Response to Psychologic Stress in Persons

Who Are Potentially HypertensiveBy ROBERT E. HARRIS, PH.D., MAURICE SOKOLOW, M.D., LEWIS G. CARPENTER, JR., PHi.).,

MERVIN FREEDMAN, PH.D., AND SAMUEL P. HUNT, M.D.

College women who are prehypertensive and matched controls were exposed to emotion-provokingsituations, psychologic tests, and psychiatric interviews. Patterns of response differentiating thegroups were observed. The prehypertensives were less well controlled, more impulsive, more ego-

centric, and generally less adaptable in the stressful situations. In the psychiatric interviews,behavior similar to that of patients with hypertension was noted. The findings suggest that pre-

hypertensives are more vulnerable in situations involving psychologic stress, and hence more

subject to the autonomic concomitants of emotion, including repetitive rises of blood pressure.

E SSENTIAL HYPERTENSION is adisease of unknown etiology. As inother diseases with obscure causes and

chronic but variable course, a number of in-vestigators have claimed that emotional fac-tors play a role in what is probably a verycomplex network of events, some of which canbe described physiologically and anatomically,and others psychologically. They have sup-ported their assertions in a series of provocativeand at times persuasive clinical psychiatricstudies. Evaluated as scientific evidence, thesestudies have three major limitations.

1. Patients with hypertension who havebeen studied by psychiatiitsts have usuallybeen referred by internists, who are likely tolefer their "interesting" and troubled patients,neurotic persons whose psychologic problemsmay have no causal connection with elevationsof blood pressure. Other hypertensive patientsmay find their way to psychiatrists because ofmultiple vague physical symptoms of a hypo-chondriac nature, again with no necessary re-lation to hypertension.

2. Clinical psychiatric methods are notablyimpressionistic and qualitative, perhaps in-evitably so by their very nature. There are nounequivocal criteria for establishing the pies-ence or absence or the degree of important

From the Departments of Psychiatry and Medi-cine, University of California School of Medicine,San Francisco, Calif.

Supported in part by a research grant (H-754)from the National Heart Institute, United StatesPublic Health Service.

874

phenomena such as unconscious hostility orneurotic conflict. To the probing eye of apsychiatrist, few people are "normal."

3. It can be argued that the neurotic prob-lems found in hypertensive patients may besecondary to other processes involved in thedisease, or to the fact that the patient knowshe has a disease which will probably shortenhis life span.

In our research, we have controlled a numberof variables related to these problems and, inaddition, have attempted to demonstrate thepresence of psychologic deviations in personswho are not now hypertensive but who prob-ably will become so. Such a demonstrationwould add weight to the argument that psyche-logic factors are relevant in the complex etiol-ogy of hypertension, since they then would beshown to exist before the onset of fixed hyper-tension.

In selecting our subjects, we have takenadvantage of the fact that an elevated bloodpressure in early adult life, even though tran-sient, has prognostic significance for the laterdevelopment of hypertension. Figure 1, adaptedfrom Hines,' demonstrates an increasing inci-dence of clinically diagnosed hypertension after10 years as a function of initial systolic anddiastolic readings. Figure 2 presents similardata for a 20-year follow-up period. Supportingdata are available from the studies of DiehlPon college students and Levy and co-workers3on army officers, as well as from certain in-surance data.

Accordingly, we selected for our experi-Circulation, Volume VII, June, 1955

by guest on July 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

ROBERT E. HARRIS, ET AL.

mental group 40 undergraduate college womenwho at the time of their initial physical ex-aminations on matriculation had shown bloodpressures which exceeded 140 systolic or 90diastolic or both.* They were matched with40 women whose pressures were less than

,I\\ SYSTOLIC~80- we

h DIASTOLIC4

Q6 60-

40-

'0 20-

0O0oSYSTOLIC <4109 ~110 120t...

IE

30 140 >150119 129 1-9 149

DIASTOLIC (74 ,75 t 80 t85 tog0 >95D79 84 89 94BLOOD PRESSURE

FIG. 1. Relation of blood pressure to incidenceof subsequent hypertension. Ten year follow-up.(Adapted from Hines.1)

SYSTOLIC (109 to1o to,120 ,t013O 140 >1

119 129 139 149

DIASTOLIC 74 1t7 ~80 tc85 to0 >95

79 84 89 94BLOOD PRESSURE

FIG. 2. Relation of blood pressure to incidence ofsubsequent hypertension. Twenty year follow-up.(Adapted from Hines.')

