_______ CHAPTER 21

Social Psychology and Health

EmergeDIAdaptatic

During th,ologists ,

(1956) desponses (related HI:

tric secrepressure)sors or cheat. coldcial psycldrome tht

physiologigists. has fstimuli ancport) ailephysiolog)acteristic 0

demiologi~

gists. hashealth andteristics astion. Thishistorical,viding insperspecti v'States is aofsmokinllevel havemarkedly;past quartchanges­increasesnorms. attduced byworking tlganizationopment w.macroscolhow stres~

social fonfect psychserved pat

Although the organism-environment adaptationperspective operates at hoth microscopic and mac­roscopic levels, the two sometimes have been es­tranged by disciplinary boundaries. The moremicroscopic perspective, characteristic of psycho-

chosocial environment (e.g., smoking, lack of exer­cise. and stress).

The importance of psychosocial influences onhealth is now a major concern of biomedical sci­ence and also a major focus of social psychologicalresearch. This convergence has led to a new per­spective on health which assumes that: (I) illnesshas multiple determinants. both biomedical andpsychosocial; (2) what is seen as a disease is notinvariant over time but changes based on sociocul­tural and biological definitions; and (3) the medicalprofession is a social institution that shapes itsmembers' views based on broad sociocultural con­siderations that go beyond scientific concerns.

It is impossible to encompass in a single chap­ter the full range of research that has evolved fromthis perspective; our review is necess~rily selec­tive. [n the first half of the chapter, we discussresearch on the psyc~osocial determinants of ill­ness, and in the second half we discuss research onthe psychosocial determinants of illness definitionand response. In each section we present an histori­cal overview. discuss recent developments, andpropose future directions.

PSYCHOSOCIAL DETERMINANTSOF ILLNESS

RONALD C. KESSLERJAMES S. HOUSE

RENEE R. ANSPACHDA VID R. WILLIAMS

Until the early part of the twentieth century.biomedical researchers attended nearly exclusivelyto organic pathogens and assumed that: (I) diseaseis a deviation from normal biological functions;(2) diseases are generic and invariant over time andspace; (3) medicine is a scientifically neutral pro­fession uninfluenced by wider social. cultural, andpolitical forces; and (4) each disease has a specificbiological cause (Mischler 1981). This last assump­tion, known as the "doctrine of specific etiology,"implied that a disease is best controlled by treatingthe culpable biological agent. The elimination ofpolio via vaccination during the 1950s dramati­cally epitomized the successful application of thismodel.

Despite such successes, some contemporarycritics argued that disease is often a normal bio­logical response to abnormal environmental de­mands rather than a biological deviation, and thathealth varies over time as a function of changingenvironmental demands (Dubos 1959). Such no­tions construed health as a state of adaptation be­tween the individual and his/her environment.These ideas have deep historical roots. They in­spired the public health movement of the midnine­teenth century to emphasize the importance of abenign physical environment for health. In theearly twentieth century. as public health advancesled to a shift in the major causes of death fromacute infectious diseases to chronic diseases. theseideas were broadened to emphasize the importanceof not only the physical environment (e.g.. cleanair, water, food, and sanitation) but also the psy-

548

physiologists and psychological social psycholo­).'lstS. has focused on how proximal environmental'>timuli and contexts (e.g., stress and social sup­port) affect individuals' behavior. mood. andphysiology. The more macroscopic approach, char­acteristic of many sociologists, demographers, epi­demiologists. and sociological social psycholo­"ists. has focused on the broader distribution ofhealth and illness in populations by such charac­teristics as age, race. gender, and geographic loca­llllO. This macroscopic approach also considershistorical and contextual influences on health. pro­viding insights not apparent from an individualperspective. The decline in smoking in the UnitedStates is a good illustration. The long-term effectsl1f smoking intervention programs at the individuallevel have been modest. Yet smoking has declinedmarkedly and steadily in the United States over thepast quarter century due to broader contextualchanges-restriction of smoking in public places,increases in costs, and a changing set of societalnorms. altitudes, and values regarding smoking in­duced by political and mass media institutionsworking through intermediate levels of social or­ganization. An important direction for future devel­opment would integrate the more microscopic andmacroscopic traditions in an effort to illuminatehow stress and adaptation are structured by broadsocial forces and how microsocial phenomena af­fect psychophysiological processes to produce ob­served patterns of health and illness.

Emergence of the Stress andAdaptation Perspective

During !.he middle of the twentieth century, physi­ologists Walter Cannon (1932) and Hans Selye( 1956) described a syndrome of physiological re­sponses (including adrenocortical secretions andrelated neuroendocrine activation, increased gas­tric secretions. and higher heart rate and bloodpressure) to a wide range of environmental stres­sors or challenges, including infectious agents,heat. cold, physical pressure and restraint, and so­cial psychological threat. Selye called this syn­drome the "general adaptation syndrome" (GAS).

CHAPTER 21 Social Psychology and H~allh 549

According to Selye. GAS originally evolved as anadaptive response to physical stressors but has be­come maladaptive in modern society. where manystressors are chronic and inescapable. In the face ofthese modern stresses. GAS can lead to what Selyecalled "diseases of adaptation," such as hyperten­sion, heart disease. ulcers. and arthritis.

Although Selye's model provides a usefulframework for understanding how psychosocialstresses promote physical illness. it does not ex­plain how the SUbjective sense of stress itself isgenerated. Subsequently. social psychologists de­veloped the framework known as the stress andadaptation model. This framework suggests thatcharacteristics of situations and individuals com­bine to create perceptions of stress or threat, whichideally elicit responses that reduce stress and pro­tect health. Failure to do so results in ill health viamechanisms such as those originally explored byCannon and Selye or those identified by more con­temporary biopsychosocial researchers. Whethersituations are experienced as stressful and how per­sons respond to them is now seen as a function ofboth preexisting personal dispositions and otheraspects of the situation. sometimes referred to asmoderating. buffering. or vulnerability factors (seeHouse 198 L Lazarus and Fol kman 1984: McGrath1970; for variations on this framework).

Since the mid-1970s, the stress and adaptationframework has stimulated major developments inthe study of the perception of and response to po­tentially stressful situations in social psychologyand related areas of medical sociology and healthpsychology. The evidence linking stress and otherpsychosocial factors to health is growing steadilyand. although as of yet not definitive. is made quiteplausible by the convergences of both laboratorystudies of animals and humans and nonexperimen­tal research on broader human populations. re­cently augmented by intervention studies.

The initial breakthroughs in research on psy­chosocial determinants of illness occurred as thepsychosocial stress and adaptation perspective wasjust emerging. Initiated by physicians rather thansocial scientists. such work has been superseded bymore psychosocially sophisticated and empirically

550 PART III Social Siruciure, Relalionshirs, and the Individual

accurate formulations. However. these initial re­search programs played a critical role in estah­lishing that psychosocial factors are important inthe etiology of morhidity and mortality and havestimulated important continuing developments inpsychosocial research.

Psychosomatic Medicine

Heavily influenced hy the hiomedical model. oneapproach to studying psychosocial factors in healthhas heen to focus on a speci fic disease and identi fyputatively distinctive psychosocial characteristicsof individuals with the disease. The earliest sys­tematic efforts of this type were made under theruhric of "psychosomatic medicine," which in­itially involved the extension of psychoanalytictheory and methods to prohlems of physical illness.Franz Alexander (ILJS()). an early leader in thisfield. expounded a psychosocial doctrine of spe­cific etiology:

1'11('1'(' ;.1' II IIwit 'T;d"II"" J!1U/jIlS/ ".I' "('1'/";11 I"I/lto­log;,."IIII;,.roOl'g"";,IIIS !Jl/\'(' " SI,,.,.i/i(' al/illi/I' /f'I'('''1'/'';11 ol'g"I/.\'. so also ('''1'/'';11 ,.1I1O/;1II1U1 ('(I/I/Ii('/,I

1'0.1',1".1'.1' ,l/w('i/;,;/;,.s "lid (/(col'dillgll' /<'IId /0 "tflid

,'''1'/'';11 ill/,.I'I/"I Ol'g""s (I', -+7),

Suhselluent work related personality attrihutes10 specific illnesses. Suppressed aggression. forexample. was hypothesized to cause cardiovasculardisease. while dependency conflicts were impli­cited in an ulcer-prone personality. This traditionhas continued in efforts to identify personalitytraits and syndromes causally associated with dis­eases as diverse as cancer (Levy and Heiden 1l)\I()l.arthritis (Anderson ILJX5). diahetes (Ohwovoriolcand Omololu ILJX6l. and heart disease (FriedmanIl)l)(); Friedman and Booth-Kewley I\lX7).

Such research has suffered. however. frommultiple methodological prohlems. Studies arcmostly cross-sectional or retrospective. rather thanprospective in design. Samples arc often unrepre­,cntat ive and analyses often fai I to control ade­quately for e\ogenous third variahlcs. which may,puriously produce associations hetween diseaseand personality. Thus it is difficult to interpret the

associations reported in this literature in a way 111.11

rigorously evaluates the influence of person~illll

on disease (Anderson ILJXS: Ohwovoriole ;11101

Omololu I\lX6).

The psychosomatic research has also ht'l'llflawed theoretically. Generally. a given person.ilillvariahle has heen studied in relation to only a Sill

gle disease. thus making it impossihle to knOllwhether that variahle may have similar assol'l;\tions with other diseases. A doctrine of specillletiology has often heen assumed rather than elll

pirically demonstrated. Recent careful reviews "I

the literature on personality and multiple dise~lsl'

outcomes. in fact. demonstrate that most "person,1Iity" variahles that show associations with one discase show similar associations with other diseascs,For example. meta-analyses hy Howard Friedm;1I1and Stephanie Booth-Kewley (ILJX7) found thaianxiety and depression are associated with coronary heart disease (CHD). asthma. arthritis. ulcers,and headaches. while a complex of variahles illllicating anger/hostility/aggression is associated withCHD. asthma. and arthritis. though prohably nolulcers and headaches. Thus. they argue that a generally "disease-prone personality" syndrome or sciof trait:-; is more likely to exist than distinctive"arthritis-prone," "ulcer-prone," and so on person­alities. We return to more contemporary researchon personality helow.

Type A or Coronary-Prone Behavior Pattern

Research initialed hy Meyer Friedman and Ra)Rosenman ( ILJ74) on what they termed the "coro­nary-pronc hehavior pattern" or "coronary-pronepersonality" played an evcnmore central role in theemergence of psychosocial theory and research onthe etiology and epidemiology of physical healthand illness. At the same time. this research illus­lI'ates the prohlems posed hy the doctrine of spe­ci fic etiology. Friedman and Rosenman were prac­ticing cardiologists who heliL'ved they sawcharacteristic patterns of hehavior in many of theirpatients. The) termed this the ·'type A" or coro­nary-prone hehavior pattern. which they conceivedas the result of an interaction hetween personality

'I,.,r

'if«Ld

: ~Jispohigh (

:; time I

via clmeth(

1 caron,~ studit

l Prone

I Heartthe tyheartchole

j F

f initialrelati(,.popult perha

J

poseswidelsugg(carre

j ity arrecen'

Itivelyheha\with,than I

ingly~ CHD1t{' teristiI'>

"doma

,~

may t;j

i tainty

1 numhlates,

I~ factor

I disOf(

1 Life Ii

Drawand Cthat I

pathothat cadapt;

dispositions and challenging situations leading tohigh degrees of competitiveness. job involvement.time urgency, and hostility. Type A, assessed firstvia clinical interviews and later via questionnairemethods, predicted both the onset and course ofcoronary heart disease in several major prospectivestudies, leading the Review Panel on Coronary­Prone Behavior and Heart Disease of the NationalHeart. Lung and Blood Institute (1981) to certifythe type A personality as a risk factor for coronaryheart disease. in the same general class as smoking,cholesterol, and blood pressure.

