The Relationship Between Social Support and Physiological ...€¦ · Kampine, 1990, for detailed reviews). The heart muscle gener-ates th e necessar y forc for circulator process

Psychological Bulletin1996. Vol. 119, No. 3, 488-531

Copyright 1996 bv the American Psychological Association, Inc.0033-2909/96/$3.00

The Relationship Between Social Support and Physiological Processes:A Review With Emphasis on Underlying Mechanisms

and Implications for Health

Bert N. UchinoUniversity of Utah

John T. Cacioppo and Janice K. Kiecolt-GlaserOhio State University

In this review, the authors examine the evidence linking social support to physiological processesand characterize the potential mechanisms responsible for these covariations. A review of 81 studiesrevealed that social support was reliably related to beneficial effects on aspects of the cardiovascular,endocrine, and immune systems. An analysis of potential mechanisms underlying these associationsrevealed that (a) potential health-related behaviors do not appear to be responsible for these associ-ations; (b) stress-buffering effects operate in some studies; (c) familial sources of support may beimportant; and (d) emotional support appears to be at least 1 important dimension of social support.Recommendations and directions for future research include the importance of conceptualizingsocial support as a multidimensional construct, examination of potential mechanisms across levelsof analyses, and attention to the physiological process of interest.

Social relationships are a ubiquitous part of life, serving impor-tant social, psychological, and behavioral functions across the life-span. More important, both the quantity and quality of social re-lationships have been reliably related to morbidity and mortality(see reviews by Blazer, 1982;Broadheadetal., 1983;Cassell, 1976;Cobb, 1976; S. Cohen & Syme, 1985; and House, Landis, & Um-berson, 1988). For instance, House et al. reviewed evidence from 6large prospective studies indicating that mortality is higher amongmore socially isolated individuals. These associations hold evenafter inclusion of standard control variables such as age and initialhealth status. Indeed, House et al. summarized evidence showingthat the association between social relationships and health is com-parable with standard risk factors, including smoking, blood pres-sure, and physical activity.

An important issue concerns the potential mechanisms re-sponsible for the epidemiological links between social relation-ships and such long-term health consequences (S. Cohen, 1988;S. Cohen & Wills, 1985; Kiecolt-Glaser &Glaser, 1989). In thepresent review, we first examine the evidence linking the posi-

Bert N. Uchino, Department of Psychology and Health PsychologyProgram, University of Utah; John T. Cacioppo, Department of Psy-chology and Brain, Behavior, Immunity, and Health Program, OhioState University; Janice K. Kiecolt-Glaser, Department of Psychiatryand Brain, Behavior, Immunity, and Health Program, Ohio StateUniversity.

We thank Timothy Smith for his comments on a draft of this article.This study was partially supported by Grants T32-MH18831,MH44660, and MH42096 from the National Institute of MentalHealth, a supplement to MH42096 from the Office of Women's Health,Grant DBS9211483 from the National Science Foundation, and theJohn D. and Catherine T. MacArthur Foundation.

Correspondence concerning this article should be addressed to BertN. Uchino, Department of Psychology, 502 Social-Behavioral SciencesBuilding, University of Utah, Salt Lake City, Utah 84112. Electronicmail may be sent via Internet to [email protected].

tive aspects of social relationships (i.e., social support) to phys-iological processes. We characterize these associations by exam-ining the influence of social support on aspects of the cardiovas-cular, endocrine, and immune systems. The literature searchwas conducted using the ancestry approach and with PsycLIT(1974-1995) and Medline (1983-1995) by crossing the key-words social support, social networks, or social integration withcardiovascular, blood pressure, endocrine, or immune. Onlystudies whose researchers directly examined the association be-tween social support and physiological function were includedin this review. Based on this research, we examined potentialmechanisms responsible for the associations between socialsupport and physiological function (S. Cohen, 1988).

We summarize the research examining social support and phys-iological processes by using both qualitative and meta-analyticprocedures. Major details regarding studies (e.g., type of supportassessment and main findings) were first characterized and ana-lyzed in tabular form. Based on this qualitative analysis, meta-analytic procedures were used primarily when (a) the pattern ofresults were equivocal and (b) there were a sufficient number ofrelatively homogeneous studies (e.g., similar paradigms) to reli-ably characterize the effects of interest. In addition, meta-analyticprocedures were used to test specific hypotheses from our qualita-tive analyses. The meta-analysis was performed using a commer-cially available software package (Mullin, 1989) that provided de-tailed results regarding combined tests of significance levels, effectsizes, tests of variability regarding significance levels and effectsizes, and a fail-safe number.1 Results of the unweighted meta-analysis are reported, but analyses weighted by sample size werealso performed and produced comparable results. To reduce the

1 The fail-safe number represents the number of unpublished nullstudies that would be needed to overturn the conclusions found in themeta-analysis. Although there is no standard fail-safe number, Rosen-thai (1984) suggests that 5k + 10, where k represents the number ofretrieved studies, represents a reasonable tolerance level.

488

SOCIAL SUPPORT AND PHYSIOLOGY 489

problem of nonindependence for studies with multiple assess-ments of social support, results were first transformed within astudy to z scores, averaged, and then entered into the meta-analy-sis. Therefore, as recommended by Rosenthal (1984), only onestatistic was included from each study. Finally, when results werereported as nonsignificant, a conservative significance level of .50was used (Mullin, 1989).

One important source of heterogeneity in the literature on so-cial support and health revolves around the conceptualization andmeasurement of support (Barrera, 1986; S. Cohen & Wills, 1985;Heitzmann & Kaplan, 1988;Orth-Gomer&Unden, 1987; Tardy,1985; Winemiller, Mitchell, Sutliff, & Cline, 1993). In the presentreview, we include diverse studies with both structural (e.g., socialnetwork) and functional (e.g., emotional support) measures of so-cial support. Structural measures of support assess the existenceand interconnection between various social relationships (e.g.,number of siblings), whereas functional measures of support as-sess the particular functions that social relationships may serve(e.g., providing emotional or informational support). The un-derlying theme of these assessments is that they conceptually mea-sure the potentially positive aspects of social relationships. Thisdiversity reflects, in part, the interest that social support has gener-ated in different areas of inquiry (e.g., sociology, psychology, andhealth). When the number of studies permitted it, we performedfocused comparisons between structural and functional measuresof support to examine if they are associated with quantitativelydifferent effects on physiological function (S. Cohen & Wills,1985). In addition, when appropriate we discuss the implicationsof both measures in research regarding social support, physiologi-cal processes, and health.

Social Support and Physiological Processes

More than 18 years have passed since the seminal reviews byCassell (1976) and Cobb (1976) on the importance of socialrelationships for health. These 2 reviews in particular have beenresponsible for generating interest in social support and its rela-tionship to psychological and physical well-being. Cobb focusedprimarily on the stress-buffering effects of social support andemphasized the informational value of social support processes(e.g., that one is cared for and loved) in fostering coping andadaptation. Similarly, Cassell viewed social relationships as po-tentially buffering the individual from life stressors but furtheremphasized the importance of physiological processes in medi-ating the effects of social relationships:

The psychosocial processes thus can be envisaged as enhancing sus-ceptibility to disease. The clinical manifestations of this enhancedsusceptibility will not be a function of the particular psychosocialstressor, but of the physicochemical or microbiologic disease agentsharbored by the organism or to which the organism is exposed.(Cassell, 1976, p. 109)

As suggested by Cassell, the associations between social supportand physical health have been found on such diverse heath out-comes (e.g., coronary heart disease, cancer, and infectiousillnesses) that there are probably multiple physiological path-ways by which social support may influence disease states. Inthis review, we focus on the cardiovascular, endocrine, and im-

mune systems as potential physiological pathways by which so-cial support influences physical health.

Correlational Studies Examining the AssociationBetween Social Support and Cardiovascular Function

Of the 81 studies whose researchers examined social supportand physiological processes, 57 focused on aspects of cardiovas-cular function. This emphasis is understandable consideringthat cardiovascular disorders are still the leading cause of deathin the United States and that social support has been linked tolower coronary heart disease (CHD) rates (House et al., 1988).Conceptually, an examination of the relationship between so-cial support and the cardiovascular system is important becauseof its implications for both the development and maintenance ofCHD. For instance, the prognostic value of tonic arterial bloodpressure in predicting cardiovascular disorders is widely ac-cepted (J. J. Smith & Kampine, 1990). Additionally, the reac-tivity hypothesis suggests that increased cardiovascular reactiv-ity to stress may be an important factor in the development ofcardiovascular disorders (see Krantz & Manuck, 1984; Ma-nuck, 1994; and Matthews etal., 1986).

Because of the relatively large number of studies examiningcardiovascular parameters, we now briefly review basic princi-ples of cardiovascular physiology. The cardiovascular system isinvolved in the transport of oxygen and the removal of carbondioxide, a critical function for every cell and organ in the body(see Larsen, Schneiderman, & Pasin, 1986; and J. J. Smith &Kampine, 1990, for detailed reviews). The heart muscle gener-ates the necessary force for the circulatory process. The vascu-lature (i.e., arteries, veins, and capillaries) serves as the vehiclefor the pumping of the heart.

The most commonly used cardiovascular measures in this re-view include heart rate, systolic blood pressure (SBP), and dia-stolic blood pressure (DBP). Heart rate, a measure of cardiacchronotropy, is usually expressed in beats per minute. It isjointly determined by the sympathetic and parasympatheticnervous systems: Sympathetic activation increases heart rate,whereas parasympathetic activation decreases heart rate.

SBP and DBP are measures of the force of blood against thearterial walls and are a function of both cardiac output and therelative state of the vasculature. Because of the importance ofblood pressure in the transport of blood, it is normally a regu-lated endpoint. SBP is associated with ventricular contraction(i.e., systole) and therefore corresponds to the peak arterialpressure. DBP is associated with ventricular relaxation (i.e.,diastole) and corresponds to the lowest arterial pressure.

For purposes of this review, it is important to distinguish be-tween tonic and phasic components of cardiovascular activity(Cacioppo, Berntson, & Andersen, 1991). Tonic or basal levelsof cardiovascular activity provide information on the tonicphysiologic state of an individual. The correlational studies ex-amining the association between social support and cardiovas-cular function have focused primarily on tonic measures. Thephasic or reactivity components of cardiovascular activity referto momentary fluctuations from tonic levels. Recent laboratorystudies, reviewed later, have focused on the possibility that so-cial support may reduce cardiovascular reactivity to acute psy-chosocial stressors.

490 UCHINO, CACIOPPO, AND KIECOLT-GLASER

An important issue to consider is the psychometric propertiesof the physiological assessments because they bear on the po-tential mechanisms linking social support to long-term physicalhealth. In this regard, it is conceptually important to distinguishbetween measurement reliability and temporal stability. Mea-surement reliability refers to the accurate assessment of thephysiological state at one point in time. In comparison, tempo-ral stability refers to a dispositional characterization of physio-logical function (i.e., stability of the physiological assessmentacross different situations and occasions). Adequate measure-ment reliability is necessary but not sufficient for temporal sta-bility. The distinction between measurement reliability andtemporal stability is important because if social support is tohave effects on disease processes with a long-term etiology, thephysiological assessments should be characterized by temporalstability. The assessment context (e.g., specific tasks), popula-tion (e.g., phobics), and techniques (e.g., specificity of tracersin radioimmunoassay) may all influence an individual differ-ence assessment of physiological function. As an example, aneedle stick is often associated with relatively short-term eleva-tions in catecholamines. Because of the measurement reliabilityof current techniques (Baum & Grunberg, 1995), the catechol-amine changes due to venipuncture would be accurately as-sessed at that point in time. However, this may be a poor indexof an individual's cathecholamine response across time andsituations.

Past researchers have examined the temporal stability of heartrate, SBP, and DBF reactivity. As reviewed by Manuck, Kaspro-wicz, Monroe, Larkin, and Kaplan (1989), measures of heart ratereactivity evidence the strongest test-retest correlation, typicallyranging from .67 to .91. SBP reactivity tends to evidence adequatetemporal stability that is slightly lower than the stability seen forheart rate, whereas DBF tends to show relatively low test-reteststability. Although many of the studies reviewed by Manuck et al.did not report data on the test-retest stability of tonic measures,the patterns of stability across heart rate, SBP, and DBF appearsimilar to that for reactivity assessments. It should be noted, how-ever, that researchers have demonstrated that the stability of thesecardiovascular assessments, including DBF, are enhanced consid-erably when assessments are aggregated across multiple timepoints and multiple tasks (Kamarck, 1992;KamarcketaL, 1992;Manuck, 1994).

There are several methodological issues related to an exami-nation of the relationship between social support and cardio-vascular function. In particular, the use of appropriate statisti-cal controls is important as many of the studies reviewed in thissection are correlational studies in which potential associationswith confounding variables may occur. For instance, social sup-port may be correlated with socioeconomic status, medicationuse, age, and other factors that may have direct influences onphysiological function. We should note that there is some dis-crepancy in the literature on whether such variables are poten-tial confounding variables or mechanisms by which social sup-port has an association with health (S. Cohen, 1988; House etal., 1988). In our tabular analyses of each study, we explicitlynote when such statistical controls were used. In addition, wediscuss the attention (or lack thereof) paid to appropriate sta-tistical controls and its implications for the mechanisms un-

derlying the relationships between social support and cardiovas-cular function.

Many of the studies on social support and cardiovascularfunction have used a correlational design with normotensive in-dividuals. Table 1 summarizes 28 correlational studies, mostof which used middle-aged and older adult samples from thecommunity. Twenty studies examined both men and women, 5examined only men, 2 examined only women, and 1 study didnot report the gender composition of the sample. Researchersof 14 of these studies explicitly assessed some aspect of familialsupport. In addition, researchers of 7 studies assessed structuralmeasures of support, of 15 studies assessed functional measuresof support, and of 6 studies assessed both structural and func-tional measures of support.

In general, the results of the correlational studies are consis-tent with the notion that higher social support is associated withbetter cardiovascular regulation (e.g., lower blood pressure). In1 of the first studies investigating the relationship between socialsupport and cardiovascular function, Kasl and Cobb (1980) ex-amined the influence of social support on blood pressurechanges in response to job termination; they reported that per-ceptions of social support were negatively related to blood pres-sure changes in response to job loss. To summarize Table 1,researchers of 23 studies reported some evidence that socialsupport was associated with better cardiovascular function, of4 studies reported no relationship (see Ely & Mostardi, 1986;Houben, Diedriks, Kant, & Notermans, 1990; Kaufmann &Beehr, 1986; and Lercher, Hortnagl, & Kofler, 1993), and of 1reported opposite effects (Hansell, 1985). A meta-analysis of21 correlational studies whose researchers reported data on theassociation between social support and blood pressure revealeda significant combined test (z = 4.22, p = .00001, fail-safe n =117.38).2 The mean effect size (r) was .08, suggesting a smallbut reliable effect across studies. None of the tests of variabilitywas significant (p > .45). Thus, the evidence for an associationbetween social support and lower blood pressure levels appearsreliable.

We coded each of the studies included in the meta-analysis asmeasuring structural or functional measures of support. Of the6 studies that assessed both types of support, we were able toseparate the effects in 4 of these studies. Therefore, data from 9studies were identified as structural, and data from 14 studies

2 The meta-analysis consisted of 21 studies that directly examined theassociation between social support and tonic blood pressure levels. In 2cases, we averaged the results reported across 2 different published stud-ies (i.e., Dressier, 1980, 1983; Janes, 1990; Janes &Pawson, 1986) be-cause data were apparently reported on the same sample. In addition,the 5 studies examining job-related social support were excluded fromthis analysis because it had been identified a priori as a feature associ-ated with inconsistent effects. In the text, we examine in detail potentialreasons why job-related support may be associated with weak effects onblood pressure.

In our initial search, we excluded 2 studies examining the relationshipbetween social support and blood pressure for methodological reasons(James, LaCroix, Kleinbaum, & Strogatz, 1984; Orth-Gomer, Rosen-gren, & Wilhelmsen, 1993). More specifically, these studies includedparticipants on cardiovascular medication but did not account for thisfactor in reporting the association between social support and bloodpressure.


were identified as functional measures of support. Focusedcomparisons between the structural and functional measures ofsupport revealed no differences in significance level (p = .29) oreffect size (p = .47). Although appropriate caution is warrantedbecause of the small number of studies contrasted, these dataare consistent with the larger literature, suggesting that bothstructural and functional measures of support predict benefi-cial effects on physical health. However, the specific psychologi-cal and behavioral mechanisms that contribute to these effectsmay differ for structural and functional measures (S. Cohen,1988). We return to a discussion of such issues later in thereview.

Researchers in 3 of the 4 studies that did not find any rela-tionship between indices of social support and blood pressureregulation measured job-related social support (Houben et al.,1990; Kaufmann & Beehr, 1986; Lercher et al., 1993). How-ever, Winnubst, Marcelissen, & Kleber (1982) and Unden,Orth-Gomer, & Elofssen (1991) also examined work-relatedsocial support and reported some effects on cardiovascularfunction. One potential reason for this discrepancy may be re-lated to the psychometric properties of the measures of socialsupport. For example, the Lercher et al. measure of social sup-port was two dichotomous questions (also see Houben et al.,1990, which also contains two questions), whereas Winnubst etal.'s measure contained five questions and Unden et al.'s mea-sure contained six questions. Although Kaufmann and Beehrdid report high internal consistencies for their job-related socialsupport measures (. 59 < internal consistencies < .88), they didnot report data on the main effects of social support on bloodpressure (only that the interaction between social support andjob stress was nonsignificant).

It is also possible that job-related social support may not berelated to blood pressure because these relationship are not assignificant for the individual, at least compared with othersources of support. Consistent with this possibility, we reviewevidence later in this article indicating that familial sources ofsupport appear to be associated with reliable effects on bloodpressure regulation.

An issue of interest concerns potential gender differences insocial support processes (Shumaker & Hill, 1991). In this re-gard, the studies in Table I generally suggest that social supportpredicts better cardiovascular regulation in both men andwomen. Many of the studies in Table 1 had the effects of genderstatistically controlled. However, 8 out of the 20 studies whoseresearchers examined both men and women did report gendereffects (Bland, Krogh, Winkelstein, & Trevisan, 1991; Cotting-ton, Brock, House, & Hawthorne, 1985; Dressier, Grell, Gal-lagher, & Viteri, 1992; Dressier, Mata, Chavez, Viteri, & Gal-lagher, 1986; Jackson & Adams-Campbell, 1994; Janes, 1990;Linden, Chambers, Maurice, & Lenz, 1993; Livingston, Le-vine, & Moore, 1991). In an illustrative study, Janes examinedboth structural (i.e., social resources) and functional (i.e.,instrumental) measures of support in Samoan men andwomen. Results revealed that both structural and functional as-pects of support predicted lower blood pressure in men andwomen. However, subsequent analyses by gender revealed thatsocial resources were a stronger predictor of blood pressure inmen, whereas instrumental support was a stronger predictor ofblood pressure in women (also see Henderson, Byrne, Duncan-

Jones, Scott, & Adcock, 1980). Therefore, these data suggestthat social support is important for both men and women butthat specific types of social support may be important as a func-tion of gender (e.g., Bland et al., 1991).

A noteworthy feature of Table 1 is the 8 studies whose research-ers examined the association between social support and cardio-vascular function in differing cultural contexts. For instance, re-search by Dressier and colleagues suggests that culturally impor-tant aspects of social support predict lower blood pressure levels(Dressier, 1980,1983,1991; Dressier, Mata, etal., 1986; Dressier,Santos, & Viteri, 1986; Dressier et al., 1992). In addition, Janesand colleagues (Janes, 1990; Janes & Pawson, 1986) have dataindicating similar associations in Samoan culture. An importantpoint of the data generated by these studies is an emphasis on theproper operationalization of social relationships within a specificcultural context. For instance, Dressier (1983) used historical-cultural analyses to argue that societal membership is an impor-tant facet of social relationships in West Indian culture. Thesecross-cultural data demonstrate the generality of the effects of so-cial support in predicting blood pressure across differing culturalcontexts.

Despite the consistency of the associations reported in Table 1,there are several important issues relating to these studies. Manyof the studies in Table 1 have conceptualized social as a unidimen-sional construct and operationalized social support in terms ofgeneral levels of social integration or perceptions of support.Clearly, social support can be conceptualized as a multidimen-sional construct (S. Cohen & McKay, 1984; Cutrona & Russell,1990), and specific dimensions of social support may be moreeffective when they meet the demands of related stressors(Curtrona & Russell, 1990). In addition, an examination of spe-cific dimensions of social support may suggest more precise mech-anisms through which social support influences health (Uchino,Cacioppo, Malarkey, Glaser, & Kiecolt-Glaser, 1995). To thispoint, researchers in only 4 studies of Table 1 have assessedmultiple functional dimensions of social support (Hanson, Isacs-son, Janzon, Lindell, & Rastam, 1988; Kaufmann & Beehr, 1986;Knox, Theorell, Svensson, & Waller, 1985; Strogatz & James,1986). In 1 study that examined two conceptually distinct func-tional aspects of social support, Strogatz and James found thatlower tangible but not appraisal support was a significant predictorof increased hypertension (i.e., DBF > 90 mm/Hg, or currentuse of hypertensive medication), particularly in low income Blackparticipants. These data are consistent with the notion that specificsupport components may be more effective when they meet thedemands of related situations.

One concern in these correlational studies was the rarity inwhich psychometric data regarding the measurement of socialsupport were reported. Only 11 studies made reference to thepsychometric properties of their scale (e.g., factor analysis andinternal consistency). Given the heterogeneity in which studiessummarized in Table 1 have conceptualized and measured so-cial support, the scales' psychometric properties are importantto examine, especially for the less validated measures ofsupport.

Relatedly, only researchers of 4 studies in Table 1 reportedany data on the temporal stability of the cardiovascular assess-

(text continues on page 499)


21%•- J«K•21

u

§seg

ird

iava

scu

lar

ass1a.%"$.0to

tf.

