egret fam te eral fp Ararcan Mec Aescton 2 apo 7 0) Yona = ‘copra 130 Nino Mel Peso, Effects of Stress Management Training and Dietary Changes in Treating Ischemic Heart Disease Dean Omish, MI tty W, Scherwitz, Phi fachelle S. Doody; Deborah Kesten, MPH; Sandra M. MeLanahan, MD; Shirley E. Brown, MD; E. Gordon DePuey, MD; Robert Sonfiemaker, MD; Cathie Haynes, RN; jerry Lester, PhD; Gay K. McAllister, RN; Robert J. Hall, MD; John A. Burdine, MD; Antonio M. Gotto, Jr, MD, DPhil © To evaluate tha short-term effects of an intervention that consists of stress management training and dietary changes in patients with ischemic heart disease (IHD), we compared the cardiovascular status of 23 patients who cecelved this Intervention with a randomized control group of 23 Patients who did not. After 24 days, pationts in the exporimental group demons! jed a 44% mean increase in duration of exercise, a 55% mean increase im total work pettormed, somewhat Improved left ventricular regional wall motion during, peak exercise, and a net change in the left ventricular jection fraction from rest to maximum exercise of +6.4%. Also, we measured 9 20.5% meen decrease In plasms cholasteral levels and 2 91.0% mean reduction In frequency of anginal episodes. In this selected sample, shortterm Improvements in cardiovascular status seom to result trom these adjuncts to conventional treatments of IHD. (JAMA 1983;249:54-59) THE ROLES of both emotional stress and diet have long been suspected in the pathogenesis of ischemic heart disesse (IHD).' Some emotions and behaviors are associated with IHD in a variety of populations; these include intense anxiety, depression, feelings From the Departs of Maicine (Ors Orsish, Halland Goto). Communty Medicine (Oe Sena ‘rane Brown). Raology (Ora Saonomater, Br ‘re, ar OuBurp, an Papin (Or Uae) an the Olfes of Stent Allee Ge Doody, Baye Cotege of Madina, Houston: The Mabie Hosa ta, Howton (Or Got; St Luks Epacopal Hes tw Hesston (ore Ha, Sutin, Sorearatar, ans DePuer): The Univeray of Teas Sshoo! of Puble Neal, Houston is Keston: Deparment of Farly Mace, Unversy of Ganectd Sebo! of Wed tine, Fomingion (Or Metananar Ate Cae ft hrsng, Tuczon (Ma Haynes) and Souths! Nentorta ne, Houston (ds MeAlistr. Or rion Is now win Maseachuoeds Geneca! Hospi) td Marre Maseat Sheet, Boston. Brow Sam ‘wth Braham and Wome Hosp end Haars Until Senoo, Boston "Read in part etore the enol acini meeting ihe Aneta Coleg st Cardona, Apel 25.28, 1982, eprint requests to Medics Reside, Massaehu- sons General Hospital Boston, MA O2t¥4 ‘minh. 54 JAMA, Jan 7, 1983—Vel 249, No, of helplessness, and “type A behav- ior,” characterized by ambitiousness, competitiveness, impatience, and a sense of time urgency." Biobehavioral techniques, eg, meditation, yoga, and progressive relaxation, may elicit what Benson has termed the “relaxa- tion response,” which may reduce cardiovascular risk factors, eg, BP’ and plasrta cholesterol levels,” inde- pendent of dietary changes. ‘The evidence linking elevated lpi levels, particularly plasma cholesterol levels, to the development of IHD is well established* Studies of vegan sabgroups in this country have dis- closed lower levels of plasma choles- terol, low-density lipoprotein (LDL), very low-density lipoprotein (VLDL), and triglyceride, a higher high-densi- ty lipoprotein (HDL)-LDL ratio, and Jower BPs when compared with matched controls from the Framing- ham study; as the intake of animal products increased, the plasma cho- lesterol level rose.” Case reports have suggested that changing to a vegan diet may reduce the frequency of angina? ‘We report the results of a random- ized, controlled study to determine if ‘a combination of training in stress ‘management and an essentially vegan diet may produce short-term im- provements in the cardiovascular sta- tus 6f patients with THD. PATIENTS AND METHODS Patient Selection ‘We audited all patient records (1977 to 1980) in the files of the nuclear cardiology and cardiac eatheterization laboratories at St Lake's Bpiseopal Hospital, The Methot ist Hospital, and the Kelsey-Seybold Clini in Houston as vell as the entire office records of two