Further research, however, has confused theinitial understanding of type A personality and itsrelation to CHD. In recent studies of high-riskpopulations, type A has failed to predict CHD,perhaps because the controlling type A style predis­poses one to adapt health-promoting behaviorswidely publicized in the 1980s. Other researchsuggests that the effects of type A may be due tocorrelates (e.g., mistrust) or components (hostil­ity and anger) (Matthews 1985). Finally, morerecent research suggests that as with other puta­tively disease-specific personality variables orbehavior patterns. type A personality is associatedwith a range of health problems and diseases otherthan CHD. As research and theory turn increas­ingly from seeking the psychosocial causes ofCHD to understanding how psychosocial charac­teristics of persons and situations affect the fulldomains of health and illness, the type A constructmay become obsolete, leaving as a legacy the cer­tainty that psychosocial variables. including anumber of key personality and situational corre­lates and components of type A, are significant riskfactors for a range of physical and psychologicaldisorders.

Life Events and Change

Drawing somewhat loosely on the ideas of Selyeand Cannon and a variety of research indicatingthat major life changes could be stressful andpathogenic. Holmes and Rahe (1967) hypothesizedthat change, whether for better or worse, requiresadaptation, and that high levels of adaptive effort

CHAPTER 21 Social Psychology and Health 551

could produce both physical and psychological dis­order. They constructed a Social ReadjustmentRating Scale, which asked individuals to indicatewhether they had experienced each of forty-threelife changes in the preceding year and assigned toeach the average rating of the amount of adjust­ment involved in a model by a panel of individualsof varied social backgrounds. The summed total ofadjustment units for an individual has been repeat­edly found to predict the onset of a wide range ofphysical and psychological disorders (cf. Cocker­ham 1986, 76-80: Mirowsky and Ross 1989).

Further research has suggested tlaws in boththeory and measures propounded by Holmes andRahe (1967). Most important, subsequent researchshows that it is only more serious negative events, notchange per se. that adversely affect health. Further,the life change weights of Holmes and Rahe pro­vide little more predictive power than a simpleunweighted sum of the number of serious negativeevents. the most serious of which (e.g., widow­hood, divorce, unemployment) also have been foundto have separate effects on morbidity and mortality,including cancer (Sklar and Anisman 1981), heartdisease (Wells 1985), and autoimmune diseasessuch as rheumatoid arthritis (Solomon 1981). Al­though research in this area is just beginning, theavailable evidence suggests that life events may bemore important in predicting the course of illness(e.g., speed of recovery, recurrence) than initialonset (Kessler and Wortman 1988).

A New Focus on Chronic Stress

For a time, the study of life changes and events wasalmost synonymous with the study of stress andhealth. During the 1980s, however, renewed atten­tion was focused on chronic stress and deprivationas determinants of illness (Mirowsky and Ross1989; Pearlin 1989). These new studies suggestthat it is the more enduring stressful sequelae ofsuch events that explain thei.. effects on health. Forexample, the adverse effects of unemployment onhealth are partly mediated by resultant financialstresses (Kessler, Turner, and House 1987), whilethe relationship between widowhood and health is

552 PART III Social StTllcture. Relationships. and the Individual

partly due to the effects of social isolation (Umber­son, Wortman. and Kessler 1992).

This renewed attention to chronic stress drawson two long-standing epidemiologic research tradi­tions as well as on laboratory and field experi­ments. The first is a tradition of research on workand health that has suggested that high levels ofphysical and psychological demands (e.g .. work­load. conflict, responsibility) can adversely affecthealth. Early work in this tradition compared ag­gregate morbidity and mortality profiles of differ­ent occupations that are comparable on all knownrisk factors other than job stress. yielding strikingevidence that indirectly implicated job stress as apowerful determinant of ill health (Kasl 197R). Amore recent approach has been to use multivariateanalysis to study the effects of job conditions onworker health at the individual level of analysis.The most persuasive studies of this sort have usedlongitudinal data to determine the effects of jobconditions on changes in health over time. Severalsuch investigations have documented significanteffects of job pressures and conflicts on mortality.coronary artery disease. peptic ulcers. diabetes, andpsychological distress (e.g.. House and Cottington1986; Karasek and Theorell 1990).

A second basis of the renewed interest in chronicstress is the persistence in the United States andmost other developed countries of socioeconomic,racial. ethnic. and gender differences in physicaland mental health, despite substantial progress inpublic health and the equalization of access tomedical care (Cockerham 1986; Marmot, Kogevi­nas, and Elston 1987). Although gaps undoubtedlyremain in access to quality and preventive care. andalthough biological factors play some role in theseaggregate differences, a growing body of researchsuggests that differences in exposure to chronicstress as well as other psychosocial risk factorsmay be central as well. For example. chronic finan­cial stress plays a significant role in explainingsocioeconomic differences in health. and socioeco­nomic factors are central to racial differences inhealth (House et al. 1992).

A major methodological problem in this re­search is that measures of chronic stress. which are

typically based on self-reports. may be affected 11\acute and chronic life conditions and thereby COli

founded with current levels of health. Resolutiollof this methodological problem will require prospective studies that measure perceived levels 01

chronic stress at several points in time and lISl'

these reports to predict subsequent morbidity andmortality. controlling for health level at the time 01

the measurement of stress. In one effort of thistype. House et al. (19R6) found that men who rc­ported high levels of chronic occupational stress attwo points in time separated by more than twoyears were three times more likely to die over thesucceeding decade than men who reported lowerlevels of job stress at either or both times, afteradjustment for age, education. a variety of healthindicators (e.g .. blood pressure and cholesterol),and health risk (e.g., smoking) at the initial point ofmeasurement. Similar research is needed on theeffects of financial. marital. parental. and otherchronic stresses.

Until such research is done. experimental stud­ies on animals and humans will continue to providethe most powerful evidence concerning the effectsof chronic stress on ill health. Several laboratoryexperiments and quasi-experimental studies haveexposed humans to mild stress (e.g .• demandinglevels of workload. responsibility. or conflict withothers) and have shown effects on a wide range ofphysiologic outcomes. including cardiovascularfunctioning (Manuck et al. ]989), neuroendocrinefunctioning (Krantz and Manuck 1984). and cellu­lar immune response (Cohen. Tyrell, and Smith1991). Although the stressors used are, for obviousethical reasons. too mild to cause serious or pro­longed health impairments. their effects recall themore marked manifestations of naturally oc<.:Urringlife crises. Animal experiments confirm data on hu­man subjects regarding the pernicious physiologi­cal effect of stress. with recent studies documentingthat long-term exposure of mice and monkeys tothreatening social situations leads to impaired im­mune response to a variety of infections (Cohen etal. 1992).

In 1960. Jackson et al. demonstrated experi­mentally that people with chronic role-related

·0

stresses alper respir,to a nasathan a ntdocumenttion throu,is unclearnificant. Fson (1991

erature. thstress is ,control foexposurestudies b)these metIpie of healow doseneutral sasubjects fand contnences andinfectivit)of negatinegativesignificarcontroll ining varialInterestin.across thestress am(i.e .. the (to life ev(once infel

Vulnerat

A consishgation re\pie who;stressfulhealth pr'health tritally. the'that havetion. vari(pabilities

,tresses are more likely than others to develop up­reI' respiratory infection when randomly exposedto a nasal spray containing viral material ratherthan a neutral solution. Although these studiesdocumented effects of stress on resistance to infec­tion through various aspects of immune function. itis unclear whether these effects are clinically sig­nificant. Further. as noted by Cohen and William­son (1991) in a comprehensive review of this lit­erature. the few studies that directly document thatstress is associated with infectious illness fail tocontrol for the confounding effects of differentialexposure or health behaviors. A series of recentstudies by Cohen et al. (1991) resolved many ofthese methodological problems by exposing a sam­ple of healthy volunteers via nasal drops to either alow dose of one of five respiratory viruses or to aneutral saline solution and then quarantining thesubjects for a full week after exposure to monitorand control their subsequent environmental experi­ences and behaviors during the period of potentialinfectivity. Results showed clearly that measuresof negative life events. perceived stress, andnegative affect assessed prior to the challengesignificantly increased risk of developing a cold.controlling for a wide range of potential confound­ing variables (including prechallenge antibodies).Interestingly. the pathways of these effects differedacross the three stress measures. with perceivedstress and affectivity increasing risk of infection(i.e .. the development of antibodies) and exposureto life events increasing risk of clinical symptomsonce infected.

Vulnerability Factors

A consistent finding across all the areas of investi­gation reviewed above is that the majority of peo­ple who are exposed to all but the most extremestressful life experiences do not develop serioushealth problems. Current research on stress andhealth tries to explain this finding and. more gener­ally. the variation in stress reactivity. The factorsthat have been examined include biogenic constitu­tion. various aspects of personality. intellectual ca­pabilities such as cognitive tlexibility and effective

CHAPTER 21 Social P,ycho!ogy and Health 553

problem-solving skills. interpersonal skills such associal competence and communication ability, andsocial resources. including financial assets andcoping styles. Because full consideration of thisdiverse array of studies is beyond the scope of thischapter. we focus on two classes of variables thathave generated intense interest over the past dec­ade: social relationships and support and disposi­tional/personality variables. especially what hasvariously been termed "control." "efficacy." or"mastery."

Social Relationships and Support. Current inter­est in the effects of social relationships and supporton health was triggered by several influential pa­pers published in the mid-1970s that reviewed di­verse studies demonstrating that such things asmarital status. geographic stability. and social inte­gration are associated with both mental and physi­cal health (Caplan 1974; Cassel 1976; Cobb 1976).A theme present in all these associations seemed tobe access to social ties and supports. Here. as in thecase of stress. the available evidence has comefrom experimental studies of animals and humans.as well as from nonexperimental studies of humanpopulations.

The presence of a familiar member of the samespecies buffered the impact of experimentally in­duced stress on ulcers. hypertension. and neurosisin rats. mice. and goats. respectively (Cassel 1976).The presence offamiliar others also reduced physi­ological arousal (e.g .• secretion of free fatty acids)in humans in potentially stressful laboratory situ­ations (Back and Bodgonoff 1967). Such effectsmay even operate across species, with affectionatepetting by humans reducing the cardiovascular se­lJuelae of stressful situations among dogs, cats,horses. and rabbits (Lynch 1979. 163-80) and eventhe arteriosclerotic impact of a high-fat diet onrabbits (Nerem. LeveslJue. and Cornhill 1980).

Nonexperimental studies of human popula­tions have devised scales to measure social integra­tion and support and demonstrated that thesemeasures are associated with health. The most in­tluential of these studies examined the effects ofsocial relationships on subselJuent mortality in pro-

lI

I,I

II:iit!

I

554 PART III Social SlruClure, Rclalinnships, and Ihe Individual

spective surveys of the general population, The

first study of this sort showed that marriage. con­

tact with family and friends. church membership.and affiliation with other social groups were allassociated with reduced mortality risk over a nine­

year follow-up period in a large sample of respon­dents living in Alameda County. California (Berk­man and Syme 1979). Subsequent reports byBlazer ( 1982) and House. Robbins. and Metzner( 19lQ) showed similar patterns in other longitudi­

nal community surveys in the United States. sincereplicated in a number of European studies (seeHouse. Landis. and Umberson 1988 for a review).All of these reports were based on secondary analy­ses and none contained a comprehensive set ofsocial support measures, Therefore. though theyprovide strong evidence that social relationshipsincrease longevity. they do not allow an estimate of

the full extent of this inlluence or an understandingof the precise components of relationships. suppor­tive or otherwise. that are involved.

Similar longitudinal studies have examined

the association between support and onset of physi­cal illness. The most rigorous of these have focusedon coronary artery disease and are reviewed byBerkman (19R5). Despite broad consistency infinding some indicator of social relationships orsupport associated with decreased morbidity risk.there are inconsistencies. For example, social tiesare associated with disease incidence but not pre­valence in some studies. while in others the onlysignificant predictors are associated with preva­lence, The effects are limited to lower-class womenin one major study and appear only among men inanother. The aspects of social relationships thatseem to promote health vary across studies as well.Inconsistent findings of this sort further obscurethe mechanisms involved in the effects of support.