B4;oi

cco §Z i< |

m

c:O

oci

al s

upport

sment-co

ndi

CO vi

1CO

^"Ec aS SQ. ***

*n c

1*JS M

vS

c,00'«&

•5-355

f61 -t2

a i1 fr•§ c" «flu -O

SI §D,£ -ojs <*> 5s •a *** § o

1? c £S a §.2 *£» ±3 «t*y K! g t>

il E-8« S S.O.>-S-i.6§Is 11•a „ C «

»l§3C M *= c

i -fieQ, £3 -ta- irg c g o.3 «J O. *-*- E 3 e o-•S I c « 9o* § 3Dco Q

a.03aflu03CO

e S

1-2& 3 •-§••03

- * "a•S '?T3

s I §<o<gf

WJ 5*i ""o.-^.ao C n•2 a v-EE E8l«§1

§i*nm

CV

rfr-

J3 <>. M• - O vo —

* S S g <='Sg -E5, c „ -a .g gc o > o 4 J e M *-n

O -3 -2 S S a, <f « ">'i-t-sfi'ssîiK 3 _ c l $ ' F , c , o C og s-g.1 g 8.e.a.a eu <

o00OV

J2XIOO

<*!"wca^

•ocaa.09c<0.

1Ic•on

_cgg

Be5J

"c

*Z

E_c

_ow;n33'>•5c

0.03Q

flu"09CO

coEg8

1^d

u.9"'•£2

3

\o«r>

G"u

c.2-oc

I>3

a T3' £ w C

t c c *o u t« 73

l^slrfl§1 sj s s 5 § gfc B ft^j 10W5

133S'>'•5co.lcErt

iety

mem

ber

!* ecgoj

aog

siblin

gs:

2 o

ri

com

munit}

0-51.5siI1"

,— N

To2,

u>Ci

o£

gc2'«uct-J5E

i ef

fect

for

nu

i'rt

Q

ft

'£

g(U

S

&8co

3-s85'-S trt" on.S £W5 tfi

2 <SW5 ""*

(£J .£™ >"~* £"O _Oc5 T3V5 CbO 01^ aS c

Si

ctio

n b

etw

eei

lose

hig

h in

s

g *S tSc x:

^cIU

E_cQ,

ISh g

25 .SE .gu oC bl

•oeegS2o(A

'S8.3<Ê

^

H.O.cu S-oa03 w D

9.-SS

vest

SB

P a

ndi

effe

ct o

f soc

kers

on S

BP

S. K *-o -a o-^ w >w c #5 I S

z

CL,CQQA."S3CO

«5

1o

ort

fro

m c

ow

rvis

ors

IIW5 «

.Sio ca

c1o

oIS-

c O ^"

E nr^- v^vo «

"a.0Sg6

"3o

|R3 OOC OsC -•

^^"

.1.<L> « i)S 2 ?o a o

5-8* m^j —- ^j py

•> k •> A> ^ > *"•9 S-0 S3" o " *•g 'S -g 5O ^ O J2

rt '31 BJ >k.« t- -ai S S-o S•> f •> 8 .Si- -* U «^5 rjo> MI u O X& 2 & P S3 0 3 g |cfl J3 w •" «t

IpflC CQ *- « 3^ *" O «9 « g £ "

S"U 9- " «a e & c -5

J i f feS.2 so .3 9 13o co o Q «0 0 , 0co co H

0-03Q

oT03CO

153 ^j

'51 •£8 .S-s a'S .Ss asO '? 'B"si's4> O *-i

|1|35 aj Ca c 5ft. Z

VOu-> c

a ae oa *.2 i^1»^" M OS

IB2^^sTf-TI-

73oc3

1

mooo>

t«u"

Q

g.

ity p in

a v

olu

nt;

sisa 2

s|o E0 U


^s^ss^^

— «c2&

Res

ults

Strt |Z S< 8

rt

CO

i!f!"c3 g

Soci

asse

ssm

-«*\a ei S& >>•5 cC „-(E|

c0

ji

•t

&r

x 8 « Si £ 5 > 1c -c -a t: *

Ana

lyse

s pe

rfor

med

sta

tistic

ally

con

trol

ling

for a

gean

d re

lativ

e bo

dy w

eigh

tN

umbe

r of t

imes

eat

ing

out p

er w

eek,

mar

riage

,nu

mbe

r of c

lose

frie

nds,

and

a re

ligio

us a

ffili

atio

asso

ciat

ed w

ith lo

wer

rate

s of

hyp

erte

nsio

nW

idow

ed h

ad h

ighe

r rat

es o

f hyp

erte

nsio

nN

umbe

r of p

eopl

e liv

ing

in h

ouse

hold

ass

ocia

ted

vin

crea

sed

hype

rten

sion

Ana

lyse

s pe

rfor

med

exc

ludi

ng d

iagn

osed

hyp

erte

nN

onhy

pert

ensi

ve w

omen

with

poo

rer s

ocia

l su

ppo

evid

ence

d m

argi

nally

hig

her S

BP

(p =

.07)

, eve

rst

atis

tical

con

trols

for a

ge, a

lcoh

ol c

onsu

mpt

ion

ciga

rette

hab

its, r

elat

ive

wei

ght,

and

educ

atio

n

go At «>.S f Q. KI^§! |s &. s -o « u

g a g g o jc>,t" c a ov p.S t/3

W5 t*-1 >i

•i « -i « «tl•«"* , r 35 -•-» — i oiS.SSw ^ O v c a 3 tn^^2l2l| "8 a§«£ E ° £ * -s a s *^ w ^ n ^ 5 d - ^ c v1 till * .iv-Lw§|||-8| S f f i l

S t c n ^ S ' S o J S o s C f c3 o - p ' S - a S g » > . g , « > J 5

llillllliill^ (2 a

>• g0 S P

<g C <=o 3 S» « gM _ t« .5 <u.2 50. g c S .S g a• o S - 2 I g I 1S^ .aS i^ l^g S e g s l ^Sg.a -s£s s^-^g^ g = & 3f S •» S 1^-â^ 1 ^ O C O P•a-SOS u - i c S e ?-i ^ m^f ""•

1 1.8 -2^5 ^1 10 00 O

s~**foo ,— ;O\ iSc. s•a §* S>»2P -SSS scS 5

„

Ana

lyse

s pe

rfor

med

sta

tistic

ally

con

trol

ling

for s

exgr

ade,

SES

, and

ach

ieve

men

tR

each

abili

ty, p

rim

ary

netw

ork

role

, and

bro

ker

netw

ork

role

ass

ocia

ted

with

hig

her S

BP

Isol

ate

netw

ork

role

ass

ocia

ted

with

low

er S

BP

0-ca

oTme/5

§ "si „a '5 _ % | "S c'la-S 1g£j i"O IS ^ ^ 'C O "* 5

itlllil!!2 | g s a a | « rf2 E 8"8«-S S'S'S- £ u u > " ° ' 0 ' S S ; gg'S o'g g g g ' g < E? js .e o 'C 'C 'C o <s• g O O C C t f J i S e Oz

(rt•8 £- 13S 2a s oS3 C1/5 $ '£c § a

I1— >>.S <N

lllls3£ 03 c/2 •*!• o

v><N

•3.0M

1OO

V~iooO"v

S

«

I

£ t?tS ^

T3 > 3° % * £

Exc

lude

d pa

rtic

ipan

ts o

n m

edic

atio

nN

umbe

r of c

onta

cts

aver

aged

acr

oss

diff

eren

tdi

men

sion

s of

soci

al s

uppo

rt n

egat

ivel

y re

late

d 1

DB

PA

ttach

men

t and

num

ber o

f acq

uain

tanc

es a

ssoc

iaw

ith lo

wer

EPI

in p

ath

mod

el p

redi

ctin

g SB

PN

umbe

r of c

onta

cts

with

acq

uain

tanc

es a

ssoc

iate

clo

wer

HR

in p

ath

mod

el p

redi

ctin

g D

BP

(tabl

e co

nii

i'o,C/5 pQ

afO5")«M

•£ e <§-ao § S'S'S

8 s „- «" H g «•8 J IB f-sf2^|lllÎ1OS c is i-t t«'35.g

111! 111! 1,^ °.S^|I g-9 «,g

llptlllllI'g1jr8|a«8|3•g S K 2 fr-SBS'-B 3 tSrX>*J

5a °^•*s* r^(^ (TJ U(M <N d> efl«r dj" c C9J r *•"»•*

•l|s?Sl 'g-ac S g-S

-111•= § j= c3**i

•ao2SOu

Jcd*••ooO- — ' IA

2s4J 0*x f™1

|.S^


co

I!in i

o

UJZ

2Qft."ffi

00 C

c 8v c

.H 3

il

a.aQeunC/3

a.

"Ss 73.1 .£

1

1 &i

.9 is


C e«a^>; . g

u

OJ

>,155

It.K

•Si•Q

Q.03Q

oTCOc/i

0-ea

y ^^ u« , c> o o t> o'« °" oo - S. "5c A <r. S g- o a

§S!C O Mc <u§ s

oo(N ^•* c

g e

73o

oU

I0000

o'i1U

JCQ.

a

cG

3 .3'

a.csaoToaan

OT3•C tnc •=8 cE oo S

g5

6

496 UCH1NO, CACIOPPO, AND KIECOLT-GLASER

T3^J3

§

§

,

<usf2

y

0.

^

Cgl< «

cd

c

support

t-conditi

c

15 ^

jlC/5 g

as1C c?

'5 c'£'«'TO 00

CU rt

cop

Q

•O3

!S)

port

unre

late

d to S

BP

or

DB

P

a3«

"aS

1

a.aQft."COC/i

c£•g fe^" 3 fcu M °

CU

^t/ju'5.£owE!rt fN*

0 II

_.2c310u

^^

2;;

„

C

10E

Hliil Hi i ilist ifli .1 sifiiii! ilpl il|i! HI HMSmil il 11 II S! I HI !f!i!

if IIlllii5i»|!i illl! I

^™ 5? S ^ ^ ^ ffl D">^Kl ^^^ c * j J— C ' H ^ 3 ^ *~ ffl "O t! ^ V- ^*

"o "pa "" t« S E. 4. Q ^"cs 'O >->>,c c R . ^ ^ — .Sii ^ " c3o *j 13 D,'*-*-^-,fc- gj-g u gi^^S §ca $'3 S"§^= § £ ^-^ H s ^ ^ ' o S 2 ^x> SS^g c ^ c ^ l x . a ^ i ^ s f e t . i i s s o i f i S a S Q

fe *•x paQ < Qft. nT09 m^ M

g t. ;g 'o » u i . ' S SC " 5 o ^ ^ ^ C - " ^ ^ <u°fl r^ a j ^ c S'« ^ "•" E o c C •"* "o S l .— -E P T^ J5 .2 C

IfilllllHlliill lllifg ' o > c J S H c 3 C 3 ' > l a ) ? j ~ ; ' £ 1 ' O 1 * ' " .ES^ -O

« § il !H'IIi i'SJi §-l s 111 iliiiifi

OT £ 5

OO

<N s— ' U

"*-S 1g § iSo" In

C _ u-i p-.

o g 53 C'PE g o e '$ $ 2 8fS y-j "*

C g.0 ^g

S s^ Ô oO u

—

^ —o r2 2^ -ac g•? 3


*?8s.c•12

^•~*u3£

X

1£

+-*C

C/J 8

Si•fl, aj1c§

CO

I!3 4.—i 53

Soci

aiss

essm

e

co

a'ea S§. "'3 .5'€ '„-a S>

Cu a

c.a?(1

>•*•a3

55

8 ° "- T3 c§ &• c' Q I. "8»'S « °o g i- ,_ <n o • 2 s> « ft-

Hi* ffr! 1 5. 1 1! !!! ill i!11 ! 1 liJlJil II

lilll 111! ?! ! H f M l t f l 1i§ 8.1? ^ cc c J o § « g § 1 g.s ags.sri ' r -g aSS-l-- g£.a.2 Sj o-S S » o • 3 - B - c S a B 82.£• ,. u -H a,' §; ° •€ •= e c .« ™ § . S g o ^ S c - c -gf^ « c >.« 3 c * j= § " •£ S S c S -r = ' § > < ' o S « g g.§ -8 § ? Q I -S s -I | a I E 4 •§ s • 8 ' 8 ^ - S - S e e c a |

I !!i l l

1 -g 1 -3 ,s a.s .E = a.E Is > i =e I E fe 8 5 3 .s -3 -a .E O.Q % 1 1 i< 5 z o £ s ^z

ft, 0, ft.co n —Q Q Q0," ft." ft."CO CQ 02™ ™ V3

§ B 1- «f « 3 g i; "2u *3 o a ¥ y ca c e s E . t o3 j ( u a i - O T 5 a « i

a c ' a ' u * * C D .c o S t c S B o o « c -s— ^ • ^ O , ^ " " C a ^ c o1*-^

it!«|1 «flj*. 11!«§••§(=" iS M S ^ E a! 3s g £ 5 2 8. .2 -a c •= ;: § 3S' " " ^ « w ' j 3 n t i a a ^ 1-^<« e ta . •» 2 — ' -5 o c u^ll|IS. f l .go |||1111 H£ • s&« ; i t « g aUllli sill l|lt i cS 'OO. j : o o o a S ^ Q ' "£ V3 C«

o\f> _2? >* Ji «

l lg i -•c E 2^ „ So, p <*- w-> ^ i« g-o^ S o^ 8 ^ 4 , -o C-E j; ng 55 § ^° o g 3 S T g,u A 2 ^3 C OO /-^ ^3 a c ,_;. £KO aa ji a c- C.W ™ ^ 00 ^~. 10 ^

'a 'a "ag i §^ ca ^g

1 S So o 5O 0 0

i, — ' 'c— « aON tl "

2 C ^

T 2cr £cro *a ff. -s <^^— OO 5JJ O ;i?(fl gj <j\ <j

O *> O "(3 O

Q 3 S

Ana

lyse

s pe

rfor

med

sta

tistic

ally

con

trol

ling

for a

ge a

ndse

xPe

rcei

ved

soci

al s

uppo

rt a

t wor

k ne

gativ

ely

rela

ted

toH

R d

urin

g sl

eep,

wor

k, a

nd l

eisu

re ti

me

whe

n al

sost

atis

tical

ly c

ontr

ollin

g fo

r phy

sica

l str

ain

and

wor

kco

ntro

l-de

man

dLo

w p

erce

ived

soc

ial s

uppo

rt a

t wor

k as

soci

ated

with

high

er r

estin

g SB

P co

mpa

red

with

hig

h pe

rcei

ved

soci

al s

uppo

rt

r« OflO C

-O-nfeEo8{1^3 ECQc/n

•E <='O ° 01* 1.&

C tS ^ J3O M O "5.-C " cft S &.22 i^sstJS 3 -^ t- 4> S13 K J4 00 t- S*s? i— . t" « **-« rCS 73 o s o Sc 'o f o >• ZO O "O C . - O

111.!!z3 ""^ • ^ TI **i § .S S 1 1crt j- O i 1> *" 'C *~" "5P 3 C * 3 n y i r : o —fce-§.s88.i«Q l-i l_h CO XJ 1-fc ,* — ,

u <

Ov0

,

"c3^jc

1

(raW

eew

!/m

«es)


?i>3

.****•«*

«J

(Us(2

Res

ults

can e£ ccia

l support

ment-co

ndition

"I

Ifaf,'3.S

£|

c00

>,

1i/5

• g g t 0 "S >> T3 -0 g _S c| 8. fe 2 3 g | „, 2 £ _ _ < 2t nil! Mil i in i ini inn i i t i i in i in! 1 | i]i& i - f t 1 Hi1 H 5 - S 4 J - S 3 «s iSâ | 1§| .s||||« 3-05. 'Sg s 8 °^B° S > > h g 8§ |o l

1 1 1 «fl i 11il?!!,! s Iliilllilli IWIitilgfiii iiilil il-iiliniii!• " ( O U f t , C L , < o , c / ) a, wi 1/3

&a."a« o M a, a,so a" n n° Irffe Q Q«!ft" e E Q &•' Bu* 3C« 1 ffl nW3 < c/3 ^

50 "g x-,C W O JS

Eg-ri| 3?E S l 1o g M« „ -s a c S o w u 5

!ii !l|l illI|il3? Ill fill? lit1J5JJI- 15' *iSI* fl*ls|| II l;| illf! Ill•8s£«!|! ^1| 8.815? litIsslsll |li |o3s l 111fl. V! U 1

^5 aj o 4Je -S .S .a ^

1|I f^Ll a! Jfisl IIS! iirfis !gss.g^ l l |=ggS|^gs^l-s s£g|s I^Bllilgil'llS 8f< E ^ 2 S^cSE l i lSE l i ^ lSa . E, |P S Z S S

x: o ^^ <• i«2 (N e 73 CN ,

- i 'i -s ^ E 'S • = • " £ "•s iSs f l i l s - a S e & o s O . - _ 8 2 S 2 = o

i*IS!fII 13 « S S2 fc « 32 '« ^1°^:^ «» S - a - B p E e E ' S^ I s s S ' S & ^ 3 > - a g ? - c . s e "» " S 3 > > S i ' c « oH « 2 ^ C j - C £ - H C - ^ - ^ ^ ^ C - H H e - C K w ^ t - sg ^ e - u E s a g E a a g a E s g § E a a | 6j= 6<J« O < U O <

fT <-v C? ^~.ON (S a\ m

CTi Ov S Ov— ON 2- ^^ ^^ * — ^-

1 ^ 3 •aS s S «je h 6 cg B | |& '& 2 3


Z<

a.«QaTeo00

fe yffl ft.

1 Sft. ^C

WO

aa3

00

2 wg;£2

P 8u .3

C/3

1ÎSD. ""

S Ssfl

§, o

I oU

•

S

e; in

-0, ,

"5 6

nt bj

ments( Dressier, 1980, 1983;Hansell, 1985; Kaufman &Beehr,1986). As a representative study, Dressier (1980) assessed SBPand DBP three times during the study, with each assessmentseparated by 20 min. Dressier reported strong average corre-lations between these reading (r= .89forSBP, r= .88 for DBP)and thus evidence for temporal stability of the tonic assessmentsover relatively short periods of time. Nevertheless, the temporalstability of the physiological measures needs examination moreconsistently in correlational studies if social support is to beassociated with the long-term development of CHD.

Finally, an important methodological issue for the studiessummarized in Table 1 concerns the lack of explicit controls forcardiovascular-altering medications (e.g., beta blockers), espe-cially because most of these studies tested middle-aged to olderadult samples. Only 10 of the studies summarized in Table 1directly reported that they had excluded individuals on cardio-vascular medication or statistically controlled for this variablein their analyses. Researchers of 3 studies examined a youngpopulation in which medication use may not have been prob-lematic. This issue is important because one might expect indi-viduals low in social support to be characterized by higherblood pressure levels, however, the use of cardiovascular medi-cations would serve to lower their blood pressure and eliminateor weaken the association between social support and tonicblood pressure levels. In fact, this may be one factor responsiblefor the relatively small, albeit reliable, meta-analytic associationreported in this review. Future studies need to be sensitive tothis potential confound.

Intervention Studies Examining the Effects of SocialSupport on Cardiovascular Function

The evidence summarized in Table 1 suggests an associationbetween social support and cardiovascular function. However,the correlational designs may limit the inferences from thesestudies. Table 2 summarizes 6 prospective intervention studieswith normotensive middle-aged and older adult samples thatexamined social relationships and cardiovascular function. Thesocial support interventions in these studies varied: 1 study pro-vided opportunities for increased social interactions in activitygroups, 2 studies used group discussions to increase social in-teractions, 1 study contained a social support education com-ponent, and 2 studies used social support to facilitate exercise,lifestyle, and stress management training. Two studies testedonly women, 3 studies both men and women, and the remainingstudy did not report the gender composition of the sample.

Because of the small number of studies and the diverse waysin which social support was operationalized, we performed aqualitative analyses of the studies in Table 2. In this regard, re-searchers of 4 of the 6 studies in Table 2 reported positive effectsof social support on cardiovascular function. In a prototypicinvestigation, Andersson (1985) studied older adults who re-ported problems with loneliness and randomly assigned themto either a small group discussion (e.g., opportunities for leisureactivities) that met four times during the intervention or a con-trol condition of individuals followed for the duration of thestudy. A 6-month follow-up revealed that the support interven-tion participants had a greater frequency of social contact and

500 UCHINO, CACIOPPO, AND KIECOLT-OLASER

"2a"3

J_U ^t32 'C,BJ JOH Co

1

O

a

S i l tw o c o

• - c" 8 e".3 I w | £o E i~. E f

Pllalli.!,g.-ot2.E g S S 3 -o00 1/1

O S^J3 U5 |~~-S E<-e w ii

o

S?

O

v. *«^

f oc


ai

S<

co

ll

I

<N

O.oaQoT09VI

S -c&> a "-S g<2S •£ g-s .a!5 5 ?1 81ft S u

cA O «3 •£ >

E Ef aM rTJ2 a

I2III& K -O "B T5

j|l!|c rt «> B..S

pjll

ca

c5o.S

fe"Q-oaQ

•S-a"S § - K StS -a S & §.sl'sasT) O U (9 2

2 &.S E^.35

I

c S C a v "tZ ™S o « I1 <A 43 — £ —S?|i-B-I«-Sl

^'•53 •55;.0 ON

is 5->8 »>< B.2

- ^^d ii£•5^ "11 TIE£ !2 « 5/3

•an s;*

§JS1


leisure activities and, more important, lower SBP and DBF. Thecontrol condition evidenced no such changes.

Only researchers of 2 of the studies conducted to date re-ported no effect of a social support intervention on blood pres-sure regulation (Arnetz, Theorell, Levi, Kallner, & Eneroth,1983; Gill, Veigl, Shuster, & Notelovitz, 1984). However, Gill etal. did not find a significant manipulation check on social sup-port, which suggests that the intervention was unsuccessful inaffecting participants' support networks. The null finding byArnetz et al. is more difficult to explain. However, familial rela-tionships may be relatively important for blood pressure regu-lation (see Potential Mechanisms). In general, stronger associa-tions might be obtained with interventions that focus on famil-ial sources of support.

Although the studies summarized in Tables 1 and 2 suggest thatsocial support influences cardiovascular function, few of thesestudies suggested its effect on established risk factors. To this point,we summarize 8 prospective (primarily intervention) studies withhypertensive patients in Table 3. Researchers of 5 of the 8 studiesin Table 3 explicitly noted that they examined both White andBlack participants. Researchers of 6 studies examined both menand women, whereas of 2 studies examined only men. Six of theinterventions used for the most part familial sources of support.However, researchers of 1 study simply assessed naturalistic socialsupport, and of 1 study used organizational social support.