groups of eardiologists. We selected patients (aged 45 to 75 years) who hhad evidence of IHD with (2) greater than 50% stenosis in one or more major eoro- nary arteries by cardiac eatheterisation oF (2) positive exercise radionuclide yentricu- lography, defined oe a reating ejection fraction that fails to rise more than 3% with exercise and/or with regional wal motion absormalities during exercise.” We excluded patients for any of the following reasons: a resting ejection frac- tion of less than 40%, cardiomyopathy, 2 ‘myoeardial infaretion or changes in car- dise medications within the preceding six months, carcinoma, a eerebral vascular accident, psychosis, or previous cororary artery bypass surgery, arless there” was angiographically verified evidence of graft ‘eshosica. Using these eritort 125 patients were ligible: Bach was sent a letter that esecibed the intervention (which began six weeks later) and a statement of informed consent. Fifty-one patients vol- unteered and were pretested during Au-gust 1960. Three patients had a resting sjeetion fraction less than 40% and were sxcluded from the study. The remaining 48 patients were randomly sanigned to the fxperimental and contol groups (2¢ patients cack), using fandom aumber {ables in balanced randomization." Dur- ing the study, one patient withdrew from cach group (before positeting, and their results are excluded from all the analy- Study Design ‘The experimental group participated in ‘a program of stress management training and dietary changes from Sapt 3 to 21, 1980, while the control group continued thelr routine activities at work and home. Patients in the experimental group were housed together in a rural environment to maximize compliance with the interve tion and as 2 component of the stress management training. The project stall prepared and served all meals at this site and trained the patienta in stress manage- ‘ment. Patients were required to consume only the food and beverages that were served to them; this was further rein forced by the relative inaccessibility of other food. (To reduce confounding infiu- fences, aerobic exercise was not 2 compo- ‘est of this intervention.) ‘Both groups were retested on all pre tervention measures between Sept 27 and Oct 18 according to the exact preinterven- ‘ian protocols. All testy were condocted at the Texas Medical Center is. Houston ‘Technicians who processed the data and physicians who interpreted the results were blinded to patient identity, testing ‘ime (before or after intervention), and group membership (experimental or con- fol). The protocol was approved by the Human Subjects Committees of Baylor Coliege of Medicine, Houston,:St Luke's Spiscopal Hospital, and the Kelsey-Sey- ‘bold Clinic. Dependent Variables Exercise Radionuclide Ventrculograpy. ‘The protocel for exercise radionuclide ven- triculography has been described in detail im other publications.” Our protocol dif- fered only slightly—we discontinued all medications for 12 hours before testing, and we tested patients in a siting position rather than in a supine position. In brief, 80 mCi of sodium pertechnetate Te 99m ‘was injected into an antecubital vein 20 ‘minutes after the injection of 6 mg of stannous pyrophosphate to label RBCs. All radioactive emissions ware collected with the patient in the 45° sitting position using a scintillation camera, A 30° resting right anterior oblique gated image and 45° resting le(t anterior oblique (LAO) sated image were collected for 2% minutes tach. Blectrocardiographic gating was used by a computer to organize the ama, Jan 7, 1983—Vol 249, Wo. + acquired data into series of images that span an average cardize cycle. Images were displayed in rapid sequence as an cendless-loop ficker-free movie so that wall motion could be evaluated, Globular ventricular fonction was assessed by determination of the ejection fraction, which is ealealated from the ratio of the radioactive emissions (counts) after back- ground correction collected from the left ‘ventricle in end-diastote (ED) minus end- systolic (ES) counts to end-diastolie counts (ED), or (ED-ESVED. The BP was obtained in the right