120/80. In the former group, which we shallcall prehypertensive, we expect a greater inci-dence of hypertension in later adult life thanin the normal controls. At present the evidencefor the prediction is presumptive; the actual

* We are grateful for the assistance given us byDr. William G. Donald, University Physician, andby Dr. Margaret Zeff, Ernest V. Cowell MemorialHospital, University of California.

incidence of essential hypertension for the twogroups will be checked by later studies.The subjects were unaware of the purposes

of the study and of the reasons why they hadbeen selected. The experimenters, interviewersand observers did not know which subjectswere prehypertensive and which were controlsuntil all the data had been recorded.

METHOD AND RESULTS

I. The Psychiatric Interview

Eighteen of the prehypertensive group and20 of the normals were interviewed by one ofus, a psychiatrist (S. P. H.), who sought toidentify their characteristic psychologic reac-tions and to infer which subjects were pre-hypertensive and which were normal. From

TABLE 1.-Accuracy of Identifications of Prehyper-tensives and Normals from Psychiatric Interview

Interview Rating

01-3

Group

Normals PrehypertensivesN = 20 N = 18

14 66 12

Ratings on 0-3 scale indicating increasing degreesof hypertensive personality features.

i2= 5.1p = less than 0.02 (single tail)

previous interviews with more than 100 pa-tients with fixed hypertension he had derivedhypotheses about characteristic psychologicfeatures associated with the disease. He ratedeach student on a scale from 0 to 3, indicatingthe degree to which she showed similaritiesto the personality pattern of fixed hyper-tensives. The results are presented in table 1.Note that 14 of the 20 normals (70 per cent)

were assigned a rating of zero, while only six(33 per cent) of the prehypertensives wereassigned a like score. If a rating of one or moreis taken to indicate the presence of hypertensivefeatures, the psychiatrist was found to identifycorrectly 26 of the 38 cases (68 per cent). Thesignificance of this finding is given by. the chisquare which is significant at the 2.0 per centlevel of confidence; that is, such a finding wouldoccur by chance only twice in 100 repetitions.

875-

by guest on July 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

RESPONSE TO PSYCHOLOGIC STRESS

Although not the major focus of this pre-

liminary report, the basis on which the psychi-atrist made his correct identifications is ofsome interest. An important factor was hisestimate of the degree of neurotic involvementin each subject. Of the eight women whom hedescribed as most normal, those with goodscholastic and social adjustment, seven were

from the control group and only one was

prehypertenisive. Of the 16 most severelydisturbed, either markedly neurotic or schizoid,11 were prehypertensive and only five were

controls. More specifically, the psychiatrist'sjudgements were based on his evaluations ofthe dynamics and economics of conscious andunconscious hostility. These involved an

estimate of the quantity of hostility latent inthe personality, the strength of the defenses,the breakthrough of hostility into conscious-ness under stress, and the degree of anxietywhen the hostility was perceived. These dataand interpretations support the hypothesesdeveloped from the clinical experience of mostpsychiatric observers.

II. Experimentally Induced Stress

In the second procedure, we exposed thesubjects to situations involving psychologicstress. The technic devised to evoke stress was

that of staged interpersonal situations or

psychodramas, interactions between two peo-

ple. One of two roles was assigned to a subject;the other was played by a staff member. Themethod has several advantages over othercontrived laboi atory methods for inducingstress:

1. Plots can be constructed which are realis-tic and approximate those life situations thatare especially meaningful to the subjects.

2. Plots can be constructed which are

directly relevant to the presumed psychody-namics of hypertension.