Evidence for the role of social support on cardiovascularfunction and risk factors comes from the prospective interven-tion studies on hypertensive individuals summarized in Table3. In an early study, Levine et al. (1979) identified 400 hyper-tensive patients and assigned them to interventions consistingof an exit interview, family support, small group, various com-binations of these groups (e.g., exit interview and familysupport), or a control condition. In the family support condi-tion, patients were asked to identify a target individual withwhom they had frequent contact (typically a spouse). Thetarget individuals were then trained to increase understanding,support, and reinforcement regarding positive management ofthe patient's hypertensive state. Results revealed that familysupport alone decreased DBF (i.e., DBF was below the hyper-tensive limits for the participant's particular age group) by 11 %at an 18-month follow-up assessment. Predictably, exposure toall intervention conditions was associated with the best bloodpressure control (28%). Subsequent follow-ups of this projectsample revealed reliable long-term effects of the social supportmanipulation on blood pressure regulation (Morisky, DeMuth,Field-Pass, Green, & Levine, 1985; Morisky et al., 1983). Ameta-analysis of studies whose researchers have used social sup-port manipulations to control blood pressure in at-risk popula-tions (Earp, Ory, & Strogatz, 1982; Erfurt, Foote, & Heirich,1991; Levine et al., 1979; Morisky et al., 1983, 1985; Stahl,Kelley, Neill, Grim & Mamlin, 1984) revealed a significantcombined test (z = 3.32, p = .0004, fail-safe n = 12.29).3 Themean effect size was r = .15, and no test of variability was sig-nificant (p > .20). These prospective data from at-risk popula-tions provide evidence that social support may have beneficialeffects on established risk factors.

Although the results of the prospective intervention studieswith normotensives and hypertensives suggest that social sup-port leads to better blood pressure regulation, there are several

issues raised by these studies. The prospective interventionstudies with hypertensives were primarily designed to affect tan-gible aspects of support. However, the social support manipula-tions may have affected other aspects of social support, includ-ing appraisal support due to the increased participation andknowledge of the support provider. Furthermore, althoughthese studies suggest tangible support may have been importantbecause of better medical adherence (e.g., Levine et al., 1979),none of the studies researchers performed statistical analyses todirectly examine the importance of this factor. Interestingly, theprospective intervention studies generally preceded the correla-tional studies summarized in Table 1 that suggest tangible fac-tors alone cannot explain the associations between social sup-port and blood pressure regulation. Therefore, the prospectivedata are only suggestive of tangible support influences on bloodpressure regulation because other unmeasured components ofsocial support may have contributed to these effects.

We should also note that aspects of several interventions withnormotensive participants might have affected other health-related processes (e.g., Andersson, 1985; Sallis, Trevorrow,Johnson, Hovell, & Kaplan, 1987). For instance, the Sallis et al.manipulation also informed participants of the harmful effectsof stress (also see Clifford, Tan, & Gorsuch, 1991). Therefore,lifestyle or behavioral changes related to stress, but not directlyinvolving social support, may have also contributed to the re-sults of these studies. Nevertheless, the prospective design of thestudies in Tables 2 and 3, along with the importance of bloodpressure regulation in hypertensive individuals, provides rela-tively strong evidence linking social support to risk factors.

Potential Mechanisms Linking Social Support toCardiovascular Function

Because of the consistency of the associations between socialsupport and cardiovascular parameters presented in Tables 1 to3, we now turn to specifying the potential mechanisms respon-sible for these covariations. In a review of potential mechanismslinking social support to health, S. Cohen (1988) suggests thatsocial support may have beneficial effects through social (e.g.,stress buffering), psychological (e.g., affective states), and be-havioral (e.g., health-promoting) mechanisms. Consistent withS. Cohen (1988), we examined the mechanisms linking socialsupport to physiological processes at different levels of analysis(also see Cacioppo & Berntson, 1992).

At a social psychological level of analysis, it appears that fa-milial sources of social support may be associated with reliableeffects on blood pressure regulation. A meta-analysis of 12 cor-relational studies whose researchers explicitly noted that they

3 The meta-analysis consisted of 4 studies that used social supportmanipulations to control blood pressure in at-risk populations. One teststatistic is entered for the 3 Levine et al. and Morisky et al. studies aboveto reduce potential problems with the nonindependent samples. ThePinto, Sirota, and Brown (1985) study was not included because nostatistics were presented for their case study. For all prospective studies,the statistic entered was the difference in blood pressure level or controlbetween the social support manipulation and a control condition duringthe final assessment, thereby providing evidence on the long-term effectsof these interventions.


I

COZ

I a.aaQ

a.COQ

£

I

fc,J3

•5"§«

.2?

c jj.a S

jjr-•S Ca %£ St. o

\^ B •r C, „ M Oimn

t ">m C-s "~ -o & a 9

zz? a-s e-g||rio"IS^.i

lif^lllP ao i 1 toCD

IS«;c '

s:&

[3

S

504 UCH1NO, CACIOPPO, AND K1ECOLT-GLASER

/rm^*»

•S

8£T3caH

Res

ults

«-o

Z §•< 8

eg

e"

ial s

uppo

rten

t-con

ditio

i

8 f iC/3 «

3

"etC

«ODc3

•*-

1•3fi<2

c00

oD

•o£

Follo

win

g th

e in

terv

entio

n, D

BF

was

with

inno

rmal

lim

its (D

BF

<: 90

mm

/Hg)

a,CDQ

ao o'3 00 c2 , «

> 1 1 I I^ S I'M s

*l!i|

Jl°i|tolll-l-e ._ 5 .S so£ 2 £ EW

EC

p ,_j

I 1•I"2 BC S3 Q,S S ^

IIS•rj C O•C 3 -^»J s j§.s •- s ?E * S os^&i•§£.£«

.S"ie•9•c!&1c °

« o E•S 3<sal «J "-

a. > T3if h co 2 a

OOo\

^

uocE

v_ w ca -Q ^ c ^«

l№ 1 .hi i\ »|, S ! t|!M111! Itssfl ls ! |i||i§! jilli111! Sis !l| fl il lit Si

!! I l l f f l I l i f ll i « s § S l ? - s i 8 s l f e i 2 •isiJi-ss Isiliiiiii ills>,'p 3 -§na ^ - S ' S ^ e i e c B B f t . . & e c g - & " § ) § ' = •=" c C S

.•3 c jj -|f ^T ra 8 ft'C E S S S o o S o s s H S - s S o ' - ' a s S 1 3

p o e g o p c . S P r t o ' S o o c w S • p o o c 2 S o 2 - 3 o o K ec o E s c g as > o g o o E o " t o o . S i s s o o o g o o o o e oli. U. Z Z OL, CU

ft. a. o.co ca coQ D Oa," ft." cCCQ co atA !S) W

Ea jj S „ _ f. S's°" S l I S j = S K •S'* 3

S g S c a - -0 o* .. g -•?>,o= t tRSSSêôg

till |l*«l li f i l lIIP II Jil lIjj l II I'll §11 1 fill

" ( U U j ' t U ' ^ J i l H « ^ : i l i n I a 3 ^ i ' 3 « ^ t j C >

§£•£•5 H a ^ g S a ^ w ' s i g o c j c o bRj i* •« *» JrU^ & S CU O

*° p2 O ^35 Jgg js g U ^j U j"(« O £ r" ^ î V5 f~- ® * * G 2 O

gl i g ^ s|||5|l|£-4'B «s > S S> !*•• c "3.m I' '2 •- •£> s '5S g c '2 J; ' g c S u i J S ' a e g CC r ^ «r1^ ° ' 2 , . l f i - ^ - S 3 ' C C ! ^S S C-^2 n S K ^ g 3 w 5 " O 3 >

^^ 5> >; O.-B m "S SJ ^ S -o J< § O OO °^ *~* 5 ""

«>'§' tt =*> «£ * o -"g E , s i " 2' S § a « "S r n ^ - . £ ' S 3 - S£ - o - a ^ e S £ _ .. c S '§ 'S'" ' t7«j 'T£?-5-

' S i n U j g M g 'So ' S ' c S C « 9 ^ 9 C M C ' K

*2 S ' *•* O ^*P ^ ^ ^ 3 3 T 3 3 * i 5 § !^ ^ § O ^ ^ ^ S u c ^ ' ^ T3 "O ^ TD * Q*

PfllH Pi li's till till( £ < e o £ w-r 2 u — •"'S 2 ~oo ? 2O • — * G Cb. . -S 5 w

grou

pR

esul

ts si

mila

r whe

n on

ly e

xam

inin

gin

divi

dual

s who

act

ivel

y pa

rtici

pate

d in

inte

rven

tion

prog

ram

s at

the

site

s

•o

= di

asto

lic b

lood

pre

ssur

e; S

BP

= sy

stol

ic b

loo

ft.CQQin

SS"gc

SI I

us s

yste

m; B

5

oScu

1ocoC3I I

05

<

IS1/3ig<uG531.1'ie•5

T3Baj

§•o•8I*yC

<u

11O<ugz«jS;


assessed social support related to family members revealed areliable combined test of significance (z = 4.17, p = .00001,fail-safe n = 64.97). The associated effect size for this analysiswas r— .12, and no test of variability was significant (p > .09).In 1 study whose researchers directly examined blood pressurein the presence of different social relationships, Spitzer, Llabre,Ironson, Gellman, and Schneiderman (1992) found that beingaround a family member was associated with lower ambulatorySBP and DBF compared with being around a friend or astranger. The prospective interventions with hypertensive pa-tients that directly used family members as sources of supportprovide convergent evidence on the importance of familialsources of support on blood pressure regulation.

Researchers of 8 of the correlational studies have directlytested the potential stress-buffering effects of social support oncardiovascular function (Dressier, 1980, 1991; Dressier, Matar,et al., 1986; Janes, 1990; Kasl & Cobb, 1980; Kaufmann &Beehr, 1986; Strogatz & James, 1986; Winnubst et al., 1982).As argued by S. Cohen and Wills {1985), one methodologicalrequirement for a test of the buffering model consists of demon-strating a significant main effect for the stress assessment to en-sure that the measure was characterized by an adequate rangeof scores and measurement reliability. However, only research-ers of 4 of these studies reported data indicating that their mea-sure of stress was associated with blood pressure. These 4 stud-ies were associated with a significant combined test (z = 3.39, p= .0003, fail safe n = 12.99) and an effect size ofr = . 18. No testof variability was significant (p > .34). In 1 illustrative study,Dressier (1980) reported an interaction between structuralmeasures of support (i.e., no. of siblings) and levels of life stressfor SBP and DBF: Individuals high in number of siblings andlow in life stress were characterized by the lowest blood pres-sure. Although Cohen and Wills suggest that buffering effectsare more likely to be found when there is a reasonable matchbetween the stressor type and support function, they also re-ported that buffering effects were sometimes found when re-searchers assessed close interpersonal relationships. Consistentwith Cohen and Wills, all 5 studies whose researchers examinedfamilial relationships (e.g., spouse and siblings) reported a sig-nificant buffering effect on cardiovascular regulation (z = 3.43,p = .0003, fail-safe n = 16.74), with an effect size ofr = .14. Notest of variability was significant (p > .55)* These studies fur-ther underscore the potential importance of examining familialsources of social support in studies of cardiovascular regulation.

The studies summarized in Table 2 suggest that structuredinteractions with others may also produce beneficial effects oncardiovascular function. However, these results may not simplybe a function of the intervention discussion because such struc-tured interactions appear to generalize to others in one's net-work (e.g., Andersson, 1985). Therefore, the studies in Table 2may produce part of their effects by increasing social compe-tence or the perceived importance of social interactions in one'ssocial network (Sallis et al., 1987).

At a more behavioral level of analysis, part of the associationbetween social support and cardiovascular function may be aresult of health-related lifestyle factors (Umberson, 1987). Forexample, social support may be associated with better cardio-vascular regulation because individuals high in social supportengage in better health practices (e.g., better diet and more

physical activity). Contrary to this position, the associations be-tween aspects of social support and cardiovascular function re-mained significant even after statistically controlling for a num-ber of health-related variables, including weight or body mass(e.g., Bland et al., 1991; Janes & Pawson, 1986; Stavig, Igra, &Leonard, 1984). However, it should be noted that many of theseresearchers have not assessed specific health-related behaviors(e.g., substance abuse). In addition, of those researchers thatdid assess specific health-related behaviors, data on the reliabil-ity or validity of their assessments were typically not reported(see Umberson, 1987).

At a psychological level of analysis, perceptions of stress, feel-ings of controllability, intrusive or ruminative thinking, feelingsof loneliness, depression, and other emotional processes (e.g.,anxiety) are potential psychological mechanisms for the associ-ations between social support and cardiovascular function(Collins, Dunkel-Schetter, Lobel, & Scrimshaw, 1993; Pierce,Sarason, & Sarason, 1991; Quittner, Glueckauf, & Jackson,1990; Russell & Cutrona, 1991; Stokes, 1985; Solomon, Miku-lincer, & Hobfoll, 1986). Unfortunately, empirical data are un-available concerning the psychological mechanisms responsiblefor the associations between social support and cardiovascularfunction reported in Tables 1-3. Future research is clearlyneeded in this area of inquiry. We return to this important pointlater in the review.

Laboratory Studies Examining the Effects of SocialSupport on Cardiovascular Function

Whereas the prior studies have focused primarily on tonicmeasures of cardiac function, many of the recent studies havebeen experimental, laboratory studies conducted under the ru-bric of the reactivity hypothesis. Briefly, the reactivity hypothe-sis suggests that exaggerated cardiovascular reactivity to stres-sors may be a pathogenic mechanism influencing the develop-ment of cardiovascular disorders (see Krantz & Manuck, 1984;Manuck, 1994; and Matthews et al., 1986). These 15 studiesare summarized in Table 4. Thirteen of these studies tested rel-atively young participants under the age of 30. One study useda middle-aged sample, and 1 study used an older adult sample.Researchers of 8 of these studies exclusively examined women,of 3 examined men, and of 4 examined both men and women.In 11 studies, researchers examined social support through ex-perimental manipulation. Researchers of the remaining 4 stud-ies assessed naturalistic levels of social support. Of these 4 stud-ies, 3 studies' researchers examined functional measures of sup-port, and 1 examined a combined index of structural andfunctional support.

The laboratory studies in Table 4 collectively suggest that so-cial support may reduce cardiovascular (or autonomic nervoussystem) reactivity to acute psychological stress. One salient fea-


4 We should note that 4 of the 5 studies' researchers who assessedfamilial sources of support also reported significant effects of their stressmeasures. Therefore, the influence of close interpersonal relationshipsand the methodological requirement suggested by S. Cohen and Wills(1985) are potentially confounded in these meta-analyses of bufferingeffects.


c y <sIIIo ^-'-P*8 a §2


1

mi.a E S la •

S -C S '$ gO « P a .9

U BO u " U

III if0,

tfS

B

1

1'•3

scus

sion

did

no

ropo

rtio

n of

tim

e ta

lkin

g du

ring

dii

a.

t to

leav

e,

c&

Is4>

•1

aj

supp

ort c

ondi

tion

sm

otio

nal r

espo

nses

to d

iscu

ssio

n (i

w

IQ.

vee.n

su

Si•ooB

T3•3

anxi

ety,

ang

er, a

nd w

ant t

o re

turn

cond

ition

s, a

nd

DBP

s-BQW5

t£X&

iwer

incr

ease

s ii

uppo

rt c

ondi

tion

asso

ciat

ed w

ith lo

on

§»s u

-C

If?° C toi S.BIs & 2

ared

with

no

suta

lkin

g di

scus

siw

ith f

rien

d pr

eh

stra

nger

pre

s<

reac

tivity

to d

iscu

ssio

n ta

sk c

omp;

(res

ults

sim

ilar

for b

oth

quie

t and

'hos

e w

ho p

erfo

rmed

stre

ssfu

l ta

sks

obse

rver

as

mor

e fr

iend

ly th

an w

it

H

OCT!•61St»>

n3

qual

ly n

onev

alri

ends

and

stra

nger

s w

ere

rate

d as

e

u,

1•o1"

•oBi3•oB8

reat

er p

eak

skir

diff

er in

task

rat

ings

of d

iffic

ulty

ai

tran

ger c

ondi

tion

ass

ocia

ted

with

gi

HR

than

alo

ne c

ondi

tion

V5

e1&

8..S

VI

B

i

w

•a

he f

rien

Bjrt

task

per

iods

thi

1,tB -5te s•3 •—•Sc

cond

ition

s di

d

tran

ger c

ondi

tion

asso

ciat

ed w

ith g

ico

nduc

tanc

e fr

om a

ntic

ipat

ory

toal

one

cond

ition

(fri

end

and

alon

e

(n

b g '

s>•&i1

£(2

II

|Sifffl^laI^H^SIllw as #

508 UCHINO, CAC1OPPO, AND KIECOLT-GLASER

a sc aa^'3 .e'•g -

"Sxs

1)S

c-ffl ft,V5 gg

a.03Qa."non

III a §qfi> .S 3 .5

a.so

ssniiti"ill si II

CL.eaCO

.• .^Ô R W O CSc o _§ '3 g;12 o 3 c,D


Res

ults

14= S CQ• - Q Q

g

i>Co<di

tlll'li11 $

i11

.§ §•"> . .•a a -a c« > '-3 o ,o'6 -a 8 -3 '•§ '- -

rt vi

o-oa o,SO BQ

efQS

0.

SB-

a

E

I'

If

if

ill•S -w

I'lles-l

e c.9 a

t/~) j™

g'ljS:3 "3"

•S s

I•c •u •O

KS

65


ture of the studies summarized in Table 4 is that strong resultsemerge with laboratory paradigms that manipulate the sup-portive functions of social relationships (Gerin, Pieper, Levy,& Pickering, 1992;Kiecolt-Glaser&Greenberg, 1984;Lepore,1995; Lepore, Allen, & Evans, 1993), rather than simply relat-ing naturalistic social support measures to laboratory stressors(e.g., Boyce & Chesterman, 1990; Tardy, Thompson, & Allen,1989). The 4 studies whose researchers manipulated the sup-portive function of social relationships and examined its effectson cardiovascular reactivity were characterized by a significantcombined test of significance (z = 4.00, p = .00003, fail-safe n= 19.62) and an effect size of r = .28. No test of variability wassignificant (p > .32). Such experimental studies provide strongevidence linking functional aspects of social support to changesin cardiovascular parameters. For example, Gerin et al. hadparticipants engage in a discussion task in the presence or ab-sence of social support. In the no social support condition, twoconfederates attacked the participant's views, while a third con-federate sat silently. The discussion task was identical for thesocial support condition, however, the third confederate now de-fended the participant's views. Results of the study revealed thatsocial support during the conflict discussion was uniformly as-sociated with lower increases in heart rate, SBP, and DBF.

Although the laboratory studies are generally consistent withthe notion that social support may reduce ANS reactivity toacute psychological stress, studies that have operationalized so-cial support by having participants perform a stressful task inthe presence of a relationship assumed to be supportive (e.g.,friend) differ in the pattern of results obtained. In such studies,although the presence of a friend is associated with reducedANS reactivity compared with the presence of a stranger(Edens, Larkin, & Abel, 1992; Snydersmith & Cacioppo,1992), results regarding the presence of a friend compared withan alone condition are inconsistent. Researchers of 7 studiesexamined participants' ANS reactivity to acute psychologicalstress in such studies. Researchers of 3 studies have reportedthat ANS reactivity to brief psychological stress was lower whena friend was present than in the alone condition (Gerin, Milner,Chawla, & Pickering, 1995; Kamarck, Annunziato, & Ama-teau, 1995; Kamarck, Manuck, & Jennings, 1990). In contrast,researchers of 3 studies found no differences between these con-ditions (Edens et al., 1992; Sheffield & Carroll, 1994; Sny-dersmith & Cacioppo, 1992), and K. M. Allen, Blascovich, To-maka, and Kelsey (1991) reported greater ANS reactivity in thefriend condition compared with the alone condition. However,as suggested by Allen et al. (1991), the presence of a friendmay have been associated with increased reactivity due to theevaluative nature of the situation, compared with Kamarck etal. (1990). Consistent with this notion, Allen et al. (1991)found that the friend condition was associated with a faster taskrate and more task errors compared with the alone condition,whereas Kamarck et al. (1990) found no such performancedifferences.

The differences between Kamarck et al. (1990) and Edens et al.(1992) are more difficult to resolve. However, one explanationmay lie in the interactions operating in the alone conditions. Al-though physically alone, participants in these conditions realizedthat their performance was being visually and physiologicallymonitored by the experimenter. All participants in both studies

were women, however, the experimenter in Kamarck et al. was aman, whereas the experimenter in Edens et al. was a woman. Thegender incongruency in the Kamarck et al. study may have con-tributed to increased evaluation apprehension and threat in thealone condition. Consistent with this reasoning, cardiovascular re-activity levels in the alone condition for Edens et al. were lowerthan those reported in the Kamarck et al. study. In fact, as sug-gested by Edens et al., had reactivity levels in the alone conditionbeen comparable with Kamarck et al., these results may have rep-licated Kamarck et al.'s. This interpretation is also consistent withthe results of Snydersmith and Cacioppo (1992) who found resultssimilar to Edens et al. using a female experimenter interacting withfemale participants.

Finally, 2 studies (Gerin et al., 1995; Kamarck et al., 1995)suggest more specific conditions under which a friend is likelyto be associated with lower cardiovascular reactivity comparedwith the alone condition during acute psychological stress. Inparticular, in both of these studies, researchers manipulated thedegree of threat or stress present while the participants per-formed the psychological tasks. Results of these studies revealedthat buffering effects of the friend condition relative to the alonecondition are likely to be found when the situation is especiallythreatening or stressful. In general, the studies summarized inTable 4 provide evidence on the boundary conditions (i.e., de-gree of evaluation and threat) under which a friend may reducecardiovascular reactivity to stress.