brachial artery using fan audible BP cuff that uses an electronic transducer. After resting measurements were completed, the patient pedaled a Bicycle ergometer st a Idad of 300 kpm/ rin (approximately 50 W), and this load ‘was inereased in 100-kpm increments in hree-minate stages antil stopping. After an initial 80-5 period to reset the RR Interval that was accepted by the eomput- cr, sequentia! LAO images were recorded during the remaining 244 minutes of each stress period. Heart rate and RP were recorded at three-minute intervals to cal- culate the rate-pressure product, Reasons for stopping exercise incladed exhaustion, severe chest sin, shertness of breath, Attainment of maximum heart rate, ECG changes (ST-segment displacement of 2 mam or more if lat or downsloping or if the J point is depressed 2 mm or more and the return to baseline is 80 ms or longer), complex ventricular archythmias, a systol ie BP greater than 250 mm Hy or a Giastolic BP greater than 140 mm Hg or a falling BP with increasing exereise. Three nuclear cardiologists reached a consensus, on global ejection fraction and regional vwall mation by viewing each patient's prointervention and _postinterveation study together as a random A-B compari- ‘Other Measorements—Plasma lipids were drawn after a 34-hour fast, doting whieh nothing but water was ingested. Blood samples were analyzed under labo-, ratary conditions standardized for the Lipid Research Clinies™ Angina frequen cy, smoking history, and medication usage Were determined by questionnaire. Intervention ‘trees Managemest Training —Streas management techniques were taught and practhoed Sve haure ger day; this time was vided equally among the diferent tech- siques. Each technique was presented as having the common purpose of increasing « patlents sense of relaxation, concentra- ton, and awareness of internal sistes to retrain physiologiesl responsiveness to emotional stress." Techniques ineladed the following. ° 1. Stretching/Relazation Esercees—Pae jonta were taught simple, nonaerobie stretching exercises. They were advised to strech slowly and getly and were care, fully monitored to avoid injury or strain. Patients were directed. to. focus their attention onthe areas being stretched and while resting to eoeentrata on. their breathing. AE the end of each dass each patient was jatroced to tense and relax muscle groups sequentialy from fest to iad, ending with a meditation. 2. Meditation Each patient was wed to sit in a comfortable position and to breathe slowly and deeply while focusing Xs station on his breathing, etarning tot when attention wandered 3. Applied Metvaion (Visualization, — Bach class began with lecture on basic physiology and anatomy of the cardiovas- Colar ston and the pxthopbyeslogy of THD to aid in constructing and maintine inga mental image. After the lecture, each vatient was instructed to. meditate, as previoosly Jesribed, Alec several ine tes of meditation, each patient was asked to visualie his beat and coronary stern, referring, whenever necessary, £0 Grawings bowed wu pene coronary angiog- raphy. With eye closed, each patient was shed. to visualize the atherosclerotic plaques being removed from the coronary aeriey wang aa image of thelr choice Each class ended withthe patients viaal- fring themselves as healthy, doing an activity that they enjoyed when they were ‘without the physical limitations of THD. 4. Emirorment—The primary reson for housing patients together in 2 roral frvionmert was to ensure compliance to he intervention. Approximately half the Patients reported thot the investigative Setting contributed to thelr perecived reduction in stress, but the others said that it was more stressful fr them to be say foom their work, hore, and family and tobe living in close quarters with & new group of persons Diet—Patiente wore served a vegan diet (Geoid of mrimal. products) exept for minimal amounts of nonfat yogurt. Alo txciuded im the diet were salt, sigan, Alcohol, and caffeine. They were served teeth Trina and vegslales, whole grains, legumes, tubers, and. soyinan products ‘The diet was verfed for nutritional ade- auscy, with an average dally intake of $140 caries, 325 mg of sodium and 82 mi of cholestrol. Partesar atexion was given to making the food attractive nd appetizing, Dally classes were given in food’ purchasing, preparation, até nati- ee Statistical Analysis ‘To compare the experimental and con- trol groups before intervention. we used Students text (two tailed) on the interval data and Fisher's exact probability test on ‘the eategorieal meagures. To assess wheth- cer the experimental group improved rela- ‘Treating Ischemic Heart Disease—Omish et al 55(eerie minke Ha) “Table {.