SOCIAL PSYCHOPHYSIOLOGY, SOCIAL CIRCUMSTANCES, AND HEALTH 1
Douglas Carroll, Ph.D.
University of Birmingham
David Sheffield, Ph.D.
East Tennessee State University
ABSTRACT trated by considering the consequent variations in life expectancy.
At age 20, given the mortality rates operating around 1980 in the
Health varies markedly with social circumstances. While we
U.K., men in higher social class groups could expect to live five
are still without a comprehensive account of the mechanisms which
years longer than men in lower social class groups (9).
underlie this variation, it is clear that psychological factors are
The association between health and socioeconomic status
involved and that key pathways may prove to be psychophysiologi-
holds not only for all-cause mortality, but is evident in a range of
cal. Thus, social psychophysiological research of the kind illus-
health outcomes. It is apparent for most major cause of death
trated in this Special Issue is ideally placed to help unravel some of
groupings, such as cardiovascular disease (10), and for objective
the mechanisms by which social circumstances impact on health.
measures of morbidity (11,12) as well as for both behavioral and
Nevertheless, the success of this sort of social psychophysiological
subjective measures of health and well-being (12,13). The associa-
enterprise most likely depends on reconceptualizing psychophysi-
ological reactivity as a situational, or psychological exposure, tion holds for women as well as men (14), is characteristic of all the
concept rather than as an individual difference concept. This shifts western countries studied in this context (15), and is manifest in
the research goal from one of identifying individuals at risk for data derived for earlier eras (16,17). The generality of the
disease to identifying the psychological exposures that put individu- phenomenon across different contexts, different health outcome
als and groups at risk. measures, and different indices of socioeconomic status makes it
highly unlikely that it is an artifact of measurement.
(Ann Behav Med 1998, 20(4):333-337) It is important to appreciate that it is not a matter of the
poorest suffering from the worst health, while everyone else enjoys
INTRODUCTION uniformly good health. Rather, it is the case of a continuous
The rather ambitious objective of this final article in the gradient of health and ill health mapped on to fine gradations in
Special Issue on social psychophysiology is to try to place the other socioeconomic status (1,3). Expressed another way, inequalities in
contributions into some kind of broad health science context. It is health persist into the better-off social groups. This has implica-
our contention that these papers, whether they explicitly acknowl- tions for the underlying mechanisms. Classical epidemiology
edge it or not, are, in fact, part of a challenging and critically tended to regard health variations as a product of occupational and
important project. Its aim is to increase our understanding of the environmental exposure to physical and chemical hazards. How-
psychophysiological pathways through which variations in the ever, it is difficult to see how differential exposures of this sort
psychological fabric of our social circumstances map on to could account for health variations among the better-off social
variations in our health. The broad mission of social psychophysi- groups. This is not to suggest that physical pathogens in the
ology is to examine the psychophysiological consequences of workplace and the wider environment are not implicated. The
social processes. As such, it is perfectly placed to help unravel the effects on health of damp and substandard housing, chemical
mechanisms by which social circumstances impact on health. hazards at work, and environmental pollutants are undoubtedly
substantial (18-20). It is just that these factors cannot explain
S O C I A L CIRCUMSTANCES AND H E A L T H health variations over the full range of the gradient.
It is now a commonplace observation that health is related to The principle of proximity (21) considers that macro social
social circumstances. The most striking example is the marked constructs, such as socioeconomic status, exert an impact on
inequalities in health contingent on socioeconomic status (1-3). A matters such as health through smaller, more proximal, mediating
strong association with all-cause mortality has been observed processes. It has been argued that the identification of these
regardless of whether socioeconomic status is measured in terms of mediating factors and their underlying mechanisms constitutes the
occupation (4), income (5), material assets (6), education (7), or most formidable intellectual challenge facing not only those whose
composite indices of deprivation (8). The magnitude of these primary research focus lies with health inequalities but also those
health variations with socioeconomic status is most easily illus- grappling with behavioral and psychological health-related pro-
cesses (1,2). Indeed, Carroll et al. (3) contended that understanding
the mechanisms underlying the health gradients contingent on
socioeconomic status "might be regarded as a key test of the
i Preparation of this manuscript was supported in part by financial support frequently evoked but imperfectly articulated biopsychosocial
to the second author from the National Heart, Lung, and Blood Institute, model of health championed by the newly formalized disciplines
Bethesda, MD (1-R29-HL-56825). of health psychology and behavioral medicine" (3, p. 36).