While research on social relationships andphysical health has focused primarily on direct ef­fects. research on social relationships and mentalhealth has been more concerned with stress-buffer­ing effects. In an influential research program. forexample. Brown and Harris (llI7R) showed that theimpact of stressful life events on depression was

substantially reduced among respondents who 10.,.1

an intimate confiding relationship with a l,il'lId "

relative. While nearly 40 percent of the .,1'"" ,\women studied without a confidant becanll' I k

pressed. only 4 percent of those with a conlld.II'1did so. This paradigm has subsequently been Il",11cated in many studies. and the general paltl'lII ,,,results clearly shows that access to a confidalll :111,1

perceived availability of crisis support arc a..,.,\l1 I

ated with a reduced impact of stressful life cn'II"

on depression and anxiety (Cohen and Wills I'His

Kessler and Mc Leod 1985).

Another line of investigation has used nwas

ures of social relationships and support to predllladjustment to specific life crises. such as widowhood (e.g., Umberson. Wortman. and Kessler 199.',and unemployment (e.g .• Kessler. Turner, and HOll'"1987). Almost all of these studies have been con

cerned with mental health outcomes. and l11o.,1have found that measures of social relationship"obtained shortly before or after a crisis are sigl1lt I

cant predictors of subsequent emotional adjuslment. Moreover. these focused studies begin 10

suggest that specific kinds of supportive ties maybe most helpful for particular problems. For exam­ple. Hirsch (1979) showed that low-density networks that facilitate contact with new people arcparticularly useful when the coping task is to obtain new information or adopt a new role. Interven­tions aimed at providing coping skills and supporthave been shown to reduce adverse health const:­quences of unemployment (Price. van Ryn. andVinokur 1992) and widowhood (Raphael 1977).

Studies of specific life events or crises alsoprovide an opportunity to examine social supportprocesses in relation to other determinants of ad­justment. such as appraisals and coping strategie~.

and in this way clarify the mechanisms throughwhich support may protect against illness. Lifecrisis studies conducted to date have not fully real­

ized this potential but have provided two very pro­vocative and consistent results. One of these in­volves miscarried support efforts. and the otherinvolves the distinction between perceived supportand received support.

The first

,'\aminatiomIV hat supportSuch studiesaujustmentl'fforts can a

'!uences. suclent (Coyne.portive actiolimes can lea\upport had1l}86). Evide

interest in de

ics of actualdeveloping tlnping suppor

This evi

support tramlinding: whilable is aSSOCi

to stress. thelion is medi:

One possibltthe perceptiopromotes adjactual receiptseveral ways

ity might lealas less threa

chological etAlternativel),"safety net" ,

efforts (Rool

ceived SLlPP!some unmea~

cially compesupport andand Swindleto evaluate e:

rod. and Clal!l}l}() ).

Personality.sonality traitillnesses pro

The first of the two results is based on focused'''lminations of exactly what supporters do andwhat support recipients think about these efforts.'illl:h studies show that while support can promote.Idjustment to stress, well-intentioned support,'Ilorts can also have unintended negative conse­'1l1~nces, such as making recipients feel incompe­kill (Coyne, Wortman, and Lehman 19H8). Sup­portive actions can also create social costs that attlllles can lead to greater emotional distress than if'lipport had not been obtained at all (LiebermanI'lX6). Evidence of this sort has led to a heightenedl'i1~rest in detailed descriptive work on the dynam­Il~ of actual support transactions, both as an aid ind~veloping theory and as a practical guide to devel­"ping support interventions (Sandler et al. 1988).

This evidence on the mixed effects of actual~lIpport transactions has led to another importantIinding: while the perception that support is avail­:Ihle is associated with good emotional adjustment10 stress, there is little evidence that this associa­lion is mediated by the actual receipt of support.One possible interpretation of this finding is thatIhe perception of support availability itself activelypromotes adjustment to stress over and above any:Iclual receipt of support. This could occur in any of,everal ways. The perception of support availabil­Ity might lead to an appraisal of stressful situationsa~ less threatening, thereby decreasing their psy­lhological effects (Wethington and Kessler 1986).\lternatively, it could provide a psychological"~afety net" that helps motivate self-reliant copingefforts (Rook 1990). The putative effects of per­leived support on health may actually be due to,orne unmeasured common cause; for example, so­lially competent people may be more able to attract,upport and to manage stressful situations (Heller,lI1d Swindle 1983). Ongoing research is attemptingto evaluate each of these possibilities (Cohen, Sher­rod. and Clark 1986; Sarason. Pierce, and SarasonIYYOl.

Personality. Although the early search for per­,tmality traits uniquely associated with particularIllnesses proved fruitless. more sophisticated cur-

CHAPTER 21 Social Psychology and Health 555

rent research has focused on personality disposi­tions that can affect a broad range of health prob­lems, either directly or as vulnerability factors.This new approach to the investigation of personal­ity effects on health has gained momentum only inthe 1980s and is therefore less well-developed thanresearch on social relationships and support. Inparticular, there are few good prospective data onpersonality and health nor have the causal path­ways linking personality to health been investi­gated in great detail. Only a few studies of person­ality and health have gone beyond main effectsanalyses to examine whether there are interactionsbetween personality and stress in predicting illhealth.

Most studies that have examined stress-buffer­ing effects of personality are either studies of men­tal illness or laboratory studies of stress and infec­tious disease. The former have documented thestress-buffering effects of self-esteem, perceivedcontrol, and hardiness and the stress-exacerbatingeffects of neuroticism and interpersonal depend­ency (Pearlin et al. 1981; Cohen and Edwards1989). The latter have shown that introversion,social skills, and negative affectivity all modify theeffects of mild stress on infection (Cohen and Wil­liamson 1991).

One of the most intriguing areas of investiga­tion in this literature concerns a personality dispo­sition variously termed self-efficacy, mastery, orcontrol. Sutton and Kahn (1984) reviewed a varietyof laboratory and field studies on this concept andfound consistent evidence suggesting that indi­viduals who have a greater chance to predict, un­derstand, and control events in their lives experi­ence less stress and fewer adverse effects of stresson their physical and mental health. In a relatedresearch program, Karasek and Theorell (1990)have shown that lack of control over one's workenvironment is a risk factor for cardiovascular dis­ease and psychological distress, both directly andthrough a tendency to exacerbate the deleteriouseffects of other occupational stresses. Pearlin andcolleagues ( 1981 ) have shown that a sense of mas­tery promotes mental health and buffers the impact

II

I1i1t

!~-f:

itI:t

Iit[!

!f

III

556 PART III Social Structure, Relationships, and the Individual

of acute and chronic stress on mental health, Amajor program of research by Rodin (19R6) andothers has demonstrated that increased control overone's social environment can promote better physi­cal and mental health, and even longer life, perhapsespecially for older persons. Langer and Rodin(1976) designed an inexpensive set of structuralinterventions in nursing homes to create opportuni­ties for mastery experiences. Markedly positive ef­fects on psychological well-being, physical health,and even the longevity of nursing home residentswere documented.

Finally, research on cancer and personalitysuggests that feelings of helplessness and hopeless­ness, as well as repression or denial of emotions,may both predispose people to the onset of cancerand exacerbate its course (Levy and Heiden 1990).Here animal studies yield especially dramatic re­sults. Animals induced to become helpless throughbehavioral restraint and repeated exposure to stress(e.g., electric shocks) have lower rates of tumorrejection, earlier appearance of tumors, and fastertumor growth than control animals exposed to im­planted tumors (e.g .. Shavit et al. 19R4: Visintainer,Volpicelli, and Seligman 1982). Although there areobviously no analogous experimental human stud­ies, several prospective studies have consistentlysupported the hypothesis. A study of Veterans Ad­ministration patients found clear evidence ofgreater repression of negative affectivity on theMMPI among men who subsetjuently developedcancer (Dattore, Shontz, and Coyne 1980), and aprospective community study in Yugoslavia overten years reported similar results (Grossarth­Maticek 1980). A prospective study by Greer, et al.(1985), subsequently replicated, found that help­lessness/hopelessness is also associated with poorprognosis in cancer patients after controlling forobjective predictors.

These epidemiologic studies provide scant in­formation on the mechanisms involved in the ef­fect of personality on onset and course of cancer.According to the most widely endorsed hypothesis,a sense of efficacy and control affects the immunesystem, which, in turn, affeels host resistance tomalignant transformation of cells. This hypothesis

is consistent with both the broader literature onpersonality and immunity (Jemmott and Locke19R4) and an observed association between person­ality and immune competence among cancer pa­tients (Levy et al. 1985). However, it is impossibleto preclude the possibility that the association isdue to an effect of illness severity on personality.Further research is needed to determine whetherbaseline measures of personality among cancer pa­tients predict changes in immune function.

Another area for future research involves in­vestigation of the structural determinants of health­related personality dispositions (Kohn et al. 1983).

In one of the few studies to examine this issue,Harburg, et al. ( 1973) found that suppressed hostil­ity was significantly increased in high-stress resi­dential neighborhoods versus low-stress areas.More research of a similar sort is needed to identifythe structural causes of personality and to specifythe mediating effects of personality on the relation­ships of these structural variables to ill health. Inaddition, research is needed to determine whetherthe effects of personality on health vary dependingon structural contexts. [n one of the rare studies toinvestigate this issue, James et al. ( 1987) found thatan active predisposition to master stress was asso­ciated with increased risk of high blood pressureamong low socioeconomic status African Ameri­cans but not among either whites or higher socioe­conimic status African Americans, presumably re­tlecting the fact that social circumstances make itunlikely that active mastery will effectively reducestress in the face of the environmental barriersthat face lower socioeconomic status AfricanAmericans. More research is needed to investi­gate other interactions between personality dispo­sitions and environmental conditions.

Psychosocial Determinants oflIIness: Overview

The past several decades of research and theorizinghave clearly established the role of psychosocialfactors in the etiology of illness. The major focu,and contribution of this work has been to estahlisha theoretical rationale and empirical evidence for a

number ofhealth risk(2) chroniships andefficacy, 01

(5) high Ie'personal rfpsychologision) is boand itself amortality.

Researneeded. Filand Jongituabout caus.tinuing neethe broad pquential fOIological mFor examplnegative lifhealth. butprecisely wterious anechronic aneof social rearguably 01

smoking, bwhat it is ative of he2produces Sl

Sociolcemphasizedfocus more,psychosoci.by a broadpeople aregender, an,1992: Pearlsearch suggcit fferencessocioeconolciation of 1

risk factorsbehaviors teVerbrugge I

number of psychosocial factors as consequentialhealth risk factors: (I) major negative life events;(2) chronic stress; (3) lack of social relation­ships and supports; (4) lack of sense of control,efficacy, or mastery over one's work and life; and(5) high levels of hostility and/or mistrust in inter­personal relations. This research also shows thatpsychological distress (e.g., anxiety and depres­sion) is both a consequence of these risk factorsand itself a risk factor for physical morbidity andmortality.

Research and theoretical development are stillneeded. First, there is need for more prospectiveand longitudinal research to increase our certaintyabout causal relationships. Second, there is a con­tinuing need for research to specify what aspects ofthe broad psychosocial risk factors are most conse­quential for health, and through what psychophysi­ological mechanisms these effects are produced.For example. we have learned that it is only majornegative life events which are most deleterious forhealth. but we are only beginning to understandprecisely what it is about those events that is dele­terious and through what pathways they affectchronic and infectious disease. We know that lackof social relationships is a risk factor for mortality,arguably of a magnitude comparable to cigarettesmoking, but we are only beginning to understandwhat it is about social relationships that is protec­tive of health and through what mechanisms itproduces such effects.

Sociological social psychologists have recentlyemphasized the need for theory and research tofocus more on the interrelations among these variouspsychosocial risk factors and how they are shapedby a broader social structural context in whichpeople are stratified along lines of race/ethnicity,gender. and socioeconomic status (Aneshensel1992: Pearl in 1989; Williams 1990). Current re­search suggests that many of the persistently largedifferences in health by gender, race/ethnicity, andsocioeconomic status can be explained by the asso­ciation of these variables with the psychosocialrisk factors just considered and the health-relatedbehaviors to which we now tum (House et al. 1992;Verbrugge 1989). In quite a different vein. research

CHAPTER 21 Socia) Psychology and Health 557

on the interplay between psychosocial and geneticfactors in the etiology of health and illness is beingreported just now. We return to these themes at theend of the chapter.