The laboratory studies linking social support to phasicchanges in cardiovascular function provide evidence that socialsupport may be important in reducing cardiovascular reactivityto psychosocial stressors. The laboratory studies provide a po-tentially important conceptual link to the studies summarizedin Tables 1-3 that focused on tonic blood pressure in middle-aged to older adult community samples. More specific, socialsupport may be conceptualized as a stable individual differencevariable (I. G. Sarason, Sarason, & Shearin, 1986), therefore,laboratory paradigms may provide a model of how social sup-port operates in everyday life. The laboratory studies suggestthat the higher cardiovascular reactivity seen in situations in-volving low social support may translate to gradual elevationsin tonic blood pressure across the lifespan. In a relevant study,Light, Dolan, Davis, and Sherwood (1992) examined the use ofcardiovascular reactivity to acute stressors in predicting subse-quent tonic blood pressure levels. Results revealed that individ-uals high in heart rate reactivity evidenced elevated tonic bloodpressure 10 to 15 years later compared with individuals low inheart rate reactivity. Significantly, these groups were initiallycharacterized by comparable tonic blood pressure levels. Thesedata are consistent with animal models of hypertension, sug-gesting that exaggerated cardiovascular reactivity influences thedevelopment of cardiovascular disorders (Folkow, Hallback,Lundgren, Sivertsson, & Weiss, 1973; Hallback & Folkow,1974; Manuck, Kaplan, & Clarkson, 1983).

Potential Mechanisms Linking Social Support toReduced Cardiovascular Reactivity

The laboratory studies are suggestive of several differentmechanisms linking social support to reduced cardiovascularreactivity. At a social psychological level, nonevaluative friends


appear to buffer cardiovascular reactivity to highly threateningpsychosocial stressors compared with the stranger or alone con-dition. One should note, however, that most of these studiesused undergraduate samples, and friends may be a particularlyimportant source of support in such younger populations.

These studies are also suggestive of the particular dimensionsor functions of social support that may reduce cardiovascularreactivity in response to certain stressors or situations. In par-ticular, the dimension of emotional support appears to be a sa-lient feature operating in these experiments to reduce cardio-vascular reactivity (Edensetal., 1992;Gerinetal., 1992, 1995;Kamarcketal., 1990,1995;Kiecolt-Glaseretal., 1985;Lepore,1995; Lepore et al., 1993; Snydersmith & Cacioppo, 1992).These findings do not suggest that other aspects of social sup-port (e.g., appraisal support) are unimportant because most ofthe studies in Table 4 have used stressors in which emotionalsupport may be especially important. Appraisal support, for in-stance, cannot be effective if there is no opportunity for discus-sion in these studies. Therefore, consistent with past research(S. Cohen & Wills, 1985; Cutrona & Russell, 1990) and thestudies summarized thus far, the laboratory paradigms suggestthat particular functions of social support are more effectivewhen there is a reasonable match with the demands of particu-lar situations. On the basis of the studies presented in Tables 1 -3, the match between the support dimension and stressor mayalso need to be considered in the context of cultural and gender-related processes.

In many of these studies, data were provided regarding po-tential behavioral mechanisms linking manipulated or natural-istic social support with lowered cardiovascular reactivity. Thestudies in Table 4 suggest that behavioral factors, such as poten-tial health-related variables (e.g., Uchino, Kiecolt-Glaser, &Cacioppo, 1992) and performance indices during the tasks(Edens et al., 1992; Gerin et al., 1992, 1995; Kamarck et al.,1990, 1995; Lepore, 1995; Lepore et al., 1993; Snydersmith &Cacioppo, 1992; Uchino et al., 1992) were not responsible forthe results obtained in these studies. Therefore, consistent withthe correlational studies reviewed earlier, the laboratory studiessuggest that behavioral factors are not necessary for associationsbetween social support and cardiovascular reactivity.

Several studies in Table 4 examined the possibility that spe-cific psychological and emotional reactions during the labora-tory tasks mediated the effects of social support on cardiovas-cular reactivity. These studies' researchers generally reportedthat participants' emotional reactions during the experimentdid not differ as a function of the social support conditions(Edens et al., 1992; Gerin et al., 1992; Kamarck et al., 1990,1995; Lepore et al., 1993; Snydersmith & Cacioppo, 1992;Uchino et al., 1992) and therefore do not appear to be respon-sible for the pattern of results in these studies. In an illustrativestudy, Lepore et al. (1993) had participants perform a speech,while in one condition a confederate made comments designedto reflect emotional support. In contrast, the confederate in thelow emotional support condition was reserved and inattentive.Results indicated that the social support speech condition wasassociated with lower SBP and DBF reactivity than the no sup-port speech condition. In addition, results revealed that theseconditions did not differ in terms of perceived stress during thespeech. Data from 2 additional studies relating naturalistic so-

cial support to cardiovascular indices during acute psychologi-cal stress reveal that results were not a function of depression(Uchino et al., 1992) or environmental stress (Knox, 1993).Whether this lack of mediation by psychological reactions re-flects methodological issues related to the retrospective natureof these assessments or the possibility that other unmeasuredpsychological processes may be important (e.g., feelings of con-trollability and intrusive thoughts) is beyond the scope of theexisting literature. Future research is necessary to elucidate thepotential psychological mechanisms operating in laboratory re-activity paradigms.

Social Support and Endocrine Function

Endocrine literally refers to the internal secretion of biologi-cally active substances. These biologically active substances, orhormones, are usually defined as substances that are releasedfrom an endocrine gland (e.g., pituitary) into the blood streamto act on a distant target tissue site (see Greenspan & Baxter,1994). Hormones can act in either an autocrine (i.e., act on thesame cells that produced them) or paracrine (i.e., act on cellsother than those that produced them) fashion. The major endo-crine glands include the thyroid, pituitary, pancreas, adrenalmedulla, adrenal cortex, testes, and ovaries.

The most commonly examined endocrine measures in stressresearch include the catecholamines (e.g., norepinephrine andepinephrine) and cortisol. Epinephrine (EPI) is produced andreleased into the blood stream from the adrenal medulla. Nor-epinephrine (NE) is also synthesized in the adrenal medullabut is largely produced in the central nervous system and pe-ripheral sympathetic nerves. Catecholamines bind to adrener-gic receptors where their effects are to increase oxygen and heatconsumption and activate glucose and fat from storage areas inthe body. Some important physiologic changes in response tocatecholamines include increased heart rate, increased myocar-dial contractility, changes in vascular resistance, and regulationof renin secretion from the kidneys.

Cortisol is a cholesterol-derived steroid, secreted from theadrenal cortex where it acts through the cytosolic glucocorti-coid receptor protein. Cortisol secretion is a result of complexregulatory processes and involves corticotropin releasing hor-mone (CRH) and ACTH. The biologic process is initiated, inpart, by the release of CRH from the hypothalamus that signalsthe pituitary gland to secrete ACTH. ACTH then signals theadrenal cortex to release cortisol and later serves to regulate theamount of cortisol secreted. Once released, cortisol has a vari-ety of metabolic effects, including increased glucose metabo-lism and down regulation of immune function.

An examination of endocrine function is important becauseof its association with both the cardiovascular and immune sys-tems. As reviewed earlier, the catecholamines play an importantrole in cardiovascular regulation. For instance, catecholaminesbind to alpha-adrenergic receptors and lead to constriction ofarterial smooth muscles. Furthermore, catecholamines appearto be toxic to endothelial cells of the arteries (Krantz & Ma-nuck, 1984). Damage to the endothelial cells of the arteries isthought to be an important initiating factor in the developmentof coronary atherosclerosis (Gorlin, 1976).

Clearly, important interactions exist between the endocrine and


immune systems (Ader, Felton, & Cohen, 1991). Lymphocytesappear to have beta-adrenergjc, glucocorticoid, and opioid recep-tors (Plaut, 1987; Sibinga & Goldstein, 1988). More significantly,hormones influence numerous aspects of immune function. Forinstance, the acute release of catecholamines produce transientincreases in natural killer (NK) cell lysis (i.e., an in vitro model ofthe functional ability of NK cells to lyse or destroy tumor or virus-infected cells; Tonnesen, Christensen, & Brinklov, 1987) and de-creases in the proliferative response to mitogens (i.e., an in vitromodel of the functional ability of lymphocytes to proliferate whenfaced with a challenge; Crary et al., 1983). In addition, glucocor-ticoids have regulatory effects on immune responses, including de-creased NK cell lysis (Gatti et al., 1987) and mitogen proliferation(Rupprecht et al., 1990). Prolactin, growth hormone, estradial,testosterone, opioids, CRH, and ACTH also have immunomodu-latory effects (see reviews by Blalock, 1989; Carr, 1991; and Cupps& Fauci, 1982).

The measurement reliability of commonly examined endocrinemeasures (e.g., catecholamines and cortisol) is well documented.For instance, one of the most accurate and cost-effective tech-niques, radioimmunoassay (RIA), uses specific binding sub-stances (e.g., antibody) with radioactive tracers to quantify endo-geneous hormones in bodily fluids (see Bissette & Ritchie, 1992;Chard, 1982 and Jaffe & Behrman, 1979). RIA has very goodspecificity and sensitivity and is characterized by low intra-assayvariability (i.e., typically less than 1% to 5%). However, the in-terassay variability may vary depending on factors such as the sta-bility of solutions and radioactive tracers, incubation temperatureand duration, and changes in the sample matrix and technicians(Bissette & Ritchie, 1992). Nevertheless, quality control proce-dures (e.g., monitoring standard curve parameters) are typicallyimplemented such that interassay variability below 10% may beachieved (Bissette & Ritchie, 1992).

Little data exist on the temporal stability of interindividualvariations in endocrine function.5 There are a variety of meth-odological issues worth considering in examining the temporalstability of endocrine measures (see reviews by Baum &Grunberg, 1995; and McKinnon, Baum, & Morokoff, 1988).First, there are diurnal variations associated with both catechol-amines and cortisol (Akerstedt & Levi, 1978; Greenspan &Baxter, 1994). Therefore, if only a limited number of samplesare drawn, it is important to collect such samples at approxi-mately the same time to reduce this extraneous and potentialconfounding source of variance. In addition, even if samples areobtained at the same time, the point in the diurnal cycle thatendocrine function is assessed may be important. For instance,cortisol has a diurnal cycle that peaks during the morning hours(Greenspan & Baxter, 1994) so that relationships with socialsupport may be more difficult to detect at this time. Thus, onemay need to examine the relationship between social supportand cortisol across the full range of the diurnal cycle.

Second, it is important to distinguish whether endocrinemeasures are assessed in plasma or urine (Baum & Grunberg,1995). This distinction is especially relevant for the assessmentof plasma catecholamines that have a relatively short half-lifeof 1-3 min. As a result, plasma assessments of catecholaminesprovide data on transient changes in endocrine function. Uri-nary assessments, however, provide information on more

chronic or long-term alterations in endocrine function but arerelatively insensitive to short-term bursts of catecholamines.

Finally, the method of obtaining samples for endocrine as-sessments may contribute significant sources of variance. Ifplasma samples are assessed, blood draws that are obtainedfrom a single needle stick can elevate catecholamines and makerelationships with social support more difficult to detect. There-fore, the use of a catheter with a rest period that allows forchanges due to venipuncture to subside is typically recom-mended (Baum & Grunberg, 1995). The use of urinary sam-ples is associated with different methodological considerations.For instance, compliance with 24-hr urine samples may beproblematic so that 15-hr overnight samples are useful (Baum& Grunberg, 1995).

As evidenced by inspection of Table 5, the association be-tween social support and endocrine function has not been wellstudied. Only 10 studies included endocrine assessments: 4 ex-amined catecholamines, 4 examined cortisol, and 2 examinedboth. In 4 studies, the investigators assessed hormones in urine,in 1 study hormones were assessed in saliva, and in 1 study hor-mones were assessed in breast milk. In addition, 4 studies as-sessed plasma endocrine measures, and only 1 study (Knox etal., 1985) mentioned that they had drawn blood from a catheterthat was inserted before a rest period. As noted earlier, this maybe important because the needle stick associated with a singleblood draw can elevate catecholamines and make relationshipsmore difficult to detect.

In regard to sample characteristics, 4 studies examined onlymen, 2 studies examined only women, and 4 studies used bothmen and women. Social support was also operationalized in sev-eral ways: 6 studies assessed naturalistic social support levels,and 4 studies manipulated social support (i.e., interventionsand laboratory reactivity studies). Of the 6 studies that assessednaturalistic social support, 4 assessed functional aspects of sup-port, 1 assessed both structural and functional aspects of sup-port, and 1 assessed a combined index of both structural andfunctional aspects of support.

Because of the small number of studies and differences in theoperationalization of social support (e.g., naturalistic socialsupport and laboratory manipulations), we performed a quali-tative analyses of the studies whose researchers examined socialsupport and endocrine function. To this point, researchers of 5of the 6 studies reported an association between social supportand catecholamines levels (Arnetz, Edgren, Levi, & Otto, 1985;Ely & Mostardi, 1986; Fleming, Baum, Gisriel, & Gatchel,1982; Knox et al., 1985; Seeman, Berkman, Blazer, & Rowe,1994). In an early study, Fleming et al. examined the influenceof perceptions of support in participants exposed to the chronicstress of Three-Mile Island and in control participants. Resultsrevealed that individuals low on social support were uniformlycharacterized by higher urinary NE levels than individuals highon social support (a similar but nonsignificant trend was foundforEPI).

Of the 6 studies whose researchers have examined cortisol

5 A literature search on PsycLIT( 1974-1994) and Medline( 1983-1994) crossing the keywords endocrine or neuroendocrine with psycho-metrics or reliability revealed no researchers had examined the tempo-ral stability of the endocrine assessments discussed in this review.


levels, researchers of 4 have found no association with socialsupport (Arnetz et al., 1983, 1985, 1987; Groer, Humenick, &Hill, 1994). Two of these studies are interventions (i.e., Arnetzet al., 1983, 1987), and 1 of these studies' data suggest the in-tervention successfully increased social contact (Arnetz et al.,1983). Although the number of studies is small, these data sug-gest that simply increasing social contact may be insufficient toaffect cortisol levels, at least at nomethetic levels of analysis.However, there are important issues that may have prevented anadequate test of the relationship between social support andtonic cortisol levels. First, idiographic analyses of cortisolchanges to brief psychological stressors suggest that suchchanges may be heightened in individuals characterized by highcardiac sympathetic reactivity (Cacioppo et al., 1995). An ex-amination of individual differences, therefore, may provide ad-ditional information about the effects of social support on cor-tisol levels.

In this regard, laboratory paradigms of social support may behelpful to examine phasic cortisol response. In a recent study,Kirschbaum, KJauer, Filipp, and Hellhammer (1995) exam-ined the influence of social support from a stranger or partner(i.e., boyfriend or girlfriend) on cortisol reactivity during acutepsychological stress. Results revealed that partner-supportedmen showed significantly lower cortisol levels than the stranger-supported or no-support conditions. In contrast, womenshowed a trend toward greater cortisol response during the part-ner-supported compared with the other two conditions, eventhough women reported feeling supported by their partners.These results, however, should be considered preliminary andrequire replication.

There is another potential reason why the intervention stud-ies summarized in Table 4 did not find a relationship betweensocial support and cortisol levels. The studies summarized ear-lier in Tables 1-3 on social support and cardiovascular functionsuggest that familial sources of social support may be impor-tant. None of the intervention studies was designed to facilitatefamilial sources of support. In addition, none of these studies'researchers examined the potential generalization of their inter-vention program to such support sources.

Three of the 4 studies that did not find a relationship betweensocial support and cortisol obtained specimens primarily dur-ing the morning hours. One potential problem with a morningassessment is that cortisol has a diurnal cycle that peaks duringthe morning hours (Greenspan & Baxter, 1994). Therefore, thehigher cortisol levels seen during morning hours may make itmore difficult to detect relationships between social supportand tonic cortisol levels. It is therefore promising that Seemanet al. (1994) examined 12-hr urinary cortisol samples and re-ported relationships between social support and cortisol levelsin men. The study by Seeman et al. is also one of the most meth-odologically rigorous because the researchers also statisticallycontrolled for a variety of potential confounding variables (e.g.,age, medication use, and body mass).

Potential Mechanisms Linking Social Support toEndocrine Function

On the basis of Table 5, there appears to be preliminary evi-dence for a relationship between social support and catechol-

amine levels. Researchers of 2 studies conducted under thestress-buffering model found no evidence for a buffering effecton catecholamine levels for participants faced with job termi-nation (Cobb, 1974) or chronically stressed Three-Mile Islandresidents (Fleming et al., 1982). Moreover, both studies' re-searchers found that the stressor was associated with significantelevations in catecholamine levels. Because of the small numberof studies, more data are needed to adequately test the stress-buffering model on catecholamine levels.

In a small number of studies summarized in Table 5, research-ers have examined psychological processes and its relationships tosocial support (Arnetz et al., 1985; Fleming et al., 1982; Kirsch-baum et al., 1995). Fleming et al. found that high social supportwas associated with lower depression and anxiety, along with lowerfeelings of alienation. However, analyses specifically examiningmediating or moderating mechanisms were not conducted be-cause they were not a primary aim of the study.

In the study by Arnetz et al. (1985), social support was con-ceptualized by examining neurotic boys' endocrine responsesto an exciting and partly violent film in either a different school(i.e., low support) or same school (i.e., high support) setting.Arnetz et al. found that the low support condition was associ-ated with greater catecholamine changes in response to the filmthan the high support group. In addition, the low support grouplater preferred to review film scenes representing feeling of se-curity, whereas the high support group preferred to review filmscenes representing danger, suggesting that increased feelings ofanxiety may be one mechanism operating in this study.

The laboratory social support study by Kirschbaum et al.(1995) examined perceptions of support and stress and foundthat these psychological mechanisms could not explain the ob-tained pattern of results. These results are consistent with thosesummarized in Table 4. All in all, more research is needed onthe relationship between social support and endocrine function,along with an examination of psychological and behavioral(e.g., health-related) mechanisms responsible for these associa-tions. Greater attention to methodological issues in examiningthe relationship between social support and endocrine function(e.g., sufficient rest period after needle stick) may also proveuseful.

Social Support and Immune Function

The immune system is the body's defense against infectiousand malignant disease (see Borysenko, 1987; Calabrese, Kling,& Gold, 1987; and Kiecolt-Glaser & Glaser, 1988b, 1995, forbasic reviews). The major organs of the immune system are thethymus, bone marrow, lymph nodes, spleen, tonsils, appendix,and Peyer's patches.

There is no single measure of immune function, and re-searchers in the field of psychoneuroimmunology (PNI) haveused various measures to index aspects of immune function. Adistinction has typically been made between quantitative andfunctional measures of the immune system (Kiecolt-Glaser &Glaser, 1988a). Quantitative measures can include absolutecounts or percentages of certain immune cells such as helper Tcells, suppressor/cytotoxie T cells, and NK cells. Quantitativemeasures are typically examined because both the number and



CTJ

1

?

1jC

t and

End

ocrin

e an

d /

|S1

^3***»•2o

Co

I

c1

3

Eg

u t;•c g

S E•o ^

U3 la

cft! CZ

Ki

c*- S

Soci

al s

uppe

ras

sess

men

t-con

d

.£CJ

•2 C

a S>~>

•1

>>

55

"w>

ed to

NE

le

s«

13

Ifi.a

15•g

s0

Z

wznC

'C

<uaZ

.22 — tr.

Perc

eptio

ns o

f soc

supp

ort p

rimar

ifr

om w

ife b

ut a

l

fII

cI

 1)S c" c , S -a 9 « o 'C «•gS c 7 3^! i 5

?! If fill's-?||clt|i||,||

! fli 111 i|| It i i '5?'^'SS 7 ^ 2 ' — - W Q "I|s|*g||§tl.2 S. >> o .y '$ "v G ' $ ' X ,2

HI'S B I &!'•£ If 1|Ii|'g'SHs^£|J ~ J

S0Z

EtulUZ

«c

5

u§

a

- s s j1 ilh1 3 1 1 f<fc y a S S1

1

fl 1 =~IS*. "HiE S ^ "5 « 2 rt

O A Jj = "o S1 (3

llllill

ii, 1 & §. s ? 1l= -5£ -ll|Mi|l«lf ilIlaMSflE.a-|~« s s g*llHlFbl< CJ S

"rt

w

_ C o o

g C-E

•oca £ u. js

- I - .§ o 1 °

C ' M ! 3 o ! 0 3 c S f c

llgllcl ||u p a | g | §• |

|l'|i-||l If

|.S 1.1 § g & 1.1 1

Z eo co

„oZ

w 2 ttj *c w 21 p 4J-o y Q ™ w- - §^ fc — ^ C •_— ^Ju ** tit c "S 2

s0.03Q0."09tfl

StfI

co —

'*3 H .£2

11(1

O iC_"oo2 cE et- .. S

1 1 " = 1

I a||^7

c

Pros

pect

ive

Asse

ssm

ent p

erio

d(b

asel

ine,

3 m

onth

s, an

d6

mon

ths)

BS fa

ctor

: so

cial

inte

ract

io(s

ocia

l int

erac

tion

and

cont

rol)

• s ?»ri "S wS^'lgi£ic.s

iK

th c

ontr

ols

ip h

ad in

cn

pc u e

o wT3 g;

M

'x *£

X ^

JZ ^1-

X "OX^ :P

o

-.8 -S18

oZ

U 05


8 2

-Si-Q

sO

sO

1 I£•

o

o o o

-3il'l|§iI la a-S E - S

E6

111

C CO

<a

u

Q "**"*"*

E C-

18— _,

.8*

516 UCHINO, CACIOPPO, AND KJECOLT-GLASER

S£

>

'!•! = ~ If 6 Ii

'iaS.SigS1* -S-Saiii^

" ' "

>,6.o g"S vi O r:

4J 43 1>< = : » £

I* I&O Ii oo —

if!K o 3

U T3 C :s r-n « S c JD S'

O 3

i 8fS s,c

leli'a D

+ rfS

|llCu o o

a

<§< 81

o oZ

oZ I o

o o

I!* Jllllfill!!