—Charactenstcs of Patients at Entry Into Teal” Ta Brpwtinanal Group Control Group Characters os) neta) Signincance Tae, yi Teen SEM) Tos tsareay ws reusT aS Man ci 19 xs Proiovs myocar! inaction 5 1 1s Prowou cororary bypass 8000) 1 1 Ns Ina i parethones a8 ranges NS Indealan P05 by Shab’ xt. Table 2.—Exerciae Radinucide Ventriculography” ‘ean Se oF Tipatmoial Control Stgnite taco Period Group * Group =" cance Daaion of exercise, 2 Bolore merendon SER raAtT —abaazaaa Aterinenenton seaazs4o aaggzssg <0" “Toni work cerormed, kom Belrenervonion zasz.824927 9888726002 44, anerstecanton 4952.72 642.4 surras seas 7000! Maximum ralegrenaure, Baler inerrenion 25052102 WASEDA Pedic. HRXEPXIO” — Aner ntrvenion 254.02197 220.83 12.0 “charges nthe expert groupware compared ith chnges i the conto rou wang an any ot 3 penarvenion daperdent meancea were the covaratan (the Bosinanerton Table 3.—Plaawa Lipid and Lipoprotein Levels = See aes Sn ma Se et ae ae aa eget level (DL), mgr “Attor intervention tsk 38.223.0 s.0008 lercceel cone asees “Changes inte apartment groupware comarsdwah change inthe col 0¥p using an anaes covaranes. Tha pinorvonton dependent Netores tive to the control group after interve tion, we conducted analyses of covariance’ on al interval data; the values of preinter- vention-dependent measures were the co- variates for the postintervention mea- sures. The ejection fraction response data ‘were further analyzed using a three-way analysis of variance; the factors were (1) group (experimental v control), (2) time (before intervention » after ‘interven- tioa), and (3) condition (rest v maximal exercise)" Group data are expressed as means SEM, Percentage changes reflect differences in mean values of the. raw Sata. RESULTS Baseline Characteristics ‘There were 0 statistically signifi cant preintervention differences (P>.05) between the experimental and control groups in age, sex, previ- ‘ous myocardial infarction, or previous coronary bypass surgéry (Table 1). 56 JAMA, Jan 7, 1983—Vol 249, No. 1 eth covariates ora posintarenon menses, Also, there were no significant prein- tervention differences in any of the reported measures, with the exeop- tion of the serum HDL level, which “was slightly higher in the experimen- ‘tal group. Eighteen patients in the experimental group and 21 patients in the control group had prior coronary angiography; they averaged 18 and 24 occluded arteries (250%) (P>.05). In the experimental group, seven persons had one-vessel disease, seven persons had two-vessel disease, and four persons had three-vessel disease; in the control group, six petsons had one-vessel disease, eight persons had two-vessel disease, and seven persons had three-vessel dis- ease. Exercise Tolerance In the experimental group, the total duration of exercise (bicycle ergome- ‘Treating lschemie Heart Di try) increased 44% (F'=20.1, P<.001), and the total work performed in- creased 55% (F=16.0, P<.0001), ‘whereas the control group was essen- tially unchanged in both measure- iments (Table 2). Both groups achieved approximately the same rate-pres- sure products (systolic BP x heart rate at peak exercise) after interven- ion as before intervention, but the experimental group performed at a ‘much higher work load before achiev- ing the preintervention rate-pressure product (Table 2). The resting heart rate did not change significantly in either group. Plasma Lipid Levels Changes in plasma lipids are out- lined in Table 3. Overall, the exper mental group showed a 205% reduc- tion in plasma cholesterol levels (22 of 23 patients had reductions, even though