Reprint Address: D. Carroll, Ph.D., School of Sport and Exercise Sciences, What might those more proximal factors be? First, let us
University of Birmingham, Birmingham, B 15 2TT, England. assume that social causation is at work here and that the association
91998 by The Society of Behavioral Medicine. between health and socioeconomic status is not the product of
333
334 ANNALS OF BEHAVIORAL MEDICINE Carroll and Sheffield
health-driven social selection (i.e. movement up and down the the relationship resists correction for the overall wealth of the state
social scale as a function of good and poor health). There are sound (33,34). What these data suggest is that the aggregate health of
reasons for assuming the direction of causality that we do and for nations and states within a nation is a function of the material gap
believing that social selection is minimally involved (22). The two between rich and poor and not just the absolute material status of
most likely causal factors are what, for brevity's sake, we might the poor or, indeed, the average standard of living.
call behavioral and psychological processes. While, as indicated, While egalitarian and less egalitarian nations and states
exposure to physical pathogens at work and in the environment is undoubtedly vary in a number of ways that could impinge on
likely to contribute to health differentials among lower socioeco- health (33), one likely pertinent variation is in terms of public
nomic groups, as well as the variation between lower and higher investment (34). Societies that tolerate substantial income inequali-
groups, it is difficult to see how this contributes to differentials in ties are generally those that also underinvest in public amenities,
health among higher socioeconomic groups. such as accessible health care. Access to health care is undoubtedly
a factor in both the health of populations and in socioeconomic
BEHAVIORAL FACTORS health inequalities. Even in countries like the U.K., with a national
For a factor to be implicated in socioeconomic health health service, different social groups treat and are treated by
differentials, it has to satisfy three basic criteria. First of all, the health care systems in different ways. For example, there is
factor has to affect health. Secondly, it has to be stratified by extensive evidence that, in terms of primary prevention (e.g.
socioeconomic status in an analogous fashion to health. Third, immunization, antenatal care, health checks, family planning,
there has to be plausible psychobiological mechanisms. Many of cancer screening of various kinds), those from lower socioeco-
the most socially problematic unhealthy behaviors meet these nomic backgrounds are less likely to avail themselves of the
criteria. Let us consider the case of cigarette smoking. It is clear provision, even when no direct costs are involved (35). However, it
that of all the putative unhealthy behaviors, smoking is the one is important to recognize that medical care makes only a limited
which most profoundly compromises health (23,24). In addition, contribution to population life expectancy (36) and that it is those
smoking prevalence is stratified by socioeconomic status (12,25,26). causes of death that are not amenable to medical care that
Finally, while gaps in knowledge remain, the mechanisms by contribute most to socioeconomic inequalities in health variation
which smoking exerts its effects on ill health are reasonably well (37).
understood and compelling (27). For present purposes, though, concern lies with another likely
Accordingly, a very parsimonious explanation of socioeco- corollary of income inequality: social cohesion. It has been argued
nomic health differentials might be constructed in terms of the persuasively that as income inequality increases so social cohesion
differential propensity of different social groups to engage in decreases (33,38), and as an expression of this, interpersonal
unhealthy behaviors, such as cigarette smoking. This proposition is relationships deteriorate, perceptions of personal control fade,
given particular weight by a recent analysis of unhealthy behaviors psychological stress increases, and negative emotions such as
across different European countries, which reveals that the magni- hostility are exacerbated. In short, the relationship between income
tude of socioeconomic health differentials for countries reflects the distribution and health provides additional reason for involving
social distribution of smoking behavior (28). However, it should psychological factors in socioeconomic health inequalities. After
also be appreciated that controlling for smoking and other un- all, as Wilkinson (32) noted, "The implication is that our environ-
healthy behaviors in studies of socioeconomic health variations ment and our standard of living no longer impact on our health
attenuates the gradient but in no way abolishes it (29,30). Other primarily through direct physical causes, regardless of our attitudes
studies have shown that a substantial socioeconomic gradient in and perceptions, but have come to so mainly through social and
health exists for individuals who have never smoked and is seen for cognitively mediated processes" (32, p. 405).