£,'£

Ij# a

c-a

«^1^ |S«2B

a !;g

I

*>aK

£a

> sfi ov = -R | G-Illl ii -s | s .s* a 11 a

o5

gou

E S

u3(2

IS

w (u

^c t2 —S o o^2 >• C^


clo t

heir

rela

tions

hips

com

pare

d w

ith c

areg

iver

s hi

gh in

NK

CA

to b

oth

rIL-2

and

r!F

N-j>

Car

egiv

ers (

i.e.,

cont

inui

ng a

nd b

erea

ved)

low

in N

KC

A to

bot

h rI

L-2

and

rIFN

-yvi

site

d th

eir p

hysi

cian

s m

ore

than

tw

ice

as o

ften

for i

nfec

tious

illn

ess

sym

ptom

sco

mpa

red

with

car

egiv

ers

high

in

able

con

tinue

Io

o I

supp

ort

ondi

t

-ii!=i!t

- - =

-in!IllHlflL

M-SslI-i-aaex c a g .5 g S 'g .S

2sUl:sI

g gSg'-a ii

asym

ptom

atic

H1V

+ ga

y m

en=

34)

i2 Ov

.Is'K-s9ma< II+ 35> o *-X - S g- S S

| S i » g• ii g E;^t s

§O

|c

£ < O 03

•sa

II

.ss;~C CT\s c£

518 UCHINO, CACIOPPO, AND K1ECOLT-GLASER

»3f

iii iiiri 15Is,3 -

s*l

£ll«:ill?2sla 8.13 S a3

11 nil HI<

isi"S.5~

= 811

ilfiilS-B S-O?sg.|3"§iS

ll E g I S 1

•as-

£ s« * - S u '?


relative balance of immune cells (e.g., ratio of helper T cellsto suppressor/cytotoxic T cells) are important in mounting aneffective immune response (Herbert & Cohen, 1993b). Func-tional measures examine the performance of certain immunecells. One common measure is the blastogenic response of lym-phocytes to the plant mitogens concanavalin A (Con A) andphytohemmaglutinnin (PHA). Blastogenesis provides an in vi-tro model of lymphocyte proliferation in response to antigens.In general, greater proliferation is interpreted as a better im-mune response. In addition, measures of NK cell lysis are takenby incubating NK cells with a tumor cell line and examiningthe ability of NK cells to lyse (i.e., destroy) tumor cells. NKcells appear to play an important role in the body's defenseagainst virally infected and malignant cells (Whiteside & Herb-erman, 1994).

The measurement reliability of immunological data variesacross assays and laboratories. Issues such as the day on whichan individual's immunological data are analyzed can contributesignificant variance for some assays. Considerable work has beendone to ensure the measurement reliability of some assays acrosslaboratories and times (e.g., percentages of T lymphocytes andsubpopulations; Gelman, Cheng, Kidd, Waxdal, & Kagan, 1993;Paxton et al., 1989; Schenker et al., 1993). Similarly, studies per-formed after implementation of quality control procedures to es-tablish reliability and reduce daily variability of NK cell lysis showthat NK activity is also characterized by adequate measurementreliability (Whiteside & Herberman, 1994). However, researchersin only a few studies have examined the temporal reliability ofcommon measures of immune function. Researchers reporteddata on the 3-week temporal stability of several quantitative andfunctional immune measures in response to a stressful speech task(Marsland, Manuck, Fazzari, Stewart, & Rabin, 1995). Marslandet al. found that most baseline quantitative measures of immunity(e.g., CD4+ and CD8+ cells) evidenced significant temporal sta-bility (.22 < r <• .75). Moreover, immune responses to acute stresswere also associated with significant temporal stability for mostquantitative measures (.25 < r <, .53), as well as changes in theproliferative response to PHA (r = .50, p < .005).

According to psychometric theory, several assessments of anindividual across times and situations are likely to provide amore accurate individual difference assessment of immunefunction. Consistent with this possibility, Fletcher, Klimas,Morgan, and Gjerset (in press) found that the proliferative re-sponse to PHA, pokeweed mitogen, and NK cell lysis were char-acterized by adequate generalizability (G) coefficients (Gs >.70), which increased when assessments were aggregated acrosstimes (Gs 21.85). Unfortunately, the high costs of immunolog-ical assays make repeated determinations impossible in manycircumstances. Although the data obtained by Marsland et al.(1995) and Fletcher et al. suggest that measures of immune re-sponse are characterized by good temporal reliability, addi-tional data are needed to examine an individual difference as-sessment of immune function.6

There are additional issues involved in the interpretation ofmeasures of immune function that may be useful to discuss. Asmentioned earlier, there is no single generally accepted measureof immune function. As a result, the interpretation of a singlemeasure of immune function as representing a down regulationmay be difficult to determine. Studies that measure multiple

aspects of immune function would be in a stronger position tomake such conclusions by examining the pattern of immunechanges across assays.

We found 19 studies whose researchers examined the associ-ation between social support and aspects of immune function(see Table 5) .7 Eight studies used only functional assays, 6 stud-ies only quantitative assays, and 5 studies both quantitative andfunctional immunological measures. Nine studies were per-formed with middle-aged or older adult samples (M age > 40).Nine studies tested both men and women, 6 only women, and 4only men. In addition, 10 studies were correlational, 7 prospec-tive, and 2 experimental interventions. Of the 17 studies whoseresearchers assessed naturalistic social support, only 4 included,in part, structural measures of support.

The results of the 19 studies summarized in Table 5 are con-sistent with the notion that higher social support is associatedwith better immune system function (e.g., greater NK celllysis). To summarize, 12 studies' researchers reported evidencethat social support was associated with aspects of immune func-tion, and 7 reported no relationship. A meta-analysis of 9 stud-ies whose researchers examined the association between socialsupport and functional immune measures revealed a reliablecombined test of significance (z = 4.38, p = .000006, fail-safe n= 54.90). The effect size associated with these studies was r =.21, and no test of variability was significant (p > .49) .*

6 A literature search on PsycLIT( 1974-1994) and Medline (1983-1994) crossing the keywords immune or immunology with psychomet-rics or reliability revealed no additional researchers who had examinedthe temporal stability of the immune assessments discussed in thisreview.

7 Our initial search produced 24 studies whose researchers examinedthe relationship between social support and immune function. We ex-cluded analyses of 6 studies in part or in whole for several reasons. Dataon the relationship between social support and white blood cell or totallymphocyte counts were not included in the present review (Arnetz etal., 1983; R. S. Baron, Cutrona, Hicklin, Russell, & Lubaroff, 1990;Mclntosh, Kaplan, Kubena, & Landmann, 1993; Thomas, Goodwin,& Goodwin, 1985) because of difficulties in interpreting the signifi-cance of these measures. In addition, we excluded 1 study that exam-ined salivary IgA (Jemmott & Magloire, 1988) because of methodolog-ical issues regarding the reliability of the salivary IgA assessment thatwas used in this study (Herbert & Cohen, 1993a; Stone, Cox,Valdimarsdottir, & Neale, 1987). One study was also excluded from thereview because the small number of participants (3) in their low stressand no social support cell precluded definitive analyses (Herrera, Alva-rado, & Martinez, 1988).

8 The meta-analysis consisted of 9 studies whose researchers directlyexamined the association between social support and functional mea-sures of immune function. To reduce problems associated with nonin-dependence, the results of Snyder, Roghmann, and Sigal (1990, 1993)were averaged to produce one test statistic because data were reportedfrom the same sample. Six studies were excluded from the meta-analysisbecause it was determined a priori that intervention studies and popu-lations with individuals having HIV were associated with inconsistenteffects. These studies are discussed in detail later in the text.

In the meta-analysis, we examined functional immune measures be-cause there were only 4 remaining studies whose researchers examinedthe association between social support and quantitative measures of im-mune function. The interested reader is referred to Table 5 for a sum-mary of the relationship between social support and quantitative im-mune measures. In addition, there were only 2 remaining studies that


Two of the studies that did not find an association betweensocial support and immunological data were interventions de-signed to facilitate social interactions (Arnetz et al., 1987; Kie-colt-Glaseretal., 1985). The Kiecolt-Glaser et al. (1985)inter-vention documented increased social contact and interactions.However, keep in mind that the sample sizes in 2 of these studiesare relatively small (n < 45.00). In addition, we reviewed evi-dence earlier suggesting that familial sources of support maybe important. It is possible that interventions aimed directly atincreasing familial contact and support may yield even largereffects sizes on immune function.

Two of the studies that did not find a significant associationbetween social support and aspects of immune function exam-ined men who were HIV+ (Goodkin et al., 1992; Perry, Fish-man, Jacobsberg, & Frances, 1992). There are important meth-odological reasons that might explain a lack of an associationbetween social support and immune function in individualsHIV+. For instance, stage of disease, age, gender, drug abuse,and health behaviors are potentially important confoundingvariables (Ironson et al., 1994).

However, 2 recent studies have reported an association be-tween social support and CD4+ counts (a marker of HIVprogression) in men with HIV (Persson, Gullberg, Hanson,Moestrup, & Ostergren, 1994; Theorell et al., 1995). In 1 pro-spective study with data across a 5-year period, Theorell et al.found that the availability of social and emotional support pre-dicted subsequent changes in CD4+ counts in a representativeSwedish sample of men with HIV. Results revealed that highand low social support groups did not differ in CD4+ countsduring the early years of the study. However, the prediction ofCD4+ counts as a function of social support was evident duringYears 4 and 5 of the study. For instance, during Year 5 of thestudy, individuals high in social support showed a —37% changein CD4+ counts, whereas individuals low in social supportshowed a -64% change in CD4+ counts (Theorell et al., 1995).Note that Perry et al. (1992) reported null findings on the rela-tionship between social support and CD4+ counts only up toYear 1 of their study. These preliminary prospective data sug-gest that social support may influence the progression of HIVinfection and provide evidence on the utility of such long-termprospective designs. An important implication of these data isthat if a researcher was to only examine the relationship be-tween social support and CD4+ counts later in the stage of dis-ease, information on the longer length of time that individualshigh in social support took to get to that stage would be lost.9

A population of particular interest in this review is olderadults because social support may be especially important forthese individuals (House et al., 1988). Alterations in immunefunction may have significant consequences in this populationas aging is associated with a down regulation of immune func-tion (Goidl, 1987; Goodwin, Searles, & Tung, 1982; Roberts-Thomson, Whittingham, Youngchaiyud, & Mackay, 1974;Schleifer, Keller, Bond, Cohen, & Stein, 1989), and infectiousillnesses are the fourth leading cause of death in the elderly

included a structural assessment of support. Therefore, focused com-parisons between structural and functional measures of support werenot performed.

(Effros & Walford, 1987). Excluding the intervention studiesdiscussed earlier, it is important to note that the association be-tween social support and functional measures of immunity isconsistent in older adults. A meta-analysis of 7 studies in mid-dle-aged to older adult populations confirmed this hypothesis (z= 4.27, p = .000009, fail-safe n = 40.30), with an effect size ofr= .23. No test of variability was significant (p> .35).

Although 9 studies' researchers examined both men andwomen, 8 studies' researchers did not report analyses aimed atexamining potential gender differences. In the only study withdata on potential gender differences, Thomas, Goodwin, andGoodwin (1985) found that the availability of a confidant wasassociated with a stronger proliferative response to PHA forwomen but not men. However, the correlations were in the samedirection, and no statistical test was performed to directly testthe difference between men and women. In addition, Thomas etal. provided a conservative test of the effects of social supporton immune function, as they statistically controlled for psycho-logical distress as well as potential health-related variables (e.g.,alcohol consumption).

Similar conceptual issues exist in the research examining so-cial support and immune function as in the research reviewedearlier. Only 2 of the studies summarized in Table 5 conceptu-alized social support as a multidimensional construct and re-ported analyses regarding a relatively specific dimension of so-cial support (R. S. Baron et al., 1990; Persson et al., 1994).Although 3 additional studies used multidimensional socialsupport measures, results were only reported on the total scale(e.g., Glaser, Kiecolt-Glaser, Bonneau, Malarkey, & Hughes,1992; Goodkin etal., 1992; Perry etal., 1992). An examinationof relatively distinct dimensions of social support may haverevealed greater specificity (Glaser et al., 1992) and strongerassociations between social support and immune function(Goodkin et al., 1992; Perry et al., 1992) due to a better matchbetween the sample needs and the support resource.

Potential Mechanisms Linking Social Support toImmune Function

The studies summarized in Table 5 suggest that social sup-port is associated with better immune function. These resultsand a recent meta-analysis conducted by Herbert and Cohen(1993a) provide converging evidence for the effects of socialsupport on physiological function. However, Herbert and Co-hen only examined aspects of immune function and focused onsocial stressors involving the loss or disruption of interpersonalresources (e.g., bereavement and marital conflict).

As in the review of social support and cardiovascular function(see Tables 1-3), the studies summarized in Table 5 suggestthat close relationships, such as familial ties, may be a particu-larly important source of social support. Researchers of 2 stud-ies assessed social support specific to close relationships (Levy

9 We should note that a typical strategy in such designs might be tostatistically control for the length of time since illness. However, if theinteraction between this covariate and social support were significant, itwould invalidate the use of this statistical control procedure (J. Cohen& Cohen, 1983) but accurately reflect the findings of Theorell et al.(1995).


et al., 1990; Thomas et al., 1985), and both found social sup-port to be related to aspects of immune function, including astronger proliferative response to PHA and greater NK cell lysis.

Researchers of 4 of the studies in Table 5 directly tested thebuffering model of social support (Goodkin et al., 1992; Kie-colt-Glaser, Dura, Speicher, Trask, & Glaser, 1991; Snyder,Roghmann, & Sigal, 1990, 1993). Of these studies, only Kie-colt-Glaser et al. (1991) reported a reliable effect of stress onimmune function. As noted earlier, S. Cohen and Wills (1985)argued that this is a methodological requirement for an ade-quate test of the buffering model. More important, Kiecolt-Gla-ser et al. found evidence for a buffering effect of social supporton immune function. However, because of the small number ofstudies, more data are needed to adequately test the bufferingmodel on immune function.

Two of the studies summarized in Table 5 conceptualized so-cial support as a multidimensional construct and reported anal-yses regarding relatively specific dimension of social support(R. S. Baron et al., 1990; Levy et al., 1990). Levy et al. exam-ined the dimension of emotional support from a spouse (or in-timate other) and emotional support from one's doctor andfound both to be associated with greater NK cell lysis in cancerpatients. In a study of spouses of cancer patients, Baron et al.used the social provisions scale and found that higher levels onall support dimensions (i.e., guidance, reliable alliances, reas-surances of worth, social integration, attachment, and opportu-nity for nurturance) were equally and significantly associatedwith a stronger proliferative response to PHA and greater NKcell lysis. As noted by Baron et al., caring for a spouse withcancer may result in a mobilization of one's support network,such that there was little differentiation among support compo-nents. Consistent with this possibility, Baron et al. reported highintercorrelations among the components of support.

Levy et al. (1990) suggest that emotional support may be onedimension of social support that is associated with immunefunction. Researchers of 4 additional studies also assessed, inpart, emotional support (also see R. S. Baron et al., 1990). Kie-colt-Glaser et al. (1991) and Esterling, Kiecolt-Glaser, Bodnar,and Glaser (1994) used a composite index of emotional andtangible support. Snyder et al. (1990, 1993) used a compositeindex of emotional and informational support. A meta-analysisof these studies revealed a significant combined test of signifi-cance (z = 4.02, p = .00003, fail-safe n = 24.90). The effect sizeassociated with this test was r = .26, and no test of variabilitywas significant (p > .44). These data suggest that emotionalsupport may be at least one important aspect of social supportin predicting immune function. Additional research is needed,however, that directly compares the predictive utility of specificdimensions of social support.

At a behavioral level of analysis, part of the association be-tween social support and immune function may be due to theireffects on potential health-related variables (Kiecolt-Glaser &Glaser, 1988b). Several researchers assessed the effects of po-tential health-related behaviors and found that the associationsbetween social support and immune function were significanteven when statistically controlling for health practices (Thomaset al., 1985; also see Theorell, Orth-Gomer, & Eneroth, 1990).These data are consistent with results reviewed earlier on socialsupport and blood pressure, suggesting that such behaviors do

not appear to be necessary for an association between socialsupport and immune function. However, these findings shouldbe taken as preliminary, given the restricted number of health-related practices assessed and the lack of reported data on thevalidity and reliability of such assessments in many of thesestudies.

Psychological factors such as levels of stress and depressionhave reliable effects on immune function (Herbert & Cohen,1993a, 1993b). Therefore, part of the association between so-cial support and immune function may be mediated by thesefactors. Researchers of 3 studies in Table 5 reported data relat-ing to potential psychological mechanisms responsible for theassociations between social support and immune function(R. S. Baron et al., 1990; Glaser et al., 1992; Kiecolt-Glaser etal., 1991). Baron et al. (1990) found that the associations be-tween social support and immune function were not mediatedby life events. In addition, Baron et al. (1990) and Kiecolt-Gla-ser et al. (1991) found that depression was not mediating theassociations between social support and immunity. Finally, Gla-ser et al. (1992) reported evidence indicating that anxiety levelswere not responsible for the associations between social supportand immune function. Therefore, although health-related be-haviors, depression, and life stress have reliable effects on as-pects of immune function, these factors do not appear to bemajor pathways explaining the associations between social sup-port and immune function.

Discussion

Social support has been linked to lower rates of morbidityand mortality from diverse disease processes and endpoints.Therefore, the major aims of this review were to examine theevidence linking social support to multiple aspects of physio-logical function and to characterize the potential mechanismsresponsible for these covariations. To the best of our knowledge,this is the first comprehensive review on this topic. The presentreview indicates that there is relatively strong evidence linkingsocial support to aspects of the cardiovascular, endocrine, andimmune systems. These data are consistent with research sug-gesting that the formation and disruption of social relationshipshave important immunological and endocrinological sequalaein nonhuman primates and humans (Coe, 1993;Gunnar, 1992;Herbert & Cohen, 1993a). More important, the physiologicalsystems reviewed may play important roles in the leading causesof death in the United States, including cardiovascular disor-ders, cancer, and respiratory illnesses.10 Conceptual and meth-

10 C. E. Smith, Fernengel, Holcroft, Gerald, and Marien (1994) con-ducted a meta-analysis on the effects of social support on various healthmeasures, including physical and stress-related outcomes. They opera-tionalized stress outcomes as reports of negative life events, conflict ordistress, and laboratory measures, such as catecholamine levels. Physi-cal health status was operationalized as subjective states, such as symp-toms and signs, and as objective data, such as weight loss, activities ofdaily living, blood pressure, blood glucose, and reports of sexual activityposthysterectomy. The results of the meta-analyses revealed effect sizeestimates ranging from .01 to .22. Smith et al. concluded that the rela-tively small effect sizes suggest that the relationship between social sup-port and health may not be significant or generalizable.

There are several issues that warrant discussion regarding the C. E.Smith et al. (1994) meta-analysis. First, Smith et al. did not present any


odological issues were also raised regarding the associations be-tween social support and physiological processes. We now turnto these issues.

One basic and recurring issue in the social support literaturerelates to the measurement of social support (Barrera, 1986; S.Cohen & Wills, 1985; Heitzmann & Kaplan, 1988; Orth-Gomer&Unden, 1987; Tardy, 1985; Winemiller et al., 1993).Questions remain about the factor structure of social supportand the temporal stability and psychometric properties of suchassessments. In the present review, few studies' researchers re-ported data pertaining to the psychometric properties of theirsocial support measures. Given the heterogeneity in the mea-sures of social support covered in this review, psychometric datamay help clarify reliable relationships.

An additional measurement issue concerns the specificsources of social support. The present review indicates that fa-milial ties appear to be an important source of social supportto consider in studies of physiological function. Social supportresearchers might gain greater specificity and prediction by ex-amining specific types of social relationships. The studies sum-marized in this review whose researchers examined cross-cul-tural and gender effects of social support are examples of suchapplications. In addition, behavioral data obtained during lab-oratory studies may amplify the relationships found betweenself-report data and physiological processes (e.g., Kiecolt-Glaser et al., 1993; Malarkey, Kiecolt-Glaser, Pearl, & Glaser,1994).

Most of the studies reviewed in this article have conceptual-ized social support as a unidimensional construct. As noted ear-lier, multidimensional assessments may allow for an examina-tion of more specific associations and mechanisms (Uchino,Cacioppo, Malarkey, Glaser, & Kiecolt-Glaser, 1995). For in-stance, Seeman et al. (1994) found that emotional support wasa more consistent predictor of neuroendocrine function thaninformational support. More important, such specificity wouldhave been lost if an aggregate measure of social support wasused. The relative importance of specific dimensions of social

data on combined tests of significance for the relationships that theyexamined. Therefore, some of the relationships between social supportmay have been statistically significant, albeit characterized by small tomoderate effect sizes. Second, the aggregation of such diverse measuresas indices of physical health and stress outcomes (see above) may ob-scure reliable relationships within particular measures (e.g., blood pres-sure and catecholamines). Finally, and perhaps most important, theselection of studies in the meta-analysis was limited. None of the majorprospective studies on social relationship and mortality were includedin the meta-analysis (see House et al., 1988, for a review). In addition,there appears to be little overlap in the studies examined in Smith et al.and the present review. The reasons for this discrepancy is unclear be-cause similar databases were searched. Unfortunately, Smith et al. didnot report the keywords used for their search. We reviewed 81 studiesfocusing on the physiological mechanisms that may be underlying theassociation between social support and health, whereas the Smith et al.review only included 61 studies, but they also examined other diverseoutcomes (i.e., psychological, stress-related, and physical outcomes).Therefore, it appears that our review on the physiological processes un-derlying the associations between social support and health was morecomprehensive than the Smith et al. review.

support was difficult to characterize in the present review, butconsistent with Seeman et al. there was evidence that emotionalsupport has reliable effects on physiological function. Given theconceptual advantages to a multidimensional approach, futureresearchers might profitably use such an approach and reportdata on both specific dimensions as well as overall levels of so-cial support.

An important issue examined in the present review relates tothe potential psychological mechanisms linking social supportto physiological function. A large literature has documentedpsychological antecedents and consequences of social support(e.g., B. R. Sarason, Sarason, & Pierce, 1989). However, thepresent review indicates that most studies' researchers have notreported data on such potential mechanisms, and when suchdata were presented, appropriate statistical procedures for ex-amining mediation and moderation were typically not per-formed (see R. M. Baron & Kenny, 1986). One interesting pre-liminary observation is that although health-related behaviors,stressful events, and depression clearly influence physiologicalprocesses in their own right, they do not appear to be majorpathways by which social support influences physiological func-tion. Future researchers might attempt to characterize the po-tential psychological mechanisms responsible for the associa-tions between social support and physiological processes basedon the larger social support literature (see S. Cohen, 1988).