THE SOCIAL PSYCHOLOGY OF HEALTHDEHAVIOR AND ILLNESS DEHAVIOR

The study of health behavior and illness behaviorencompasses how people perceive, define, and acttoward symptoms, how they utilize medical care,how they act to promote health and produce risks,and how they adhere to medical regimens. Interestin health behavior and illness behavior grew out ofa set of practical problems in medicine and publichealth concerning the fact that many people delayseeking medical attention, even in the face of seri­ous and life-threatening symptoms (Leventhal,Meyer, and Nerenz 1980; Rodin 1985), while oth­ers seek medical help for complaints with no dis­cernable organic basis (Mechanic 1992b). Further­more, many patients refuse to do what is seeminglyin their rational self-interest. continuing to smoke,drink, and overeat despite the warnings of physi­cians and health educators (Sackett and Haynes1976). Finally, large numbers of patients, perhapsas many as 50 percent. fail to comply with medicaladvice even when this noncompliance endangerstheir lives (Conrad 1985; Haynes, Taylor, andSackett 1979; Tebbi et al. 1986). These behaviorpatterns result in increased morbidity and mortality(Sackett and Haynes 1976), contribute to escalat­ing medical costs (Fuchs 1974), and frustrate thosewho provide care (Mechanic 1992b).

These observations could not be explained us­ing the traditional biomedical model and led healthresearchers to distinguish analytically between twoorders of phenomena: disease-an organic and bio­logical process; and illness-a psychological, so­cial, and cultural process that includes symptomrecognition, decision making, and utilization. Itwas argued that illness could not be reduced todisease (e.g., Barondess 1979; L. Eisenberg andKleinman 1981). While early work expanded themedical model to include psychosocial factors inpatient behavior, the research questions remained

558 PART III Social Structure. Relationships. and the Individual

largely medical: understanding why patients pro­

crastinate. take risks. or fail to follow medical ad­vice to modify these medically inappropriate be­haviors (Schneider and Conrad 1983). In thisframework. medical judgments about appropriateactions were the "gold standard" against whichactual patient behaviors were judged and foundwanting. From a social psychological standpoint.this medical orientation excluded many importantempirical issues from consideration (Mechanic1978; Zola 1972).

Increasingly aware of these limitations. a num­ber of social psychologists carried out patient­centered analyses of health behavior and illnessbehavior that moved in three broad directions:( I ) away from abstract models of rational choicesabout health care and toward understanding thelogic of lay theories and representations of illnessproblems: (2) away from an exclusive focus on the

decision to seek medical care and toward an em­phasis on patterns of health and illness behaviorthat do not invol ve physicians: and (3) away froma focus on individual characteristics as detenni­nants of health and illness behavior toward under­slanding how the social environment, including thehealthcare system, shapes these behaviors. Severalsocial psychological perspectives contributed tothis work. After brietly reviewing these perspec­tives, we examine how they are retlected in empiri­cal research on health behavior. illness behavior,

and adherence to medical regimens.

Theoretical Perspectives

Culture. Social Structure, and Patient Behavior.Social psychologists with a cultural or social struc­tural orientation have found that social groups dif­fer in their responses to symptoms and patterns ofcare seeking and have attributed these differencesto cultural orientations or structural constraints.Researchers have examined ethnic differences inthe response to pain and symptoms as wcll as eth­nic, socioeconomic. and gender differences in utili­zation patterns. Early research in this area oftenused purely correlational, cross-se<.:tional designsthat postulated reasons (often <.:ultural) for groupdifferences. rather than demonstrating them em-

pirically (Cockerham 1986: Mechanic 1978). Morerecent work has moved toward explanatory ap­proaches, emphasizing that the structure of socialand healthcare institutions. as well as the culture ofcare seekers. shape health and illness behavior.

Integrative Models, Several e<.:lectic approachesdepict illness behavior as resulting from an inter­play of biologi<.:aL sociocultural. and psychologicalfactors. The best-known example of this approachis Mechanic's theory of help seeking. Basing hismodel on a large body of theory and research.Me<.:hanic identifies ten cognitive. social, and psy­chological factors that influence the decision toseek help: ( I) the visibility and salience of symp­toms: (2) the extent to which symptoms are per­ceived as serious: (3) the extent to which they dis­rupt sodal activities; (4) the frequency andpersistence of symptoms; (5) the tolerance level of

those who experien<.:e symptoms: (6) available in­formation, knowledge, and cultural assumptions:(7) basic needs that lead to denial: (8) other needsthat compete with illness responses; (9) wmpetinginterpretations that can be assigned to symptoms:and ( 10) the available resources, physical proxim­ity of care, and l:Osts of taking action (Mechani<.:197X). The strength of this approach is its effort to

develop a comprehensive, biopsychosocial ap­proach to care seeking. However, as Mechani<.:himself acknowledges, the model is physician­centered, focusing on the decision to seek conven­tional medical <.:are (S<.:hneider and Conrad 1983),though it could be broadened to include othersources of care.

Cognitive Models of Decision Making. Socialpsychologists have examined the cognitive proc­esses at work in decisions to take preventive action.to seek help, and to follow medical advice. Untilre<.:ently, most psychologists employed one of anumber of rational choice models that assume thaIpeople make health behavior choices on the basisof cost-benefit ratios. Perhaps the best-known ra­tional <.:hoice theory is the health belief model.According to this model. people decide to takepreventive action or follow medical advice on thebasis of their subjective beliefs about the severity

of the i1lnling ill. andaction (e.g1958: Ros.

This r\ufficienttradition. ,to considelions on in\implify t~

ing rationa

'tandards c1985: Tverics argue t~

theories, prhow decisil

quate desC'made.

More remotion in(1977). for,

making occfonnation pthey are aw.have hope,they have ercause of theCIS Ions, peollive copingddherence tl\ive avoidan([Janic).

Other hthe role of c,ymptoms aPeople who (,If them by n,entations st,lIons about t~

l'llurse, and plIl,m, and GI,ymptom proprocessor ofliCOl'S perspe1,lins a medic:Illfomlation pII'y common-

Ill' the illness. their susceptihility or risk of hecom­ing ill. and the costs. benefits. and harriers to taking.lction (e.g .. Becker and Maiman 1975; Hochhaum195~; Rosenstock and Kirscht 1979).

This model has heen criticized for giving in­,ufficient attention to the role of cultural values.tradition. and emotion in decision making. failingto consider cognitive and organizational limita­tions on information processing that lead people to,implify the decision-making process. and eqllat­II1g rational choices with those that conform 10 the,tandards of Western medicine (Garro I<.JX5; GoodI\)X5; Tversky and Kahneman 1974). In short. crit­ICS argue that this model. like other rational choicetheories. provides a prescriptive. idealized view ofhow decisions should he made rather than an ade­quate description of how decisions actually aremade.

More recent approaches consider the role ofemotion in health decisions. Janis and Mann(1\)77). for example. propose that optimal decisionmaking occurs when people engage in "vigilant in­formation processing'" which can occur only whenthey arc aware of the risks allached to each choice.have hope of finding an alternative. and helievethey have enough time to del iherate. However. he­cause of the anxiety-provoking nature of health de­c'isions. people often resort to one of several defec­tive coping strategies. including unconflictedadh('rence to their present course of action. defcn­,ive avoidance (procrastination). or hyperv igilance[panic).

Other health psychologists have focused on[he role of cognitive schemas in making sense of,ymptoms and deciding what to do ahout them.People who experience bodily changes make sense\1f them by means of common-sense illness repre­,cntations stored in memory and include attrihu­tlons ahout the identity. causes. consequences. timeC:l)urse. and potential for cure (Leventhal. Zimmer­man. and Gutmann 19X-l). Cognitive models of,ymptom processing depict the patient as an activeprocessor of information and emphasize the pa­tiL'nt's perspective. However. this approach main­[~\ins a medical orientation by emphasizing flaws inInformation processing. stressing: the need to mod­Ifv common-sense representations to bring them

CHAPTER 2t Social p,yL'iwlnilY anu Hcalth 559

into line with medical ones. and neglecting envi­ronmental and structural factors .

Phenomenological Analyses ofDecision Making.Phenomenological approaches. developed in an­thropology and social psychology. provide themost patient-centered perspectives on health he­havior and illness hehavior. These perspectivesconsider patterns of care outside the professionalsector and attempt to discover the logic of patientdecisions rather than analyzing them in terms ofmodels the researcher formulates a priori. Thesemodels include everyday ideas ahout the etiology.anticipated course. and consequences of a partiClI­Jar illness. Explanatory model research has goneheyond cognitive approaches to explore how mod­els of doctors and patients collide in medicalencounters. how explanatory models relate tohroader cullUralthemes. and how they assume dif­krent forms in different cultural contexts (Klein­man I<.JXO).

Other cognitive anthropologists have devel­oped formal models of decision making in whichthe researcher el icits from patients their actual con­siderations in making medical decisions. developsa formal model of the criteria used in health deci­sions. and tests the model's validity hy comparingit to the aClllal decisions of community memhers(Garro IYX5). Sociologists have developed a phe­nomenological approach to the experience of ill­ness that examines how patients notice somethingis wrong. develop lay thenj'ies and explanations.decide to seek help. manage relationships withsigni ficant others and health professionals. andcope with the stigma attached to their illness. Re­searchers emphasize the importance or studyingnonhospitalized patients and examining self-medi­cation practices that do not involve professionals(Schneider and Conrad I<.JX3).

Empirical Applications

Health Behavior. Health behavior refers to theactions of well people that have consequences fortheir future health. such as smoking. diet. exercise.and substance ahuse (Mechanic !Yl)O). A numherof programs have heen developed to change health

560 PART III Soci~1 SlrUClllr~" R~I~lion,hips. ~llLllhe Indl\ldll,li

behaviors. Most are based on psychological mod­els of hehavior change and. with a few notableexceptions. have generally not succeeded in effect­ing lasting changes in health behavior (Mechanic1990. 1992a). In fact. a review of the long-termoutcomes of many programs shows that as many asthree of every four people who successfully changerisk behaviors arc unable to sustain these changesfor as long as one year (Brownell et al. 1(86).

Critics have attributed these failures to the in­accurate assumptions of some psychological mod­els of health behaviors and have proposed newideas about ways to modify interventions. Findingsconcerning the importance of cognitive repre­sentations. for example. have led to the sugges­tion that future interventions identify and alterparticipants' representations of risk. provide themwith self-regulation skills. and encourage them toperceive that they can effect change (Leventhal.Zimmerman. and Gutmann 1(84). Other healthpsycilOlogbts have called for programs that en­hance perceptions of control. This recommenda­tion is hased on research showing that participantsin weight reduction programs are more likely tosucceed when they attribute change to their ownetlorts (Rodin 11)85). Conversely. participants areless likely to relapse pennanently when they attrib­ute lapses to situational causes (Marlatt and Gor­don 1(85).

Some sociological social psychologists havealso suggested that traditional interventions fail be­cause they treat health behavior as an individualrather than a social phenomenon (Mechanic [990;Syme and Alcalay 1(82). In particular. health be­haviors may have other meanings that interfereWith behavior change (Mechanic 1(90). Smoking,for example. can he a mark of status and a symbolof defiance in adolescent culture (lessor, Donovan.and Costa )990: Osgood et al. 1988: Rodin 1985).Successful interventions to modify health behav­iors need to consider such symbolic meanings. Thiscan be done by providing the individual with re­sources to resist interpersonal pressures. An exam­ple is Michelson's (1986) Social Skills TrainingProgram. a wide-ranging program that teachesadolescents to evaluate health behavior options

and to resist interpersonal pressures to engage inrisk hehaviors. Alternately. one may attempt tochange the cultural meanings of health behaviors.One of the most successful examples of this ap­proach is the antismoking movement in the UnitedStates. a movement that changed the symbolicvalue of smoking in the middle class (Mechanic!1)90: Syme and Alcalay 1982) and resulted in asubstantial long-term reduction in smoking (War­ner 1(77).