3. The amount of stress may be varied bythe protagonist (staff member).

4. The method seems to have an inherentpower to produce a high level of motivationor ego-involvement in the subjects.

In the first of the two psychodramas usedhere, the subject was instructed that she was a

college student, that she was going to a dance

that evening, and that she was to pick up anevening gown which she had left at a drycleaner's. In the experimental room, on aslightly raised platform, she was met by astaff member who played the role of the pro-prietor. He informed her that her gown wasnot ready, that it was not even in the shop.He met her every question or demand withstubborn indifference. In the second psycho-drama the subject was instructed that a num-ber of fellow students had asked her to callon the dean of women of the college to requestthat an examination be postponed. She wasalso told that the dean was an old friend of hermother. The dean was played by a professionalactress of considerable skill and versatility.*She met the student initially with warmacceptance, then became critical and personallyattacking and rejecting.

It was clear that these situations wereemotion-provoking for the subjects, from thefact of visible autonomic responses, fromdisturbances of verbal and motor behavior,and from retrospective accounts related by thesubjects after the psychodramas were com-pleted.Two or more staff members observed each

student and recorded their impressions by avariety of methods. One method utilized a listof 389 adjectives selected to cover a wide rangeof personality and behavior. t The observerschecked as few or as many words as theythought were descriptive of each subject.These data were analyzed by comparing thefrequency with which each adjective waschecked for the prehypertensive and for thenormal groups. The differences in frequencywere then tested for statistical significance andtwo lists compiled: one list of those 43 ad-jectives checked significantly more often forthe prehypertensives, and another of 42

* We wish to express our indebtedness to Mrs.Grace M\Lacouillard, whose facile ingenuity contrib-uted greatly to the successful execution of the emo-tion-provoking situations; and to Mrs. Julia Lee forkeeping the observers unaware of the subjects'identity.

t Adapted from H. G. Gough: Predicting successin graduate training: A progress report. Berkeley,The University of California Institute of PersonalityAssessment and Research, 1950. (Mimeographed.)

876

by guest on July 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

ROBERT E. HARRIS, ET AL.

checked more often for the controls. The firstlist, describing the prehypertensives, is shownin table 2.

.N\ote first that the majority of these ad-jectives have a negative tone; that is, theydescribe personality in unfavorable terms;for example, there are the words obnoxious,irritable, unsympathetic, and self-centered. Sec-ond, many of these adjectives may be describedas referring to inadequate control of behavioror emotionality, for example such words aschangeable, erratic, excitable, impulsive, un-stable, and restless. A third characteristic ofthis list is that the words describe an ego-

TABIE 2. Adjectives Checked Significantly MoreOften for the Prehypertensives than for the Norinals

by One or Two Observers

aggressiveawkwardblusterychangeablecleverdespondentemotionalerraticexcitablefearfulgraveguardedheadstrongidealistic

imprudentimpulsiveinfantileintenseirritableloudnaggingnervousnoisyobnoxiousopinionatedpeculiarpersistentpessimisticprecise

providentprudishqueerresentfulrestlessself-centeredsourstubbornsulkytalkativetemperamentalunsympatheticunstablewary

centric kind of person, for example, self-centered, unsympathetic, stubborn, tempera-mental, and headstrong. Fourth, there are wordswhich suggest moodiness, for example despond-ent, pessimistic, resentful, sour. In general one

may describe this list of words by saying thatthe observers found the subjects in this groupunattractive, and they tended to describethem in negative, unflattering terms.

It should be noted that not all of theseadjectives were applied by the observers toall of the subjects. Hence there may be ap-parent contradictions within the list. Forexample, there are pairs of words which are

approximately opposite in meaning: persistentand changeable, wary and headstrong, guardedand impulsive. All of these words are on roughly

the same continuum, that of control of emo-tionality and behavior, but at opposite ends.All may be thought of as representing the ex-tremes of over- and under-control, and contrastsharply with the more adaptive, less effortfulcontrol found in the normals. The inferencewould be that the prehypertensives betraytheir greater degree of emotional upset, someby behaving erratically, others by a tense kindof repressive control. Both kinds of behaviorindicate tension which is unresolved and aretherefore inadequate. The important thing isthe fact of greater emotional upset. The pre-hypertensives are less efficient in meeting these

TABLE 3. Adjectives Checked Significantly More Oftenfor the Nor.tmals than for thre Prehypertensives by

One or Two Observers

adaptableaffectedaffectionateapatheticappreciativebroad-mindedcautiousconciliatoryconscientiousconservativeconsiderateconsistentcontentedconventional

cooperativedependableeasy-goingefficientevengentlehumanekindliberallikeablemannerlymeticulouspeaceablepleasure -seeking

poisedpreciserelaxedrespectfulsimplesociablesophisticatedsympathetictactfultendertranquiltrustingunambitiouswitty

emotion-provoking situations in both a psycho-logic and (presumably) a biologic sense.