causes of death not regarded as smoking-related (31). In sum, it is What might these psychological factors be? We have alluded
evident that the differential prevalence of unhealthy behaviors to some candidates already, and recent reviews (1,2,39) have
accounts for some of the gradient. At the same time though, it is identified, among others, the following: psychological stress, the
also clear that health inequalities persist when variations in absence of social support, lack of personal control, and hostility.
unhealthy behaviors are taken into account. All have been generally implicated in health (39--43), and evidence
also points to inequalities in their distribution among different
P S Y C H O L O G I C A L FACTORS socioeconomic groups (12,44,45). Thus, they meet two of the three
criteria that we suggested were necessary for establishing a factor
In order to understand more fully the factors underlying
as contributing to socioeconomic health variations. What remain to
socioeconomic health variations, it would seem essential, then, to
be determined, though, are the underlying mechanisms.
cast the explanatory net wider into the differing ecologies of
different social groups. In particular, two considerations implicate
psychological aspects of those ecologies as key factors in the MECHANISMS
socioeconomic health gradient. The first has already been men- Broadly speaking, there are two general sorts of pathways
tioned: the persistence of health variations into the better-off social through which psychological factors might contribute to socioeco-
strata. nomic health differentials: behavioral and psychophysiological.
The second arises from recent research indicating that health For example, there is evidence that smoking and unhealthy dietary
status is a function of relative as well as absolute socioeconomic behavior increase in frequency during periods of psychological
position. For example, analyses of life expectancy in western stress (46,47). Similarly, the absence of social support and the
countries reveal a slight correlation with overall national wealth ownership of a hostile disposition are both linked with a whole
but a strong correlation with a measure of income distribution (32). range of unhealthy behaviors (48,49). However, we have already
Analogous explorations of all-cause and specific-cause mortality considered the contribution of social variations in unhealthy
in the states of the U.S. reveal that these outcomes are also behavior to the socioeconomic health gradient, concluding that
substantially related to measures of income distribution and that while the differential propensity of different social groups to
Social Psychophysiology VOLUME 20, NUMBER 4, 1998 335
behave unhealthily explains some of the variation in health social factors may affect the magnitude of responses to traditional,
contingent on socioeconomic status, a substantial proportion of the non-social stressors.
variation remains unexplained. This makes it likely that there are If the accounts of the reactivity hypothesis above can be
direct psychophysiological pathways involved and that the sorts of characterized as individual difference approaches, the conceptuali-
psychological factors identified above directly disrupt human zation that underlies much of what is reported in this Special Issue
psychophysiological function. The prime candidates here are is perhaps best characterized as the psychological exposure
hemodynamic, hemostatic, neuroendocrine, and immunological approach. The distinction is most clearly drawn by Christenfeld
pathways, although the papers that make up this Special Issue and his associates (66), who contrast these different versions of the
focus on the hemodynamic pathway. There are good reasons for reactivity hypothesis as the personality approach and the social
this. Hemodynamic activity has been more fully investigated in the psychology approach. The distinction is an important one for
context of psychological factors, such as stress, and the perturba- behavioral medicine, since the two approaches countenance very
tions produced by stress are generally considered a good bet as a different research agenda and very different remedial strategies.
component in the etiology and/or expression of cardiovascular From the former perspective, the research agenda remains very
disease (50--52). Nevertheless, we would do well not to neglect much as is (i.e. to try to relate individual differences in hemody-
other pathways. Psychological stress, for example, has also been namic reactivity to individual differences in disease risk and
associated with changes in hemostatic markers of atherogenesis outcome). In contrast, a psychological exposure approach would
(53), as well as with alterations to neuroendocrine activity (54) and necessarily switch research attention from individual variations to
to the adhesiveness of protective lymphocytes on the vascular situational variations. In this case, concern would lie with identify-
walls (55). ing psychological circumstances, exposure to which provoked
large magnitude hemodynamic perturbations. Disease, it is pre-
REACTIVITY HYPOTHESIS sumed, is a consequence of life course exposure to such provoca-
tions. It can be seen that this approach gives emphasis to the likely
The papers in this Special Issue are linked not only by their
circumstantial correlates of disease over the likely individual
general concern with hemodynamic reactions in different psycho-
correlates. It terms of remediation, attention would also shift from
logical contexts, but also by the manner in which most of them
the individual to the provoking situation.