Prior research on social support processes has been focusedon support as a situational variable, however, social supportmay also be conceptualized as a stable individual differencesvariable (I. G. Sarason et al., 1986). It is therefore possible thatmeasures of social support may be influenced by personalityfactors that are correlated with social relationships. As a result,social support may be associated with physiological functionthrough personality processes that influence perceptions of sup-port (Bolger & Eckenrode, 1991). However, consistent with re-search indicating that social support predicts objective healthoutcomes, even after statistically controlling for personality fac-tors (House et al., 1988), social support appears to predictphysiological function even after statistically controlling for theeffects of personality variables or their confounding mechanism(e.g., Kiecolt-Glaser et al., 1991; Uchino et al., 1992; Uchino,Kiecolt-Glaser, & Cacioppo, 1994). Note that personality fac-tors such as neuroticism and extraversion appear to have alarger effect on more subjective measures of health, such as self-reported symptomology (Watson & Pennebaker, 1989). How-ever, the personality dimension of trait hostility is related to ob-jective measures of health (Barefoot, Dahlstrom, & Williams,1983; Shekelle, Gale, Ostfeld, & Paul, 1983) and appears tohave an important interpersonal component (T. W. Smith &Frohm, 1985; T. W. Smith, Pope, Sanders, Allred, & O'Keeffe,1988). Future researchers could profitably examine the in-terface between social relationships and personologic factors,especially in reference to physical health (McGonigle, Smith,Benjamin, & Turner, 1993; T. W. Smith, in press).

In the present review, we have focused on the positive aspectsof social relationships. There is a growing body of literature in-dicating that negative aspects of social relationships are inde-pendent of positive aspects of support (Ruehlman & Karoly,1991) and an important predictor of psychological functioning(Pagel, Erdly, & Becker, 1987; Rook, 1984; Rook & Pietromo-


naco, 1987). In addition, in their meta-analysis Herbert andCohen (1993a) suggest that such negative aspects of social rela-tionships may have important effects on immune function.Therefore, the assessment of both positive and negative aspectsof social relationship by future researchers might prove worth-while. In a recent study demonstrating the use of examiningboth positive and negative aspects of social relationships, Lep-ore (1992) found that conflict with roommates or friends was asignificant predictor of psychological distress. More important,positive aspects of support served to attenuate this association.The separability of positive and negative components is an im-portant and emerging conceptual theme in several areas of psy-chological inquiry (Cacioppo & Beratson, 1994; Kiecolt-Glaseret al., 1993; Watson, Clark, & Tellegen, 1988).

An important issue in the present review is an examinationof the psychometric properties of the physiological assessmentsbecause they bear on the potential mechanisms linking socialsupport to long-term physical health. Measurement reliabilitytends to be high in the assessment of cardiovascular, endocrine,and immune function. Furthermore, prior research has docu-mented the temporal stability of the cardiovascular assessments(e.g., Manuck, 1994). However, little data currently exist onthe temporal stability of assessments of endocrine and immunefunction (cf. Fletcher et al., in press; Marsland et al., 1995). Inthis regard, the lack of prospective studies examining the linkbetween social support and physiological function is notewor-thy. Prospective studies may prove helpful in examining thetemporal stability and association between social support, phys-iological processes, and potential mechanisms over appreciableperiods of time.

The effects of social support appear most reliable on bloodpressure, catecholamines, and aspects of both the cellular andhumoral immune response. The laboratory reactivity studieswith young individuals suggest that social support may lowercardiovascular reactivity to psychosocial stressors (e.g., Gerinet al., 1992; Lepore et al., 1993). Cardiovascular reactivity toacute psychological stressors is thought to reflect, in part, sym-pathetic-adrenergic activation (Light, 1981). In addition,exposure to acute psychological stressors appears to activate thesympathetic adrenal medullary (SAM) system as evidenced byincreased plasma catecholamines levels (Cacioppo et al., 1995;Manuck, Cohen, Rabin, Muldoon, & Bachen, 1991; Sgoutas-Emch et al., 1994). Therefore, one model of these data suggeststhat social support may be related to long-term health througha common sympathetic-adrenergic mechanism.

We also reviewed evidence on the association between socialsupport and tonic blood pressure levels. These studies testedmiddle-aged to older individuals and suggested that social sup-port was associated with lower blood pressure levels. As notedearlier, heightened cardiovascular reactivity in the young maytranslate to gradual elevations in tonic blood pressure levelsover the years (e.g., Light et al., 1992). This developmental se-quence involving mediation by a sympathetic-adrenergicmechanism is consistent with animal models of hypertension(Folkow et al, 1973; Hallback & Folkow, 1974).

Sympathetic-adrenergic activation may not only have im-plications for the development of cardiovascular disorders butmay also prime significant coronary events in diseased popula-tions (Kamarck & Jennings, 1991; Krantz et al, 1991; Rozan-

ski et al, 1988). For instance, Krantz et al. (1991) examinedstress-induced myocardial ischemia in coronary disease pa-tients. Results revealed that patients with more severe myocar-dial ischemia during mental stress were characterized by greaterSBP reactivity (also see Rozanski et al, 1988). These data areconsistent with research indicating that social support also pre-dicts survival after the diagnosis of cardiovascular disorders(e.g., Berkman, Leo-Summers, & Horwitz, 1992).

Activation of a sympathetic-adrenergic mechanism may alsoinfluence immune system functioning. Direct sympathetic in-nervation exists for both primary and secondary lymphoid or-gans (see review by Felten & Felten, 1991), and lymphocyteshave beta-adrenergic receptors (Khan, Sansoni, Silverman, En-gleman, & Melmon, 1986; Williams, Snyderman, & Lefkowitz,1976). More significantly, catecholamines appear to influenceaspects of the cellular and humoral immune response (Madden& Livnat, 1991; Sanders & Munson, 1985). In addition, labo-ratory studies suggest that short-term alterations in immunefunction to acute psychological stress are largest in individualsshowing greater cardiac-sympathetic reactivity (Cacioppo etal, 1995).

It is important to separate the short-term and longer termeffects of sympathetic activation because they may have differ-ent implications for the effects of social support on the etiologyof disease progression. Although the data on this issue are lim-ited, chronic stress may be associated with increased sympa-thetic activity (Baum, 1990; Irwin et al, 1991) and a down reg-ulation of beta-adrenergic receptors (Dimsdale, Mills, Patter-son, Ziegler, & Dillon, 1994). Such changes appear to beassociated with decrements in immune function (Irwin et al,1991). For instance, Irwin et al. (1991) found elevated levels ofa sympathetic neurotransmitter, neuropeptide Y, in caregiversof patients with Alzheimer's disease compared with matchedcontrols and reported that levels of neuropeptide Y were nega-tively related to NK cell lysis. Given the long-term nature ofchronic stress, the repeated activation of a sympathetic-adren-ergic mechanism may have relatively long-term effects on theimmune system.

In addition to the evidence indicating the importance of sym-pathetic-adrenergic influences, glucocorticoids have importanteffects on many aspects of immune function (Cupps & Fauci,1982). As reviewed earlier, there are important reasons whyprior studies may not have found a relationship between socialsupport and tonic cortisol levels. In addition, cortisol reactivityin the form of "acute bursts" may be important in explainingvariance between social support and health. Consistent withthis reasoning, preliminary data from our laboratory suggestthat cortisol reactivity to acute psychological stress predictedolder adult reactions to an influenza vaccination (i.e., influenzavirus induced interleukin-2 levels in vitro) given 7 to 8 monthsearlier (see Cacioppo, 1994). To date, little idiographic data ex-ist on the influence of social support on cortisol reactivity toacute psychosocial stressors. In addition, an examination of re-ceptor functioning may prove helpful in obtaining an integratedunderstanding of not only the association between social sup-port and hormones of the hypothalamic pituitary-adrenal axis(Seeman & Robbins, 1994) but also hormones of the SAM(Mills & Dimsdale, 1993).

Future research will help to understand the potential impor-


tance of examining specific patterns of sympathetic-adrenergicactivity (Cacioppo et al, 1992). For instance, research on neuro-endocrine-immune interactions during acute psychological stresssuggests that cardiac-sympathetic reactivity, as assessed by pre-ejection period, appears to be closely linked to activation of theHPA, whereas catecholamines do not appear to operate throughsuch a mechanism (Cacioppo et al,, 1995; Manuck, Cohen,Rabin, Muldoon, & Bachen, 1991; Sgoutas-Emch et al., 1994;Uchino, Cacioppo, Malarkey, & Glaser, 1995). Research designsthat incorporate assessments across the cardiovascular, endocrine,and immune systems may help elucidate more specific pathwaysresponsible for the effects of social support across physiologicalprocesses. To this point, it should be noted that of the 81 studiesreviewed in this article, only 5 obtained simultaneous measuresfrom two of the different physiological systems (e.g., cardiovascu-lar and endocrine) and that none of the studies obtained measuresacross all three systems. It is also noteworthy that there is a sur-prising dearth of research on social support and neuroendocrineprocesses in humans. The neuroendocrine system almost cer-tainly serves as an important gateway between personal relation-ships and health; even daily psychological stressors can provokethe release of pituitary and sympathetic adrenomedullary hor-mones that have multiple effects, including alterations in the car-diovascular and immune systems (e.g., Cacioppo et al., 1995; Ma-larkey etal., 1994).

We should note that unidimensional conceptualizations ofphysiological processes may be at least as significant an ob-stacle to illuminating the mechanisms underlying the healthbenefits of social support, as are unidimensional conceptual-izations of social support. Cardiovascular reactivity, for in-stance, has tended to be treated as a unidimensional (and oc-casionally as a unidirectional) construct ranging from low tohigh, reflecting individual differences in adrenergic reactivityto daily stressors and behavioral challenges. Although adren-ergic activity exerts predominant control over the vasculatureand cardiac inotropy, cardiac chronotropy (and, thus,cardiac output) is a joint function of sympathetic and vagalactivity. Moreover, vagal as well as sympathetic influences onthe heart are evident both at rest and in response to dailytasks, challenges, and stressors (e.g., M. T. Allen & Crowell,1989; Grossman, Stemmler, & Meinhardt, 1990; see reviewsby Cacioppo, 1994; and Forges, 1992), and the vagal andsympathetic outflows to the heart can vary reciprocally, non-reciprocally (e.g., coactivation), or independently (Berntson,Cacioppo, & Quigley, 1991, 1993; Cacioppo, Uchino, &Bernstson, 1994). An individual's classification as high incardiovascular reactivity ignores possible individual differ-ences in the autonomic origins of this reactivity. Variationsin the autonomic origins of cardiovascular reactivity havegenerally been relegated to the error term, a practice that hasobscured relationships between autonomic, endocrine, andimmune responses and may obscure relationships betweenautonomic processes and social, behavioral, or health out-comes.

As the case for heart rate, similar conceptual issues relateto the assessment of blood pressure. Blood pressure is a func-tion of both cardiac output and peripheral resistance. Recentresearch has demonstrated considerable individual differ-ences in the underlying mechanisms of blood pressure re-

sponses (Kasprowicz, Manuck, Malkoff, & Krantz, 1990;Sherwood, Dolan, & Light, 1990). A specifiable subset of in-dividuals achieve blood pressure changes primarily throughvascular mechanisms (i.e., peripheral resistance), whereas adifferent subset of individuals achieve comparable bloodpressure changes primarily through hemodynamic mecha-nisms (i.e., cardiac output). The impedance cardiograph(see Sherwood, Allen, et al., 1990) provides noninvasive esti-mates of cardiac output and total peripheral resistance andshould prove useful in future investigations of the underlyingbasis of both tonic and phasic blood pressure.

We have focused on the cardiovascular, endocrine, and im-mune systems as potential windows through which to examinethe long-term health consequence of social relationships. Theintervention studies examining hypertensive patients providedata to show how social support could have important healthconsequences. Animal models have demonstrated the predic-tive utility of cardiovascular reactivity in the development ofcardiovascular disorders (Folkow et al., 1973; Hallback & Fol-kow, 1974; Manuck et al., 1983). Less is known, however, aboutthe long-term effects of heightened cardiovascular reactivity inhumans, although supportive evidence is mounting (e.g., Lightet al., 1992; Manuck, 1994). Research on social support in at-risk populations and continued theoretical development in thereactivity literature may help to clarify the link between cardio-vascular function and physical health.

Similarly, animal models have demonstrated the importanceof alterations in the endocrine and immune systems on infec-tions and tumor growth (Habu, Akamatsu, Tamaoki, & Oku-mura, 1984; Lewis et al., 1983; Sheridan, Feng, Bonneau, Ma-larkey, & Hughes, 1991). In addition, some studies have linkedalterations in immune function with various health conse-quences in humans (Cogen, Stevens, Cohen-Cole, Kirk, & Free-man, 1982; Fletcher, Baron, Ashman, Fischl, & Klimas, 1987;Kiecolt-Glaser & Glaser, 1995; Lumio, Welin, Hirvonen, & We-ber, 1983; Murasko, Weiner, & Kaye, 1988), including cancer,infectious illness, and HIV progression. Future studies on socialsupport and physiological processes that include measures ofphysical health status would be helpful in clarifying the links toactual health outcomes.

As noted by House et al. (1988), the effects of social supporton health may be particularly important for older adults:

Changes in marital and childbearing patterns and in the age struc-ture of our society will produce in the 21st century a steady in-crease of the number of older people who lack spouses or chil-dren—the people to whom older people often turn for relatednessand support. Thus, just as we discover the importance of socialrelationships for health, and see an increasing need for them, theirprevalence and availability may be declining, (p. 544)

Chronological aging is typically associated with changes in thecardiovascular (Uchino et al., 1992) and endocrine systems(Meites, Goya, & Takahashi, 1987), a down regulation of theimmune system (Burns & Goodwin, 1990), and declines inphysical health (Effros & Walford, 1987; Kart, Metress, & Me-tress, 1992). Therefore, alterations in these compromised phys-iological systems may have significant health consequences inolder populations. More important, the studies summarized inthis review are consistent with the notion that social support


may moderate such physiological processes in older adults.Given the importance of close affective social bonds in theseadults (Antonucci & Akiyama, 1987; Carstensen, 1992) andimpending demographics shifts that may curtain the availabil-ity of familial sources of support, facilitating the developmentof close emotional bonds among them may be particularlyimportant.

The studies summarized in this review may have relevancefor the aging process more generally. Research on chronologicalage and physiological function suggests that such physiologicalchanges are not a biological invariant with aging (E. L. Smith,1984) and that social factors may play a role in the aging process(Szklo, 1979; Uchino et al., 1992). The results of this reviewsuggest that social support has beneficial effects on physiologicalprocesses across different age groups. The net effect of such pro-cesses may be to biologically age the individual at a slower rate.Consistent with this hypothesis, we have found that social sup-port predicts age-related differences in blood pressure. Morespecifically, individuals low in social support were characterizedby age-related increases in blood pressure, whereas individualshigh in social support were characterized by low and compara-ble blood pressure across the ages (Uchino et al., 1992; Uchino,Cacioppo, Malarkey, Glaser, & Kiecolt-Glaser, 1995).

In conclusion, understanding what precisely are the relation-ships between social support and health outcomes in specificpopulations will likely change as refinements are made in theconceptualization and assessment of theoretical constructs,such as social support, construals and coping responses, behav-iorally relevant physiological events, and so on. As researchmoves from a description of associations to a delineation of un-derlying mechanisms, questions of causative factors take onadded importance. Considerable research exists, for instance,demonstrating that neurochemical events can influence socialprocesses and that social processes influence neurochemicalevents. Studies of mating behavior in the ring-necked dove indi-cate that social behaviors (e.g., male strutting and cooing) trig-ger hormonal changes (e.g., increased production of estrogen inthe female dove), which predispose the female toward a new setof behaviors (e.g., courtship and copulation), which results inadditional reciprocal influences between hormones and socialbehaviors until the newborns are reared (Erikson & Lehrman,1964; Lehrman, 1964; Martinez-Vargas & Erikson, 1973). Re-ciprocal influences between social and physiological events havebeen found in primates, as well. Testosterone levels in male pri-mates, for example, have been found to promote sexual behav-ior, whereas the availability of sexually receptive females, inturn, has been found to influence testosterone levels in maleprimates (Bernstein, Gordon, & Rose, 1983; Bernstein, Rose,& Gordon, 1974; Rose, Gordon, & Bernstein, 1972; see Caci-oppo & Berntson, 1992). Indeed, understanding the function ofhormones may have been far more rudimentary if not for anal-yses of their effects on social behavior and for the effects of socialbehavior on hormonal changes. There are, of course, physiolog-ical mechanisms underlying these phenomena, but the identi-fication and understanding of these mechanisms are betterserved by systematic investigations within and across multiplelevels of analysis rather than by an exclusive focus on any onelevel of analysis. Thus, the research reviewed here specifies a setof associations in need of mechanistic explanations but repre-

sents only the first wave of research describing associations be-tween social processes and health.

References

Ader, R., Felton, D. L., & Cohen, N. (1991). Psychoneuroimmunology(2nd ed.). New York: Academic Press.

Akerstedt, T., & Levi, L. (1978). Orcadian rhythms in the secretion ofcortisol, adrenaline, and noradernaline. European Journal of ClinicalInvestigations, 8, 57-58.

Allen, K. M., Blascovich, J., Tomaka, J., & Kelsey, R. M. (1991). Pres-ence of human friends and pet dogs as moderators of autonomic re-sponses to stress in women. Journal of Personality and Social Psy-chology, 61, 582-589.

Allen, M. T., & Crowell, M. D. (1989). Patterns of autonomic responseduring laboratory stress. Psychophysiology, 26, 603-614.

Andersson, L. (1985). Intervention against loneliness in a group of el-derly women: An impact evaluation. Social Science and Medicine,20, 355-364.

Antonucci, T. C, & Akiyama, H. (1987). Social networks in adult lifeand a preliminary examination of the convoy model. Journal ofGer-ontohgy,42,5l9-52T.

Arnetz, B. B., Edgren, B., Levi, L., & Otto, U. (1985). Behavioural andendocrine reactions in boys scoring high on Sennton neurotic scaleviewing an exciting and partly violent movie and the importance ofsocial support. Social Science and Medicine, 20, 731-736.

Arnetz, B. B., Theorell, T., Levi, L., Kallner, A., & Eneroth, P. (1983).An experimental study of social isolation of elderly people: Psychoen-docrine and metabolic effects. Psychosomatic Medicine, 45, 395-406.

Arnetz, B. B., Wasserman, J., Petrini, B., Brenner, S., Levi, L., Eneroth,P., Salovaara, H., Hjelm, R., Salovaara, L., Theorell, X, & Petterson,I. (1987). Immune function in unemployed women. PsychosomaticMedicine, 49, 3-12.

Barefoot, J. C., Dahlstrom, W. O., & Williams, R. B. (1983). Hostility,CHD incidence, and total mortality: A 25 year follow-up study of 255physicians. Psychosomatic Medicine, 45, 59-63.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator dis-tinction in social psychological research: Conceptual, strategic, andstatistical considerations. Journal of Personality and Social Psychol-ogy, 51, 1173-1182.

Baron, R. S., Cutrona, C. E., Hicklin, D., Russell, D. W., Lubaroff,D. M. (1990). Social support and immune function among spousesof cancer patients. Journal of Personality and Social Psychology, 59,344-352.

Barrera, M. (1986). Distinctions between social support concepts,measures, and models. American Journal of Community Psychology,14,413-445.

Baum, A. (1990). Stress, intrusive imagery, and chronic distress.Health Psychology, 9, 653-675.

Baum, A., & Orunberg, N. (1995). Measurement of stress hormones.In S. Cohen, R. C. Kessler, & L. U. Gordon (Eds.), Measuring stress:A guide for health and social scientists (pp. 175-192). New %rk:Oxford University Press.

Berkman, L. F, Leo-Summers, L., & Horwitz, R. I. (1992). Emotionalsupport and survival after myocardial infarction. Annals of InternalMedicine, 117, 1003-1009.

Bernstein, I. S., Gordon, T. P., & Rose, R. M. (1983). The interactionof hormones, behavior, and social context in nonhuman primates. InB. B. Svare(Ed.), Hormones and aggressive behavior (pp. 535-561).New \brk: Plenum Press.

Bernstein, I. S., Rose, R. M., & Gordon, T. P. (1974). Behavioral andenvironmental events influencing primate testosterone levels.Journal of Human Evolution, 3, 517-525.

Berntson, G. G., Cacioppo, J. T., & Quigley, K. S. (1991). Autonomic


determinism: The modes of autonomic control, the doctrine of auto-nomic space, and the laws of autonomic constraint. Psychological Re-view, 98, 459-487.

Berntson, G.G., Cacioppo, J. X, & Quigley, K. S. (1993). Cardiac psy-chophysiology and autonomic space in humans: Empirical perspec-tives and conceptual implications. Psychological Bulletin, 114, 296-322.

Bissette, G., & Ritchie, J. C. (1992). Radioimmunoassay methods. InC. B. Nemeroff (Ed.), Neuroendocrinology (pp. 39-50). London:CRC Press.

Bland, S. H., Krogh, V., Winkelstein, W., &Trevisan, M. (1991). Socialnetwork and blood pressure: A population study. PsychosomaticMedicine, 53, 598-607.

Blalock, E. J. (1989). A molecular basis for bidirectional communica-tion between the immune and neuroendocrine systems. PhysiologicalReviews, 69, 1-31.

Blazer, D. G. (1982). Social support and mortality in an elderly com-munity population. American Journal of Epidemiology, 115, 684-694.

Bolger, N., & Eckenrode, J. (1991). Social relationships, personality,and anxiety during a major stressful event. Journal of Personality andSocial Psychology, 67,440-449.

Borysenko, M. (1987). The immune system: An overview. Annals ofBehavioral Medicine, 9, 3-10.

Boyce, T. W., & Chesterman, E. (1990). Life events, social support, andcardiovascular reactivity in adolescence. Developmental and Behav-ioral Pediatrics, 77,105-111.

Broadhead, W. E., Kaplan, B. H., James, S. A., Wagner, E. H., Schoen-bach, V. J., Crimson, R., Heyden, S., Tibblin, G., & Gehlbach, S. H.(1983). The epidemiologic evidence for a relationship between socialsupport and health. American Journal of Epidemiology, 117, 521-537.