Turning now to those adjectives which theobservers checked more frequently for thenormals than for the prehypertensives, wefind a quite different list of words, as tabulatedin table 3.Note first that most of these words are

positively toned, that is, they are "nice"things to say about people, quite in contrastto those checked for the prehypertensives.For example, there are words like kind, adapt-able, likeable, cooperative, sociable, tactful.Second, note the absence of words denotingimpulsiveness and, in their place, words sug-gesting tempered control, like consistent, even,adaptable, efficient. Third, many more words inthis list than in the prehypertensive list describe

877

by guest on July 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

RESPONSE TO PSYCHOLOGIC STRESS

relations with people, for example such wordsas gentle, affectionate, cooperative, considerate,sympathetic, trusting, humane. Fourth, thesewords describe persons with social poise, forexample tactful, poised, relaxed, tranquil,sophisticated, contented.

Clearly the groups behaved differently inthese stressful situations, and consistent per-sonality pictures emerge. An important over-all difference between the groups, but onewhich is difficult to document briefly, is thatthe prehypertensives were less effective incarrying out their assigned tasks, that is, ingetting the dry cleaner and dean to complywith their requests. This is evidenced in partby the fact that although the prehypertensives

they are less effective in achieving their ends.Further, they are likely to create an unfavor-able social impression, just as they did in ourobservers.

III. Self-Descriptions

In order to obtain a third and entirelydifferent kind of data, each subject was askedto check on the same list of adjectives used bythe observers those words which she thoughtdescribed herself. The data were analyzed inthe same way as the observers' checks, and twolists compiled. These suggest the subjects'perceptions of themselves and their behavior.

Table 4 lists those adjectives checked morefrequently by the prehypertensives than by

TABLE 4.-Adjectives Checked Significantly MoreOften by Prehypertensives than by Normals-

Self-Checks

adventurous foolish rashartistic frugal reflectiveawkward guarded repentantclever hotheaded self-possessedcompassionate humble self-reproachfulconceited hurried self-reprovingconfused indifferent shycurious nervous simpledaring original spendthriftdocile passive submissivedreamy peculiar tenderegotistical preoccupied tensefearful prudent tranquil

were checked on rating scales as being more

involved in the situations, they were alsochecked as being less well controlled andorganized than the normal subjects, who tookthe scenes much more "in stride." In playingtheir roles, the staff members tried to be con-

sistent from one subject to the next, makingeach scene roughly equally stressful. Appar-ently the prehypertensives interpreted thesituations as more stressful than did the nor-

mals and behaved as if these situations were

more emotion-provoking. At the risk of over-

generalizing from one situation, a semilabora-tory one at that, one may say that the pre-hypertensives are likely to bring more anxietyto real-life problem situations, that they be-come more emotionally involved, and that

TABLE 5. Adjectives Checked Significantly MoreOften by Normals than by Prehypertensives-

Self-Checks

adaptablebossycharitablecollectedefficientfair-minded

frankhumaneindependentlevel-headedmature

meticulouspersistentreasonablethoroughthoughtfultolerant

the controls. In general this list is similar tothe observers' checks for the prehypertensives.There is the same negative feeling tone tomany of the words; and several of the words,for example, hurried, hotheaded, rash, daring,adventurous, tense, describe the same defectof control noted by the observers. Other wordswith a negative feeling tone are awkward, con-fused, peculiar, and foolish. Also there are wordssuggesting moodiness and egocentricity. Ingeneral the agreement with the observers'check list is striking, except for one importantdifference. Note such words as submissive,docile, passive, and humble, which imply a lackof dominance not noted by the observers.