most were not hypersholes- terolemic), while the control group did not change (F=198, P<.0001) ‘Triglycerides also were signiticantly reduced in the experimental group but not in the control group (F'=6. P03). While the HDL levels d creased in the experimental group, the total Sholetra/ EDL ratio showed no significant differences be- tween the groups. Studies of vegstar- jan subgroups in this country and elsewhere have shown both a lower total plasma cholesterol and a lower HDL level, yet a lower incidence of coronary heart disease than the gen- eral population. The total cholesterol/ HDL ratio may be a better indicator of coronary risk than the HDL level pense" Frequency of Angina ta the experimental group, the reported frequency of angina episodes per week decreased from 101+20 before intervention, to 16:05 after two weeks of the intervention,, to 0.9203 after intervention stopped. ‘The control group rermained essential- ly unchanged from 8.021 episodes per week before intervention to 1521 after intervention (F=25.1, P<.0001). Exercise Radionvetiea Ventriculography test provides an accurate, non- invasive measure of left ventricular function at rest and during exercise. In interpreting the global left ventrie- LOmish at al§ 00) cL aag} 3 sad tof 9 19} Changes in tat wentrcularsintion traction ta ‘ercent trom rest to maximum exercise (AEF) Dotore and ater intervention, Each bar is obtained by subtracting AEF,, trom AEFn., ‘were: BEF equals change in Seeton fraction ftom rest 10 peak exercise. Experimental ‘group, arrow diagonal lines; control rou, ‘wide diagonal ines. lar ejection fraction dats, ~ the change in ejection fraction from rest to peak exercise. (AHF) is more refle tive of the degree of myocardial is- chemia than the absolute values of ejection fraction at rest or at peak exercise. Most patients with multive: sel THD are onable to increase their ejection fraction more than 5 absolute percent from rest to peak exercise (EF) and/or they exhibit new regional wall-motion abnormalities during exercise that are not present at rect Ejection Fraction Response Before igtervantion, the ejection fraction response to exercise was abnormal for both the experimental and control groups. In the experimen- tal group, there was a slight decrease in the mean ejection fraction from rest (58.5:22%) to maximal exercise (619221%), AEF,.--0.6:1.5%; in the control-group, there was a slight rise from rest (585+1.7%) to maxi mal exercise (614+22%), AEFq =42615% (Figure). Those inter- group preintervention differences, were not statistically significant (P>-05). However, after intervention, the mean ojection fraction response to exercise (AEF) of the experiments! group was significantly improved when compared with the control group (Fa160, P<.0001, three-way analysis of variance). In the experi- ‘mental. group, there Was an increase in the mean ejection fraction from rest (53.8:12.4%)'to maximal exercise (842.7%), AEF_m+58%. In the JAMA, Jan 7, 1983—-Vel 249, No. 1 “Table 4.—hecieaion Changes During intervention Tae Medication URE? Dlncorlinaed "-Radoced Dovaue "© ho Change > Wet Taviog control group, there was less rise from rest (56.3:+18%) to peak exer- cise (67.2:12.4%) than jn the preinter~ vention studies, AEF,..=+0.9%. In the experimental group, ABE,.— AEF, (+58)-(-08)= 464%. Of these 23 "patients, 19 showed im- provement, one was unchanged, and three showed a slight decline (~1%, =1%, and -2%) in the mean ejec- tion fraction response from rest to peak exercise, In the control group, AEF .