Burns, E. A., & Goodwin, J. S. (1990). Immunology and infectiousdisease. In C. K. Cassel, D. E. Risenberg, C. B. Sorensen, & J. R.Walsh (Eds.), Geriatric medicine (pp. 312-329). New York:Springer-Verlag.

Cacioppo, J. T. (1994). Social neuroscience: Autonomic, neuroendo-crine, and immune responses to stress. Psychophysiology, 31, 113-128.

Cacioppo, J. X, & Berntson, G. G. (1992). Social psychological contri-butions to the decade of the brain: Doctrine of multilevel analysis.American Psychologist, 47, 1019-1028.

Cacioppo, J. X, & Berntson, G. G. (1994). The relationship betweenattitudes and evaluative space: A critical review, with emphasis on theseparability of positive and negative substrates. Psychological Bulle-tin, 775,401-423.

Cacioppo, J. X, Berntson, G. G., & Andersen, B. L. (1991). Psycho-physiological approaches to the evaluation of psychotherapeutic pro-cess and outcome, 1991: Contributions from social psychophysiology.Psychological Assessment, 3, 321-336.

Cacioppo, J. X, Malarkey, W. B., Kiecolt-Glaser, K. G., Uchino, B. N.,Sgoutas-Emch, S. A., Sheridan, J. F., Berntson, G. G., & Glaser, R.(1995). Heterogeneity in neuroendocrine and immune responses tobrief psychological stressors as a function of autonomic cardiac acti-vation. Psychosomatic Medicine, 57, 154-164.

Cacioppo, J. X, Uchino, B. N., & Berntson, G. G. (1994). Individualdifferences in the autonomic origins of heart rate reactivity: Xhe psy-chometrics of respiratory sinus arrhythmia and preejection period.Psychophysiology, 31, 412-419.

Cacioppo, J. X, Uchino, B. N., Crites, S. L., Snydersmith, M. A., Smith,G., Berntson, G. G., & Lang, P. J. (1992). The relationship betweenfacial expressiveness and sympathetic activation in emotion: A criti-cal review, with emphasis on modeling underlying mechanisms and

individual differences. Journal of Personality and Social Psychology,62, 110-128.

Calabrese, J. R., Kling, M. A., & Gold, P. W. (1987). Alterations inimmunocompetence during stress, bereavement, and depression: Fo-cus on neuroendocrine regulation. American Journal of Psychiatry,144, 1123-1134.

Carr, D. J. (1991). The role of endogeneous opioids and their receptorsin the immune system. Proceedings of the Society of ExperimentalBiology and Medicine, 198, 710-720.

Cartensen, L. L. (1992). Social and emotional patterns in adulthood:Support for socioemotional selectivity theory. Psychology and Aging,7,331-338.

Cassell, J. (1976). The contribution of the social environment to hostresistance. American Journal of Epidimiology, 104, 107-123.

Chard, X (1982). An introduction to radioimmunoassay and relatedtechniques. In T. S. Work & E. Work (Eds.), Laboratory techniques inbiochemistry and molecular biology (pp. 1-284). New York: ElsevierBiomedical Press.

Clifford, P. A., Tan, S., & Gorsuch, R. L. (1991). Efficacy of a self-directed behavioral health change program: Weight, body composi-tion, cardiovascular fitness, blood pressure, health risk, and psycho-social mediating variables. Journal of Behavioral Medicine, 14, 303-323.

Cobb, S. (1974). Physiologic changes in men whose jobs were abol-ished. Journal of Psychosomatic Research, 18, 246-258.

Cobb, S. (1976). Social support as a moderator of life stress. Psychoso-matic Medicine, 38, 300-314.

Coe, C. L. (1993). Psychosocial factors and immunity in nonhumanprimates: A review. Psychosomatic Medicine, 55, 298-308.

Cogen, R. B., Stevens, A. W, Cohen-Cole, S., Kirk, K., & Freeman,A. (1982). Leukocyte function in the etiology of acute necrotizingulcerative gingivitis. Journal of Periodontology, 54, 402-407.

Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlationanalysis for the behavioral sciences. Hillsdale, NJ: Erlbaum.

Cohen, S. (1988). Psychosocial models of the role of social support inthe etiology of physical disease. Health Psychology, 7, 269-297.

Cohen, S., & McKay, G. (1984). Social support, stress and the bufferinghypothesis: A theoretical analysis. In A. Baum, S. E. Taylor, & J. E.Singer (Eds.), Handbook of psychology and health (pp. 253-267).Hillsdale, NJ: Erlbaum.

Cohen, S., & Syme, S. L. (1985). Social support and health. New York:Academic Press.

Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buff-ering hypothesis. Psychological Bulletin, 98, 310-357.

Collins, N. L., Dunkel-Schetter, C., Lobel, M., & Scrimshaw, S. C. M.(1993). Social support in pregnancy: Psychosocial correlates of birthoutcomes and postpartum depression. Journal of Personality and So-cial Psychology, 65, 1243-1258.

Cottington, E. M., Brock, B. M., House, J. S., & Hawthorne, V. M.(1985). Psychosocial factors and blood pressure in the Michiganstatewide blood pressure study. American Journal of Epidemiology,727,515-529.

Crary, B., Borysenko, M., Sutherland, D. C., Kutz, I., Borysenko, J. Z.,& Benson, H. (1983). Decrease in mitogen responsiveness of mono-nuclear cells from peripheral blood after epinephrine administrationsin humans. Journal of Immunology, 130, 694-697.

Cupps, X R., & Fauci, A. S. (1982). Corticosteroid-mediated immuno-regulation in man. Immunological Review, 65, 133-155.

Cutrona, C. E., & Russell, D. W. (1990). Type of social support andspecific stress: Toward a theory of optimal matching. In B. R. Sarason,I. G. Sarason, & G. R. Pierce (Eds.), Social support: An interactionalview (pp. 319-366). New York: Wiley.

Dimsdale, J. E., Mills, P., Patterson, X, Ziegler, M., & Dillon, E. (1994).


Effects of chronic stress on beta-adrenergic receptors in the homeless.Psychosomatic Medicine, 56, 290-295,

Dressier, W. W. (1980). Coping dispositions, social supports, and healthstatus. Ethos, 8, 146-171.

Dressier, W. W. (1983). Blood pressure, relative weight, and psychoso-cial resources. Psychosomatic Medicine, 45, 527-536.

Dressier, W. W. (1991). Social support, lifestyle incongruity, and arte-rial blood pressure in a southern Black community. PsychosomaticMedicine, 53, 608-620.

Dressier, W. W., Grell, G. A., Gallagher, P. N., & Viteri, F. E. (1992).Social factors mediating social class differences in blood pressure in aJamaican community. Social Science and Medicine, 35, 1233-1244.

Dressier, W. W., Mata, A., Chavez, A., Viteri, F. E., & Gallagher, P.(1986). Social support and arterial blood pressure in a central Mexi-can community. Psychosomatic Medicine, 48, 338-350.

Dressier, W. W., Santos, J. E., & Viteri, F. E. (1986). Blood pressure,ethnicity, and psychosocial resources. Psychosomatic Medicine, 48,509-519.

Earp, K. L., Ory, M. G., & Strogatz, D. S. (1982). The effects of familyinvolvement and practitioner home visits on the control of hyperten-sion. American Journal of Public Health, 72, 1146-1154.

Edens, J. L., Larkin, K. T., & Abel, J. L. (1992). The effects of socialsupport and physical touch on cardiovascular reactions to mentalstress. Journal of Psychosomatic Research, 36, 371-382.

Effros, R. B., & Walford, R. L. (1987). Infection and immunity in rela-tion to aging. In E. A. Goidl (Ed.), Aging and the immune response(pp. 45-65). New York: Marcel Dekker.

Ely, D. L., & Mostardi, R. A. (1986). The effects of recent life eventsstress, life assets, and temperament pattern on cardiovascular risk fac-tors for Akron city police officers. Journal of Human Stress, 12,11-91.

Erfurt, J. C, Foote, A., & Heirich, M. A. (1991). Worksite wellnessprogram: Incremental comparison of screening and referral alone,health education, follow-up counseling, and plant organization.American Journal of Health Promotion, 5,438-448.

Erikson, C., & Lehrman, D. (1964). Effects of castration of male ringdoves upon ovarian activity of females. Journal of Comparative andPhysiological Psychology, 58, 164-166.

Esterling, B. A., Kiecolt-Glaser, J. K., Bodnar, J. C., & Glaser, R.(1994). Chronic stress, social support, and persistent alterations inthe natural killer cell response to cytokines in older adults. HealthPsychology, /J, 291-299.

Felton, S. Y, &Felten, D. L. (1991). Innervationoflymphoid tissue. InR. Ader, D. L. Felton, & N. Cohen (Eds.), Psychoneuroimmunology(pp. 27-69). New York: Academic Press.

Fleming, R., Baum, A., Gisriel, M. M., & Gatchel, R. J. (1982). Medi-ating influences of social support on stress at Three Mile Island.Journal of Human Stress, 8, 14-22.

Fletcher, M. A., Baron, G. C., Ashman, M. R., Fischl, M. A., & Klimas,N. G. (1987). Use of whole blood methods in assessment of immuneparameters in immunodeficiency states. Diagnostic Clinical Immu-nology, 5,69-81.

Fletcher, M. A., Klimas, N. G., Morgan, R., & Gjerset, G. (in press).Lymphocyte proliferation. In N. Rose (Ed.), Manual of clinical labo-ratory immunology. Washington, DC: American Association forMicrobiology.

Folkow, B., Hallback, M., Lundgren, Y, Sivertsson, R., & Weiss, L.(1973). Importance of adaptive changes in vascular design for estab-lishment of primary hypertension, studied in man and in spontane-ously hypertensive rats. Circulation Research, 32-33 (Suppl, 1), 2-13.

Gatti, G., Masera, R., Cavallo, R., Sartori, M. L., Delponte, D., Carig-nola, R., Salvadori, A., & Angeli, A. (1987). Studies on the mecha-nism of cortisol inhibition of human natural killer cell activity:

Effects of calcium entry blockers and calmodulin antagonists. Ste-roids, 49, 601-616.

Gelman, R., Cheng, S.-C, Kidd, P., Waxdal, M., & Kagan, J. (1993).Assessment of the effects of instrumentation, monoclonal antibody,and fluorochrome on flow cytometric immunophenotyping: A reportbased on 2 years of the NIAID DAIDS flow cytometry quality assess-ment program. Clinical Immunology and Immunopathology, 66,150-162.

Gerin, W., Milner, D., Chawla, S., & Pickering, T. G. (1995). Socialsupport as a moderator of cardiovascular reactivity in women: A testof the direct effects and buffering hypotheses. Psychosomatic Medi-cine, 57, 16-22.

Gerin, W, Pieper, C., Levy, R., & Pickering, T. G. (1992). Social sup-port in social interaction: A moderator of cardiovascular reactivity.Psychosomatic Medicine, 54, 324-336.

Gill, A. A., Veigl, V. L., Shuster, J. J., & Notelovitz, M. (1984). A wellwoman's health maintenance study comparing physical fitness andgroup support programs. Occupational Therapy Journal of Research,4, 286-308.

Glaser, R., Kiecolt-Glaser, J. K., Bonneau, R., Malarkey, W, & Hughes,J. (1992). Stress-induced modulation of the immune response to re-combinant hepatitis B vaccine. Psychosomatic Medicine, 54, 22-29.

Goidl, E. A. (1987). Aging and the immune response. New Vbrk: MarcelDekker.

Goodkin, K., Blaney, N. T., Feaster, D., Fletcher, M., Baum, M. K.,Mantero-Atienza, E., Klimas, N. G., Millon, C., Szapocznik, J., &Eisdorfer, C. (1992). Active coping style is associated with naturalkiller cell cytotoxicity in asymptomatic HI V-1 seropositive homosex-ual men. Journal of Psychosomatic Research, 36, 635-650.

Goodwin, J. S., Searles, R. P., & Tung, K. S. K. (1982). terminologicalresponses of a healthy elderly population. Clinical and ExperimentalImmunology, 48, 403-410.

Gorlin, R. (1976). Coronary artery disease. London: Saunders.Greenspan, F. S., & Baxter, J. D. (1994). Basic and clinical endocrinol-

ogy. Norwalk, CT: Appleton & Lange.Groer, M. W., Humenick, S., & Hill, P. D. (1994). Characterizations

and psychoneuroimmunologic implications of secretory immuno-globulin A and cortisol in preterm and term breast milk. Journal ofPerinatal and Neonatal Nursing, 7, 42-51.

Grossman, P., Stemmler, G., & Meinhardt, E. (1990). Paced respira-tory sinus arrythmia as an index of cardiac parasympathetic toneduring varying behavioral tasks. Psychophysiology, 27,404-416.

Gunnar, M. R. (1992). Reactivity of the hypothalamic-pituitary-adre-nocortical system to stressors in normal infants and children. Pediat-rics, 90,491-497.

Habu, S., Akamatsu, K., Tamaoki, N., & Okumura, K. (1984). In vivosignificance of NK cells on resistance against virus (HSV-1) infec-tions in mice. Journal of Immunology, 133, 2743-2747.

Hallback, M., & Folkow, B. (1974). Cardiovascular responses to acutemental "stress" in spontaneously hypertensive rats. Acta PhysiologicaScandinavica, 90, 684-693.

Hansell, S. (1985). Adolescent friendship networks and distress inschool. Social Forces, 63, 698-715.

Hanson, B. S., Isacsson, S. O., Janzon, L., Lindell, S. E., & Rastam,L. (1988). Social anchorage and blood pressure in elderly men—Apopulation study. Journal of Hypertension, 6, 503-510.

Heitzmann, C. A., & Kaplan, R. M. (1988). Assessment of methods formeasuring social support. Health Psychology, 7, 75-109.

Henderson, S., Byrne, D. G., Duncan-Jones, P., Scott, R., & Adcock,5. (1980). Social relationships, adversity, and neurosis: A study ofassociations in a general population sample. British Journal of Psy-chiatry, 136, 574-583.

Herbert, T. B., & Cohen, S. (1993a). Depression and immunity: Ameta-analytic review. Psychological Bulletin, 113, 472-486.


Herbert, T. B., & Cohen, S. (1993b). Stress and immunity in humans:A meta-analytic review. Psychosomatic Medicine, 55, 364-379.

Herrera, J. A., Alvarado, J. P., & Martinez, J. E. (1988). The psychoso-cial environment and the cellular immunity in the pregnant patient.Stress Medicine, 4, 49-56.

Hoff, P. S., & Lowenstein, A. J. (1994). Professional support: An essen-tial component in a cardiovascular risk reduction program. Journalof Gerontological Nursing, 29, 28-32.

Houben, G. J., Diederiks, J. P. M., Kant, Y., & Notermans, J. V. H.(1990). Rationalization in garages in the Netherlands and its effectson occupational health. Work and Stress, 4, 179-189.

House, J. S., Landis, K. R., & Umberson, D. (1988, July 29). Socialrelationships and health. Science, 241, 540-545.

Ironson, G., Friedman, A., Klimas, N., Antoni, M., Fletcher, M. A.,LaPerriere, A., Simoneau, J., & Schneiderman, N. (1994). Distress,denial, and low adherence to behavioral interventions predict fasterdisease progression in gay men infected with human immunodefi-ciency virus. International Journal of Behavioral Medicine, I, 90-105.

Irwin, M., Brown, M., Patterson, T., Hauger, R., Mascovich, A., &Grant, I. (1991). Neuropeptide Y and natural killer cell activity:Findings in depression and Alzheimer caregiving stress. FASEB Jour-nal, 5,3100-3107.

Jackson, L. A., & Adams-Campbell, L. L. (1994). John Henryism andblood pressure in Black college students. Journal of Behavioral Med-icine, 17, 69-79.

Jaffe, B. M., & Behrman, H. R. (1979). Methods of hormone radioim-munoassay. New York: Academic Press.

James, S. A., LaCroix, A. 2.., Kleinbaum, D. G., & Strogatz, D. S.(1984). John Henryism and blood pressure differences among Blackmen: II. The role of occupational stressors. Journal of BehavioralMedicine, 7, 259-275.

Janes, C. R. (1990). Migration, changing gender roles and stress: TheSamoan case. Medical Anthropology, 12, 217-248.

Janes, C. R., & Pawson, I. G. (1986). Migration and biocultural adap-tation: Samoans in California. Social Science and Medicine, 22, 821-834.

Jemmott, J. B., Ill, & Magloire, K. (1988). Academic stress, social sup-port, and secretory immunoglobin A. Journal of Personality and So-cial Psychology, 55, 803-810.

Jung, J. (1990). Global versus health-specific social support and matchof preferred and perceived social support levels in relationship tocompliance and blood pressure in hypertensives. Journal of AppliedSocial Psychology, 20, 1103-1111.

Kamarck, T. W. (1992). Recent developments in the study of cardio-vascular reactivity: Contributions from psychometric theory and so-cial psychology. Psychophysiology, 29, 491-503.

Kamarck, T. W., Annunziato, B., & Amateau, L. M. (1995). Affiliationmoderates the effects of social threat on stress-related cardiovascularresponses: Boundary conditions for a laboratory model of social sup-port. Psychosomatic Medicine, 57, 183-194.

Kamarck, T. W., & Jennings, R. J. (1991). Biobehavioral factors insudden cardiac death. Psychological Bulletin, 109, 42-75.

Kamarck, T. W., Jennings, J. R., Debski, T. T, Glickman-Weiss, E.,Johnson, P. S., Eddy, M. J., & Manuck, S. B. (1992). Reliable mea-sures of behaviorally-evoked cardiovascular reactivity from a PC-based test battery: Results from student and community samples.Psychophysiology, 29, 17-28.

Kamarck, T. W., Manuck, S. B., & Jennings, J. R. (1990). Social sup-port reduces cardiovascular reactivity to psychological challenge: Alaboratory model. Psychosomatic Medicine, 52, 42-58.

Kart, C. S., Metress, E. K., & Metress, S. P. (1992). Human aging andchronic disease. Boston: Jones & Bartlett.

Kasl, S. V., & Cobb, S. (1980). The experience of losing a job: Some

effects on cardiovascular functioning. Psychotherapy and Psychoso-matics, 34, 88-109.

Kasprowicz, A. L., Manuck, S. B., Malkoff, S. B., & Krantz, D. S.(1990). Individual differences in behaviorally evoked cardiovascularresponse: Temporal stability and hemodynamic patterning. Psycho-physiology, 27, 605-619.

Kaufmann, G. M., & Beehr, T. A. (1986). Interactions between jobstressors and social support: Some counterintuitive results. Journalof Applied Psychology, 71, 522-526.

Khan, M. M., Sansoni, P., Silverman, E. D., Engleman, E. G., & Mel-mon, K. L. (1986). Beta-adrenergic receptors on human suppressor,helper, and cytolytic lymphocytes. Biochemical Pharmacology, 35,1137-1142.

Kiecolt-Glaser, J. K., Dura, J. R., Speicher, C. E., Trask, O. J., & Glaser,R. G. (1991). Spousal caregivers of dementia victims: Longitudinalchanges in immunity and health. Psychosomatic Medicine, 53, 345-362.

Kiecolt-Glaser, J. K., & Glaser, R. (1988a). Immunological compe-tence. In E. A. Blechman & K. D. Brownwell (Eds.), Handbook ofbehavioral medicine for women (pp. 195-205). New York: PergamonPress.

Kiecolt-Glaser, J. K., & Glaser, R. (1988b). Methodological issues inthe behavioral immunology research with humans. Brain, Behavior,and Immunity, 2, 67-78.

Kiecolt-Glaser, J. K., & Glaser, R. (1989). Interpersonal relationshipsand immune function. In L. Carstensen & J. Neale (Eds.), Mecha-nisms of psychological influence on physical health (pp. 43-59). New\brk: Plenum Press.

Kiecolt-Glaser, J. K., & Glaser, R. (1995). Psychoneuroimmunologyand health consequences: Data and shared mechanisms. Psychoso-matic Medicine. 57, 269-274.

Kiecolt-Glaser, J. K., Glaser, R., Williger, D., Stout, J., Messick, G.,Sheppard, S., Ricker, D., Romisher, S. C., Briner, W., Bonnell, G., &Donnerberg, R. (1985). Psychosocial enhancement of immunocom-petence in a geriatric population. Health Psychology, 4, 25-41.

Kiecolt-Glaser, J. K., & Greenberg, B. (1984). Social support as a mod-erator of the aftereffects of stress in female psychiatric inpatients.Journal of "Personality and Social Psychology, 93, 192-199.

Kiecolt-Glaser, J. K., Malarkey, W. B., Chee, M., Newton, X, Cacioppo,J. T., Mao, H., &Glaser, R. (1993). Negative behavior during maritalconflict is associated with immunological down-regulation. Psycho-somatic Medicine, 55, 395-409.

Kirschbaum, C., Klauer, T, Filipp, S. H., & Hellhammer, D. H. (1995).Sex-specific effects of social support on cortisol and subjective re-sponses to acute psychological stress. Psychosomatic Medicine, 57,23-31.

Knox, S. S. (1993), Perception of social support and blood pressure inyoung men. Perceptual and Motor Skills, 77, 132-134.

Knox, S. S., Theorell, T, Svensson, J. C., & Waller, D. (1985). Therelation of social support and working environment to medical vari-ables associated with elevated blood pressure in young males: Astructural model. Social Science and Medicine, 21, 525-531.

Krantz, D. S., Helmers, K. E, Bairey, N., Nebel, L. E., Hedges, S. M.,& Rozanski, A. (1991). Cardiovascular reactivity and mental stress-induced myocardial ischemia in patients with coronary artery dis-ease. Psychosomatic Medicine, 53, 1-12.

Krantz, D. S., & Manuck, S. B. (1984). Acute physiologic reactivity andrisk of cardiovascular disease: A review and methodologic critique.Psychological Bulletin, 96, 435-464.

Larsen, P. B., Schneiderman, N., & Pasin, R. D. (1986). Physiologicalbases of cardiovascular Psychophysiology. In M. G. H. Coles, E. Don-chin, & S. W. Forges (Eds.), Psychophysiology: Systems, processes,and applications (pp. 122-165). New York: Guilford Press.