conceptualize what has come to be called the reactivity hypothesis.
However, it is important in recasting the reactivity hypothesis
Briefly, the reactivity hypothesis considers that the hemodynamic
in this way not to regard individuals as passive instruments of their
perturbations contingent on different psychological exposures
psychosocial environment. They have agency, and life course
contribute to cardiovascular disease.
exposure to physiologically provocative circumstances very much
However, reactivity has typically been regarded as a trait.
reflects that agency. The transactional model makes this clear
From the conventional perspective, individual differences in the
(67--69). A primary concern of the transactional model is with
magnitude of hemodynamic reactions to particular psychological
those aspects of psychological disposition, such as anger and
exposures dictate individual differences in risk for cardiovascular
hostility, that have been implicated in cardiovascular disease
disease. Most of the prospective tests of the reactivity hypothesis
outcomes. From this sort of perspective, chronically angry and
have embarked from this perspective. Considered together, they
hostile individuals actively structure their worlds in ways that
can hardly be regarded as providing powerful evidence in favor.
increase both the frequency and the intensity of their exposures to
Some find no relationship between the magnitude of hemodynamic
physiologically provocative challenges. An additional feature of
reactions and prospective health outcomes (56-58), and for those
the transactional account, and one that is very evident in this
that do, the association is usually slight (59,60) and/or inconsis-
Special Issue, is the emphasis on studying reactions to social and
tently manifest (61,62). In addition, the one published study to
interpersonal challenges rather than asocial and impersonal ones.
examine variations in the magnitude of reactivity with socioeco-
The former are more characteristic of the exposures which perturb
nomic status found that, if anything, the gradient was opposite to
physiological systems in daily life and provide more scope for the
that predicted on the basis of the socioeconomic gradient for
expression of individual characteristics such as anger and hostility.
cardiovascular disease (63).
Others have argued that for large-magnitude hemodynamic
reactions to have pathological impact, they must be frequently CONCLUSIONS
evoked (52,60). From this perspective, it is the interactive influ- The health variations contingent on socioeconomic circum-
ences of reactivity magnitude and provocation exposure history stances are substantial. Indeed, there is evidence that in countries
that contribute to disease. Supporting this version of the reactivity like the U.S.A. and the U.K., the gradient of health inequalities has
hypothesis are data showing that for animals maintained on fatty been getting even steeper in recent years (70-72). Accordingly,
diets, it was the combination of chronic exposure to psychological there is an urgent need to understand the source of socioeconomic
provocations and reactivity magnitude that predicted the extent of health variations. The failure of social differences in the propensity
atherosclerosis (64). It is important to recognize, though, that this to behave unhealthily to account fully for health variations, the
interactive approach still retains the notion that it is individual continuous nature of the gradient, and the relationship between
differences in reactivity magnitude that confer differential risk for health and income distribution within and between western
disease. It is simply asserting that for the reactive individual to countries implicates psychological factors. As yet, we can point
develop disease he/she has to be exposed to something that only to broad categories of the sort of psychological exposures
provokes a reaction. Even from these perspectives, social pro- likely to be involved. Nevertheless, it is probable that many of the
cesses are relevant to reactivity in at least two ways, as Smith and psychological exposures that have been invoked in this context
Gerin describe in their introduction to this Special Issue (65). First, exert their effects on health through psychophysiological path-
most traditional tasks used to elicit responses do not have an ways. As the papers in this Special Issue attest, an important
explicit social component and thus do not capture one of the most project for social psychophysiologists is to characterize the
important aspects of daily stress: the interpersonal aspect. Second, psychophysiological perturbations elicited by these exposures.