~AEF ;=(40.9)-(42.6)=-L 7%, Of these 28 patients, nine showed improvement, one was unchanged, and 18 showed a decline (F=125, P<001, analysis of covariance, Fig 1) is the mean ejection fraction ‘osponse from rest to peak exercise. Before the intervention, 20 patients in the experimental group aid 16 ‘patients in the control group demon- strated abnormal responses in the left ventricular ejection fraction to peak exercise (AEF <5%). After the intervention, only ten patients in the ‘experimental group showed abnormal jection fraction responses to exercise compared with 17 patients in the control group. Regional Wall Motion ‘The experimental group also showed some improvements in re- gional wall motion at peak exercise after intervention when compared with the control group. Before inter- vention, there were 13 patients in the experimental group with new region- al wall-motfon sbnormalities during peak exercise (not present at rest). After intervention, despite a much higher level of exercise, five of these patients displayed some improvement in segional wall motion when com- pared with their preintervention studies, seven were unchanged, and one was worse (P-<.05). Before inter- vention, 15 patients in the control group had new regional wall-motion abnormalities during maximal exer- cise; after intervention, two of these abnormalities were somewhat im- proved, six were unchanged, and sev- en were worse (P>.05). Medication Changes Tt was necessary to discontinue antihypertensive medications and/or A-blockers in eight patients and reduee dosages in ten others in the experimental group (Table 4) because of the appearance of medication side fleets and/or hypotension (diastolic BP <70 mm Hg), although none reported an increase in compliance with medication regimens during the intervention. No medication changes were made in control group patients. In the experimental group, proprano- lol hydrochloride therapy was re- duced in those taking It from @ mean dosage of 80.7 to 40.7 mg/day; mean dosage in the control group remained ‘nehanged (91.2 mg/day). Multivariate Analysis im the experimental grovp, there was a significant reduction in weight during the intervention, from TSAR 129229 kg, whereas the control group remained essentially unchanged from 825229 to 83.2+30 kg P=%9.3, P05).2 Other studies have demonstrated that the resting ejec- tion fraction tends to be lower when it is measured a second time, whether by radionuclide ventriculography” or by coronary angiography.” This bio- logie variability may be rélated to greater familiarity with the proce- dures at the second study. Also, if the increase in the postintervention change in the ejection fraction from rest to peak exercise were caused only dy the decrease in Testing ejection fraction, then the control group should hare displayed a similar increase in the postintervention change in the ejection fraction from est to peak exercise, but it did not. The changes are not likely to be because of changes in BP (afterload), since neither systolic or diastolic BP changes were significant predictors of the variance in ejection fraction response, It is possible that the change:in ejection fraction response may be caused by factors other than a reduction ia myocardial ischemia. Aerobie exercise was not a compo- nent of this intervention, although some patients began to walk more as they became less symptom-limited. ‘This may be responsible in part for the increases that we measured in total duration of exercise and total work performed. Although aerobic exercise has many cardiovascular benefits, itis not likely to be a factor in the short-term improvements in ‘exercise radionuclide ventriculogra- phy that we measured.” We do not know why the conditions af the patients improved. There is some evidence to speculate that the apparent improvements in the experi- mental group may have occurred through currently accepted mecha- nisms of IHD, although.we did not study this. Emotional stress may lead to myocardial ischemia both by way of coronary artery spasm and by increased platelet aggregation within coronary arteries’* Stress may lead ta coronary spasm mediated either by direct a-adrenergic stimulation or secondary to the release of thrambox- wie A, from platelets, perhaps through increasing circulating cate- eholamines or other madiators.