The importance of facilitating social and struc­tural conditions is not limited to symbolic mean­ings. It is also important to consider the functionsof health behaviors in the lives of the people whosebehaviors we seek to change. Alcohol and tobacco,for example. appear to be used as coping resourcesby many people. This raises the question ofwhether interventions to change the structural con­ditions that lead to chronic stress make more sensethan interventions aimed at removing the copingresources used by people to manage chronic stress.FurthernlOre. if interventions do attempt to removethese coping resources there is a need to providealternate resources. See House and Cottington(1986) and Williams (1990) for a discussion. Asimilar question can be raised about the logic ofattempting to change individual health behaviorwhen powerful economic interests continue to pro­mote risk taking (McKinlay 1990: Symeand Alca­lay 1(82). It is noteworthy in this regard that statelicensing boards permit more retail outlets for thesale of alcohol in poor and African Americanneighborhoods than in more affluent areas (Rabowand Watts 1(82). Furthennore, more than 70 per­cent of billboards in the United States that adver­tise tobacco and alcohol are targeted to AfricanAmericans (Hacker. Collins. and Jacobson 1(87).It is difficult to avoid the conclusion. based onthese results. that there are systematic structuralforces at work that impede individual efforts toreduce the problems of substance use among disad­vantaged sectors of American society. Based onthis conclusion. there is a growing belief that struc­tural change is needed to guarantee the success ofwidespread health behavior change (McKinlay1990 ).

",

J..

i:

Illness Bways peotions andsymptom:and in res(Mechaniissues intenninantthat infl Ul

recognizeinfluenceations inmatic emwith illne

Rese,are substations to mand sociacreating tl1986). Foethnic groto symptcand TursIAn especgenerallysymptom~

nize and1986). Ththe lowerwomen arbodily Cl

comes froradiograplstudy selfvisits forwomen wwere nearreport reCI

Reseathe ways cof bodilytenns. Re:arrive at ato illness (neutralize.symptoms

!r

II

I~

Illness Behavior. Illness behavior refers to theways people define and respond to bodily sensa­tions and experiences that might be seen as signs orsymptoms of illness, both before seeking treatmentand in response to the recommendations of healers(Mechanic 1986). Important social psychologicalissues in the study of illness behavior include de­terminants of initial symptom recognition, factorsthat influence how symptoms are interpreted oncerecognized, and social and individual variables thatinfluence willingness to adopt the sick role. Vari­ations in these ways of responding can have dra­matic effects on the social impairment associatedwith illness.

Research on illness behavior shows that thereare substantial individual differences in predisposi­tions to monitor bodily sensations and that culturaland social experiences play an important part increating these dispositions (Hansell and Mechanic1986). For example, early research discovered thatethnic groups differ dramatically in their responsesto symptoms and perceptions of pain (Sternbachand Tursky 1965; Zborowski 1952; Zola 1966).An especially intriguing result is that womengenerally seem more sensitive than men to bodilysymptoms and are therefore more likely to recog­nize and seek help for health problems (Kessler1986). This sex difference is more pronounced atthe lower end of symptom severity, suggesting thatwomen are more likely than men to monitor subtlebodily complaints. An interesting illustrationcomes from the work of Davis (1981), who usedradiographic examination data on osteoarthritis tostudy self-reported knee pain and recent doctorvisits for knee pain. These data documented thatwomen with objective evidence of osteoarthritiswere nearly twice as likely as comparable men toreport recent doctor visits for this problem.

Research on symptom sensitivity has exploredthe ways cognitive schemas are used to make senseof bodily sensations and define them in illnessterms. Research also has shown that most peoplearrive at a definition of their bodily changes as dueto illness only after concerted efforts to normalize,neutralize, or minimize the significance of theirsymptoms, a practice that may account for com-

CHAPTER 21 Social Psychology and Health 561

mon delays in seeking help (e.g., Davis 1971;Mechanic 1972; Schneider and Conrad 1983).These interpretations often involve the use of so­cial networks. Parents, spouses, friends, and evenphysicians often collaborate in the normalizationprocess (Davis 1971; Schneider and Conrad 1983).

The literature on illness representations docu­ments many dramatic cases that illustrate thisprocess. For example, people who are alone whenthey first experience a mild heart attack commonlydelay calling an ambulance due to uncertaintyabout what has happened to them. Instead, beforecalling an ambulance they will call a friend or rela­tive, describe the symptoms, and ask whether theother person thinks it was really a heart attack.This delay is associated with a dramatically in­creased risk of long-term cardiac damage (Alonzo1986).

Illness behavior research also has examinedhow groups differ in utilization patterns. Begin­ning with Koos's (1954) early work on socioeco­nomic differences in help seeking, research hasdocumented that the poor use health services lessfrequently than those more favorably situated inthe social structure. Although expanded public in­surance coverage has dramatically decreased so­cioeconomic differences in overall utilization, con­tinuing differences in utilization patterns suggest atwo-class system of health care. While higher so­cioeconomic groups use private physicians, thepoor use a public healthcare system of outpatientclinics and hospital emergency rooms (Cockerham1986). Early cultural explanations attributed thesedi fferences to a greater tendency of poor patients tonormalize or neutralize symptoms (Koos 1954) or,alternatively, to a present-oriented culture of pov­erty that led poor patients to eschew prevention anddelay seeking help until emergencies arose (e.g.,Kosa, Antonovsky, and Zola 1969). However,more recent "systems" or "culture of medicine"explanations suggest that the highly alienating,impersonal, bureaucratic atmosphere and low qual­ity of care in the public healthcare system leadspoor patients to view medical care as a measureof the last resort (e.g., Dutton 1978; Reissman)981).

562 PART 1lI Social Structure, Relationships, and the Individual

Current research has moved away from an ex­clusive focus on the decision to seek medical careand toward a focus on care that extends beyondconventional medicine (e.g., Garro 1985). Whileearly studies viewed the use of alternative practi­tioners as a deviant pattern of utilization confinedto lower socioeconomic classes or ethnic enclavesusing parochial referral networks, it has becomeclear based on more recent studies that many pa­tients with chronic medical problems are turning toalternative practitioners, such as chiropractors andacupuncturists, who offer hope of symptomatic re­lief and more personal attention (D. Eisenberg et al.1993; Kotarba 1983).

Although most research on illness behaviorcontinues to study patient responses from the view­point of the medical establishment, a growing num­ber of studies are questioning professional defini­tions of illness and giving more attention to thepatient viewpoint (e.g., Roth and Conrad 1987).Whereas earlier studies invidiously contrasted layexplanations with professional ones, several con­temporary researchers portray both lay and medicaldecisions as socially constructed. These analysessuggest that professional diagnosis and treatmentdecisions are influenced by the cultural assump­tions of providers, the perceived characteristics ofpatients, and the social setting in which decisionsare made (Todd 1989).

Another current trend in illness behavior re­search is to focus on patient self-care. This researchhas shown that many people diagnose and treattheir own symptoms (Zola 1983). Furthermore, pa­tients who are receiving medical treatment forchronic conditions commonly search for patterns intheir symptoms, make note of antecedents to flare­ups, develop and test hypotheses, and sometimeseven devise strategies thought to control symptomexpression (e.g., Schneider and Conrad 1983).

Treatment Adherence. Patient failure to followmedical advice is a widespread phenomenon thatreduces the effectiveness of therapy (Rodin andSalovey 1989) and may significantly increase mor­bidity and mortality (Sackell and Haynes 1976).Buckalew and Sallis (1986) estimated that roughly

one-third of the 750 million new prescriptions Willten each year in the United States and the UlIil~'d

Kingdom are not taken at all and another one-thirdare taken incorrectly. Nonadherence for medicItion and lifestyle changes recommended to trcalchronic conditions is estimated at roughly 50 percent (Haynes, Taylor, and Sackett 1979). Surpri,ingly, these high rates exist even for those paticllhwith life-threatening and seriously disabling condltions. For example, Tebbi et al. (1986) found lhalonly 50 percent of adolescent cancer patients tooktheir prescription medications as directed, whlkConrad (1985) found that only 50 percent of Cpl

leptics took their medications correctly.Determinants of adherence include a wide va­

riety of individual and environmental factors (Dr·Malleo and DiNicola 1982). As in the case of ill­ness behavior, social and symbolic meanings 01medications figure importantly in adherence. Patients view medications alternatively as an indicator of the degree of their disorder, a ticket to normality that can increase self-reliance, a symbol 01

dependence, or a reminder of deviance and stigma.Nonadherence is powerfully affected by thc~~'

meanings. For example, some epileptic paticlII,stop taking their medication against medical advin~when they view drugs as a symbol of their dependence and wish to reassert their independence orwhen they view drugs as a reminder of differcntness and want to escape the stigmatizing connola­tions (Schneider and Conrad 1983).

Illness representations also have been shownto play a prominent role in adherence. Researchconsistently shows that there are major discrepallcies between the illness representations of patienhand of healthcare providers and that these diffcr­ences play an important part in adherence. LevclIthaI et al. (1984), for example, proposed that mall}patients with chronic, asymptomatic illnesses farlto follow physician advice because they use all"acute disease schema" to interpret their medicalproblems. That is, they believe their disease to tx·caused by external agents that are short-terlll,symptomatic, and treatable by medications that fl'

move the symptoms and cure the disease. WhclItheir experiences in treatment clash with these pcr

ceptions. thic'nt with th(iutmann (I,ive patient~

pressure beiaches and f.uf the patie,panse to tIllcnts defie,orne to stolhigh blood~tn acute discIlcnts with ;drop out of'hese. there idistribution (and investig,ense represt1986).

The litt:shows clearlelicit informi!lness reprebetween the~

recommendatient's theoricate in theseadherence (\experimentallive show th.,icians chanpatients. Inudocumented 1

lion style wita single two­ccnt increaseamong patienlphysicians. Snumber of ott\cntions (Me

Research,hows that atIllunication st()f informatioquestions-inI969: Freem(

"l'ptions. they often terminate treatment. Consis­Il'nt with this perspective. Meyer. Leventhal. andl Jutmann ( 1985) found that 90 percent of hyperten­,ive patients believed they could "feel" their bloodpressure being elevated by such symptoms as head­al'hes and face tlushing, This perception led someDf the patients to adjust their medications in re­sponse to these feelings, even when the adjust­Illents defied physician instructions, It also led,orne to stop taking medications when feelings ofIligh blood pressure persisted. The persistence of~m acute disease schema may also explain why pa­tients with asymptomatic, chronic illnesses oftendrop out of treatment. Based on findings such asthese. there is much current interest in studying thedistrihution of illness representations (Bishop 1987)~md investigating ways to modify these common­sense representations to promote adherence (Cleary1986).

The literature on doctor-patient interactionsshows dearly that it is critical for the doctor todicit information about patient expectations andillness representations, to confront discrepancieshetween these cognitions. and to explain treatmentrecommendations in a way congruent with the pa­tient's theories of illness. Doctors who communi­cate in these ways have consistently higher rates ofadherence (Whitcher-Alagna 191'3). Furthermore.~xperimental interventions based on this perspec­tive show that adherence can he improved by phy­sicians changing their style of interacting withpatients. Inui. Yourtee, and Williamson (1976)documented that changes in physician communica­tion style with hypertensive patients resulting froma single two-hour training session led to a 30 per­cent increase in effective blood pressure controlamong patients experimentally assigned to the trainedphysicians. Similar results have been reported for anumber of other cognitively-based adherence inter­ventions (Meichenbaum and Turk 1987).