336 ANNALS OF BEHAVIORAL MEDICINE Carroll and Sheffield
This project might also usefully examine neuroendocrine, hemo- (14) Arber S: Gender and class inequalities in health: Understanding the
static, and immunological markers in addition to the customary differentials. In Fox J (ed), Health Inequalities in European Coun-
hemodynamic variables. Indeed, few of the current contributions tries. Aldershot, UK: Gower, 1989, 250-279.
stray much beyond heart rate and blood pressure, in spite of the (15) Fox J (ed): Health Inequalities in European Countries. Aldershot,
UK: Gower, 1988.
ready availability of, for example, impedance cardiograph systems.
(16) Davey Smith G, Carroll D, Rankin S, Rowan D: Socioeconomic
Further, most studies have focused on perturbations during short- differentials in mortality: Evidence from Glasgow graveyards.
lived psychological exposures. Evidence from recent prospective British Medical Journal. 1992, 305:1554--1557.
and psychophysiologic studies provides good reason to also focus (17) Morrison AS, Kirstiner J, Molha A: Life cycle events in 15th century
on recovery from and adaptation to both these acute exposures Florence: Records of the Monte Delle Doti. American Journal of
(73-76) and more chronic exposures (77-79). Some years ago Epidemiology. 1977, 308:1135-1139.
now, the late Paul Obrist urged psychophysiologists to be good (18) Hunter D: The Diseases of Occupations. London: Hodder and
biologists (80). His counsel still holds and the prefix "social" Stoughton, 1955.
affords no exemption. A second key aspect of this project, also well (19) Martin CJ, Platt SD, Hunt S: Housing conditions and health. British
Medical Journal 1987, 294:1125-1127.
illustrated in this Special Issue, is to help refine our understanding
(20) Platt SD, Martin CJ, Hunt S, Lewis CW: Damp housing, mould
of just which exposures provoke marked psychophysiological growth, and symptomatic health state. British Medical Journal.
reactions and what, if anything, attenuates reactivity in different 1989, 298:1673-1678.
contexts. The broad social psychophysiological project illustrated (21) House JS: Social structure and personality. In Rosenberg M, Turner
here is undoubtedly facilitated by the sort of reconceptualization of RH (eds), Social Psychology: Sociological Perspectives. New York:
reactivity from an individual difference concept to a situational, or Basic Books, 1981.
psychological exposure, concept. This shifts the goal from one of (22) Blane D, Bartley M, Davey Smith G: Social selection: What does it
identifying individuals at risk of disease to identifying psychologi- contribute to social class differences in health? Sociology of Health
and lUness. 1993, 15:1-15.
cal circumstances and exposures that put individuals, and for that
(23) Doll R, Peto R: Mortality in relation to smoking: 20 years'
matter whole social groups, at risk. It is likely that what mainly observation on male doctors. British Medical Journal. 1976, ii:469-
differentiates socioeconomic groups is not their tendencies to react 473.
differently to the same psychological exposures, but different (24) Reid DD, Hamilton PJS, McCartney P, Rose G: Smoking and other
psychological exposure profiles throughout the course of their risk factors for coronary heart disease in British civil servants.
lives (3). Lancet. 1976, ii:979-984.
(25) Davey Smith G, Shipley MJ: Confounding of occupation and
smoking: Its magnitude and consequences. Social Science and
REFERENCES Medicine. 1991, 32:1297-1300.
(26) Pugh H, Power C, Goldblatt P, Arber S: Women's lung cancer
(1) Adler NE, Boyce T, Chesney MA, et al: Socioeconomic status and mortality, socioeconomic status, and changing smoking patterns.
health: The challenge of the gradient. American Psychologist. 1994, Social Science and Medicine. 1991, 32:1105-1110.
49:15-24. (27) Davey Smith G, Phillips AN, Neaton JD: Smoking as "independent"
(2) Carroll D, Bennett P, Davey Smith G: Socioeconomic health risk factor for suicide: Illustration of an artifact from observational
inequalities: Their origins and implications. Psychology and Health. epidemiology? Lancet. 1992, 340:709-712.