*" Both thromboxane A; and catechol- amines are potent constrictors of arteriak smooth mustle and powerful endogenous stimulators of platelet aggregation.” ‘ven a single high-fat, high-choles- terol meal may cause short-term ezhancement of platelet reactivity.” ‘These changes may result from a shift in the thromboxase/prostscy- clin balance to favor thromboxane pradudtion; some evidence supports this. Cholesterol-enriched platelets release more arachidonic acid from platelet phospholipids than cholester- ol-depleted platelets, and the coneer- sion, of released arachidonic acid to platelet thromboxane fs higher ia cholesterol-rich platelets than in ‘Tremtng Ischemic Heart Disease—Omnish ot slthose that are cholesterol depletes In animals with atherosclerosis in- duced by high-cholesterol diets, plate- lets synthesize thromboxane A, in increased amounts." Since cholesterol is contained only in foods of animal origin, a vegan diet may shift the balance away from thromboxane for- mation, which would make both coro- nary spasm and platelet aggregation Jess likely to occur. In general, animal protein has been found to increase the level of plasma cholesterol, even in experiments in which the cholesterol and fat have ‘been removed from the protein.” Some studies suggest that plant pro- tein may be hypocholesterolemic.” Changes in free fatty acid (FFA) metabolism may have contributed to the observed improvements. Myocar- dial oxygen consumption is influenced by the substrate supply to the heart. Use of excess FFAs increases myocar- dial oxygen consumption and de- creases left ventricular wark, left ventricular systolic pressure, aortic pressure, epicardial motion, and exer- cise duration; these effects also are seen during myocardial ischemia.” Excess FFAs during ischemia result in even greater deteriorations in hemodynamic and metabolic fune- tions” A diet that contains a large proportion of animal products results in high blood levels of FFAs. The opposite is true with diets low in animal produets, probably by shifting from nonearbokydrate to earbohy- drate energy sources during physical or emotional stresses.” Emotional stress also increases blood levels of FFAs, primarily by way of eatechol- amine stimulation; furthermore, eats cholamines sensitize the heart to the oxvgen-wasting effects of FFAs.” ‘A combination of stress manage- ment training and an essentially vegan diet produced short-term im- provement in cardiovascular status {as measured by a variety of end- points) when compared with a nonin- tervention control group. Interpre- tation and generalization of these findings must be tempered with esu- tion, since the patient population is selected and the sample size is rela- tively small. The intervention is safe and compatible with conventional treatments af THD. Major fnaing for this tay was prov Gerad Bn srs The sey oy lob supported by grants from the National Hear. JAMA, Jap 7, 1983—Vol 249, No. 1 sod Blood Vessel Research and Demonsteation Center, Hash Callege of Medicine (grant HL 1729 from the National Hear, Lang, od Blood Tnatitate, National Tnatiteter of Heaithh the bile Health Service Biomedical Research Supe port Grast REOSOL10, the Natwin Founda on: the Abraham Student Aid Foondstion: the FFannla Bank; the Kayser Foundation he Pranz heim Synergy Trust Tranace Companies; Amax Petroleus Merril Lynch, Tac The Holmes Cen- ter, Pentaol Petroleum: Seagull Pesce, Raling M. Raia, N. Consing, and. ‘We sre indebted to K Aime and Lelsre ‘Resourses Group for housing the patients to & Satehidananda for instructing us i the inter= ‘venting; to MC Sholman for designing and pre- Dating the meals, and toD. Clark .Calten, c E-Glatk, her asiatants to C.iyers, 8 Posnds, SE Bletendort, MA, Nava, and A. Farle Tor technical assistance to-A. Leaf, MD, HL Brown, MD,'D. Munford, MD, R Rosenthal, MD, A ‘Taji, MD, C. Vallbon, MD, 8. Oraleh, MB, 8 Jit MD, sod F, Shin MB, or tea, sistance to J. King, L Harden, Br Newall Sele, and O. Schotemann for sdmiiatrative asistanee; to H. Gruppe and. ‘Mishel fr thee support tothe cardiac fellows ‘and nurses who helped make the study posible, 4nd to the private physicians bo refered patients to thio study, capeially D. Hochells, 1D, and his aoiates References 1 Harvey W, cad by Eastwood MR, Treve: lyan TE Stress’ and. coronary heart dleaae, J Peychosom Bas 1711880 2. 