Research on doctor-patient interaction alsoshows that other characteristics of physician com­munication style--quantity of information. qualityof information. and willingness to let patients askLjuestions-importantly affect adherence (Davis1969: Freemon et al. 1971: Roter and Hall 191'9:

CHAPTER 21 Social Psy<'ho[ogy and Heallh 563

Svarsted 1976), Many studies demonstrate thatphysicians frequently fall short on all of these di­mensions. There is currently a good deal of interestin the structural determinants of these aspects ofphysician communication behavior aimed at un­derstanding why many doctors interact in ways thatproduce the very nonadherence they find so trou­blesome. Some researchers argue that the logic ofdifferential diagnosis. coupled with the demandsassociated with rapidly processing information inbureaucratic organizations. drastically limits whatdoctors can accomplish in the medical interview(Cicourel 191'1), Others have suggested that ashealth care comes to be delivered increasingly in acolleague-dependent context of referral. physiciansbecome oriented to the wishes of colleagues ratherthan the wishes of their patients. Subsequently.effective doctor-patient communication is im­paired (Freidson 1970). A related observation isthat providers' cultural assumptions can makethem less inclined to encourage active patient par­ticipation. For example. Anspach (1993) observesthat many providers have a "common-sense socialpsychology" that underestimates patients' compe­tence to participate in medical decisions and over­estimates the likelihood of deleterious psychologi­cal consequences that can result from participatingin medical decisions. Providers' cultural assump­tions also include social schemata that lead profes­sionals to underestimate the ability of lower-classHispanic patients to participate in certain medicaldecisions (Anspach 1993). In short, physicians cur­tail patient participation. thereby discouraging ad­herence. less because of conscious choice than be­cause of organizational restraints and misguidedcultural assumptions.

The Social Psychology of Health Behavior andIllness Behavior: Overview

A growing body of research shows that psychoso­cial factors are crucial to understanding and modi­fying health and illness behavior. Researchers dif­fer as to whether they examine psychosocialprocesses in an effort to change patient behavior inways that will improve health outcomes or take a

564 PART III Social Structure, Relationships, and the Individual

··1··'····!.·..··•

...

patient-centered approach that examines a broaderset of issues concerning lay representations ofhealth problems and other meanings of health andillness behaviors, Researchers also focus variablyon processes involved in the creation of health andillness behaviors or on the impact of macrosocialcontexts on these behaviors, Unlike most otherareas of social psychological research, there aremany opportunities here for developing and imple­menting large-scale interventions that can be usefulboth in applying social psychological knowledgeand advancing the knowledge base, A challenge forthe future is to integrate research on the structuraldeterminants of health and illness behaviors withresearch on the mechanisms linking meaning struc­tures to behavioral responses and to devise meth­ods of intervening at a more structural level thanwe have at present.

RETROSPECT AND PROSPECT

Theory and research on social psychology andhealth have developed remarkably over the past

NOTE

Work on this chapter was partially supported by grantsK02 MH00507 from the National Institute of MentalHealth and PO I AG0556I and R29 AG07904 from theNational Institute on Aging. The authors are indebted to

REFERENCES

Alexander, Franz. 1950. Psychosomatic Medicine; ItsPrinciples and Applications. New York: Norton.

Alonzo, Angelo A. 1986. The impact of the family andlay others on care-seeking during life-threateningepisodes of suspected coronary artery disease. So­cial Science and Medicine 22: 1297-1311.

Anderson, Karen O. [985. Rheumatoid arthritis: Reviewof psychological factors related to etiology effectsand treatment. Ps."c1wlogical Bulletin 98:358-387.

Aneshensel. Carol. 1992. Social stress: Theory and re­search. Annual Rel'iew ofSociologv 18: 15-38.

Anspach. Renee R. 1993. Deciding Who Lires; FatefitlChoi,'es in the Intensh'e-Care Nurser\'. Berkeley,CA: University of California Press.

few decades, contributing to a major reorienta­tion in the way the scientific community antisociety think about the nature, causes, and courseof illness. The social environment and its socialpsychological consequences have become recog­nized as central to understanding and improvinghealth.

Much of this work has focused on individual­level variables. Much remains to be clarified antilearned at this level of analysis. Greater emphasison macrosocial factors is already apparent in newIines of research on the macrosocial determi­nants of health and illness behaviors (Anehensel1992; Mirowsky and Ross 1989; Pearlin 1989;Williams 1990). Such new work is beginning todocument systematic structural patterns of a broadarray of psychosocial determinants of health andhealth-related behaviors and provides importantavenues for intervention (House et al. 1992). Noproblem is more central to our well-being thanhealth, and in no area does social psychology havegreater opportunities for advancement over thecoming years.

Karen Cook. Gary Fine. Catherine Ross. and especiallyDavid Mechanic for their constructive comments on pre­vious drafts.

Back, Kurt W., and Morton D. Bodgonoff. 1967. Bufferconditions in experimental stress. Behavioral Sci­ence 12:384-390.

Barondess, Jeremiah A. 1979. Disease and illness: Acrucial distinction. American Journal of Medicine66:375-376.

Becker, Marshall H.: and Lois A. Maiman. [975. So­ciobehavioral determinants of compliance withhealth and medical care recommendations. MedicalCare 13:10-24.

Berkman, Lisa F. [985. The relationship of social net­works and social support to morbidity and mortality.pp. 241-262 in Social Support all£! Health, ed. Shel­don Cohen and Leonard Syme. New York: Academi<:.

.,••.' .

" .. "

~ ."~

I""

Berkman,netw(

year tArneI',

Bishop, Gsymp127­

Blazer, DaelderlofEpl

Brown, Ge:cial CDisor

Brownell,stein,and pi765-1

Buckalew,ance,cal P.I

Cannon, \\

York:Caplan, Ge

MentaCassel. Jot

host n104:H

Cicoure[, I

microAdmn~\'ard ,gies. eRoute

Cleary, Palresear,

tidisci/Plenul

Cobb, Sidrlife Sll

Cockerhamed. En

Cohen, Sheality cship bAdmnStress.

Cohen, She:Stepheaffiliatman P'

Berkman, Lisa F., and S. Leonard Syme. 1979. Socialnetworks, host resistance. and mortality: A nine­year follow-up study of Alameda county residents.American Journal of Epidemiology 109: 186-204.

Bishop, George D. 1987. Lay conceptions of physicalsymptoms. Journal of Applied Psychologv 17:127-146.

Blazer. Dan G. 1982. Social support and mortality in anelderly community population. Amerimll Jourl/alof Epidemiology 115:685-694.

Brown, George W., and Tirril O. Harris. ed. 1978. So­cial Origins of Depression: A StudY of" PsychiatricDisorder in Women. New York: Free Press.

Brownell, Kelly D., G. Alan Marlatt, Edward Lichten­stein, and G. Terence Wilson. 1986. Understandingand preventing relapse. American PHchologist 41:765-782.

Buckalew, L. W., and R. E. Sallis. 1986. Patient compli­ance with medication perception. Journal of Clini­cal Psychology 42: 161-167.

Cannon, Walter B. 1932. The Wisdom of the Body. NewYork: Norton.

Caplan, Gerald, ed. 1974. Support Systems in CommunityMental Health. New York: Behavioral Publications.

Cassel, John. 1976. Contribution of the environment tohost resistance. American Journal of Epidemiology104: 107-223.

Cicourel, Aaron V. 1981. Notes on the integration ofmicro and macro levels of analysis. Pp. 1-40 inAdvances in Social Theory and Methodology: To­ward an Integration of Macro- and Micro-Sociolo­gies. ed. K. Knorr-Cetina and A. Cicourel. London:Routeledge & Kegan Paul.

Cleary, Paul D. 1986. New directions in illness behaviorresearch. pp. 343-354 in Illness Behavior: A Mul­tidisciplinary Model. ed. S. McHugh. New York:Plenum.

Cobb. Sidney. 1976. Social support as a moderator oflife stress. Psychosomatic Medicine 38:300-314.

Cockerham, William C. 1986. Medical Sociology. 3rd.ed. Englewood Cliffs, NJ: Prentice Hall.

Cohen, Sheldon, and Jeffrey R. Edwards. 1989. Person­ality characteristics as moderators of the relation­ship between stress and disorder. Pp. 235-283 inAdvances in the Investigatioll of PsychologicalStress. ed. R. W. J. Neufeld. New York: Wiley.

Cohen. Sheldon, Jay R. Kaplan, Joan E. Cunnick, andStephen B. Manuck. 1992. Chronic social stress,affiliation and cellular immune response in nonhu­man primates. Psychological Science 3:301-304.

CHAPTER 21 Social Psychology and Heahh 565

Cohen. Sheldon. Drury R. Sherrod, and Margaret S.Clark. 1986. Social skills and the stress-protectiverole of social support. Jourl/al of Persollality andSocial Psychology 50:963-973.

Cohen. Sheldon, D. A. J. Tyrell, and A. P. Smith. 1991.Psychological stress and susceptibility to the com­mon cold. New England Journal of" Medicine 325:606-612.

Cohen, Sheldon, and Gail M. Williamson. 1991. Stressand infectious disease in humans. PsychologicalBulletin 109:5-24.

Cohen. Sheldon, and Thomas A. Wills. 1985. Stress.social support, and the buffering hypothesis. Psy­chological Bulletin 98:310-357.

Conrad. Peter. 1985. The meaning of medications: An­other look at compliance. Social Science Medicine20:29-37.

Coyne, James, Camille B. Wortman. and Darrin Leh­man. 1988. The other side of support: Emotionalover-involvement and miscarried helping. Pp. 305­330 in Marshalling Social Support, ed. B. H. Got­tlieb. Beverly Hills: Sage.

Dattore, Patrick 1.. Franklin C. Shontz, and LolafayeCoyne. 1980. Premorbid personality differentiationof cancer and noncancer groups: A test of the hy­pothesis of cancer proneness. Journal of Consult­ing Clinical Psychology 48:388-394.

Davis. Milton. 1969. Variations in patients' compliancewith doctors' advice: An empirical analysis of pat­terns of communication. American Journal ofPub­lic Health 58:274-288.

---. 1971. Variation in patients' compliance withdoctors' advice: Medical practice and doctor-pa­tient interaction. Psychiatry and Medicine 2:31-54.

---. 1981. Sex differences in reporting osteoarth­ritic symptoms: A sociomedical approach. Jour­nal of Health and Social Behavior 22:298-309.

DiMatteo, M. Robin, and D. Dante DiNicola, ed. 1982.Achieving Patiellts Compliance. Elmsford, NY:Pergamon.

Dubos. Rene. 1959. Mirage of Health. New York: Har­per & Row.

Dutton, D. B. 1978. Explaining the use of health serv­ices by the poor: Costs, attitudes, or delivery sys­tems'? American Sociological Review 43:348-368.

Eisenberg, David M., Ronald C. Kessler, Cindy Foster,Frances E. Norlock, David R. Calkins, and ThomasL. Delbanco. 1993. Unconventional medicine in theUnited States: Prevalence, costs, and pattems of use.Nell' Englalld JO/lrnall!f"Medicille 328:246-252.

566 PART 1II Social Structure. Relationships. and the Individual

Eisenberg, Leon. and Arthur Kleinman. 1981. The Rele­

ral/ce of Social Sciel/ce .tilr Medicil/e. Boston:

Reidel.Freemon. B., Vida Negrete, Milton Davis, and Barbara

Korsch. 1971. Gaps in doctor-patient communica­tion. Pediatric Research 5:298-311.

Freidson, Eliol. 1970. Professioll oj' Medicille, NewYork: Dodd-Mead.

Friedman, Howard S.. ed. 1990. Persollalitl' allli Dis­elise. New York: Wiley.

Friedman, Howard S.. and Stephanie Booth-Kewley.1987. The disease-prone personality: A meta-ana­lytic view of the construcl. Americall P.ITclwlogist~2:539-555.

Friedman, Meyer, and Ray H. Rosenman. 1974. Type ABehal'ior alld Your Heart. New York: Knopf.

Fuchs. Victor R. 1974. Who Shall Live ~ Health. Eco­lIomics. alld Social Choice. New York: Basic

Books.Garro. Linda. 1985. Decision-making models of treat­

ment choice. Pp. 173-1 XX in Illness Belllll'ior: AMultidisciplillary Model. ed. S. McHugh and T. M.Vallis. New York: Plenum.