1993, 8:295-316. (28) Cavelaars AEJM, KunstAE, Mackenbach JP: Socioeconomic differ-
(3) Carroll D, Davey Smith G, Bennett P: Some observations on health ences in risk factors for morbidity and mortality in the European
and socieconomic status. Journal of Health Psychology. 1996, Community: An international comparison. Journal of Health Psychol-
1:23-36. ogy. 1997, 2:353-372.
(4) Townsend P, Davidson N: The Black Report. Harmondsworth, UK: (29) Berkman LF, Breslow L: Health and Ways of Living: The Alameda
Penguin, 1982. County Study. Oxford, UK: Oxford University Press, 1983.
(5) Rogot E, Sorlie PD, Johnson NJ, Schmit C: A Mortality Study ofl.3 (30) Marmot MG, Shipley MJ, Rose G: Inequalities in health--Specific
Million Persons by Demographic, Social, and Economic Factors: explanations of a general pattern? Lancet. 1984, i: 1003-1006.
1979-1985 Follow-Up U.S. National Longitudinal Mortality Study. (31) Davey Smith G, Shipley MJ, Rose G: The magnitude and causes of
Washington, DC: National Institutes of Health, 1993. socioeconomic differentials in mortality: Further evidence from the
(6) Goldblatt P (ed): Longitudinal Study: Mortality and Social Organiza- Whitehall study. Journal of Epidemiology and Community Health.
tion. London: HMSO, 1990. 1990, 44:265-270.
(7) Kitagava EM, Hauser PM: Differential Mortality in the United (32) Wilkinson RG: Income distribution and mortality: A 'natural'
States: A Study of Socioeconomic Epidemiology. Cambridge, MA: experiment. Sociology of Health and lllness. 1990, 12:391--412.
Harvard University Press, 1973. (33) Kaplan GA, Pamuck ER, Lynch JW, Cohen RD, Balfour JL:
(8) Carstairs V, Morris R: Deprivation and Health in Scotland. Aber- Inequality in income and mortality in the United States: Analysis of
deen, UK: Aberdeen University Press, 1991. mortality and potential pathways. British Medical Journal. 1996,
(9) Haberman D, Bloomfield DSF: Social class differences in mortality 312:999-1003.
in Great Britain around 1981. Journal of the Institute of Actuaries. (34) Kennedy BP, Kawachi I, Prothrow-Stith D: Income distribution and
1998, 115:495-517. mortality: Cross-sectional ecologic study of the Robin Hood index in
(10) Marmot MG, Davey Smith G: Socioeconomic differentials in health: the United States. British Medical Journal. 1996, 312:1004-1007.
The contribution of the Whitehall studies. Journal of Health (35) Benzeval M, Judge K, Whitehead M (eds): Tackling Inequalities in
Psychology. 1997, 2:283-296. Health: An Agenda for Action. London: King's Fund, 1995.
(11) Blaxter M: Health and Lifestyle. London: Tavistock, 1990. (36) McKeown T: The Role of Medicine. Oxford, UK: Blackwell, 1979.
(12) Marmot MG, Davey Smith G, Stansfeld D, et al: Health inequalities (37) Carroll D, Davey Smith G: Health and socioeconomic position: A
among British civil servants: The Whitehall II study. Lancet. 1991, commentary. Journal of Health Psychology. 1997, 2:275-282.
337:1387-1392. (38) Wilkinson RG: Unfair Shares. Ilford, UK: Barnardos, 1993.
(13) Chamberlain K: Socioeconomic health differentials: From structure (39) Williams DR: Socioeconomic differentials in health: A review and
to experience. Journal of Health Psychology. 1997, 2:399--411. redirection. Social Psychology Quarterly. 1990, 53:81-99.