Lehrman, D. (1964). The reproductive behavior of ring doves. Scien-tific American, 211,48-54.

Lepore, S. J. (1992). Social conflict, social support, and psychologicaldistress: Evidence of cross-domain buffering effects. Journal of Per-sonality and Social Psychology, 63, 857-867.

Lepore, S. J. (1995). Cynicism, social support, and cardiovascular re-activity. Health Psychology, 14, 210-216.

Lepore, S. J., Allen, K. A., & Evans, G. W. (1993). Social support low-ers cardiovascular reactivity to an acute stressor. Psychosomatic Med-icine, 55,518-524.

Lercher, P., Hortnagl, J., & Kofler, W. W. (1993). Work noise annoy-ance and blood pressure: Combined effects with stressful workingconditions. International Archives of Occupational and Environmen-tal Health, 65, 23-28.

Levine, D. M., Green, L. W., Deeds, S. G., Chwalow, J., Russell,R. P., & Finlay, J. {1979). Health education for hypertensive patients.Journal of 'the American Medical Association, 241, 1700-1703.

Levy, S. M., Herberman, R. B., Whiteside, T, Sanzo, K., Lee, J., &Kirkwood, J. (1990). Perceived social support and tumor estrogen/progesterone receptor status as predictors of natural killer cell activ-ity in breast cancer patients. Psychosomatic Medicine, 52, 73-85.

Lewis, J. W., Shavit, Y., Terman, G. W., Nelson, L. R., Gale, R. P., &Liebeskind, J. C. (1983). Apparent involvement of opioid peptidesin stress-induced enhancement of tumor growth. Peptides, 4, 635-638.

Light, K. C. (1981). Cardiovascular responses to effortful active coping:Implications for the role of stress in hypertension development. Psy-chophysiology, 18, 216-225.

Light, K. C., Dolan, C. A., Davis, M. R., & Sherwood, A. (1992). Car-diovascular responses to an active coping challenge as predictors ofblood pressure patterns 10 to 15 years later. Psychosomatic Medicine,54,217-230.

Linden, W., Chambers, L., Maurice, J., & Lenz, J. W. (1993). Sexdifferences in social support, self-deception, hostility, and ambula-tory cardiovascular activity. Health Psychology, 12, 376-380.

Linn, B. S., Linn, M. W., & Klimas, N. G. (1988). Effects of psycho-physical stress on surgical outcomes. Psychosomatic Medicine, 50,230-244.

Livingston, I. L., Levine, D. M., & Moore, R. D. (1991). Social inte-gration and Black intraracial variations in blood pressure. EthnicityandDisease, 1, 135-149.

Lumio, J., Welin, M.-G., Hirvonen, P., & Weber, T. (1983). Lympho-cyte subpopulations and reactivity during and after infectious mono-nucleosis. Medical Biology, 61, 208-213.

Madden, K. S., & Livnat, S. (1991). Catecholamine action and immu-nologic reactivity. In R. Aden D. L. Felten, & N. Cohen (Eds.), Psy-choneuroimmunology (pp.283-310). New York: Academic Press.

Malarkey, W. B., Kiecolt-Glaser, J. K., Pearl, D., & Glaser, R. (1994).Hostile behavior during marital conflict affects pituitary and adrenalhormones. Psychosomatic Medicine, 56,41-51.

Malcolm, A. T., & Janisse, M. P. (1991). Additional evidence for therelationship between Type A behavior and social support in men. Be-havioral Medicine, 17, 131-134.

Manuck, S. B. (1994). Cardiovascular reactivity in cardiovascular dis-ease: "Once more unto the breach." International Journal of Behav-ioral Medicine, 7,4-31.

Manuck, S. B., Cohen, S. C., Rabin, B., Muldoon, M. E, & Bachen,E. A. (1991). Individual differences in cellular immune response tostress. Psychological Science, 2, 111-115.

Manuck, S. B., Kaplan, J. R., & Clarkson, T. B. (1983). Behaviorallyinduced heart rate reactivity and atherosclerosis in Cynomolgusmonkeys. Psychosomatic Medicine, 45, 95-108.

Manuck, S. B., Kasprowicz, A. L., Monroe, S. M., Larkin, K. T, &Kaplan, J. R. (1989). Psychophysiologic reactivity as a dimension of

individual differences. In N. Schneiderman, S. B. Weiss, & P. Kauf-mann (Eds.), Handbook of methods and measurements in cardiovas-cular behavioral medicine (-pp. 365-382). New York: Plenum Press.

Marsland, A. L., Manuck, S. B., Fazzari, T. V., Stewart, C. J., & Rabin,B. S. (1995). Stability of individual differences in cellular immuneresponses to acute psychological stress. Psychosomatic Medicine, 57,295-298.

Martinez-Vargas, M. C., & Erikson, C. J. (1973). Some social and hor-monal determinants of nest-building behaviour in the ring dove(Streptopelia risoria). Behaviour, 45, 12-37.

Matthews, K. A., Weiss, S. M., Detre, T, Dembroski, T. M., Falkner,B., Manuck, S. B., & Williams, R. B. (1986). Handbook of stress,reactivity, and cardiovascular disease. New\brk: Wiley.

McGonigle, M. M., Smith, T. W., Benjamin, L. S., & Turner, C. W.(1993). Hostility and nonshared family environment: A study ofmonozygotic twins. Journal of Research in Personality, 27, 23-34.

Melntosh, W. A., Kaplan, H. B., Kubena, K. S., & Landmann, W. A.(1993). Life events, social support, and immune response in elderlyindividuals. International Journal of Aging and Human Development,37, 23-36.

McKinnon, W., Baum, A., & Morokoff, P. (1988). Neuroendocrinemeasures of stress. In H. L. Wagner (Ed.), Social psychophysiologyand emotion (pp. 43-64). New York: Wiley.

McNaughton, M. E., Smith, L. W., Patterson, T. L., &Grant, I. (1990).Stress, social support, coping resources, and immune status in elderlywomen. Journal of Nervous and Mental Disease, 178, 460-461.

Medline [Database]. (1983-1995, June 1). Bethesda, MD: NationalLibrary of Medicine [Producer], Available from: SilverPlatter.

Meites, J., Goya, R., & Takahashi, S. (1987). Why the neuroendocrinesystem is important in aging processes. Experimental Gerontology,22, 1-15.

Mills, P. J., & Dimsdale, J. J. (1993). The promise of adrenergic recep-tor studies in psychophysiologic research. II: Applications, limita-tions, and progress. Psychosomatic Medicine, 55, 448-457.

Morisky, D. E., DeMuth, N. M., Field-Pass, M., Green, L. W., & Lev-ine, D. M. (1985). Evaluation of family health education to buildsocial support for long-term control of high blood pressure. HealthEducation Quarterly, 12, 35-50.

Morisky, D. E., Levine, D. M., Green, L. W, Shapiro, S., Russell, P., &Smith, C. R. (1983). Five-year blood pressure control and mortalityfollowing health education for hypertensive patients. American Jour-nal of Public Health, 73, 153-162.

Mullin, B. (1989). Advanced BASK meia-analysis. Hillsdale, NJ:Erlbaum.

Murasko, D. M., Weiner, P., & Kaye, D. (1988). Association of lack ofmitogen-induced lymphocyte proliferation with increased mortalityin the elderly. Aging: Immunology and Infectious Disease, I, 1-6.

Orth-Gomer, K., Rosengren, A., & Wilhelmsen, L. (1993). Lack of so-cial support and incidence of coronary heart disease in middle-agedSwedish men. Psychosomatic Medicine, 55, 37-43.

Orth-Gomer, K., & Unden, A. (1987). The measurement of social sup-port in population surveys. Social Science and Medicine, 24, 83-94.

Pagel, M. D., Erdly, W. W., & Becker, J. (1987). Social networks: We getby with (and in spite of) a little help from our friends. Journal ofPersonality and Social Psychology, 53, 793-804.

Paxton, H., Kidd, P., Landay, A., Giorgi, J., Romenberg, N., Walker, E.,Valentine, F., Fahey, J., & Gelman, R. (1989). Results of the flowcytometry ACTG quality control program: Analysis and findings.Clinical Immunology and Immunopathology, 52, 68-84.

Perry, S., Fishman, B., Jacobsberg, L., & Frances, A. (1992). Relation-ships over 1 year between lymphocyte subsets and psychosocial vari-ables among adults with infection by human immunodeficiency vi-rus. Archives of General Psychiatry, 49, 396-401.

Persson, L., Gullberg, B., Hanson, B. S., Moestrup, T, & Ostergren,


P. O. (1994). HIV infection: Social network, social support, and CD4lymphocyte values in infected homosexual men in Malmo, Sweden.Journal of Epidemiology and Community Health, 48, 580-585.

Pierce, G. R., Sarason, I. G., & Sarason, B. R. (1991). General andrelationship-based perceptions of social support: Are two constructsbetter than one? Journal of Personality and Social Psychology, 61,1028-1039.

Pinto, R. P., Sirota, A. D., & Brown, W. A. (1985). Behavioral instruc-tion for a patient's relative to increase antihypertensive compliance.Psychosomatics, 26, 677-679.

Plaut, M. (1987). Lymphocyte hormone receptors. Annual Review ofImmunology, 5, 621-669.

Porges, S. W. (1992). Vagal tone: A physiological marker of stress vul-nerability. Pediatrics, 90, 498-504.

PsycLIT [ Database ]. (1974-1995, June). Washington, DC: AmericanPsychological Association [Producer]. Available from: SilverPlatter.

Quittner, A. L., Glueckauf, R. L., & Jackson, D. N. (1990). Chronicparenting stress: Moderating versus mediating effects of social sup-port. Journal of Personality and Social Psychology, 59, 1266-1278.

Roberts-Thomson, I. C, Whittingham, S., Youngchaiyud, U., &Mackay, I. R. (1974). Aging, immune response, and mortality. Lan-cet, ii, 368-370.

Rook, K. S. (1984). The negative side of social interaction: Impact onpsychological well being. Journal of Personality and Social Psychol-ogy, 46, 1097-1108.

Rook, K. S., & Pietromonaco, P. (1987). Close relationships: Ties thatheal or ties that bind. Advances in Personal Relationships, 1, 1-35.

Rose, R. N., Gordon, T. P., Bernstein, I. S. (1972, November 10).Plasma testosterone levels in the male rhesus: Influences of sexualand social stimuli. Science, 178, 643-645.

Rosenthal, R. (1984). Mela-analytic procedures of social research.London: Sage.

Rozanski, A., Bairey, N., Krantz, D. S., Friedman, J., Resser, K. J., Mo-rell, M., Hilton-Chalfen, S., Herstrin, L., Bietendorf, J., & Berman,D. S. (1988). Mental stress and the induction of silent myocardialischemia in patients with coronary artery disease. New England Jour-nal of Medicine, 318, 1005-1012.

Ruehlman, L. S., & Karoly, P. (1991). With a little flak from myfriends: Development and preliminary validation of the Test of Neg-ative Social Exchange (TENSE). Psychological Assessment, 3, 97-104.

Rupprecht, R., Wodarz, N., Kornhuber, J., Schmitz, B., Wild, K.,Braner, H. U., Muller, O. A., & Riederer, P. (1990). In vivo and invitro effects of glucocorticoids on lymphocyte proliferation in man:Relationship to glucocorticoid receptors. Neuropsychobiology, 24,61-66.

Russell, D. W., & Cutrona, C. E. (1991). Social support, stress, anddepressive symptoms among the elderly: Test of a process model. Psy-chology and Aging, 6, 190-201.

Sallis, J. E, Trevorrow, T. R., Johnson, C. C., Hovell, M. E, & Kaplan,R. M. (1987). Worksite stress management: A comparison of pro-grams. Psychology and Health, 1, 237-255.

Sanders, V. M., & Munson, A. E. (1985). Norepinephrine and the anti-body response. Pharmacological Reviews, 37, 229-248.

Sarason, B. R., Sarason, I. G., & Pierce, G. R. (1989). Social support:An interactional view. New York: Wiley.

Sarason, I. G., Sarason, B. R., & Shearin, E. N. (1986). Social supportas an individual difference variable: Its stability, origins, and rela-tional aspects. Journal of Personality and Social Psychology, 50, 845-855.

Schenker, E. L., Hultin, L. E., Bauer, K. D., Ferbas, J., Margolick, J. B.,& Giorgi, J. V. (1993). Evaluation of a dual-color flow cytometryimmunophenotyping panel in a multicenter quality assurance pro-gram. Cytometry, 14, 307-317.

Schleifer, S. J., Keller, S. E., Bond, R. N., Cohen, J., & Stein, M. (1989).Major depressive disorder and immunity: Role of age, sex, severity,and hospitalization. Archives of General Psychiatry, 46, 81-87.

Schlesinger, M., & Yodfat, T. (1991). The impact of stressful life eventson natural killer cells. Stress Medicine, 7, 53-60.

Seeman, T. E., Berkman, L. F., Blazer, D., & Rowe, J. W. (1994). Socialties and support and neuroendocrine function: The MacArthur stud-ies of successful aging. Annals of Behavioral Medicine, 16, 95-106.

Seeman, T. E., & Robbins, R. J. (1994). Aging and the hypothalamic-pituitary-adrenal response to challenge in humans. Endocrine Re-views, 15, 233-260.

Sgoutas-Emch, S. A., Cacioppo, J. T., Uchino, B. N., Malarkey, W,Pearl, D., Kiecolt-Glaser, J. K., &Glaser, R. (1994). The effectsof anacute psychological stressor on cardiovascular, endocrine, and cellu-lar immune response: A prospective study of individuals high and lowin heart rate reactivity. Psychophysiology, 31, 264-271.

Sheffield, D., & Carroll, D. (1994). Social support and cardiovascularreactions to active laboratory stressors. Psychology and Health, 9,305-316.

Shekelle, R. B., Gale, M., Ostfeld, A. M., & Paul, O. (1983). Hostility,risk of coronary heart disease, and mortality. Psychosomatic Medi-cine, 45, 109-114.

Sheridan, J. E, Feng, N., Bonneau, R., Malarkey, W, & Hughes, J.(1991). Restraint stress differentially affects anti-viral cellular andhumoral immune responses in mice. Journal of Neuroimmunology,31, 245-255.

Sherwood, A., Allen, M. T, Fahrenberg, J., Kelsey, R. M., Lovallo,W. R., & Doornen, L. J. P. van. (1990). Methodological guidelinesfor impedance cardiography. Psychophysiology, 27, 1-23.

Sherwood, A., Dolan, C. A., & Light, K. C. (1990). Hemodynamics ofblood pressure responses during active and passive coping. Psycho-physiology, 27, 656-668.

Shumaker, S. A., & Hill, D. R. (1991). Gender differences in socialsupport and health. Health Psychology, 10, 102- 111.

Sibinga, N. E. S., & Goldstein, A. (1988). Opioid peptides and opioidreceptors in cells of the immune system. Annual Review of Immunol-ogy^, 219-249.

Smith, C. E., Fernengel, K., Holcroft, C., Gerald, K., & Marien, L.(1994). Meta-analysis of the association between social support andhealth outcomes. Annals of Behavioral Medicine, 16, 352-362.

Smith, E. L. (1984). Special considerations in developing exercise pro-grams for the older adult. In J. D. Matarazzo, S. M. Weiss, J. A. Herd,N. E. Miller, & S. M. Weiss (Eds.), Behavioral health: A handbook ofhealth enhancement and disease prevention (pp. 525-546). NewYork: Wiley.

Smith, J. J., & Kampine, J. P. (1990). Circulatory physiology. London:Williams &Wilkins.

Smith, T. W. (in press). Assessment and modification of coronary pronebehavior: A transactional view of the person in social context. InA. J. Goreczny (Ed.), Handbook of health and rehabilitation psychol-ogy. New \brk: Plenum Press.

Smith, T. W, & Frohm, K. D. (1985). What's so unhealthy about hos-tility? Construct validity and psychosocial correlates of the Cook andMedley Ho Scale. Health Psychology, 4, 503-520.

Smith, T. W, Pope, M. K., Sanders, J. D., Allred, K. D., & O'Keeffe,J. L. (1988). Cynical hostility at home and work: Psychosocial vul-nerability across domains. Journal of Research in Personality, 22,525-548.

Snyder, B. K., Roghmann, K. J., & Sigal, L. H. (1990). Effects of stressand other biopsychosocial factors on primary antibody response.Journal of Adolescent Health Care, 11, 472-479.

Snyder, B. K., Roghmann, K. J., & Sigal, L. H. (1993). Stress and psy-chosocial factors: Effects on primary cellular immune response.Journal of Behavioral Medicine, 16, 143-161.


Snydersmith, M. A., & Cacioppo, J. T. (1992). Parsing complex socialfactors to determine component effects: I. Autonomic activity andreactivity as a function of human association. Journal of Social andClinical Psychology, 11, 263-278.

Solomon, Z., Mikulincer, M., & Hobfoll, S. E. (1986). Effects of socialsupport and battle intensity on loneliness and breakdown duringcombat. Journal of Personality and Social Psychology, 51, 1269-1276.

Spitzer, S. B., Llabre, M. M., Ironson, G. H., Gellman, M. D., & Schnei-derman, N. (1992). The influence of social situations on ambulatoryblood pressure. Psychosomatic Medicine, 54, 79-86.

Stahl, S. M., Kelley, C. R., Neill, P., Grim, C. E., & Mamlin, J. (1984).Effects of home blood pressure measurement on long-term BP con-trol. American Journal of Public Health, 74, 704-709.

Stavig, G. R., Igra, A., & Leonard, A. R. (1984). Hypertension amongAsians and Pacific Islanders in California. American Journal of Epi-demiology, 119, 677-691.

Stokes, J. P. (1985). The relation of social network and individualdifferences in loneliness. Journal of Personality and Social Psychol-ogy, 45,981-990.

Stone, A. A., Cox, D. S., Valdimarsdottir, H., & Neale, J. M. (1987).Secretory IgA as a measure of immunocompetence. Journal of Hu-man Stress, 13, 136-140.

Strogatz, D. S., & James, S. A. (1986). Social support and hypertensionamong Blacks and Whites in a rural, southern community. AmericanJournal of Epidemiology, 124, 949-956.

Szklo, M. (1979). Epidemiologic patterns of blood pressure in children.Epidemiologic Reviews, 1, 143-169.

Tardy, C. H. (1985). Social support measurement. American Journalof Community Psychology, 13, 187-202.

Tardy, C. H., Thompson, W. R., & Allen, M. T. (1989). Cardiovascularresponses during speech: Does social support mediate the effects oftalking on blood pressure? Journal of Language and Social Psychol-ogy, 5,271-285.

Theorell, T, Blomkvist, V., Jonsson, H., Schulman, S., Berntorp, E., &Stigendel, L. (1995). Social support and the development of immunefunction in human immunodeficiency virus infection. Psychoso-matic Medicine, 57, 32-36.

Theorell, T, Orth-Gomer, K., & Eneroth, P. (1990). Slow-reacting im-munoglobin in relation to social support and changes in job strain: Apreliminary note. Psychosomatic Medicine, 52, 511-516.

Thomas, P. D., Goodwin, J. M., & Goodwin, J. S. (1985). Effects ofsocial support on stress-related changes in cholesterol level, uric acidlevel, and immune function in an elderly sample. American Journalof Psychiatry, 142, 735-737.

Tonnesen, E., Christensen, N. J., & Brinklov, M. M. (1987). Natural

killer cell activity during cortisol and adrenaline infusion in healthyvolunteers. European Journal of Clinical Investigation, 17, 497-503.

Uchino, B. N., Cacioppo, J. T, Malarkey, W., & Glaser, R. (1995). In-dividual differences in cardiac sympathetic control predict endocrineand immune responses to acute psychological stress. Journal of Per-sonality and Social Psychology, 69, 736-743.

Uchino, B. N., Cacioppo, J. T., Malarkey, W. M., Glaser, R., & Kiecolt-Glaser, J. K. (1995). Appraisal support predicts age-related differ-ences in cardiovascular function in women. Health Psychology, 14,556-562.

Uchino, B. N., Kiecolt-Glaser, J. K., & Cacioppo, J. T. (1992). Age-related changes in cardiovascular response as a function of a chronicstressor and social support. Journal of Personality and Social Psy-chology, 63, 839-846.

Uchino, B. N., Kiecolt-Glaser, J. K., & Cacioppo, J. T. (1994). Con-struals of preillness relationship quality predict cardiovascular re-sponse in family caregivers of Alzheimer's disease victims. Psychologyand Aging, 9, 113-120.

Umberson, D. (1987). Family status and health behaviors: Social con-trol as a dimension of social integration. Journal of Health and SocialBehavior, 28, 306-319.

Unden, A.-L., Orth-Gomer, K., & Elofsson, S. (1991). Cardiovasculareffects of social support in the work place: Twenty-four hour ECGmonitoring of men and women. Psychosomatic Medicine, 53, 50-60.

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and val-idation of brief measures of positive and negative affect: The PANASscales. Journal of Personality and Social Psychology, 54, 1063-1070.

Watson, D., & Pennebaker, J. W. (1989). Health complaints, stress, anddistress: Exploring the central role of negative affectivity. Psychologi-cal Review, 96, 234-254.

Whiteside, T. L., & Herberman, R. B. (1994). Role of human naturalkiller cells in health and disease. Clinical and Diagnostic LaboratoryImmunology, 1, 125-133.

Williams, L. T, Snyderman, R., & Lefkowitz, R. J. (1976). Identifica-tion of b-adrenergic receptors in human lymphocytes by ( — ) [~3H]alprenolol binding. Journal of Clinical Investigation, 57, 149-155.

Winemiller, D. R., Mitchell, E. M., Sutliff, J., & Cline, D. J. (1993).Measurement strategies in social support: A descriptive review of theliterature. Journal of Clinical Psychology, 49, 638-648.

Winnubst, J. A. M., Marcelissen, F. H. G., & Kleber, R. J. (1982).Effects of social support in the stressor-strain relationship: A Dutchsample. Social Science and Medicine, 16, 475-482.

Received February 14, 1995Revision received July 21, 1995

Accepted July 24, 1995 •

Documents

The Relationship Between Social Support and Physiological ...€¦ · Kampine, 1990, for detailed reviews). The heart muscle gener-ates th e necessar y forc for circulator process