Good. Byron 1. 19X5. Explanatory models and care­seeking: A critical account. Pp. 161-173 in 111111'.1'.1'

Behl/l'ior: A Multidisciplinary Model. ed. S.McHugh and T. M. Vallis. New York: Plenum.

Greer, Steven. Keith Pettingale, Tina Morris, and 1.Haybittle. 1985. Mental attitudes to cancer: An ad­ditional prognostic factor. Lallcet 3:750.

Grossarth-Maticek. Ronald. 1980. Social psychotherapyand course of the disease: First experiences withcancer patients. Psvchorherapy and Psyclwsol1lllf­

iI'S 33: 129-138.Hacker, George A., Ronald Collins, and Michaellacob­

son. 1987. Marketillg Boo:e to Blacks. Washing­ton. DC: Center for Science in the Public Inter­

esl.Hansell. Steven. and David Mechanic. 1986. The so­

cialization of introspection and illness behavior.Pp. 253-260 in Illness Behal'ior: A Multidiscipli­lIaTl' Model. ed. S. McHugh. New York: Plenum.

Harburg, Ernest. lohn C. Erfurt. L. Havenstein. C.Chape, W. Schull, and M. Anthony Schork. 1973.Socioecological stress. suppressed hostility, skincolor. and black-white male blood pressure. Psy­clwsomatic Medicille 35:276--296.

Haynes, R. Brian. D. W. Taylor. and David L. Sackett,ed. 1979. Compliallce ill Hl'lIltli Carl'. Baltimore:lohns Hopkins University Press.

Heller. Kenneth, and Ralph V. Swindle. 19X~, ,>".

networks. perceived social support, and l "1'11with stress. Pp. 87-103 in Prevelltive PI'vcil"/,, ..

Research and Practice ill Comlllullity IlIte1"1'1'1I11' ./.

ed. R. D. Feiner, L. A. lason, 1. Moritslll-!I1,'"'1S. S. Farber. New York: Pergamon.

Hirsch. Barton 1. 1979. Psychological dimensl"ns "I

social networks: A multi method analysis. ,\"" /I

call Journal of COIllIllUllit\, P.n·dwlogy 7:26,~ .'"

Hochbaum. G. 1958. Public participation in 111""1' ,01

screening programs: A sociopsychological sln"lDHEW publ. no. [PHS] 572. Washington. 1)(

U.S. Government Printing Office.

Holmes, Thomas H.. and Richard H. Rahe. 1967. III,'social readjustment rating scale. Joun/al ofP.II1 il",\,(II/llitic Re.l'earch II :213-218.

House. lames S. 19X I. Work Stress lIlId Social SUI'I'II/ I

Reading. MA: Addison-Wesley.

House. lames S., and Eric M. Cottington. 1986. I kalrlland the workplace. Pp. 382-415 in Applicatio/ls IIISocilll Sciellce to Clillical Medicine lind HI'alrl,Polin', ed. L. H. Aiken and David Mechanic. N",\;

Brunswick. Nl: Rutgers University Press.

House, lames S.. Ronald C. Kessler. A. Regula Herwl-!.Richard P. Mero. Ann M. Kinney. and Martha JBreslow. 1992. Social stratification. age, ;11111

health. Pp. 1-32 in Aging. Health Behaviors, II/ltl

Health Owcollles, ed. K. Warner Schaie, DanBlazer. and lames S. House. Hillsdale, Nl: J-:r I

baum.House. lames S., Karl R. Landis. and Debra Umberson,

1988. Social relationships and health. Science 2~ I'5~0-545.

House. lames S., Cynthia Robbins. and Helen I ..Metzner. 1982. The association of social relation­ships and activities with mortality: Prospective evi­dence from the Tecumseh Community HealthStudy. American Journal of Epidellliology 116:123-140.

House, lames S.. Victor 1. Strecher. Helen L. Metzner.and Cynthia A. Robbins. 1986, Occupational stressand health among men and women in the TecumsehCommunity Health Study. Journal of Health andSocial Behavior 27:62-77.

Inui, Thomas S.. E. L. Yourtee, and 1. W. Williamson.1976. Improved outcomes in hypertension afterphysician tutoriais: A controlled trial. Anllals ofIllternal Medicine 84:646--651,

lackson, G. G .. H. F. Dowling. T. O. Anderson, L. Riff,1. Saporta, and M. Turck. 1960. Susceptibility and

immunit:[ions: Th

cille 53:iI:lIlles, Sherm

Ramsey.ism. andC((II Jour

J<mis. Irwin L

A IISI'£'hlcO/lllI/itill

krnmott. lohchosociahuman s

much dc78-108.

lessor. RichaCosta, I'and ado

Healt" Hand F. L(

Karasek. RobtWork. Ne

Kasl. Stanis/a'

to the sllWork, ellWiley.

Kessler. Ronahealth seJMllilidis£'Thomas I

Kessler. Rona

support aPp.219­Cohen an

Kessler. RomHouse. I'ship bet...logical M

Kessler. Ronal

psycholol86 in HmH. E. FreNl: Prent

Kleinman. AnCOlltext I.

fornia PreKohn, Melvin

A. Miller1983. W..III/pact of

immunity to common upper respiratory viral infec­tions: The common cold. Annals of Intel"l1al Medi­cine 53:719-738.

James, Shennan A., David S. Strogatz, S. Wing, and D.Ramsey. 1987. Socioeconomic status. John Henry­ism, and hypertension in blacks and whites. Ameri­can Journal of Epidemiolog.l' 126:664-673.

Janis, Irwin L.. and Leon Mann. 1977. Decision making:A psychological analysis of' conflict, choice. andcommitment. New York: Free Press.

Jemmott, John B., and Steven E. Locke. 1984. Psy­chosocial factors, immunologic mediation andhuman susceptibility to infectious disease: Howmuch do we know'! P.Hclwlogical Bulletin 95:78-108.

Jessor. Richard. John E. Donovan, and Frances M.Costa. 1990. Personality, perceived life chances.and adolescent health behavior. Pp. 25-41 inHl'IIlth Ha:ards in Adolescence, ed. K. Hurrelmannand F. Lose!. New York: Aldine de Gruyter.

Karasek. Robert A.. and Thores Theorell. 1990. HealthvWork. New York: Basic Books.

Kasl, Stanislav V. 1978. Epidemiological contributionsto the study of work stress. Pp. 3-48 in Stress atWork, ed. C. Cooper and R. Payne. New York:Wiley.

Kessler, Ronald C. 1986. Sex differences in the use ofhealth services. Pp. 136-148 in Illness Behavior: AMultidisciplinary Model. ed. S. McHugh andThomas M. Vallis. New York: Plenum.

Kessler. Ronald C.. and Jane D. McLeod. 1985. Socialsupport and mental health in community samples.Pp. 219-240 in Social Support and Health. ed. S.Cohen and L. Syme. New York: Academic.

Kessler. Ronald C.. J. Blake Turner. and James S.House. 1987. Intervening processes in the relation­ship between unemployment and health. Pn'cho­logical Medicine 17:949-96/.

Kessler. Ronald c., and Camille Wortman. 1988. Socialpsychological factors in health and illness. Pp. 69­86 in Handbook of Medical Sociology. 4th ed .. ed.H. E. Freeman and S. Levine. Englewood Cliffs:NJ: Prentice Hall.

Kleinman, Arthur. 1980. Patiellts and Healen' in theColltext of' Culture. Berkeley: University of Cali­fornia Press.

Kohn. Melvin L.. Canni Schooler, Joanne Miller, KarenA. Miller. Carrie Schoenbach, and R. Schoenberg.1983. Work and Personality: An Inquiry into theImpact ofSocial Stratification. Norwood. NJ: Ablex.

CHAPTER 21 Social Psychology and Health 567

Koos, E. 1954. The Health of'Regionvilie. New York:Columbia University Press.

Kosa. J., A. Antonovsky. and Irving K. Zola. ed. 1969.Poverty and Health: A Sociological Analysis. Cam­bridge. MA: Harvard University Press.

Kotarba. Joseph. 1983. Chronic Pain: Its Social Dimen­sions. Beverly Hills: Sage.

Krantz, David S.. and Stephen B. Manuck. /984. Acutepsychophysiologic reactivity and risk of cardiovas­cular disease: A review and mcthodologic criti4ue.p,\·.I'c!wlogiclil Bulletill 96:435-464.

Langer. Ellen 1.. and Judith Rodin. 1976. The effects ofchoice and enhanced personal responsibility for theaged: A field experiment in an institutional setting.Journal of' Per.wllolin· and Social Psycholo!?\' 34:191-198.

Lazarus. Richard S.. and Susan Folkman. 1984. StreH.Appraisal and Coping. New York: Springer.

Leventhal, Howard, Daniel Meyer, and David R. Ner­enz. 1980. The common-sense representation onillness danger. Pp. 7-30 in Medical Psvclwlog.l'.vol. 2, ed. S. Rachman. New York: Pergamon.

Leventhal, Howard. Richard Zimmennan, and Mary Gut­mann. 1984. Compliance: A self-regulation per­spective. Pp. 369-436 in Handbook of' BellllvioralMedicine. ed. W. D. Gentry. New York: Guilford.

Levy, Sandra M.. and Lynda A. Heiden. 1990. Person­ality and social factors in cancer outcome. Pp. 254­279 in Per.wlIlalitr and Disl'llse. ed. Howard S.Friedman. New York: Wiley.

Levy, Sandra M.. Ronald B. Herbennan, A. Maluish,Bernadene Schlien. and Marc Lippman. 1985. Prog­nostic risk assessment in primary breast cancer bybehavioral and immunological parameters. HealthPsychologr 4:99-113.

Liebennan, Morton A. 1986. Social supports: The con­se4uences of psychologizing-A commentary.JOl/mal of' Consulting lind CliniCIII Psychology 54:461-465.

Lynch, James J. 1979. nIl' Broken Heart: Medical Con­sequences of'Loneliness. New York: Basic Books.

Manuck, Stephen B.. Alfred L. Kasprowicz. Scott M.Monroe, Kevin T. Larkin, and Jay Kaplan. 1989.Psychophysiologic reactivity as a dimension of in­dividual differences. Pp. 365-382 in Handbook ofRe.\'earch Methods in Cardiovascular BehavioralMedicine, ed. N. Schneidennan. S. M. Weiss, andP. G. Kaufmann. New York: Plenum.

Marlatt. G. Alan. and Judith R. Gordon, ed. 1985. Re­lapse Prl'l'entioll: Maintenance Strategies in the

568 PART III Social Slruclure, Relarionships, and the Individual

Treatlllellt 01" Addictive Behlll'iors. New York:Guilford.

Mannol, Michael G.. M. Kogevinas, and Mary A. El­ston. 1987. Social/economic status and disease.Pp. 111-135 in Annual Rel'iew 01" Public Health,vol. 8, ed. L. Breslow, J. E. Fielding, and L. B.Lave. Palo Alto, CA: Annual Reviews.

Matthews, Karen A. 1985. Assessment of type A behav­ior. anger. and hostility in cardiovascular disease.Pp. 153-183 in Measuring P,I'ychosocial Variahlesin Epidemiologic Studies 01" Cardio\'{/scular Dis­ease: Proceedin/{s 01" a Workshop, ed. A. M. Ost­feld and E. D. Eaker. NIH publ. no. 85-2270.Washington DC: National Institute of Health.

McGrath, Joseph. 1970. Social alld PsYcho!ogicalFactors ill Stress. New York: Holt, Rinehart &Winston.

McKinlay, John B. 1990, A case for refocussing up­stream: The political economy of illness. Pp. 510­520 in The Sociologv ol"Health and I/Il1ess: CriticalPerspectil'es, ed. Peter Conrad and R. Kern. NewYork: SI, Martin's.

Mechanic, David. 1972. Social psychologic factors af­fecting the presentation of bodily complaints. NeH'

Englalld }ournal of Medicine 286: 1132-1139.---. 1978. Medical Sociology. New York: Free

Press.---. 1986. The concept of illness behaviour: Cul­

ture, situation, and personal predisposition. P,I"\'­chological Medicine 16:1-7.