Table 5 lists those adjectives checked moreoften by the normals than by the prehyper-tensives. Note here that the words suggest apositive self-evaluation, for example the wordsindependent, mature, level headed, collected, andadaptable, also that they describe a steady kindof control, for example the words efficient,thorough, persistent. and collected. A third

878

by guest on July 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

ROBERT E. HARRIS, ET AL.

characteristic of these words is an impliedbenevolent dominance in social relations, wordssuch as humane, tolerant, charitable, fair-minded, and reasonable.Thus we have evidence that the self-ap-

praisals of the prehypertensives and the con-

trols are quite different and, further, thatthey are congruent with the impressionscreated in the observers of the stressful situa-tions. The validity of the differences noted bythe observers is therefore reinforced. Addi-tionally these self-appraisals provide groundfor generalizing beyond the experimentalsituations. The prehypertensives, thinking ofthemselves in these negative terms, as moodyand egocentric and as less well controlled thanother persons, would probably behave lesseffectively in ordinary everyday problemsituations and would more often experienceanxiety and frustration.

SUMMARY

In summary, it is clear that by technics ofpersonality study, college women who have a

high actuarial probability of becoming hyper-tensive in later life can be differentiated fromnormal controls. Our data indicate that theprehypertensives responded characteristicallyto a psychiatric interview and to interpersonalsituations involving psychologic stress. In thestressful situations they behaved less effec-tively, were less well controlled, less poised,and socially they created a less favorable im-pression than the conti ols. The validity of theseinterpretations was reinforced by the agree-

ment between the observers' descriptions andthe subjects' own self-appraisals. Finally, ina psychiatric interview they responded simi-larly to patients with clinically diagnosedhypertension, and the personality patternsdescribed were like those noted by the observersduring stress.

It is suggested, therefore, that the prehyper-tensives are less well able than normal personsto handle stressful or frustrating situations (ofwhich the initial college physical examinationis one) without becoming emotionally upset,and that therefore they are more likely to besubject to the autonomic accompaniments ofemotion, including repetitive rises in bloodpressure. Such repetitive rises may well beone of the conditions, an important even thoughperhaps not a necessary or sufficient one, inthe network of events leading to fixed hyper-tension.

SUMARIO ESPAROL

Mujeres colegiales prehipertensas y controlesemparejados fueron expuestas a situacionesprovocadoras de emociones, examenes psico-l6gicos y entrevistas psiquiatricas. Modelos derepuesta diferenciando un grupo del otro seobservaron. Las prehipertensas eran menoscontroladas, mas impulsivas, mas egocentricasy generalmente menos adaptables en las situa-ciones tensas. En las entrevistas psiquialtricascomportamiento similar al de pacientes conhipertension se observo. Los hallazgos sugierenque los prehipertensos son mas vulnerables ensituaciones que envuelven tensi6n psicologicay naturalmente mals sujetos a las emocionesautonomicas concomitantes, incluyendo repeti-das alzas en la presi6n arterial.

REFERENCES1 HINES, E. A.: Range of normal blood pressure and

subsequent development of hypertension.J.A.M.A. 115: 271, 1940.

2 DIEHL, H. S., AND HESDORFFER, M. B.: Changesin blood pressure of young men over a sevenyear period. Arch. Int. Med. 52: 945, 1933.

3 LEVY, R. L., WHITE, P. D., AND STROUD, W. D.:Transient hypertension: The relative prognosticimportance of various systolic and diastoliclevels. J.A.M.A. 128: 1059, 1945.

879

by guest on July 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

FREEDMAN and SAMUEL P. HUNTROBERT E. HARRIS, MAURICE SOKOLOW, LEWIS G. CARPENTER, JR., MERVIN

Response to Psychologic Stress in Persons Who Are Potentially Hypertensive

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1953 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.7.6.874

1953;7:874-879Circulation. 

http://circ.ahajournals.org/content/7/6/874the World Wide Web at:

The online version of this article, along with updated information and services, is located on

  http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document.

Permissions and Rights Question and Answer Further information about this process is available in therequested is located, click Request Permissions in the middle column of the Web page under Services.the Editorial Office. Once the online version of the published article for which permission is being

can be obtained via RightsLink, a service of the Copyright Clearance Center, notCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:

by guest on July 14, 2018http://circ.ahajournals.org/

Dow

nloaded from