2 Hackott TP, Rosenbaum JF: Ematio, poy chiatric disorders, and the heart, in Braunwald 1 (ed) Heare Dicezse Philadelphia, WB Saun- ers Go, 1980, pp 1822-1988, 3 Benson ff Sprtemie hypertension and the relasation response, N Brol J) Mod 197% 235:152-115, {Pash Gk Reduction of strum eee and blood preseure in hypertensive patients by ‘behavior modifeation. JR Coll Gen Prot 1876, mas, 5. Cooper Ms, Aygen MM A relaxation tach sique in the masagement of ypercholster- lamia. J Humen Stree 1970124-2 ‘Shakelle RE, Shyrock AM, Paul O, et ab seram cholestrol; sud death from coro nary heart disease: The Western Blecri Sad 1B Bagl Med tae ane Pasa tend ppsetl i vera ed entra N Bul ded te sh. 3 Sees i, Homer Be Kats BE Bod pre sate in vegarans 4mm J Epon Jean 9. ls FR, SandreT Angi and vega it An Hear rata TofS Ofte Sn at Ena radionele imaging apres coronary tetery doese’ mY Co ah Seneca Golden A Balti New York, ae niles Palette Go ne, 196% pp "IE Boer facta Sa ren MY ah se radondie dnantiograpy in ine OF ola to rpm sent and aree ‘corcae in pant th sorararyragy ae ‘ease. N Engl J Med 1977,296:899-644. ined ik. Mental of Labrory Operon, Lipit ‘Rewarch Clinic Program vol Did tad “pepo alee U8 Dep of Heath ise don tod Wellare polation. NIH) 6 Goreramet rate Oe, 9. Te Orn Or Wina/heitInteracons For sete and er oor ah Youn 2 ‘resting lschemie Heart Dis D: Stress, Diet and Your Heart. ‘New Yor Hale achat Wey fee He NE. Me NE SPSS Statistical Peckage fone Set Sens A ee Yor tran Book Co, 1975, op TE lon WS fet Biometil Comte Pro. grams Darkeley, University of California Pree, iors, pi. 18 Casteld WP, Doyle JF, Gordon 7, et al: HL cholesterol and’ other i 1S. Oreak DM, Cotto AM, Miller RR, et ak Bflects of a vegetarian diet and selected yors technlgara tn the treatment af earonary Beart Aisease abstracted. Clin ee TOPEIOA, ‘2h Here RM, Leveeque Md. Commi JP. Socal environment as a factor fo dietinduced ‘theropelronia: Sconce 19802081475 167 ‘21, Marshall 8, Wisenberg G, Schelbert Ry et at Bier of eral propranolal on ret, exercise ‘ad postexercise Tet venteicular performance ‘in normal subjects and patients wih coronary arty disco. Oi ease Sa 2 Upiow MT, Rerych SK, Newman GE, et ‘The reproduclblty af radiosuclige sngiarsp rmeasgremeats of left ventricular fancion_in Sormal subjects at rest and Waring exerio, (Grenation 181;g>128-132. "A. Mednolty JH, Rremkan EL, Rosch J eta Spontaneaus changes in left ventricular fanetion between sequential studies. aon J Cardia! 1974; rae, 2 Verani MS, Hartuag GH, Harris JK, et a fects of exereie training on left ventricular pacformance and myocardial persion in ps: ent with coronary artery dlesse, Ame J Cor it 10-07-608 ‘2%, Oliva PB. Dathoph cardial infartion, 198, sezae 250. "28. Schiffer P, Hartley LH, Sehulman CL, et ak Evidence fr emotionally induced soromary srteral spasm in patents with angina pectoris. Br Beart 19eO Ace s6 ‘fi Hatt JI, Pant Ke Intravasclar platelet on’ inthe east induced by st Giredaion a:50-958 ‘2 Willams DO, Most AS: Responsiveness of ty coronary eirealatlon to brief vs sustained (eadronergic Wisalaton. Cireulation 18083. ie ‘, Moneads S, Vane JR: Arachidonic aad metabolites and the interactions between plate: leis and blood vessel walls A” Bagl J ied wibpoen east 30. Haft JE Role of blood platelets in coronary srtaty disease. doe J Cardiol 197948.1197-1205. 31. Stuart Mi, Gerrard Jal, White 1G: Bec gic on padcton of hramborane B ty platelets fe wtro. "Biol J Med 1080 mee ‘2. Zmuaa b, Dembinska-Kiee A, Chybowskt ‘Avot ak Seperznentalathereaceree te rabbit ology of acute mpo- nm Inlre Med 198 Carroll KK: Dictaryprotsin in relation to seana Golstarol Ievea and aterosleroxs ‘itr Roo 62-5. UL. Sitori GR, Agradi B, Conti Ft at: Soybean-protein diet fa the treatment of typeTT Iyperipopreteinemia, Lana! 1977-25-27 ‘Slmoneen §) Kiekehos JK: The effect of free faty acs on myoeardlal oxygen eonsump- ‘Son duriog steal pacing and ctacholarsine Infuslon in men. Cieuation 19TSAEOL. 138, Liedtke AS, Nellis 5, Weey TB: Bact of {ove fatty acids on mechanieal ad matabolie foes i normal and acheme myocardium ia ‘sine, Cire Hoy 197466261. Fe Martin -B, Robinson 3, Robertshaw D: Infloene of det on leg uptake of glucose daring navy exercan, Ave J Clie Nutr 9082-7. BS Vite Hi, Mjos OD: Inisence of re fatty acids on myocardial oxygen, consimption. 20d [Sthomie injury. im J Cardit 19816061 26. Ornish ot al 59