 1 0 1 0 Hypotension 0 1 Subtotal 7 1 Total 16 1 control group and 8 in the treated associated predominantly with cor- patients (Table 4). Five deaths onary artery disease. Eleven pa- associated with class A or hyper- tients in the placebo group and 6 in tensive events (see list of assessable the treated group had either a docu- events at the end of the communi- mented myocardial infarction or a cation) were cerebral hemorrhage "sudden death." Cerebrovascular in four and dissecting aortic an- thrombosis as opposed to hemor- eurysm in one, all occurring in the rhage was the cause of death in control group of patients. Deaths three control patients and in one resulting from class B events were treated patient. The remaining JAMA, Aug 17, 1970 . Vol 213, No 7 Morbidity in Hypertension 1147 Table 6.-Nonterminating Class B Events Type CVA, thrombosis orTIA* et/G-heart failuret Myocardial infarction Atria. __- _.__ - __. Heart-block Se;u3m creatinine, persistent, ~PYl.7,lll~~~ _"_ -- PrOtL...r ..-, -.. --~%tent, >1+ r... Total - . `Cerebrovascular accident. either a thrombosis (clinical diagnosis) or transient ischemic at- tabk with objective neurological signs. tcontrolled by administration of digitalis and short-term diuretics. Table 7.-Classification of Morbid Events by Diagnostic Categories Diagnosis Total Events Terminating Events A 7 I * Control Treated Control Treated Cerebrovascular accident 20 5 12 1 Coronary artery disease 13 11 11 6 11 0 5 0 ension 4 0 4 0 Congestive heart failure "Accelerated" hypert Renal damage Other Total 3 0 1 0 5 6 2 2 56 22 35 9 -__ death in the treated group was caused by a rupture of an ather- osclerotic aneurysm of the aorta. OTHER CLASS A EvsWrs-Nine patients in the control group as opposed to none in the treated group developed nonfatal class A events (Table 5). Five of the pa- tients had congestive heart failure which could not be controlled by administration of digitalis, sodium restriction, and the intermittent ad- ministration of diuretics. In the four remaining patients there was one instance of each of the following complications: dissecting aortic an- eurysm, subarachnoid hemorrhage, multiple striate retinal hemorrhages, and acute hypertensive encepha- lopathy with accompanying neuro- logical signs. OTHER TERMINATING EVENTS.- Additional organic complications, which did not fulfill the criteria for class A events but which were nevertheless of sufficient severity to require terminating protocol treat- ment occurred in eight patients of which seven were in the control group. These are listed in Table 5 under the subtitle "treatment fail- ure." 4 ur WC : associated with cerebrovascr ' i ccidents diagnosed clinically as t.~r +mbosis rather than hemorrhage but which resulted in such severe incpacity that the pa- tients were urable to attend the clinic. Two additional control pa- tients were removed from the study because (J the appearance of a single str ate retinal hemorrhage associated in one with symptoms suggesting acute hypertensive en- cephalopathy, and, in the other, with questionable early papillede- ma. The remaining control patient exhibited increasing azotemia. One patient in the treated group was removed from the study because of hypotension following a myocardial infarction which resulted in his in- ability to tolerate the antihyperten- sive regimen. It is noteworthy that of the 17 nonfatal terminating events (class A and others) 16 oc- curred in the control group and only one in the treated patient (Table 5). Nonterminating (Class B) Events. -Class B events include organic complications which require no or only temporary suspension of proto- 1148 JAMA, Aug 17. 1970 . Vol 213. No 7 Morbidity in Hypertension co1 treatment (see list of assessable events listed at the end of the com- munication) . Objectively demon- strable atherosclerotic complications predominate as class B events, but the category also includes conges- tive heart failure responsive to rou- tine therapy other than administra- tion of antihypertensive drugs and certain less severe manifestations of renal disease. Nonfatal class B events occurred in 21 of the control patients and in 13 of the treated patients (Table 6). Six patients developed conges- tive heart failure controllable by digitalis and short-term administra- tion of diuretics. It is noteworthy that all six of these patients were in the control group. Also, the in- cidence of nonterminating cere- brovascular accidents was twice as great in the control as in the treated patients. However, nonfatal myo- cardial infarction occurred in five of the treated patients as opposed to two of the control group. The in- cidence of atria1 fibrillation and conduction defects was essentially the same in the two groups. Life-Table Analysis.--The bene- fit of treatment is more precisely analyzed using life-table methods (Fig 2). This method has the fol- lowing advantages: (1) it adjusts for the fact that patients enter the study at different times and thus are observed for varying lengths of time; (2) the method adjusts for any differences in losses to observa- tion between the control and treated groups; and (3) most important, it determines whether the benefit of treatment occurs early or late or is continuing through time. The dis- tance separating the control and treatment lines is a measure of the degree of benefit. It is clear from Fig 2 that the benefit of treatment manifested it- self early and continued throughout the entire five years of follow-up. The life-table analysis of either ler- minating or all morbid events indi- Table 8.4ncidence of Morbid Events With Respect to Level of Prerandomization Blood Pressure P Prerandomiration Blood Pressure, mm Hg Systolic 5 cells per high power field centrifuged sedi- ment) not due to primary renal or lower urinary tract diseases. 1152 JAMA, Aug 17, 1970 . Vol 213, No 7 Morbidity in Hypertension PrInted and Pubiished ,n the United Stales of America