--. 1990. Promoting health. Socien' (Jan.-Feb.):17-22.

--. 1992a. Health and illness behav ior. Pp. 795­800 in Encvclopedia 01" Sociology. ed. E. F. Bor­gatta and M. L. Borgatta. New York: Macmillan.

---. 1992b. Health and illness behavior and patient­practitioner relationships. Social Science and Medi­cine 34: 1345-1350.

Meichenbaum, Donald, and Dennis C. Turk, ed. 1987.Facilitating Treatment Adherence: A Practitioner'sGuide. New York: Plenum.

Meyer, Daniel. Howard Leventhal. and Mary Gutmann.1985. Common-sense models of illness: The exam­ple of hypertension. Health Psw'hology 4: 115-135.

Michelson. Larry. 1986. Cognitive-behavioral strategiesin the prevention and treatment of antisocial disor­ders in children and adolescents. pp. 275-310 inPrel'ention of Delinquellt Behavior, ed. J. Burchardand S. Burchard. Newbury Park, CA: Sage.

Mirowsky. John. and Catherine Ross. 1989. Soc;o!Cau.les 01" P.I·ychologiml Distress. New York: A IIIine de Gruyter.

Mischler, E. G. 1981. Critical perspectives on thl'biomedical model. Pp. 1-23 in Social Contexts ofHealth. I/Illess, a/l(1 Patient Care. ed. E. G. Mis·chler. L. Amara Singham. S. T. Hauser, R. Liem, S.

D. Osherson. and N. G. Waxler. Cambridge, UK.Cambridge University Press.

Nerem, Robert M.. Murina J. Levesque, and J.Frederick Comhil!. 1980. Social environment asa factor in diet-induced atherosclerosis. Science 20X:1475-1476.

Ohwovoriole, A. E.. and C. Butler Omololu. 1986. Pcrsonality and control of diabetes mellitus. Briti.lh}oufllal ol"Medical P.I·YchologY 59:101-104.

Osgood, D. Wayne, Lloyd D. Johnston, Patrick M.O'Malley, and Jerald G. Bachman, 1988. The gener·ality of deviance in late adolescence and early adult­hood. AlIlerical1 Sociologiml Review 53:81-93.

Pearl in. Leonard I. 1989. The sociological study ofstress. }ournal 01" Health and Social Behavior 30:

241-256.Pearlin, Leonard L Elizabeth G. Menaghan. Morton A.

Liebennan, and Joseph T. Mullan. 1981. The stressprocess. }oufllal 01" Hw!th alld Social Behavior 2~:

337-356.Price, Richard H.. Michelle van Ryn, and Amiram Vi·

nokur. 1992. Impact of a preventive job searchintervention on the likelihood of depression amon!!the unemployed. }oufIlallJI" Health and Social Be·hl/\'ior 33: 158-167.

Rabow, Jerome. and Ronald K. Watts. 1982. Alcoholavailability, alcohol beverage sales and alcohol-re·lated problems. }oufIlal 01" Studies on Alcohol 4-':767-801.

Raphael. Beverley. 1977. Prevention intervention wnhthe recently bereaved. Archil'es 1~1" General PSI'chiatf\' 34: 1450-1454,

Reissman, C. K. 1981. Improving the use of health services by the poor. Pp. 541-546 in The Sociologv o!Health alld I/Illess. ed. P. Conrad and R. Kern. NCIIYork: SI, Martin's.

Review Panel on Coronary-Prone Behavior and HeartDisease. 1981. Coronary-prone behavior and coronary heart behavior and coronary heart disease: ;\critical review. Circulatioll 63:1199-1215.

Rodin, Judith. 1985. The application of social psychoIogy. Pp. 805-881 in Handhook of Social Psycho!

ogl', e(

RandOl--.19~

controlRodin. Judit

ogy.AIRook. Karer

com parchologiport: AIG. Sara:

Rosenstock.peoplePsycho,hen. an,

Roter. Debr:doctor-,Heliith

Roth. Julius.ellce w/,

in the S,CT: JAJ

Sackett, Davplilll1C1:'Johns H

Sandler, lrwi

Kallgrerdata to rgram 1'01!ing Soned. B. H

Samson, IrwSarason.esses: Aing, activ01" Socia,

Schneider. Jolep,I·Y. Pt

Selye. Hans.McGraw

Shavil, J., J.beskind.pressiveity. Sciel

Sklar. Lawrerand canc

Solomon, Gelfactors irease. pan

IIi

ogr. ed. G. Lindzey and E Aronson. New York:

Random House.

--. 1986. Aging and health: Effects of the sense ofcontrol. Scit'llct' 233: 1271-1276.

Rodin. Judith. and Peter Salovey. 19R9. Health psychol­ogy. Allllual Rt'I'it'H' or PSI'cllOlog\, 40:533-579.

Rook. Karen S. 1990. Social relationships as a source ofcompanionship: Implications for older adults' psy­chological well-being. Pp. 219-250 in Social Sup­port: All IIIIt'ractiollal Vit'u', cd. B. R. Sarason. IrwinG. Sarason. and Gregory R. Pierce. New York: Wiley.

Rosenstock. Irwin M.. and John P. Kirsch!. 1979. Whypeople seek health care. Pp. 161-IR8 in Ht'alth

Psrchology: A Halldbook. ed. G. C. Stone. F. Co­hen. and N. E. Adler. San Francisco: Jossey-Bass.

RoteI'. Debra L.. and Judith A. Hall. 1989. Studies ofdoctor-patient interaction. All/lual Rt'I'it'H' o(Puhlic

Ht'alth 10: 163-80.

Roth. Julius. and Peter Conrad. ed. 1987. Tht' Erpt'l'i­(,I/et' alld Mallagt'ult'llt orChrollic 1I111t'ss. Researchin the Sociology of Health Care. vol. 6. Greenwich.CT: JAI Press.

Sackett. David L.. and R. Brian Haynes, ed. 1976. CO/ll­

plimlct' with Thl'rapt'IlIic Rt'gilllt'IIS. Baltimore:

Johns Hopkins University Press.Sandler. Irwin N.. J. C. Gersten, K. Reynolds. Carl A.

Kallgren. and R. Ramirez. 198R. Using theory anddata to plan support interventions: Design of a pro­gram for bereaved children. Pp. 53-R3 in Marslllll­

lillg Social Support: Fomllits. Proet'sst's, alld EftlxtS.

ed. B. H. Gottlieb. Newbury Park. CA: Sage.Sarason. Irwin G.. Gregory R. Pierce. and Barbara R.

Sarason. 1990. Social support and interactional proc­esses: A triadic hypothesis. Special issue: Predict­

ing. activating and facilitating social support. lOllnralof Socialalld Persollal Rt'latiollshiIJ,1 7:495-506.

Schneider. Joseph. and Peter Conrad. 1983. Hal'i/lg Epi­

It'psY. Philadelphia: Temple University Press.Selye, Hans. 1956. The Strns or ure. New York:

McGraw-Hill.Shavit. J.• J. Lewis. G. Tennan. R. Gale. and J. Lie­

beskind. 1984. Opioid peptides mediate the sup­

pressive effect of stress on natural killer cytotoxic­ity. Sciellce 223:188-190.

Sklar. Lawrence S.. and Hymie Anisman. 19R I. Stressand cancer. PsYchological Bulletill 89:369-406.

Solomon. George F. 1981. Emotional and personalityfactors in the onset and course of autoimmune dis­ease. particularly rheumatoid arthritis. Pp. 159-182

CHAPTER 21 Social Psychology and Health 569

in Psrc1lO/lcllroi/lllllHllologl·. cd. R. Ader. NewYork: Academic.

Sternbach. R. A.. and B. Tursky. 1965. Ethnic differ­

ences among housewives in psychophysical andskin potential response to electric shock. Psrcho­IJ!lI'siology 1:241-246.

Sutton. Robert I.. and Robert L. Kahn. 1984. Prediction.understand ing. and control as antidotes to organ­izational stress. Pp. 272-2R5 in Hal/dbook of 01'­

galli:atiOlIllI BelllIl'ior. cd. J. Lorsch. Cambridge,MA: Harvard University Press.

Svarsted, Bonnie. 1976. Physician-patient communica­tion and patient conformity with medical advice.Pp. 220-238 in The Crml'tll or Bureaucratic Ml'di­

cil/I'. ed. David Mechanic. New York: Wiley.

Syme. Leonard. and Rina Alcalay. 19lC. Control ofcigarette smoking from a social perspective. AIl­IIl/al RI'I'it'w or Pllblic Health 3: 179-199.

Tebbi. Cameron K.. K. Michael Cummings. Michael A.Zevon. L. Smith. Mary E. Richards. and Janis C.Mallon. 1986. Compliance of pediatric and adoles­cent cancer patients. Callca 58: 1179-1184.

Todd. Alexandra. 1989. Illtimatt' Adl·t'/'.wril's: Cultural

C,m/lin "etll'l'l'lI Doctors alld WOII/I'll Patil'IIts.

Philadelphia: University of Pennsylvania Press.Tversky, Amos. and Daniel Kahneman. 1974. Judgment

under uncertainty: Heuristics and biases. Sciellet'

IR5: I 124--1 13 I.Umberson. Debra. Camille B. Wortman. and Ronald C.

Kessler. 1992. Widowhood and depression: Ex­plaining long-tenn gender differences in vulner­ability. loumal or Health a/ld Social Bl'lll/l'ior 33:10-24.

Verbrugge. Lois M. 1989. The twain meet: Empiricalexplanations of sex differences in health and mor­tality. lOllmal of Ht'alth illld Social Behal'ior 30:2X2-304.

Visintainer. Madelon A.. Joseph R. Volpicelli. and Mar­tin E. Seligman. 19X2. Tumor rejection in rats afterinescapable shock. Science 216:437-439.

Warner. Kenneth E. 1977. The effects of the anti-smok­ing campaign on cigarette consumption. American

loumal oj' Puhlic Health 67:645-650.Wells. James A. 19R5. Chronic life situations and life

change events. Pp. I05-12X in Measl/ring Ps\,­

clwsocial Varia!JIe in Epidemiologic Studies oj'

Cardiol'lls(,l/Iar Discase. ed. A. M. Ostfeld and

E. A. Eaker. Washington. DC: U.S. Department ofHealth and Human Services.

Itf

IIJ

!riI

IrI

!ff

titt;

~,l

tII

1

f

t

IIf!

I

III!}

570 PART III Social Struclure. Relationships, and the Individual

Wethington. Elaine. and Ronald C. Kessler. 1986. Per­ceived support, received support, and adjustment tostressful life events. Journal of Health and SocialBehavior 27:78-89.

Whitcher-Alagna, Sheryle. 1983. Receiving medicalhelp: A psychosocial perspective on patient reac­tions. Pp. 131-161 in New Directions in Helpinx,cd. A. Nadler, 1. D. Fisher, and B. M. DePaulo.New York: Academic.

Williams, David R. 1990. Socioeconomic differentialsin health: A review and redirection. Social Psvcho/­OX." Quarterll' 52:81-99.

Zborowski. M. 1952. Cultural components ill 1<",1"" •

to pain. Journa/ of Socia/ Issues 8: 16-.\0.Zola, Irving K. 1966. Culture and syrnpllllll' \ ..

analysis of patients' presenting complaints. I,." "can SocioloXical Review 31 :615---630.

---. 1972. Studying the decision to sec a d'h I'"

Review, critique and corrective. Pp. 21 h .' III ,"

Advances in Psychosomatic Medicine. vol. X. <'d /

Lipowski. Basel: Karger.---. 1983. Socio-Medical Inquiries: Recoil," 1/"".

Reflections, and Reconsiderations. Philadl'll'llI.1Temple University Press.

This chaptepeets of sothat occur ihave historlective behbehavior relem-solvingcollective a,tions. to bephenomenamovementsumbrella coamong sch(1992; Tum!theoretical 'cial movetr.events, suclcrowd acti(of both of 1

tations WOl

we focus pdo includelective betthemes anechapter.2

As witlis ambiguittion of socient theoretiwhat dlfferlmost concements: chaI