Social Psychophysiology VOLUME 20, NUMBER 4, 1998 337
(40) Barefoot JC, Larsen S, v o n d e r Leith L, Schroll M: Hostility, cardiovascular responses to stress on the progression of carotid
incidence of acute myocardial infarction, and mortality in a sample atherosclerosis? American Journal of Public Health. 1998, 88:389-
of older Danish men and women. American Journal of Epidemiol- 394.
og): 1995, 142:477-484. (61) Markovitz JH, Raczynski JM, Wallace D, Chettur V, Cbesney MA:
(41) Boscarino JA: Disease among men 20 years after exposure to severe Cardiovascular reactivity to video game predicts subsequent blood
stress: Implications for clinical research and medical care. Psychoso- pressure increases in young men: The CARDIA study. Psychoso-
matic Medicine. 1997, 59:605-614. matic Medicine. 1998, 60:186-191.
(42) Karasck RA, Theorrell TG, Schwartz JE, Pieper C, Alfredsson L: (62) Matthew KA, Woodall KL, Allen MT: Cardiovascular reactivity to
Job, psychological factors, and coronary heart disease: Swedish stress predicts future blood pressure status. Hypertension. 1993,
prospective findings and U.S. prevalence findings using a new 22:479-485.
occupational influence method. Advances in Cardiology. 1982, (63) Carroll D, Davey Smith G, Sheffield D, Shipley MJ, Marmot MG:
29:62-69. The relationship between socioeconomic status, hostility, and blood
(43) Orth-Gomer N, Johnson JV: Social network interaction and mortal- pressure reactions to mental stress in men: Data from the Whitehall
ity. A six-year follow-up of the Swedish population. Journal of II study. Health Psychology. 1997, 16:131-136.
Chronic Disease. 1987, 40:949-957. (64) Manuck SB, Kaplan JR, Adams MR, Clarkson TB: Effects of stress
(44) Barefoot JC, Peterson BL, Dahlstrom WG, et al: Hostility patterns and the sympathetic nervous system on coronary artery atherosclero-
and health implications: Correlates of Cook-Medley Hostility scores sis in the cynomolgus macaque. American Heart Journal 1988,
in a national survey. Health Psychology. 1991, 10:18-24. 116:328-333.
(45) Berkman LF, Syme SL: Social networks, host resistance, and (65) Smith TW, Gerin W: The social psychophysiology of cardiovascular
mortality: A nine-year follow-up of Alameda County residents. response: An introduction to the special issue. Annals of Behavioral
American Journal of Epidemiology. 1979, 109:186-204. Medicine. 1998, 20(4):243-246.
(46) Schachter S: Pharmacological and psychological determinants of (66) Christenfeld N, Glynn LM, Kulik JA, Gerin W: The social construc-
smoking. In Thorton RE (ed), Smoking Behavior: Physiological and tion of cardiovascular reactivity. Annals of Behavioral Medicine.
Psychological Influences. Edinburgh, UK: Churchill Livingstone, 1998, 20(4):317-325.
1978, 208-228. (67) Smith TW: Concepts and methods in the study of anger, hostility, and
(47) Steptoe A, Lipsey Z, Wardle J: Stress, hassles, and variations in
health. In Siegman AW, Smith TW (eds), Anger, Hostility and the
alcohol consumption, food choice, and physical exercise: A diary
Heart. Hillsdale, NJ: Lawrence Erlbaum, 1994, 23-42.
study. British Journal of Health Psychology. 1998, 3:51-64.
(68) Smith TW: Assessment and modification of coronary-prone behav-
(48) Dean K: Self-care components of lifestyle: The importance of
ior: A transactional view of the person in social context. In Goreczny
gender, attitudes, and their social situations. Social Science and
AJ (ed), Handbook of Health and Rehabilitation Psychology. New
Medicine. 1989, 29:137-152.
York: Plenum Press, 1995, 197-217.
(49) Doherty WL, Schrott HG, Metcalf L, Lasiello-Vallas L: Effect of
(69) Smith TW, Rhodewalt F: On states, traits, and processes: A
spouse support and health beliefs on medication adherence. Journal
transactional alternative to the individual difference assumptions in
of Family Practice. 1983, 17:837-841.
Type A behavior and physiological reactivity. Journal of Research in
(50) Carroll D: Health Psychology: Stress, Behaviour and Disease.
London: Taylor and Francis, 1992. Personality. 1986, 20:229-251.
(51) Krantz DS, Manuck SB: Acute psychophysiologic reactivity and risk (70) McCarron P, Davey Smith G, Wormsley J: Deprivation and mortality
of cardiovascular disease: A review and methodologic critique. in Glasgow: Changes from 1980 to 1992. British Medical Journal
Psychological Bulletin. 1984, 96:435-464. 1994, 309:1481-1482.
(52) Manuck SB: Cardiovascular reactivity in cardiovascular disease: (71) Pappas G, Queen S, Hadden W, Fisher G: The increasing disparity of
"Once more unto the breach." International Journal of Behavioral mortality between socioeconomic groups in the United States, 1960
Medicine. 1994, 1:4-31. and 1986. New England Journal of Medicine. 1993, 329:103-109.
(53) Markovitz JH, Matthews KA: Platelets and coronary heart disease: (72) Phillimore P, Beattie A, Townsend P: Widening inequalities of health
Potential psychophysiologic mechanisms. Psychosomatic Medicine. in northern England 1981-91. British Medical Journal 1994,
1991, 53:643-668. 308:1125-1128.
(54) Kirschbaum C, Hellhammer DH: Salivary cortisol in psychobiologi- (73) al'Absi M, Bongard S, Buchanan T, et al: Cardiovascular and
cal research: An overview. Neuropsychobiology. 1989, 22:150-169. neuroendocrine adjustment to public speaking and mental arithmetic
(55) Mills E Berry C, Dimsdale J, et al: Lymphocyte subset redistribution stressors. Psychophysiology. 1997, 34:266-275.
in response to acute experimental stress: Effects of gender, ethnicity, (74) Brosschot JF, Thayer JF: Anger inhibition, cardiovascular recovery,
and the sympathetic nervous system. Brain, Behavior, and Immunity. and vagal function: A model of the link between hostility and
1995, 9:61-69. cardiovascular disease. Annals of Behavioral Medicine. 1998, 20(4):
(56) Carroll D, Davey Smith G, Sheffield D, et al: Blood pressure 326-332.
reactions to the cold pressor test and the prediction of future blood (75) Linden W, Earle TL, Gerin W, Christenfeld N: Physiological stress
pressure status: Data from the Caerphilly Study. Journal of Human reactivity and recovery: Conceptual siblings prematurely separated?
Hypertension. 1996, 10:777-780. Journal of Psychosomatic Research. 1997, 42:t17-135.
(57) Carroll D, Davey Smith G, Willemsen G, et al: Blood pressure (76) Sheffield D, Davey Smith G, Carroll D, Shipley MJ, Marmot MG:
reactions to the cold pressor test and the prediction of ischaemic Blood pressure recovery to psychological stress and prediction of
heart disease: Data from the Caerphilly study. Journal of Epidemiol- future blood pressure: Data from the Whitehall II study. Psychophysi-
ogy and Community Health. 1998, 52:528-529. ology. 1997, 34:$8.
(58) Coresh J, Klag MJH, Mead LA, Liang K-Y, Whelton PK: Vascular (77) Carroll D, Cross G, Harris MG: Physiological activity during a
activity in young adults and cardiovascular disease: A prospective prolonged mental stress task: Evidence for a shift in the control of
study. Hypertension. 1992, 19:218-223. pressor reactions. Journal of Psychophysiology. 1990, 4:261-269.
(59) Carroll D, Davey Smith G, Sheffield D, Shipley M, Marmot MG: (78) Lepore SJ: Problems and prospects for the social support-reactivity
Pressor reactions to psychological stress and prediction of future hypothesis. Annals of Behavioral Medicine. 1998, 20(4):257-259.
blood pressure: Data from the Whitehall U study. British Medical (79) Miller SB, Ditto B: Cardiovascular responses to an extended
Journal. 1995, 310:771-776. aversive video game task. Psychophysiology 1988, 25:200--208.
(60) Lynch JW, Everson SA, Kaplan GA, Salonen R, Salonen JT: Does (80) Obrist PA: The cardiovascular behavioral interaction As it appears
low socioeconomic status potentiate the effects of heightened today. Psychophysiology. 1976,13:95-107.
David Sheffield
